Protecting Electrical Equipment: Good practices for preventing high altitude electromagnetic pulse impacts 9783110639285, 9783110635966

Citation preview

Vladimir Gurevich Protecting Electrical Equipment

Also of Interest Signals and Systems B. Guo, J. Zhu, 2018 ISBN 978-3-11-059541-3, e-ISBN (PDF) 978-3-11-059390-7, e-ISBN (EPUB) 978-3-11-059296-2

Analog Electronic Circuit B. Ning, 2018 ISBN 978-3-11-059540-6, e-ISBN (PDF) 978-3-11-059386-0, e-ISBN (EPUB) 978-3-11-059319-8

Signal Processing and Data Analysis T. Qiu, Y. Guo, 2018 ISBN 978-3-11-046158-9, e-ISBN (PDF) 978-3-11-046508-2, e-ISBN (EPUB) 978-3-11-046513-6

Signals and Systems Volumes 1+2 W. Zhang, 2017 ISBN 978-3-11-054409-1

Magnetic Flux Leakage S. Huang, W. Zhao, 2016 ISBN 978-3-11-047701-6, e-ISBN (PDF) 978-3-11-048015-3, e-ISBN (EPUB) 978-3-11-047971-3

Vladimir Gurevich

Protecting Electrical Equipment |

Good Practices for Preventing High Altitude Electromagnetic Pulse Impacts

Author Dr. Vladimir Gurevich Israel Electric Corporation POB 10, 31000 Haifa Israel [email protected]

ISBN 978-3-11-063596-6 e-ISBN (PDF) 978-3-11-063928-5 e-ISBN (EPUB) 978-3-11-063606-2 Library of Congress Control Number: 2019934269 Bibliographic information published by the Deutsche Nationalbibliothek The Deutsche Nationalbibliothek lists this publication in the Deutsche Nationalbibliografie; detailed bibliographic data are available on the Internet at http://dnb.dnb.de. © 2019 Walter de Gruyter GmbH, Berlin/Munich/Boston Cover image: dani3315 / iStock / Getty Images Typesetting: VTeX UAB, Lithuania Printing and binding: CPI books GmbH, Leck www.degruyter.com

About the Author

Vladimir Gurevich was born in the city of Kharkov (in the Ukraine) in 1956. In 1978 he graduated from the Faculty of Electrification at the P. Vasilenko Kharkov National Technical University. From 1980 to 1983 he was a postgraduate student. In 1986 he defended his PhD thesis: ‘Quasistatic Switching and Regulating Equipment with HighVoltage Decoupling’ at the National Technical University, Kharkov Polytechnical Institute. He worked as a Lecturer and acting Associate Professor at the P. Vasilenko Kharkov National Technical University and as Chief Engineer and Director of the Scientific Technical Enterprise ‘Inventor’ (in the city of Kharkov). He led a number of projects to develop new types of equipment, which were conducted at the behest of the Defence Sector Industries in the USSR. Following the collapse of the USSR he has been engaged in developing and administering the production of protection and automation equipment for electrical power systems. He now works for the Israel Electric Corporation as a Senior Specialist, and Head of section at the Central Electric Laboratory. Since 2006 he has been Professor Emeritus at the P. Vasilenko Kharkov National Technical University. Since 2007 he has been an expert on the TC-94 Committee of the International Electrotechnical Commission (IEC) and since 2017 – on SC77C Committee; International Editorial Board member of 5 scientific and technical journals; member of International Association of Engineers (IAENG). V. I. Gurevich has written 15 books, more than 120 patents and more than 200 scientific and technical articles.

https://doi.org/10.1515/9783110639285-201

Our vulnerability is increasing daily as our use of and dependence on electronics continues to grow in both our civil and military sectors Dr. William R. Graham, Chairman of the Commission to Assess the Threat to the United States from Electromagnetic Pulse (EMP)

The current state of EMP protection is random, disoriented and uncoordinated.

Dr. George H. Baker, Prof. Emeritus James Madison University

...its potentially paralyzing effects on military forces and civilian critical infrastructures were deeply understood only by a small number of nuclear strategists and specialists. Dr. Peter Vincent Pry, Executive Director of the Task Force on National and Homeland Security

Much of the available information is not specifically applied to electric utilities, making it very difficult for utilities and regulators to understand effective options for protecting energy infrastructure. Robin Manning, Vice president for transmission and distribution for the Electric Power Research Institute (EPRI)

Our power grid is very vulnerable. It’s very much on edge. Our military knows that.

Roscoe Bartlett, Ex-Congressmen

It would be “suicidally optimistic” to assume that an EMP attack that inflicted a statewide blackout would not also cause cascading grid and infrastructure failures at least regionally. Dr. William Radasky, Founder and President of the Metatech Corporation

Annotation V. Gurevich Protecting Electrical Equipment. Good Practices for Preventing High Altitude Electromagnetic Pulse Impacts. – De Gruyter, Berlin, 2019. This unusual book recounts the history and development of military nuclear programs in the USSR and USA; the role of intelligence services in the development of nuclear weapons; discovery of electromagnetic pulses (EMP) caused by high-altitude nuclear explosion; and numerous tests of nuclear weapons. The book contains numerous previously secret documents and photos that have been recently declassified and approved for public release. Using approachable language for the nonspecialist in nuclear physics, the book describes the formation process of EMPs caused by high-altitude nuclear explosions (HEMP) and discusses numerous factors affecting the HEMP strength and other of its parameters. Also discussed is the impact of nuclear EMPs on electronic components and devices and also on electrical power equipment. The main part of the book describes only practical (and not theoretical, as in hundreds of existing scientific reports) protective methods and protection means, as well as methods for effective evaluation of the protective measures. Due to its breadth of scope, timeliness, depth of coverage and the practical importance of described protective means, the book may be considered as an encyclopedia of HEMP, having no equal on the book market. The book is intended for electrical engineers dealing with the development, design and operation of electronic and electrical equipment, and it will also be useful for university lecturers and students. Much interesting material will be found here, also appealing to fans of the history of technology. Theme of the book corresponds to the “Executive Order on Coordinating National Resilience to Electromagnetic Pulse”, signed by President D. Trump at March 26, 2019.

https://doi.org/10.1515/9783110639285-202

Introduction The electromagnetic pulse resulting from a high-altitude (30–400 km) explosion of a nuclear charge (HEMP) is a quite strange and extraneous phenomenon in the context of both physical processes and informational contradictions. Initially, the phenomenon was detected as a side-effect of the first nuclear-test explosion in 1945. That side-effect disrupted the registering and the measuring equipment and prevented obtaining numerous important data about the parameters of a nuclear explosion. For a long time, it has been impossible to gather clear and accurate EMP parameters due to the correlative impact on the equipment. Numerous misunderstandings and calculation mistakes made by the leading physical scientists of that time hindered the efforts to build HEMP’s theoretical models. However, recently, it appears that the theory of the process actually had been developed in 1925, well ahead of the detection of the phenomenon. The informational contradictions result from the fact that, during the decades since the initial detection of HEMP, the phenomenon has been described in meticulous detail in hundreds of extensive reports, unclassified only 20 or 30 years ago (today the reports are freely available on the internet) many of them describing HEMP impact on electronics. However, over the decades, the phenomenon has not become known or clear to the majority of civil experts in a critical national industry such as the electrical power infrastructure. It seems like they live in a parallel universe: While there exist large groups of military experts working on this problem, the leaders in the electric power industry, as well as engineers and technicians, in the best case, have only appeared to have heard something about a problem related to the electromagnetic pulse. However, the problem is really severe: HEMP acting on the unprotected microelectronic and microprocessor components of the control, telecommunication and relay protection systems commonly used in the modern power industry, can cause a power disaster over a vast area. Consequently, this will result in the suspension of the water supply, sewerage systems, communication, etc. The carelessness of government agencies, and primarily the leaders of the electrical power industry, endangers the strategic national interests and provokes potential enemy countries to develop specialized nuclear weapons with intensified HEMP (socalled Super-EMP). More than ten years ago, in 2005, I raised for the first time the question of electromagnetic security in the electrical power industry in an article published in the journal Public Utilities Fortnightly and, later that year, in articles published in Russian-language technical publications. In Russia, after my first publications, I was confronted with total misunderstanding and furious antagonism by electrical power engineering experts. They were unaware of the problem and perceived it as something outrageous and contradictory to their conventional perception of the world. https://doi.org/10.1515/9783110639285-203

X | Introduction The Western world, including the US, perceives it somewhat differently. Today, there are dozens of governmental and non-governmental entities in the US working on the protection of the national infrastructure against HEMP. There are scores of booklets alarming about the consequences of HEMP impact on the housewife’s level. A lot of such populistic scary stores can be found on Internet sites and in on-line book shops. Unfortunately, numerous of those entities produced publications by academics about the damage resulting from the global collapse that inevitably results from HEMP impact. Fundamentally, the only difference between those reports and the above booklets is that the reports are written with addition small amount of technical and economical details for frighten the Congress and the Senate, not the housewives. It may seem strange, but in parallel and completely independently of the above, there is a pattern of complete disregard and concealment of the problem in some of the professional journals in the field of the electric power industry. However, this is not a harmless pattern of behavior because the absence of appropriate knowledge about the existence of the HEMP problem by specialists in the field of the industry leads to hazardous decisions, sharply increasing its vulnerability. This is a very dangerous trend, but the people who have a certain commercial interest in all respects support it, limiting the transmittal of knowledge about the dangerous effect of the HEMP on digital microelectronic devices for protection, automation and control. Very often the apologists of digitalizing everything, including everything in the electric power industry, simply manipulate public opinion. “Digitalization – Yes or No?”—appears very popular international journal “Protection, Automation & Control World” (PAC World) in September 2018 issue (in the section “Last word”, p. 98), and continue: “This is a question that many people in the electric power industry are asking themselves”. Furthermore, the journal editor invented his own interpretation of the term “digitization” and asked the reader to accept it and say “Yes”: “Digitalization is the use of digital technologies to change a business model and provide new revenue and value-producing opportunities; if we just think about this definition, it is clear that the answer should be Yes”. “Many other questions cross our minds” – writes the editor next – “but they don’t stop us from moving forward and taking advantage of the digital technology. . . And it is the time to get to the office, bring your team together and say – Let’s go digital!” Paraphrasing the author of these lines: “the dangers to HEMP are off interest to us and they will not stop us on our way forward”. Such continued expanding uncontrolled use of digital microelectronic technologies in the electric power industry that are unprotected against HEMP is a tragedy of a national scale that must be prevented. This is the main mission of the book. Also, there are numerous companies thriving on specialized expensive training materials only developed to scare the staff of power-generating companies. Are such

Introduction

| XI

publications and trainings helpful when it comes to the protection of the national infrastructure? The answer is obvious. This answer motivated me to write a code of practice actually helpful to a power-system staff willing to protect their facilities against HEMP, instead of just being afraid of HEMP’s consequences. Certainly, to understand the nature of all the recommendations made in the book and to apply them appropriately under a specific electric equipment operation, the elementary theory of HEMP must be explained. That is why it is included in the book. Finally, to make the book both interesting and inspiring, I included recondite historical facts about the creation and testing of nuclear weapons the USSR and the US, as well as photocopies of the previously classified documents of the USSR Committee for State Security (KGB) and the CIA. Finally, I hope my readers will find much new, interesting and useful information in this book. Important document “Executive Order on Coordinating National Resilience to Electromagnetic Pulse”, recently signed by the President D. Trump gives hope that the book will be in demand by many specialists responsible for protecting the infrastructure against HEMP. Author

Contents About the Author | V Annotation | VII Introduction | IX 1 1.1 1.2 1.3 1.4 1.5 1.6 1.7

Electromagnetic pulse—a parcel from the past | 1 Introduction | 1 History of HEMP | 1 The issues of theoretical physics | 9 People’s Commissariat for Internal Affairs (NKVD) as the primary “designer” of the first Soviet nuclear explosive | 12 Thermonuclear bombs | 29 Nuclear test explosions | 36 The status of HEMP protection | 53 Bibliography | 59

2 2.1 2.2

A contemporary view of HEMP for electrical engineers | 60 Is the contemporary view up to date? | 60 The basic physical processes | 60 Bibliography | 77

3 3.1 3.2 3.3 3.4 3.5

HEMP simulators | 79 HEMP simulators: principle of operation | 79 Classification of HEMP simulators | 81 Foreign HEMP simulators | 82 HEMP simulators available in Russia and Ukraine | 88 Portable HEMP simulators | 93

4 4.1

The vulnerability of electronic equipment to HEMP | 95 Electronic equipment is the most important component of the modern infrastructure | 95 The vulnerability of discrete electronic components to HEMP | 96 Vulnerability of integral circuits (microchips) to HEMP | 99 Vulnerability of microprocessors to HEMP | 104 Vulnerability of computers to HEMP | 107 Conclusions | 108 Bibliography | 109

4.2 4.3 4.4 4.5 4.6

XIV | Contents 5 5.1 5.2 5.3 5.4

6 6.1 6.2 6.3 6.4 6.5 6.6 6.7

7 7.1 7.2 7.3 7.4 7.5 7.6

8 8.1 8.2 8.3 8.4

Electronic components for HEMP protection system | 110 Testing of low-power protective components under the low pulse voltages | 110 Testing of low-power protective components under the high pulse voltages | 114 Testing of powerful protective components under conditions close to reality | 118 Conclusions | 125 Bibliography | 126 External protection of power systems’ electronic equipment from HEMP | 128 Introduction | 128 Analysis of capability of conventional building materials to weaken electromagnetic emission | 128 Composite construction materials with improved electrical conductivity | 132 Materials absorbing electromagnetic emission | 138 Another method for depression of HEMP electromagnetic field strength inside the power industry facilities containing the electronics | 142 Reducing electronic equipment vulnerability to HEMP with architectural solutions | 145 Conclusions | 146 Bibliography | 146 The issues of electronic equipment grounding at the power facilities | 149 Types of electromagnetic interference at power facilities | 149 Challenges of the conventional grounding systems | 150 Differences between lightning and HEMP | 154 Grounding of electrical equipment as the main protective means for HEMP | 160 Protection devices for HEMP | 161 New method for grounding electronic equipment mounted inside the cabinets | 162 Bibliography | 169 The issue of control cables selection for HEMP-protected electric facilities | 171 Introduction | 171 Designs and features of shielded control cables | 171 Evaluation of control-cable shielding effectiveness | 175 Choosing control cables | 177

Contents | XV

8.5

Conclusion | 178 Bibliography | 178

9 9.1 9.2 9.3 9.4 9.5 9.6

Grounding of control-cable shields | 179 Introduction | 179 Shielding principles | 179 Interference types and grounding options for cable shields | 180 Problems and contradictions | 181 Factors impacting the effectiveness of shield groundings | 182 The suggested method of shield grounding | 185 Bibliography | 187

10 10.1 10.2 10.3 10.4 10.5 10.6

HEMP filters | 189 Introduction | 189 Do the filters really protect from an electromagnetic pulse? | 189 The frequency range of filters | 193 Feasibility of HEMP equipment protection with filters | 193 Protection of equipment from HEMP high-frequency noise | 195 Protection of the equipment from the HEMP-generated pulse overvoltage | 196 Ferrite filters | 197 Conclusions | 210 Bibliography | 211

10.7 10.8

11 11.1 11.2 11.3 11.4

12 12.1 12.2 12.3 12.4 12.5 12.6 12.7

High-voltage insulation interfaces | 212 Introduction | 212 High-voltage link for transmitting discrete commands in relay protection, automation and control systems | 212 Usage reed-switch-based high-voltage interfaces in HEMP susceptibility tests | 218 Design features of high-voltage isolation interfaces | 219 Bibliography | 221 Improvement of the resilience of industrial cabinet-installed electronic equipment to HEMP Impact | 222 Introduction | 222 New cabinets for electronic equipment | 222 Retrofitting existing cabinets equipped with glass doors | 225 Enhancement of the cabinet cable entries | 228 Voltage pulse suppression | 233 Retrofitting grounding systems of electric cabinets | 236 Conclusion | 237

XVI | Contents Bibliography | 237 13 13.1 13.2 13.3 13.4 13.5

14 14.1 14.2 14.3 14.4 14.5 14.6 14.7 14.8 14.9 14.10 14.11 14.12

15 15.1 15.2 15.3 15.4

16

Basic principles of direct-current auxiliary-power system (DCAPS) protection | 238 Introduction | 238 Protection of DCAPS operating equipment from HEMP | 238 Backup-power supplies for DCAPS systems | 240 Mobile substations and features to protect their DCAPS from HEMP | 245 Direct-current auxiliary-power systems of power plants | 251 Bibliography | 252 Protection of telecommunication systems in electric power facilities from HEMP | 253 Introduction | 253 Ways to solve the problem | 254 The use of fiber-optic communication lines | 254 Protection telecommunication equipment with galvanic couplings | 255 New devices for protecting existing telecommunication equipment | 260 Protection of the communication cabinets | 263 The general concept for communication-equipment protection | 265 Retrofitting grounding systems of cabinets containing the electronic equipment | 266 Retrofitting open-patch panels | 267 Protection of the power supply system | 267 Retrofitting the facility (room) containing the critical kinds of communication equipment | 267 Conclusion | 268 Bibliography | 268 Improvement of HEMP resilience of automatic fire-suppression systems | 269 Introduction | 269 Firefighting systems for power facilities | 269 Improvement of automatic firefighting system’s resilience to HEMP | 273 Conclusion | 278 Bibliography | 278 Protection of diesel generators from HEMP | 279

Contents | XVII

16.1 16.2 16.3 16.4 16.5 16.6

17 17.1 17.2 17.3 17.4 17.5 17.6 17.7

18 18.1 18.2 18.3 18.4 18.5 18.6

19 19.1 19.2 19.3 19.4 19.5 19.6

Introduction | 279 Increasing resilience of medium- and high-capacity DGs | 279 Protection of DGs stored and de-energized outdoors | 280 Protection of DGs connected to consumer network | 284 Active protection method for diesel-generator controller | 288 Conclusion | 295 Bibliography | 295 Features of HEMP resilience-test methods for power system electronics | 296 Introduction | 296 Features of testing equipment on a HEMP simulator | 296 Test objectives | 297 Features of the test procedure | 298 Test modes and test-pulse parameters | 300 Performance criteria | 302 Conclusion | 303 Bibliography | 304 Methods and means of evaluation of the effectiveness of HEMP protection of the installed power-system | 305 Introduction | 305 Testing of equipment resilience to direct impact of the HEMP electrical field (E1-component) | 305 Equipment for HEMP filter testing | 307 Equipment designed for evaluation of the effectiveness of building, room and cabinet shielding | 311 Pulse voltage generators | 313 Conclusion | 315 Bibliography | 315 Features of testing digital protective relays resilience to HEMP | 317 Use of performance criterion during the electromagnetic compatibility (EMC) test of electronic equipment | 317 Features of using performance criterion during the HEMP resilience test of digital protective relays (DPR) | 317 Criticism of the DPR testing method used | 318 Analysis of the result of the second independent trial of the same type of DPR | 320 Analysis of the result of the third independent trial of the same type of DPR | 323 Conclusions | 331

XVIII | Contents Bibliography | 332 20 20.1 20.2 20.3 20.4 20.5

21 21.1 21.2 21.3 21.4 21.5

A A.1 A.2 A.3 A.4 A.5 A.6 B B.1 B.2

Establishment of inventory of electronic equipment’s replacement modules as a way to improve survivability of the power system | 334 Optimization of inventory of electronic equipment replacement modules | 334 The problem of the traditional mode of SPTA storage | 335 Requirements for protective containers | 336 Protective containers available on the market | 337 Conclusion | 340 Bibliography | 341 The problem of impact of geomagnetically induced currents on power transformers and it solution | 342 Geomagnetically induced currents generated by solar storms | 342 Geomagnetically induced currents generated by HEMP | 352 The effect of the E3 component of HEMP on electric power equipment | 353 Protection of power equipment from geomagnetically induced currents | 354 Conclusions | 362 Bibliography | 363 Standards on HEMP | 365 Standards of International Electrotechnical Commission (IEC) | 365 Standards of Institute of Electrical and Electronics Engineers (IEEE) | 366 Standards of European Commission | 366 Standards of International Telecommunication Union (ITU) | 366 Military Standards (USA) | 366 NATO Standards | 367

B.3

EMP and its Impact on Power System (List of Reports) | 369 EMP Theory | 369 Geomagnetically Induced Currents and its Impact on Power System | 369 EMP Impact on Power System | 370

C

European Projects related to Protection against HEMP | 375

Index | 377

1 Electromagnetic pulse—a parcel from the past 1.1 Introduction High-Altitude Electromagnetic Pulse (HEMP) addressed in this book is one of the damaging effects of nuclear explosions, more specifically, high-altitude nuclear explosion, when the nuclear weapon explodes at the altitude of 30–400 km, i. e., in the ionosphere or even in near space (The International Aeronautical Federation has set the altitude of 100 km (a so-called Karman line) as the borderline between the atmosphere and space). Obviously, explosions occurring at this altitude cannot result in serious terrestrial destruction or people’s death. So why would somebody use it as an effective tool of war? Effecting what? Why do people talk about it 70 years after the first trial of the nuclear weapon? Let us try to answer these questions and in order to do it, we shall need to begin with the history of HEMP discovery.

1.2 History of HEMP The history of HEMP originates from the moment of the first test of a nuclear weapon. Unlike other correlative physical phenomena, it was not predicted in advance. It was rather discovered “with the point of the pen”. Furthermore, HEMP was not immediately understood and explained upon its discovery, and conclusions drawn from some subsequent proposed theories turned out to be erroneous. The first nuclear test explosion was performed within the frameworks of the famous ‘Manhattan project’ in USA at the Alamogordo range, 200 miles south from Los Alamos (New Mexico) on July 16, 1945. The 20 kt explosive (informally called “gadget”), made of Pu-239 (Figures 1.1 and 1.2), was mounted on a 33-meter steel tower (Figure 1.3), while the administration and data collection center (Figure 1.4) was located at a safe distance inside a protected building. The word “Trinity” derives from the Latin word “trinitas”. This concept is used in Christian religion to denominate the triality of the Lord: the Father, the Son and the

Figure 1.1: Delivery of the world’s first nuclear bomb to the testing site. https://doi.org/10.1515/9783110639285-001

2 | 1 Electromagnetic pulse—a parcel from the past

Figure 1.2: Hoisting up a tower (left) and preparation of the nuclear explosive mounted on a tower (right) at Trinity range.

Figure 1.3: Steel tower with the mounted “gadget”.

Figure 1.4: Administration and data-collection center.

1.2 History of HEMP

| 3

Holy Spirit. Thus, Robert Oppenheimer, (Figure 1.6) the research advisor for the project and the manager of Los Alamos laboratory, gave this name to the project under the influence of the religious poetry of John Donne (English poet of the 17th century). Simultaneously, the Conference of the Three Heads of Government was held in Potsdam. Upon receiving an encrypted notice with a coded phrase “Delivery was successful”, the US President Harry S. Truman suddenly saw himself as the ‘Ruler of the World’ and informed Joseph Stalin that the USA had invented a new type of weapon. The reason for such openness was clear: it was the USA’s habit to hold negotiations “from a position of strength”. However, this message did not surprise Joseph Stalin as much as Truman may have expected. Truman could hardly imagine that Stalin had known that the Americans were preparing and rushing to test the first nuclear bomb, so he took corresponding measures. In 1946, after the test, the Los Alamos laboratory issued a secret report LA-1012, which remained classified for more than 30 years. It was not until May 1976 that a declassified version of this report (LA-6300-Y, Figure 1.5) was issued, even though some data were deleted. Yet, the initial and general information about HEMP was published in the released version of the report. This information was published in Section 1.7 (Figure 1.5) and consisted of only several lines. This section suggested that the problems with sensor-signal transfer has been anticipated based on the theory of Enrico Fermi. Fermi determined that a nuclear explosion created X-ray emission, leading to ionization of the air over a large area and resulting in a high difference of potentials in the atmosphere (similar to lightning discharges). That is why all signal cables were thoroughly shielded and buried in the ground. However, despite all preventive measures, many types of data-acquisition equipment were knocked out of service during the explosion by a high-voltage pulse. It should be noted that electronic equipment of that time was made of vacuum tubes (Figure 1.4), i. e., it was rather primitive, unsophisticated and far less sensitive compared to modern micro-processor-based measuring equipment. This was the first-ever registered impact of EMP on electronic equipment. Further missions (Crossroads, Sandstone, Greenhouse, Buster-Jangle, TumblerSnapper, Ivy, Upshot-Knothole, Castle, Teapot, Wigwam and Redwing) involved a series of explosions and enabled research on various aspects of ground-surface, abovewater and underwater nuclear explosions, including the measurement of EMP. Upon explosion, the data-acquisition equipment was frequently knocked out of service, and so many parameters of the explosion remained unrecorded. This was attributed to the inferiority of equipment and their low reliability. Similar cases of equipment damage by a powerful electromagnetic-pulse field were registered during nuclear weapon testing conducted by Great Britain in 1952– 1953 (Maralinga range, South Australia). British physicists called this phenomenon “radio-flash”. A concept similar to what we call electromagnetic pulse today was also used in some early publications of Soviet scientists.

4 | 1 Electromagnetic pulse—a parcel from the past

Figure 1.5: Declassified version of “Trinity” project’s report.

In 1958, US scientists and military experts resumed EMP research within the Hardtack1 mission, which involved 35 nuclear explosions of various types. Some of those explosions were high-altitude explosions. The first high-altitude nuclear explosion in the US (Yucca project, Figure 1.7) was performed using a large stratospheric balloon, launched from the Essex aircraft carrier (USS Boxer CVS-21), see Figure 1.8. It was a relatively low-capacity (1.7 kt) explosive weighing about 100 kg. In addition to the explo-

1.2 History of HEMP

| 5

Figure 1.6: Robert Oppenheimer, research advisor for the project (left), and Leslie Groves, two-star general and director of the project (right), near remnants of one of the tower’s supports, dissolved during the explosion.

sive, the air-balloon carried containers with data-acquisition equipment attached to it on a long rope. The total weight of equipment carried by the air-balloon was 346 kg. The explosive was detonated 26 km above ground between Eniwetok and Bikini atolls on April 28, 1958. This explosion was expected to offer possibilities of using nuclear explosive as a means capable of impacting missile and air-defense electronic systems. In addition to data-acquisition equipment placed in five evenly spaced containers under the nuclear explosive (suspended on the same rope), a US Army research laboratory also registered HEMP parameters from two tracking stations located in the towns of Wotho and Kusale (100 and 460 miles from Bikini, respectively). Moreover, two long-range B-36 bombers equipped with data-acquisition equipment remained in the air at a safe distance from the point of explosion. Initially, the plan was to launch the balloon from the ground. However, since even a light breeze could significantly impact such a large air-balloon, a decision was made to relocate the launch to a ship. In order to avoid any shift of the balloon and its ballast at the time of launching, the ship was supposed to move in the windward direction at the wind’s speed. During the preparation stage, a total of 76 air-balloons with ballast dummies were launched from various locations. The Hardtack mission involved two additional high-capacity, high-altitude nuclear explosions coded “Teak” and “Orange”. For the purpose of the “Teak” project, the thermonuclear explosive W-39 (3.8 Mt) was delivered to the altitude of 77 km by a Redstone missile (Figure 1.9), which was launched from one of Johnson Atoll’s islands on July 31, 1958. The outcome of this nuclear explosion was a powerful EMP, which could not be recorded in the testing area due to the failure of data-acquisition

6 | 1 Electromagnetic pulse—a parcel from the past

Figure 1.7: Report on Yucca project within Hardtack-1 mission declassified in 1997.

equipment, errors in preliminary calculations and predictions made by Hans Bethe (the famous theoretical physicist) regarding EMP. Nonetheless, a powerful geomagnetic storm, much stronger than geomagnetic agitations caused by magnetic storms on the Sun, and some flashes in the sky were registered far away from the explosion epicenter in the observatory of Apia town (Upolu

1.2 History of HEMP

| 7

Figure 1.8: Preparation (left) and launch (right, top-right corner) of the air-balloon with nuclear explosive W-25 attached to it (from the deck of Essex aircraft carrier).

Figure 1.9: Missiles (left to right): Thor, Redstone (USA) and Soviet SS-4 Sandal, which were used during tests to deliver nuclear devices to the altitude of the explosion.

island in the Pacific Ocean)—the capital of Samoa, located 3,200 km from Johnson Atoll. During the “Orange” project, (August 11, 1958), a nuclear explosive of the same type was delivered by the same missile from the same range, but to an altitude of 43 km. Unfortunately, the search for any additional data regarding the EMP effect in this project in published or declassified literature was unsuccessful. Some authors have claimed recurring problems with data-acquisition equipment, whereas others have suggested that information about this project is still classified. However, fragmentary data published in partially declassified (1999), to be more precise, “sanitized” (“sanitized version” is what is written on the front page), report ITR-1660 regarding

8 | 1 Electromagnetic pulse—a parcel from the past

Figure 1.10: Available pages of partially declassified (“sanitized”) ITR-1660 report (some pages have been deleted completely) regarding the Hardtack-1 mission.

Hardtack-1 in 1959 (Figure 1.10), suggest that many registered EMP parameters appeared to be different from what was expected or what can be derived from the theory. In particular, the EMP intensity from a high-altitude explosion was three orders of magnitude higher than from a ground-surface explosion. Furthermore, some pieces

1.3 The issues of theoretical physics | 9

of measuring equipment (similar to what happened during the very first testing of a nuclear explosion) were knocked out of service, and all the data were lost. Since the registered data did not comply with the predicted theoretical data, they were simply disregarded until testing of nuclear explosives was resumed in 1962 (Fishbowl mission). These tests returned data similar to those obtained during the Hardtack-1 mission, which were initially disregarded as seemingly erroneous. One of the most renowned projects within this mission was Starfish Prime, which was deployed on July 9, 1962. The mid-range missile GRM-17 (“Thor”), see Figure 1.9, delivered a thermonuclear explosive W49 (1.44 Mt) to a 1,100-km altitude above Johnson Atoll in the Pacific Ocean. Upon detachment, the explosive descended to 400 km, where it was remotely triggered. An unexpected outcome of this explosion was registered in Hawaii, located 1,445 km from the explosion epicenter: simultaneous failure of 30 lines of street lighting, actuation of multiple alarms and disruption of telephone systems. It took several tests, in which the recorded HEMP parameters did not comply with the theory, to understand that there was something wrong with the theory.

1.3 The issues of theoretical physics Arthur Compton from Cambridge University suggested the fundamentals of HEMP theory when studying the scattering of X-rays by paraffin atoms. In 1923, he found that X-rays scattered in paraffin have higher wavelengths compared to those falling onto them. Classical electrodynamics of J. J. Thompson could not explain this phenomenon. A new physical phenomenon had been discovered: the effect of elastic scattering of short-wave electromagnetic emission (X-rays and gamma rays) on unbound atoms’ electrons, accompanied by elongation of the wavelength and emergence of additional unbound electrons. In other words, Compton experimentally found that X-rays reflected by electrons act as though they consisted of separate particles (Figure 1.11). Thus, he was the first to affirm Einstein’s theory regarding the existence of light quanta. This discovery indirectly affirmed the corpuscular theory of electromagnetic emission, more specifically—light (which had earlier been rejected by Max Planck and Niels Bohr). In 1927, A. Compton was awarded the Nobel Prize for this discovery. During World War II, A. Compton was a key member of the Manhattan project, developing the world’s first nuclear weapon. He was the director of a so-called Metallurgic Laboratory, which produced weapon-grade plutonium for the first nuclear bombs. However, the disputes regarding quantum physics in general, and the quantum theory of light in particular, did not subside. According to a prominent physicist, Academy Fellow, Yakov Borisovich Zeldovich [1], an experimental work suggesting that Compton’s theory was incorrect appeared back in 1933. Physicists were confused. New, more thorough work was necessary (the first and the best in this list belongs to brothers

10 | 1 Electromagnetic pulse—a parcel from the past

Figure 1.11: Arthur Compton and illustration of his theory in relation to HEMP.

A. I. and A. I. Alikhanov and L. A. Artsymovich) to restore the truth and rehabilitate Compton’s formula. Compton’s theory was used to explain experimentally discovered powerful electromagnetic pulses near the ground surface during high-altitude nuclear explosions. According to this theory, the interaction of powerful X-ray emission, caused by highaltitude nuclear explosion in the air, with unbound electrons of air atoms leads to the emergence of a large number of additional unbound electrons, (Compton’s electrons) captured by the magnetic field of the Earth. This powerful stream of electrons rapidly descending towards the Earth’s surface induces an upsurge of the electric field near at the surface. However, when using Compton’s effect for theoretical substantiation of HEMP parameters, some false theories were implemented. For example, in 1957 Hans A. Bethe, the famous American physicist and Nobel Prize winner (Figure 1.12) published (in a secrete report, which is classified until now) a theory of electromagnetic pulses based on

Figure 1.12: Founders of HEMP theory: Hans A. Bethe, A. S. Kompaneets, V. Gilinskiy.

1.3 The issues of theoretical physics | 11

extrapolation of experimental data obtained during ground-surface trials of nuclear explosives. Nuclear explosions in 1958 and 1962 previously described showed that the values of the electromagnetic-field density of EMP during a high-altitude explosion, obtained using Bethe’s theory, are 1,000 times lower than actual values. In the USSR, the HEMP theory was also addressed simultaneously with H. Bethe’s theory, but independently from him. While the theory of H. Bethe was based on experimental data, obtained during ground-surface nuclear explosion, Soviet physicists did not have such an opportunity and their calculations were solely theoretical. The first model was developed by Prof. Alexander Solomonovich Kompaneets, famous Soviet theoretical physicist, Doctor of Physics and Mathematics (see Figure 1.12), a representative of the Kharkov physics school and a student of its founder— worldwide renowned L. D. Landau. Since 1946 and until the last day of his life, A. S. Kompaneets worked at the Institute of Chemical Physics of Academy of Science of USSR. Due to the beginning of his work under the Soviet nuclear project, the Council of Ministers of USSR approved a top-secret Decree No. 973-40ts (dated April 30, 1946): “On assistance to the Institute of Chemical Physics of Academy of Science of USSR”. One of the sections of this decree was called “Establishment of a special sector on studying the theory of nuclear chain reactions and explosions at the ICP of Academy of Science of USSR”. There were six departments within the Special Sector. A. S. Kompaneets was the Head of Theoretical Physics Department. The “radio emission of a high-capacity explosion” (today called HEMP) was one of the principal fields of A. S. Kompaneets’ activity. He developed a, internationally renowned equation that was later named after him (Kompaneets’ equation). The equation described changes in the photon-gas emission at Compton’s scattering in lowdensity plasmas [2]. That became a cornerstone equation used by physicists all over the world. They used this equation until American physicist Victor Gilinski, Head of Physics Department of the Rand Corp. (USA) proved [3] the erroneousness of Kompaneets’ theory (Figure 1.13).

Figure 1.13: Introduction to V. Gilinski’s work [3] with criticism of A. S. Kompaneets’ theory.

12 | 1 Electromagnetic pulse—a parcel from the past

1.4 People’s Commissariat for Internal Affairs (NKVD) as the primary “designer” of the first Soviet nuclear explosive Long before Americans started testing nuclear explosives, Soviet intelligence of NKVD started looking for any work in this field in all the countries, especially in the USA and Great Britain. Soviet intelligence operation in the field of nuclear bomb was codenamed “Enormoz”. This name was given to the project by the 1st Directorate of NKVD of USSR in 1941. The overall management of the “Enormoz” operation was performed by the Head of the 1st Directorate (foreign intelligence) of NKVD–NKGB, 3rd rank Commissioner of homeland security, Pavel Fitin. The operation itself was developed by the Head of the 3rd (UK–USA) Department of the 1st Directorate, Commissioner of Homeland Security, Gike Ovakimian, who worked as a New York spymaster until 1941 which involved the cooperation of the Rosenbergs. Major Leonid Kvasnikov, the deputy of the New York spymaster, was appointed to perform the mission. It should be noted that he was one of the founders of Soviet scientific and technical intelligence, when he worked as the Head of the 3rd Division of the 3rd Department of the 1st Directorate in 1939. While the office of L. Beria was fully involved in “procuring” nuclear secrets, the Red Army’s Intelligence Service (RAIS) had only had poor chances to learn about the nuclear issue. This is evidenced by a secret letter of the Head of the Second Directorate of Intelligence Service to the Head of Special Department of Soviet Academy of Science, M. P. Yevdokimov (Figure 1.14).

Figure 1.14: Letter of RAIS No. 137955s dated May 7, 1942 to the Special Department of Soviet Academy of Science. In this secret letter, Panfilov, Deputy Head of the 2nd Directorate of the Main Intelligence Directorate of the Red Army and Commissar General Kiselev, asked the head of the Special Department of the USSR Academy of Sciences Evdokimov about the plausibility of reports regarding the possibility of using uranium nuclear-fission energy for military purposes and about Niels Bohr laboratory in Copenhagen.

1.4 NKVD as the primary “designer” of the first Soviet nuclear explosive | 13

The Special Department of Soviet Academy of Science was dealing with defense issues. The background of Mikhail Prokopiyevich Yevdokimov, the Head of the Department, was as an engineer–metallurgist, so he was not familiar with atomic physics. That is why he forwarded the letter to Director of Radium Institute of Academy of Science, Academy Fellow, V. G. Khlopin. Two weeks later the Academy Fellow wrote the answer: The Academy of Science does not have any data regarding the work progress of foreign laboratories on the use of internal energy released upon uranium fission… Moreover, there were no publications on this issue in scientific literature (which we have access to) during the last year. This situation makes me think that the corresponding work is of top significance and thus, it is performed in absolute secrecy.”

Despite the passive attitude of the Intelligence Service and Academy of Science of USSR, the intelligence activity of NKVD was successful. In fact, in spite of the extremely difficult defense emergency of the USSR during that time, the State Defense Committee issued the top-secret Resolution No. 2352ts “Regarding uranium handling” in September of that year (Figure 1.15).

Figure 1.15: Resolution GKO-2352ts to resume work on nuclear energy research in order to build a U-bomb. This top-secret document of the USSR State Defense Committee (GKO) is entrusted to the USSR Academy of Sciences (academician Ioffe) to resume work on nuclear energy; create a special laboratory; develop and create the necessary equipment. Submit by April 1, 1943 Report on the possibility of creating a uranium bomb.

14 | 1 Electromagnetic pulse—a parcel from the past

Figure 1.16a: Extracts from Resolution GKO-7102 ts/pi regarding extraction and processing of uranium. In this top-secret document of December 8, 1944 under the signature of I. Stalin, the NKVD is charged with organizing all the necessary work for the extraction and processing of uranium ores, including the erection of mines and factories in the USSR, the creation of a uranium research institute, the purchase abroad of the necessary equipment and technical literature. Separately indicated was an increase of 50 % of the nutritional standards of prisoners working in these mines and factories.

In 1944, the State Defense Committee issued a “Top Secret/Particular Importance” (ts/pi) Resolution No. GKO-7102 ts/pi (SDC-7102 top secret/particular importance) regarding the extraction and processing of uranium (Figures 1.16a, 1.16b). Management of extraction and processing of uranium was again entrusted to L. Beria and his department, including creation of a Research and Development Institute within NKVD, i. e., the Institute of Special Metals (InSpecMet of NKVD).

1.4 NKVD as the primary “designer” of the first Soviet nuclear explosive | 15

Figure 1.16b: Continued. Extracts from Resolution GKO-7102 ts/pi regarding extraction and processing of uranium.

Two weeks later, after the Hiroshima bombing, the State Defense Committee issued the Resolution No. GKO-9887ts/pi dated August 20, 1945 (Figure 1.17). This resolution stipulated the creation of a Special Committee chaired by Lavrentiy Beria that was supposed to be responsible for the “management of all activities on the use of uranium’s internal energy”. The committee was granted extraordinary powers and unrestricted financing. The First General Directorate (FGD) of NKVD became the executive body of the Special Committee. Two more bodies were established within FGD, i. e., Scientific and Research Council (SRC) and Bureau No. 2. Department “S”, established as part of P. Sudoplatov’s team (Figure 1.18), became the operational administration of Bureau No. 2. The most important strategic materials, including 200 pages from the “Enormoz” case, were transferred to this department. Among other things, it is interesting to review clause 13: “Comrade Beria shall endeavor to organize foreign intelligence work to obtain more comprehensive technical and economic information about uranium industry and nuclear bombs. Moreover, he is appointed to manage every aspect of intelligence work performed by intelligence agencies in this field.” In December of that year, the Resolution of the Council of People’s Commissars No. 3117-937ts/pi stipulated creation (again as part of NKVD) of a separate The Ninth Directorate of Special Institutes chaired by a two-star general A. Zaveniagin, and further subordination of laboratories and factories dealing with “Problem No. 1” to it (please note that the creation of the nuclear bomb was coded as “problem no. 1” even in the-top secret document). Indeed, the NKVD became the principal implementer of the nuclear project in the USSR (Figure 1.19).

16 | 1 Electromagnetic pulse—a parcel from the past

Figure 1.17: Extracts from Resolution of the State Defense Committee (GKO) No. GKO-9887ts/pi. In this top-secret document of August 20, 1945, the USSR GKO decided to organize a special committee to supervise all work on the use of uranium atomic energy under the guidance of the NKVD Chairman L. Beria.

Lavrentiy Pavlovich Beria was a disciplined and proactive performer; since he was an excellent administrator, he became very involved in the work. Soviet intelligence has always watched this issue very closely. However, the intensity of their work increased significantly after Hiroshima and Nagasaki. The best spies were moved to this project. Yakov Terletsky, Doctor of Physics and Mathematics, a lieutenant colonel, organized and coordinated all the incoming materials and reported during the Scientific Research Committee (SRC) meetings. Initially, the chairman of SRC was the People’s Commissar in charge of ammunition, one of the first triple Heroes of Socialist Labor, two-star general Boris Vannikov, while his deputy (and then the chairman) was Academy Fellow Igor Kurchatov, who

1.4 NKVD as the primary “designer” of the first Soviet nuclear explosive | 17

Figure 1.18: Top-secret Order of NKVD No. 001094 ts/pi signed by L. Beria regarding establishment of Special Department “S” as part of NKVD chaired by two-star general P. Sudoplatov for perform special tasks on the uranium problem.

chaired the SRC until the end of his life. In addition, the following people were part of SRC: Beria’s deputies Vasily Machnyov and Avraamiy Zaveniagin, as well as Academy Fellows Abram Ioffe, Abram Alikhanov, Isaac Kikoin, Vitaly Khlopin and Yuliy Khariton. Leonid Romanovich Kvasnikov was called back from New York at the end of 1945 to work on the project. In 1947, he chaired scientific and technical intelligence of NKVD and remained in this position until his resignation in 1966, disregarding various reforms and the renaming of agencies. Obviously, the Soviet Union was intensively working on the nuclear project, but it lagged behind the USA in terms of practical deployment. Specialists from almost all the fields of science and industry (provided the issues addressed were within their competence) were invited to the meetings of the Special Committee (Figure 1.20).

18 | 1 Electromagnetic pulse—a parcel from the past

Figure 1.19: Resolution of Council of People’s Commissars No. 3117-937ts/pi. In this top-secret document, the Council of People’s Commissars of the USSR decided to create a Directorate of Special Institutions “on Problem No. 1” within the structure of the NKVD (Directorate No. 9) and to transfer the laboratories “A” and “G” to support this directorate, as well as organize new laboratories “B” and “C” in the structure of the NKVD to use in them as imprisoned specialists, as well as German specialists, who are subject to isolation from living among the prisoners of war.

Among other issues, the Minutes of the Special Committee No. 9 (Figures 1.21a–1.21c) stipulated that the committee consisting of the leading nuclear scientists of the country was expected to analyze all the materials concerning the explosions in Hiroshima and Nagasaki. In these minutes, factory 813 designates the Production Facility “Mayak”, while factory 817 designates the “Ural Electrochemical Plant”. Intelligence officers from NKVD collected information from the leading physicists and nuclear experts, including those working at the National Laboratory in Los

1.4 NKVD as the primary “designer” of the first Soviet nuclear explosive | 19

Figure 1.20: Heads of the USSR’s nuclear project, employees of NKVD: Commissar of Internal Affairs L. Beria; People’s Commissar in charge of ammunition B. Vannikov (he was not NKVD’s employee, but he worked with it very closely); Head of Directorate of Foreign Intelligence of NKVD P. Fitin; 3rd -rank Commissar of Homeland Security (two-star general), Deputy of Commissar of Internal Affairs, Head of The Ninth Directorate of NKVD A. Zaveniagin; Head of Department “C” of NKVD P. Sudoplatov; Head of Department of Scientific and Technical Intelligence, colonel L. Kvasnikov; expert of Department of Scientific and Technical Intelligence, Doctor of Physics and Mathematics, Professor, Lieutenant Colonel Ya. Terletsky.

Alamos, such as Klaus Fuchs, Ted Holl, Morton Sobell and David Greenglass. They were contacted by New York spymasters of NKVD, Alexander Feklisov and Anatoliy Yatskov, as well as by Harry Gold and the Cohen couple (recruited US citizens). Robert Oppenheimer, research advisor of the Manhattan project and director of the Los Alamos laboratory, who sympathized with socialists and adhered to left-wing views, was an object of special attention from Soviet intelligence service. Grigory Heifets, spymaster of Soviet intelligence acting as a vice-consul of USSR in SanFrancisco, managed to establish a close contact with him. Simultaneously, the wife of Vasily Zarubin (Soviet spymaster in New York), Yelizaveta Zarubina (Major of NKVD) became acquainted with Oppenheimer’s wife Kathrine, who used to be a member of the US Communist party. Upon Zarubina’s request, Kathrine convinced the “fathers” of a nuclear bomb—Enrico Fermi and Leo Szilárd—to permit some specialists recruited by Soviet intelligence to participate in the Manhattan project. The profusion of Soviet spies around the “Manhattan Project” was remarkable. About 200 Soviet intelligence officers worked to collect data about the US nuclear bomb. Foreign intelligence involved 14 especially valuable agents, i. e., foreigners who participated in the US nuclear project and made a significant contribution to the development

20 | 1 Electromagnetic pulse—a parcel from the past

Figure 1.21a: Top-secret minutes of the Special Committee’s Meeting No. 9. In this top-secret order signed by L. Beria and dated November 30, 1945, the creation of the Engineering and Technical Council for the design and construction of factories and special equipment for the use of “intraindustrial resources” (the encrypted name for atomic energy) and for the production of “product 180” (the encrypted name for enriched uranium) were authorized. Also the creation of a commission to study the effects of the use of atomic bombs in Hiroshima and Nagasaki was approved, including a limited list of persons allowed to discuss the problem.

of the Soviet nuclear bomb. The list includes renowned physicist Klaus Fuchs, the Rosenbergs, (sentenced to death in the electric chair), as well as deep-cover agents Leontina and Morris Cohen. During the several-year-long mission, Soviet intelligence officers undoubtedly scouted huge amounts of secret documents, which consisted of 12,000 pages in aggregate. Soviet scientists who developed a nuclear bomb in the so-called “Laboratory No. 2” hardly imagined that all those pieces of information, and sometimes even research findings (Figure 1.22), were obtained by NKVD. They really thought that these bites of information came from some R&D centers within the country, who were working with them simultaneously on the same problem.

1.4 NKVD as the primary “designer” of the first Soviet nuclear explosive | 21

Figure 1.21b: Minutes of the Special Committee’s Meeting No. 9. Continued.

Figure 1.21c: Appendix to Minutes of the Special Committee’s Meeting No. 9.

22 | 1 Electromagnetic pulse—a parcel from the past

Figure 1.22: A page from Soviet intelligence report. This is a top secret document intended for the scientific supervisor of works in the field of the atomic problem I. Kurchatov. The document presents technical information obtained by Soviet intelligence on the design of the US atomic bomb based on uranium-235.

Shortly afterwards, nuclear experts received a rather clear assignment: replicate the US bomb with minimum changes and supplements. In order to do this, the Cabinet Council of USSR issued a Decree No. 805-327ts/pi dated April 9, 1946 to establish a special engineering department—Constructor Bureau “CB-11” (Figure 1.23). It is amusing to see today how they used vague terms, such as “jet engine”, to denominate a nuclear bomb in a top-secret document meant to be stored in a special folder, or “Plant 550 of the Ministry of agricultural machine-building”, located in a detached village Sarov in Mordovia (since 1946, the name ‘Ministry of agricultural machine-building’ was assigned to former People’s Commissariat in charge of ammunition). By the way, CB-11 was called “facility 550” in many additional governmental documents. In fact, the first nuclear bomb was denominated in the documents as RDS-1 (meaning “jet engine named after Stalin”). The abbreviation RDS with different numbers was used for a long time and was assigned to multiple nuclear explosives developed by CB-11 (today,

1.4 NKVD as the primary “designer” of the first Soviet nuclear explosive | 23

Figure 1.23: Decree of the Cabinet Council of USSR of April 9, 1946. This top-secret decree was intended to organize a special constructor’s bureau (CB-11) to create “jet engines” (the encrypted name of the atomic bomb) under the guidance of Professor Y. Khariton and to organize research work on the nuclear explosion at the institute Chemical Physics Academy of Sciences of the USSR.

this facility is called “All-Russian Research and Development Institute of Experimental Physics”). In February 1947, a Resolution of the Cabinet Council of USSR granted CB-11 the status of a high-security facility and converted its territory to a closed restrictedaccess area. The place-name Sarov was excluded from toponyms of the Mordovia

24 | 1 Electromagnetic pulse—a parcel from the past Autonomous Soviet Socialist Republic and was deleted from all records (it was not used on geographic maps and public mass media). Later on, this village (today this is a restricted-access area) existed under different fictional names, (even in secret documents) such as Base No. 112, Gorky-130, Arzamas-75, Kremlyov, Arzamas-16 and Moscow-300. In 1945, simultaneously with intensive intelligence activity performed by NKVD, the office of A. Zaveniagin found German specialists (metallurgists, chemists and physicists) and delivered them to the USSR. The list included: Nikolaus Riehl, Manfred von Ardenne and others. In total, 70 specialists were delivered to the USSR in 1945. This number increased to 300 people by 1948. Later on, A. Zaveniagin was in charge of “German” laboratories. Simultaneously with this, P. Ya. Meshik and I. K. Kikoin were responsible for the search for technological equipment, ore reserves and excavated uranium raw material over the entire territory of Germany controlled by the USSR. In total, 220 tons of uranium compounds (pure metal equivalent) were found by mid-1946. In the middle of 1946, a geologic exploration was performed by 320 geologic parties all over the USSR. As a result, in addition to the Taboshar uranium field, they started excavation of uranium compounds in the Krivoy Rog basin, Estonia and the Transbaikal region. They resumed excavation in Jáchymov (Czech Republic) and started development in Saxony of the mines of the future SDAG Wismut. By 1947, the USSR owned enough resources and scientific data to start building its own nuclear bomb. However, technical information stolen from the Americans was sufficient to start this job. That is why L. Beria insisted on simple copying of the “Fat Man”—the American bomb dropped on Nagasaki. Outraged, Peter Kapitsa (who was involved into the nuclear project on the recommendation of I. Kurchatov) wrote a personal letter to J. Stalin, where he challenged Beria’s opinion. As a result, P. Kapitsa was dismissed from the project, and Beria’s opinion became the law for bomb designers. As Yu. B. Khariton (the research supervisor of Soviet nuclear project and chief engineer of a nuclear bomb) confessed: “the first Soviet nuclear explosive was built based on American sample.” The Chief of NKVD’s Department “S”, who was responsible for procuring information regarding development of the US nuclear bomb, twostar general P. A. Sudoplatov also wrote that “the first Soviet bomb (RDS-1) was a copy of American plutonium bomb dropped on Nagasaki to the smallest details.” At the same time, Sudoplatov emphasized that “the design of the first US nuclear bomb was available in the USSR 12 days after the bomb’s assembly.” As Academy Fellow Ya. Zeldovich joked, this bomb was “full drawn” from the Americans. Thus, it can truly be said that the first Soviet nuclear bomb was developed (and here “development” means not only theoretical development, but all the issues related to the construction of new factories, establishment of production of necessary parts, ore extraction, etc., as well as the procurement of drawings and descriptions of the nuclear bomb) inside the belly of NKVD. I. V. Kurchatov had a separate office at NKVD Directorate in Moscow (Lubyanka Street, bldg. 2) to make his work more comfortable (Figure 1.24).

1.4 NKVD as the primary “designer” of the first Soviet nuclear explosive | 25

Figure 1.24: Building of NKVD (KGB) Directorate in Lubyanka Street (Moscow).

Figure 1.25: Founders of Soviet nuclear physics, whose scientific efforts provided theoretical substantiation and deployment of the first Soviet nuclear bomb: Igor Vasiliyevich Kurchatov, Yuliy Borisovich Khariton, Yakov Borisovich Zeldovich, Lev Davidovich Landau, Abram Isaacovich Alikhanov, Georgy Nikolayevich Flerov, Konstantin Antonovich Petrzhak, Isaac Konstantinovich Kikoin.

Of course, we should not belittle the role of the leading Soviet nuclear experts who participated in building of the first nuclear bomb (Figure 1.25), though it was NKVD’s game and especially its Chairman Lavrentiy Beria.

26 | 1 Electromagnetic pulse—a parcel from the past Here is what Yu. B. Khariton (research supervisor of the project and Chief engineer of the first Soviet nuclear bomb) wrote in his book [4]: Beria quickly instilled the necessary amplitude and dynamism to the work. This person, regarded as embodiment of evil in the contemporary history of the country, possessed mega-energy and fitness to work simultaneously. Our specialists, who interacted with him, could hardly miss his intellect, will and motivation. They become convinced that he was the top orchestrator, who can get the job done. Surprisingly, but Beria, who sometimes manifested offensive behavior, could (depending on situation) be polite, tactful and just a normal human being. The meetings that he chaired were always efficient, business oriented and never exhaustingly long. He was a master of unexpected and extraordinary solutions. M. A. Sadovskiy found himself at an absolutely different meeting chaired by Beria. There were about 30 people in his Kremlin office in Moscow. They discussed preparation of the range for the first thermonuclear explosion. Speakers tried to explain, how machines will be located, which facilities need to be built and how as well as which experimental animals need to be brought to the field in order to study the outcomes of damage effects. Suddenly, Beria went unhappy getting more and more anxious. He was interrupting speakers, changed people that were supposed to report to him, asked strange questions, which were difficult to answer. Eventually, he blew up and (according to M. A. Sadovsky) being completely unsatisfied he shouted: I will do it myself!” Then Beria started talking nonsense. Step by step we understood the following from his monologue: he wants the explosion to destroy everything on the range. It had to be frightening! When the meeting was over, the participants left depressed. That was the first time, when Mikhail Alexandrovich understood that dealing with Beria is not a joke… Beria worked quickly, he didn’t neglect site visits and got familiar with the results of performed work personally. When we performed our first nuclear explosion, he was the Chairman of the State Commission. Despite his exceptional position in the Party and the Government, Beria managed to allocate time for personal communication with people he was interested in, even though they didn’t have any official honors or top-notch titles. He met with A. D. Sakharov (PhD in Physics and Mathematics at that time) several times as well as with O. A. Lavrentyev (just demobilized sergeant from the Far East).

Beria demonstrated tolerance and appreciation when he needed any specialist for any work, even though people from his department thought that person was suspicious. When the Security Service decided to remove L. V. Altshuhler (who explicitly favored genetics and disliked Lysenko) from the site as a security risk, Yu. B. Khariton called directly to Beria and said that this employee was very beneficial in his job. The conversation was short and the almighty person asked just one question after a long silence: “Do you really need him?” And having received a positive answer he said: “Alright” and hung up. The subject was closed. According to many veterans of the nuclear industry, had this project been under Molotov’s control, it would have been difficult to expect any quick success performing such tremendous work. We can treat Beria’s personality differently, and there is no doubt that he is guilty of the deaths of millions of innocent Soviet people. At the same time, we should recognize: his personal drive and organizational abilities, as well as severe reprimands during the meetings and far-from-idle threats to shoot people in case of failure; thou-

1.4 NKVD as the primary “designer” of the first Soviet nuclear explosive | 27

sands of prisoners building plants and laboratories for the nuclear project; and powerful scientific and technical intelligence, enabling the prevention of perhaps an even larger global disaster. Indeed, the fact that the USSR tested and built its nuclear bomb much earlier than was expected in the USA and UK (1953–1954), cooled down some American war hawks and helped to avoid deployment of multiple (totally 13) plans for preventive nuclear bombing of the USSR. For example, according to the “Dropshot” plan, 300 nuclear bombs (50 kilotons each) were supposed to be dropped on 100 large industrial centers of the USSR in order to demolish 85 % of its industrial and military potential, including 25 bombs on Moscow; 22 on Leningrad; ten on Sverdlovsk; eight on Kiev; five on Dnepropetrovsk; two on Lvov, etc. In addition, the additional plan envisaged dropping of 250,000 tons of ordinary bombs. Estimates suggested that roughly 60-million Soviet citizens would have died in the event of such bombing. Luckily, those plans were destined to remain on paper. In the summer of 1949, the completed product “501” (code name of the first Soviet nuclear bomb) was dismantled, packed and sent from CB-11 to “training range No. 2” (the former name of the Semipalatinsk nuclear testing range) by a special train. Final elements of the bomb were mounted onto its central part during the night of August 29, and at 7 a. m. it was successfully detonated. The bomb’s capacity equivalent to TNT was 22 kt, length was 3.7 m; diameter was 1.5 m; and its weight was 4.6 tons (Figure 1.26). Similarly to the American “Trinity” project, the nuclear explosive was mounted on a steel tower of the same height (33 meters), see Figure 1.27.

Figure 1.26: Prototype of RDS-1 bomb displayed at the All-Russian Research and Development Institute of Experimental Physics and Academy Fellow Yu. B. Khariton in 1993.

Upon successful testing of the bomb, L. Beria wrote a joyful report to Stalin (Figure 1.29). Afterwards, the Special Committee prepared a special Resolution of the Cabinet Council of USSR “Regarding testing of the nuclear bomb” (Figure 1.30). The majority of the scientists who engineered the bomb were granted governmental

28 | 1 Electromagnetic pulse—a parcel from the past

Figure 1.27: Nuclear explosive RDS-1 and the tower where it was mounted for detonation.

Figure 1.28: Sensational headlines of American newspapers about USSR’s testing of the nuclear bomb.

awards. The project supervisor I. V. Kurchatov received special laurels from J. Stalin (Figure 1.31). Successful testing of the Soviet nuclear bomb (which was named “Joe-1”, derives from Josef Stalin, in the USA) was a surprise to Americans, and the US newspapers screamed about this sensation at the top of their voices (Figure 1.28). When awarding Kurchatov after the successful bomb testing, J. Stalin said: “If we delayed building of the nuclear bomb for 1–1.5 years, we would have “tried” its effect on ourselves.”

1.5 Thermonuclear bombs | 29

Figure 1.29: Chirograph of L. Beria to J. Stalin. In this top-secret letter from the head of the NKVD, L. Beria reported to J. Stalin on the successful creation and testing of the first Soviet atomic bomb.

1.5 Thermonuclear bombs Long before RDS-1 testing in 1945, NKVD started obtaining intelligence information about a new type of nuclear bomb—a so-called “super-bomb”—that was under development in the USA. This led to establishment of a small research group at the Institute of Chemical Physics of Academy of Science of USSR, chaired by Yakov Zeldovich. In the spring of 1948, Klaus Fuchs delivered an especially large and detailed file about the super-bomb to Soviet intelligence. To work on this, an additional group was formed (intended to help Zeldovich’s group), lead by Igor Yevgeniyevich Tamm at the Physics Institute of Academy of Science of USSR (PIAS) in the summer of 1948. Tamm’s students—Andrey Dmitriyevich Sakharov and Vitaliy Lazarevich Ginzburg (Figure 1.32)—were also part of this group.

30 | 1 Electromagnetic pulse—a parcel from the past

Figure 1.30: Minutes of the Special Committee’s meeting regarding approval of the draft of Resolution of the Cabinet Council of USSR “Regarding testing of a nuclear bomb” signed by L. Beria.

After visiting CB-11 and becoming familiarized with the design of RDS-1 (in the fall of 1948), D. Sakharov decided to add a layer of light-weight elements into a uranium container surrounding the plutonium core. Thus, a scheme was utilized in which fission and fusion fuel were “layered”, a design known as the “Sloika” (Russian layered cake). A similar design was much earlier theorized by Edward Teller in the USA. At the end of August 1946, Teller proposed a new bomb configuration that he dubbed the “Alarm Clock.” The scheme alternated spherical layers of fissionable materials and thermonuclear fuel (deuterium, tritium and possibly their chemical compounds). In September 1947, Edward Teller proposed the application of a new thermonuclear fuel in the “Alarm Clock”: lithium-6 deuteride, which was supposed to greatly enhance the production of tritium during the explosion, thereby substantially increasing the thermonuclear combustion efficiency. Still unbeknownst to the American team, in March 1948 in London, Klaus Fuchs provided a Soviet agent with information on the progress

1.5 Thermonuclear bombs | 31

Figure 1.31: Extract from Resolution of Cabinet Council of USSR regarding awarding Kurchatov I. V. for “outstanding merits to Soviet Motherland”. This top-secret decree of the Council of Ministers of the USSR of October 29, 1949 refers to awarding the atomic energy research supervisor I. Kurchatov a considerable monetary prize, a personal house and a dacha, and awarding him with honorary titles of the USSR, etc.

Figure 1.32: Founders of the theory of Soviet thermonuclear bomb Academy Fellows: Igor Yevgeniyevich Tamm, Andrey Dmitriyevich Sakharov and Vitaliy Lazarevich Ginzburg.

of the “Classical Super” (Klaus Fuchs departed from Los Alamos on June 15, 1946, but research and development on the hydrogen bomb continued). This design was expected to significantly increase the blast capacity. The other participant of the group—V. Ginzburg—suggested substituting heavy water (suggested by D. Sakharov) with a promising lithium-6 deuteride (as in” Alarm Clock”).

32 | 1 Electromagnetic pulse—a parcel from the past In 1949, the Directors of PIAS S. I. Vavilov and Yu. B. Khariton informed L. Beria about a so-called Sakharov’s “sloika”, so a new R&D plan to build a thermonuclear bomb (RDS-6) during 1949–1950 was approved during the summer meeting of CB-11 in 1949. This plan envisaged putting efforts both into Sakharov’s “sloika”—RDS-6s and into “Tube”—RDS-6t (the so-called project “Classic Super”, stolen in the USA). After Harry Truman announced that the USA would start working on a “thermonuclear or super-bomb” (January 31, 1950) as a response to successful testing of a nuclear bomb in USSR, the Special Committee called a meeting chaired by Beria, who resolved to accelerate the efforts to create its own super-bomb. In order to implement this solution, a group headed by I. A. Tamm, which consisted of A. D. Sakharov, Yu. A. Romanov and Tamm himself, was moved to CB-11 to work on RDS-6s (“sloika”). Two members of his group were Jews, namely Vitaliy Ginzburg and Yefim Fradkin, who stayed at PIAS as they faced problems with getting access to top-secret jobs. Semen Belenkiy was also not moved, due to health issues. The group of Ya. Zeldovich kept on working on foreign “Tube” at the Institute of Chemical Physics of Academy of Science of USSR until the end of 1953 (in other words, almost six years), when these efforts were recognized as meaningless. The Americans had drawn this conclusion four year earlier, based on the calculations of American mathematician Stanislav Ulam, who proved the unfeasibility of the “Classic Super” project using mathematical methods in 1950. Edward Teller (Figure 1.33), an American physicist, used this conclusion to define new principles for the thermonuclear bomb’s design in 1951. This design of a thermonuclear bomb was called the Ulam–Teller’s design. Overpressure for hydrogen-3 and deuterium in this design was achieved by focusing reflected radiation after a preliminary blast of a small nuclear explosive inside the super-bomb, rather than by a detonation wave caused by ordinary chemical explosives. On November 1, 1952, the USA tested Ivy Mike (Operation Ivy)—a thermonuclear device equivalent to ten megatons of TNT, based on the Ulam–Teller principle. The

Figure 1.33: The “fathers” of American thermonuclear bomb: Edward Teller, Stanislav Ulam and Richard Garwin.

1.5 Thermonuclear bombs | 33

tests were conducted on Elugelab Island in the Enewetak Atoll (Marshall Islands in the Pacific Ocean). However, “Mike” was not a deliverable bomb. It was a huge unit (Figure 1.34), developed by Richard Garwin. It was as big as a two-storey house and weighed 74 tons. In addition to that, it used liquefied deuterium in a cryogenic state, so it did not have any chance to become a real weapon.

Figure 1.34: The world’s first high-capacity thermonuclear explosive Ivy Mike based on Ulam–Teller’s principle.

After this test, the USSR unified all its efforts to the thermonuclear bomb. Neither Stalin’s death nor Beria’s arrest halted this work. Eventually, on August 12, 1953, the first Soviet thermonuclear bomb was tested in Semipalatinsk. Its capacity was “just” 400 kilotons, much more than that of nuclear bombs, but it still fell short when compared to the American “Mike”. Of course, the Soviet RDS-6s was a full-fledged bomb (Figure 1.35), rather than a fixed installation. Nonetheless, it did not have enough capacity (did not reach even 1 Mt) to comply with the established requirements.

Figure 1.35: The first Soviet thermonuclear bomb RDS-6s (“sloika”).

34 | 1 Electromagnetic pulse—a parcel from the past A new era in Soviet thermonuclear bomb’s history started at the beginning of 1954 due to a situation of a double dead-end: recognition of the hopelessness of American “Tube” on the one hand, and impossibility to increase the capacity of Sakharov’s “sloika” on the other. However, in order to establish a new focus area, it was necessary to prove the hopelessness of the previous two and end both of them, even though a vast amount of time and financial resources had been wasted. It was not so easily achievable even after Stalin’s death and Beria’s arrest and required the collective efforts of many scientists, including the authority of Academy Fellow Lev Landau. The new focus area, justified by D. Sakharov and Ya. Zeldovich, was about exploiting the idea of “nuclear crimping”, i. e., use of an auxiliary nuclear bomb to crimp the “sloika”, rather than using the ordinary chemical explosive, as was the case with RDS-6s. In fact, this idea was not new and was touched upon in Ulam–Teller’s theory. Yet, it still needed to be deployed, and it actually was a new design of CB-11. On November 22, 1955, a Tu-16 bomber dropped a bomb with a designed capacity of 3.6 megatons over the Semipalatinsk range. There were some deaths caused by this test; the radius of destruction reached 350 km; Semipalatinsk itself was also damaged. It was the beginning of the nuclear arms race. The Resolution of Cabinet Council of USSR approved the plan of batch construction of nuclear bombs in the country (Figure 1.36). Simultaneously, the second nuclear weapon center—Scientific and Research Institute 1011—was established in order to speed-up the engineering and building of nuclear weapons. The new institute immediately received part of CB-11’s work (Figure 1.37), and the work schedule was established.

Figure 1.36: Resolution of Cabinet Council of USSR No. 142-84ts “Regarding plan of building of nuclear and thermonuclear bombs”. In this top-secret document of January 22, 1955, a plan for the production in the USSR of atomic and thermonuclear bombs and missile charges in 1955 is given.

1.5 Thermonuclear bombs | 35

Figure 1.37: Extract from the Resolution of Cabinet Council of USSR No. 586-362ts “Regarding measures to set up operation and speed up commissioning of R&D Instituted-1011”. In this topsecret decree of the Council of Ministers of the USSR dated March 24, 1955, refers to the creation of the new research institute No. 1011 to accelerate the production of nuclear and thermonuclear weapons.

The new institute was located in a small town—Snezhinsk—situated in the eastern foot-hills of the middle of the Urals, somewhere between Sverdlovsk and Chelyabinsk on the Sinara lake shore. Today, this institute is known as the All-Russia R&D Institute of Technical Physics (ARRDITP) or Chelyabinsk-70. The area where it is located is called “Closed territorial and administrative division Snezhinsk”; there are 60,000 people living in this area. Some top-notch specialists from CB-11 were moved to the new institute. For instance, Fishman D. A., the first Deputy of Chief Design Engineer of CB-11, became the first Deputy of the Chief Design Engineer of R&D Institute-1011 (Figure 1.38).

Figure 1.38: Fishman David Abramovich, the Hero of Socialist Labor, Doctor of Technical Science, Professor, Prize Winner of Lenin’s Award and two USSR State Awards, Honored Scientist and Engineer of RSFSR, the first Deputy of Chief Design Engineer of CB-11 and R&D Institute-1011.

36 | 1 Electromagnetic pulse—a parcel from the past

1.6 Nuclear test explosions The first high-altitude explosion and the first measurement of their damage-effect parameters, including those of HEMP, were conducted in the USSR during testing of nuclear surface-to-air missiles for air defense, developed in compliance with the Resolution of the Cabinet Council of USSR No. 33389-1426ts/pi dated August 9, 1950 (Figure 1.39). The first air-defense system based on surface-to-air missile (SAM) equipped with a nuclear warhead was the S-25 “Berkut” (“SA-1 Guild” in NATO classification). One

Figure 1.39: Resolution of Cabinet Council of USSR No. 33389-1426ts/pi dated August 9, 1950, signed by J. Stalin and with a signing statement from L. Beria. This top-secret decree of the Council of Ministers of the USSR with J. Stalin’s visa constitutes the beginning of the development and creation of anti-aircraft missiles with atomic warheads.

1.6 Nuclear test explosions | 37

Figure 1.40: SAM ZUR-215 type of S-25 air-defense system (“Berkut”).

of the modifications of these air-defense systems used the missile type 215 (ZUR-215), Figure 1.40, with a nuclear warhead fitted with a RDS-9 explosive. This nuclear explosive was also used in the battlefield missile system, Frog-3 (2K6) developed by the R&D Institute-1 (Moscow Institute of Thermal Engineering). A missile launcher on the amphibian tank PT-76 was engineered by Artillery Central R&D Institute-58. A special warhead 3N14 with a 901A4 (RDS-9) explosive was developed by CB-11 for the Frog-5 missile. Due to restrictions imposed by a nuclear explosive, the warhead featured increased maximum diameter and was shaped differently as compared to the Frog-3 missile with an external-blast warhead (Figure 1.41).

Figure 1.41: Missile launcher of battlefield missile system Frog-3 featuring external-blast warheads (left) and Frog-5 featuring nuclear warheads (right).

An explosive with a capacity of 10 kt was fitted onto a body (maximum diameter of 540 mm) with a cone fairing and a truncated cone-shaped tail part. The weight of 3N14 was 503 kg. Due to the large over-caliber of the warhead, Frog-5 was 10.6-m long, and its total weight was 2.3 tons. This technical solution—an increased diameter of the warhead (i. e., use of a socalled over-caliber warhead) to fit a nuclear explosive—was copied from a similar-

38 | 1 Electromagnetic pulse—a parcel from the past

Figure 1.42: Missile launcher with Honest John (MGR-1) with an over-caliber warhead for a nuclearwarhead option.

class American missile, Honest John (MGR-1) equipped with a nuclear warhead W7 (20 kt), see Figure 1.42. Obviously, “borrowing” military secrets in the US, both in the field of nuclear ammunition and in the field of missile engineering, was so much more efficient and beneficial that it still occurs in the modern era. For example, it is worth mentioning that the newest American supersonic missile, engineered by Defense Advanced Research Project Agency—DAPRA (the US agency for developing prospective devices). This missile is now being tested both in the USA and in Russia (Figure 1.43) by the Central R&D

Figure 1.43: The newest American supersonic missile at the R&D Air-Defense Center of Central R&D Institute of Aerospace Defense Forces in Tver.

1.6 Nuclear test explosions | 39

Institute of Aerospace Defense Forces (CRDIADF), more specifically by its Air-Defense R&D Center in the town of Tver. It is known that modern radar facilities are equipped with detection systems based on both target’s Doppler features and its so-called “radar portrait”. So, this R&D center is busy “drawing the portrait” of the newest American missile, which has not been commissioned yet, similar to many other examples of American missiles and military planes. As for the previously mentioned Frog-3, the author had a chance to study this system and subsequent 9K52 (“Frog-7”) during military training when the author was a student of Kharkov Technical Institutes. Later, the author was exposed to them during annual military field training at a missile battalion of the 25th division of the land forces named after Chapayev. The shape of Frog-3 was rather strange (Figure 1.44): it featured a two-chamber solid-fuel engine 3ZH6 (like two missiles connected in series). This was a single-stage engine, but with two chambers placed one after the other inside a common body. The head chamber had a set of tube jets placed in the middle of the body at a specific angle to twist the missile around the axis, when in flight, while the tail chamber was equipped with ordinary jets.

Figure 1.44: Unusual shape of the Frog-3.

In other words, the missile seemingly had two propulsion engines working simultaneously to increase its power. This was especially important for a heavier missile with a nuclear warhead. On the other hand, the availability of two separate solid-fuel propellant sticks required additional pre-launch preparation of the missile in order to reduce scatter of the engine’s parameters during its operation, i. e., reduction of the probable circular error—CEB (measurement of target accuracy). In order to achieve this, it was necessary to raise the guided launching system with a missile into a vertical position and then return to a required position based on calculations. Tube jets in the middle of the missile’s body made it look like a fire ball when it was flying. Though the two propellant sticks were working simultaneously, their capacity was still low at the launch, thus, once leaving the guide ways, the missile “sank” to about half a meter above the ground and then started its flight from this altitude. No wonder the CEB of this missile was up to 2 km with a flight range of up to 32 km. Resolution of the Cabinet Council of USSR No. 342-135ts/pi, dated January, 1952 (Figure 1.45), stipulated the beginning of construction of plant No. 933. Later, it bore other names, such as: enterprise P. O. Box 17, then enterprise P. O. Box G-4146. This plant (contemporary name “Instrument Making Plant”) located in the town of Tryokhgorniy, in the Chelyabinsk region (code names: “Zlatoust-20”, then “Zlatoust36”) started building nuclear explosives, engineered by CB-11, including the RDS-9.

40 | 1 Electromagnetic pulse—a parcel from the past

Figure 1.45: Top-secret Resolution of Cabinet Council of USSR No. 342-135ts/pi dated January 24, 1952, signed by J. Stalin, refers to the creation of special plants No. 933 for the production of atomic weapons and auxiliary equipment.

The first missile launch with a real nuclear explosive was made from the Kapustin Yar range (official name: The 4th State Central Cross-Branch Range of Russian Federation— the 4th CCBR; military unit 15644), located in the northwest part of the Astrakhan region, close to a small town called Znamensk (closed territorial and administrative division “CTAD Znamensk”), Figure 1.46. The range was established on May 13, 1946, as a central range for the Ministry of Armed Forces of USSR (former name of the Soviet Ministry of Defense) to test the first Soviet ballistic and SA missiles. On February 2, 1956, the missile type R-5M (8K51) (SS-3 Shyster in NATO classification) with a nuclear warhead was launched from this range. This missile was engineered at the Experimental Design Bureau (EDB) of R&D Institute-88 (named today as the Central Research and Development Institute of Machine Building, in Korolyov in the Moscow region) chaired by S. P. Korolyov. Travelling 1,200 kilometers eastwards, the missile delivered

1.6 Nuclear test explosions | 41

Figure 1.46: Photo of the missile type R-5M (SS-3 Shyster in NATO classification) with a crossed-out “Top Secret” label during pre-launch at the Kapustin Yar range.

the warhead to a desert near the town of Arkalyk (Kazakhstan), where it exploded. The capacity of nuclear explosive was about 3 kilotons. The first high-altitude nuclear explosion, code named “ZUR-215”, was performed in the USSR on January 19, 1957, (i. e., before a similar US test within the “Yucca” project in April of 1958) with the detonation of a RDS-9 nuclear explosive with a capacity of 10 kt, mounted on SAM 215 (Figure 1.40), It was launched from the Kapustin Yar range. The purpose of this mission was to study the damage effects of high-altitude nuclear explosions on airplanes, flying in a tight formation (two radio-controlled targets—IL-28 bombers—as well as ground facilities that were located at distances of 500 m and 1,000 m from the point of explosion). The plane’s radar beacon (transponder) was used as a target for missile aiming (the transponder’s signal was caught by the SAM aiming system’s radar). The transponder was previously dropped from the plane and gradually descended on a parachute. Upon reaching the transponder, the warhead was detonated at the altitude of about 10 km. Instruments equipped with telemetry devices for data transfer, placed in 16 specially cylinder-shaped containers, were used to acquire data of the nuclear explosion’s damage effects. These containers had been previously dropped from planes in such a way as to have some of them at approximately the explosion’s altitude (at various distances away from it) and some of them at other altitudes. The actual position of containers at the moment of nuclear explosion and position of the explosion point itself were determined by photo surveys from ground facilities (from four directions). Additional ground facilities were established to measure the parameters of the impact wave, light emission, (spectrum, integrated flux and time), as well as direct nuclear radiation. Dummy models of wooden buildings were built close to the zero point and in other locations. Both target planes were downed by the explosion: the first plane, flying away from the burst point, caught on fire, while the other plane, which was flying almost directly towards the impact wave, lost a wing.

42 | 1 Electromagnetic pulse—a parcel from the past Instruments mounted on these planes did a good job, and the data were transferred to the ground via telemetry. Later, these results were used to determine the criteria and damage area for planes in the case of a nuclear explosion. Ground observation stations did not register any noticeable impact of the explosion on the wooden buildings and their windows. Further nuclear tests, conducted on November 1 and November 3, 1958 (supposed to be carried out at the altitude of 20–25 km) were a failure due to defects in the safety and arming units (SAU) of the warheads. SAU is a system for arming, safeguarding (in case of emergencies) and self-destruction. The SAU of nuclear warheads features multi-level arming. For example, in the case of Frog-7A with ground-to-ground missiles (9M21B) equipped with 9N32 warhead (which the author studied), the first safeguarding stage is deactivated upon leaving the guide way and detachment of the control cable; the second upon reaching design conditions of the on-board power source (ampule battery); the third upon reaching a certain velocity of the missile, (determined by the airstream pressure sensor) and the fourth upon a signal from the Doppler-radar altimeter. Other types of missiles may employ different control parameters for the SAU’s safeguarding stages. For example, the SAM-215 type employs an air-pressure transducer, which determines the altitude of the missile. The failures of tests carried out on November 1 and November 3, 1958, were a result of fault actuation of these SAU sensors, which led to warhead explosion at 12 km instead of planned 20–25 km. The next successful high-altitude nuclear explosion was performed three years later (mission “Thunderstorm”). The long break is attributed to numerous negotiations between the USA and the USSR regarding conclusion of a test-ban treaty, as well a temporary unilateral obligation, which either the USA or the USSR accepted from time to time. After a collapse of the talks, a new high-altitude nuclear explosion was performed above the Kapustin Yar range on September 6, 1961. The target for the guiding system’s radar of the SAM-215 type, which carried a nuclear explosive, was represented by a passive corner reflector, which was delivered to a 20-km altitude (the design point of the explosion) by an air-balloon. This air-balloon also carried instrument containers to measure gamma rays and other parameters of the nuclear explosion. In order to conduct additional measurements of γ- and β-emissions, two SAM-207 type missiles (built at the Tushinskiy Machine-Building Plant) were launched from an S-25 system (SA-1 Guild) into the atomic cloud, but they carried measuring instruments instead of explosives in their nosecones. The first tests of EMP-affected radar systems’ efficiency and that of electronic equipment placed into containers were carried out during this “Thunderstorm” mission. The tests were performed by tracking the “measuring” of 207-type missiles under EMP impacts occurring from nuclear explosion. Another successful high-altitude nuclear explosion was performed above the Kapustin Yar range on October 6, 1961 (“Thunder” mission). This mission was carried out to study the damage effects of a high-altitude nuclear explosion for the benefit of Anti-Ballistic Missile Defense (ABM Defense). A middle range R-5M (SS-3 Shyster), see

1.6 Nuclear test explosions | 43

Figure 1.43, raised a nuclear explosive with a capacity of 40 kt to about 40 km, where it eventually was detonated. Four heat-resistant steel spherical containers (50 cm in diameter) with data acquisition equipment were used to record the damage effects. The containers were attached directly to the missile beneath special fairings. Upon actuation of separation charges (after receiving a radio signal at the set time), the containers were discharged from the missile, stretching steel ropes with distance sensors mounted 20-m apart. These sensors measured the distance of the containers from the burst point (it was 140–150 meters at the moment of the explosion). After the explosion, the containers fell to the ground. Later, they were located using special devices. In addition, similarly to the already mentioned case, two 207-type missiles with dataacquisition equipment were launched into the atomic cloud. Similarly to the previous case, both main damage effects of nuclear explosion and EMP impact on electronic equipment placed into containers were studied. Unlike in the USA, all technical reports on the previously mentioned tests are still classified in Russia. Thus, it is not possible to obtain any new findings regarding HEMP even today. Trying to make up for the three years that had been lost, the USSR carried out a series of additional high-altitude nuclear explosions (in addition to “Thunderstorm” and “Thunder”) during 1961–1962 (mission “K”—Kosmos). Similarly to previous tests, missiles were launched from the Kapustin Yar range. However, the blast took place above the Sary-Shagan range in Kazakhstan (the State R&D and Testing Range No. 10 of the Ministry of Defense of USSR—SRDTR-10; administrative center—Closed territorial and administrative division CTAD Priozyorsk) rather than above Kapustin Yar. This means that the missiles with nuclear warheads were flying from the Astrakhan region (Russia) to Kazakhstan, literally above the heads of Soviet civilians. On October 27, 1961, two middle-range missiles type R-12 (SS-4 Sandal), see Figure 1.47, with low-capacity nuclear explosives (1.2 kilotons) were launched from Kapustin Yar. These missiles were engineered at EDB-586 (CB “Yuzhnoye”, Dne-

Figure 1.47: Placement of an R-12 missile (SS-4 Sandal) on a launching cradle.

44 | 1 Electromagnetic pulse—a parcel from the past propetrovsk) under the direction of Chief Design Engineer M. K. Yangel. These explosives were detonated 150 km above the ground (mission “K-1”) and 300 km above the ground (mission “K-2”), i. e., almost in space. The purpose of these tests was to determine the effect of the impact of space nuclear bursts (including HEMP) on radiocommunication equipment and radars, as well as to study the physical processes accompanying space bursts and check the possibility to detect them. On October 22, October 28 and November 1, 1962, three more high-altitude explosions were carried out: “K-3”—at 290 km; “K-4”—at 150 km and “K-5”—at 59 km. These bursts employed thermonuclear high-capacity explosives (300 kilotons). On October 22, the plan envisaged launching the R-9A (8K75) missile (SS-8 Sasin in NATO classification) in addition to SS-4 Sandal and “measuring” 207-type SAM, launched from Sary-Shagan towards it. The missile was engineered by EDB-1 of R&D Institute-88 under the direction of S. P. Korolyov, whereas the control system was developed and produced by the Kharkov Scientific and Production Center “Electropribor” (EDB-692; P. O. Box 67; SPC “Khartron”). The batch production of these missiles was set up at Kuybyshev (Samara town) plant “Progress” (now “Missile and Space Center “Progress”), see Figure 1.48. The SS-8 Sasin was expected to start from the Tyura-Tam range (other names: R&D and Testing Range of R&DIP-5; military unit 11284; “Taiga” facility; Baikonur) within the framework of the second stage of its flight development tests and to pass as close to the burst point as possible (Figure 1.49). The purpose of the test was to look into the reliability of radio-control equipment, evaluate the accuracy of motion measurements and determine the impact of a nuclear explosion on the level of incoming signals (both on onboard and ground receivers of missile radio-control system). In other words, they researched HEMP impact on a missile’s control system. Thus, the test of October 22 served several purposes. First, it was another check of the reliability of the nuclear-weapon carrier. Second, it was a check of the explosive itself. Third, there was a need to explore the damage effects of a nuclear explosion and its effect on various types of military machinery, including missiles and military satellites. Fourthly, there was a need to check the fundamentals of the ABMdefense system “Taran”, developed by V. N. Chelomey. The system was expected to knock down enemy missiles by a series of nuclear explosions along their way. The development of the ABM-defense system “Taran” was stipulated by the Resolution of the Central Committee of CPSU and the Cabinet Council of USSR dated May 3, 1963. EDB-52 (chaired by V. N. Chelomey) and the Radio Engineering Institute of Academy of Science of USSR (chaired by Academy Fellow A. A. Mints) were appointed as the principal engineering organizations. The system was supposed to employ long-range UR-100 (8K84) missiles (SS-11 mod. 1 SEGO in NATO classification), equipped with thermonuclear warheads with a capacity equal or above 10 Mt. Later, this project was severely criticized and thus was never deployed. Calculations offered by Director of Applied Mathematics Institute and Academy Fellow M. V. Keldysh were among the reasons for the failure. The calculations suggested that, considering the stated specifications of

1.6 Nuclear test explosions | 45

Figure 1.48: Batch assembly of strategic R-9A (8K75) missiles (SS-8 Sasin) at Kuybyshev plant “Progress”.

the “Taran” system, it would excessively anti-ballistic missiles. For example, in order to intercept 100 “Minuteman” missiles (American intercontinental ballistic missiles), the system would need 200 missiles of the SS-11 mod. 1 SEGO type. The prospect of detonating hundreds of thermonuclear bombs with a capacity of 10 Mt each over your own country did not seem attractive either for scientists or for the political elite of the country. However, back in 1962, i. e., before even making a decision regarding the development of the “Taran” system, there was an attempt to confirm the principal idea of V. N. Chelomey, i. e., whether or not it is possible to intercept long-range missiles by means of high-capacity high-altitude nuclear explosions. That is why the SS-8 Sasin missile was launched to the nuclear-weapon burst point. But on October 22, the launch was a complete failure. The combustion chamber of the primary stage broke down 2.4 seconds after ignition and the missile fell 20-m from the launching cradle.

46 | 1 Electromagnetic pulse—a parcel from the past

Figure 1.49: Map of Kazakhstan with the trajectories of SS-4 Sandal (from Kapustin Yar), 207 missiles (from Sary-Shagan) and SS-8 Sasin (from Baikonur) launched within the “K-3” project.

Since the attempt to test interception of strategic missiles failed, another test was carried out a week later, i. e., on October 28 (K-4). This test was almost an exact copy of the previous one both in terms of its set up (except for a lower planned altitude of the nuclear explosion) and in terms of its results. The missile rose to several dozen meters, and the combustion chamber of the primary stage broke down. The missile sank and collapsed on the launching cradle, resulting in severe damage. Nonetheless, other parts of the research ended successfully, since interception of a missile by means of high-altitude nuclear explosion was not the only aim of the tests. In addition to studying the damage effects of a nuclear explosion, the aim of the tests was to obtain experimental data regarding geophysical phenomena accompanying highaltitude explosions, particularly, the explosions’ impact on the ionosphere. Parameters of artificial van Allen belts occurring in space were also measured. In order to do this, “Kosmos-3”, “Kosmos-5” and “Kosmos-7” satellites were launched before the tests. Another objective was to determine HEMP impact on radars, radios and cable communication systems. Some investigations were made as part of the objective to establish a system of detection and control of nuclear explosions. Addressing all those issues during “K” missions required significant ground and satellite observations and measurements to be carried out. For this purpose, many different radio-technical aids were used. Up to 20 radar stations featuring various wavelength ranges observed the explosion area from various (up to ten) directions. Radio signals of satellites and missiles, scattered through ionized areas, were recorded by ground receiving facilities (some of them were duplicated). All permanent observation stations located in various regions of the country performed ionospheric observations. Observations of the impact of atmosphere ionization on the performance of communications equipment at various wavelengths were conducted on specially developed radio waves, capable of passing through the ground zero of the nuclear explosion, as well as on multiple

1.6 Nuclear test explosions | 47

permanent radio lines of various lengths. Several permanent radar telescopes from various observatories were also involved into observation of space radio emission. Additional explosion—“K-5”—was performed on November 1, 1962. The purpose of this explosion was to confirm some previously obtained data, particularly, those of optical observations and measurements. Actually, the development of “K-5” and the parameters of its impact were close to what had been predicted (based on previous observations performed during “Thunder” and “Thunderstorm” missions). However, improved and more sophisticated investigation during “K-5” made it possible to expand the pool of experimental data regarding explosion development at such altitudes. A. N. Shchukin, two-star general, Academy Fellow, and Deputy Chairman of the Military-Industrial Commission (MIC) of the Cabinet Council of USSR, scientifically supervised the wide array of full-scale and model experiments, as well as theoretical research related to these missions. The research findings suggested that high-altitude nuclear explosions were accompanied by electromagnetic-pulse (EMP) emissions over a broad range of frequencies with amplitude significantly higher than ground explosions of the same capacity. It was also found that the registration of nuclear explosion’s EMP was possible at higher (up to 10,000 kilometers) altitudes above the explosion epicenter. The performance of geomagnetic measurements confirmed that an observer located almost anywhere on the globe could identify powerful nuclear explosions occurring at 100– 150 km altitudes. However, it should be noted that precise and consistent interpretation of much data was found to be more complicated than initially assumed, and it required the development of comprehensive models and calculation methodology for high-altitude nuclear explosions. Unfortunately, technical reports with the results of tests and measurements of all these explosions are still classified in Russia. Some indirect data can be obtained from partially declassified multi-page CIA reports related to these tests (Figure 1.50). The label “Top Secret” is crossed out, but the sources of information have been deleted and nuclear explosives are marked with “JOE” plus corresponding figures. Missile types are denoted according to American classification (e. g., SS-1, SS-2, etc.). The only source of technical information regarding HEMP (studied under the “K” project as one of the damage effects of nuclear explosion) is represented by a report of the Principal (at that time) of the Central Physical and Technical Institution of the Ministry of Defense of Russian Federation (military unit 51105; CR&DI-12; Sergiyev Posad-7; facility “Ferma”; the Federal State-Owned Enterprise “The 12th Central Research and Development Institute” of the Ministry of Defense of Russian Federation), Brigadier General Loborev V. M, which was formed subject to the direction of the USSR authorities during “Perestroyka” and “reset” 32-years later after completion of project “K” [5], as well as his article co-authored with other employees of the same Institute, published 36 years later after completion of project “K” [6].

48 | 1 Electromagnetic pulse—a parcel from the past

Figure 1.50: A page from CIA’s report regarding high-altitude nuclear explosions performed in the USSR under the “K” project.

Both the report [5] and the article of V. M. Loborev described the consequences of the “K-5” nuclear explosion’s EMP impact on the civil infrastructure of Kazakhstan, located within the area of exposure. Some interesting factors are worth to be noted: CR&DI-12 was founded in 1950 subject to the Directive of the General Staff of the Armed Forces of USSR No. 553343 dated April 26, 1950. Later, the Institute became part of the 12th Chief Directorate of the Ministry of Defense of USSR (military unit 31600, Moscow, Znamensky Lane, 19) to research the damage effects of nuclear explosion and their impact on ammunition, military machinery, buildings and people; elaborate recommendations for the armed forces to protect the staff and machinery from nuclear weapons’ effects and to provide scientific and methodological supervision of special weapon testing. One of the institute’s objectives was to provide scientific substantiation of the country’s protection against nuclear attack. Looking at the institute’s building in Sergiyev Posad (Moscow region), the building of its branch office

1.6 Nuclear test explosions | 49

in St. Petersburg (military unit 70170, Novoselskaya Street, 39E), as well as its experimental facilities featuring dozens of test benches, one can say that it’s rather large. The following test benches were involved with HEMP: – pulse voltage generator PVG-10; – simulators of electromagnetic pulses SEMP-B and SEMP-BM; – “Arterit” unit intended for testing large-scale military machinery for the impact of powerful electromagnetic fields; – “Zenith” unit to test military machinery and ammunition in regard to the EMP impact; Despite such facilities and resources, there is no word about the institute’s staff involvement and any significant achievements, either on the institute’s web-site or in the scientific literature published in the USSR on the topic of nuclear testing, or in declassified documents on the nuclear project published in USSR, or in the memoirs of participants of those events. This cannot be explained by mere confidentiality. First of all, this topic was declassified long ago and secondly, employees of other scientific institutions (e. g., All-Russian R&D Institute of Experimental Physics from Saratov town) have long been publishing the articles on HEMP issues in the public domain [7, 8]. Publications of the Chief Research Fellow of CR&DI-12, Doctor of Engineering Science, Professor, Balyuk Nikolay Vasiliyevich are, perhaps, the only exception to this rule. However, most of these publications regarding HEMP are not based on original research and development efforts of the institute. They are rather paraphrase information previously published in dated American reports. One of the typical examples is the book [9] published in 2013. The major portion of the text and illustrations are just a translation and paraphrase of some American 30-year old reports. Other examples include his articles in the EMC Technologies Journal co-authored with the Principle of CR&DI-12, Rear Admiral Pertsev S. F. and the first Deputy of the Chief of the General Staff of the Military Forces of Russian Federation, two-star general Burutin A. G. [10, 11]. Those are just general arguments and descriptions of equipment available at CR&DI-12 that are well known to the specialists in this field. Who will benefit from the general arguments of generals and admirals, instead of specific recommendations and inventions on how to protect critical infrastructure of the country? The only reference is made to the participation of the office staff in the Leningrad branch in measuring parameters of one of the underwater nuclear test explosions, as well as the just mentioned report and article of the Institute’s Principal Loborev V. M. Quite a short list for 68 years, isn’t it? The number of classified thesis papers accessible to a very limited circle of employees would hardly justify the existence of the institute. Official literature with research findings regarding HEMP’s impact on civil infrastructure was published by the institute staff in English. Why? Soviet (and now Russian) civil specialists still lack information regarding HEMP’s impact on civil infrastructure. At the same time, one of the objectives of the institute upon its foundation was to develop means of protection from nuclear weapons, meaning HEMP, for the country.

50 | 1 Electromagnetic pulse—a parcel from the past

Figure 1.51a: Panorama of the main building of CR&DI-12 in Sergiyev Posad (Moscow region).

Figure 1.51b: Branch office of CR&DI-12 in St. Petersburg.

Figure 1.51c: Experimental and testing facility of CR&DI-12 (Moscow region).

Don’t they have enough information to provide to civil specialists in the field of the electric power industry, communication and water supply, as well as other specialists responsible for the country’s infrastructure? Reviewers from CR&DI-12, who review authors’ articles from time to time, state that it is prohibited to publish articles addressing the necessity to intensify work in

1.6 Nuclear test explosions | 51

the field of protection of civil infrastructure from HEMP and to offer specific technical solutions intended to ensure this type of protection. Why? It seems that the institute staff is afraid that civil specialists responsible for the country’s infrastructure will find out about this serious problem and eventually ask why nothing has been done in this field after all these years! Eventually, somebody will have to answer this question. The Institute responsible for this matter for 68 years now will continue its cotton-wool existence, while a minimum number of people are aware of the problem. But, let us get back to the outcomes of HEMP’s impact on civil infrastructure during “K-5” tests. According to the data published by V. M. Loborev [12] (Figure 1.52), HEMP impact caused failures in the operation of air-defense radar located about 1,000 km away. Breakdowns of ceramic insulators resulting in short-circuits were observed on 35-kV electric overhead power lines. Electromagnetic pulses caused fires due to short circuits in electric appliances. A power generator was knocked out of service at one power plant; and relay protection was triggered resulting in switching the power generator off at another power plant. A small geomagnetic component of HEMP induced a short-current pulse with an amplitude of several thousand amps, as well as a long (more than 20-sec) current pulse, rated 4 amps. This led to diesel-generator damage and triggering of protection devices mounted over a 570-km above-ground telephone line. There is also information about some breakages of electronic equipment that occurred at Baikonur Cosmodrome.

Figure 1.52: Damaged electric equipment affected by HEMP during nuclear high-altitude test explosion performed under “K-5” project in Kazakhstan in 1962 (based on data published in V. M. Loborev’s report presented at EUROEM International conference in France in 1994).

The US Nuclear Energy Agency took advantage of the new policy of Soviet President M. S. Gorbachev and proclaimed the principles of “perestroika”, “reset” and “glasnost” (when the doors of previously well-protected Soviet enterprises were opened

52 | 1 Electromagnetic pulse—a parcel from the past

Figure 1.53: First page (partially) of the Resolution of Cabinet Council of USSR No. 1559-699ts regarding establishment of nuclear range on Novaya Zemlya islands.

both literally and figuratively and when the financing of many defense projects was ended). The agency provided grants (in 1992, when the new President B. N. Yeltsin was in power) of $288,500 to prepare a technical report with analysis of the effects occurring during underwater nuclear explosions, performed by the USSR in the Arctic and during high-altitude nuclear explosions under the “K” project in Kazakhstan. This job was given to CR&DI-12, as its staff at the Leningrad branch office were recording and measuring the parameters of the Arctic nuclear explosions on the Novaya Zemlya islands (a nuclear range established subject to Resolution of the Cabinet Council of USSR No. 1559-699ts dated July 1954, see Figure 1.52; other names: Facility-700; Range No. 6; Arkhangelsk-55 and Arkhangelsk-56). Whereas the staff at the headquarters (Sergiyev Posad) was responsible for high-altitude nuclear explosions performed under the “K” project, in fact, the report presented by V. M. Loborev and his article were a result of this job performed with funding from American grants. In 1995, upon finishing work and preparation of the report (17 paragraphs long), two executives of CR&DI-12—the principal, two-star general, Professor, V. M. Loborev and the chief research worker, colonel, Doctor of Technical Science, V. M. Kondratyev— were invited to the leading nuclear centers: Lawrence Livermore National Laboratory and Los Alamos National Laboratory to lecture on the topic of electromagnetic effects of high-altitude nuclear explosions. Since neither V. Loborev nor V. Kondratyev spoke any English, they were constantly accompanied by two interpreters. V. Kondratyev lectured about HEMP history and its recording because he was a senior specialist in this field and his doctoral thesis had been devoted to HEMP. He suggested that Soviet scientists registered HEMP (though many sensors were knocked out of service) during the very first ground testing back in 1949. This phenomenon was assumed to be

1.7 The status of HEMP protection

| 53

used as the means of registration of remote nuclear explosions rather than as a kind of weapon capable of ruining electronic equipment. Then, V. Kondratyev reported that, since the USSR did not find any American data about HEMP, scientists thought that topic was top secret in the US. Later, it became obvious that there were no such data in the US. “Then—Kondratyev said—when we received your data, particularly the EMP formula, we noticed that they were in full compliance with ours, and thus we concluded that you took our data from our classified reports. We are familiar with calculation models of Radasky, Baum and Longmeyer, but we couldn’t check them as we didn’t have high-capacity computers like those that you’ve got in the USA. So, our methods were different.” According to V. Kondratyev, the HEMP theory was elaborated in the USSR back in 1961–1962. This was the merit of scientists from the Ministry of Communication of the USSR, who analyzed the results of HEMP impact on communications systems. Later, V. Kondratyev listed damages to electric equipment and communication systems that occurred as a result of high-altitude nuclear explosion in Kazakhstan in 1962 (see Figure 1.49). The question was how they protected communication lines after testing. He answered that, subject to the initiative of the Ministry of Communication, aboveground wires were replaced by underground cables. While answering further questions, V. Kondratyev mentioned breakages of military diesel generators and substations as a result of pulse overvoltage, as well as the actuation of relay protection on power lines. Answering the question about breakages of communication lines during other tests, V. Kondratyev mentioned that those breakages occurred during low-altitude explosions. As for breakages of power lines during other tests, he said that he was not aware of such data. Answering the question, i. e., whether the direction of wires (north–south; west–east) influences their susceptibility to breakages, V. Kondratyev said that it definitely does. Subsequently, there was a series of questions regarding more general issues of nuclear testing and nuclear safety. Those were addressed by V. Loborev. Most of the answers were like this: “I am only the Principal of the Institute and not the President of the country”, “Ask my government, not me”… One of the comments from the American side that was given during one of the discussions is noteworthy: “We are not interested in protection of our national power system, since we think that any hurricane can create more serious problems in power industry”. This phrase was articulated back in 1995. It is known that today this point of view of the US is no longer shared (except by Dr. Saul Rabinowitz of the Electric Power Research Institute).

1.7 The status of HEMP protection So, what is the situation today? It should be acknowledged that nobody is taking any serious steps towards protection of civil infrastructure from HEMP, either in Russia or in any other post-Soviet country. The major institute in this field—CR&DI-12—that has both equipment and staff, performs rare tests of some samples of military machines

54 | 1 Electromagnetic pulse—a parcel from the past for HEMP resistance, ignoring critical types of civil infrastructure such as power and water supply systems. It looks like this situation is accepted, and nobody cares about protecting Russia’s infrastructure from HEMP. The only exception is the attitude of the Ministry of Communications which has always been interested in the HEMP impact onto civil communications systems since the onset of experiments with nuclear explosions. Several dissertations and reports on this topic emerged a bit later. There are also regulatory materials, e. g., “Provisions for resistance of common communication network’s equipment, appliances and devices to IE and HEMP in Russia”, approved by the decision of the State Commission for Telecommunication of the Ministry of Communications of the Russian Federation No. 143 dated 31.01.1996 (not available in the public domain) and others, as well as a set of Russian standards, such as: State Standard (GOST) R 52863-2007; Protection of Information; Protected automated systems; Testing for resistance to intentional powerful electromagnetic impacts; General requirements; GOST R 53111–208; and Sustainability of operation of communication networks for general use. Unfortunately, there is nothing similar in the electric power industry, which is as important as communication. The opposite situation has come about in the US. They have commercialized this issue, and now it represents a well-functioning business. Dozens of professional consultants who frighten people with HEMP consequences appeared there during the last 10–20 years. Dozens of books, hundreds of reports (a list of major reports available in the public domain is given in the appendix) have been published on this topic. Dozens of private and state-owned organizations have received orders to conduct research in this field. Here is a nonexhaustive list of them: – Metatech Corp. – Department of Homeland Security (DHS) – EMP Commission of Congress – North American Electric Reliability Corp. (NERC) – Department of Energy – Department of Defense (DoD) – Critical Infrastructure Partnership Advisory Council (CIPAC) – Electric Infrastructure Security Council (EICS) – Defense Science Board (DSB) – US Strategic Command (USSTRATCOM) – Defense Threat Reduction Agency (DTRA) – Defense Logistics Agency (DLA) – Air Force Weapons Laboratory – FBI – Sandia National Laboratories – Lawrence Livermore National Laboratory (LINL) – Oak Ridge National Laboratory – Idaho National Laboratories

1.7 The status of HEMP protection

– – – – – – – – – – – – – – – – – – –

| 55

Los Alamos National Laboratories Martin Marietta Energy Systems, Inc. National Security Telecommunications Advisory Committee Federal Emergency Management Agency (FEMA) National Academy of Science Task Force on National and Homeland Security EMPrimus SARA Inc. Neighborhood of Alternative Homes (NOAH) EMPact America Federal Energy Regulatory Commission (FERC) Electric Power Research Institute (EPRI) NASA U. S. Northern Command (NORTHCOM) SHIELD Act EMP Grid EMP Technology Holding Strategic National Risk Assessment (SNRA) Walpole Fire Department

It appeared that the EMP topic is nothing else than a wonderful long-playing tool for the government budget peculation. And it looks like nobody wants this “process” to end by concrete actual actions aimed at protection of electric power-supply systems. To support this, let me cite one of the former authorities of the US Department of Defense, Dr. Ashton Carter: “Army, Navy and Strategic Command continue to think that they need to think about the problem”. Executive Director of the Task Force on National and Homeland Security, Dr. Peter Vincent Pry, was more specific, when speaking on this topic he said: “The problem is not the technology. We know how to protect against it. It’s not the money, it doesn’t cost that much. The problem is the politics. It always seems to be the politics that gets in the way”. So, it becomes clear why nothing specific has been done anywhere in the world regarding protection of the infrastructure (and the electric power industry, specifically) from HEMP and why all the efforts are limited by multi-page reports about investigations, presentations, workshops, conferences and other types of pleasant leisure times in a circle of colleagues. The fact is that these multiple “participants of the process” are not interested in finishing the long-term investigation process, but prefer to keep the topic “afloat” in order to receive governmental financial funding. The author is personally aware of the business and knows the employees of one such company in the US. The company makes hundreds of thousands of US Dollars merely frightening the executives of power companies with “bogeyman” about HEMP. Once the contract is concluded, they give specialists in power and water-supply companies (far away from

56 | 1 Electromagnetic pulse—a parcel from the past the HEMP problem) a song and dance about EMP and general information, which one can easily procure on the Internet free of charge. Considering that there are hundreds of manufacturing companies all over the world that promote expensive (sometimes unnecessary) HEMP protection aids, which do not always possess the declared features (who can check?!), one can reach the conclusion that HEMP is an excellent business today! Many authors indirectly support this business because their books are nothing but “bogeymen”, intended to frighten laymen. They do not contain any technically significant information, and thus, they are not interesting for specialists. Nonetheless, they create an atmosphere of fear and despair in society, while the problem can easily be resolved, provided those responsible for the country’s infrastructure are willing to do something with it. Unfortunately, it is very common to hear (both in Russia and in the US) that protection of the country and its infrastructure from HEMP is the army’s domain and not that of civil specialists. On the other hand, army specialists assume that their responsibility is to ensure HEMP protection of military equipment and ammunition and not of civil infrastructure. Moreover, they insist that the only efficient protection from electromagnetic pulses from a nuclear explosion is represented by the national Air and Missile Defense Systems (ADS and MDS), where more budgeted funds need to be invested. This attitude of Military-Industrial Commission (MIC) representatives becomes clear when comparing the relatively low cost of the means of HEMP protection for the most important elements and systems of the national infrastructure with the costs for development and production of an efficient multi-level missile shield, which protects the whole country. There is also lack of understanding (or unwillingness to understand) that military aids, such as ADS and MDS, are often incapable of ensuring protection of the infrastructure from various types of modern nuclear explosive carriers. For example, modern strategic carriers are equipped with rather sophisticated aids to overcome existing and perspective ADS and MDS of an enemy. Simpler, smaller-sized and short-range missiles can be similarly “successful”. These missiles are fitted into standard sea containers on ships, near the coast line or even in ports (Figure 1.54) and are capable of carrying nuclear charges over hundreds of kilometers, while ascending to a dozen kilometer altitudes. They are the sources of EMP invulnerability to any MDS, both existing and potentially developed due to their capability of concealed approach to a target, unexpectedness of launch, ultra-short approach time and changing trajectory during cruising. The possibility of a concealed approach by tactical warhead missiles of a shortaction radius to a target in order to avoid its being hit by MDS, on the one hand, and removing it from the regulation of international treaties, on the other hand, has long been known to military specialists, and the attempts to develop these systems started immediately upon the creation of relatively small nuclear warhead missiles. For example, in 1961, the US airborne units received “Little John” (MGR-3) missiles, which were equipped with free-flight missiles capable of carrying nuclear warheads.

1.7 The status of HEMP protection

| 57

Figure 1.54: Containers resting on ships and in ports where tactical ballistic nuclear-warhead missiles can fit are invulnerable to MDS.

Light-weight launching units of this system could be delivered by CH-47 “Chinook” helicopters, both in the bay and with an external lift. The Soviet Union quickly appreciated the perspectives of these systems, and, based on the Decree of the Council of Ministers of USSR No. 135-66ts dated February 5, 1962, it started developing the tactical missile complex “FROG-7” (9K53) with 9M21B missiles (with nuclear warhead) and 9M21B1 (with thermonuclear warheads) and the launching unit 9P114 constituting a light-weight self-propelled platform with a carburetor 45-hp engine M-407 from the “Mosckvich” car. Later, several modifications of such systems were introduced, which allowed transportation by MI-6 and MI-10 cargo copters. The helicopter was expected to deliver the missile with its launching unit behind enemy lines. The rest of the way, where necessary, could be covered on wheels, and then it could suddenly strike a missile from a position which the enemy did not anticipate, which converts it from a tactical complex into a strategic. The efforts of the “FROG-7” development reached the stage of experimental testing. However, this resulted in many obstacles including high “windage” of a helicopter

58 | 1 Electromagnetic pulse—a parcel from the past carrying a launching unit and consequently a high drifting rate, as well as the inappropriate flying range of fully loaded helicopters. As a result, the efforts to develop this complex were stopped in 1965. Modern technological developments have made it possible to return to this idea and deploy it successfully. For example, let us take the Israeli missile system LORA (LOng Range Attack) manufactured as a container with four missiles (Figure 1.55). Its shape is reminiscent of containers of the Russian system Club-K, with the similar number of missiles 3M54K (SS-N-27 Sizzler in NATO classification) (Figure 1.55).

Figure 1.55: Container-based launching units of missile complexes Club-K (top) and LORA (bottom).

Club-K is a Russian container-based missile unit, which can fit into a standard 20- or 40-foot sea container. This unit is intended for targeting above-water and ground targets. The unit can be installed on the coast, various classes of vessels, railways and truck platforms. The complex can be used with ground launching units, as well as sea, railway and truck platforms. It can use various anti-ship missiles, as well as missiles for hitting ground targets. All the missiles included in the complex are cruising, flying at a relatively low altitude of 5–50 m and are not intended to be equipped with nuclear warheads, while

Bibliography | 59

LORA is equipped with a tactical ballistic missile, which can fly as high as 45 km and can carry a high-capacity nuclear weapon to a distance of up to 300 km. Today, there are hundreds of millions of standard containers circulating all over the world, see Figure 1.50. Who knows which of them are just containers and which of them carry missiles. Despite the fact that Israeli LORA is actually the only full-fledged container system that can secretly approach the coast of a country on a container ship and hit its territory with an electro-magnetic pulse, the fact of the existence of this system makes it possible conclude that the statements of MIC representatives about efficient protection of advanced MDS against HEMP and that they should continue to receive additional investments, are not true, and in fact are a way of deceiving public opinion. In practice, an army will not be able to ensure efficient protection of the power systems of cities and settlements from HEMP and thus, electric engineers should assume the leading role to take care of such protection.

Bibliography [1]

Zeldovich Ya. B. Interaction of Unbound Electrons with Electromagnetic Emission. Success Phys. Relat. Sci., 1975, Vol. 115, No. 2. [2] Kompaneets A. S. Radio Emission of Nuclear Explosion. Exp. Theor. Phys. J., Vol. 35, No. 6(12), pp. 1538–1544, 1958. [3] Gilinsky V. Kompaneets Model for Radio Emission from a Nuclear Explosion. Memorandum RM-4134, Rand Corporation, August 1964. [4] Khariton Yu. B., Smirnov Yu. N. Myths and Reality of Soviet Nuclear Project. Arzamas-16: ARR&DIEP, 1994. p. 72. [5] Seguine, Howard, “Memorandum for Record, Subject: US-Russian Meeting” at Lawrence Livermore National Laboratory, February 14–15, 1995. [6] Greetsai V. N., Kozolovsky A. H., Kuvshinnikov V. M., Loborev V. M., Parfenov Y. V., Tarasov O. A., Zdoukhov L. N. Response of Long Lines to Nuclear High-Altitude Electromagnetic Pulse (HEMP). IEEE Trans. Electromagn. Compat., 1998, Vol. 40, No. 4, pp. 348–354. [7] Bashurin V. P., Gaynullin K. G., Golubev A. I. et al. Some Theoretical Computation Models and Software to Study Electrodynamic Effects, Accompanying Nuclear Explosions. Collection of Research Papers. The Issues of Mathematic Modeling, Numerical Mathematics and Informatics. Arzamas-16: ARR&DIEP, 1994, pp. 117–130. [8] Boriskin A. V., Zolotov V. A., Kravchenko A. S., et al. Mobile Simulators of Electromagnetic Pulses Employing Magnetic Cumulation Generators. Appl. Mech. Techn. Phys., 2000, Vol. 41, No. 3, pp. 6–12. [9] Akbashev B. B., Balyuk N. V., Kechnev L. N. Protection of Telecommunication Facilities from Electromagnetic Impacts. Griphon, Moscow, 2013. p. 472. [10] Burutin, A. G., Pertsev S. F., Balyuk N. V. Experimental and Testing Facilities of the Ministry of Defense of Russian Federation. EMC Technol., 2010, Vol. 1, pp. 33–37. [11] Burutin A. G., Pertsev S. F., Balyuk N. V. Weapon and Electromagnetic Factors. “Voyenny Parad” (Military Parade) Magazine, 2009, Vol. 6, pp. 14–16. [12] Loborev V. V. Up to Date State of the NEMP Problems and Topical Research Directions. Electromagnetic Environments and Consequences: Proceedings of the EUROEM 94 International Symposium, Bordeaux, France, 30 May–3 June 1994, pp. 15–21.

2 A contemporary view of HEMP for electrical engineers 2.1 Is the contemporary view up to date? It is worth noting that a contemporary view of HEMP is based on the results of the nuclear tests and research studies carried out more than 50 years ago, and since then has not changed significantly. During these years, the civilian standards (without any references) were taken from the old classified test reports and research, which is the reason that there are no references. For example, all the basic data and curves given in the International Electrotechnical Commission (IEC) Standard IEC 61000-2-9 [1] are indicated in such old reports, while the standard contains no references to them. Also, most of the later declassified reports prepared by various organizations in 1980s and 1990s contain a many images, graphs and tables taken from those same reports. The most recent books (such as [2]) on that subject contain nothing but free interpretations and paraphrases of data taken from the subsequently declassified reports [3]. With this in mind, the referenced sources may prove to be the multiple restatements, rather than the real origins of information.

2.2 The basic physical processes The physical processes accompanying the high-altitude explosions with nuclear yields are very complicated, and the power industry experts are not necessarily obliged to know every detail. Moreover, such a detailed description of all physical processes with complex mathematical formulas may frighten the readers and cause them skip this material. So, to avoid this, here you will find the so-called simplified HEMP theory optimized for electrical engineers, rather than for nuclear physicists. According to the IEC classification (taken from the classified military standard), the HEMP contains three components: E1, E2, and E3, see Figure 2.1. E1 is the fastest and the shortest HEMP component produced by the powerful X-radiation (γ-quanta or X-ray photons) generated by the explosion with the nuclear yield. Upon the explosion, the X-radiation knocks out unbound electrons (the so-called Compton scattering electrons) from air atoms, see Figure 2.2. Then, the electrons are captured by the Earth’s magnetic field and gyrate to the Earth’s surface with a speed close to the velocity of light. The directed flow of electrons is the electrical current generating the magnetic field. The rapid flux of magnetic field (from 0 to the peak value) generates the high-power pulse of electric field described by Maxwell’s equations. Upon the nuclear explosion, the strength of the electric field near the Earth’s surface may reach 50 kV/m. https://doi.org/10.1515/9783110639285-002

2.2 The basic physical processes | 61

Figure 2.1: Parameters of HEMP components E1, E2, and E3 according to (MIL-STD-2169 and IEC 61000-2-9).

Figure 2.2: Emission of Compton free electrons generated by the aerial nuclear explosion.

This interaction between the super-fast negative electrons and the magnetic field generates the electromagnetic wave concentrated by the Earth’s magnetic field and is directed from the sky to the ground. According to IEC, the full duration of a HEMP pulse amounts to 1 ms (1,000 nanoseconds). The E1 component is conditioned by the very intensive electromagnetic field provoking the very high overvoltages within the electric chains. Near the Earth’s surface at medium latitudes, the E1 component creates pulse voltages up to 50 kV/m with the power density of about 6.6 MW per square meter. The E1 component accounts for most of the electronic equipment failures caused by the overvoltage and the breakdowns of the p-n-junction in the semiconductor elements and their internal insulation. The

62 | 2 A contemporary view of HEMP for electrical engineers regular discharge arresters, optimal for the atmospheric (lightning) overvoltage protection, may be too slow to respond to the E1 component and to protect the equipment appropriately. It should be emphasized that the Compton model [4] is based on the presumptions disputed by certain authors because they were not derived from current electrodynamics principles. However, today this model is generally accepted since it is the only available model. The Thompson classical electrodynamics assumed that the light is a wave by nature. An electron affected by such a wave should fluctuate with a frequency equal to the field frequency (i. e., the wavelength of the incident light) and radiate the secondary (scattered) waves of the same frequency. So, in the case of Thompson scattering, this process should not contain waves of varying frequencies. However, the research into X-ray scattering in paraffin made by Arthur Compton, see Figure 2.3, demonstrated that X-rays scattered in paraffin have longer wavelength than the initial scattered rays. In other words, radiation with both initial wavelength and longer waves was detected.

Figure 2.3: Arthur Holly Compton, Nobel Prize winner in physics.

Arthur Compton proposed the theoretical interpretation of this phenomenon (later it was independently proposed by Peter Debye) based on the corpuscular theory of light formulated A. Einstein in 1905. Indeed, assuming that light radiation is a flow of particles—(corpuscles) photons—the Compton effect results from the elastic collisions between X-ray corpuscles—photons and free electrons from the substance. Since the light atoms of scattering substances (the paraffin in the experiment) are weakly connected to the nucleus, they can be deemed as free. Upon the collision, the photon transfers part of its energy to the electron according to the law of conservation of energy. During this process, the partial loss of the photon energy is registered as the radiation-frequency decrease (wavelength increase) in the course of the experiment.

2.2 The basic physical processes | 63

Such a wavelength increase was named the Compton Shift. In 1927, A. Compton was awarded the Nobel Prize in Physics for this discovery, confirming the double nature (wave-particle) of light. The E2 component is an intermediate HEMP component (due to the increase in speed and wavelength) appearing as the secondary effect of the Compton electrons flowing within the Earth’s magnetic field. The E2 parameters have much in common with the electromagnetic pulses of aerial origin (i. e., generated by the lightning). The strength of the E2 field can reach 100 V/m. Since the E2 component is similar to lightning and there are well-proven lightning protection technologies available, it is deemed that protection against the E2 component is very simple. The E3 component (or a geomagnetic effect of HEMP) is very different than the two other HEMP components. It is a very slow pulse, lasting up to tens or hundreds of seconds and generated by the Earth’s magnetic field shift and its following restoration. The E3 component is similar to the geomagnetic storm provoked by a very intensive solar burst. Geomagnetically induced currents are generated by the magnetic disturbances within the Earth’s magnetosphere and their flow in the ground. The E3 component is based on the magnetohydrodynamic effects of the interaction between the plasma products of the nuclear explosion and high-temperature ionized air with the magnetic field of the Earth. This effect has two stages called “blast wave” and “heave” and is characterized by different mechanisms of generation and duration, see Figure 2.4. The first stage lasts between to 1–10 seconds and is produced by the expansion of large plasma substances generated in the thin air (at high altitude) under the influence of the Earth’s magnetic field at the time of the explosion. This phenomenon is accompanied with the complex interaction between plasma ions, magnetic field, gamma and X-radiation, leading to the generation of the eddy electric field. These physical phenomena result in the significant disturbance of the Earth’s magnetic field, increasing with the increase of the explosion power and altitude. The second stage is characterized by the heave and hike of the ionized air (actually, plasma) overheated due to the explosion. When the ionized plasma crosses the Earth’s

Figure 2.4: Two stages of a magnetohydrodynamic effect of HEMP [5]: a) “blast wave” b) “heave”.

64 | 2 A contemporary view of HEMP for electrical engineers magnetic-field lines, the air layer is polarized and generates the high-power electric field, creating the high circulating currents within the ionosphere, in turn. Those processes are relatively slow. The second stage of explosion lasts 10–300 seconds. Consequently, all these thin-air processes generate the relatively slowly varying magnetic field (from one to tens of volts per km) near the Earth’s surface, see Figure 2.5.

Figure 2.5: Variation of the horizontal component of the electric field near the Earth’s surface upon the HEMP E3 impact. Left—pulse registered during the test explosion within the Starfish Prime Project (1962 year); right—standard pulse (according to IEC 61000-2-9 [1]).

Despite the low strength of the E3-generated electric field, it induces rather high electric currents with a very low frequency (less than 1 Hz) into the long metal objects (such as pipes, rails, power transmission lines). Such quasiconstant currents are dangerous for the electric-power equipment not designed to operate under constant currents (transformers, generators). It should be noted that despite its danger regarding electronic and electrical power equipment, the energy of HEMP is rather low—less than 1 % of the energy released from the nuclear explosion. In any case, the HEMP energy is less than the energy released in the lightning strike, see Figure 2.6. Thus, from the 1980s onwards, various countries of the world worked strenuously to create the so-called nuclear Super-EMP with high-power electromagnetic radiation. Basically, the works has been performed in two directions: creating a core-shell made of a special substance additionally radiating the high-energy γ-rays (X-ray photons) under the influence of the nuclear-explosion neutrons, and focalization of such γ-radiation. According to the experts, Super-EMP can significantly intensify the E1 component and create, near the Earth’s surface, a field with a strength of hundreds and even thousands of kilovolts per meter. Moreover, military officials make no secret that the government and military control systems, as well as national infrastructures,

2.2 The basic physical processes | 65

Figure 2.6: Strength of lightning energy vs. HEMP energy.

Figure 2.7: Generation of the electric field pulse near the Earth’s surface by an aerial nuclear explosion.

including power, water supply, etc., would be the primary targets for such a weapon in case of a conflict. Figure 2.7 has been repeatedly published in numerous reports and standards. Originally, it was taken from top-secret report AD-A955391 [6] prepared 40 years ago, see Figure 2.8. Since the E1 component is generally considered as the most dangerous for electronic and electrotechnical equipment, let us take a closer look at its properties and parameters. The atmosphere has a special so-called source (or deposition) region located 20– 40 km above the ground in the stratosphere. Here, the maximum number of X-ray elec-

66 | 2 A contemporary view of HEMP for electrical engineers

Figure 2.8: One page of fragmentary declassified (more accurately, the “sanitized” version with crossed-out labels “Top Secret” at the top and at the bottom of each page) version of report AD-A955391 (“Capabilities of Nuclear Weapons”, DNA-EM-1, 1978); Chapter 7 of this report describes HEMP and its characteristics.

trons is generated under the influence of the X-radiation from above and the secondary electrons knocked out the air atoms by Compton electrons. Each Compton electron has energy of about 1 Megaelectron Volt (MeV) and generates 30,000 secondary pairs of electron-ions knocked out of the air atoms along the way. Such pairs generate the just mentioned source region [7].

2.2 The basic physical processes | 67

Figure 2.9: Expansion of Compton electron region in proportion to the nuclear yield and increasing altitude: top—for a 1 Mt yield; bottom—for a10 Mt yield [8].

This region remains virtually steady with altitude, but its radius increases significantly in proportion to the increasing yield and altitude of the explosion, see Figure 2.9 [8]. However, the atmosphere varies in density at different altitudes, thus changing the properties of the generated electric field. It turns out that, due to a high-altitude nuclear explosion in very thin air, the number of atoms is low so therefore the number of generated free electrons is also low. Moreover, the free electrons must travel to the Compton electron-source region such a great distance that most of the free electrons are able to recombine, and the electric-field pulse near the ground weakens. On the other hand, for low-altitude nuclear explosion in the thick air (i. e., under the Compton electron-source region), the number of such electrons decreases, the free electron travels to the ground, impedes, and is accompanied by the intensified recombination, thus further weakening the electric-field pulse near the ground. According to some data reflected in the declassified reports, there is a certain optimal nuclear-explosion altitude that makes possible reaching the maximum electrical-field strength near the Earth’s surface, see Figure 2.10. Nevertheless, the X-radiation propagates along a straight line and does not depend on the curvature of the Earth since, from the explosion point, the EMP impact region radius is limited to the distance from the explosion point to the horizon. Def-

68 | 2 A contemporary view of HEMP for electrical engineers

Figure 2.10: Dependence of the electrical-field strength (E1) near the Earth’s surface at the height of the nuclear explosion.

Figure 2.11: Dependence of HEMP impact region on the explosion height.

initely, the increase in the nuclear-explosion altitude increases the region of HEMP impact on the surface electric equipment, see Figure 2.11. However, the impact region is not equal to the damage region since the expansion of the HEMP impact region due to the explosion height increase is accompanied by a weakening of the electric-field strength near the Earth’s surface. Moreover, it is obvious that the strength of such an electromagnetic field depends strongly on the yield of a nuclear explosion, see Figure 2.12. Also, the diagram shown in Figure 2.12 demonstrates that the optimal explosion height is hardly a constant value. It increases in proportion to the yield of the nuclear explosion.

2.2 The basic physical processes | 69

Figure 2.12: Dependence of the HEMP electric-field strength on the height of the explosion and the yield of the explosion.

However, it so happens that there is even more to note. The experimental measurements of the electric-field strength at various distances to the explosion epicenter have shown a very strange tendency, see Figure 2.13 [6]. Contrary to expectations and common sense, it transpired that the field strength is minimal immediately at the explosion epicenter (region A) while reaching its maximum somewhere outside the epicenter (region B, Figure 2.13). Additionally, the electric-field strength varies in amplitude, inasmuch as, near the ground, the E1-pulse shape and extent varies, see Figure 2.14. Thus, a certain average curve of the E1 component was calculated, see Figure 2.14. Today, it is used as a standard curve of the E1 pulse of 2.5/23 (2.5/25) nanoseconds and an amplitude of 50 kV/m, see Figure 2.15. What is 2.5/25 nanoseconds? This is a special value characterizing the pulse shape. It is considered to reveal the relationship between the pulse rise time (front edge), calculated as the time when the pulse rises from 10 % to 90 % (2.5 nanoseconds) of amplitude value and the pulse-full width at half maximum amplitude—FWHM (23 nanoseconds or 25 nanoseconds in certain standards), see Figure 2.16. However, it take considerable time to make the proper calculations. Various authors have proposed significantly different pulse parameters, see Figure 2.17, Table 2.1. Finally, the version described in the MIL-STD-464A and IEC 61000-2-9 Standards was generally accepted, and the HEMP E1 pulse has the generally accepted shape shown in Figure 2.15. However, it is not all that simple because the pulse energy significantly depends on the pulse shape. This indicates that the lower pulse amplitude in region A, see

70 | 2 A contemporary view of HEMP for electrical engineers

Figure 2.13: Distribution of the electric-field strength (E1 of HEMP) at various distances to the epicenter (for a nuclear yield detonated at 100–500 km with the epicenter located between 300° and 60° north latitude) [6].

Figure 2.14: E1 electric-field-pulse shape and extent in various regions at the ground surface.

Figure 2.13, does not mean that the energy has a lower value in the same region. The research carried out in [13] shows that it is most likely that this energy does not decrease, see Figure 2.18, as wider and smoother low-amplitude pulses have the same energy as shorter and steeper high-amplitude ones.

2.2 The basic physical processes | 71

Figure 2.15: Standard shape of HEMP E1 pulse [6].

Figure 2.16: Meaning of “full width at half maximum” (FWHM) characteristic. Table 2.1: HEMP parameters proposed by various authors at different times. Parameter

Peak Field, kV/m Rise Time, ns FWHM, ns Energy Density, J/m

Reference Bell Labs 1960 A [9]

Baum 1992 B [10]

Leuthäuser 1994 C [11]

VG95371-10 1995 D [12]

IEC 61000-2-9 1996 E [1]

50 4.6 184 0.891

50 2.5 23 0.114

60 1.9 23.8 –

65 0.9 24 0.196

50 2.5 23 0.114

It is worth noting here that values F (Fall Time) indicated by authors [13] (a trailing edge or a pulse-decay time) are not used to describe HEMP parameters (or lightning pulse). Instead, the so-called “full width at half maximum” or FWHM parameter is used. Moreover, even if the pulse-decay time (or time when the signal value decreases

72 | 2 A contemporary view of HEMP for electrical engineers

Figure 2.17: HEMP shape proposed by various authors at different times.

Figure 2.18: Calculated HEMP pulses of different shapes with the same energy.

from 90 % to 10 % of an amplitude) is applied, the values shown in the diagram do not correspond to the diagrams themselves. Requesting all three authors of the article [12] for the explanation of the situation was unsuccessful: No answers were received. Fourier transformation of a standard curve of the E1 pulse (see Figure 2.19), according to [3], shows that within the range of 10 kHz to 1 MHz, the electric-field strength stays relatively constant and maximum, and quickly decreases (almost by a factor of 100) when the frequency rises from 1 MHz to 100 MHz and decreases even faster for frequencies above 100 MHz, see Figure 2.19.

2.2 The basic physical processes | 73

Figure 2.19: Distribution of the electric-field strength within the frequency range under Fourier transformation of a standard E1 pulse.

Figure 2.20: Distribution of energy within the HEMP frequency range [1].

Thus, the HEMP frequency range in IEC 61000-2-9 [1] is defined as within 100 kHz– 100 MHz, where the pulse-energy release reaches 96 %, see Figure 2.20. In [14], see Figure 2.21, an unusual but clearly evident definition of HEMP frequency range is depicted. And if one were to narrow the power range up to 90 %, the frequency range of the HEMP will decrease up to limits of 100 кГц–10 MHz, see Figure 2.22 [15]. Such essential compression of the frequency range of the HEMP in comparison to some statements in which the HEMP frequency range reaches up to 1 GHz, is very important, because this range determines the major characteristics of shielding materials, elements and designs principles, and also the demands on them. Obviously, some part of the HEMP spectrum is also at frequency of 1 GHz detected, however, the contribution of this part to total energy of a HEMP is so insignificant that it can be neglected.

74 | 2 A contemporary view of HEMP for electrical engineers

Figure 2.21: HEMP frequency range according to [13].

Figure 2.22: Energy distribution within the HEMP frequency range according to [15].

Why does the electric field emanate so strangely (see Figure 2.13) from the nuclear explosion epicenter? Since EMP generation depends largely on the magnetic field of the Earth, the answer is obvious: because of the Earth’s magnetic field. The Earth’s magnetic field has a rather interesting structure and shape. Prevalently, the Earth’s magnetic poles do not match the geographical poles and show a tendency to slowly shift. Secondly, the magnetic field has a different value at various points of the ground surface. The weakening of the magnetic field is accompanied by a decrease in HEMP intensity. Thirdly, the vectors of the horizontal and vertical components of the magneticfield induction have certain angles. The angle between the geographical and magnetic meridians at the defined point of the Earth’s surface, demonstrating the difference between the magnetic compass readings and the true north direction at this point, is called the magnetic declination (or magnetic variation). The angle of the compass needle vertical deflection under the influence of the Earth’s magnetic field is called the magnetic inclination (or magnetic dip). Also, in the Northern Hemisphere, the needle

2.2 The basic physical processes | 75

Figure 2.23: A map of the isolines of the main magnetic field (µT) on the surface of the Earth.

tip pointing to the North goes down (to the Earth’s surface), in the Southern hemisphere, it goes up. The magnetic inclination value is measured by a special device known as the inclinator. Figure 2.23 shows a map of isolines of the general magnetic field intensity on the surface of the Earth. An isoline is a line of points, where the measured characteristic has the same value at every point on the line. Since, as was shown above, the Earth’s magnetic field is directly involved in HEMP generation and this magnetic field is heterogeneous. The calculations presented in [16] demonstrate that the same yield of nuclear explosion, detonated at the same height, create at the Earth’s surface an electromagnetic pulse that differs significantly in the amplitude of the electric field and in energy, depending on the location (geomagnetic latitude) of the epicenter of the explosion, see Figure 2.24. Since the HEMP pulse shape of 2.5/25 (2.5/23) nanoseconds described previously is related to the pulse of the voltage applied to the apparatus, the situation becomes even more complicated. However, the current pulse starting to flow under the applied voltage pulse has a completely different shape: 10/100 nanoseconds (IEC 61000-5-3, IEC 61000-2-10), see Figure 2.25. The shape of the current pulse depends heavily on the load conditions, i. e., on current circuit inductance and capacitance. It is obvious that a change of length of a cable (wire) will change both current parameters: the shape of the current pulse and it amplitude, Figure 2.26 [17]. However, in order to simplify the situation, the particular shape of the average standard current pulse with the parameters of 10/100 nanoseconds was accepted. The situation is even more complicated with the buried electric cables because the ground is a semiconducting environment providing partial reflection of the decreasing electromagnetic wave, partially shunting the EMP. Obviously, the degree of the ground

76 | 2 A contemporary view of HEMP for electrical engineers

Figure 2.24: Variation of HEMP electric-field strength as a function of latitude in the Northern hemisphere for an explosive yield of 10 kt and detonation altitude of 200 km.

Figure 2.25: Standard shape of HEMP-current pulse.

impact on the weakening of HEMP acting on the cable significantly depends on ground conductivity and the cable burial depth, see Figure 2.27. According to the information just mentioned, it should be clear that HEMP values generally accepted in standards are averaged and generalized and do not reflect reallife and real-device values. The only good thing here is that in most cases the standards show the worst-case values, so hopefully, the real-life HEMP would be less severe than that described in standards. Nevertheless, in 1985, the United States Department of Defense prepared the special standard MIL-STD-2169, repeatedly amended and cor-

Bibliography | 77

Figure 2.26: Dependence of current magnitude on cable length (1–10 m) at the HEMP impact.

Figure 2.27: Dependence between the cable burial depth and the shape of the pulse applied by a standard HEMP in the ground with a conductivity of σ = 10−2 mOhm per meter.

rected, where all just-mentioned relationships were represented as graphic charts designed to calculate the HEMP impact on objects operating under the various conditions. Unfortunately, today this standard remains classified.

Bibliography [1] [2]

IEC 61000-2-9 Electromagnetic Compatibility (EMC) Part 2: Environment – Section 9: Description of HEMP environment – Radiated disturbance, 1996. Akbashev B. B., Baluk N. V., Kechiev L. N. Protection of Telecommunications against Electromagnetic Disturbances. Gryphon, Moscow, 2013. 472 p. (by Russian).

78 | 2 A contemporary view of HEMP for electrical engineers

[3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13]

[14] [15] [16] [17]

Report No. EP 1110-3-2 Electromagnetic Pulse (EMP) and Tempest Protection for Facilities, Engineering Department of the Army, Washington, 1990. Seiler L. W., Jr. A Calculational Model for High Altitude EMP. Air Force Institute of Technology. Report AD-A009208. Wright-Patterson Air Force Base, Ohio. March 1975. Study to Access the Effects of Magnetohydrodinamic Electromagnetic Pulse on Electric Power Systems. Report ORNL/sub-83/43374/1/v, Oak Ridge National Laboratory, 1985. Philip J. Dolan’s Capabilities of Nuclear Weapons, DNA-EM-1 Chapter 7, page 7-1 (change 1 page updates, 1978), Report AD-A955391. Report AD-A144408. Evaluation of Methodologies for Estimating Vulnerability to Electromagnetic Pulse Effect. Washington, 1984. Glasstone S., Dolah P. J. The Effect of Nuclear Weapons. US Department of Defense, Energy Research and Development Administration, Washington, 1977. EMP Engineering and Design Principles, Electrical Protection Department, Bell Telephone Laboratories, 1975. Baum C. E. “From the Electromagnetic Pulse to High-Power Electromagnetics,” Proc. IEEE, Vol. 80, No. 6, June 1992, pp. 789–817. K-D. Leuthäuser, “A Complete EMP Environment Generated by High-Altitude Nuclear Bursts: Data and Standardization, ” Theoretical Note 364, Air Force Phillips Laboratory, February 1994. VG95371-10 from Bundesamt für Wehrtechnik und Beschaffung, Germany (replaces Edition 1993-08). Ghandehari M. B., Lotfi-Neyestanak A. A., Naser-Moghadasi M. Electromagnetic Pulse Coupling Inside a Rectangular Enclosure with an Aperture. J. Electromagn. Anal. Appl., 2011, Vol. 3, pp. 84–89. Viel J. Testing for Immunity to EMP. Compliance, 2010, 1 July. High-Altitude Electromagnetic Pulse (HEMP Testing. Test Operation Procedure (TOP 01-02-620). U.S. Army Test and Evaluation Command, November 10, 2011). Leuthauser K. D. A Complete EMP Environment Generated by High-Altitude Nuclear Bursts, Theoretical Note 363, October 1992. Jianguo Z., Xin Z. Coupling Effect of Transmission Lines by HEMP Based on CST. General Assembly and Scientific Symposium (URSI GASS), Beijing, China, 16–23 Aug, 2014.

3 HEMP simulators 3.1 HEMP simulators: principle of operation Upon signing the nuclear-test ban treaty covering nuclear tests in the atmosphere, in space and underwater (later, it was extended to an unconditional overall test-ban treaty in 1996) by the USSR, the US and Great Britain, the leading countries started designing and building HEMP simulators to be able to test the effects of their ammunition, see Figure 3.1. Since the E1 component of HEMP causes the most significant effect on sensitive electronic equipment of control, communication and navigation systems, and since it is very difficult to build protection against it, the simulators were expected to model this effect. A modern typical HEMP simulator consists of two major parts: a source of highvoltage pulses and an antenna system, which creates electric-field pulses (matching E1 component) within the operational volume of the simulator, where a test object is placed (Figure 3.2). Usually, a source of high-voltage pulses is represented by a pulse voltage generator (PVG) assembled according to the Marx design. This generator consists of a set of high-voltage capacitors, charged in parallel by an external power source and controlled discharge tubes, connected in such a way that, upon their internal breakdown, the capacitors become connected in series, and the output electrodes of the generator receive voltage equal to the sum of all voltage rates from all capacitors (Figure 3.3). The picture of a Marx generator (Figure 3.3) shows small sparks at multiple discharge gaps and one powerful electric arc between major electrodes.

Figure 3.1: Assembly of one of the first HEMP simulators in the US. https://doi.org/10.1515/9783110639285-003

80 | 3 HEMP simulators

Figure 3.2: Typical design of a HEMP simulator.

Figure 3.3: Simplified circuit diagram and desktop experimental working model of a Marx generator. C1-C4—capacitors; G1-G3—discharge tubes; R2-R6—resistors, equalizing voltage distribution on capacitors.

Figure 3.4: Real Marx PVG.

Real full-sized Marx PVGs are built based on the same design, but obviously employ much larger and more expensive elements (Figure 3.4).

3.2 Classification of HEMP simulators | 81

Figure 3.5: PVGs rated for several million volts.

PVGs rated for several million volts are rather massive (Figure 3.5). In order to obtain a PVG of a reduced size, they are sometimes placed into large basins made of insulation material and filled with insulating mineral oil or into special air-tight chambers filled with sulfur hexafluoride gas (SF6 ). Discharge tubes used in these designs are produced with either a pneumatic or hydraulic drive, which makes possible adjusting the gaps between discharge tubes’ electrodes, i. e., their breakdown voltage and, consequently, the amplitude of the aggregate output pulse of the PVG.

3.2 Classification of HEMP simulators Antenna systems of HEMP simulators are very diverse. According to IEC 61000-4-32, there are three provisional types of HEMP, i. e.: – guided-wave; – dipole; – hybrid Actually, this classification was first offered by a renowned specialist in this field, Dr. Carl Edward Baum, research fellow of Air Force Weapons Lab., Kirtland Air Force Base, Albuquerque, see Figure 3.6. At some time later, it became part of the standard. Many American HEMP simulators were built under C. E. Baum’s supervision.

82 | 3 HEMP simulators

Figure 3.6: Carl E. Baum.

Guided-wave simulators also have other names, used in technical literature, such as: – two parallel-plate transmission line; – open parallel-plate waveguide It should be noted that another standard of the same Committee (SC77C) of the same International Electrotechnical Commission (IEC 61000-4-25) offers completely different classifications based on some minor differences in parameters of a generated pulse, such as its rise time (leading edge) and width. According to IEC 61000-4-25, Type I simulator features the following specifications: rise time: 2.5 ± 0.5 ns; width: within 25–75 ns; frequency range: 1–200 MHz. Type II simulator features the following specifications: rise time: 2–10 ns; width: 25–500 ns; frequency range: 1–100 MHz. This classification was not received positively by specialists, so we shall continue using the one offered in IEC 61000-4-32.

3.3 Foreign HEMP simulators Guided-wave antenna systems are the simplest in terms of their design and the most widely implemented. Figure 3.2 depicts a simulator employing this kind of antenna system. Design of a real simulator is very similar to this simplified layout (Figure 3.7). The bottom “plate” of this simulator comprises a metal grid placed in a concrete foundation, while the top “plate” comprises rows of stretched wire, supported by insulated supports. Such a type creates a vertical electric field (between the plates). Moreover, it is not affected by reflection of a wave from the ground surface. The latter is important for testing of objects located above the ground during HEMP impact (e. g., planes and missiles). At the same time, testing of ground objects enclosed into such a simulator (which otherwise would have been affected by HEMP, reflected from the ground surface) cannot be considered fully-fledged.

3.3 Foreign HEMP simulators | 83

Figure 3.7: Guided-wave HEMP simulator.

Figure 3.8: Dipole simulator.

Other simulators—the so-called dipole simulators—feature partial reflection of the incident wave from the ground or water surface (Figure 3.8). These are mainly used to test ground equipment, as well as planes parked on airfields during HEMP impact (Figure 3.8) and ships (Figure 3.9). Figure 3.9 shows that the simulator is located alongside, and a small distance from the object was being tested. This makes possible modelling both a vertical and partially horizontal component of HEMP. Dipole simulators are less efficient in terms of converting pulse energy into an electric field. Since such simulators reproduce mostly vertical components of the field, they are often denoted as “VPD” (Vertical Polarized Dipole).

84 | 3 HEMP simulators

Figure 3.9: Dipole simulator EMPRESS II (USA), placed on a mobile sea platform (barge) to test a US Navy combat ship.

Figure 3.10: Design of a hybrid simulator (belongs to RAFAEL Company, Israel).

The third type of simulators—hybrid (Figure 3.10)—enables modelling both a horizontal component of E1 and a HEMP wave reflected from the ground surface. However, since its primary purpose is simulation of a horizontal component of E1, these simulators are denoted as “HPD” (Horizontal Polarized Dipole). The design of this simulator consisted of a generating module, strung in the middle and radiating cylinder-shaped antennas connected to it on both sides. Such simu-

3.3 Foreign HEMP simulators | 85

Figure 3.11: Testing of a plane and a helicopter on a hybrid simulator.

lators are mainly used to model HEMP’s impact on flying planes and helicopters, when a horizontal component prevails under real conditions (Figure 3.11). Since the pulse generator employed in this simulator was raised above the ground surface, its capacity was rather restricted and significantly less than that of other simulators. Obviously, this explains why such simulators are not so widely used. The advent of HEMP simulators and the onset of their use brought about many reports, mostly made-up by military research laboratories and devoted to parameters and specifications of these simulators. Today, these reports are available in open US archives for those willing to expand their knowledge in this field (Figure 3.12). The world’s largest HEMP simulator was designed and built in the US in the 1970s and commissioned early in 1980. It was designed by the Air Force Weapons

86 | 3 HEMP simulators

Figure 3.12: Some open domain reports regarding HEMP simulators.

Laboratory—AFWL at Kirtland Air Force Base, New Mexico. The simulator was named using the acronym of the AFWL Transmission Line Aircraft Simulator, i. e., ATLAS. At that time, the unit was still classified, and thus its second designation was “TRESTLE”. The huge platform of this simulator could easily accommodate a long-range bomber B-52 and was completely made of wood (15,000 cm3 of high-strength wood) (see Figure 3.13) in order to eliminate any reflection of the electromagnetic wave from

3.3 Foreign HEMP simulators | 87

Figure 3.13: The world’s largest HEMP simulator “ATLAS” during B-52 testing. Not the truck approaching the plane on the upper picture and imagine the size of the simulator. There are supports with stretched radiating antennas around the testing area.

metallic structural parts affecting HEMP. Even the securing bolts were made of highstrength wood. Reflection from the ground was reduced to nothing since the working volume of the simulator was located high above the ground surface. The length and width of the working platform of this simulator is 300 meters; its height is about 200 meters.

88 | 3 HEMP simulators

3.4 HEMP simulators available in Russia and Ukraine Large HEMP simulators were built by three institutions in Russia: CR&DI-12 of the Ministry of Defense (Sergiyev Posad); its branch office CR&DI-26 of the Ministry of Defense (1, Gangutskaya Street, St. Petersburg); and the All-Russia Electrotechnical Institute n. a. Lenin (12, Krasnokazarmennaya Street, Moscow). CR&DI-12 was previously discussed. What is CR&DI-26 in the Ministry of Defense? It turned out that this institute has nothing to do with physics, nuclear projects and HEMP. This is an R&D institute that belongs to a military building organization. Its branch office in Leningrad used to be the Central R&D Laboratory of the Soviet Navy, and afterwards it was an independent R&D Institute-12 of Naval Forces. In 1961, a strange idea came into some peoples’ minds: they converted the independent R&D institution, researching explosions, high-voltage pulse equipment, HEMP impact on marine electric equipment and other physics-related issues in the military domain, into a branch office of the military Design and Construction Institute (CR&DI-26 of the Ministry of Defense, Balashikha town, Moscow region, Gagarin residential area, 31). Later, CR&DI-26 and its branch office in St. Petersburg became the branch office of the 31st State Design and Construction Institute for Special Building of the Defense Ministry of Russian Federation (established by the Directive of the General Staff of the Red Army No. Org/5/315546, dated December 22, 1944, under the name “Central Design and Engineering Institute of the Red Army”, 19, Smolensky Blvd., bldg 1, Moscow), Figure 3.14. Today, this “unhappy” branch office is called the “Research and Development Center of the 26th Central R&D Institute of the Defense Ministry of Russian Federation”. This R&D Center of the 26th CR&DI of the Defense Ministry of RF operates its own range of impact and electromagnetic effects, located in Pesochny village, Vyborg district, Leningrad region.

Figure 3.14: State Design Institute for Special Building of the Ministry of Defense of Russian Federation.

3.4 HEMP simulators available in Russia and Ukraine

| 89

Figure 3.15: HEMP simulator at the R & D Center of the 26th CR & DI of the Defense Ministry of RF SEMP 12-3 (according to IEC 61000-4-32); length = 170 m. There is a powerful Marx generator at the end of the simulator (left), which looks like a high cylindrical tower.

This range accommodates a very large HEMP simulator, suitable for testing of strategic intercontinental missiles (Figure 3.15). In addition to CR&DI-12 and R&D Center of the 26th CR&DI of the Defense Ministry of RF, there is another simulator, which belongs to the High-Voltage Research and Development Center of the All-Russia Electrotechnical Institute—HVR & DC of AREI n. a. Lenin (12, Krasnokazarmennaya Street, Moscow). This simulator is called “Allure” and is located at the range of the HVR & DC of AREI near Istra town, Moscow region (Figure 3.16). There is a smaller HEMP simulator at the State Missile Center n. a. Academy Fellow V. P. Makeyev—JSC “Makeyev SMC” (1, Turgoyak Hwy, Miass town, Chelyabinsk region), see Figure 3.17. This holding is the major designer of liquid- and solid-fuel strategic missiles in Russia. It was established subject to the Resolution of the Council of Ministers of USSR dated December 16, 1947, as part of plant No. 66 in Zlatoust town (Zlatoust Machine-Building Plant) under the name “Special Design Bureau for Longrange Missiles” with laboratories and testing facilities; since 1948, Special Design Bureau No. 385 (SDB-385). In 1955, SDB-385 was moved to Miass town, Chelyabinsk region. Today, JSC “Makeyev SMC” includes: JSC Krasnoyarsk Machine-Building plant, JSC Miass Machine-Building Plant, Public Corporation R&D Institute “Hermes”, JSC Zlatoust Machine-Building Plant. The simulator owned by JSC “Makeyev SMC” can create an electromagnetic pulse with the field density of up to 200 kV/m. Obviously, based on the working volume of the simulator and the size of the Marx generator, it is suitable to test some individual electronic parts, but not large missiles.

90 | 3 HEMP simulators

Figure 3.16: HEMP simulator of HVR & DC of AREI.

As for former USSR countries, apart from Russia, the simulators were also built in Ukraine, e. g., the Research and Design Engineering Institute “Molniya” (RDEI “Molniya”) of the National Technical University “Kharkov Polytechnic” (47, Shevchenko Street, Kharkov), whereas its testing facilities are located in Andreyevka village (Balakleya district, Kharkov region). The history of this institute dates back to 1954, when the special Decree of the Cabinet Council established an R&D laboratory of high-voltage equipment and current converters (R&DL HVE and CC) as part of the Electrical Power Transmission Department of the Kharkov Polytechnic Institute. In 1974, R&DL HVE and CC was converted into the Experimental Design Bureau of High-Voltage Pulse Equipment—EDB HVPE. Comprehensive, full-size testing of space systems, aviation and military machinery took place at the range of EDB HVPE. The range was operating in three shifts. A hotel and canteen were built to accommodate representatives of various enterprises accompanying their equipment during the tests. This is the place where the first (and initially unsuccessful) testing of the control system of SS-18 “Satan” (15A14), built at Enterprise

3.4 HEMP simulators available in Russia and Ukraine

| 91

Figure 3.17: HEMP simulator of JSC “Makeyev SMC” K107.0100.000. There’s a multistage Marx generator on the left.

Figure 3.18: HEMP simulators owned by RDEI “Molniya”.

No. 67 (currently, SPC “Khartron”, Kharkov) took place. In 2007, RDEI “Molniya” tested lightning protection for aviation structures made of polymer composite materials under a contract with Boeing (US).

92 | 3 HEMP simulators Among the various simulators at RDEI “Molniya”, there are very large units, such as GINT-12-30, which is a length of 254 meters (Figure 3.18). Some specifications of large stationary HEMP simulators are shown in Table 3.1. Table 3.1: Some specifications of large stationary HEMP simulators operating in various countries. HEMP Simulator Type Guided-wave ALECS APES Atlas-1 (TRESTLE) VPBW NOTES ERU-2M SEMP 6M-2M Pulse M SEMP 12-3 SEMP 1.5 IEMP-10 GIN 1.6-5 IEMI M5M (2 variants) GINT 12-30 VEPES SEMIRAMIS VERIFY INSIEME SIEM-2 SSR France Telecom RAFAEL DM-1200 TDRI EMIS-III-TL DIESES SAPIENS-2 NEMP DREMPS Dipole VPD-I EMPRESS-I EMPRESS-II (VPD-II) US Navy VPD EMIS-III-VPD VPD

Country

Voltage Amplitude (MV)

Electric Field (kV/m)

Risetime (ns)

Simulator’s Length (m)

1992 1989 1991 Switzerland 1999 Early 1990s 1972 1986 France 1996 Israel 1989 China 1985 Japan 1999 Holland 1992 Germany 1981 Sweden 1990 Czech 2004 Canada Mid 1990s

1 4 6–8 – 1 1 6 0.6 2.4 1.5 2.5 1.6 2.5 0.7 4.5 0.8 0.1 0.6 1 2.8 2 0.8 2 1.2 0.3 0.5 1 1 0.45 0.6

100 >100 50 50 100 100 100 100 30 20–100 140 150 – 330 120 100 62 100 100 100 100 75 200 120 50 50 100 50 40–50 55

10 6 20 1–2 3–5 2.5–25 9 5 >20 5–12 20–40 5–10 20–40 5–10 5–10 8 10 1 4 10 1–5 2.5 5 10 6 10 1–7 5 2.5–5 5

100 189 400 161 85 30 80 15 170 100 110 48 – 23 254 55 10 20 120 180 106 50 130 54 3.6 50 120 90 30 100

Early 1970s Early 1970s Late 1980s Late 1970s Early 1980s Early 1980s 2001

1.6 1.5 7 4 – 0.5 0.4

10 3 25 36 – 2 10

5 8–15 10 10 5 5 1.2

– – – – – – –

USA

Russia

Ukraine

USA Holland Germany

Year

Mid 1960s 1970 Early 1980s 2005 2005 1982 1982 Early 1990s 1992 1998 1970 1976 1992

3.5 Portable HEMP simulators | 93 Table 3.1: (continued) HEMP Simulator Type Hybrid TEMPS & AESOR ATHAMAS-II-HPD (USAF) ATHAMAS-I US Navy-HPD DPH RRAFAEL HPD SPERANS NEMPS WIS

Country

USA

Year

Early 1970s Mid 1970s

– Mid 1970s France 1980 Israel 1991 Sweden 1984 Switzerland 1985 Germany 1999

Voltage Amplitude (MV)

Electric Field (kV/m)

Risetime (ns)

Simulator’s Length (m)

7 4

52 33

4–12 8–12

300 150

– 5 4 0.6 0.2 4 0.36

– 46 50 9 4 60 10

– 2 1–5 5 2.5 10 1.2

– 150 150 30 150 60 30

3.5 Portable HEMP simulators The relevance of large HEMP simulators has recently decreased significantly as computer software capable of HEMP impact modeling became available. Thus, now there is no need to conduct expensive full-size testing. Simultaneously with this, portable, relatively cheap (US$100,000–150,000) HEMP simulators became available. These are intended for in-house testing (Figure 3.19) of relatively small electronic devices (Figure 3.20).

Figure 3.19: Design of in-house test bench with portable HEMP simulator.

94 | 3 HEMP simulators

Figure 3.20: Portable in-house HEMP simulators, from top to bottom: Elite Electronic Engineering Inc (USA); Montena (Switzerland); Applied Physical Electronics (USA).

4 The vulnerability of electronic equipment to HEMP “...our vulnerability is increasing daily as our use of and dependence on electronics continues to grow in both our civil and military sectors.” Dr. William R. Graham, EMP Commission’s Chairman

4.1 Electronic equipment is the most important component of the modern infrastructure Today, it is impossible to imagine that any part of the country’s infrastructure operates without electronic equipment based on microprocessors and computers. This is especially true for critical systems such as the power supply, the water supply and sewage, communication and transportation networks. Modern electronic equipment uses lower level electrical signals (compared to the electrically powered equipment that was used in the beginning of the previous century or even electronic equipment of the previous generation that was based on electron vacuum tubes) and features much more intricate and branched internal architecture. This is the key to the sharp increase of modern electronic equipment’s susceptibility to High-Altitude Electromagnetic Pulse (HEMP) and that of the entire country’s infrastructure. When analyzing the HEMP’s impact on the infrastructure, special emphasis is always placed on electronic equipment. For example, the comprehensive report of the US Congress commission dealing with the HEMP’s impact on critical parts of the country’s infrastructure addresses primarily the issues of electronic equipment’s susceptibility. Indeed, the report begins with the analysis of susceptibility of such electronic systems. The first chapter of this multipage report is called “The SCADA system. Susceptibility of SCADA and its effect on critical infrastructure” [1]. Supervisory Control and Data Acquisition (SCADA) is a hardware and software package intended for realtime collection, processing, displaying and compression of information in complex and branched systems of various purposes, the components of which are remotely located and autonomously operate automatically. These include the power and watersupply systems etc. According to the report, analysis of computers, local area networks and common use of the network hardware revealed that where the length of connection cables is 7–60 m, the magnitude of induced current can reach as high as 100–700 A, causing inoperable computer equipment. The analysis revealed that multiple duplication and backing-up of equipment is not efficient, that the volume of reasonable reserve does not correspond to the assortment of spare parts and that the number of operating units is not adequate. https://doi.org/10.1515/9783110639285-004

96 | 4 The vulnerability of electronic equipment to HEMP The report also found that an electromagnetic pulse (EMP) impact can disable the majority of electronic equipment in civil communication and telecommunication systems. In this regard, mobile communication systems will be much more affected than the ordinary communication lines. Even weak EMP can lead to a several-day failure of mobile communication systems, whereas ordinary electrical lines will be out of order for several hours. These estimates were obtained based on the model of EMP damage and further recovery of mobile communication systems. The report also assesses the EMP resistance of modern electrical locomotives, equipped with a microprocessor control system. The newer models use three computers to control all the subsystems. Failure of any of them will cause the locomotive to stop. Malfunctioning of the locomotive’s equipment starts at an EMP electric-field’s density of 4–8 kV/m, whereas outright failure occurs at 20–40 kV/m. The committee concluded that it takes several days or weeks to restore the railway service due to equipment repair, and several months in case it is necessary to substitute the damaged computer systems. The testing of 37 cars manufactured in 1986 through 2002 and 18 trucks manufactured in 1991 through 2003, regarding the impact of standard HEMP with the field density of up to 50 kV/m, revealed that approximately 10 % of the vehicles will immediately become inoperable and thus create emergencies at electric-field densities of 25–30 kV/m and higher. Traffic-control centers will also become inoperable, thus making the situation in large cities even worse. The resulting power outages will result in shutdown of the majority of traffic lights. The list of disasters that can happen from HEMP impact on the country’s infrastructure is even longer. However, specialists should, in our opinion, be more interested in addressing more specific issues at the level of the electronic equipment’s hardware components. This would enable better understanding of the processes underlying the operation of the electronic equipment and how it can be protected. During the previous 50 years, there has been much research devoted to the HEMP susceptibility of various electronic components, as well as integral microchips, microprocessors, microcontrollers, computers and computer networks. In the next section, the findings of this research are summarized.

4.2 The vulnerability of discrete electronic components to HEMP For discrete electronic components, such as diodes, transistors, capacitors, resistors, etc., the failure resulting from the impact of an electromagnetic pulse is (though not always!) irreversible and attributable to the breakdown of their internal structure under the pulse of an applied reverse voltage, or that of a forward current exceeding specific threshold values. The technical report written by the staff of the Air Force Research Laboratory [2] makes a profound analysis of various theoretical models of disruption of diodes and

4.2 The vulnerability of discrete electronic components to HEMP

| 97

transistors’ semiconducting structure, discusses the limitations of various models and provides related critical judgments. Two major theoretical models enjoyed the actual performance: thermal (Wunsch–Bell model) and electro-thermal, (Ward model and Barush–Budenstein model) named after their authors. The report suggests that the thermal Wunsch–Bell model does not fully reflect the physical processes disrupted within the semiconducting structure, and in particular, in the case of nanoseconds, the attempts to use the Wunsch-Bell model are incorrect. However, the fact is that, in the case of quick reduction of pulse power below a certain threshold limit, the semiconductor device can resume its functionality. Thus, there can be two types of malfunctioning in semiconductor components and (which is even more important and relevant) electronic devices, i. e., reversible (upset) and irreversible (destruction), see Figure 4.1.

Figure 4.1: Upset and destruction in semiconductor devices and their constituents caused by the external impact pulse.

In the case of an upset, there is no physical damage and destruction to the internal structure of a semiconductor device or a microchip, and its functionality can be fully restored automatically after some delay without any intervention or manually after a system reboot (with regard to a computer system). The destruction results in physical damage of the semiconductor devices’ internal structure, which is related to the damage to p-n junctions (blowing) of a semiconductor structure (see Figure 4.2) [3] or puncturing of insulation between individual elements in the integrated system (see Figure 4.3) [4]. Destruction makes it impossible to restore the functionality of a device, and thus the element (device) needs to be replaced. In our opinion, theoretical arguments, about an approximate description of what happens when a semiconductor structure of diodes and transistors fails, offer theoretical rather than practical interest because the range of parameters obtained as a

98 | 4 The vulnerability of electronic equipment to HEMP

Figure 4.2: Puncture and blowing of microprocessors’ semiconductor structures affected by the electromagnetic pulse of 7.5/180 nanoseconds (these pictures were taken using a microscope).

Figure 4.3: Destruction of the 7400 series integral microchip (logical element AND-NO based on Schottky transistors) caused by the puncture between P1 and P2 points of the circuit and shortcircuiting of R1 resistor.

result of experiments, even for similar groups of diodes and transistors, is so broad that they neglect the significance of arguments about the accuracy of the various models. To confirm this, Figure 4.4 shows experimental data obtained in [5] for a widely used general-purpose bipolar transistor 2N2222, manufactured by many companies for dozens of years.

4.3 Vulnerability of integral circuits (microchips) to HEMP

| 99

Figure 4.4: Dispersion of destruction parameters of a group of 2N2222 transistors during the testing.

Obviously, the accuracy of one or another theoretical model has no practical value in the case of such significant dispersion of experimental data. Nevertheless, the damage of semiconductor diodes and transistors is approximately associated with thermal processes and the damage of p-n-junctions. Thus, it is obvious that the power of a pulse capable of damaging the semiconductor structure will depend on the duration (width) of this pulse. These dependencies, obtained experimentally for various types of semiconductor devices, are shown in Figure 4.5 [5]. It can be clearly seen that the area in Figure 4.5, where the damage of certain types of transistors or diodes takes place, is not a straight line, but rather an area containing many specifically scattered dots. These dots correspond to experimental data similarly to that shown in Figure 4.4. The straight lines represent averaged and generalized trends determined by these dots.

4.3 Vulnerability of integral circuits (microchips) to HEMP For more complicated microelectronic elements, such as integral microchips, there are also upsets (Figure 4.6) [6], in addition to irreversible destruction that actually occur frequently, unlike the very rare upsets of discrete semiconductor elements.

100 | 4 The vulnerability of electronic equipment to HEMP

Figure 4.5: Power of a destruction pulse as a function of its duration (for diodes and transistors with maximum operating voltage of 40–500 V and current rates of 0.1–1 A).

It is interesting to note that the response of microchips with transistors built with CMOS technology (Complementary Metal-Oxide-Semiconductor) is completely different from that of microchips with bipolar transistors. Microchips based on bipolar transistors (unlike CMOS) rarely experience operating upsets; rather, they are prone to immediate irreversible destructions (this feature of bipolar transistors was already mentioned). The circuits on CMOS technology were invented by Frank Wanlass at Fairchild Semiconductor in 1963. Nowadays, the speed of switching, density of installation and energy consumption provided by CMOS technology are unattainable for technologies based on bipolar transistors. As a result, logical microchips based on CMOS transistors have almost completely displaced microchips based on bipolar transistors. Consequently, the upsets of microchips under the impact of external electric field are now more relevant as ever before. A relatively high resilience of individual microchips to the pulse of an external electric field can be explained by the small dimensions and small distances between the terminals. Smaller distances result in a relatively small potential difference occurring between conducting elements impacted by the pulse of an external electric field. However, considering that, upon HEMP impact, the external wires and cables connected to inputs of electronic devices (including microchips) will suffer high induced voltages applied directly to inputs and outputs of microchips, the situation becomes more serious.

4.3 Vulnerability of integral circuits (microchips) to HEMP

| 101

Figure 4.6: Upsets and destructions of various types of logical microchips as affected by the pulse of external electric field.

Table 4.1 [7] shows the data regarding destruction of various microchips upon direct application of voltage to their inputs. During the trials, the amplitude and duration of applied pulses changed, and these values were fixed. Other parameters (current, power, energy) were not strictly fixed and were dependent on the amplitude of apTable 4.1: Parameters of destructive pulses applied to terminals of some types of microchips. 8T15 LINE DRIVER 8T16 RECEIVER 8T15 OUTPUTS

8T16 INPUT

PULSE WIDTHS 25 ns 100 ns VOLTAGE, V CURRENT PULSE, A POWER, W ENERGY, mcJ VOLTAGE, V CURRENT PULSE, A POWER, W ENERGY, mcJ

150 9 1350 33,7 220 1,5 330 8,25

140 2 280 28 175 0,8 140 14

1 mcs 60 0,55 33 33 125 0,14 17,5 17,5

102 | 4 The vulnerability of electronic equipment to HEMP Table 4.1: (continued) 5404 HEX INVERTER TTL 54L30 NAND GATE TTL

30 0,5 15 15 20 0,8 16 16

µA747 LINEAR OP AMP

PULSE WIDTHS 25 ns 100 ns

1 mcs

µA747 INPUT

NO FAILURES to 500 V 65 100 200 6,5 9 12 422 900 2400 10,5 90 2400

54L30

µA747 OUTPUT

VOLTAGE, V CURRENT PULSE, A POWER, W ENERGY, mcJ

MINIMUM VALUES CMOS CD4000-SERIES CD4001A

CD4016

CD4049

CD4050

CD4050

CD4071

VOLTAGE, V CURRENT PULSE, A POWER, W ENERGY, mcJ VOLTAGE, V CURRENT PULSE, A POWER, W ENERGY, mcJ VOLTAGE, V CURRENT PULSE, A POWER, W ENERGY, mcJ VOLTAGE, V CURRENT PULSE, A POWER, W ENERGY, mcJ VOLTAGE, V CURRENT PULSE, A POWER, W ENERGY, mcJ VOLTAGE, V CURRENT PULSE, A POWER, W ENERGY, mcJ

120 3 360 9 90 4 360 9

1 mcs

50 1 50 5 50 2 100 10

5404

VOLTAGE, V CURRENT PULSE, A POWER, W ENERGY, mcJ VOLTAGE, V CURRENT PULSE, A POWER, W ENERGY, mcJ

PULSE WIDTHS 25 ns 100 ns

PULSE WIDTHS 25 ns 100 ns 350 10 3500 87,5 150 2 300 7,5 150 15 2250 56,2 170 13 2210 55,2 120 4 480 12 80 5,2 416 10,4

150 7 1050 105 120 4 480 48 25 6 150 15 60 7,5 450 45 60 4 240 24 150 0,3 45 4,5

1 mcs 60 1,2 72 72 20 2 40 40 12 3 36 36 20 3 60 60 24 2 48 48 250 0,4 100 100

4.3 Vulnerability of integral circuits (microchips) to HEMP

| 103

plied voltage and duration of pulses. Their values were obtained as a result of measurements. This data suggests that damages to microchips occur at relatively low voltage, much lower than the real values that can occur from HEMP impact in the absence of adequate protection from pulse overvoltage. Moreover, it should be noted that the susceptibility of microchips to the pulse of an applied voltage decreases with an increase of the pulse duration. As shown in [8], the dispersion of parameters responsible for the damage of semiconductor devices from the impact of short voltage pulses is stipulated by the lognormal distribution (see Figure 4.7). This is the case when pulse voltage is directly applied to the microchip’s terminals. Figure 4.7 shows that the voltage withstands the

Figure 4.7: Log-normal distribution of damage of 7400 type microchips caused by voltage pulse lasting for 44 and 0.44 microseconds.

104 | 4 The vulnerability of electronic equipment to HEMP microchip before the damage significantly increases with the decrease of pulse duration, which is in agreement with the data from Figure 4.5 and Table 4.1. However, Figure 4.7 also shows that the dispersion of the damaging voltage level is rather broad. Moreover, this dispersion applies to a microchip of the same type.

4.4 Vulnerability of microprocessors to HEMP Microprocessors have a much lower electric strength. The development of electronics follows a trend of continuous reduction in the size of separate elements, an increase of density of internal arrangement and complexity of integral circuits and microprocessors, reduction of insulation layers and, consequently, reduction of operating voltage. A good description of this trend is given by Moore’s Law [9], which was elaborated by one of the founders of Intel Company Gordon Moor in 1965. According to the law, the number of transistors in microprocessors doubles every other year, whereas their capacity grows exponentially (Figure 4.8). This law has now been in effect for decades.

Figure 4.8: Increase in the number of transistors on the microprocessor chip during the previous 40 years. There is a logarithmic scale on the vertical axis, so a straight line corresponds to the exponential law.

4.4 Vulnerability of microprocessors to HEMP

| 105

Figure 4.9: Information on microprocessors (inclusive up to Pentium III) and the levels of pulse electric fields resulting in their malfunction.

See Figures 4.9 and 4.10 containing more detailed information on microprocessors up to Pentium III level [3]. For comparison, microprocessor Intel Pentium IV HT 661, B1 data: operating frequency 3,600 MHz, power voltage 1.25 V at a structure size 0.065 µ (structure size determines the minimum size of the elements that can be formed on a semiconductor

106 | 4 The vulnerability of electronic equipment to HEMP

Figure 4.10: The probability of microprocessor upsets when converting from an older generation to a newer generation.

plate during the photolithography process, which lies at the foundation of microprocessor production). This high density of elements per a unit of the chip’s area leads to tremendous heat production and, consequently, necessitates the use of an especially efficient cooling system. Obviously, this and related limitations may soon slow down the continuous growth of the transistors’ number and violation of Moore’s Law. However, today’s level of placement density (which affects the insulation distance between the elements) and the level of supply voltage and operating frequency are self-evident: Susceptibility of modern microprocessors to HEMP has become much higher than it was 20–30 years ago. Further research has shown that the susceptibility of microprocessors to electromagnetic pulse depends significantly on the pulses’ frequency specifications (Figure 4.11) [3]. Figure 4.11 shows that a microprocessor’s susceptibility to a short pulse with a high rise time (which is more typical for the ultra-wide band (UWB) pulse generated by special directed sources of emission rather than for a wider HEMP pulse with a clearly lower rise time) is much higher. Regardless of lower energy of a UWB pulse, the efficiency of its impact (in terms of energy and voltage) on a typical electronic system proves to be higher. The first disruptions in the case of a UWB pulse occur at a field density of 12 kV/m, whereas, in the case of HEMP, this happens at 30 kV/m. Based on the above [3], it is concluded that the measure of induced energy is not the total energy of a pulse, but the energy in a certain frequency range.

4.5 Vulnerability of computers to HEMP

| 107

Figure 4.11: Probability of upsets of Rocky-518 HV circuit board with a Pentium-MMX 233 MHz processor as a function of various frequency specifications of a pulse.

4.5 Vulnerability of computers to HEMP Complex systems based on microprocessors, such as computers, are even more susceptible to HEMP. However, when testing free-standing personal computers equipped with a standard metal casing acting as a rather efficient shield, which is intended to weaken the external electromagnetic field, very high values for pulse amplitude in this field were obtained in [3] that caused destruction (Figure 4.12). The motherboard is much larger than individual elements of a computer. Thus, the potential difference occurring on the motherboard under the pulse impact of an external electric field is much higher than that between the inputs and outputs of smaller elements.

Figure 4.12: Pulse amplitudes of damaging electric fields (7.5/180 nsec) for various PC parts.

108 | 4 The vulnerability of electronic equipment to HEMP

Figure 4.13: The level of susceptibility (failures) of the simplest computer network for various network and cables configurations.

It is clear that the situation between a free-standing and a shielded PC and a computer network, with its long communication cables, will be completely different (Figure 4.13) [3].

4.6 Conclusions The above analysis shows that susceptibility of electronic equipment to HEMP significantly depends on both the dispersion of the components’ parameters (even of the same type) and on the variability of external conditions and the impacting parameters of a pulse. This means that it is absolutely impossible to forecast the susceptibility level of electronic equipment from a real nuclear explosion. There is another conclusion that can be made: Real testing of a certain type of electronic equipment on a test-bench, modeling HEMP, can be applied with their findings interpreted to a specific test sample only. They cannot be applied to similar units of other types, no matter how similar they are to the one being tested in terms of parameters and functionality. Another conclusion that can be drawn, based on the just cited analysis, is that the resistance of individual electronic components to a pulse of an external electric field is very high, much higher than the real field of a HEMP. The problem is that this

Bibliography | 109

field induces a significant voltage in the long wires and cables, and, since they are connected to the electronic equipment, they pose a serious danger. Thus, the efforts to protect electronic equipment from HEMP should be aimed at prevention of highvoltage pulses flowing directly into it via the connected cables.

Bibliography [1] Report of the Commission to Assess the Threat to the United States from Electromagnetic Pulse (EMP) Attack. Critical National Infrastructures. 2008. [2] Yee J. H., Orvis W. J. Martin L. C. Theoretical Modeling of EMP Effects in Semiconductor Junction Devices. Report AFWL-TR-82-91, Air Force Weapon Laboratory, Kirtland Air Force Base, 1983. [3] Camp M., Garbe H. Susceptibility of Personal Computer Systems to Fast Transient Electromagnetic Pulses. IEEE Trans. Electromagn. Compat., 2006, Vol. 48, No. 4, pp. 829–833. [4] Camp M., Garbe H., Nitsh D. Influence of the Technology on the Destruction Effects of Semiconductors by Impact of EMP and UWB Pulses. IEEE International Symposium on Electromagnetic Compatibility, Minneapolis, USA, 19–23 Aug. 2002, pp. 87–92. [5] Wunsch D. C., Bell B. R. Determination Failure Levels of Semiconductor Diodes and Transistors Due to Pulse Voltages. IEEE Trans. Nucl. Sci., 1968, Vol. 15, No. 6, pp. 244–259. [6] Nitsh D., Camp M., Sabath F., Haseborg J. L., Garbe H. Susceptibility of Some Electronic Equipment to HPEM Threats. IEEE Trans. Electromagn. Compat., 2004, Vol. 46, No. 3, pp. 380–389. [7] Van Keuren E. Effects of EMP Induced Transients on Integrated Circuits. IEEE International Symposium on Electromagnetic Compatibility, San Antonio, TX, USA, 7–9 Oct. 1975. [8] Jenkins C. R., Durgin D. L. An Evaluation of IC EMP Failure Statistic. IEEE Trans. Nucl. Sci., 1977, Vol. 24, No. 6, pp. 2361–2364. [9] Moore G. E. Cramming More Components onto Integrated Circuits. “Electronics”, April 19, 1965, pp. 114–117.

5 Electronic components for HEMP protection system 5.1 Testing of low-power protective components under the low pulse voltages The Western technical literature categorizes the protected equipment as the front-door coupling devices and back-door coupling devices. The front-door coupling devices are immediately exposed to the external electromagnetic radiation. The back-door coupling devices suffer from the electromagnetic radiation entering through the “backdoor” (inadequate screens, gaps, input, etc.). Since the substation equipment is located in the cabinets installed inside the buildings weakening the external radiation, only the back-door coupling devices are considered in this chapter. In the lecture of Prof. P. K. Skorobogatov, see Figure 5.1, we found very interesting information regarding the ability of some commonly available protective components to ensure the adequate protection.

Figure 5.1: Peter Skorobogatov, Doctor of Science, Professor of Automation and Electronics Department, Moscow Engineering Physics Institute, author of more than 200 scientific works and inventions, including more than ten learning aids, and the title slide of his lecture.

However, the lecture contained odd statements and this initiated a check of the originality of the material presented by Prof. Skorobogatov as his own research. Our suspicion was verified because it turned out that the lecture of the honored professor, who specialized in the field of pulse overvoltage protection, was a complete copy of the graduation paper by Tony Nilsson, a student of Linköping University, see Figure 5.2, and it contained no references to that work, as is expected. Later, Tony Nilsson published other works as an employee of SAAB Communication and the Swedish Defense Research Agency (FOI). https://doi.org/10.1515/9783110639285-005

5.1 Testing of low-power protective components under the low pulse voltages | 111

Figure 5.2: Graduation paper by Tony Nilsson.

So, let us discuss some interesting results of the research studies made by Tony Nilsson. First of all, let us give the precise definition to the parameter called “protection component response time”. According to the cited work, protection component response time is the time period from the moment of the overvoltage pulse applied to the protection component until the moment when the pulse is chopped. See Figure 5.3. The graduation paper was devoted to the evaluation of the response time of a standard surface-mount varistor type EV18N0402L (breaking voltage 46–60 V, max clamping voltage 110 V, capacitance approx. 20 pF). As we see from Figure 5.4, the varistor’s voltage amplitude varies from the applied pulse amplitude (even of the shortest pulse) and decreases to its actuation voltage level. Immediately, the amplitude of this voltage is significantly lower than the value supposed to remain on the actuated varistor (residual voltage)—the so-called clamping voltage (110 V). This could be explained by the fact that the varistor clamping voltage is limited by the pulse of current with a certain amplitude flowing through it. For

112 | 5 Electronic components for HEMP protection system

Figure 5.3: Graphic depiction of the definition of protection component response time.

Figure 5.4: The time of response of a standard varistor type EV18N0402L to the short pulse of applied voltage.

varistor type EV18N0402L, this current value equals to 10 A. If the test results show that the clamping voltage amplitude is significantly lower than the nominal value, this could happen because the current that flowed through the varistor upon the test was significantly lower than the nominal value. The phase shift between the applied pulse and varistor pulse can be explained by the capacitance and inductance of the test scheme and the varistor itself. While these values are low, they still impact the nanosecond-range pulses. Within such interpretation, the protection component response time definition, given in the work of T. Nielsen, is not quite correct. For back-door coupling devices, it must be considered that the overvoltage pulse affecting the sensitive equipment enters through the cables characterized by a much higher inductance and capacitance than the parameters existing during the evaluation. Therefore, it is clear that the real input HEMP pulse will not be that steep and short. Also, the varistors protecting the sensitive inputs of the electronic equipment

5.1 Testing of low-power protective components under the low pulse voltages | 113

installed in the cabinets will be much larger and more powerful than the miniature surface-mount device (SMD) used during the test. Moreover, their capacitance will be higher. The more powerful modular varistors, see Figure 5.5, have higher capacitance ensuring the partial absorption of the pulse energy. In practice, it is important to arrange such varistors as close as possible to the protected device and connect them to the device with short conductors of minimum inductance.

Figure 5.5: Certain types of the powerful modular varistors.

Figure 5.6: Gas-discharge tube-arrester type CG75L response to the short pulse of the applied voltage.

Also, the test included the evaluation of the 75 V-break-down voltage gas-discharge tube (GDT) arrester type CG75L operation (according to Tony Nilsson’s work), see Figure 5.6. As we see in Figure 5.6, during Tony Nilsson’s test, the 75-V break-down voltage GDT was not actuated under 130 V. Accordingly, Tony Nilsson concluded that GDT were not suitable for the protection of the equipment against the short pulses due to their insufficient quickness. While it can be agreed that the GDT are rather troublesome when it comes to protection against HEMP, it should be noted that Tony Nilsson’s evaluation of the tested

114 | 5 Electronic components for HEMP protection system

Figure 5.7: The breakdown voltage of the certain GDT arresters (from manufacturer’s data sheet), V.

component parameters was wrong. For instance, while the CG75L arrester’s nominal voltage equals 75 V, it has much higher break-down voltage values with a fast rise time of the pulse, see Figure 5.7: up to 400 V (upon rise time 100 V/µs) or even 650 V (upon rise time 1000 V/µs). Undoubtedly, due to such break-down voltage values, the arrester was not able to respond to the steep and short pulse of 130 V. Therefore, it was not due to the long actuation time of the GDT (while true), but it resulted from the significant increase in the break-down voltage amplitude, accompanying the increasing pulse rise time. Poor awareness of this feature of the GDT resulted in the incorrect testing. While such an error is excusable when it comes to a student, it is a major mistake for the professor of the Moscow Engineering Physics Institute to present that data within the educational process. As for the possibility of using such components for HEMP protection (considering that its rise time is much fast than 1 kV/µs), it is highly questionable because, despite their long response time, their actuation threshold is too big to ensure the reliable protection of the electronic equipment, especially if operating under 220 V/250 V or 380 V/400 V. Other components evaluated in Tony Nilsson’s work were not sufficiently powerful to be used for the power-system electronics protection. Thus, they are not considered here.

5.2 Testing of low-power protective components under the high pulse voltages The INHA University and Agency for Defense Development of South Korea [1] published another interesting article discussing the comparative effectiveness of various HEMP protection devices. Authors of the article evaluated the operation of the GDT, Metal-Oxide Varistor (MOV), and TVS-diode. However, they have been far too dismissive of the test condi-

5.2 Testing of low-power protective components under the high pulse voltages | 115

tions. In particular, they considered the significant increase in the GDT break-down voltage upon the applied pulse with the fast rise time and consequently used the test pulse generator with an amplitude exceeding 1,000 V. Nonetheless, the test results were the same as Tony Nilsson’s: the GDT did not operate and thus provided no protection for the unit, see Figure 5.8. At the same time, its voltage was absolutely equal to the output voltage of the test pulse generator with a slight phase shift resulting from the connection capacitance and inductance.

Figure 5.8: GDT tube-arrester response upon the applied pulse of a high amplitude.

Nevertheless, under the adequately high applied voltage, this was caused by the excessively long response time of the arrestor to the applied pulse. This is another testament to the fact that the GDT arrestor is not suitable for protection against HEMP. The graphs representing the results of varistor type LVSL20260C (this type is indicated in [1]) test are shown in Figure 5.9. Its breakdown voltage is 30.9–40.9 V, the clamping voltage is 67 V and the capacitance is 300 pF. Finally, Figure 5.10 graphically represents the results of TVS-diode test. A lowpower TVS-diode type SMAJ15A was used for that test, see Figure 5.10. Its breakdown voltage was 16.7–18.5 V, the clamping voltage was 24.4 V and capacitance was 300– 400 pF approximately. The comparison of oscillograms of the varistor’s and the TVS-diode’s voltage (see Figures 5.9 and 5.10) confirms that the TVS-diode’s voltage amplitude reaches 600 V, while the varistor’s voltage amplitude does not exceed 400 V. Based on this observation, the authors of the article [1] conclude that the varistor is more advantageous than

116 | 5 Electronic components for HEMP protection system

Figure 5.9: Varistor type LVSL20260C response to the applied voltage pulse.

Figure 5.10: TVS-diode type SMAJ15A response upon the applied voltage pulse.

the TVS-diode, regardless of the fact that such values (600 and 400 V) do not correspond to either th breakdown voltage or clamping voltage of the tested devices, and exceed tenfold both the breakdown voltage and clamping voltage. This alone should get the authors’ attention, but it did not. The issue is that the first voltage surge on the protection component (such as the varistor or TVS-diode) with maximum amplitude appears not to depend on the protec-

5.2 Testing of low-power protective components under the high pulse voltages | 117

tion component response time exactly at its moment of action (circled on the oscillograms), after the relatively equal delay both for the varistor and the diode suppressor. This delay is attributable to the component capacitances and the capacitance ability to absorb the pulse energy. In addition, the value of the amplitude of first pulse voltage on these components is not related to their responsiveness, breakdown voltage or clamping voltage. It relates only to the capacitance absorption ability. To compare, examine the GDT voltage oscillogram found in Figure 5.8. As we can see, the GDT voltage amplitude is very similar to the pulse generator voltage amplitude. It is attributable to the fact that the GDT capacitance (and its absorbing ability) is a hundredfold lower than that of the varistor or suppressor. In reality, the protective component capacitance can be fully charged with the operating-circuit voltage containing of this component, thus eliminating any capacitance absorbing effect. Besides, there are circuits (especially high-frequency circuits, such as telecommunication systems) where the high capacitance of a protection component is absolutely unacceptable. Thus, in the context of protection against the short voltage pulse with the fast rise time (such as HEMP pulse), the varistor has no significant advantage over the TVS-diode, so the findings shown in [1] are misinterpreted. Additionally, TVS-diodes (and not varistors) are widely used to protect the sensitive electronic circuits from the static discharges characterized by the fast rise time such as HEMP’s. According to the analysis of other technical publications providing the comparative evaluation of the varistor and the TVS-diode’s ability to ensure the adequate protection against the short voltage pulses of nanosecond range, TVS-diodes are preferable due to the faster response. For example, in [2], TVS-diodes are called the fast-response protective components and the varistor slow-response protective components. In [3], TVS-diodes are called the fast-response protective components and the varistors medium-fast protective components. The work of [4] attributes the fantastic data to the TVS-diode response rate of 0.01 nanoseconds and states that varistors respond approximately 50–100 times slower. In [5], it is stated that TVS-diodes are much faster. It is confirmed in [6]: Varistors respond in 80 µs, while the TVS-diode reacts in 5 nanoseconds. However, there are some unpublished reports of experiments on the suitability of varistors to ensure the HEMP protection that indicate the successful results of such experiments in defiance of numerous confirmations of their insufficient response rate. Results of all those research studies depend heavily on the actual test conditions because even the small piece of wire or the long component terminal has a great impact on the parameters of the pulse applied to the protective component, as well as on its ability to ensure adequate pulse protection for the equipment. It should be considered that if the protection components are installed in the standard cabinets and connected to the inputs of the protected equipment, with the pieces of wires characterized by the inductance significant for such a short pulse, the influence of the in-

118 | 5 Electronic components for HEMP protection system ductance becomes inevitable and must be considered in the comparative evaluation of performance of various types of protection components. Due to such ambiguity and the lack of the definitely confirmed data, we made our own research.

5.3 Testing of powerful protective components under conditions close to reality Under actual operating conditions of the industrial electronic equipment located inside the metal cabinets with long cables connected to its inputs and outputs, the parameters of circuits exposed to HEMP are significantly different from those existing under the isolated laboratory conditions. Thus, the tests were performed on the model in some way corresponding to the real conditions, see Figure 5.11.

Figure 5.11: The appearance of the model with the tested components installed and the resultant test pattern.

5.3 Testing of powerful protective components under conditions close to reality | 119

During the model test, the tested protective component (varistor—MOV and suppressor—TVS) and the length of the conductor (0.1 m and 1.0 m) were changed. The varistor type B72220S0600K101 was tested, with the rated voltage of 60 V (85 V DC) and the clamping voltage of 165 V, the capacity of 3600 pF and the equal in power TVSsuppressor type PTVS10-076-TH, with a breakdown voltage of 85–95 V, clamping voltage of 140 V, and the capacity of 5600 pF. The model included the standard terminal, ordinary insulated wire and printed board. The protective components (varistors and TVS-diodes) were installed on this board. Obviously, such an arrangement of the model makes the high-frequency parameters (capacity, inductance, wave impedance) far from perfect and do not correspond to the pulse-generator output and the oscilloscope input characteristics. It was impossible to simultaneously record the signal sent by the generator and the signal remaining on the protective component with the oscilloscope, so as to assess the properties of the protective components using both signals and to compare the signals as planned. Thus, during the test, the calibration pulse was initially recorded after the protective component was unsoldered and removed. Then, the protective component was returned, and the signal was repeatedly recorded without any changes to the circuit. The recorder oscillograph traces are depicted in Figure 5.12. The calibration pulse sent to the model with the protective component removed kept the high rise time within the nanosecond range, while the pulse width increased up to hundreds of nanoseconds. Both tested components (MOV and TVS) cut the inputpulse amplitude down to the level approximately equal to their breakdown voltage. Upon that, the pulse amplitude rise time of the components changed significantly and decreased by a factor of five, (approximately) probably under the influence of the capacity of the protective components. Figure 5.13 shows the results of the test of the protective components when the long wire is connected to the input. As we see in the records, the calibration pulse rise time did not changed, while the protective components pulse rise time decreased even more compared to the short-wire arrangement. As before, both protective components were able to respond in proper time and to limit the input-pulse amplitude. Compared to the previous test, the voltage-limiting level was a little higher due to the increase of the input-pulse amplitude and, thus, of the current flowing through the protective components after their breakdown. Finally, the varistor with a long wire was tested, see Figure 5.14. The tests were performed with the test-pulse amplitude increased up to 2 kV. The resulted oscillograph trace demonstrates that the varistor clamping voltage is significantly lower than the applied pulse amplitude (2 kV), meaning that the varistor successfully responded and catted that pulse. However, it is obvious that the actual varistor clamping-voltage amplitude significantly exceeded the nominal reference value of 165 V for the first time. What does that mean? To answer this question, we need to understand the nature of the clamping voltage existing on the pulse-protective component. This characteristic is indicated by the manufacturer. Logic suggests that it must be a voltage remaining

120 | 5 Electronic components for HEMP protection system

Figure 5.12: Oscillograph traces recorded for the test of the two protection component types: TVS– diode (TVS) and varistor (MOV) on the model with the short conductors (0.1 m long); RT-pulse rise time.

5.3 Testing of powerful protective components under conditions close to reality | 121

Figure 5.13: Oscillograph traces recorded in the test of the two protection component types: suppressor (TVS) and varistor (MOV) on the model with the long conductors (0.1 m long); RT-pulse rise time.

122 | 5 Electronic components for HEMP protection system

Figure 5.14: Oscillograph traces of varistor operation under the test pulse of 2 kV amplitude. FWHM (Full Width at Half Maximum)—width of the pulse at the middle of the amplitude.

on the protective component after its breakdown. Thus, this is the voltage applied to the equipment protected by this component. Such is indeed the case. But why did the clamping voltage significantly exceed the value shown in the specification during the test? Since varistor properties are far from perfect, the manufacturers use a trick and indicate that the clamping voltage occurs under the much lower current (1 % or less) compared to the varistor design value in their specifications, see Table 5.1. Additionally, since the voltage drop on the protective component depends on the current flowing through it, it is clear that the clamping voltage should be low for the low current value. During the test just described, the current pulse flowing through the varistor upon the applied voltage of 2 kV exceeded the current value used by the manufacturer to measure the clamping voltage. Thus, the real clamping voltage of the varistor exceeded the rated value. Nevertheless, it means that, under the actual operating conditions with an unknown voltage amplitude and the current flowing through the varistor after its breakdown, it is not possible to define the voltage remaining on the varistor and on the protected equipment itself! Under the pulse currents of several kA (used as a nominal value when designing the powerful varistors), the clamping voltage on the varistors may reach several kilovolts! Under the impact of intensive HEMP, the effectiveness of the protection is hardly predictable without any relation to its response rate.

5.3 Testing of powerful protective components under conditions close to reality | 123 Table 5.1: Current values used by manufacturers for clamping voltage measurement. Type

Peak of Surge Current for 8/20 µs Pulse Waveform, A

Peak Current Used at Max. Clamping Voltage Measurement for 8/20 µs Pulse Waveform, A

V5E50P MOV-20D680K V20E50P B72225S4301K101 V25S300P B722240B0321K101 V321BA60

800 2,000 10,000 20,000 22,000 40,000 50,000

5 20 100 150 100 300 200

TClamp2512N SP03-6 AK1 (series) AK3 (series) PTVS-3 (series) PTVS-10 (series) AK15 (series)

120 150 1,000 3,000 3,000 10,000 15,000

100 100 1,000 3,000 3,000 10,000 15,000

Percent of Peak Current Used for Clamping Voltage Measurement, % MOV 0.6 1 1 0.8 0.5 0.8 0.4 TVS 83 67 100 100 100 100 100

TVS-diodes are free from such disadvantages since, with some minor exceptions (special diode types), the manufacturers’ specifications show the value of the clamping voltage existing under their rated maximum pulse current (see Table 5.1). By the way, the use of such an unequal approach to varistors and TVS diodes is fixed in the standards [7, 8] so it is impossible to find fault with the manufacturers in this situation. On the other hand, these standards were written more than 30 years ago, when only low-power TVS diodes and Zener diodes based on the avalanche effect were on the market. Apparently, the time has come to correct these standards in connection with the advent of new technologies and new elements in order to bring the parameters of various types of protective elements to a “common denominator” that is understandable for users. Even the group of in-parallel varistors, see Figure 5.15, does not help [9]. It confirms the definitive advantage of TVS-diodes, but they cost almost a hundred times exceed the cost of the varistors. Will it work if we try to limit the current flowing through the varistor upon the breakdown in order to reduce the varistor voltage drop? The solution should not be purely theoretical, rather it should be suitable for the real electronic apparatus cabinets and cover the whole HEMP frequency range. Such a solution exists—it consists of split snap-on ferrite beads [10] (that do not require cable cutting up) built into the latched plastic frame installed on the multicore control cables entering the cabinets containing electronic equipment, see Figure 5.16.

124 | 5 Electronic components for HEMP protection system

Figure 5.15: Volt-amps diagram of low-voltage protective components: TVS-diode and varistor (MOV) and groups of 6 in-parallel varistors (MOV) [9].

Figure 5.16: The arrangement of ferrite elements (FE) in the electronic equipment cabinet. Left— before installation, right—after installation.

Selected ferrite beads were mounted on the model just described and tested together with a varistor, see Figure 5.17.

5.4 Conclusions | 125

Figure 5.17: Design of the test-bed used to test the interoperation of a varistor and the set of six ferrite beads.

The tests show a rather significant positive effect of ferrite beads on varistor performance, see Figure 5.18.

5.4 Conclusions 1.

The test shows that due to their responsiveness, both varistors and TVS-diodes based on the avalanche effect are capable of being used for providing the basic protection of the cabinet-enclosed industrial electronic equipment against HEMP. 2. Powerful TVS-diodes are higher quality and more reliable in terms of protection than varistors. However, since varistors are less expensive, an alternative is needed when it comes to wide application. 3. Such an alternative solution may be realized with the combination of varistors and ferrite beads, and such ferrite beads must be connected in front of the varistors on the cable at its entrance to the cabinet. 4. Additional ferrite beads installed on the control cable ensure the raising edgesteepness reduction and reduce the amplitude of the current flowing through the protective component after its breakdown. This makes it possible to reduce the voltage drop on such a component, thus significantly improving the effectiveness of the varistor providing the equipment protection.

126 | 5 Electronic components for HEMP protection system

Figure 5.18: The voltage on the varistor (MOV) without ferrite beads (top) and with ferrite beads (bottom).

Bibliography [1]

Han S. M., Huh C. S., Choi J. S. A Validation of Conventional Protection Devices in Protecting EMP Threats. Prog. Electromagn. Res., 2011, Vol. 119, pp. 253–263.

Bibliography | 127

[2] [3]

Protection Technology Comparison Chart. Protek Devices, 2013. TND335/D. Transient Overvoltage Protection, ON Semiconductor, Semiconductor Components Industries, LLC, 2008. [4] AN 1826/0104. Transient Protection Solution: Transil™ Diode versus Varistor /Bremond A., Karoui C. STMicroelectronics. [5] Goldman S. J. Selecting Protection Devices: TVS Diodes vs. Metal-Oxide Varistors, Power Electron., June 1, 2010. [6] Catalog EMP Connectors Inc., USA, California, 2005 (www.conesys.com/catalogs/EMP/ empcat1.pdf). [7] IEEE Std. C62.33-1982. Test Specifications for Varistor Surge-Protective Devices, 1882. [8] IEEE Std. C62.35-1987. Test Specifications for Avalanche Junction Semiconductor Surge-Protective Devices, 1987. [9] Howell T. Comparing Circuit Protection Technologies for 48 V DC in High Surge Environments. Protection Engineers Group Conference, Dallas, March 14–16, 2017. [10] Gurevich V. The Problem of Correct Choice of Ferrite Beads. Electr. Eng. Electromechanics, 2016, Vol. 2, pp. 71–75.

6 External protection of power systems’ electronic equipment from HEMP 6.1 Introduction As a rule, electronic equipment in power systems, such as digital protection relays (DPR), programmable logic controllers (PLC), automation systems, telecommunication, etc., is mounted in special control cabinets and connected to a grounding system. The cabinets are located in a control room (relay room), which is situated in the substation building. The same “layer-based” principle should be applied to protect this equipment from the devastating impact of an electromagnetic pulse caused by a highaltitude nuclear explosion (HEMP). Among these “layers”, there are, for instance: the electronic equipment’s grounding systems; EMP filters installed in the inputs of this equipment; and the principles and means of control cabinets’ protection [1–3], which are described in another chapters. This chapter addresses the issues of protecting the substation’s building and the control room, which is referred to here as “external protection”. Why is another “layer” – external protection – necessary, in addition to those “layers” discussed earlier? There are several reasons for this. Firstly, it can be difficult to provide efficient HEMP protection with only one “layer”. Sometimes it is even impossible because HEMP comprises both an electric field in the air with a very high level (up to 50 kV/m at the ground surface), a highvoltage pulse applied directly to the input of the electronic equipment, and also powerful interference at the grounding system, which penetrates directly into sensitive electronic elements, etc. Secondly, individual electronic devices (DPR or PLC) are not insulated from another devices and systems at any facility in the electric power industry (such as a substation or power plant). They are connected to many other devices, which are often far away from each other. Thus, in this situation, efficient protection of each individual electronic device can be very expensive from a practical point of view, compared to “external protection” of the entire room, or even a building by using special construction materials. The aim of this article is to discuss available options for providing HEMP protection for entire buildings or rooms and an evaluation of the efficacy of such protection.

6.2 Analysis of capability of conventional building materials to weaken electromagnetic emission Practically significant experimental research into the capability of various construction materials to weaken the electromagnetic emission was conducted in the 1990s https://doi.org/10.1515/9783110639285-006

6.2 Analysis of capability of conventional building materials | 129

by The American National Institute of Standards and Technologies. In 1997, they published a full report with the findings of their research [4]. Unfortunately, the frequency range that was explored in their research, does not cover the real range of HEMP frequencies (100 kHz–100 MHz). However, their findings (Table 6.1) are sufficient for general evaluation of the situation. Table 6.1: Efficiency of electromagnetic emission weakening by common construction materials. Building materials Concrete not rebar reinforced, thick: 102 mm 203 mm 305 mm Reinforced concrete wall, thick 203 mm: rebar dia. 19 mm, dist. between rebars—70 mm rebar dia. 19 mm, dist. between rebars—140 mm Concrete blocks with hollow cavities, thick: 203 mm 406 mm 609 mm Regular dry lumber, diameter: 38 mm 76 mm 152 mm Bricks, thick: 1 brick (89 mm) 2 bricks (178 mm) 3 bricks (267 mm)

Attenuation, dB 500 MHz 1 GHz 7–11 17–25 31–45

11–14 22–28 33–45

26 23

30 27

8 13 26

12 17 28

2 1.5 4.5

3 3 6

0 3.5 4

3.5 5.5 7

Reinforced concrete is expected to provide the best results, though the improvement (compared to non-reinforced concrete) is not as significant as expected (note that 20 dB correspond to 10 times the weakening of emission amplitude). Obviously, any changes of parameters of both concrete and its rebar (Figure 6.1) can significantly impact the shielding quality of a building. How? This question is addressed in many research papers [5–8]. Tables 6.2–6.7 show some general data that reflect the impact of changes in various parameters of the most common construction material on its shielding properties. The findings represented in Tables 6.1–6.6 can be used to analyze and evaluate the influence of one or the other concrete’s parameter on the shielding efficiency and to select the most suitable (for specific conditions) means to improve shielding efficiency. A comparison of findings presented by various authors returned a fair match (differences do not exceed 15 %). At the same time, it is necessary to understand that the

130 | 6 External protection of power systems’ electronic equipment from HEMP

Figure 6.1: Reinforced concrete with various types of rebar. Table 6.2: Power attenuation value of concrete samples with varying moisture contents. Attenuation, dB (ratio) for frequency Moisture content, % 0.2 5.5 12 Ratio of attenuation change for spot frequencies and different moisture content 0.2 % to 12 % Average ratio of attenuation change by moisture content influence

200 MHz

500 MHz

1 GHz

3 (1.41) 11 (3.55) 18 (7.94) 5.6

4 (1.58) 18 (7.94) 32 (39.8) 25.2

4 (1.58) 20 (10) 35 (56.2) 8.75

13.2

Table 6.3: Reinforced concrete with mesh interval of 100 mm and various types of rebar. Attenuation, dB (ratio) for frequency Rebar diameter, mm 10 30 50 Ratio of attenuation change for spot frequencies and different rebar diameter 10 to 30 mm Average ratio of attenuation change at rebar diameter influence

100 MHz

200 MHz

500 MHz

1 GHz

25 (17.8) 40 (100) 70 (316) 18

20 (10) 35 (56.2) 60 (1000) 100

12 (3.98) 30 (31.6) 55 (562) 141

5 (1.78) 20 (10) 45 (178) 100

90

6.2 Analysis of capability of conventional building materials | 131 Table 6.4: Power attenuation of reinforced concrete of various thicknesses with mesh interval of 100 mm and rebar diameter of 20 mm. Attenuation, dB (ratio) for frequency Concrete thick, mm 600 1000 1500 Ratio of attenuation change for spot frequencies and different concrete thick 600 to 1500 mm Average ratio of attenuation change at concrete thick change 600 to 1500 mm

100 MHz

500 MHz

1 GHz

60 (1000) 60 (1000) 60 (1000) 1

25 (17.8) 30 (31.6) 40 (100) 5.6

15 (5.62) 22 (12.6) 30 (31.6) 5.6

4

Table 6.5: Power attenuation of 1000 mm thick reinforced concrete with 12 % moisture with one or two layers of rebar with mesh interval of 100 mm and rebar diameter of 20 mm. Attenuation, dB (ratio) for frequency 1-layer rebar 2-layer rebar Ratio of attenuation change for spot frequencies at transition from 1-layer rebar to 2-layer rebar Average ratio of attenuation change at transition from 1-layer rebar to 2-layer rebar

100 MHz

500 MHz

1 GHz

70 (3,160) 110 (316,000) 100

55 (562) 65 (1,780) 3.2

40 (100) 42 (126) 1.26

35

Table 6.6: Power attenuation of strengthened reinforced concrete with varying distances between two layers of rebar with a mesh interval of 100 mm and rebar diameter of 20 mm. Attenuation, dB (ratio) for frequency Distance between two layers of rebar, mm 30 100 200 Ratio of attenuation change for spot frequencies for various distances between two layers of rebar 30 to 200 mm Average ratio of attenuation change for various distances between two layers of rebar 30 to 200 mm

100 MHz

500 MHz

1 GHz

20 (10) 25 (178) 25 (178) 17.8

27 (22.4) 40 (100) 40 (100) 4.5

45 (178) 58 (251) 65 (1780) 10

10.8

findings are connected with a specific material in a specific environment. Real materials and real conditions of the use of these materials can differ significantly from those used in the experiment, which results in a significant change of shielding efficiency. Moreover, these changes can occur and gradually alter over time for the same material due to natural reduction of water content from that in the new concrete, which can significantly impact the shielding efficiency (Table 6.2).

132 | 6 External protection of power systems’ electronic equipment from HEMP

6.3 Composite construction materials with improved electrical conductivity Composite materials with improved electric conductivity (intended for electromagnetic shielding) were first developed and studied back in the 1970s (in other words, about 50 years ago) [6, 7]. Since that time, many new composite construction materials have emerged, which are mostly based on concrete with various additives to improve its electric conductivity and, consequently, its shielding capacity. These materials have been developed in China, India, USA, Russia and many other countries, and the results of their testing are published in many scientific papers, such as [8–16] and others. The following products are used as additives for concrete: conductive powder (mostly coal, graphite and metal), carbon filament, carbon nanotubes and short pieces of steel wire. The recipes of such composite construction materials have been patented in many countries (US patents: 2868659, 3207705, 5346547, 5422174, 6214454, 6503318, 6821336, 7578881, 8067084, 8617309, 8968461, 9278887 and others; patents of Russia: 2545585, 2345968, 2234175, 2405749, 2291130 and others; patents of China: 1282713, 1293012, 1298663, 1844025, 101030454, 1313410, 103979853, and others). The portion of graphite-carbon mixture can reach 25–35 %, and in some cases as high as 75 % of the total material weight. Clearly, this portion of graphite and carbon in a concrete mix determines: 1) the rather high price of the final product; and 2) reduces its mechanic strength. In Russia, it has recently become fashionable to use shungite as a conductive filler for construction materials. Shungite is a natural-mineral composite material, consisting of fine-grained crystal silicate particles in an amorphous carbon matrix, see Figure 6.2. Shungite is excavated from the Zagozhyn deposit (Republic of Karelia). Construction materials based on magnesium–shungite mixtures are produced in Russia by “Al-

Figure 6.2: Natural shungite.

6.3 Composite construction materials with improved electrical conductivity | 133

fapol” company (patent of Russia No. 2233255). Shungite mineral of the third group is used as construction material. This group represents a natural composite material, which consists of 26–30 % of carbon and 56–60 % of silicate particles. Being a construction material, shungite mixtures are applied on surfaces of shielded premises as a finishing layer. In order to improve shielding efficiency, it is possible to use multilayered shielding by combining a plaster compound with a metal mesh. According to developers, the price of plaster–shungite mixtures is comparable to the price of usual construction materials. Testing of radio-shielding properties of this material [17–19] in a broad range of frequencies was performed on a wall model made of plywood with a layer of shungite plaster “Alfapol ShT-1” applied onto it (15-mm thick), see Table 6.7. Table 6.7: Shielding efficiency of shungite plaster “Alfapol ShT-1”. Frequency range, MHz Attenuation, dB

3–30 6–10

30–300 8–14

300–1200 12–16

The dry mixture consists of natural shungite powder, active magnesium oxide in the form of caustic dead-burned magnesite (MgO) powder and a modifying additive. In order to prepare the plaster, a dry mixture is mixed with an aqueous solution of bishofite (MgCl2 ). The same type of plaster (RES-1) is produced by the Ukrainian company “Rudus” at 1.5 USD/kg of dry mixture together with a liquid polymer solution (Figure 6.3). Also manufactured is a protective concrete based on shungite. However, there is no adhesion between shungite particles and cement stone, and thus it is very difficult to obtain a homogeneous mixture when mixing shungite powder with cement. Consequently, these particles can be considered as conditional pores in the cement monolith. In connection with this, in order to prepare the concrete mixture, special

Figure 6.3: Shungite plaster RES-1 produced by the Ukrainian company “Rudus”.

134 | 6 External protection of power systems’ electronic equipment from HEMP magnesium cement (10 %) is used, which features higher adhesion to shungite (85 %). Magnesium cement is a type of construction binding material based on magnesium oxide. The latter is obtained from magnesite (a widely spread mineral: magnesium carbonate MgCO3 ) by heating at high temperatures with further grinding. Unfortunately, the cost of such concrete is not known, and its shielding capacity is described in promotional materials in an inaccurate and vague way. The University of Nebraska in the US developed an improved recipe of conductiveconstruction reinforced concrete (US Patents 8968461, 9278887), see Table 6.8. This is marketed as material specially designed for HEMP protection. Table 6.8: Composition of conductive reinforced concrete according to US patents 8968461 and 9278887. Components

Percent

Type I cement Silica fume Sand and gravel taconite sand Taconite aggregate Carbon particles (0.7-mm max particle size) Carbon particles (2-mm max particle size) Carbon particles (10-mm max particle size) Graphite powder (0.15-mm max particle size) Water Steel fiber (1 in.) Steel fiber (1.5 in.) Steel fiber (fine) steel shavings Superplasticizer (High-range water reducer)

20.9 % 1.1 % 20.1 % 23.5 % 2.0 % 5.6 % 8.0 % 1.2 % 10.1 % 3.3 % 2.7 % 1.0 % 0.4 %

Taconite, which accounts for 23 % of this concrete mixture, is a banded iron formation (one of the forms of iron ore). The recipe does not mention that the concrete is strengthened with several layers of steel mesh. The component analysis of this construction material shows that its recipe does not contain any revolutionary or breakthrough approach. However, the authors of these patents (the primary author was Professor of Civil Engineering at NebraskaLincoln University Dr. Christopher Y. Tuan) launched a powerful international advertisement campaign, where they advertised the unbeatable (in their opinion) features of their unique construction material (known as EMB3). They established a separate marketing company that promoted this material and obtained corresponding investments into the project. The main advantage of this material (according to the author) is lower material cost (by 60 %) compared to similar materials and higher concrete strength (by 28 %). According to the author, he managed to reduce the cost of the material to 300 USD per

6.3 Composite construction materials with improved electrical conductivity | 135

Figure 6.4: Erection of protected building using EMB3.

cubic yard (about 0.8 cubic meters), whereas the cost of common concrete is 120 USD per cubic yard. When constructing a building using this composite material (Figure 6.4), pouring of rebar is performed layer by layer as each layer contains a specific filler. The promotional material [20] suggests a unique feature for these construction materials (Figure 6.5). But this promotion does not show the specific (per cm of thickness) attenuation (attenuation rate) introduced by the material. In other words, it is unclear what the thickness and water content of material are (for which the chart is built). Indeed, the water content of concrete will significantly change during drying out, and simultaneously its resistance will change resulting in changes of the shielding capacity, see Table 6.9 [21]. Personal communication with the manager of Omni-Threat Structures, who represent this material in the market, revealed that the attenuation rate introduced by the material is about 10 dB/in (∼4 dB/cm), which is actually not so prominent. The representatives of the company do not possess the data on how conductivity and shielding efficiency change as a function of water content in the material. It should be noted that if a feature represented in Figure 6.5 is true, it means that the efficiency of electromagnetic emission shielding by new concrete is close to the shielding efficiency of HEMP-protected premises covered with copper sheeting, which is commonly used for internal lining (Figure 6.6).

136 | 6 External protection of power systems’ electronic equipment from HEMP

Figure 6.5: The feature of conductive EMB3 concrete (upper chart) being advertised and the requirements to shielding to ensure HEMP protection according to MIL-STD-188-125-1 standard. Table 6.9: Changes in concrete resistance with time [21]. Measured resistivity (Ohm⋅m) over time for material: Conventional concrete mix Conductive mix 1 Conductive mix 2

Day 10

2 Months

2 Years

Change ratio for 2 years

38 2.3 2.0

265 6.0 5.2

720 95 38

19 41 19

Figure 6.6: HEMP-protected premises lined with copper sheeting.

The copper’s high conductivity enhances its capability to reflect the falling electromagnetic wave, especially in the range of relatively low frequencies (100 kHz– 100 MHz), which are specific for HEMP (Figure 6.7).

6.3 Composite construction materials with improved electrical conductivity | 137

Figure 6.7: Efficiency of electromagnetic-field shielding by copper sheeting, which is used for lining protected premises.

The conductive concrete just mentioned also features high electric conductivity and thus, based on the feature mentioned in the advertisement material (see Figure 6.5); its shielding capacity is determined by its ability to reflect the electromagnetic wave. Usually, in order to check the premises’ shielding efficiency, a receiver of emissions with a directional antenna is located inside, while a transmitter with the second directional antenna is located outside facing the first antenna. The difference between the emitted and received signals determines the level of weakening, stipulated by the wall’s shielding effect (Figure 6.8).

Figure 6.8: Measuring the premises shielding efficiency.

138 | 6 External protection of power systems’ electronic equipment from HEMP The copper lined walls provide excellent results of measurements. A good result was also obtained when using the same method of measuring for a new American concrete. But here at least two questions arise: The first: common electronic equipment of power-industry facilities can resist the electromagnetic field’s density of 10 V/m in accordance with the request of general standards on electromagnetic compatibility. This means that in order to reach the level acceptable for electronic equipment, the field density should be reduced from 50 kV/m to 10 V/m (which is a factor of 5,000). But 100 dB (and even more) provided by this construction material over a broad frequency range means a reduction factor of 100,000! Do we really need this protection in practice, considering that the construction material being offered costs three times more than common construction material (and since the erection technology is more complicated, perhaps, it costs yet even more)? The second question is related to the material’s properties. To be precise, with its ability to weaken electromagnetic emission through reflection of the wave falling onto it. The problem is that this common method of HEMP protection is not acceptable to the electric power industry because the sources of electromagnetic emission, including HEMP, are located not only externally, but also in the interior of these facilities. The internal sources of emission are represented by cables running from the outside into the protected rooms. The number of these cables is too large to run each core through a special HEMP filter. The electromagnetic wave, emitted by a cable inside the shielded room, reflects from a lined surface of a wall (ceiling, floor) and falls on sensitive electronic equipment located within the protected premises. Moreover, since the reflection angles can be unpredictable, this can result in emission strengthening and concentration of electromagnetic energy in the area of sensitive electronic equipment. Unfortunately, all the common shielding-sheet conductive materials (e. g., films, fabrics, paints, varnishes, etc.) can reflect the electromagnetic wave falling onto them (this is actually the property that underlies their protecting ability). Therefore, they can be efficient in protecting a small group of electronic devices, located in an enclosed area with cables protected from over-emission interiorly, e. g., cabinets with electronic equipment and filters, mounted on cables coming from outside. These cabinets, having a capacity to reflect an electromagnetic wave falling onto them and can ensure highly efficient protection of electronic equipment, provided that there is no internal source of emission. What is the solution for large premises containing many different types of highlysensitive equipment and with many cables coming from outside? It seems that there is also a solution for this situation.

6.4 Materials absorbing electromagnetic emission There are many recipes for construction materials with radio-absorbing properties in [22–24]. Common drawbacks of all these materials are their low efficiency at low fre-

6.4 Materials absorbing electromagnetic emission

| 139

quencies (hundreds of kilohertz to hundreds of megahertz) and complexity of making mixtures and its high cost. But the recipes, based on ferrite powder and developed in Japan, that ensure high absorption at low frequencies, are far too expensive for broad use. It is known from the theory of spreading electromagnetic waves that porous materials contain a low index of EMP reflection due to the similar values of impedance of the material’s surface and the air surrounding it. Thus, porous construction materials are rather suitable as radio-absorbing materials. The Patent of Russia No. 2234175 describes radio-absorbing porous concrete, which consists of highly-porous glass or ceramic pellets, coated with ferrite and (or) conductive material, and ferrite and (or) conductive powder with an adhesive agent. The patent No. 102627436 suggests mixing of cement with porous pumice particles and adding technical carbon (1–2 %). The recipe of one of the best porous concretes is described in Patent of Russia No. 2545585, which suggests producing radio-protecting construction concrete based on cement, sand and porous pelleted filler. The radio-absorbing properties of concrete are based on adding carbon-containing filler, which constitutes a structured gel (up to 40 % of the cement’s volume), consisting of 5–10 % of aqueous solution of polyvinyl alcohol (51–63 %), sodium lignosulfonate (4–7 %), 25 % ammonia solution (9–12 %) and electrical carbon (24–30 %). The specific (per cm of concrete thickness) absorption described in the patent exhibits the adequate properties of this material (Figure 6.9), provided it is sufficiently thick. The Patent of Russia No. 1840794 suggests strengthening the radio-absorbing properties of concrete by modifying the clay pellets—a common construction material

Figure 6.9: Specific absorption of concrete according to Patent of Russia No. 2545585.

140 | 6 External protection of power systems’ electronic equipment from HEMP that is widely used as a concrete filler. In order to do this, the clay pellets are coated (by pouring) with two layers of a special suspension, with further drying at 80–90 degrees. The composition of the suspension is: water (81.5 %), sulfanol (1.9 %), soot (11.9 %), liquid glass (4.7 %). This composition of suspension ensures 50 % soot content in the final coating. According to the developer, the radio-absorbing clay pellets provide repeated re-reflection of a falling electromagnetic wave, thus resulting in its full absorption. The reflection index of such radio-absorbing material does not exceed 10 % the absorbing capacity of material exceeds 0.6 Wt/cm2 (unfortunately, without any reference to the frequency range). Concrete is not the only material featuring radio-absorption. There are other construction materials that feature the same properties. One of these materials is called Ceramopen™ (Figure 6.10) developed by the Russian company “Cerapen”. This material is based on glass-ceramic foamed ceramics that features attenuation of more than 30 dB, whilst the reflection from a flat surface (−13 dB) is not more than 5 %. Additionally, Ceramopen features low water absorption (less than 1 %) and good heat-insulating properties.

Figure 6.10: Radio-absorbing construction material based on foamed ceramics CeramopenTM .

Another form of foamed construction material with radio-absorbing properties is foam glass. The radio-absorbing properties of the foam glass are provided by the porous structure of the material and availability of a carbon-containing component, which is used as a gas-forming agent during foam glass production (Patents of Russia No. 2255058, 2255059, etc.). Compared to conventional absorbers, the foam glass materials feature such important benefits as mechanical strength and low weight (Figure 6.11). The world leader of foam glass production is Pittsburgh Corning Company in the US. It produces the foam glass product under the “Foamglass” trademark. Other large foam glass producers are Chinese Gansu Pengfei Insulation Materials Co. Ltd., Gomel Glass Factory in Belarus, ICM Glass in Russia and others.

6.4 Materials absorbing electromagnetic emission

| 141

Figure 6.11: Foam glass in the form of construction blocks and pellets.

The recipes, properties, production technologies and research findings for foam glass are described elsewhere ([25–27] and others). The price of foam glass panels with dimensions 600 × 450 × 50 mm produced by Pittsburgh Corning is about 30 USD per m2 . In order to ensure additional strengthening of the radio-absorbing properties of foam glass, zinc oxide and graphite, some other chemical elements [25] need to be added. Foam glass is produced by baking this mixture at 750°C for 30 minutes with further cooling. The obtained material demonstrated minimum reflection index—15.6 dB at 12.0 and 12.4 GHz frequency. It is recommended to control the absorbing capacity of the obtained material by adjusting the zinc oxide inclusion. The foam glass pellets (see Figure 6.10) coated with carbon-containing material, in addition to carbon contained in the pellets, is used for pouring between walls made of thin insulation material (up to single-face laminate). The obtained finishing panels will be attached to the walls of the protected premises. According to the Engineering and Marketing Center of “Vega” Corporation, who represent this technology, a panel like this provides specific attenuation of electromagnetic field of at least 6-dB per centimeter of the filling layer (at 4-GHz frequency). This enables a 1,000 reduction factor of the electromagnetic field at 5-cm thickness [28]. Pelleted foam glass without additional carbon coating costs about USD 140 per m3 . Unfortunately, “Vega” Corporation did not answer my question regarding the price of panels with pelleted foam-glass filling and a special coating. Ferrite is the most efficient radio-absorbing material for the frequency range corresponding to HEMP. Ferrite is a magnetic material that is a chemical compound of metal oxides (nickel ferrite (NiFe2 O4 ), zinc ferrite (ZnFe2 O4 ) and others), which are artificially produced as magnetic materials. Polycrystalline ferrites are produced according to ceramic technology. The properly shaped elements are made of ferrite powder, obtained from a mixture of initial ferrite-producing components and an adhesive agent. Then these elements need to be baked at 900–1,500°C in the air or in a special gas environment. The ferrite radio-absorbing material is produced in the form of finishing tiles, which will be attached to the interior walls of the premises. The standard tile size is 100 × 100 mm. Some manufacturers use these tiles to make lining panels of dimen-

142 | 6 External protection of power systems’ electronic equipment from HEMP

Figure 6.12: Ferrite lining panels and their standard specification.

sions 300 × 300 mm and even 600 × 600 mm. These panels feature good absorption properties in the required frequency range (Figure 6.12). These panels are produced by many companies, such as ETS-Lindgren, Samwha Electronics, Holland Shielding Systems BV, Fire-Rite Products Corp., Riken Environmental System Co., Global EMC UK Ltd, Cuming Microwave Corp., Amidon Associates, Inc., Pioneer EMC Ltd., Siepel. Unfortunately, the cost of such panels is rather high. For example, a panel of 300 × 300 × 5.5 mm costs more than USD 50, whereas one of 600 × 600 × 6.7 mm will cost as high as USD 220. One tile of 100 × 100 mm costs six–nine USD, depending on the manufacturer. This cost may prevent a broad use of this material for lining premises at power plants and substations. However, this material can be used as a radio-absorbing partition inside the cabinets with especially sensitive and critical electronic equipment.

6.5 Another method for depression of HEMP electromagnetic field strength inside the power industry facilities containing the electronics As just mentioned, electromagnetic emission of HEMP, which can penetrate through walls into the premises of power industry facilities with electronic equipment, is not the only source of emission affecting the electronic equipment. Another very powerful source is represented by a pulse electromagnetic field, re-radiated by hundreds of cables running into the premises with electronic equipment from the outside. These cables act as antennas, absorbing HEMP energy from a large area and delivering it to the premises containing electronic equipment. It is impossible to mount an expensive HEMP filter at each core of each multi-core cable if there are hundreds of them. However, it is possible to reduce the level of cable-emitted pulse without these filters. To do this, the cables need to be closed with electromagnetic screens right at the place

6.5 Another method for depression of HEMP electromagnetic field strength

| 143

Figure 6.13: Non-perforated metal cable trays preventing re-emission from cables entering the protected premises from outside.

Figure 6.14: Construction elements for enclosed solid cable trays.

of their entrance into the protected premises. It is recommended to use solid (nonperforated) cable trays as the screens (Figure 6.13). Both solid cable trays with lids and their construction elements are available in the market (Figure 6.14). The latter are made of aluminum, galvanized steel, steel with single-layer spraying of protecting insulation layer and galvanized steel with additional powder spraying.

144 | 6 External protection of power systems’ electronic equipment from HEMP Steel trays feature the best protecting properties at HEMP frequencies. The cost of galvanized steel (steel thickness of 1.2 mm) trays ranges from USD 2 to USD 15 per m of length (depending on the width and height). Steel trays with powder spraying of insulation coating are a bit more expensive. For example, a tray of 50 × 50 × 1.2 mm would cost about USD 4 per m, whereas a tray 200 × 50 × 1.2 mm would cost more than USD 7 per m. Plastic trays with metal coating are less costly, but they are not suitable for this purpose. Due to their very thin metal coating, they work exclusively at very high frequencies, which fall outside the HEMP frequency range. Reflection of an electromagnetic pulse, emitted by power cables located in the metal tray into the inner area of the tray, can result in an unfavorable effect of this emission onto adjacent control cables, located in the same tray and connected to electronic equipment. That is why power and control cables must be run in separate trays, while control cables must be split into groups and placed in separate sections of sectioned cable trays (Figure 6.15). If hard cable trays are inconvenient to use, flexible metal tubing can be used, especially for control cables with a small cross section (Figure 6.16).

Figure 6.15: Sectioned cable trays.

Figure 6.16: Flexible electromagnetic screens made of flexible metal tubing.

6.6 Reducing electronic equipment vulnerability to HEMP with architectural solutions | 145

6.6 Reducing electronic equipment vulnerability to HEMP with architectural solutions According to [29], modern buildings built from common construction concrete and metalized glass (widely used for UV and IR reflection) weaken electromagnetic emission by 13–14 dB more than the old buildings. This weakening can reach 20–25 dB in the frequency range from 800 MHz to 18 GHz. Unfortunately, this frequency range is far away from the HEMP range. However, in an earlier publication [30], it was discussed that for not very modern concrete buildings the electromagnetic field attenuation was specified as 20–25 dB at frequency range 1 to 100 MHz, and 10–15 dB at frequency 100 kHz. In the IEC 61000-4-36 [31] standard is shown (as the trend) the shielding ability of reinforced concrete buildings depending on the electromagnetic emission frequency, including for the frequency range corresponding to the HEMP, Figure 6.17. The data obtained from numerous measurements.

Figure 6.17: The trend of shielding ability of reinforced concrete buildings depending on the electromagnetic emission frequency (IEC 61000-4-36).

From Figure 6.17, it can be seen that, in the frequency range characterized the HEMP (105 –108 Hz), the attenuation level by the building can change 2,000 times from the beginning to the end of the frequency range! With such changes, accurate measurements of attenuation by a specific building will provide little useful information. Apparently, one can use some averaged data for the minimum attenuation characteristic of the maximum frequency. If the frequency of 10 MHz is taken as the maximum frequency (90 % of the energy of HEMP is allocated in the range 100 kHz–10 MHz), then the minimum attenuation that can be taken into account is 20 dB. The further reduction of penetrating electromagnetic emission can be obtained by placing the critical electronic equipment into the internal building premises having no exterior walls and windows (Figure 6.18).

146 | 6 External protection of power systems’ electronic equipment from HEMP

Figure 6.18: An example of floor layout with internal protected premises “A”.

Moreover, this protected room should not have any windows and should be equipped with metal doors with conductive rubber sealers along the perimeter. The inner surface of the walls in this room can be lined with some shallow thickness (3–4 cm) panels coated with radio-absorbing powder. Another option: the walls of these inner insulated premises should be made of radio-absorbing foam-glass blocks.

6.7 Conclusions None of the previously mentioned approaches to external protection of rooms full of electronic equipment from HEMP is ideal from both efficiency and cost-effectiveness standpoints. Therefore, the most sensible in our opinion is to ensure comprehensive protection based on joint use of inexpensive construction materials with the required properties, each of which provides at least partial protection. For example, the array of measures includes: concrete reinforced with metal rebar (as material that partially reflects an electromagnetic wave falling on its external surface); panels coated with inexpensive radio-absorbing powder (as inner lining); sectioned unperforated metal cable trays (as shields protecting the internal area of the premises from re-emission from cables entering from outside); placement of highly sensitive and critical electronic equipment in the internal premises of the building that have no windows but have metal doors, and mounting of this equipment in special metal cabinets with radioabsorbing ferrite partitions interiorly. In any case, the consumer must balance all the “pros” and “cons” of each option of external protection considering his/her financial position. This article should assist the consumer in making the optimum solution.

Bibliography [1]

Gurevich V. The Issues of Electronic Equipment Grounding at the Power Facilities. Int. J. Res. Stud. Electr. Electron. Eng., 2017, Vol. 3, No. 1, pp. 11–19.

Bibliography | 147

[2] [3] [4] [5]

[6]

[7]

[8] [9] [10] [11]

[12]

[13] [14] [15] [16] [17]

[18]

[19]

[20]

Gurevich V. Use of LC-filters to Protect Equipment from Electromagnetic Pulse: Is It Real Necessity or “Business as Usual”. Int. J. Sci. Res. Eng. Trends, 2017, Vol. 3, No. 4, pp. 85–89. Gurevich V. Accessible Method of Electronic Equipment Grounding at the Power Facilities. Int. J. Res. Stud. Electr. Electron. Eng., 2016, Vol. 2, No. 2, pp. 13–18. NISTIR 6055. Electromagnetic Signal Attenuation in Construction Materials, NIST Construction Automation Program, Report No. 3, National Institute of Standards and Technology, 1997. Hyun S. Y., Du J. K., Lee H. J., Lee K. W., et al., Analysis of Shielding Effectiveness of Reinforced Concrete Against High-Altitude Electromagnetic Pulse. IEEE Trans. Electromagn. Compat., 2014, Vol. 56, No. 6, pp. 1488–1496. Allen L., Walker W. F., Siarkiewicz K. R. An Investigation of the Electromagnetic Properties of Advanced Composite Materials. International Symposium on Electromagnetic Compatibility, IEEE, 1976, pp. 174–179. Casey K. F. Electromagnetic Shielding by Advanced Composite Materials. Final report AFWL-PD-76-l74 prepared by Kansas State University for Air Force Weapons Laboratory, Kirtland Air Force Base, New Mexico, 1997, 144 p. Chiou J.-M., Zheng Q., Chung D. D. L. Electromagnetic Interference Shielding by Carbon Fiber Reinforced Cement. Composites, 1989, Vol. 20, No. 4, pp. 379–380. Chung D. D. L. Electrically Conductive Cement-Based Materials. Adv. Cem. Res., 2004, Vol. 16, No. 4, pp. 167–176. Wen S., Chung D. D. L. Electromagnetic Interference Shielding Reaching 70 dB in Steel Fiber Cement. Cem. Concr. Res., 2004, Vol. 34, No. 2, pp. 329–332. Fei G., Bihua Z., Cheng G., Hailin C. Analysis of Shielding Effectiveness of Conductive Cement-based Materials in HEMP Environment. International Conference on Microwave and Millimeter Wave Technology, Nanjing, China, 2008. Ogunsola A, Reggiani U., Sandrolini L. Shielding Properties of Conductive Concrete against Transient Electromagnetic Disturbances. IEEE International Conference on Microwaves, Communications, Antennas and Electronics Systems (IEEE COMCAS 2009), Tel Aviv, Israel, 2009. Muthusamy S., Chung D. D. L. Carbon-Fiber Cement-Based Materials for Electromagnetic Shielding. ACI Mater. J., 2010, Vol. 107, No. 6, pp. 602–610. Krause A. P., Nguyen L., Yuan C. Y., Blasey J. D. Conductive Concrete: A Shielding Construction Material. J. Radiat. Eff. Res. Eng., 2013, Vol. 31, No. 1, pp. 151–157. Giri D. V., Tesche F. M. Electromagnetic Attenuation through Various Types of Buildings. Asia-Pacific Symposium on Electromagnetic Compatibility (APEMC), Melbourne, 2013. Venkatraman K., Tamizharasan V. Electrically Conductive Concrete. Int. J. Innov. Res. Sci. Eng. Technol., 2015, Vol. 4, No. 8, pp. 6738–6743. Research of Shielding of Electromagnetic Fields in the Frequency Range 10 kHz–1.2 GHz by Magnesium-Shungite Mixture “ALFAPOL-ShT-1”. Report. St. Petersburg branch-office FGUP “Scientific and Technical Center “Atlas” of FSS of Russia”. St. Petersburg, 2005. Baydin F. N., Nikitina V. N., Safronov N. B. Electrophysical Specifications and Radio-Shielding Properties of Magnesium-Shungite Composite Construction Materials. Proceedings of the Ninth Russian Scientific and Technical Conference on Electromagnetic Compatibility of Technical Means and Electromagnetic Safety, St.-Petersburg, 2006, pp. 292–294. Nikitina V. N., Lyashko G. G., Potseluyeva L. N., Radio-Shielding Properties of Magnesium-Shungite Construction Materials. Yearbook of Russian National Committee on Protection from Non-Ionized Emissions. RUDN, Moscow, 2006, pp. 109–113. EMP Shielding Shotcrete Structures for Multi-Threat Protection. Omni-Threat Structures, An American Business Continuity Group Company, Miami, 2017.

148 | 6 External protection of power systems’ electronic equipment from HEMP

[21] Yehina S., Qaddoumi N., Hassan M, Sewaked B. Conductive Concrete for Electromagnetic Shielding Applications. Adv. Civ. Eng. Mater., 2014, Vol. 3, No. 1, pp. 270–290. [22] Bowen G., Donghai D., LeFan W., Jiayu W., Rui X. The Electromagnetic Wave Absorbing Properties of Cement-Based Composites Using Natural Magnetite Powders as Absorber. Mater. Res. Express, 2017, Vol. 4, pp. 1–6. [23] Smirnov D. O. Composite Radar-Absorbing Materials Based on Ferrite-Magnetic Compositions. Thesis of Technical Science Ph. D., Moscow Power Engineering Institute, 2009. [24] Hongtao G., Shunhua L., Yuping D., Ji C. Cement Based Electromagnetic Shielding and Absorbing Building Materials. Cem. Concr. Compos., 2006, Vol. 28, pp. 468–474. [25] Chen K., Li X., Lv D., Yu F., Yin Z., Wu T. Study on Microwave Absorption Properties of Metal-Containing Foam Glass. Mater. Sci. Eng., 2011, Vol. 176, pp. 1239–1242. [26] Kazmina O. V., Vereshchagin V. I., Abiyaka A. N. Foam Glass Crystalline Materials Based on Natural and Artificial Raw-Materials. TPU, Tomsk, 2014, 246 p. [27] Dushkina M. A. Development of Compositions and Production Technologies of Foam Glass Crystalline Materials Based on Silica Raw-Materials. Thesis of Technical Science Ph. D. Tomsk Polytechnical University, Tomsk, 2015, 197 p. [28] Gulbin V. N., Kolpakov N. S., Alexandrov Yu. K., Polivkin V. V. Radio-Protecting Construction Materials. Knowl.-Intensive Technol., 2014, Vol. 15, No. 3, pp. 17–25. [29] Rodrigues L. I., Nguyen H. C., Jorgensen T. K., Sorensen T. B., Mogensen P. Radio Propagation into Modern Buildings: Attenuation Measurement in the Range from 800 MHz to 18 GHz. Vehicular Technology Conference (VTC Fall), IEEE 80th , 2014. [30] Buga N. N., Kontorovich V. Y., Nosov V. I. Elecromagnetic Compatibility of Radio-Electronic Equipment. Radio and Communication, Moscow, 1993, 240 p. [31] IEC 61000-4-36 Electromagnetic Compatibility (EMC) – Part 4-36: Testing and Measurement Techniques. IEMI Immunity Test Methods for Equipment and Systems, 2014.

7 The issues of electronic equipment grounding at the power facilities 7.1 Types of electromagnetic interference at power facilities Let us discuss the four major types of interference: – Conducted – Inductive – Capacitive – Electrostatic Conducted interference spreads through direct electric contact between electric circuits. Such interference can be subdivided as follows: – wire–earth interference—voltage is applied between each conductor and the earth. It is also known as asymmetrical, in-phase or common mode (CM) interference. – wire–wire interference—voltage is applied between discrete electric circuits or between the elements of the same electric circuit. It is also known as symmetrical, out of phase or differential mode (DM) interference. Unlike DM interference, CM interference usually does not result in electronic-equipment faults. However, it can make such equipment inoperative due to electric disruption of internal insulation (or p-n-junctions) in the electronic chips and the microprocessors. On the other hand, when electronic equipment is fully isolated from the earth, pulse interference and voltage spikes to earth (CM interference) cannot affect such equipment. This is similar to the situation when high voltage differential to earth does not kill the birds sitting on high-voltage wires. The DM interference is not connected to the availability or lack of grounding. As for inductive interference that spreads through the electromagnetic fields, it is known that efficient protection against this kind of interference is ensured by putting sensitive electronic equipment into closed metal enclosures (acting as the “Faraday cage”). It is also known that Faraday-cage grounding does not affect its ability to attenuate the inductive interference. The situation with capacitive interference is more complicated because this kind of interference spreads through the capacitance between the adjacent wires and closely spaced metal enclosures, as well as the above mentioned elements and the earth. Grounding of wire screens and metal enclosures (same as the capacitance reduction) enables eliminating the CM capacitive interference and has almost no effect on the DM capacitive interference. Electrostatic interference may appear due to static-charge accumulation on the enclosure insulated from the earth and periodic discharges to the earth. Grounding of https://doi.org/10.1515/9783110639285-007

150 | 7 The issues of electronic equipment grounding at the power facilities the enclosure prevents the static-charge accumulation. However, in order to eliminate electrostatic (or partially even capacitive) CM interference, low-resistance grounding is not necessary. It is sufficient to connect the enclosure to the grounding system through a high-resistance resistor (about 100 MOhm). Obviously, grounding of the electronic equipment (digital protective relays—DPR, for example) enclosure cannot solve all types of interference that come to the internal electronic elements via the cables and the wires connected to the DPR. Nonetheless, the internal “ground” of the electronic circuit of many DPR types that acts as a circuit of “zero (reference) potential” is usually connected to the metal enclosure and, consequently, to the external grounding system. Sometimes, the separate protective grounding (signal-reference subsystem) is used to protect sensitive electronics. Such a subsystem is connected to the common ground system at one point (which does not really change the idea of grounding). At the same time, it is supposed that various electronic devices interconnected through the data channels and the electrical cables share zero (reference) potential. This prevents the problem of highly sensitive electronic equipment experiencing electromagnetic interference from generating additional differences of potentials between the zero potential circuits, if they are not grounded. This is the approach according to [1]: “provides a reference potential for relaying and other equipment, and serves to protect equipment and personnel from power-frequency disturbances.” This is a common approach and practice to ensure the electronic equipment EMC. However, as mentioned in [2], “many interference problems occur because designers treat the ground as ideal and fail to give proper attention to the actual characteristics of the grounding system.”

7.2 Challenges of the conventional grounding systems It is known that the grounding system potential to earth can increase upon a lightning strike. At the same time, it is thought that, if all the electronic devices share a common potential of the grounding system, i. e., there is no difference in potentials between the reference potential circuits of various devices, such an increase in common potential and its difference from zero appearing simultaneously in all the devices cannot cause them to fault. The whole theory of grounding is based on this assumption, prescribing to maintain minimum resistance of the grounding system’s elements, as well as use of equipotential planes and other measures aimed at preventing the difference of potentials between zero-potential circuits separated from each other (and hence grounded at different locations), but interconnected through data channels and electric cables. In practice, this is realized for the majority of electronic devices used in the electric-power industry, for example, digital protection relays (DPR) placed in metal housings and mounting inside metal cabinets. A modern, properly designed (in accordance with traditional approaches) system of grounding of such devices is effectuated by a multipoint through the use of an equipotential surface, Figure 7.1.

7.2 Challenges of the conventional grounding systems | 151

Figure 7.1: Arrangement of multipoint grounding with equipotential plane.

The metal elements of relay protection cabinet can be used as an equipotential surface, see Figure 7.2. In many DPR designs, the functional modules located on separate printed circuit boards have special sections of printed conductors with an increased width and are covered with a layer of silver. When the board is installed in the housing, they make contact with special springs, ensuring the contact of these printed conductors with the grounded case of the DPR, Figure 7.3. These sections of the printed conductors are connected to the circuits of the internal “ground” of the electronic circuit, which acts as a “zero (reference) potential” circuit. Thus, high-sensitivity elements of electronic circuits are directly connected to an external grounding system. In this regarding, the processes running in a single electronic device during the potential rise in a zero potential circuit are not considered. Any electronic circuit contains many nonlinear elements, as well as capacitance and inductance elements connected to the zero potential circuit. Due to this, the voltage at various points of the circuit does not synchronously fully rise upon the circuit potential spikes. To better understand this process, imagine a plate supporting weights of different mass that are attached to this plate by means of springs of various rigidities. If we lift the plate gradually (i. e., gradually increase the potential energy), the potential energy of all the elements resting on this plate increases simultaneously. However, if we lift the plate abruptly, the elements do not simultaneously change their positions and potential energies. In addition, mechanical connections (if any) between such elements can be broken. Thus, the availability of an equipotential plane and the maintaining of zero difference between the zero potential circuits of different devices do not guarantee the absence of faults in highly sensitive electronic equipment. In real life, when electronic devices are located in spacious facilities, it is very difficult and sometimes even impos-

152 | 7 The issues of electronic equipment grounding at the power facilities

Figure 7.2: Arrangement of DPR multipoint grounding with equipotential plane 1—DPR in metal casings; 2—bounding strip; 3—metal cabinet’s element that performs the role of the equipotential plane.

sible to maintain zero difference of potentials between zero potential circuits upon a lightning strike, regardless of all the wiles and appreciation of the grounding system. According to [3], if there are many electric devices and separate pieces of electronic equipment are significantly distanced from each other and grounded at their particular locations, the high potential difference (up to 10 kV or even more) inevitably appears between the grounding points upon a lightning strike. High voltage between the distanced grounding points occurs due to a voltage dip on the grounding-system elements under the lightning currents and short-circuit currents flowing through such elements. If there are two DPRs significantly distanced from each other and their ports are interconnected through the Ethernet, see Figure 7.4, the voltage applies to such DPR nodes (telecommunication ports) least protected against the high-voltage pulses. Ac-

7.2 Challenges of the conventional grounding systems | 153

Figure 7.3: The printed circuit board of DPR with cleared sections of printed wiring (1 and 2), which contact the grounded casing by means of a special spring.

Figure 7.4: Connection diagram of two DPR (1 and 2) located at significant distances from each other with a uninsulated communication channel (twisted pair for Ethernet network).

cording to [4]: “the larger the territory of the protected site, the higher the problem probability”. Today, Ethernet networks realized with the twisted-pair multicore copper cable increasingly supersede the expensive fiber optic lines in the relay protection systems. This global tendency originated in an effort to cut the costs. Due to the low electrical strength of communication ports, they are not tested (unlike other DPR’s inputs and outputs) with high pulse voltage at all (IEC 60255-5, 60255-22-5) or tested with the low voltage (IEC 60255-22-1, 60255-22-4). According to [3], the level of pulse overvoltage appeared at the lightning discharges and applied to the electronics circuit exceeds the accepted levels for industrial electronics by several times. The basic standard regulating the electronic grounding systems [5] states: “Upon the lightning discharge or the emergency, the difference of potentials on the electronics and the grounded equipment may be very high and endanger the electric safety or result in the damage of the electronics”. It is quite clear that the existing grounding systems do not actually acts as equal energy surfaces with zero potential capable of ensuring the reliable operation of electronic circuits, and do not act as the return lines sending the signals back to the DPR. Moreover, all input and output circuits of DPRs (excluding the communication ports)

154 | 7 The issues of electronic equipment grounding at the power facilities are well-isolated from the ground and other electrical installations as follows: powersupply circuits are isolated with the inner power-supply transformer; analog inputs (currents and voltages) are isolated with the insulation of inner current and voltage transformers; logic inputs are isolated with optocouplers; the output circuit is isolated with the output electromechanical relay insulation. Additionally, the operability of the DPR’s internal electronic circuits is not related to the availability of the external grounding system. As for the protection of the sensitive electronic circuits of DPRs from external electromagnetic fields, with the metal enclosure acting as a so-called Faraday cage, it must be clear that the effectiveness of such protection is not related to the availability of the grounding system. Therefore, grounding of the DPR’s enclosure does not influence the level of the enclosure screening effect. On the other hand, if the noise signals enter the enclosed DPRs electronic circuits through the cables, how can the enclosure grounding prevent the noise effects (especially if the noises are in a differential mode)? The answer is obvious: It cannot! Even more, based on the just mentioned reality, it is fair to say that function grounding of the DPR’s enclosure only exacerbates an issue and reduces the relay protection immunity, as the real overvoltage levels applied through the grounding circuits to the various inner circuits (even if they are well-isolated) of the well-separated DPR can significantly exceed the acceptable levels, even without due regard to the weakened isolation of communication ports. Consequently, on the one hand, modern electronic devices used in the power industry do not require the availability of functional grounding. But, on the other hand, the electronics grounding is frequently considered as the basic means of protection from the electromagnetic noise mandatory to ensure its normal operation under the real operation conditions of electromagnetic noise. Just taking the title alone [6], it equates the electromagnetic compatibility (EMC) with automation and control-system grounding. According to [7]: “For many decades there have been many huge myths surrounding the word “ground”, both in circuit design and in EMC design”. This statement makes me wonder if the traditional view about the grounding of electronics is correct, even disregarding the HEMP issues.

7.3 Differences between lightning and HEMP The electromagnetic pulse (EMP) occurring when lightning (LEMP) hits grounded facilities (either a tree, tower, building or a lightning rod) is a natural phenomenon that has been known for as long as mankind exists. During the previous century, this phenomenon was well studied, and this made possible adoption of some methods and techniques that are widely used as protection from EMP. The destructive impact of both types of EMP on the objects is similar and is effectuated by two factors: the very high amplitude of the voltage pulse applied to the object and the high pulse current flowing through this object, as well as other secondary

7.3 Differences between lightning and HEMP

|

155

EMP outcomes related to these two factors, which are dangerous and damaging to electronic and electrical equipment. This similarity of destructive impact resulted in the fact that the lightning-protection methods and techniques, which have been properly researched and tested, started to be applied to HEMP. An example would be the fundamental principle of protection against the lightning: compulsory grounding of objects through the minimum possible resistance and the use of gas-discharge tubes and filters that divert the pulse’s energy to the ground. Is this really true? Are the specifications of LEMP and HEMP so similar that they allow identical methods and techniques of protection? The nature of the HEMP effect on the grounding system differs significantly from the effect of the lightning discharge by the spectral energy density, the time rise, the level of currents, the repetition of the pulses and the area covered [8]. LEMP is a local electric breakdown of the gas space (air) between two electrodes featuring a high potential difference (up to billion volts) between them: a cloud and the earth (or an object located on the earth and featuring the earth’s potential), Figure 7.5. As a result of this breakdown in air, a high-temperature (up to 30,000 °C) plasma conductive channel occurs through which the accumulated charge is discharged. The current of this discharge can reach 100–200 thousand amperes or more.

Figure 7.5: The area of lightning and high altitude nuclear explosion impact.

HEMP is a distributed electric field, which covers a large area and affects the objects located hundreds and thousands of kilometers away from the explosion epicenter due to spatial relocation of charged particles, e. g., electrons and ions that appeared as a result of complex physical processes, which are caused by the nuclear explosion in the atmosphere. Therefore, comparing these completely different physical phenomena with each other is not correct at all. The structure of the HEMP field is not uniform and can be conditionally split into three component parts: E1, E2 and E3. E1 is a very short pulse of the electric field

156 | 7 The issues of electronic equipment grounding at the power facilities shaped in 2/25 ns with the field gradient of 50 kV/m near the ground surface. E2 is a weaker pulse of the electric field with a duration from several to dozens of milliseconds. E3 is a very long low-voltage pulse of the electric field, which has to do with various processes in the ionospheric medium. This can last up to several minutes and causes the occurrence of significant quasi-DC currents in long-distance conductive media, such as rails, pipes, cables and wires. E1 is the most powerful, destructive and complex pulse (from the standpoint of protection) with both vertical and horizontal polarized parts. Thus, when using the term HEMP in this chapter, we mean E1 as it’s the main component. Compared to LEMP, HEMP is less powerful (Figure 7.6) and significantly shorter (Figure 7.7), but because it covers a large area and affects thousands of facilities simultaneously, it is more dangerous than LEMP. As just stated, both LEMP and HEMP can relocate over a distance and reach the ground surface in various ways. In the case of LEMP’s relocation through the ionized

Figure 7.6: Spectral density of LEMP and HEMP energy.

Figure 7.7: Differences in time parameters of LEMP and HEMP.

7.3 Differences between lightning and HEMP

|

157

channel represented by a single or even branched cord, the situation is more or less clear. However, in the case of HEMP, the situation is much more complicated. First, the shape of HEMP’s electric field near the ground surface develops in accord with the Earth’s magnetic field, and it is rather uneven. Second, the electromagnetic wave reaches the ground surface at a specific angle and thus, the electric field near the ground surface possesses both vertical and horizontal components. Third, part of electromagnetic energy falling onto the ground surface at an angle, will be reflected and can consolidate with the energy falling onto the ground. Fourth, if, when lightning strikes a lightning rod and its grounding system, the task is to minimize the resistance of the grounding-system elements in order to reduce the voltage drop across the conductive elements as lightning current flows through them, then in the case of HEMP, a high-voltage pulse is applied simultaneously to all the elements of the grounding system, acting as a huge antenna, absorbing energy from a large area. The smaller the resistance of the elements of this antenna and the larger its dimensions, the higher its absorption capacity and the greater part of the energy is delivered directly to the equipment connected to this “antenna”. This is a pulse voltage with an amplitude of hundreds of kilovolts. The main principal differences between HEMP and LEMP are summarized in Figure 7.8. These differences between LEMP and NEMP make it possible to assume that they are different in their effect on the objects located on the ground surface. Indeed, if we take a ten-meter metal rod, push one of its ends into the soil (vertically) and attach to it a current sensor, when lightning hits the open end of the rod, the sensor will register high-amplitude current flowing through the rod because the grounded end has zero (conditionally) potential, while the upper end acquires the high (relative to ground) potential of the lightning. When we have the bottom end of the rod well insulated from the ground surface and install it vertically, then there will be no current in the rod, even if we assume that lightning hits it because there is no potential difference between the rod’s ends (different capacitance values of the rod’s ends relative the ground can be neglected due to their low levels). If HEMP impacts the same insulated rod, there will be a high potential difference between its ends (theoretically, dozens of kilovolts), and the current sensor will register the relatively high amplitude current pulse flowing through it. Moreover, high potential difference occurs between the rod’s ends, even if it is located horizontally relative to the ground surface. What happens if we ground one of the ends of this horizontal rod? It is a much more complex case because HEMP penetrates into the soil and induces gradients directly in the soil. This effect takes into consideration the model of a power transmission line with a grounded neutral to study the HEMP effect. In such a model, the voltage on the open second end of the line to the ground will depend on the transmission line height above the ground, its length and soil conductivity [9]. But

158 | 7 The issues of electronic equipment grounding at the power facilities

Figure 7.8: The principal differences between HEMP and LEMP.

this model is not the case with insulated ends of the rod, and in our case the grounding of one of its ends does not affect the voltage gradient between the ends. The same effect will occur in a single electronic device installed in a cabinet in a control room with fully electrical insulated (without considering capacity to ground) control cables connected to its inputs. The electric field affecting these cables has nothing to do with the ground and its potential. In other words, such cables with a potential difference induced at its ends by HEMP acts as a EMP source insulated from the ground for electronic devices. It works as a charged accumulator battery in an insulated body. What happens, when only one pole of the accumulator battery is grounded? Just nothing! Neither with the accumulator battery, nor with the insulated load that receives power from this accumulator. So, why would something happen if we ground the HEMP-affected small local object in a control cabinet with electronic devices inside? This question is very important

7.3 Differences between lightning and HEMP

|

159

and highly relevant because it directly affects the efficiency of the equipment intended to ensure protection against HEMP. According to [10]: “The early-time E1 HEMP waveform also couples efficiently to short lines (1–10 meter) connected to equipment (power, signal lines, etc.) and can induce large voltages and currents that can be conducted to the inside of the equipment”. In this sentence, there is no relation to ground. The difference between the lightning and HEMP is also mentioned in other publications. For example, [11] directly states that “a “ground” is commonly thought of as a part of a circuit that has relatively low impedance to the local earth surface. A particular ground arrangement that satisfies this definition may, however, not be optimum and may be worse than no ground for EMP protection”; [12] suggests that: “techniques used for protecting equipment from slowly rising lightning transients are not necessarily effective against the fast-rising EMP-induced transients”; [13] suggests that: “Ideally, grounding would keep all system components at a common potential. In practice, because of possible inductive loops, capacitive coupling, line and bonding impedances, antenna ringing effects, and other phenomena, large potentials may exist on grounding circuits. The choice of grounding concept is therefore important in the EMP protection philosophy”. Unfortunately, this document is limited to the fact statements and offers no special grounding systems that can address the controversy between the need to have a grounding system and the danger that it can cause in case of HEMP impact. The authors of [14] are even more categorical: they make a straightforward suggestion that the grounding system has an adverse effect on the operation of electronic equipment: “If we don’t take special measures, both types of grounding systems (internal and external) can become the main source of interferences caused by the HEMP. External grounding devices cannot act as “bypass” for HEMP as the latter is very spacious and deeply penetrates into the soil. Moreover, absence of the loop of external grounding does not exclude a possibility of the HEMP’s interferences to occur”. A similar conclusion is also made in [8]: “a HEMP can induce very large currents and voltages onto the ground network “antenna”. These ground network currents and voltages pose a serious threat of damage to electronic components…” The inconsistency of the situation is clearly highlighted in the fundamental work [15] of more than 1,000 pages. The title of Section 4.1.1 speaks for itself: “Grounding May Not Be the Solution; Rather, It Could Be Part of the Problem” In addition, there are two contradictory ideas on page 935: – “The primary effect of the HEMP is, therefore, the production of large voltages or currents in large structures and conductors such power lines, buried cables, and antennas, as well as in facility grounding systems” – “The goal of all grounding and bounding techniques is to redirect the HEMP-induced currents to the earth.” The first idea suggests that the grounding system is a receptive capacity that absorbs HEMP energy, while the second idea states that the grounding system is a source of high voltage that brings HEMP energy to the grounded equipment.

160 | 7 The issues of electronic equipment grounding at the power facilities Unfortunately, it is a very difficult to study this phenomenon in an open area testsite simulator (OATS) suitable for simultaneous testing the group of electrical control cabinets with cable connection between them because most such simulators contain a Marx generator and two electrodes: one grounded mesh and another one—an insulated mesh placed above the grounded mesh at a height of 5–20 m. A simulated electrical pulse field is applied directly between these two electrodes, between the upper electrode and the ground. In such a simulator, well grounding the equipment under test (that is a low impedance connection the shields and metal shells of equipment to the down electrode) will always play the role of effective protection means as in lightning testing. Such a model is more suitable for lightning impact testing than for the effects of HEMP.

7.4 Grounding of electrical equipment as the main protective means for HEMP Various standards (both civil and military), as well as various guidelines and recommendations, justify the necessity of compulsory grounding of all types of electronic and electrical equipment as the main protection means against HEMP. But why, if the grounding system does not act as an opposite electrode with an opposite charge for HEMP (unlike a lightning strike)? According to [16] In general, the reason for grounding are varied, and it would be presumptuous to attempt to specify grounding procedures without first establishing the reasons for grounding and the goals that the grounding system should achieve. These reasons and goals are usually based on system functional, safety and RF interference considerations as a consideration in the ground-system design, at least one more goal has been added (EMP hardness), but the reason for grounding may remain unchanged. The basic reason for providing a “ground” in electronic equipment is to establish a firm reference potential against which signal and supply voltage are measured (or established).

Such considerations are a reason for the standard recommendation about standard grounding methods in all documents related to NEMP, despite that the grounding is not a clear and proven protection means against NEMP. But the functional and safety considerations and reference potential necessity for electronic equipment have also another direct grounding solution. At the same time, it is obvious that the branched and spatially distributed grounding system acts as a huge antenna for NEMP, absorbing energy from a large area and delivering it directly to sensitive electronic equipment via the grounding circuits. Of course, the energy level will be partially lowered by the conductive soil. However, the part that finds its way into the system will be enough

7.5 Protection devices for HEMP

|

161

to result in a dangerous potential rise directly in electronic circuits of highly sensitive microprocessor-based equipment (such as digital protection relays—DPR): – “Many elements of a facility can act as efficient collectors and provide propagation paths for EMP energy. EMP can couple to structures such as power and telephone lines, antenna towers, buried conduits, and the facility grounding system” [13]; – “Based upon coupling calculations it is appears that levels up to 10 kV may be coupled to horizontal buried lines in a substation yard (although 20 kV is possible under some scenarios)” [10]; – “A “ground” is commonly thought of as a part of a circuit that has relatively low impedance to the local earth surface. A particular ground arrangement that satisfies this definition may, however, not be optimum and may be worse than no ground for EMP protection” [11]. – “For HEMP protection, however, the grounding system is considered a potential path for transient penetration into the system and a means of distributing transients throughout the interior” [16]. – “Grounded metal structures, like the earthing system itself, work as antennas absorbing high-frequency energy” [17].

7.5 Protection devices for HEMP Usually, devices designed for protecting equipment from HEMP overvoltage are connected between the circuits to be protected and the grounding system (common mode protection), see Figure 7.9 and Figure 7.10. Thus, they are intended to protect sensitive inputs of equipment solely from pulses featuring higher amplitudes relative to the ground and divert energy from the input to the ground. However, when the grounding system does not represent the area of reverse potential or zero potential for HEMP, where will the pulse energy be diverted? And when a similar pulse occurs at the grounding electrode simultaneously with high voltage pulse occurring on the input of a filter or a device protecting from overvoltage, how will this filter weaken HEMP?

Figure 7.9: Simplified design of various LC-filters against NEMP and devices protecting from pulse overvoltage with parallel elements that divert impulse energy from the input to the ground. VR— varistors, GDT—gas discharge tube.

162 | 7 The issues of electronic equipment grounding at the power facilities

Figure 7.10: Filters protecting from HEMP pulse applied to equipment input terminals relative to the ground.

These questions are still waiting to be answered. Thus, the specialists invite active discussions concerning this problem because “grounding may not be a solution; rather it could be part of the problem” [15].

7.6 New method for grounding electronic equipment mounted inside the cabinets If the grounding causes so many problems, should we not ground the electronic equipment at all? For some reasons, a report of the special working group C4.206 CIGRE discussing the protection of the power system’s electronic equipment from Intentional Electromagnetic Impacts [18] does not entirely cover the issue of electronic equipment grounding. This way of grounding (more correctly “non-grounding”) is called “floating ground” in the technical literature, e. g., in basic American military standards for grounding systems MIL-HDBK-419A [19], from which it migrated to other military standards and instructions, e. g., [13]. This type of grounding is described in [19] and [13] as follows: The effectiveness of floating ground systems depends on their true isolation from other nearby conductors, i. e., to be effective, floating ground systems must really float. In large facilities, it is often difficult to achieve a completely floating system, and even if complete isolation is achieved it is difficult to maintain such a system. In addition, a floating ground system suffers from other limitations. For example, static charge buildup on the isolated signal circuits is likely and may present a shock and a spark hazard. In particular, if the floated system is located near high voltage power lines, static buildup is very likely. Further, in most modern electronic facilities, all external sources of energy such as commercial power sources are referenced to earth grounds. Thus,

7.6 New method for grounding electronic equipment mounted inside the cabinets | 163

a danger with the floating system is that power faults to the signal system would cause the entire system to rise to hazardous voltage levels relative to other conductive objects in the facility. Another danger is the threat of flashover between the structure or cabinet and the signal system in the event of a lightning stroke to the facility. Not being conductively coupled together, the structure could be elevated to a voltage high enough relative to the signal ground to cause insulation breakdown and arcing.

Alternatively, the specialists of Associated Power Technologies (APT)—one of the leading manufacturers of powerful power sources—state that [20]: Grounded systems can present their own set of problems… The dark side to grounding systems is that they can create issues that affect the rest of an electronic system… Small ground loops inject noise onto a system and cause interruption or loss of communication on data lines such as RS-232 or GPIB interfaces. Large ground loops can damage electronic equipment and even pose safety hazards if the ground loop currents are large enough.

Based on the analysis carried out in [20], these specialists conclude that the floating ground is a preferred measure of electronic systems protection, emphasizing that this approach is standard for the products manufactured by their company. So, is there any way to resolve this controversial situation? It can be suggested that a solution is to use a special (advanced) floating ground, i. e., the local area of reference zero potential (local ground), which is established by, for example, a special copper bus 3, resting on insulators in a metal cabinet 1 with the electronic equipment 2 inside, Figure 7.11. The separate bodies of each electronic device 2, insulated (with minimum capacitance from the cabinet) by means of insulators 4 should be connected to a copper bus 3. Copper buses of various cabinets located

Figure 7.11: Suggested grounding system: special floating ground 1—metal cabinets with electronic equipment; 2—electronic devices; 3—copper bus establishing the local area of common (reference) zero potential (local ground); R—high-resistance high voltage resistors (100 MOhm); S—contactbreaker with electrical locker on the cabinet’s door; L—choke; G—facility’s grounding system.

164 | 7 The issues of electronic equipment grounding at the power facilities

Figure 7.12: Example of a grounding system “special floating ground”. 1—metal structure of the cabinet; 2—electronic device, e. g, DPR, 3—copper bus establishing the zero potential area; 4— insulators; 5—screening enclosure (Figure 7.14); 6—choke; 7—contact breaker; 8—high-voltage highresistance resistor (Figure 7.13).

Figure 7.13: High-voltage high-resistance resistors.

close to each other in one room should be connected to an external copper bus located coaxially in a shielded enclosure 5, Figure 7.12. The zero potential local area should be connected to the external grounding system only via a high-voltage high-resistance resistor 8 (Figure 7.12), with a resistance rating of 100 MOhm that prevents accumulation of static discharge on electronic device enclosures. As just mentioned, modern electronic devices generally do not require functional grounding, so the solution offered does not influence the operating performance of electronic equipment. However, protection of personnel from a dangerous potential resulting from insulation damage upon the internal damage of equipment and cables inside the cabinet was solved in the presented concept by other measures. First, the metal cabinet can be grounded through a powerful high-frequency choke wrapped by a copper bus. Such a choke should have very low resistance under

7.6 New method for grounding electronic equipment mounted inside the cabinets | 165

Figure 7.14: Casing screen (screening enclosure) for a copper bus.

Figure 7.15: Powerful high-frequency chokes.

direct current and 50-Hz alternating current, but, at the same time, it should ensure high resistance to a lightning pulse and especially high resistance to a very short HEMP pulse, see Figure 7.15. Second, the cabinet should be equipped with a three-phase contact-breaker with high-power terminals and a handle, located outside the cabinet, see Figure 7.16 and 7.17. The contact-breaker should be closed when personnel are working inside the cabinet. One terminal should be used to connect the internal insulated copper bus to the external grounding system; the second terminal provides direct connection between the cabinet and the external grounding system; the third is used to switch off the cabinet’s door locking mechanism to ensure that the door can be opened only after the contact-breaker is closed. It is also possible to use a two-pole circuit breaker with additional auxiliary contact for these purposes. Moreover, (if necessary) a temporary ground attached to the internal insulated copper bus can be used when someone is working inside the cabinet. However, it seems not to be necessary. However, the time when the staff is working directly on activated DPR is incommensurably short compared to the total DPR operation time. This results in reasonable suggestion to provide short-time grounding of DPRs only when the staff is working on them. From the technical point of view, this idea can be easily realized by so-called “position” (or “end” or “limit”) switches installed on the relay cabinet door to ground

166 | 7 The issues of electronic equipment grounding at the power facilities

Figure 7.16: High-power three-pole contact-breaker with a visible circuit disconnection.

Figure 7.17: Two-pole contact-breaker with auxiliary contact.

DPR enclosures and controllers when the door opens. Modern position (limit) switches are highly reliable, well-protected (against mechanical damage and environmental conditions) devices widely available on the market, see Figure 7.18. They are extensively used in critical industrial systems, road and air transport, military hardware and in staff-protection systems. Various types of such switches can changeover currents of 10–16 A, at voltages of 400–690 V and can have variously complexity arrangements: from one NO (NC) contact to several groups of changeover contacts. In order to

7.6 New method for grounding electronic equipment mounted inside the cabinets | 167

Figure 7.18: Heavy-duty “position” (or “end” or “limit”) switches.

improve the reliability of DPR enclosure grounding upon opening of the relay cabinet door, two limit switches with parallel connected contacts can be installed. To increase the electric strength of contact spacing of an electromagnetic pulse, the switches rated to the maximum operating voltage (660–690 V) must be used. If there are two similar internal contacts available on the switch, they can be connected in-series. If permanent grounding is replaced with connective grounding, all metal enclosures of electronics must be connected with a flexible copper bus to the common metal bus located in the cabinet and insulated from it with small plastic insulators. Such a common bus should be connected to the cabined grounding bar through the limit switch contacts that closed upon door opening. In specific cases, if it seems reasonable, the mentioned common metal bus can be wired to the cabinet grounding bus manually during work in the open cabinet. Such a wire must be insulated, permanently connected to the cabinet grounding and be equipped with an easily detachable connection to the common metal bus. The inevitable presence of the capacitive couplings between all the just-mentioned components and the ground creates a certain problem. Under high frequencies, such capacitances have low reactance and thus reduce the effectiveness of the isolation from the ground.

168 | 7 The issues of electronic equipment grounding at the power facilities In practice, it is difficult to calculate the capacitance of all devices located inside the cabinet and of the cables interconnecting such devices. However, it is possible at least to estimate the grounded capacitance of the cabinet structure and the DPR. The simple estimation made for a cabinet of dimensions 0.5 m × 0.5 m installed on 10-cm-high insulators on the grounded metal floor provides a capacitance of 25 pF. The DPR’s enclosure of 25 cm × 25 cm separated with a 1-cm thick and 1-m wide insulator has the capacitance of 10 pF. Such two in-series capacitances result in the DPR’s total grounded capacitance of about 7 pF (or 10 pF to be on the safe side). Such values alone do not say anything because the operational frequency plays the important role, and the frequency makes a great difference. Fourier transformation of a standard-shaped HEMP pulse gives a rather wide frequency range, see Figure 7.19. According to IEC 61000-2-9 [21], 96 % of HEMP energy is attributable to a 100 kHz– 100 MHz frequency range, while 90–100 % to a kHz–10 MHz range [22]. For a maximum

Figure 7.19: Fourier transformation of a standard-shaped HEMP pulse.

Bibliography | 169

frequency of 100 MHz, the calculated capacitive resistance is only 160 Ohm. This value is too low to confirm the effectiveness of the measures taken to isolate the electronics from the ground. Progressing through the frequency range to the minimum, the same capacitance has 1.6 kOhm under 10 MHz, 16 kOhm under 1 MHz and 160 kOhm for 100 kHz. Such values of the reactance applied to the circuit are higher than the values ensured by the widely-accepted direct connection to the minimum impedance. They can significantly weaken the HEMP impact on the equipment, see Figure 7.20.

Figure 7.20: Dependence of current induced by HEMP to a ten-m-long conductor on the circuit resistance.

To further improve the effectiveness of the sensitive electronics isolation from the ground, one can consider creating the common zero potential for the electronics located inside the various cabinets (and check its operability when the copper bus 3 is not available). Preliminary figures indicate that the operability of modern electronics with fully isolated inputs and outputs should not depend on the availability of such a common zero potential. To ensure the full isolation for DPRs, all communication channels must be optical. The combination of the above measures (the isolation from the ground (even partially due to the interfering capacitive couplings) and replacement of all communication channels between the DPRs located inside the different cabinets to optic fiber lines) can achieve the maximum level of effectiveness for HEMP protection.

Bibliography [1] [2]

Report ORNL/Sub/83-43374/2. Impact of Nominal Nuclear Electromagnetic Pulse on Electric Power Systems. Oak Ridge National Laboratory, 1991, 94 p. Duff W. G. Designing Electronic Systems for EMC: Grounding for the Control of EMI. Interference Technology, November 6, 2011.

170 | 7 The issues of electronic equipment grounding at the power facilities

[3]

[4] [5] [6]

[7] [8] [9] [10]

[11] [12] [13]

[14] [15] [16] [17]

[18] [19] [20] [21] [22]

Kuznetsov M. B., Matveev M. V. Protection Against Secondary Manifestations of Lightning and the Providing of EMC Resilience to Microprocessor Based Equipment at the Facilities of the Oil and Gas Industry. Energoexpert, 2007, Vol. 2, pp. 61–65. Whitaker J. C. Electronic Systems Maintenance Handbook, Second Edition. CRC Press (Taylor & Francis Group), Boca Raton–New York–London, 2001, 624 p. IEEE Std. 1100-2005. IEEE Recommended Practice for Powering and Grounding Electronic Equipment, 2005, 589 p. Barreto R. M. EMC = Grounding on Automation and Control Systems. Applications to Eliminate Electromagnetic Interference in Industrial Plants. Interference Technology, 2013, March 19, pp. 14–19. Armstrong K. Planes Need to be Grounded to Something? Interference Technology, July 1, 2013. Pamphlet No. 1110-3-2 Electromagnetic Pulse (EMP) and Tempest Protection for Facilities. Engineering and Design. Department of the Army, U. S. Army Corps of Engineers, 1990, 467 p. Scharfman W. E., Vance E. F. Electromagnetic Pulse Coupling and Propagation to Power Lines: Theory and Experiments. Final report AFWL-TR-73-287, Stanford Research Institute, 1974. High-Impact, Low-Frequency Event Risk to the North American Bulk Power System. A Jointly-Commissioned Report of North American Electric Reliability Corp., and U. S. Department of Energy, November 2009 Workshop. NERC, 2010. The Effects of Nuclear Weapons. U. S. Department of Defense and Energy Research & Development Administration, 1977, 660 p. AD-A009 228 Electromagnetic-Pulse Handbook for Electric Power Systems. Stanford Research Institute, 1975, 341 p. TM 5-690 Grounding and Bounding in Command, Control, Communications, Computer, Intelligence, Surveillance and Reconnaissance (C4ISR) Facilities. Headquarters Department of the Army, Washington, DC, 2002. Ricketts L. W., Bridges J. E., Miletta J. EMP Radiation and Protective Techniques. Wiley, New York–London–Sydney–Toronto, 1976, 380 p. Joffe E. B., Lock K. S. Grounds for Grounding. A circuit-to-System Handbook. IEEE Press, Wiley, 2010, 1064 p. Vance E. F. Electromagnetic-pulse Handbook for Electric Power Systems. Report AD-A009 228, Stanford Research Institute, for Defense Nuclear Agency, 1975, 341 p. Nelson S. D., Larson D. J., Kirkendall B. A. Project FOOTPRINT: Substation modeling and simulations for E1 pulses. Report LLNL-TR-741344 Lawrence Livermore National Laboratory, contract DE-AC52-07NA27344, 2017. Protection of High Voltage Power Network Control Electronics against Intentional Electromagnetic Interferences (IEMI). Report WG C4.206 CIGRE, 2014, 27 p. MIL-HDBK-419A Grounding, Bonding, and Shielding for Electronic Equipment and Facilities, U. S. Department of Defense, 1987, 404 p. To Float or Not to Float? Analysis of a Floating vs. Grounded Output. Associated Power Technologies Inc. (www.aptsources.com/resources/pdf/Floating%20Output.pdf). IEC 61000-2-9 Electromagnetic compatibility (EMC) – Part 2: Environment – Section 9: Description of HEMP environment – Radiated disturbance. Jianguo Z., Xin Z. Coupling Effect of Transmission Lines by HEMP Based on CST. General Assembly and Scientific Symposium (URSI GASS), 16–23 Aug., 2014, Beijing, China.

8 The issue of control cables selection for HEMP-protected electric facilities 8.1 Introduction Protection of electronic equipment (such as control, automation and relay protection) installed in electric power facilities, from HEMP has recently become very relevant. Nowadays, there are technical solutions including special shielded cabinets, filters, arresters, etc., that provide HEMP protection for the highly sensitive electronic equipment of power plants and substations [1, 2]. However, local protection of this equipment at the level of an individual cabinet or even a room is facing a serious problem, involving wasted efforts. And this problem is called control cables. They run over long distances and act as huge antennas absorbing HEMP energy from a large area and delivering it to the interior space of protected rooms and cabinets, i. e., directly to the inputs of sensitive electronic equipment. Thus, only special cables should be used for HEMP-protected facilities, or otherwise all the efforts will be wasted.

8.2 Designs and features of shielded control cables The fact that low-voltage low-frequency control cables used in this application (and, preferably, power cables too) should be shielded is obvious and does not require additional explanation. Indeed, there are numerous designs for shielded cables, and not all of them provide the best protection from HEMP. So, it is appropriate to address various designs for low-voltage shielded cables and their features. But first of all, it should be noted that there are two kinds of screens used for shielding of cables: 1. Foil: screens made of thin (30–100 µm) copper or aluminum foil or made of metalcoated (usually aluminum) plastic strips—single strips wrapped around the cable, or consisting of several separate overlaying strips. 2. Braid: screens made of interlaced (usually in the way of a “French braid”) thin copper (usually tinned) wires, which look like a solid flexible hose, wrapped around a cable, Figure 8.1. This screen is much thicker than those made of foil. The specifications of these two kinds of screens are different, Figure 8.2. For example, a foil screen will hardly be efficient at low frequencies because electromagnetic waves penetrate much deeper into metal than only the foil thickness. Furthermore, at higher frequencies, foil screens are much more effective than braid screens because the surface resistance of the latter is too great for the high frequency. https://doi.org/10.1515/9783110639285-008

172 | 8 The issue of control cables selection for HEMP-protected electric facilities

Figure 8.1: A braided screen of a cable—“French braid”.

Figure 8.2: Shielding effectiveness of two kinds of screens.

Figure 8.3: Required shielding effectiveness of HEMP-protected equipment according to MIL-STD188-125-1 [3].

Since STD-188-125-1 [3] (Figure 8.3) requires effectiveness shielding of HEMP-protected equipment over a quite broad range of frequencies, it is obvious that none of the justmentioned kinds of screens meets these requirements.

8.2 Designs and features of shielded control cables | 173

Figure 8.4: Various designs of shielded low-voltage cables and approximate averaged and generalized values of shielding effectiveness. a, b—with a single-layer screen (1); c, d, e—with double-layer compound screen: with foil (2) and braid (1); f—with triple-layer and h—with four-layer compound screens; i—with four-layer braid; O—outer casing of a cable, 5—internal inter-screen insulation.

Fortunately, there are many compound screens that combine the advantages of both types (Figure 8.4). Apart from design differences, the shielded cables also differ in the manner of placement of the individually insulated cores inside the cable because this can also affect the cable’s protection from electromagnetic interference. To enhance protection,

174 | 8 The issue of control cables selection for HEMP-protected electric facilities

Figure 8.5: Shielded control cable with twisted pairs.

individual pairs of cables are twisted with each other many times along the whole length of the cable (Figure 8.5). The just-mentioned design features of cables are often indicated on their marking: – UTP—Unshielded Twisted Pair; – F (Foiled)—a cable with common outer foil screen; – F/UTP (Foiled/Unshielded Twisted Pair)—a cable with common outer foil screen and unshielded twisted pairs; – FTP (Foiled Twisted Pair)—a cable without outer screen, but with foiled twisted pair; – U/FTP (Unshielded overall/Foiled Twisted Pair)—same as for the previous; – STP (Shielded Twisted Pair)—a cable without the outer screen, but with braid shielded twisted pairs; – S/FTP (Screened overall/Foiled Twisted Pair)—a cable with common outer braid screen and with twisted pairs with individual foiled screens; – F/FTP (Foiled/Foiled Twisted Pair)—a cable with common outer foiled screen and with twisted pairs with individual foiled screens; – SF/FTP (Screened Foiled/Foiled Twisted Pair)—a cable with a double screen (braid + foil) and with foiled twisted pairs; – SF/UTP (Screened Foiled/Unshielded Twisted Pair)—a cable with a double outer screen (braid + foil) and with unshielded twisted pairs; – SF (Screened Foiled)—same as for the previous. Other designs are very rare; they are not intended for control cables, but rather for coaxial high-frequency cables. There are two other design features of shielded control cables that have to do with the quality of the outer braid, which can be more or less enclosing (Figure 8.6). Apparently, the higher the braid thickness, the better the shielding capacity. The maximum thickness of braid filling is 95 %. The second design feature of the braid is related to its surface resistance: the lower the resistance, the higher the shielding capacity. This is the reason why high-quality braids are made of tinned copper, which is distinguished (tin color) from non-tinned copper (copper color).

8.3 Evaluation of control-cable shielding effectiveness | 175

Figure 8.6: Outer cable braids with various filling thicknesses (in percent).

Due to the high diversity of designs and because various types of cables are used at electric power facilities, a question arises if the overall shielding effectiveness can be measured.

8.3 Evaluation of control-cable shielding effectiveness The method of measurement of shielding effectiveness is described in IEC 62153-4-4 [4]. It is based on measuring the correlation between the power of a high-frequency signal delivered from an external generator to a conductive core of a cable and the power of a signal emitted from the outer surface of the cable’s screen to a special enclosed chamber. The Vector Network Analyzer (VNA), which consists of a high-frequency signal generator (port 1) and a receiver (port 2) of a signal weakened by the screen, is very convenient to measure the correlation between the delivered (on to a cable) and measured emitted signals (Figure 8.7). The cost of one of the lowest-cost VNAs (type Planar TR 1300/1), which works with an external PC (it is necessary to work both with VNA and with special software calculating attenuation caused by the cable’s screen based on data transferred from VNA), is 2,900 USD. Based on the layout of the testing bench shown in Figure 8.7 (indeed, this is a metal tube with two special locks and connectors at both ends), its cost should not be significant. But the reality turned out to be different. A kit of tubes of various lengths, along with a kit of connectors (Figure 8.8) and PC software to calculate attenuation imposed by the screen is sold at 20,000 USD. Apparently, this high cost results from the lack of competitors in the market as this measuring kit is manufactured exclusively by the Bedea company (Germany) in cooperation with the Rosenberger company (also Germany). There is something similar promoted by the Japanese company Keycom (SEM03 trademark), but they have not replied to requests for information regarding this equipment.

176 | 8 The issue of control cables selection for HEMP-protected electric facilities

Figure 8.7: Layout for testing the effectiveness of cable shielding according to the IEC 62153-4-4 standard.

Thus, the total price of the testing kit (with VNA and PC) makes the feasibility of this purchase very questionable. Is it so important to know the effectiveness of cable shielding, provided that the standard measuring method and the formulas used for it stipulate testing of a symmetric coaxial cable with a single conductive core with one screen? What will be the accuracy of measurement of a compound cable that contains the screens of twisted pairs in addition to the outer screen? Thus, considering all the just mentioned matters, we suggest using an approximate estimation of the cable shielding efficiency based on cable designs and their features described earlier.

8.4 Choosing control cables | 177

Figure 8.8: A kit of testing tubes (type CoMeT 90/1) with a kit of locks and connectors manufactured by the Bedea (Rosenberger) company to test cable-shielding effectiveness with an outside diameters 6–22 mm.

8.4 Choosing control cables It is obvious that facilities being designed should employ exclusively new cables with double (at least) shielding capacity, such as S/FTP, SF/FTP, SF/UTP or SF with the filling thickness of outer braid (compulsory tinned!) of not less than 85 %. Cables of such design are manufactured by many large manufacturers, such as Belden, Hosiwell, DDA, Helukabel, Elettrotekkabel and Huanghe Cable Group, among many others. In addition to that, many cable companies offer customized cable production. If it was impossible to obtain a cable from one of the manufacturers mentioned here (with the required section and the number of cores), it is possible to have this cable individually produced. For old facilities with existing cables that are not planned to being replaced, one can take into account the approximate estimation of shielding effectiveness of these old cables. This estimation may be based on the mentioned criteria (for critical sites with critical equipment only). Additional protection means will be planned, taking into consideration the estimation of existing shielding effectiveness of these cables. At the same time, it is unacceptable to use both protected and unprotected cables connected to various input terminals of the same piece of equipment or leading into the same control cabinet. Another important parameter, which has nothing to do with shielding effectiveness, but which is one of the important indicators of the cable’s reliability (and needs

178 | 8 The issue of control cables selection for HEMP-protected electric facilities to be considered during cable selection) is the electric strength of the cable’s insulation. For HEMP-protected electric facilities, insulation of a cable should sustain oneminute test voltage of not less than 2 kV AC between conductive cores, and between them and the common outer screen. As a rule, this insulation sustains short pulses of less than 1 µsec and the amplitude of up to 8 kV (such as HEMP’s pulse) without any damage. According to IEC 61000-4-25 [5] requirements, a pulse voltage of this amplitude should be used for testing equipment located inside ordinary reinforced-concrete production facilities to shield from HEMP impact.

8.5 Conclusion Use of shielded control cables in electric units that need to be protected from HEMP is very important. Due to a high diversity of shielded cables available on the market, the problem of proper cable selection becomes very relevant. The choice can be based either on accurate measurements of shielding effectiveness of various cable samples (using of special measurement tools) or an approximate evaluation based on information regarding the unique features of various cable designs. Due to the high cost of tools and devices measuring shielding effectiveness and the imperfect correspondence of the standard measurement method with some important cable designs, the second method of evaluation is preferable. Information regarding shielded cables’ designs, their features, marking and estimated screens’ efficiency provided in this article enable a customer to browse through the diversity of types and kinds of control cables available on the market and to make a correct choice.

Bibliography [1] Gurevich V. Cyber and Electromagnetic Threats in Modern Relay Protection. Taylor & Francis Group, Boca Raton, 2015, 205 p. [2] Gurevich V. Protection of Substation Critical Equipment against Intentional Electromagnetic Threats. Wiley, London, 2017, 240 pp. [3] MIL-STD-188-125-1 High-Altitude Electromagnetic Pulse (HEMP) Protection for Ground Based C41 Facilities Performing Critical. Time-Urgent Mission. Part 1 Fixed Facilities, 2005. [4] IEC 62153-4-4 Metallic Communication Cable Test Methods – Part 4-4: Electromagnetic Compatibility (EMC). Shielded Screening Attenuation, Test Method for Measuring of the Screening Attenuation as up to and above 3 GHz. [5] IEC 61000-4-25 Electromagnetic Compatibility (EMC) – Part 4-25: Testing and Measurement Techniques. HEMP Immunity Test Methods for Equipment and Systems.

9 Grounding of control-cable shields 9.1 Introduction There are ongoing debates as to the number of grounding points necessary for controlcable shields. These debates arise in various specialized forums (in various languages) and on the pages of professional journals. Why? Perhaps, the reason is that the practical experience of equipment use goes beyond theoretical speculation. Sometimes the best results are obtained in the case of unilateral grounding of shields. However, there are situations when bilateral grounding of shields does a better job.

9.2 Shielding principles Shielding is a common practice to increase the noise resistance of equipment. Generally speaking, the electromagnetic shield is represented by a metal partition (barrier) between the source of electromagnetic emission and the protected area, see Figure 9.1.

Figure 9.1: Operation of a metal shield. 1—metal partition (shield); 2—electromagnetic-wave energy impacting the shield; 3—part of the energy reflected from the shield’s surface; 4—part of the energy reflected from the boundary layer created by the shield’s wall and the outside environment; 5—part of the energy converted into the current in the metal; 6—the balance of energy penetrating through the shield into the protected area.

It is obvious from the figure that part 3 of energy 2 impacting the shield is reflected from the surface back into the spacing, while the other part 4 penetrates into metal and is reflected from the boundary layer created by the shield’s wall and the outside environment. Another part of energy 5 is converted into an electric current inside the metal, and the balance of energy 6 remaining after all these conversions finds its way into the protected area as noise.

https://doi.org/10.1515/9783110639285-009

180 | 9 Grounding of control-cable shields

9.3 Interference types and grounding options for cable shields Grounding of control-cable shields is considered to be an efficient practice to weaken this interference. There are two major concepts of grounding for control-cable shields: on one side of a cable and on both sides of a cable, see Figure 9.2.

Figure 9.2: Common grounding practices of control cable shields: on one side of a cable: protects against capacitive interferences (C); on both sides of a cable: protects from capacitive (C) and inductive (L) interferences.

Obviously, both practices have different features and specifications regarding various types of interference. There are four major types of interference: – Conductive – Inductive – Capacitive – Electrostatic Each of them is subdivided into two types: – wire–earth interference—voltage is applied between each conductor and the earth. It is also known as asymmetrical, in-phase or common mode (CM) interference; – wire–wire interference—voltage is applied between separate electric circuits or between the elements of the same electric circuit. It is also known as symmetrical, out-of-phase or differential mode (DM) interference. Conductive interference spreads through the direct electric contact between electric circuits. Thus, shielding of control cables is nonessential with this type of interference in these electric circuits. Capacitive interference spreads via capacitance between the central cores of a cable and the earth; between the shield and the earth; and between the shield and the central cores. Grounding of the cable’s shield at one or two points will shunt the capacitance between the shield and the ground. On the other hand, it will also bring the “earth” closer to the central core, thus increasing the capacitance between this core and the earth. This expedites the capacitive interference’s penetration from the earth to the central cores. However, apart from interference spreading through the earth’s

9.4 Problems and contradictions | 181

circuits, there is noise coming from the adjacent cables, from high-voltage wires, powerful high-voltage switching apparatus and other sources of electromagnetic interference. When this interference is in-phase, i. e., creates a potential relative to the earth, the grounding of the cable’s shield at one point assist elimination of this interference completely. For example, there is an unshielded cable in a common cable tray, and occasionally a significant pulse voltage occurs at its cores relative to the earth. Subsequently, single-point grounding of the adjacent control-cable shield will ensure efficient protection of the control cable’s central cores from pulse noise arising on the unshielded cable. However, when the just-mentioned pulse voltage arising in a unshielded cable causes the pulse current flow (the most common situation) that generates the pulse magnetic field around it (differential inductive interference), obviously that single-point grounding of the adjacent control cable’s shield will have no effect, and the noise will be conducted into the central cores of the control cable. Grounding of the shield at both end-points establishes a closed circuit for the current conducted in the shield, and this weakens the inductive interference impact on the central cores of the control cable. Static interference resulting from accumulation of a static charge on equipment parts, insulated from the earth with further discharging and breakdown of insulation to the earth, are not dangerous for cables because the charge flows freely to the earth through the existing insulation resistance and does not accumulate. The examples just discussed show that single-point grounding of control-cable shields protects the central cores from capacitive in-phase (relative to earth) interference only, while grounding at both ends will ensure protection of the central cores from any type of interference. Apparently, based on these thoughts, it is mostly recommended to use this type of grounding of control-cable shields.

9.4 Problems and contradictions However, it is not all that simple! In reality, the grounding system is not so ideal. If the cable is long enough and the current flowing through the grounding system is significant, there will be a high difference of potentials between the shield’s grounding points located far away from each other. According to [1] this difference of potentials in real grounding systems can reach 10+ kV upon the lightning strike. And this is not the most frightening situation that can happen. Upon impact from a spatially distributed electric field of a high-altitude electromagnetic pulse of nuclear explosion (HEMP), with the field gradient up to 50 kV/m near the soil surface on the grounding system (acting as a huge antenna), the difference of potentials can reach as high as tens of kilovolts at the coupling points of the long cables’ shields. When this voltage is applied directly (i. e., through a direct contact) to the shield, it will result in high amplitude current flowing through it, and this can induce significant current in the cable’s cores directly connected to electronic elements in the equipment.

182 | 9 Grounding of control-cable shields So, which type of control-cable shield grounding is more preferable? The majority of official documents, such as standards, guidelines and instructions of both civil and military application [2–8], suggest straightforwardly that the grounding of shields should be executed on both sides of a cable. Even though these documents are well-known to specialists that operate electronic equipment of power systems (particularly, digital protection relays—DPR) all over the world, there is an ongoing debate as to the number of grounding points needed for control-cable shields. These debates arise in various specialized forums (in various languages) and on the pages of professional journals. Why? Perhaps the reason is that the practical experience of equipment use is much broader than just theoretical speculation. Sometimes, unilateral grounding of shields gives the best results. However, there are situations when bilateral grounding of shields does a better job. What’s the matter? Personal communication with the top world specialists in this area, who are the authors of fundamental studies [9–11], could not have adequately clarified the situation. So, I tried alone to analyze the situation and find the answer to the question first mentioned above. As a result of thorough analysis of dozens of publications on this topic, including fundamental writings, where the issue of cable shields’ grounding is discussed thoroughly and comprehensively (for example, in [11] a separate 119-page chapter 7 is devoted to this topic), I came to the discouraging conclusion that there is no (and cannot be any) single, comprehensive answer to the question mentioned previously. Moreover, it is not even possible to articulate any accurate general recommendation regarding selection of any specific shields grounding practice that will be clear and suitable for practical application by the power systems’ staff.

9.5 Factors impacting the effectiveness of shield groundings This situation occurred due to the fact that available guidelines, recommendations, articles and even standards, which suggest a certain type of control cables’ shield grounding, justify the choice based on a very limited number of factors that really impact the interference–resistance of electronic equipment by considering only one of them and neglecting the others. This is why we have an ongoing debate among energy industry workers regarding their preferences in terms of specific ways of shields’ grounding and the reference to personal experience, which often contradicts the experience of other participants in the discussion. Which factors are we talking about? 1. Sensitivity of various types of electronic equipment to interference of various types, frequency, durations and amplitudes is not the same. Thus, one interference may cause faults in the equipment’s operation, while the other (even more powerful) that has another pulse frequency or duration will not result in fault conditions of the same equipment. This also means that the same interference affecting inputs of various electronic devices via the various cores of a multicore cable can cause fault conditions in some devices, but have no effect in others.

9.5 Factors impacting the effectiveness of shield groundings | 183

2.

Pulse current flowing through the shield of one cable may impact the current flowing in the central core of the same cable and shields of adjacent cables running parallel in a common cable tray. Alternatively, current flowing in the central cores of unshielded cables may affect the current in the shields of shielded cables, if both types are running in the same cable tray. 3. Various types of cable trays: metal or metallized plastic, open or closed—all of them differ in their ability to weaken electromagnetic interference. 4. Some parameters of the shield, such as inductive resistance to current flowing through the shield, as well as capacitive resistance between the central cores and the shield, between the central cores and the earth, and the shield and the earth, are significantly dependent on the frequency of interference or duration and the increasing of the leading edge of the pulse interference. 5. Various types of shields: single-, two-, three-, four-layer (see Figure 9.3), made of foil only, made of braid only, combined (braid + foil); twisted pair cable only, twisted pair cable with various kinds of shields—all of them differ in their shielding ability at different frequencies.

Figure 9.3: Cables with double (a), triple (b) and four-layer (c) combined (braid + foil) shielding capability.

6.

There is also a relationship between the shield thickness and the frequency of an interference because, depending on the frequency, the electromagnetic wave can penetrate into the shield to different depths (the so-called skin effect), which is comparable with the shield’s thickness (0.1–0.2 mm), see Table 9.1. 7. The shielding ability of the shield is also a function of the braid-filling degree of a cable protected by this shield. There are shields with 60–90 % of filling degree. In other words, the same interference can affect the equipment differently depending on the type of cable used. 8. The length of a shielded cable affects the absorbing ability of the shield upon the impact of electromagnetic field. What is especially important is the ratio of the

184 | 9 Grounding of control-cable shields Table 9.1: The depth of electromagnetic wave penetration into copper. Frequency, MHz

Depth of penetration, mm

0.001 0.01 0.1 1.0 10 100 1,000

2.09 0.66 0.21 0.066 0.02 0.0066 0.0021

wavelength to the cable length. In other words, interference of different frequencies (i. e., electromagnetic waves of different lengths) can affect the same cable differently; and alternatively, the same interference can affect the cables of alternative lengths differently. 9. The status, type and parameters of a grounding system can have a significant effect on the efficiency of grounded-cable shields. The previous example discussed the connection of the long cable shield to a real (rather than theoretical) grounding system at both ends. 10. In general, the proximity of a cable and even some of its parts to noise sources, as well as its direction in relation to these sources, plays a significant role. Considering this limited list of factors influencing the efficiency of control cable shields, one can draw the conclusion that there is a lack of data to reach an informed decision regarding selection of a particular type of the shield’s grounding. The lack of a single factor, such as parameters of pulse interference that affects the cable, makes it impossible to make a straightforward decision. In addition, it becomes obvious that even the general estimated model (supposing it would be possible to build it) will be useless in practice due to the lack of input data for specific conditions. Thus, I think a conclusion about a specific practice of grounding the control-cable shield could be drawn based on the experience of operation of the specific types of equipment under the specific conditions. Let us address another aspect of this topic, i. e., the issue of what should be considered a “dangerous” interference for the electronic equipment of power systems. For example, is a single pulse with a duration of several milliseconds (lightning) or several nanoseconds (HEMP) dangerous for such a common type of power system’s electronic equipment such as a digital protection relay (DPR), with a typical response time of 20–40 milliseconds? It is unlikely that it will be dangerous because this interference will not have enough time to significantly affect a long-duration process of data processing and the actuation of a DPR’ necessary function. However, what happens if this “interference” has an amplitude of dozens of kilovolts? Now that we are not

9.6 The suggested method of shield grounding

| 185

talking about a soft failure in the data processing software, it is rather an irreversible damage of the internal electronic components. As previously mentioned, these two are the most powerful, but short-duration types of interference that penetrate into the control cable by means of direct contact (from the grounding system to the shield, provided it is grounded at both ends), and then they go from the shield to the internal cores inductively. This means that the short-pulse interference itself, lasting as long as the lightning charge or HEMP, is not dangerous for electronic equipment (at least for DPR), if its amplitude remains low.

9.6 The suggested method of shield grounding Based on this discussion, the article offers an unusual method of grounding the control-cable shield. This method presupposes shield grounding at both ends, but one of them should be connected across a high-frequency choke (see Figure 9.4) that features specific inductive resistance.

Figure 9.4: The proposed way for grounding control-cable shields.

On the one hand, this suggestion contradicts all the canons stating that even insignificant increase of the shield grounding circuit’s inductive resistance reduces the shielding efficiency at high frequencies. However, nobody questions this: indeed, a choke included into the grounding circuit reduces the shielding efficiency from inductive noise (that is not very dangerous) at high frequencies (i. e., in case of very short pulses). While on the other hand, the most dangerous high-power interference penetrating the shield from the grounding system by means of direct contact will be significantly suppressed. Regardless of their simplicity and low price, these ferrite rings (filters, in essence) are very efficient in weakening the high-frequency current, see Figure 9.7. For this type of grounding, both ordinary devices (see Figure 9.5) are attached to the cable’s break connecting the shield to the grounding system and the modular ferrite rings in a plastic holder with a locker (see Figure 9.6) that are put on a wire and do not require its breaking and function as a high-frequency choke. However, when using ferrite rings, some specific features of ferrite rings described in [12] need to be

186 | 9 Grounding of control-cable shields

Figure 9.5: Conventional high-frequency chokes.

Figure 9.6: Ferrite rings in plastic holders with a locker.

considered. In order to obtain the required frequency response, several ferrite rings of various types can be attached to one wire. As for protection from low-power interference, this method of grounding controlcable shields sits in between the two basic approaches. Thus, in some specific cases, it can be more efficient compared to conventional methods of grounding, while in some other cases its efficiency is lower. However, under any circumstances, the protective high-frequency choke will prevent penetration of the most powerful and dangerous interference (lightning or HEMP) from the grounding system to the cable. In the case of continuous low-frequency noise (usually this is a rather powerful 50-Hz interference) in the shield, which leads to its excessive heating, an alternative method of limiting the low-frequency current in the shield by installing a capacitor in its grounding conductor (see Figure 9.8) can be used in addition to the choke previously offered. This compound band-pass filter that consists of capacitance and inductance connected in series will efficiently suppress both low-frequency inductive noise and very short-duration powerful pulse interference of a conductive type coming from the grounding system to the shield.

Bibliography | 187

Figure 9.7: Full resistance (Z) of a ferrite filter as a function of frequency of some ferrite types (typical).

Figure 9.8: Compound grounding of a control-cable shield by means of capacitance and inductance that constitute a band-pass filter.

Bibliography [1] [2] [3] [4] [5]

[6]

Kuznetsov M. B., Matveyev M. V. Protection from Secondary Effects of Lightning and Ensuring EMC of DPR Equipment at Oil-Gas Facilities. Energoexpert, 2007, Vol. 2, pp. 61–65. Industry Standard 56947007-29.240.044-2010 “Methodological Guidelines on Electromagnetic Compatibility at Electrical Grid Facilities”, The standard of Public Company FGC UES, 2010. Ruling Document 34.20.116-93 “Methodological Guidelines on Protection of Secondary Circuits and Electric Substations from Pulse Noise”, RJSC UES of Russia, 1993. IEEE Std. 1100-2005. IEEE Recommended Practice for Powering and Grounding Electronic Equipment, 2005, 589 p. TM 5-690 Grounding and Bounding in Command, Control, Communications, Computer, Intelligence, Surveillance and Reconnaissance (C4ISR) Facilities. Headquarters Department of the Army, Washington, DC, 2002. MIL-HDBK-419A Grounding, Bonding, and Shielding for Electronic Equipment and Facilities, U. S. Department of Defense, 1987, 404 p.

188 | 9 Grounding of control-cable shields

[7]

MIL-HDBK-1857 Grounding, Bonding and Shielding Design Practice, U. S. Department of Defense, 1998, 176 p. [8] Theory of Shielding and Grounding of Control Cables to Reduce Surges. General Electric Company, Power System Management Business Department, 1973. [9] Tsaliovich A. Electromagnetic Shielding Handbook for Wired and Wireless EMC Applications. Springer, New York, 1999, 682 p. [10] Tsaliovich A. Cable Shielding for Electromagnetic Compatibility. Springer, New York, 1995, 469 p. [11] Joffe E. B., Lock K. S. Grounds for Grounding. A Circuit-to-System Handbook, Wiley, Chichester, UK, 2010, 1065 p. [12] Gurevich V. I. The Problem of Correct Choice of Ferrite Beads. Electr. Eng. Electromech., 2016, Vol. 2, pp. 71–73.

10 HEMP filters 10.1 Introduction Protection of modern electronic equipment, particularly in the power industry from high-altitude electromagnetic pulse (HEMP), is an important and relevant problem in the current stage of machinery development due to the reasons discussed earlier [1]. The main methods of protection from HEMP impact in high-sensitive equipment include thorough electromagnetic shielding of the equipment and attached cables, as well as suppression of the pulse by means of special filters, Figure 10.1, that connect the equipment to external units and systems.

Figure 10.1: Typical circuit diagram of a power HEMP filter consisting of a set of LC-circuits.

There are several filters on the market manufactured by a number of companies, such as: ETS-Lindgren, MPE, Meteolabor-EMP, European EMC Products Ltd., Captor Corp., LCR Electronics, API Technologies, Astrodyne TDI Corp., Fi-Coil, EMI Solutions Pvt. Ltd and RFI Corp., among others. One would think: what’s the problem; do you want to protect your equipment from HEMP? Install these filters and sleep well! But the question is whether one can really sleep well after installation of these filters.

10.2 Do the filters really protect from an electromagnetic pulse? When selecting a filter that can efficiently suppress HEMP, one unexpectedly faces the problem: all the mentioned companies promote their filters as highly efficient means of protection from HEMP and declare that they conform to the military standard MILSTD-188-125 [2]. However, parameters of testing pulses applied to the filters significantly differ from those detailed in the standard. For instance, the standard stipulates that tests should be conducted with current pulses of 20/500 nanoseconds of undetermined amplitude with applied loads of 60 ohms, while the manufacturers test their filters at 8/20 µs with an applied load of one ohm. Why? Because the pulse of 8/20 µs is https://doi.org/10.1515/9783110639285-010

190 | 10 HEMP filters a standard pulse produced by all types of testing equipment intended for testing the resistance to lightning charges, while special equipment is required to test by current pulses of 20/500 nanoseconds with an applied load of 60 ohms; and the filter manufacturers do not possess it. This problem is directly addressed in [3]. Another strange thing is that MIL-STD-188-125 stipulates the testing of objects by current pulses in two modes: with current flowing between all inputs that are connected together and the ground (common mode) and between each separate input and the ground (wire-to-ground mode). However, in the case of a high-altitude nuclear blast, a high-voltage pulse can be applied not only between equipment’s inputs and the ground (this mode is referred to as “common mode” in other standards) but also between different inputs insulated from the ground (“differential mode”). MILSTD-188-125 does not contain these tests; they are addressed in other standards. Due to this, some companies promoting their filters as HEMP filters do not equip them with elements that limit pulse voltage, referring to the same MIL-STD-188-125, while they declare that the filters ensure full protection from HEMP, since they were tested by current pulses with amplitudes of several thousand amperes and recognized as conforming to MIL-STD-188-125. Indeed, the standard represents the limiters of pulse voltage as absolutely separate elements, which have nothing to do with filters, see Figure 10.2. With this approach, the filters really do not have to protect from overvoltage at inputs. On the other hand, is it possible to proclaim that these filters ensure full protec-

Figure 10.2: Design of inlet box for connecting of external cable to a unit protected from HEMP (according to MIL-STD-188-125).

10.2 Do the filters really protect from an electromagnetic pulse?

| 191

tion from HEMP? Obviously, since manufacturers understand the problem, they equip their filters with certain elements that protect from pulse overvoltage and are installed at inputs. In their opinion, these filters now enjoy the full right to be called filters that protect from HEMP. However, a close-up examination of the protecting elements used in those filters brings into question their efficiency. The most widespread and cheapest limiters of pulse voltage used in the filters are gas-discharge tubes and zinc-oxide varistors, see Figure 10.3. It is known that these are relatively “slow” elements, which do well suppressing standard pulses of 8/20 µs, but fail to actuate under the short high-voltage E1 pulse of HEMP featuring 2/25 nano-seconds [4] (or 5/50 nanoseconds according to [5]). Having read that the selected filter is intended for HEMP protection and was tested by current pulses of 8/20 µs according to MIL-STD-188-125, it is unlikely that the consumer will look for the standard and check whether this is the required pulse rate. However, it is early to challenge the manufacturers. It is directly discussed in [6] that insufficiently slow varistors and even slower gas-discharge tubes installed in filters seem to be unsuitable for HEMP protection; they just protect from lightning charges and switching overvoltages.

Figure 10.3: Filters of the MPE Company with protective elements consisting of voltage dependent resistors—VDR (zinc-oxide varistors) and gas-discharge tubes (GDT) at inputs.

Oh dear, filters intended to protect from HEMP employ voltage suppressors intended to protect from… lightning, but not from HEMP! Nevertheless, some of MPE’s filters with protective varistors are referred to as filters intended for protection from the E1 component of HEMP in the promotional materials. However, a more detailed analysis of parameters of these filters revealed that they do not differ (except for the name in the heading) from all other filters of this company, i. e., they are protected from lightning charge rather than from the E1 component of HEMP. Otherwise, it should be accepted that the parameters of lightning charge do not differ from E1 component’s parameters, which is absolutely not true.

192 | 10 HEMP filters During a discussion with a representative of one company regarding the feasibility of varistors’ use in filters, intended for protection from HEMP, a new argument was presented. The representative declared that regardless of the fact that a varistor itself is not deemed as a quick element, its efficiency becomes enough to protect from the E1 component when connected with L-C elements of the filters. Verification of this argument showed that it is also not accurate. Some publications [7, 8] suggest that connection of short external conductors having very low conductivity to the protecting element reduced their response time. It appears that the response time of the protecting element to the applied voltage pulse is very dependent both on the element’s casing design and on the configuration (length) of its pins [7, 8]. Moreover, [8] suggests that the configuration and length of the external pins determine the response time of protecting element. With this in mind, some manufacturers of protective elements often indicate in their promotional materials not the response time of a fully assembled element in the casing with pins, but that of the material used in the manufacturing of this protective element [8]. At the same time, manufacturers work on improvement of the configuration of protective elements, and they do it rather successfully [9]. The just-stated comments reveal that unbiased data about responsiveness of a certain type of protective elements can only be obtained by conducting one’s own independent trials of ready-made items offered on the market, although some indirect data about the results of these trials [10] enable making a preliminary comparative evaluation. For instance, according to [10], dynamic resistance and response time of protective elements based on avalanche diodes (TVS diodes) is almost ten times lower than that of a varistor. I have not conducted my own trials of diodes and varistors responsiveness in order to confirm or contradict these data, but the fact that protective elements based on avalanche diodes rather than varistors are used to protect electronic equipment from high-voltage electro-static discharges (and they belong to nanoseconds range, i. e., the closest to HEMP in terms of time parameter) is self-explanatory. Another problem related to the use of suppression elements is the circuit of connection of these elements employed in many filters, see Figure 10.1, where each element is connected between an input and the grounded casing of the filter. This type of connection results in the choosing of two elements connected in series between two input terminals of the filter. They stipulate double clamping voltage, which can be dangerous for the electronic circuit being protected. The technical requirements for equipment resistance to overvoltages and testing procedures are described in IEC 61000-4-4 [5] and IEC 61000-4-25 [11]. Electrical Fast Transient (EFT) is meant to happen under the testing pulse of this voltage, i. e., a quick pulse, parameters and testing procedures of which are described in IEC 61000-4-4. The procedure of the selection of testing-pulse parameters based on these standards for a specific example—digital protective relays (DPR)—is discussed in [12]. For this case, the amplitude of EFT pulse voltage amounted to 8 kV. In our opinion, these tests should be conducted on filters equipped with elements protecting from pulse voltage and intended for HEMP protection. This should be done

10.3 The frequency range of filters | 193

in addition to tests stipulated by MIL-STD-188-125, and the testing voltage should be applied both between inputs and the ground and between separate inputs.

10.3 The frequency range of filters Another problem is related to the amplitude and frequency features of the filters. The typical specification of a high-quality filter intended for HEMP protection is shown in Figure 10.4. Is there any relationship between parameters of real filters and this typical specification?

Figure 10.4: Typical insertion loss performance of high-quality HEMP filters.

As can be seen from Figure 10.4, the frequency range of such filters extends as far as 40 GHz. Of course, the wider the frequency range of the filter, the more expensive it is. This raises the question of whether such a wide frequency range is just an advertising trick designed to justify the high cost of a filter because, according to the IEC 61000-29 standard, the main energy of the HEMP (or more precisely 96 %) is allocated in the frequency range 100 kHz to 100 MHz. Another feature of the adopted connection schedule (between each input and the ground) of some separate internal elements of the filters is characterized by not only a double clamping voltage on the suppressor (as just mentioned), but also by double capacitance and double inductance connected between two inputs compared to inductance and capacitance between each input and the ground. This leads us to conclude that the frequency specifications of filters for a pulse applied between two inputs will not be the same as those for a pulse applied between an input and the ground. How far will these specifications be applicable for HEMP protection?

10.4 Feasibility of HEMP equipment protection with filters Today, the special LC-filters are deemed as basic and necessary means of HEMP protection for all kinds of electrical and electronic equipment. Thus, everyone seemed

194 | 10 HEMP filters

Figure 10.5: Certain types of EMP filters produced by several companies.

to feel no shade of doubt that such filters are effective and dozens of manufacturers produce hundreds of such filter types, see Figure 10.5. Besides communication, control, and monitoring electronics, such filters are applied for the power electric equipment protection, see Figure 10.6.

Figure 10.6: Powerful EMP filters designed for 800-A and 1200-A power circuits.

Independently, such a wide application of the filters on all kinds of equipment requiring HEMP protection, and the availability of the kinds of the filters on the market (including filters in the shape of wall tube insulators; filters embedded in multi-pin connectors, etc.) convinces the developers of the necessary equipment requiring protection to use such filters. As a result, there are even diesel generators equipped with EMP filters available on the market, see Figure 10.7. So, how do such filters provide the EMP protection for the equipment? To understand this, let us consider the characteristics of HEMP, see Figure 7.19. According to the characteristics shown in Figure 7.19, it is clear that HEMP is a combined noise impacting the equipment due to its high voltage and frequency. Consequently, EMP filters must protect the equipment both from pulse overvoltage and

10.5 Protection of equipment from HEMP high-frequency noise

| 195

Figure 10.7: EMP Engineering diesel generators equipped with a powerful EMP filter.

high-frequency noise. However, most EMP filters are very limited in operating voltage (up to several hundred volts), while the amplitude of EMP voltage induced to the cables and applied to electronics inputs reaches dozens of kilovolts. As already mentioned, due to this fact, the filters must be equipped with the additional highamplitude pulse-overvoltage protection according to MIL-STD-188-125, see Figure 10.2. It seems that all the aspects just mentioned are predictable and confirm the effectiveness of the widely accepted method of HEMP protection by filters. However, it should be considered that pulse overvoltage and high-frequency noise are absolutely different in physical action and severity, and thus require different means of prevention. So, are the filters really effective as the basic means of HEMP protection, despite the widely held belief? Let us consider both aspects separately.

10.5 Protection of equipment from HEMP high-frequency noise In fact, every item of modern industrial and power equipment is tested according to the EMC (electromagnetic compatibility) standards. These standards require checking the equipment’s immunity to the high-frequency emission and the high-frequency voltage applied to the equipment inputs (between the various inputs, as well as the joined inputs and the grounded enclosure). For example, Standard IEC 61000-4-12 requires testing the equipment under the high-frequency voltage of 1 MHz and an amplitude of 2.5 kV. Standard IEC 61000-4-4 and IEC 61000-4-5 requires applying the short pulses (i. e., high-frequency signals) with an amplitude up to 4 kV to the circuits of the tested equipment. Standard IEC 61000-4-3 requires checking the equipment’s immunity to

196 | 10 HEMP filters high-frequency emission within the frequency range up to 2 GHz. Thus, the standards should ensure the equipment immunity to the high-frequency noises including the HEMP’s range. However, certain parameters of pulse noise generated by HEMP differ from the parameters of the noise simulated during the EMC standards compliance test. In particular, HEMP noise has a little shorter pulse compared to the standard EFT (Electrical Fast Transient) pulse simulated upon the EMC compliance test. However, it is very unlikely that the regular industrial and power generation relay protection, control, and monitoring equipment compliant with EMC noise-immunity standards will not be able to withstand the nonperiodic single and very short EMP pulse (at limited amplitude). This is due to the fact that the parameters of the test of the immunity to the most problematic and the fastest HEMP E1 pulse significantly differ from the regular EMC test pulse only in terms of higher amplitudes of the test pulses. According to the regular civil standards, every industrial electronic device must be tested on EMC, i. e., withstand pulses with parameters equal to HEMP’s, provided that their amplitude does not exceed the amplitude stated in the standards. It follows from this that the special additional filters designed to suppress the periodic high-frequency noise are not required for ensuring protection against the single and very short electromagnetic pulse if its amplitude is limited by the other means.

10.6 Protection of the equipment from the HEMP-generated pulse overvoltage When it comes to protection against the high-amplitude pulse overvoltage applied by HEMP to the equipment circuits, the situation is different. Current EMC requirements and compliance test methods are not even close to the actual overvoltages affecting the equipment under EMP. However, there are well-known and widely used methods and means of pulse-overvoltage protection invented a long time ago, such as zinc-oxide varistors, see Figure 10.8. TVS-diodes, see Figure 10.9 (transient-voltage suppressor), based on avalanche diodes give little benefit over the varistors in terms of the response speed.

Figure 10.8: Certain types of powerful varistors designed for pulse-overvoltage protection of AC and DC circuits.

10.7 Ferrite filters | 197

Figure 10.9: Certain types of especially powerful TVS-diodes.

Varistors and TVS-diodes are characterized by the so-called remaining or clamping voltage, i. e., by the voltage remaining on the varistor and on the in-parallel connected protected unit after the varistor actuation. Since the regular EMC standards require the industrial equipment to withstand 2.5–4 kV overvoltage pulses, the powerful varistors (or TVS-diodes) with a clamping voltage of 500–600 V are able to ensure the reliable protection of the equipment against HEMP, without the need to use any addition HEMP filters. Why is this so important? That is why: – HEMP filters are expensive, especially the filters designed for the power generation equipment with a high current rating. – If there are many circuits to be protected, according to the standard method, you need to install many HEMP filters, meaning that you need a great deal of free space (which is often not available) and much money. – Since the EMP filters should be connected in-series to the multicore control and power cables, it makes the filter installation process expensive and complex, provided that there are many circuits to be protected. Compact and inexpensive varistors (or even the more expensive TVS-diodes), connected in-parallel to the units to protect control and power circuits against HEMP, make such an installation process more inexpensive and simpler.

10.7 Ferrite filters In particular cases, the electronics very sensitive to the noises may require filters to be used as additional means of protection. For the brand-new equipment, it can be recommended to use the solid ferrite beads and cylinders as the additional means of protection (control cables and cylinders installed in the cabinets must be routed through such ferrite beads). For the existing electric installations with the connected control cables, such a scheme is very difficult and cumbersome, and thus in such cases the solid ferrite beads and cylinders can be replaced with less effective dismountable

198 | 10 HEMP filters latched ferrite beads (cylinders) built into the plastic frame. This type of ferrite bead does not require the availability of disconnected control cables and should be slipped over the protected control cable and fixed with the latch, see Figure 10.10 [13].

Figure 10.10: Dismountable ferrite beads designed to be slipped on the cable for protection of the very sensitive electronics against electromagnetic noises.

Such filters are distinctly different from the above LC-filters since they are simpler, cheaper, readily available and mountable (do not require cutting every core of a multicore cable). To protect all the cable’s cores at the same time, you only need to slip such a dismountable ferrite bead (or a couple of them) onto the multicore cable. The impedance of the winding consists of one or several turns of a control cable run through the ferrite ring and is too low both for low-frequency operating signals and for commercial frequency-alternating current. At the same time, it is very high for high-frequency (short pulse) signals within the selected frequency range that depends on the number of turns, material and the size of the ring itself. As a result, the pulse current flowing through the wires of the control cable after activation of the protective element (that limits the amplitude of the pulse to a level of several hundred volts) will be limited by the increased impedance of these wires to a short current pulse. Thus, ferrite filters can be considered as an affordable and inexpensive addition to protective elements that limit the amplitude of the HEMP. FE-based filters are manufactured by numerous companies as listed in Table 10.1. Frequency ranges shown in Table 10.1 do not correspond to the certain filter type, but they describe the frequency area provided by certain companies. Actual frequency ranges of certain filter types are much lower than shown in Table 10.1. Figure 10.11 shows examples of frequency ranges of various types of materials used in FE manufactured by Fire-Rite Products Corp. Despite the apparent simplicity and cheapness (1–10 USD) ferrite filters are not as simple as they seem. Their effectiveness depends on the numerous parameters, such as: material type, equivalent frequency of current pulse to be attenuated, size

10.7 Ferrite filters | 199 Table 10.1: Frequency response of FE-based filters manufactured by various companies. Manufacturer Fire-Rite Products Corp. Ferrishield Ferroxcube Kitagawa Inc. Murata NEC/Tokin Parker Chomerics Laird TDK Leader Tech, Inc Wurth Elektronik Emicore Corp.

Frequency Range, MHz 0.2–1000 30–2450 0.2–200 0.15–100 Miniature, for PCB installation 0.1–300 30–200 30–2,000 10–500 1–2,450 1–1,000 And also miniature, for PCB installation 0.1–1,00

Figure 10.11: Frequency ranges of various material types (designated by numbers) used in FE Manufactured by Fire-Rite Products Corp.

of FE, number of turns of the conductor passed-through the FE, the value of the DCcomponent in the conductor, temperature, etc. The frequency response of the filter depends on several parameters, primarily on the FE material type. Manganese-zinc ferrites (Mn-Zn) with relative magnetic permeability (magnetic inductive capacity) μ = 600–20,000 are usually used for frequency range of 0.1–two MHz, and for one MHz–2.45 GHz, nickel zinc ferrites (Ni–Zn) with relative magnetic permeability μ = 15–2000 are used. Various ferrite mixes are also used during the manufacture. Apart from the frequency response the impedance—is another very important parameter of FE-based filter. It defines the level of noise suppression. To a large extent, the impedance of FE-based filter is also determined by the type of material used and by the operating frequency, see Figure 10.12.

200 | 10 HEMP filters

Figure 10.12: Dependency of the impedance of FE-based filter on the material type and frequency.

Because an FE-filter has inductance, capacitance and active resistance, see Figure 10.13, it is apparent that filter frequency response and impedance also depend on FE size (particularly on its length, see Figure 10.14).

Figure 10.13: Equivalent circuit of FE-based filter.

As shown in Figure 10.14, the filters having longer FE always have higher impedance, all other parameters being equal. This results from the higher inductance of filters with long FE. To a large extent, the impedance of FE-based filters also depends on the number of wire turns passing through the FE, see Figure 10.15. As shown in Figure 10.15, the starting impedance of the filter with several wire turns is great compared to the oneturn filter. However, the further increase in noise frequency makes filters with several turns less effective compared to filters with one turn, which can result from the higher capacitance of the filter with several turns. Also, FE-based filters have another unpleasant characteristic: their properties depend on the value of the DC-component of the passing current, see Figure 10.16. This results from the change of FE magnetic properties under the existence of the DC-component of the current. The presence of inductance and capacitance in the filter equivalent circuit (see Figure 10.13) provokes the hazard of resonance under the certain frequencies. This

10.7 Ferrite filters | 201

Figure 10.14: Dependence of filter impedance Z on the ferrite element length L made of two types of material (43 and 61) manufactured by the Fire-Rite Products Corp.

can lead to another problem of such filters—such as amplification of noise instead of its attenuation.

202 | 10 HEMP filters

Figure 10.15: Typical dependency of filter impedance on the number of urns (designated as 1–3) passing through the FE.

Figure 10.16: Influence of the DC-component on filter characteristics.

10.7 Ferrite filters | 203

Figure 10.17: Test unit designed to check the FE-based filter effectiveness.

So, if there are so many factors influencing the filter parameters, what are the basics to correctly select the filter ensuring effective protection against electromagnetic noise under a wide frequency range? This is tricky. This is especially due to: 1) unavailability of standards unifying the procedure for measuring the parameters of such filters; and 2) different methods of measurement used by the various manufacturers. Due to all these problems, it is almost impossible to compare the parameters of filters produced by the various manufacturers. Manufacturer data can be used only for preliminary filter selection. Afterwards, it is required to test the noise suppression effectiveness within the full frequency and current range needed for the customer. This test can be done on the unit consisting of the high-frequency pulse generator simulating the noise signal within the actual frequency range and of the receiving unit (Figure 10.17), such as an oscilloscope, spectral analyzer or electronic voltmeter with the extended frequency range. It is most convenient to carry out this test with the help of a Vector Network Analyzer (VNA) containing both a source and a signal receiver in one device. Such a study of the effectiveness of filters based on the FE of some types for the frequency range 300 kHz to 100 MHz was performed by the author using a Planar TR1300/1 vector network analyzer connected to a computer, see Figure 10.18. As ferrite elements (FE), Wurth Elektronik split snap-on ferrite beads were tested, built into the plastic frame with Star-Tec Snap 74271222 latches designed for installation onto the multicore control cable, with the outer diameter up to 12 mm. These FE are rated for one MHz–one GHz according to the manufacturer and cost approximately 6 USD each. Our test showed that a single FE was not capable of providing any useful noise suppression. To get the significant attenuation up to 10 dB (threefold current attenuation and tenfold power attenuation of the noise signal) within the ten MHz–100 MHz range, three similar ferrite beads installed in-series on the cable were required. However, as can be seen on the curves, such ferrite beads are not able to ensure effective noise elimination within the lower frequency ranges. Therefore, in order to improve the overall effectiveness, it seems reasonable to install the additional ferrite beads designed specifically for such a low-frequency range. Usually, the lower limit of the frequency range of such ferrite beads is 150–300 kHz and the upper limit is 30–100 MHz, according to the manufacturers’ specifications. Nevertheless, despite the numerous different ferrite bead characteristics indicated in the manufacturers’ specification, the ability of such ferrite beads to attenuate the

204 | 10 HEMP filters

Figure 10.18: The effectiveness of high-frequency ferrite bead filters.

Figure 10.19: The effectiveness of filters based on the one, two, and three low-frequency ferrite beads.

noise within the certain frequency range is unknown, and we were forced to perform additional tests. The resulting attenuation level provided by the above low-frequency ferrite beads are shown in Figure 10.19. Fair-Rate low-frequency split snap-on ferrite beads built into the plastic frame with latches type 0475164181 made of material 75 designed for installation onto the

10.7 Ferrite filters | 205

multicore control cable with the outer diameter up to 12 mm were tested. These ferrite beads are rated for the 200 kHz–30 MHz range according to the manufacturer and cost approx. 10 USD each. For illustration purposes, Figure 10.20 shows the resulting performance of three in-series low-frequency ferrite beads installed on the cable and acting near the frequency range lower limit. As can be seen in Figure 10.20, the most significant attenuation the low-frequency ferrite beads produced is within the frequency range where the high-frequency ferrite beads are not effective.

Figure 10.20: Resulting performance of three in-series low-frequency ferrite beads types 0475164181 installed on the cable and acting near the lower limit (up to 10 MHz) of the frequency range.

At first glance, it confirms the previous conclusion regarding the usefulness of the combination of high- and low-frequency ferrite beads installed on the same protected cable. The resulting performance of the full set of six in-series ferrite beads installed on the cable is shown in Figure 10.21. The comparative tests with the samples of ferrite bead type M93RS260130295, manufactured by the Chinese company Emicore Corp., were also performed. According to the company promotion, beads of this type are made of the new M93 material designed especially for the medium-frequency range, see Figure 10.22. As can be seen in the resulting frequency characteristics, see Figure 10.22, the samples of Emicore Corp. ferrite beads are not very effective within the low-frequency range. However, when the frequency rises above ten MHz, they show an increasing signal attenuation, while other bead samples trend towards a decrease of attenuation. As in the previous tests, it was assumed that the combination of the three new samples with three low-frequency Fair-Rate or Wurth Elektronik beads must ensure the best result. Indeed, three medium-frequency Emicore Corp. ferrite beads, enforced with three low-frequency Fair-Rate or Wurth Elektronik ferrite beads, ensure the significant correction of the performance of the initial samples, as can be clearly seen from Figure 10.23.

206 | 10 HEMP filters

Figure 10.21: The effectiveness of the filter based on the combination of three high-frequency and three low-frequency ferrite beads.

As can also be seen in Figure 10.23, low-frequency Fair-Rate ferrite beads are more effective compared to Wurth Elektronik devices within the frequency range up to one MHz. Also, the comparative test within the full frequency range was conducted, see Table 10.2 and Figure 10.24. As can be seen in the resulting data, Emicore Corp. samples demonstrated better results within the full frequency range, while Fair-Rite Products Corp. samples showed better results within the initial section of the range. In other words, six Emicore Corp. ferrite beads of the same type appeared to be much better than the combination of three low-frequency and three high-frequency ferrite beads of different types. Attenuation to 18 dB provided by the set of Emicore Corp. ferrite beads within the wide frequency range means an eightfold reduction of noise-signal amplitude and sixtyfold decrease of the signal power. These figures are satisfactory considering, in particular, that these beads are very inexpensive and widely available. It is possible that the set of six medium-frequency Fair-Rite Products Corp. ferrite beads of type 31 is no less efficient than the set of six Emicore Corp. devices. However, since discuss the industrial (not military) use and a very large number of ferrite beads, the price becomes the most significant factor. Emicore Corp. ferrite beads are much less expensive than Fair-Rite Products Corp. ones, while the quality is very high. Besides, many American companies working within the EMC field or manufacturing the EMC-related products are subjected to different export restrictions and must receive

10.7 Ferrite filters | 207

Figure 10.22: Attenuation within the frequency range up to ten MHz ensured by the sets of three different ferrite beads.

from the buyers an official statements confirming that their products will not be used for military or nuclear-weapon purposes. Clearly, it will be hardly possible to obtain such devices if we want to use them for HEMP protection. Besides the very sensitive electronics with low-resistance inputs, such beads can be used in all cases when the protected unit (even the high-current one) has low inner resistance, such as in a battery bank. Upon the HEMP voltage pulse, the voltage drop on the unit can be insufficient to actuate an in-parallel varistor, while the current flowing through such a unit can be very high due to the low inner resistance of the unit. However, the ferrite beads should be avoided in high-current DC circuits due to the saturation effect just mentioned. In such a case, a special high-current choke with an

208 | 10 HEMP filters

Figure 10.23: Correction of the resulting performance of the filter based on the combination of three Emicore Corp. ferrite beads enforced with three low-frequency Fair-Rate (top) and Wurth Elektronik (bottom) ferrite beads.

Table 10.2: Attenuation ensured by the various types and combinations of ferrite beads. The number of ferrite beads is shown in brackets.

Emicore M93RS26013O295-0B3 (3) + Fair-Rite 0475264181 (3)

Emicore M93RS26013O295-0B3 (3) + Wurth Elektronik 74272722 (3)

Fair-Rite 0475264181 (3) + Wurth Elektronik 74271722 (3)

0,5 1 10 50 100

Attenuation, dB Emicore M93RS260130295-OB3 (6)

Frequency MHz

−4 −8.5 −19.4 −19.5 −18

−6.3 −12.3 −15.5 −14.3 −10.2

−3.9 −9.1 −15.7 −14.5 −10.4

−4 −11.6 −11.2 −12 −10.4

10.7 Ferrite filters | 209

Figure 10.24: Attenuation within the full frequency range ensured by the sets of six different ferrite beads.

impedance ignorable under the DC current and large under the high-frequency signal (short pulse) should be connected in-series to the unit. The typical applications of the chokes are discussed in Chapter 12.

210 | 10 HEMP filters Since, according to Chapter 5, the ferrite bead can also ensure the effective protection using the inexpensive powerful varistors instead of the very expensive powerful TVS-diodes, this alone is the most important advantage of ferrite beads.

10.8 Conclusions Are we going to dig that deep on the question in order to protect our electronics against HEMP? Why shouldn’t we believe the manufacturers certifying the high efficiency of their products? There is also another point: even if we do not believe the manufacturers, in most cases we have no chance to check the real effectiveness of such a filter. Additionally, it is quite clear how effective the sensitive equipment protection will be, if when impacted in a critical situation the filter will be incapable of ensuring protection against HEMP. Today, each manufacturer is free to decide if the surge protective devices should be embedded into the filter, should the products include inexpensive components with mismatched parameters and should the filters be tested upon the standard “lightning” pulse or upon the pulse corresponding to the military standard. Nobody is going to check it! This happens since there is no special standard describing the requirements to the HEMP-filter parameters and structure, test methods and quality criteria. Considering the critical importance of the problem, such a situation is unacceptable and should be resolved as soon as possible. From all this, it follows that despite the common practice, it is better to stop using the special expensive filters to ensure the basic protection of electric and electronic equipment against HEMP, and replace them with the regular pulse-overvoltage protection equipment, such as varistors or TVS-diodes. Combined with other well-known methods and means of protection, they are capable of ensuring the adequately reliable protection, remove the “surprises” peculiar to the promoted filters and significantly streamline and make less expensive HEMP protection. As for the very sensitive electronics or power circuits with very low inner resistance, the pulse overvoltage protection can be amended by the surge current-limiting devices based on the ferrite beads or the powerful high-frequency chokes. Regarding the ferrite beads, the faster response time of TVS-diodes compared to the varistors’ (stated in certain sources) pales into insignificance, since the ferrite bead lengthens the high-voltage pulse and reduces the pulse rise time significantly (see Chapter 5), thus ensuring the reliability and effectiveness of the inexpensive varistors. Therefore, the ferrite beads become a rather basic additional means of protection and should be applied as the mandatory components installed on the multicore cables, in addition to the varistors installed on each input of the equipment such control cables are connected to.

Bibliography | 211

Bibliography [1] [2] [3]

[4] [5] [6] [7] [8] [9] [10] [11] [12] [13]

Gurevich V. Cyber and Electromagnetic Threats in Modern Relay Protection. CRC Press (Taylor & Francis Group), Boca Raton, New York, London, 2014, 222 p. MIL-STD-188-125-1 High-Altitude Electromagnetic Pulse (HEMP) Protection for Ground-Based C4I Facilities Performing Critical, Time-Urgent Missions; Part 1: Fixed Facilities. A. J. Nalhorczyk, HEMP Filter Design to Meet MIL-STD-188-125 PCI Test Requirements. IEEE. 10th International Conference “Electromagnetic Interference & Compatibility”, 26–27 Nov., 2008, pp. 205–209. MIL-STD-461F Requirements for the Control of Electromagnetic Interference Characteristics of Subsystems and Equipment, 2007. IEC 61000-4-4 Electromagnetic Compatibility (EMC) – Part 4-4: Testing and Measurement Techniques. Electrical fast transient/burst immunity test, 2012. Application Notes Cat. 1: HEMP Filter Maintenance and Monitoring. Rev. 1. MPE Ltd., December, 2012. Surge Protective Device Response Time, Application Note 9910-0003A, Schneider Electric, August 2011. Power Quality Surge Protective Devices (SPD), Application Notes: Response Time Ratings, DET-733 (8/10), General Electric. Surface Mount Power TVS Diodes Deliver Optimal Protection for Power Supply. Application Note, Bourns, Inc, 7/14.e/ESD1435. S. J. Goldman, Selecting Protection Devices: TVS Diodes vs. Metal-Oxide Varistors, Power Electronics, June 1, 2010. IEC 61000-4-25:2001 Electromagnetic Compatibility (EMC) – Part 4-25: Testing and Measurement Techniques. HEMP immunity test methods for equipment and systems. Gurevich V., Problems in Testing Digital Protective Relays for Immunity to Intentional Destructive Electromagnetic Impacts. Compon. Technol., 2014, Vol. 12, pp. 161–168. Gurevich V. The Problem of Correct Choice of Ferrite Beads. Electr. Eng. Electromech., 2016, Vol. 2, pp. 71–75.

11 High-voltage insulation interfaces 11.1 Introduction High-voltage insulation interfaces based on reed switches comprise a relatively new isolating interface type developed by the author [1]. Like the other high-voltage isolating interface types [2], the reed-switch-based interface is aimed at the transmission of the protection and discrete control command and signals between the installations components with high differential potential. It is designed for operation under operating voltages of 10–100 kV DC between the input and the output. The reed switch-based interface is a quick-response element fitted with high-voltage insulation separating the coil (input) and the sealed-contact-reed switch (output). In the past, such interfaces (designated by different local numeric codes) were widely used in many military power (radars, transmitters, lasers, etc.) and special electrophysical equipment. Since HEPM generates pulse voltages of many dozens of kilovolts induced between the different installation components, the reed-switch-based interfaces can be very effective to ensure HEMP protection.

11.2 High-voltage link for transmitting discrete commands in relay protection, automation and control systems Relay protection and automation devices, providing protection, automation and control (PAC), exchange the high number of discrete commands (such as ON-OFF signals from the so-called “dry contacts”) transmitted between the separate PACs, as well as between such devices and actuators (such as disconnectors, circuit breakers, etc.) within the same substation or between remote substations. Such commands are transmitted through the special communication channels. Between the remote substations, the communication is made by the high-voltage overhead transmission lines (HVOTL), with two sets of high-voltage transceivers connected to each side of HVOTL through the special connecting devices (power-line carrier equipment). Often, the conductors designed for construction of new HVOTL contain internal optical fibers, covered with special steel shields, to be used for establishing communication channels between the remote substations. Encoded radio-relay links, and recently, Ethernet-based communications, are also used for this purpose. Generally, the communication between PACs and telecommunication equipment, as well as local communications between various devices of the same substation, is realized through the special communication links based on various operation principles, see Figure 11.1.

https://doi.org/10.1515/9783110639285-011

11.2 High-voltage link for transmitting discrete commands | 213

Figure 11.1: Communication links for transmitting PACs discrete commands based on various operation principles.

For example, there are links designed to transform discrete input signals generated by dry contacts into coded optical signals to be transmitted through the optical fiber, and then decoded and restored, see Figure 11.2, while other devices communicate over the local Ethernet network. What is the advantage of a communication fiber-optic link over the common control multicore copper cable? First, it is the cost savings because the fiber-optic cables are less expensive than the copper ones. Second, it is the improved immunity to the external electromagnetic noises compared to ordinary control cables. While the situation is seemingly clear and straightforward, however, it is not all that simple. Certainly, the above advantages of the fiber-optic cables are obvious. However, the fiber-optic cable is the only component of the comprehensive set of equipment ensuring fiber-optic communication. Two other components of the set (coding and decoding devices) are not that inexpensive and are immune to the external electromagnetic noises [1]. Moreover, their reliability is also questionable due to their complexity. The complexity of

214 | 11 High-voltage insulation interfaces

Figure 11.2: Contact-closure communication fiber-optic link diagram.

such microprocessor-based equipment is shown in Figure 11.1. As can be seen, at least three microprocessors are located on the circuit boards of the multiplexer FOCUS mentioned in [1]. On the other hand, the adequate copper control cable with a properly grounded shield [2] provides the same high noise immunity as the optic fiber. Today, there are multicore control cables consisting of twisted-pair wires each screened with the individual shield available on the market, see Figure 11.3. Additionally, such cables are wrapped into the common foil screen covered with a copper braided screen with 85 % coverage and protected with an external plastic cover.

Figure 11.3: Multicore control cable IBI0508P series (Hosiwell) with eight twisted-pair wires and a three-layer screen.

As for the reliability, it hardly needs saying that copper cable is more reliable than the sophisticated microprocessor-based fiber-optic link. However, optic fiber also provides full galvanic isolation between connected circuits, while the common shielded cable is not capable of providing it, even when such shielding is very effective. Today, equipping PACs with highly efficient galvanic isolation becomes extremely important due to the necessity to ensure power-system electronics immunity to high-

11.2 High-voltage link for transmitting discrete commands | 215

altitude electromagnetic pulse (HEMP) [3, 4] generating an electric field of very high strength: up to 50 kV/m near the ground surface. For this purpose, to ensure the reliable transmission of contact closure commands on the PACs, combined with high-level galvanic circuit isolation developed by author, the high-voltage isolation interfaces are based on reed switches (“gerkotrons”), see Figure 11.4.

Figure 11.4: High-voltage isolation interfaces built on reed switches (“gerkotrons”) designed to work under pulsed voltages up to 50 kV between input and output.

Initially, these devices were developed for military and electrophysical applications [5], according to military standard MIL-STD-202 for electrical and electronic components, and as such are highly reliable. There are many designs for such devices and each developed with various properties, parameters and functions [6]. However, the simplest gerkotrons are suitable for the purpose previously mentioned, see Figure 11.5. They should contain a reed switch separated with HV isolation from the control coil or equipped with an auxiliary high-power switching element for direct control of the HV circuit-breaker trip coil (transfer trip). These devices are equipped with very robust elements. For example, the miniature reed switch type KSK-1A85 (MK23-85) has the best combination of switching parameters and sizes: switching current up to 1 A, voltage up to 1,000 V (under switching power of 100 W), withstanding voltage between contacts up to 1,500 V, operation time one ms, balloon diameter 2.7 mm, balloon length 21 mm. The Compact thyristor type GS60-16io1 is able to switch (short-time) up to 75 A (pulse current up to 1,500 A) and has the withstanding voltage of 1,600 V. The HV isolation body of the gerkotrons is made of a special plastic type Ultem1000, combining the range of outstanding isolation, mechanical and climatic parameters, together with a high epoxy compound adhesion. In the devices, the control coil and switching elements are protected with varistors. The outputs are made of special flexible wire cable type 178-8195 (made by Reynolds Industries), to withstand the test

216 | 11 High-voltage insulation interfaces

Figure 11.5: Diagrams and elements of two simplest types of gerkotrons: top—gerkotrons used for transmission of contact-closure commands between PACs, bottom—higher power gerkotrons used for direct control of the circuit-breaker trip coil (transfer trip). RS—reed switch; VS—HV thyristor; RU —zinc-oxide varistor.

voltage of 50 kV: the outer isolation diameter is only 2.54 mm and the conductor core section is 0.6 mm2 . Additionally, the wire surface finishing ensures high epoxy compound adhesion. The gerkotron operation time is one ms (it can take a little more time due to the control coil inductance)—it is too small to negatively affect the PACs operation, while it is too large to result in false operation under the high-power, but very short-time HEMP pulse. That is, except for HV galvanic circuit isolation, the gerkotron also acts as the impulse filter preventing the high-power disturbing signals of less than one ms from entering the inputs of electronic PACs. See one of the application scenarios in Figure 11.6—the gerkotrons are used to ensure the HV isolation of input and output circuits of PACs (Digital Protection Relays— DPR, as an example) from external circuits during the contact-closure commands transmission and receipt. In this case, the gerkotrons are used instead of fiber-optic communication links. They are installed together with a multicore copper control cable with a combined multilayer-shield grounded on the one side with the capacitor, and on the other side, with the HF choke, according to the recommendations in [2]. All cabinet cable entries used to transmit discrete commands between PACs are additionally protected by high-power bidirectional transient voltage suppressors (TVS), see Figure 11.7. In this situation, the transient voltage suppressor diodes (or TVS-diodes) are preferred over the traditional varistors, as TVS-diodes have significantly shorter response

11.2 High-voltage link for transmitting discrete commands | 217

Figure 11.6: Gerkotrons used to ensure HV isolation of DPR input and output circuits from external circuits during the contact-closure commands transmission and receipt. IM—input module with the set of gerkotrons; OM—output module with the set of gerkotrons, TVS—additional protection elements (high-power bidirectional transient-voltage suppressors) preventing the penetration of surge voltages into the relay protection cabinet.

time and are able to adequately respond to E1 component of HEMP, having the very short pulse rise time of 2 ns. Previously, TVS-diodes had low power and were used for protection of electronic components located on the printed circuit boards. Today, two world leading firms, such as Bourns and Littelfuse, offer the high-power protection built on TVS-diodes, see Figure 11.7. As already shown in Chapter 5, in spite of the higher performance of diode suppressors (or TVS-diodes) noted in technical literature, ordinary inexpensive varistors are also suitable for protecting equipment against HEMP. It is easier to mount a small varistor in the internal space of the herkotrons than a larger TVS-diode, which may be important for miniaturization of the herkotrons. In addition, there are no highly sensitive low-voltage electronic components in gerkotrons, even with an additional powerful thyristor, that would require the use of especially fast-acting diode suppressors. The usage of the proposed technical solution in the PAC systems will allow increasing the HEMP resilience of discrete-command transmission equipment.

218 | 11 High-voltage insulation interfaces

Figure 11.7: High-power bidirectional transient voltage suppressors 58 V, 15 kA (for input circuits protection—coils direction) and 430 V 10 kA (for output circuits protection—reed switches or thyristors direction) manufactured by Bourns (left) and Littelfuse (right) companies.

11.3 Usage reed-switch-based high-voltage interfaces in HEMP susceptibility tests The high-voltage isolation interfaces can be more sophisticated than shown on Figure 11.5. In particular, the devices can include in addition the adjustable threshold function (Figure 11.4). Such function permits to use the device as over-current protection relays controlling the current level flowing through the high-potential circuit and transferring output trip signal to the earth potential (Figure 11.8). Such adjustment is ensured by the rotation of the isolated sealed contact ampule (1—see Figure 11.8; 10—see Figure 11.9) around its axis and the following capsule detention at the certain location by applying the fixing element (4—see Figure 11.8; 8— see Figure 11.9). Since such devices can operate as the adjustable-threshold current detectors, they can be especially advantageous when installed in the HEMP-test model system (see Chapter 3). The state of various devices used in such a model and located within the area of HEMP impact can be realized with the auxiliary electronic equipment located outside the HEMP area, provided that there are HV galvanic circuit-isolation components designed to withstand the voltage of dozens kV. The reed switches based highvoltage interfaces, with both adjustable and not adjustable thresholds, can significantly simplify this task, especially upon the test of the relay protection devices.

11.4 Design features of high-voltage isolation interfaces | 219

Figure 11.8: The design of the adjustable-threshold compact reed-switch-based high-voltage interface to withstand voltage 50 kV consisting of the sealed reed-switch contact moving eccentrically against the control coil poles. 1—prominent part of the rotating ampule enclosure; 2—mounting flanges; 3—magnetic core; 4—anchoring screw; 5—winding coil; 6—coil terminals; 7—core poles; 8—pole fixing screw; 9—reed switch; 10—insulating spacers; 11—magnetic shunt.

11.4 Design features of high-voltage isolation interfaces The devices just mentioned are equipped with robust elements. For example, with the miniature vacuum reed switch, see Table 11.1, and thyristor type GS60-16io1 is capable of switching (momentarily) the current up to 75 A (pulse currents up to 1500 A) under the voltage up to 1,600 V, despite its small size. The reed-switch isolation interface trips in one ms (it can take a little more time due to the control coil inductance)—this value is too small to negatively affect the relay protection and automation system operation, while it is too high to result in a false trip under the high-power, but very short-time, HEMP pulse. Specifically, except for the HV galvanic circuit isolation, the reed-switch-based device also acts as the surge absorber preventing the high-power disturbing signals

220 | 11 High-voltage insulation interfaces

Figure 11.9: The adjustable-threshold reed switch designed to withstand pulse voltages of 70 kV consisting of the reed switch located outside the coil (an angle exists between its longitudinal axis and the longitudinal axis of the coil). 1—plastic mushroom capsule with the reed-switch slot; 2—epoxy compound filling the coil-enclosure body; 3—coil; 4—magnetic core; 5—reed switch; 6—plastic nut; 7—epoxy compound filling the rotating enclosure of the reed switch; 8—rotating enclosure fixing nut; 9—reed-switch terminals; 10—reed-switch rotating enclosure. Table 11.1: Parameters of certain types of high-voltage vacuum miniature reed switches suitable to use in the high voltage interfaces. Reed switch Parameter/Type Contact kind Switching voltage, V Switching current, A Switching power, W Dielectric strength, V Operate time, ms Release time, ms Dimensions, mm Sensitivity, AT

MRA 5650G

KSK 1A75

HYR 2016

HYR 1559

MARR 5

KSK 1A85

NO 1,000 1 100 1,500 0.6 0.05 D = 2.75 L = 21 20–60

NO 1,000 0.5 10 1,500 0.5 0.1 D = 2.3 L = 14.2 15–40

NO 1,000 1 25 2,500 0.8 0.3 D = 2.6 L = 21 15–70

NO 1,500 0.5 10 1,500 0.4 0.2 D = 2.3 L = 14.2 15–50

NO 1,000 0.5 10 2,000 0.75 0.3 D = 2.66 L = 19.7 17–38

NO 1,000 1 100 4,000 1.0 0.1 D = 2.75 L = 21 20–60

Bibliography | 221

of less than one ms from penetrating the inputs of electronic relay protection and automation devices. It is recommended to produce the isolating components of all discussed reedswitch interfaces from molding thermoplastic material type ULTEM-1000 (Polyetherimide—PEI). It is the translucent amber-colored material characterized by a distinctive cluster of mechanical, thermal (−55 + 170 °C) and electrical (33 kV/mm, tgδ = 0.0012) features, low water absorption (0.25 % within 24 hours), high tolerance to radiation of various types and the relatively good adhesion to epoxy compounds. Also, it is recommended to use epoxy resin STYCAST 2651-40 (Emerson & Cumming). It is a bicomponent black-colored compound characterized by good dielectric (18 kV/mm, tgδ = 0.02) features, low water absorption (0.1 % within 24 hours), a wide range of operating temperatures (−75+175 °C), very low liquid-state viscosity and good adhesion to metals and plastics. More importantly, within the wide temperature range, the linear expansion coefficient of this compound is close to the ULTEM-1000 coefficient. CATALYST-11 should be used as the solidifying agent. It should be noted that it is prohibited to cover the reed switch immediately with the epoxy resin. Preliminarily, the reed switch must be covered with a layer of damping material to absorb the mechanical stress appearing during the epoxy-resin solidification.

Bibliography [1] Gurevich V. The Optoelectronic Transformers are the Panacea or Particular Solution for Particular Problem Only? Electr. Power News, 2010, Vol. 2 (by Russian). [2] Gurevich V. Grounding of Control Cable Shields: Do We Have a Solution? Energize, 2017, Vol. 4. [3] Gurevich V. Cyber and Electromagnetic Threats in Modern Relay Protection. Taylor & Francis Group, Boca Raton–London–New York, 2015. [4] Gurevich V. Protection of Substation Critical Equipment against Intentional Electromagnetic Threats. Wiley, 2017. [5] Gurevich V. Protection Devices and Systems for High-Voltage Applications. Marcel Dekker, New York–Basel, 2003. [6] Gurevich V. Electronic Devices on Discrete Components for Industrial and Power Engineering. Taylor & Francis Group, Boca Raton–London–New York, 2008.

12 Improvement of the resilience of industrial cabinet-installed electronic equipment to HEMP Impact 12.1 Introduction There are hundreds of types of sophisticated electronic equipment employed by power plants and substations. These include: digital protection relays (DPR), automation, telecommunication and communication devices, which ensure functionality of the modern electric-power industry. These types of equipment are designed based on microelectronic and microprocessor-based devices, which are very sensitive to external electromagnetic impacts. This is especially true for a powerful and devastating highaltitude electromagnetic pulse (HEMP), which creates a pulse of electric field with a strength of up to 50 kV/m at the ground surface. Due to expanded use of microelectronic and microprocessor-based equipment in the electric-power industry (while this is the core of infrastructure of any country of the world), its susceptibility to HEMP is also rising. This is also appreciated by military men who improve the methods of nonnuclear EMP creation, as well as to try to increase the EMP component during a high altitude (30–400 km) nuclear explosion (so called “super-EMP”). Blasting of nuclear ammunition at this altitude does not directly affect the public. However, it can cause the country’s infrastructure to be inoperable for a long time, and consequently it is very attractive to military men. Thus, development of methods and techniques to protect sensitive electronic equipment from HEMP becomes very relevant.

12.2 New cabinets for electronic equipment Previously, electromechanical-protection relays of old power plants and substations were mounted on special panels facing personnel and featuring an open back side, where the circuits were connected (Figure 12.1). Upon transition to digital protection relays (DPR), this principle was initially retained. The same principle was used to mount other types of electronic equipment. However, sometime later, another principle of mounting was introduced: electronic equipment was mounted in special cabinets. Today, this principle of electronicequipment installation is deemed the most promising, convenient and justified. Nowadays, a typical relay room of a power plant or a substation accommodates several rows of clean and beautiful cabinets. However, the problem is that this beauty conceals the lack of functionality because the beautiful cabinets with glass doors (Figure 12.2) do not provide protection of the electronic equipment inside from electromagnetic impact, though the idea of a closed cabinet envisages protection of internal space from adverse external impacts. https://doi.org/10.1515/9783110639285-012

12.2 New cabinets for electronic equipment |

223

Figure 12.1: Relay protection panels: electromechanical (above) and microprocessor-based (below).

Figure 12.2: Contemporary way of electronic equipment mounting in cabinets.

At the same time, there are many specially designed metal cabinets (Figure 12.3) that ensure significant weakening of electromagnetic emissions.

224 | 12 Improvement of the resilience of industrial electronic equipment

Figure 12.3: Specially designed metal cabinets ensuring significant weakening of electromagnetic emissions.

Standard cabinets made of sheet steel, containing no windows or cracks, significantly reduce the electromagnetic pulse. However, the use of galvanized mounting panels are used to produce the cabinets, as well as special conductive sealers and gaskets, which result in significant improvement of their efficiency, since zinc priming enables equalizing the potentials over a large area (electrical resistivity of steel is 0.103–0.204 ohm×mm2 /m, while the electrical resistivity of zinc is 0.053–0.062 ohm× mm2 /m). Aluminum has even lower resistance (0.028 ohm×mm2 /m). That is why some companies produce all-in-one cabinets made of special Aluzinc-150 alloy (Aluzinc® is a registered trade mark of Arcelor Group of companies). This steel has a special covering that consists of 55 % aluminum, 43.4 % zinc and 1.6 % silica. The surface of the cabinet with this covering provides a high level of reflectance of electromagnetic radiance wave. Sarel company (now, Schneider Electric Ltd., UK) manufactures and supplies such cabinets to many countries around the world. Other companies also manufacture similar cabinets which ensure HEMP protection. The list of these companies includes: R. F. Installations, Universal Shielding, Eldon, Equipto Electronics, ATOS, MFB, European EMC Products, Amco Engineering, Addison, Line Technics, Efore, Raymond EMC, ETS-Lindgren, Delancer Shielding, Holland Shielding Systems, and Trusted Systems, among many others. These cabinets usually ensure 80–90 dB

12.3 Retrofitting existing cabinets equipped with glass doors | 225

weakening of emission at 100 kHz–one GHz frequencies. Many of these cabinets are rather expensive, large and heavy (Figure 12.3). They are more suitable to protect military equipment rather than civil facilities. However, multiple protected control cabinets produced by Schneider Electric are mostly acceptable and (regardless of their simplicity and affordability) feature fairly good specifications (Figure 12.4).

Figure 12.4: Schneider Electric’s simple protected control cabinet made of Aluzinc® alloy and its specifications: 1—weakening of vertical component of electromagnetic field; 2—weakening of horizontal component.

12.3 Retrofitting existing cabinets equipped with glass doors The just cited guidelines on how to use special control cabinets are applicable to new installations only. They cannot be used for existing substations with unprotected control cabinets. So, how do we proceed? Luckily, there are all the necessary materials in the market that can significantly improve the level of protection of existing control cabinets without the need to replace them. Firstly, the door glass should be made impermeable for electromagnetic emission. There are several ways to do this: – changing glass to metal in the door; – changing ordinary glass to conductive glass; – priming of ordinary glass with special conductive coating. There are myriad sources of the transparent conductive glass available on the market (to be specific, glass with a conductive external layer). The list of manufacturers includes Latech™, Visiontek Systems, Holland Shielding Systems B. V. and Techinstro, among many others. This glass is produced by adding special substances to the outer layer of the glass during production. Usually, several types of alloying additives are

226 | 12 Improvement of the resilience of industrial electronic equipment used: indium tin oxide (ITO), fluorine-doped tin oxide (FTO) and, rarely, doped zinc oxide. There are also polymer conductive panels based on poluyacetylene, polyaniline, polypyrrole, polythiophen, etc. [1]. The highest shielding effect is provided by three-layer glass with an internal thin copper net layer produced by Holland Shielding Systems B. V (Figure 12.5).

Figure 12.5: Shielding effect of glass with an intermediate copper net layer (1) and glass with conductive outer layer (2).

Alternatively, various conductive varnishes, paints, sprays (also readily available on the market) can be used (Figure 12.6). These can be easily applied on glass doors. Additionally, special door seals made of conductive rubber (Figure 12.7) can be laid along the doors’ perimeter to ensure better protection of electronic equipment cabinets.

Figure 12.6: Conductive varnishes, paints and sprays.

12.3 Retrofitting existing cabinets equipped with glass doors | 227

Figure 12.7: Door seals made of conductive rubber.

In case there are ventilation blinds in the cabinet, they need to be equipped with special panels (Figure 12.8) consisting of a dense metal grid or a set of metal tubes with a certain diameter-to-length ratio (the so-called “waveguides-below-cutoff”), which allow air circulation, but prevent an electromagnetic wave from entering the cabinet.

Figure 12.8: Ventilation panels protected from electromagnetic penetration.

The honeycomb structure panels (waveguides-below-cutoff), see Figure 12.8 (left), are preferable as they are less prone to dust accumulation. They are made of round or rectangular (sometimes hexagonal) section pipes welded alongside. Such ventilation panels are manufactured by many companies: Huaming Electronic Equipment, Holland Shielding Systems, Chomerics, Kemtron, Parker Chomerics, Solany EMC and Major Products, among others. It is commonly known that the hollow metal pipe acts as a waveguide that conducts the high-frequency electromagnetic wave. But in order to have this pipe act as a waveguide, it should have specific geometric dimensions that are related to the wavelength. If the dimensions of the pipe (waveguide) are different, it can cause significant wave decay (up to 80–100 dB). Specifically, it does not conduct the electromagnetic wave. The waveguides that do not conduct electromagnetic waves at a frequency rate lower than the defined value (also known as the cutoff frequency) are called waveguides-below-cutoff. The size of waveguides-below-cutoff (i. e., the size of

228 | 12 Improvement of the resilience of industrial electronic equipment

Figure 12.9: Correlation between the index of electromagnetic emission weakening (K) by waveguides-below-cutoff and their geometrical dimensions and frequency. L—length of round section waveguide; D—diameter of the waveguide; f —frequency of emission; fC —cutoff frequency.

pipes used to produce the honeycomb structure panels in our case) is determined using known formulas, see Figure 12.9. The curves in Figure 12.9 (determined by [2] using these formulas) show that the ability of waveguides-below-cutoff to weaken electromagnetic emissions is maintained over a wide range of frequencies up to the cutoff frequency. In order to ensure the reliable operation of the waveguide-below-cutoff, it is necessary to select this cutoff frequency with a one-and-a-half period margin in relation to the maximum working frequency. Apart from these measures, it is advisable to split the interior of the cabinet into several decks separated by aluminum panels with a minimum number of small openings (cut-outs) for cables. The methods of improvement of HEMP-resistance described are fully applicable to cabinets containing power supply equipment, which consist of battery chargers, converters, etc. This kind of equipment is usually placed into cabinets without any doors or those having many unprotected ventilation blinds.

12.4 Enhancement of the cabinet cable entries A cabinet protected from electromagnetic emission is a very important, but not an exclusive, issue that needs to be addressed in order to protect highly sensitive electronic equipment. Another problem is represented by multiple control and power cables entering the cabinet. These cables act as antennas absorbing HEMP energy from a large area and

12.4 Enhancement of the cabinet cable entries |

229

delivering it to the interior space of a cabinet and directly to the inputs of sensitive electronic equipment. Thus, the cables make it necessary to address at least two additional issues: protection of the cabinet’s interior space from over-emission from cables entering from the outside and their screens (in case they are shielded), as well as protection of the electronic equipment inputs. A design of a new power facility can provide for special cabinets, shielded cables, more efficient placement of equipment inside the cabinets and the mounting of special protection of this equipment from HEMP. However, in the case of existing facilities, the situation is much more difficult because it is not possible to change either the cabinets or the cables. So, let us address this, which is the most difficult situation. There are two main principles of cable entry (Figure 12.10) into control, relay protection and communication cabinets: when external cables are laid through the whole cabinet and then they are connected directly to the electronic equipment’s terminals; and when external cables are connected to intermediate terminals at the cabinet’s entrance. Cable harnesses entering the cabinets may consist of shielded and unshielded cables; cables delivering high-frequency signals and cables delivering analogue DC signals; as well as DC and AC power-supply cables. The problem is that it is impossible to use a certain common protection for all cables due to a variety of signals delivered by them. Consequently, the first thing to do is to split a common harness into individual harness groups with similar signals. Then, if we deal with option (a)—see Figure 12.10—we need to provide separate shielding for each of the separated groups of external cables laid inside a cabinet. This can be achieved by a detachable flexible screen, containing conductive fabric with a longitudinal zip (produced by Kemtron) or by wrapping separate cable harnesses with conductive tape having an adhesive layer (Figure 12.11). The openings in the cabinet’s bottom (at the cables’ entrance) should be tightly sealed with conductive fabric (Figure 12.12). If these openings are large, place a metal grid underneath to hold the fabric in place. DC or AC power supply cables are the most problematic from the standpoint of electromagnetic compatibility. Thus, they need to have protection facilities ensuring the highest level of HEMP weakening. This facility includes high-frequency chokes, embedded into positive and negative cores (or Live and Neutral), see Figure 12.9, as well as voltage pulse suppressors. The chokes are small and should be located at the cable’s entry into the cabinet (i. e., on the cabinet’s bottom) underneath the conductive fabric. These chokes (Table 12.1, Figure 12.13) have very low DC resistance (milliohm) and high impedance for a short EMP pulse. One choke encapsulated with epoxy compound costs about 100 USD. The inductance of these chokes necessary for the effective suppression is calculated taking into account its reduction during the flow of direct current through them. This influence is the reason why it is recommended to use a special type of choke (Figure 12.13), which requires cutting the cable, and not some other type that does not

230 | 12 Improvement of the resilience of industrial electronic equipment

Figure 12.10: There are two main principles of cable entry into a cabinet: a) external cables are laid through the whole cabinet and then they are connected directly to electronic equipment’s terminals; b) external cables are connected through intermediate terminals at the cabinet’s entrance.

require such cuts. For the cabinets pictured in Figure 12.10b, this poses no problem. For the cabinet in version Figure 12.10a, it will be necessary to cut the cable where it is inserted into the cabinet for connecting the choke. However, considering that we are talking only two cores (“+” and “-” of DC power supply), this should not be a problem for option Figure 12.10a either. Portable ferrite filters that do not require cutting of cables (Figure 12.14) can successfully be used for other groups of cables with low-frequency signals (in the range of up to several kilohertz) and with analogue DC signals (4–20 mA), with low currents of industrial frequency and with logic signals of relay protection.

12.4 Enhancement of the cabinet cable entries |

231

Figure 12.11: Flexible detachable screen with a zip containing conductive fabric (above) and different types of copper, aluminum and graphite conductive tapes with an adhesive layer (below).

Figure 12.12: Conductive fabric.

Figure 12.13: Chokes produced by CWS Company with a helical type winding and a core made of special alloy to weaken EMP pulse in a power-supply cable. Open (left) and encapsulated with epoxy compound (right).

Moreover, these filters should be mounted in those places where cables enter the cabinet, in other words, on its floor and under the conductive fabric. The features of these filters and the methods of their selection have been discussed in [3]. It would be appro-

232 | 12 Improvement of the resilience of industrial electronic equipment Table 12.1: Main parameters of a high-frequency choke for supply circuits of cabinets with electronic equipment for supervision, control, relay protection and telecommunication. Type

EK55246341M-40AH

Current Rating, A

Inductance at DC Bias 40 A, µH

40

162

Impedance DC, 1 MHz mΩ (calculated), kΩ 14.5

1 GHz (calculated), MΩ

1.0

1.0

Dimensions, mm

58 × 58 × 35

Figure 12.14: Portable ferrite filters that do not require cutting of cables.

priate to remind the reader that this means that three filters with different frequency specifications are placed onto the same group of cables. These three filters would provide the most efficient weakening of EMP in the required frequency range. Ferrite filters with the required parameters are produced by various companies. Nonetheless, the most appropriate units that ensure cable protection from EMP are those produced by Fair-Rite Corp. and Wurth Elektronik (Table 12.2), even though they are far from inexpensive. The previous table shows data for filters with maximum internal diameter, because using a filter for a group of cores (cables) is the most cost effective. A good example would be a set of three filters of the first type (from Table 12.2) that provide efficient EMP weakening in the required frequency range, when mounted on each group of cores (cables).

12.5 Voltage pulse suppression

| 233

Table 12.2: Specifications of ferrite filters with maximum internal diameter to be mounted on control cables entering protected cabinets. Catalog Number Filter Kind

Fair-Rite Corp. 0475176451 0431176451 0461176451 0431177081 2631181381

Round Cable Snap Assembly Split Round Cable Core

Wurth Elektronik 74271251 Round Cable Snap Assembly Laird LFB360230-300 Non-Split 28B1417-200 Round Cable 28B2400-000 Core

Cable Diameter Maximal, mm

Frequency Range, MHz

Minimal Impedance, Ω

Maximal Impedance, Ω

Cost, USD

18 18 18 25 35

0.2–5.0 1–300 100–1,000 1–250 1–300

30 47 110 45 60

130 380 480 375 530

17 8 9 13 13

25

10–1,000

110

160

31

23 23 35.5

0.5–5 25–300 25–300

24 52 60

19 230 285

4 1.7 6

In case the cables are connected as depicted in Figure 12.10b, it would be much easier if the cores were disconnected from the terminal for the period of ferrite-filter mounting (at the cable’s entry near the terminal). In this case, inexpensive, non-split ferrite rings and cylinders (Table 12.2) can be used. It should be kept in mind that the filters should not be used in low-power communication circuits that work in the same frequency range as the filters.

12.5 Voltage pulse suppression Another common problem that needs to be solved to protect any type of electronic equipment from EMP is suppressing of the pulse overvoltage at its input terminals. Not all types of nonlinear elements used to protect from switching or lightning overvoltage are suitable for HEMP protection because this pulse is very short and can increase rapidly. The most suitable (in terms of features) are those elements that implement avalanche breakdown of p-n junction in solid-state devices. This is the principle adopted in so-called Transient Voltage Suppressor Diodes (TVS-diodes). Until recently, these elements had some limitations in terms of their power and that is why there were not used in power circuits. Nevertheless, two renowned competing companies—Littelfuse and Bourns—started manufacturing powerful TVS-diodes suitable for this purpose (Table 12.3).

234 | 12 Improvement of the resilience of industrial electronic equipment Table 12.3: Main parameters of some types of powerful TVS-diodes. Type

Littelfuse AK10-430C AK6-430C AK3-430C Bourns PTVS10-430C PTVS6-430C PTVS3-430C

Nominal Voltage, V

Avalanche Breakdown Voltage, V

Clamping Voltage, V

Peak Pulse Current, A

430 430 430

440–490 440–490 440–490

625 625 625

10,000 6,000 3,000

430 430 430

440–490 440–490 440–490

580 580 580

10,000 6,000 3,000

Unfortunately, these components are still produced for mounting on printed circuit boards only, rather than as separate devices in a plastic body to be mounted in a control cabinet on a standard DIN-rail, with screw connections to external cables. But this problem can also be solved by using a simple and small plate with printed wiring (Figure 12.15).

Figure 12.15: Multi-channel voltage pulse-suppression device containing TVS-diodes and standard terminals mounted on a small printed circuit board.

Recommended types of TVS-diodes: AK10-430C (Littelfuse) or PTVS10-430C (Bourns) to protect AC and DC power supply circuits 230–250 V and TVS-diodes: AK3-430C (Littelfuse) and PTVS3-430C (Bourns), to protect supervision and control circuits, as well as logic DPR inputs operating at 250 V. These elements are not inexpensive (about 120 USD for the first group and about 50 USD for the second), so they should only be used

12.5 Voltage pulse suppression

| 235

to protect critical types of equipment located at critical facilities of the electric-power industry. In order to protect the cabinet’s equipment mounted as shown in Figure 12.10b, the printed circuit board with TVS-diodes needs to be installed near the input terminal, whereas the outputs of the diodes should be connected by means of the shortest possible conductors to terminals parallel to input cores of cables of the same circuits (differential mode). In order to protect the cabinet’s equipment mounted as shown in Figure 12.10a, the printed circuit board with TVS-diodes needs to be installed near the input terminals of the internal equipment and connected to them following the same principle. Moreover, long input cables running through the cabinet must be shielded by means shown in Figure 12.11. Unfortunately, the described powerful TVS-diodes cannot be used to protect telecommunication equipment located in the same cabinets due to the high capacitance of their p-n-junctions. Indeed, connection of these diodes to low-power high frequency circuits of telecommunication systems can result in their malfunctioning and thus it is unacceptable. To address this issue, a protection device has been designed (Figure 12.16), which employs other elements (Table 12.4, Figure 12.17 [4]). This device is connected parallel to the inputs of telecommunication equipment being protected and features relatively a low price (about 3 USD for a suppressor and about 2 USD for a resistor).

Figure 12.16: Protection device for the low-voltage high-frequency circuits of telecommunication equipment. Table 12.4: Main parameters of SP03-6 transient voltage suppressor. Max. Stand-Off Voltage, V Min. Avalanche Breakdown Voltage, V Max. Leakage Current, µA Max. Clamping Voltage, V Peak Pulse Current (8/20 µs), A Peak Pulse Power (8/20 µs), W Max. Junction Capacitance, pF

6 6.8 25 20 150 2,800 25

236 | 12 Improvement of the resilience of industrial electronic equipment

Figure 12.17: Components used in the protection device for telecommunication systems: Above— Transient voltage suppressor, type SP03-6 (Table 12.4), below—high voltage low-resistance antisurge resistors of various types, see their parameters in Table 12.5. Table 12.5: Main parameters of some types of high-voltage anti-surge resistors. Type

Manufacturer

ASRM2 PPR200 RT/RL AZ SPO250

Stackpole Electronics Firstohm HVR ARC Engineering Ohmite HVR International

Max. Pulse Voltage and Power

Dimensions, mm

5 kV; 2 W 20 kV; 2 W 7.5 kV; 2 W 2.5 kV; 2 W 5.7 kV; 2 W

Dia. = 5; L = 15 Dia. = 5; L = 13.5 Dia. = 8; L = 18 Dia. = 26; L = 30 Dia. = 13; L = 15

12.6 Retrofitting grounding systems of electric cabinets The issues of a conventional grounding system of electronic equipment mounted in electric cabinets and specifications of its grounding under HEMP impact has been addressed in [5] in full detail. Herewith, the reader is be reminded that a conventional grounding system does not provide HEMP protection, but rather acts as a source of

12.7 Conclusion | 237

dangerous pulse impact on the electronic equipment. For this reason, it should be designed as stipulated in the recommendations suggested in [6].

12.7 Conclusion Information and specific recommendations suggested in this article, as well as technical data of elements and materials, make it possible for developers and engineers to design power facilities with their electronic equipment protected from HEMP. It also enables upgrading and retrofitting of existing equipment, while incurring affordable costs, and thus, to significantly improve the level of power facilities’ protection from HEMP.

Bibliography [1] Skotheim, T. A. Reynold, J. Handbook of Conducting Polymers, CRC Press, 1998. [2] Ivko A. Screening of Radio-Electronic Equipment as Means of Electromagnetic Compatibility Implementations. Modern Electron., 2015, Vol. 8, pp. 86–90 (by Russian). [3] Gurevich V. The Problems of Correct Choice of Ferrite Beads. Electr. Eng. Electromech., 2016, Vol. 2, pp. 71–73. [4] Gurevich V. Protection of Telecommunication Systems in Electric Power Facilities from Electromagnetic Pulse (EMP). Int. J. Adv. Comput. Technol., 2017, Vol. 6, No. 9, pp. 2446–2450 (Vol. VI, Issue IX). [5] Gurevich V. Is the Electric Equipment Grounding the Basic Protection Means Against HEMP? Int. J. Res. Stud. Electr. Electron. Eng., 2018, Vol. 4, No. 1, pp. 1–8. [6] Gurevich V. The Issues of Electronic Equipment Grounding at the Power Facilities. Int. J. Res. Stud. Electr. Electron. Eng., 2017, 3, No. 1, pp. 11–19.

13 Basic principles of direct-current auxiliary-power system (DCAPS) protection 13.1 Introduction A direct-current auxiliary-power system (DCAPS) is the most important component of any substation. All other substation systems and equipment (such as power equipment, relay protection, automation, control, communication, emergency, etc.) rely upon its operability. DCAPS failure makes the whole substation completely inoperable and “invisible” for the central control room. Therefore, DCAPS—above all others— needs the special facilities to ensure its operation if impacted by HEMP. Such facilities can be categorized as follows: 1) facilities protecting operational equipment against HEMP; and 2) DCAPS redundant power supplies starting after HEMP in order to restore substation operability upon DCAPS failure.

13.2 Protection of DCAPS operating equipment from HEMP Primarily, the special protection measures are required for electronic battery chargers (BCs) supplying power to DC-current-carrying buses, feeding numerous end users and ensuring battery floating charge. A regular BC is the metal cabinet containing many electronic elements and cables (input AC-supply triple cable, output two-core DC cable and emergency control cable). From the protection point of view, such a cabinet is basically identical to a control, relay protection or an automation cabinet, therefore it can be protected from HEMP using the same means and methods as was proposed in [1–3] earlier. Briefly, it should be noted that such means include measures to improve the cabinet’s screening ability, special screened control cables, HEMP ferrite beads installed on control cables, surge overvoltage-protection elements and improvement of the cabinet ground system, including usage of the so-called “special floating ground” and grounding capable of being disconnected [4]. Considering the fact that besides BCs, DCAPS also contains a DC control power distribution panel, AC distribution panel feeding the BC and a battery bank, the DCAPS protection must extend beyond the BCs cabinet protection. Furthermore, DC current flowing through the power cables creates a strong neutralizing impact on inductive elements of standard HEMP filters and significantly reduces their effectiveness. Consequently, DCAPS protection has certain peculiar features. Therefore, the following additional measures for DCAPS protection can be suggested: 1. Special chokes acting as HEMP filters in power cables should be connected to the BC and battery bank. https://doi.org/10.1515/9783110639285-013

13.2 Protection of DCAPS operating equipment from HEMP

2. 3.

| 239

Block varistors with high dissipated-pulse energy should be installed on DC buses and in the triple AC power circuit of the BC. Block varistors with medium dissipated pulse energy should be installed in the BC cabinet and on the battery bank.

One possible type of DCAPS scheme equipped with the mentioned protections is shown in Figure 13.1.

Figure 13.1: Layout of protection elements in DCAPS. 1—DC distribution panel; 2—AC distribution panel; 3—battery bank; 4—charger (BC); R1–R8—block varistors with high dissipated energy; R9– R15—block varistors with medium dissipated energy; L5, R16, K—elements of the grounding system of the cabinet and internal zero potential circuit (local ground) of charger (so-called “special floating ground”, in detail presented in [4]).

Here, the chokes installed in the battery-bank circuit limits the HEMP current-pulse amplitude, which otherwise can be too high due to the very low internal impedance of batteries. The need to install the chokes in the BC–DC circuit arises as the external DC circuit is connected directly to the inner rectifier bridge of the BC through thyristors, which are prone to failure upon the high-amplitude current pulse if the chokes are

240 | 13 Basic principles of direct-current auxiliary-power system (DCAPS) protection not installed. Such chokes can be less powerful than the chokes used in the batterybank circuit. External AC cables are connected to the BC input transformer windings, characterized by the very high impedance upon the very short current pulse, so such a circuit does not require any additional current-limiting chokes. These chokes can be of lower power than those used in the battery circuit. Diode suppressors are mounted on a separate printed circuit board and are equipped with terminals for connecting external conductors, see Figure 13.2, by means of which the suppressors are connected in parallel with the input cables on the main terminal.

Figure 13.2: Installation of high-power TVS-diodes on a small printed circuit board with standard terminals.

As was shown in the previous chapter, much less expensive varistors can be used instead of diode compressors. External AC cables are connected to the windings of the input-transformer BC, which have a high impedance for a very short current pulse, so there is no need for additional current-limiting chokes in this circuit. All other activities must comply with the just-cited general recommendations for the protection of electronic equipment located in cabinets.

13.3 Backup-power supplies for DCAPS systems DCAPS backup-power supply is designed to restore the DCAPS operability upon HEMP in case of failure of operating DC current feeding the BC, or in case of damage of some DCAPS components despite the implemented protection measures. Two alternatives for independent redundant power supplies, with minimum power required to ensure operability of critical DCAPS loads of 5 kW are considered: 1) based on the fuel cell, and 2) based on the diesel generator.

13.3 Backup-power supplies for DCAPS systems | 241

Fuel cell (FC) is the chemical source of current converting the fuel energy directly into the electricity without the combustion stage. As for the amount of energy, the FC is significantly more effective (efficiency up to 40–50 %) compared to combustion engine (efficiency up to 12–15 %). Besides, the fuel cell is environmentally friendly as it does not inject contaminators into the atmosphere in contrast to the combustion engine. But, are these useful FC properties important in case of redundant power sources, since they are regularly inactive and actuated only upon the emergency and for a limited time only? The answer is quite obvious. On the other hand, the FC has certain drawbacks. The most common types of FC use the reaction of water formation to generate the energy (the hydrogen contained in electrochemical cell reacts with oxygen contained in the air). However, hydrogen is an explosive gas requiring the application of special storage and transportation methods, as well as the availability of airborne concentration transducers and certain approvals. Special metal-hydride containers, see Figure 13.3, used instead of regular hydrogen balloons, substantially increase the cost of the plant, which is already several times more expensive than the diesel-generator set of the same power.

Figure 13.3: Author standing at a 5-kW fuel cell (left) and hydrogen-storage metal-hydride containers (right).

If the FCs and hydrogen balloons are installed outside the building, the protection against unauthorized access, acts of terrorism and property abuse should be ensured in order to prevent a cataclysmic plant explosion. Also, the FC contains many electronic transducers, microprocessors and other microprocessor-based devices controlling its operation and preventing emergency modes. Typical 5-kW FC generates low-DC voltage (about 40 V). Then, it should be converted into standard 230VAC using the electronic inverter, or into standard 250VDC

242 | 13 Basic principles of direct-current auxiliary-power system (DCAPS) protection using the converter. Due to the numerous complex electronic systems, the FC is highly vulnerable to electromagnetic impacts—both to single HEMP and remote impacts of independent directed sources of high-frequency energy. Altogether, it is necessary to install FCs and balloons outside the building in a special thick-walled steel box, ensuring explosion protection of the FCs upon various intentional destructive mechanical and electromagnetic impacts, see Figure 13.4.

Figure 13.4: The FCs with balloons, inverter, and batteries enclosed in special steel box of 2 m × 3 m × 2 m: 1—set of standard 150-atm hydrogen balloons; 2—electrochemical cell; 3—inverter and batteries.

Furthermore, an electrochemical cell of 5-kW FC contains more than 20 liters of electrolyte, which is usually characterized by high reactivity and chemical hazard (such as concentrated KOH solution). During the long periods of electrochemical cell inoperability, such a chemical reagent must be stored separately and poured into the cell just before its activation. Moreover, before the FC activation, its electronic systems should be started using a battery with a constant functioning charger. All these features, or rather problems, make the FC hardly suitable to be used as a substation DCAPS redundant power supply. In contrast to the FC, a regular 5–7-kW diesel generator is a compact, low-cost, low-maintenance device and thus a perfect choice to be used as a DCAPS redundant power supply, see Figure 13.5. Such diesel generators are equipped with manual starters and do not require a battery. They can be easily mounted inside the substation building, requiring only an exhaust stack going through the building wall, see Figure 13.6. Compact 5–7 kW diesel generators do not contain electronic control and monitoring systems, therefore they are much less vulnerable to HEMP compared to the FC. However, considering the fact that HEMP generates an electric-field strength up to 50 kV/m near the ground surface, such diesel generators must be protected against its coils and wire-insulation failure. A light metal container with a tight cover (door) made of 5–6 mm thick aluminum is the perfect choice to ensure the protection of a standby

13.3 Backup-power supplies for DCAPS systems | 243

Figure 13.5: Compact diesel-generator sets of 5–7 kW.

Figure 13.6: Installation of a compact diesel generator inside the substation building.

diesel generator stored at the substation. During the storage period, all external cables and the external exhaust stack must be disconnected from the diesel generator. A 5-kW diesel generator consumes only about one liter of fuel per hour. Therefore, one plastic container with 200 liters of fuel, see Figure 13.7, ensures a week of uninterrupted operation, while a 1,000-liter container ensures a month of uninterrupted operation.

244 | 13 Basic principles of direct-current auxiliary-power system (DCAPS) protection

Figure 13.7: Plastic containers used for storing diesel fuel.

In order to improve the availability, a redundant power supply based on a diesel generator can also be equipped with a standby BC (to replace the main BC upon failure). Compact low-power BCs with a high-frequency intermediate link are manufactured by numerous companies. For example, two compact BCs Series ADC7480 manufactured by the Finnish company Powernet, see Figure 13.8, connected in parallel and equipped with all sorts of systems for stabilization, adjustment, protection and load sharing between devices, and energized by 5–7-kW diesel generator, can supply enough power to the medium-size substation, or ensure operation of critical equipment and keep the battery bank operable at the large substation.

Figure 13.8: Battery charger Series ADC7480 with output voltage of 0–300 V and output current up to 14 A.

Such a BC should be stored in a disconnected state in a diesel-generator container. The cost of the redundant power supply based on a 5-kW diesel generator and equipped with a double BC with the total current up to 25 A costs approximately 10,000 USD, which is near 10 times less than the cost of the FC-based power supply of the same output power. Moreover, if you use diesel generators, you do not need any special hydrogen-storage facilities or installation and operation approvals. The cost of one choke and one block varistor is approximately 30–50 USD.

13.4 Mobile substations and features to protect their DCAPS from HEMP

| 245

13.4 Mobile substations and features to protect their DCAPS from HEMP In addition to stationary substations, the use of mobile substations (MS) is widespread in the world practice, too. Figure 13.9. MS direct-current auxiliary-power system (DCAPS) differs from the stationary substation DCAPS, and such differences must be considered during development of survivability measures.

Figure 13.9: Several types of mobile substations.

One of the differences is a very limited space the DCAPS of MS occupies. This particularly explains the application of nominal DC control voltage of 60 V instead of 250 V, as well as the usage of a 60-V auxiliary-power supply, consisting of in-series 12-V compact sealed batteries for MS DCAPS in-feeding. However, five in-series 12-V batteries require a floating-charging voltage of approximately 67 V, as against 60 V. On the other

246 | 13 Basic principles of direct-current auxiliary-power system (DCAPS) protection hand, some types of electronic equipment (such as communication and data-transfer systems) used in MS are designed to operate under nominal voltage of 48 V, with the possibility to increase it up to a 60-V maximum, therefore the battery charger (BC) must provide two output voltages: 67 V for battery charging and 60 V to supply power to electronic equipment installed in MS. Such BCs built on an adjustable thyristorbased diode bridge were manufactured and previously used in MS. They were complex, expensive and large. Later, switched-mode power supplies with high-frequency links enables the development of very small BCs (Figure 13.8) and improve the MS DCAPS survivability. One option of protected MS DCAPS design is shown in Figure 13.10. Powerful diode suppressors TVS1–TVS3 are mounted on the same simple printed circuit board, as shown in Figure 13.2. Moreover, TVS1 is of the same type (AK10-430C or PTVS10-430C), and TVS1 and TVS2 suppressors can be of the type AK6-076 or PTVS6-076 (76 V, 6 kA). Instead of suppressors on the same printed circuit board, much lower-cost varistors can be mounted.

Figure 13.10: One option of MS DCAPS design. CH1 and CH2—battery chargers; VL—voltage limiter; R1–R3—high-power block varistor; L1—high-frequency chokes; L7, K, R4—elements of cabinet and electronic module enclosure grounding, so-called “floating ground”, detailed in [4].

13.4 Mobile substations and features to protect their DCAPS from HEMP

| 247

Figure 13.11: Compact BCs with intermediate high frequency link. Variable output voltage: 60 V or 67 V, output current up to 30 A.

Figure 13.12: A voltage limiter (VL) built on a high-power Darlington transistor (VT1) and the appearance of transistors with varied power rating.

As we see from this arrangement, DCAPS consists of two BCs (the first outputs 67 V and the second outputs 60 V) and a voltage limiter (VL) ensuring a 67-V voltage decrease to the safe level of 60 V, see Figure 13.10. A VL can be simple in design and built on a base of a high-power Darlington transistor, see Figure 13.12. Such transistors with a maximum current up to 100 A and a voltage of 300– 600 V are mounted in different enclosures designed for different dissipated energy. For example, a transistor type ESM3030DV dissipates up to 225 W, a transistor type QM100HY-H dissipates 620 W, and a transistor type KS624540 dissipates 1,500 W. When selecting a transistor, the voltage redundancy must be ensured to improve

248 | 13 Basic principles of direct-current auxiliary-power system (DCAPS) protection the resistance to surge overvoltage and to ensure the coordination with metal-oxide varistor (MOV) parameters. Selected output voltage of a VL is a little lower than 60 V (57–58 V, for example) to keep it in a closed mode (de-energized) by reverse voltage 60 V of BC CH1, and exclude it from the DCAPS operation when connected in parallel with CH1 of 60-V output voltage. The VL comes into operation only upon the AC control power failure and supplies the power to consumers from the battery bank. Given that, in the absence of the floating charge 67 V from CH2, the battery-bank voltage decreases immediately, the requirements of the VL dissipated energy are not that high, and even the small transistor in ISOTOP-type enclosure (e. g., type ESM3030DV, see Figure 13.12) installed on a radiator can safely cope with this task. Alternatively, DCAPS can be built on a set consisting of a battery charger 230/67 V, a DC/DC converter 67/60 V and battery bank connected between them, see Figure 13.13.

Figure 13.13: The design of MS DCAPS based on BC 230/67 V (AC/DC) and converter 67/60 V (DC/DC). L1–L6—chokes; RU1-RU3—varistors; L7, K, R4—elements of cabinet and electronic module enclosure grounding, so-called “floating ground”, detailed in [4].

13.4 Mobile substations and features to protect their DCAPS from HEMP

| 249

In this scheme, the TVS1-type suppressor AK10-430C or PTVS10-430C, TVS2 and TVS3 suppressors are of the type AK6-076 or PTVS6-076. These elements can be mounted on a simple printed circuit board, as shown in Figure 13.2. On the same board can be mounted much lower-cost varistors. Very different kinds of battery chargers and DC/DC converters for 1.5 kW of power are available on the market (Figure 13.11). From very low-cost construction (160 USD) with fan cooling, produced by Chinese companies (a), to more expensive (700 USD), but also with fan cooling and rectangular characteristics of current limiting (b), and up to very expensive (5,000 USD), with natural convection cooling and a wide range of operating temperatures more suitable for battery charger output current characteristics, etc., Figure 13.14.

Figure 13.14: Three kinds of battery chargers and DC/DC converters: a) very cheap (Guangzhou DongLong Electronics Co., China, Hangzhou Reacher Technology Co., China), b) inexpensive (Absopulse Electronics Ltd., Canada); c) very expensive (BC type B4786 and converter type C4746, produced by Powerbox Pty Ltd., Australia).

For a correct choice of a BC type, its current limiting characteristic (current regulation mode) must be considered, see Figure 13.15. To use the BC in the floating mode (voltage-stabilization mode) only, any type of characteristics may be used. But for initial charging of a discharged accumulator battery, with very low internal impedance, “C” type characteristics that determine voltage and current for each point would be much more suitable. Concerning “B” or “D” type characteristics, the output voltage of BC may quickly drop to zero at current reach to limit setting point (with these characteristics, the voltage level for this point is not determined at all and may reach zero).

250 | 13 Basic principles of direct-current auxiliary-power system (DCAPS) protection

Figure 13.15: Most widely used kinds of current regulation modes (current limiting characteristics) for switching mode based power supply, chargers and converters. A—Fold-Back; B—Constant Current Mode; C—Fold-Forward; D—Constant Power.

From this state, the BC may convert to “Hiccup Mode” (interrupted charging). Therefore, the possibility for the BC of a specific type to work with a fully discharged battery must be examined before continuing. In addition to the measures and circuit solutions described in this chapter, it should be recalled that all recommendations for the protection of cabinets with electronic equipment and recommendations for grounding remain in force for the type of equipment in question. In addition, on compact mobile substations with their very small area and a limited number of control cables, an additional measure of protection from EMP can be the placement of internal cables in flexible metal hoses, Figure 13.16.

Figure 13.16: Flexible metal hoses of different diameters designed for routing internal MS cables.

One of the effective organizational measures aimed at increasing the stability of the DCAPS of mobile substations can be the creation of a stock of plug-in modules (see Chapter 20 of this book). In this case, these spare plug-in modules are a spare BC and a converter (or two BCs and a voltage limiter) in a carefully closed aluminum container directly on the substation.

13.5 Direct-current auxiliary-power systems of power plants | 251

13.5 Direct-current auxiliary-power systems of power plants The systems of DCAPS on power plants fundamentally do not differ from DCAPS substations. However, the power of the BC, see Figure 13.17, and the capacity of batteries in power plants is usually much higher than in substations, which is associated with a much larger number of powerful DC consumers. Some design features of the cabinets of charge-charging units, see Figure 13.17, do not change anything in the general principles of ensuring the protection of equipment from HEMP and in the choice of hardware for this. For example, the controllers installed on the doors must either be moved to the interior of the cabinets or covered with additional metal shields. A current-limiting choke in the power circuits of chargecharging units with significantly more power should be chosen than for the same units installed in substations. However, this is not a problem, since such chokes are also produced by the company CWS, see Figure 13.18.

Figure 13.17: Design of power plant’s BC. The figures indicate the “weaknesses” of cabinets that allow unobstructed access of HEMP inside the cabinet.

Figure 13.18: High-power cored HCS series choke for currents up to 1,000 A manufactured by CWS Company under special technology using special material.

252 | 13 Basic principles of direct-current auxiliary-power system (DCAPS) protection Table 13.1: Key parameters of powerful HCS Series chokes with a core. Choke Type

HCS-631M-450A HCS-301M-1000A

Max. Continuous Current, A

Inductance at Current 300 A, µH

450 1,000

249 244

Resistance (Impedance) DC, 1 MHz, 1 GHz, mΩ kΩ MΩ 4.5 4.5

1.56 1.53

1.56 1.53

Dimensions, mm 393 × 118 × 97 431 × 228 × 101

Bibliography [1] Gurevich V. Protection of Substation Critical Equipment against Intentional Electromagnetic Threats. Wiley, London, 2017, 240 p. [2] Gurevich V. Cyber and Electromagnetic Threats in Modern Relay Protection. Taylor & Francis Group, Boca Raton, 2015, 205 p. [3] Gurevich V. Accessible Methods Resilience of Power System Electronics to HEMP. Int. J. Res. Stud. Electr. Electron. Eng., 2016, Vol. 2, No. 2, pp. 13–18. [4] Gurevich V. The Issues of Electronic Equipment Grounding at the Power Facilities. Int. J. Res. Stud. Electr. Electron. Eng., 2017, Vol. 3, No. 1, pp. 11–19.

14 Protection of telecommunication systems in electric power facilities from HEMP 14.1 Introduction Telecommunication systems of electric power facilities play important roles in data transmission, telemetry, remote control and communication. At the same time, these systems are the most sensitive to, and the least protected from, high-altitude electromagnetic pulse (HEMP), compared to other important electric and electronic systems used in the electric power industry. This situation cannot be accepted as normal and thus calls for corresponding actions. Unfortunately, the most well-known technical means are often very expensive, and regardless of advertised claims, they do not always ensure reliable HEMP-protection of telecommunication systems The high-altitude electromagnetic pulse from a nuclear explosion (HEMP), which creates an electric-field pulse with a density of up to 50 kV/m at the ground surface, is a powerfully affective action aimed at rendering the country’s infrastructure to be inoperable. The military leaders of various countries perceive HEMP as a very efficient potential type of non-lethal weapon. On the one hand, the widespread use of microelectronic and microprocessor-based equipment in all types of modern systems (primarily those of electric power industry) causes great vulnerability to such infrastructural impact, whereas on the other hand, this equipment in civilian sectors is completely unprotected [1]. Among the various systems used in the electric power industry, the telecommunication system, which includes data transmitting, telemetry, remote control and communication systems, features the highest level of susceptibility to HEMP. In fact, modern sophisticated electronic telecommunication equipment uses very low (compared to other electric systems used in the electric power industry) voltages (not exceeding 3.5–5 V). Thus, it has a low level of insulation of all the input and output circuits. That is why the usual EMC standard requirements for low-voltage electric and electronic equipment that stipulate (among other things) testing with high-voltage pulses (2 kV and 4 kV) are not applicable to telecommunication ports, if they are available in these devices. This provision is included in all the standards that stipulate these tests. Telecommunication channels of the electric power industry are used to transfer real-time data regarding emergency modes between digital protection relays, and to perform remote control of high-voltage circuit breakers that determine the status of electric grid. Thus, the relevance of HEMP-protection of telecommunication systems becomes obvious.

https://doi.org/10.1515/9783110639285-014

254 | 14 Protection of telecommunication systems in electric power facilities from HEMP

14.2 Ways to solve the problem In the majority of situations, the issue has been resolved by conversion from galvaniccoupled circuits and copper-conductor cables to fiber-optics communication line (FOCL); nevertheless, the problem of protecting multiple microprocessor based terminal devices that convert electric signals into optical and vice versa is still relevant. However, there are electric power facilities in which telecommunication systems equipped with highly-sensitive electronic devices with galvanic-coupled circuits are still employed. This raises a question regarding the modes of re-designing of such systems in order to improve the level of its protection from HEMP. There are some common measures to improve HEMP-resistance of equipment. These do not depend on the selected mode of re-designing. Predominantly, these measures deal with upgrading electronic equipment cabinets that provide efficient protection of internal equipment from electromagnetic emission. Additionally, they ensure protection and backup of the power supply system. These measures and the means of their adoption were discussed earlier [2]. This article discusses the technical means that protect data transmission channels.

14.3 The use of fiber-optic communication lines An obvious solution would be to equip the existing telecommunication equipment with optical links that consist of converters of incoming and outgoing electric signals into optical signals and vice versa, and to transfer these signals between the converters via a fiber-optic cable. Various converters of electric signals into optical ones, and vice versa, suitable for telecommunication systems are readily available on the market (Figure 14.1). Thus, the problem of protection of data-transfer channels using these converters can easily be solved.

Figure 14.1: Various converters of electric signals into ones, and vice versa, suitable for telecommunication systems.

14.4 Protection telecommunication equipment with galvanic couplings |

255

14.4 Protection telecommunication equipment with galvanic couplings It is more challenging to protect the existing telecommunication equipment with galvanic coupling via copper-conductor cables. Standard HEMP-protection (stipulated by standards and offered by multiple manufacturers) for this equipment is composed of special filters that efficiently suppress electric signals above a certain frequency level. However, [3] suggests that the use of special expensive filters to suppress a single short pulse lasting for parts of microseconds is absolutely unnecessary. Additionally, the frequency range of many modern communication and data-transfer systems falls within the HEMP spectrum that which should be suppressed by these filters, whereas the filters themselves are often represented by low-voltage devices that do not allow application of high-voltage pulses to their inputs. Thus, telecommunication equipment needs to be protected from the impact of high-voltage pulses only. There are devices incorporating the elements that significantly reduce their impedance in case of higher (compared to nominal) voltages applied to them. They protect electronic equipment from high-voltage pulses and include: – Gas Discharge Tubes (GDT); – Metal Oxide Varistors (MOV); – Thyristor Surge Suppressors (TSS; Sidac); – Transient Voltage Suppressors (TVS-diode). Comparison of the best in class (based on our survey) elements based on the aggregate of key parameters that make them appealing for use in telecommunication systems is provided in Table 14.1. Response time (reaction time) of the element is one of the most important indicators, which is rarely explicitly indicated in catalogs. This is caused with many reasons, in particular, with the dependence of this time on the speed of voltage pulse rise and on the shape and the lengths of leads of specific elements. If this time is indicated in catalogs, it does not make a lot of sense because the manufacturers often use the Table 14.1: Some main parameters of protective elements of various kinds. Parameter\Kind (group) of element Best type of element in the kind (group) Max. Operating voltage, V Min. Activation voltage, V Residual (clamping) voltage, V Max. Pulse power, W Max. Pulse current, A (2/10 µs) Reaction time Capacitance between electrodes, pF

GDT 2020-15T

MOV V05E1 1P

TSS TISP 4011H1 BJ

TVS S03-6

– 60 (650) 52 – 4,000 – 2

11 18 36 – 500 – 1,300

5.25 10.5 3 – 500 – 110

6 6.8 15 2,800 150 – 25

256 | 14 Protection of telecommunication systems in electric power facilities from HEMP semi-finished product (in fact, they use the material from which the element is manufactured without leads and covering) to reduce it. Furthermore, the response time of the element in real circuits will depend on the parameters of a circuit that it is protecting. It is known, however, that TVS-diodes feature the best response time (several nanoseconds). In regards to response time, they are followed by thyristor surge suppressors with their dozens of nanoseconds, followed by varistors with response times of several dozens of hundreds of nanoseconds. The last in this series are gas discharge tubes (GDT) with a response time of 0.2–0.5 ms (the rise time of the HEMP voltage pulse is several nanoseconds and the length of the current pulse amounts to dozens–hundreds of nanoseconds). Other disadvantages of gas discharge tubes include high actuation voltage and residual voltage. Moreover, the actuation (gas breakdown) voltage of the lowest-voltage gas discharge tubes increases sharply with the increase of steepness (decrease in rise time) of the applied voltage pulse. For example, according to IEC 61643-311 [4], the minimum GDT’s discharge voltage rises from 75 V to 650 V if the rate of applied voltage increase as one kV/ µs. Obviously, this value will be even higher for HEMP pulses with its high steepness (rate of increase). Now it becomes clear that GDTs themselves cannot ensure protection of electronic equipment from HEMP. Due to this, various HEMP-protection devices marketed by some manufacturers seem very weird as their main (and often the only) element protecting from overvoltage is constituted by GDTs, Figure 14.2.

Figure 14.2: Circuit diagram of a device protecting an Ethernet network from HEMP based on a gas discharge tube (GDT) manufactured by MPE Company.

One of the manufacturers explained upon our query that they are aware that GDTs cannot provide protection from HEMP, but it is preferable to use these imperfect protecting devices rather than not to use any at all. This proves that we should not rely only on promotional literature. We need to conduct a thorough analysis of the internal structure of the offered device and the applied hardware components. Varistors that are widely used in electric engineering are also not suitable for telecommunication systems, however, the reason is different: they are not suitable due to their high capacitance (for low-voltage elements). High capacitance connected to

14.4 Protection telecommunication equipment with galvanic couplings |

257

Table 14.2: Recommended [5] maximal capacitance of protective elements. Signal Type E1A-232 E1A-422 E1A-423 E1A-485 E1 USB Telecom modem

Maximal Data Rate, bit/s

Recommended Maximal Capacitance, pF

19.6 kbps 10 Mbps 100 kbps 5 Mbps 2.048 Mbps 12.5 Mbps 60 kbps/1.5 Mbps

50 5 50 25 25 5 25

high-frequency circuits of telecommunication systems results in significant distortion and weakening of the useful signal. Thus, it is not acceptable to use high-capacitance protection elements in these systems. Table 14.2 shows the maximum permissible capacitance values for various signals recommended in [5]. Gas discharge tubes feature the best parameters from the minimum capacitance point of view (i. e., minimum impact on the circuit being protected). This feature, combined with high switching capacity (discharge currents can reach several or even dozens of kilo amperes) does not allow the developers of protecting equipment to disregard them completely. Nevertheless, it is necessary to look for workarounds to use them to protect telecommunication equipment. Furthermore, according to many manufacturers of protecting devices, such a workaround has been found. The idea was to combine high current, but a slow gas discharge tube, with a fast but low-current suppressor (Figure 14.3).

Figure 14.3: Design of an electric circuit of one channel and the actuation oscillogram of Series 3414 protecting device manufactured by HUBER+SUHNER Company [6].

258 | 14 Protection of telecommunication systems in electric power facilities from HEMP However, this technical solution is rather puzzling. Transient voltage suppressors (TVS-diode in the diagram) are known to actuate (i. e., switch into a conductive lowimpedance state upon increased voltage pulse impact) much quicker than gas discharge tubes (GDT in the circuit diagram). But upon the TVS suppressor’s actuation, the gas discharge tube will never actuate due to a low residual voltage on open TVS. This voltage is not enough for gas breakdown in the GDT (minimum GDT breakdown voltage is about 60 V [4]). Absence of conditions for GDT actuation is also confirmed by an oscillogram, which clearly shows that the voltage in this circuit does not ever the reach minimum voltage necessary for GDT breakdown. Another attempt to solve the problem was made by introducing additional resistors into the circuit (Figure 14.4).

Figure 14.4: A circuit diagram of a compound two-stage protective device with additional resistors R.

The idea of the developers was that, when a high-voltage pulse with high steepness of the leading edge arrives at the input of this device, the first one to actuate would be the TVS, which would limit the voltage amplitude of the device being protected. Furthermore, current flowing through it will result in the voltage drop on R resistors. The total voltage drops on resistors connected in series and the TVS suppressor should be sufficient for GDT breakdown. This will bypass the input of the device after its actuation and take the current off the TVS. Thus, developers expected the device to combine advantages of a TVS (fast response) with the high switching capacity of a gas discharge tube, while the total capacitance of a device was expected to remain low. This design became very popular in the many various types of protecting devices manufactured by various companies (Figure 14.5). Similar designs with GDT in the first stage (sometimes with different non-crucial changes and additions) are used in many protective devices, promoted as special HEMP-protecting tools, such as those of Meteolabor and many other companies. But deeper analysis of the situation reveals the hopelessness of this technical solution of a HEMP protector. This is connected with a short duration of the HEMP voltage pulse (up to several dozens of nanoseconds). The action of this short pulse will end prior to gas discharge of the tube’s actuation. Thus, the GDT is not important, and the lack of, or its availability, will not affect operation of the protective device. Some manufacturers use chokes instead of resistors such as depicted in Figure 14.4. The idea is to delay the process of the voltage rise on a TVS suppressor; bring

14.4 Protection telecommunication equipment with galvanic couplings |

259

Figure 14.5: A sample of compound two-stage protective device designed as shown in Figure 14.4 manufactured by the industry. GDT—gas discharge tubes; R—resistors; TVS—transient-voltage suppressor diode.

the moment of its actuation closer to the origination of discharge in the gas discharge tube; and thus limit time for heavy current flow through the suppressor. These chokes, featuring high impedance for a short pulse, will also limit the amplitude of current flowing through the TVS. However, the problem is that these chokes will present significant attenuation in a useful high-frequency signal that falls into the megahertz range. Thus this idea is not very suitable for telecommunication equipment. Another problem, or more correctly, a paradox, is the fact that various measures of equipment protection that weaken the HEMP’s impact will result in a reduction of the HEMP’s pulse current amplitude. Comparatively long cables with copper cores of the small section used in telecommunication systems (i. e., with relatively high impedance) can additionally limit the HEMP’s current amplitude. When the current amplitude flowing through the TVS suppressor and low-resistance resistors R (resistance of several ohms) is not sufficiently high, the voltage drop on them may not achieve the value required for GDT breakdown, i. e., 650–700 V and higher (at the high rate of voltage increase applied to the gas discharge tube at HEMP impact), while a wider current pulse (due to chokes’ inductance affect) will go through the suppressor causing thermal overload of its internal structure and even its destruction. Unfortunately, these debates cannot be either confirmed or contradicted with the figures due to the lack of real initial data about a HEMP pulse in each specific case and specific location of each equipment, the level of its protection, etc. Also, there are no data about the parameters of each copper couple of a telecommunication system’s multicore cable running through the various intermediate connections. However, a probability of unpredictable behavior of rather expensive devices that are extensively promoted as reliable means of protection conforming to MIL-STD-188-125, MIL-STD-461F standards, should alarm the specialists. At the same time, there is a question of how these devices have passed the conformity tests, if according to the above discussion they will not work as intended by their manufacturers. A deeper

260 | 14 Protection of telecommunication systems in electric power facilities from HEMP analysis reveals that there is a pitfall here as well. Indeed, manufacturers of these devices test them using a standard lightning current pulse of 8/20 milliseconds, instead of using a HEMP current pulse of 20/500 nanoseconds, as prescribed by the standards, i. e., the test pulse is flatter and longer. As an excuse, the manufacturers state [7] that it is very difficult to simulate a HEMP pulse, and, in order to do, so special expensive equipment is required. At the same time, generators of a standard lightning current pulse are readily available on the market and they are easy to use. Since the lightning current pulse is much wider than a HEMP pulse, its energy is even stronger than that of the HEMP pulse, thus it creates higher loads for a protecting device. Then they suggest [7] that, if a device withstood the test with a more powerful lightning current pulse, it will definitely withstand the short HEMP pulse. But the advocates of this test method bashfully conceal that the behavior of a gas discharge tube under long- and shortpulse impact will be absolutely different. Gas discharge tubes are reliable under the rather long lightning current pulse featuring a relatively flat leading edge, whereas, in the case of a much shorter and steeper leading edge of a HEMP, they will not have sufficient time to actuate due to: – their natural “sluggishness”; – a sharp increase of dielectric strength of gas contained in a GDT and consequently due to a sharp increase of its breakdown voltage.

14.5 New devices for protecting existing telecommunication equipment In our opinion, a solution is to use simple, very low-cost, non-recyclable, but predictable protective devices, based on transient voltage suppressors (TVS-diodes) that feature all the parameters necessary for efficient protection of telecommunication systems, such as: fast response time, low capacitance and low actuation voltage. In case of a HEMP impact, the internal p-n-junction of a TVS will breakdown because it is affected by a high current pulse flowing through it, whereas the circuit that it protects will be bypassed (short-circuited). Given the fact that a HEMP event is extraordinary and a pulse is single, non-repeating, this algorithm for protecting the device’s operation is quite acceptable as it will protect the equipment from the HEMP impact and will allow it to return to operation by just disconnecting the damaged protective device during the recovery period, which is inevitable in the case of a global HEMP impact. The only technical issue is to ensure selective action of the protective device. In other words, TVS breakdown should occur under a HEMP impact only and not under the impact of other, weaker repeating overvoltage transients. This selectivity can be achieved primarily by selecting quite powerful TVS and secondly by limiting the current flow through it by means of a resistor. Analysis of parameters of available TVS with actuation voltage and capacitance values suitable for

14.5 New devices for protecting existing telecommunication equipment | 261

Figure 14.6: Transient voltage suppressor diode (TVS) SP03-6 type. Table 14.3: Main parameters of suppressor SP03-6 type. Max. Stand-Off Voltage, V Breakdown Voltage, V Leakage Current, µA Clamping Voltage, V Peak Pulse Current (8/20 µs), A Peak Pulse Power (8/20 µs), W Junction Capacitance, pF

6 6.8 25 20 150 2,800 25

telecommunication systems, revealed that S03-6 type TVS-diodes (Figure 14.6) manufactured by Littelfuse (USA) are the most powerful devices among the others. They are more powerful compared to the TVS of other manufacturers, with the same operating voltage and capacitance values and that allow flowing of pulse currents up to 150 A (see Table 14.3). One small chip like this protects a single twisted pair from HEMP of both common (in relation to the reference potential) and differential (between conductors) modes. The price of one element is about 2 USD, however, in the case of wholesale purchase, it less than one USD. Resistance of a resistor connected with a suppressor in series (see the circuit diagram in Figure 14.6, where the resistor is connected in series with each input) should be about 20 ohms, in order to limit the maximum permissible current pulse flowing through a suppressor in case the pulse transient interference with an amplitude of several kilovolts impacts the protective device’s input. The current-limiting resistors should be non-inductive and should be intended for pulse current Table 14.4, Figure 14.7. This design of a protective device makes it very simple in terms of engineering (Figure 14.8) and inexpensive. The same principle can be used to protect the inputs of sensitive equipment connected through a socket (Figure 14.9). These simple devices can be produced by any manufacturer of printed circuit boards at a very affordable price. A range of Chinese companies will quickly produce the required quantity of these devices with excellent quality and at minimal price. The

262 | 14 Protection of telecommunication systems in electric power facilities from HEMP Table 14.4: Main parameters of high-voltage impulse low-resistance resistors of different types. Type

Manufacturer

ASRM2 PPR200 RT/RL AZ SPO250

Stackpole Electronics Firstohm HVR ARC Engineering Ohmite HVR International

Peak Voltage and Peak Power

Dimensions (D – diameter; L – length), mm

5 kV; 2 W 20 kV; 2 W 7.5 kV; 2 W 2.5 kV; 2 W 5.7 kV; 2 W

D = 5; L = 15 D = 5; L = 13.5 D = 8; L = 18 D = 26; L = 30 D = 13; L = 15

Figure 14.7: High voltage impulse low resistance resistors of some types (Table 14.4).

Figure 14.8: A drawing of a printed circuit board of the offered protective device for six twisted pairs that includes TVS and current-limiting resistors. The circuit board should be coated with a highvoltage varnish.

latter is very important for civil branches of the electric power industry and production sector because the high cost of HEMP protection is still a key factor that restrains practical adoption of such protection.

14.6 Protection of the communication cabinets | 263

Figure 14.9: An example of a simple protective device for two twisted pairs (for E1 signal) and a diagram of one channel (for single pair) for circuits connected to the equipment via a socket.

14.6 Protection of the communication cabinets The electronic equipment used in a power sector is specifically complex and diversified: It consists of a large number of various types of apparatus enclosed in the special communication cabinets, see Figure 14.10a–14.10c. Upon examination of this cabinet (see Figure 14.10a–14.10c), it is obvious that there is a problem protecting the electric power facility’s communication system.

Figure 14.10a: Communication cabinet.

264 | 14 Protection of telecommunication systems in electric power facilities from HEMP

Figure 14.10b: MOSCAD-RTU cabinets.

Figure 14.10c: Digital Distribution Frame (DDF) and Main Distribution Frame (MDF). Left—installed in the cabinet, right—wall-mounted open panel.

It is clear that the new communication systems should be designed along optoelectronic and fiber optical communication lines (FOCL) only, while all the electronic equipment should be enclosed in the specially protected cabinets and powered from protected power supplies through the protected cables.

14.7 The general concept for communication-equipment protection

| 265

Coincidently, the existing equipment (see Figure 14.10a–14.10c) designed and installed with disregard for HEMP sustainability requirements, creates a much more complicated problem. The problem cannot be solved by replacing the copper cables with the fiber-optic cables equipped with the special additional converters since, in such a case, the sensitive electronics remain installed in the glass door cabinets, providing zero protection against the electromagnetic pulse. Such cabinets are powered from unprotected power supplies, have dozens of entering cables (some of which have no shields), etc. Plus, it is hardly possible, if possible at all, and very expensive to replace with the fiber-optic cables the whole telemetry system, as SCADA (Supervisory Control and Data Acquisition) contains numerous RTU (Remote Terminal Unit) sensors distributed over a large area. Thus, other means must be used to ensure the protection of existing communication systems.

14.7 The general concept for communication-equipment protection The communication system is the one component of the electrical complex providing electric power generation, transmission and distribution. Thus, the common approach taken to the all other components and parts of this complex can be used for this component, too: Only critical parts of the apparatus that are located at critical facilities and provide critical functions, mandatory for making energy generation, transmission or distribution possible should be HEMP-protected. The facilities to be protected should be determined by the power facility staff beforehand and included in the special reconstruction plan. The determination of critical parts of equipment can be really problematic because many types of such equipment are used to maintain service of both critical and noncritical power facilities at the same time. In fact, several inputs of the same multichannel device can be connected to the critical external electric chains, while the remaining can be connected to noncritical external electric chains. If only the critical inputs are protected, the multichannel electronic device can become fully inoperable upon the electromagnetic pulse penetrating into the device through the unprotected input. Thus, critical and noncritical electric chains must be separated into different groups, and each group should be provided with a separate communication device. This problem is the most relevant to SCADA systems containing many cabinets and electronic modules (e. g., MOSCAD, see Figure 14.10b), processing and transferring the signals coming from the remote telemetry sensors (RTU) to the central control room. Several of these signals are very important, while others can be temporarily ignored in case of emergency after the HEMP impact. Thus, all important RTU chains must be switched to the inputs of the electronic module assigned as critical, and the modules themselves must be located in the cabinets that are determined as critical. Meanwhile,

266 | 14 Protection of telecommunication systems in electric power facilities from HEMP the protection of the most critical RTU sensors is a separate challenge. Measures for protection of such cabinets are further considered in the next chapters. In the same way as in the case discussed earlier with the relay protection and automation equipment, special HEMP-shielded cabinets should be provided for the projected and newly introduced telecommunications equipment, and the existing cabinets should be upgraded using all the common principles discussed above.

14.8 Retrofitting grounding systems of cabinets containing the electronic equipment Problems of a conventional grounding system of electronic equipment, installed into the electric cabinets, and the special aspects of a grounding system upon the HEMP are detailed in [9]. Unfortunately, these special aspects are hardly ever considered. Often, the cabinets containing the identical communication apparatus have grounding systems of very different designs (e. g., connections of potential-equalizer bars, see Figure 14.11). From the point of view of common sense, such differences are hard to explain. Also, the same goes for the rationale for isolating these potential-equalizer bars from the cabinet as the electronic metal enclosures are mounted on the cabinet’s metal elements directly, without any insulation on one side, and these enclosures are bound to the insulated equalizer bars on the other side. According to the recommendations listed in [9], the enhanced communication cabinet’s grounding system should be configured as a special floating ground, see Figure 14.11.

Figure 14.11: Design of communication system cabinet grounding: 1—potential equalizer bar; 2— external grounding system cable; 3—insulator insulating the potential equalizer bar from the cabinet body; 4—elements connecting the potential equalizer bar with the cabinet body; 5—flexible links connecting the electronic devices enclosures to the potential equalizer bar.

14.9 Retrofitting open-patch panels | 267

14.9 Retrofitting open-patch panels If DDF and MDF patch panels (see Figure 14.10c) are open and installed directly on the wall, they require special measures to be taken to ensure protection from an electromagnetic pulse. One of the options is to install special aluminum two-piece shields: the plate being slightly bigger than the wall-mounted DDF and MDF panels located behind the plate and a flanged box enclosing the panel from the front. The crevicefree joint between the box and the plate is made as the metal-metal contact or using the conductive rubber spacers. All low-frequency cables should be protected with lowcost and rather effective ferrite beads, while the high-frequency twisted pairs of critical cables should be protected with the special protecting elements based on TVS-diodes, see [8]. These protective elements are installed on the small panels and are connected in parallel to the same DDF and MDF panel connectors using the twisted-pair conductors.

14.10 Protection of the power supply system The protection of a communication apparatus power supply is performed by the traditional methods using varistors and chokes. The principles of building such a power supply system are universal and have been described previously, including a small diesel generator, which must be installed directly in the room near the battery charger. This diesel generator power of 3–5 kW must be assembled with a pre-exhausted exhaust pipe, and should also be equipped with a compact rectifier with the function of current limiting and voltage stabilization (i. e., the usual small switching battery charger unit—BC). This small diesel generator, together with the BC, must be disconnected from all external electrical circuits and enclosed in a closed aluminum container. Such a redundant power source is required on the chance that the external ACpower source feeding the whole communication system is damaged by the high-power HEMP.

14.11 Retrofitting the facility (room) containing the critical kinds of communication equipment Generally speaking, many external cables enter the facilities, and they may contain unprotected windows. It is clear that the facility designed and built as unprotected cannot be fully protected. However, it is possible to significantly attenuate the level of electromagnetic radiation entering the facilities through the cables, the windows and the doors. In order to do this, the cables entering the facility must run in the special solid metal cable trays starting immediately from the penetration points, while the windows and the doors must be shielded in the same manner as the cabinet glass doors, or be fully enclosed with aluminum panels or the conductive curtains.

268 | 14 Protection of telecommunication systems in electric power facilities from HEMP

14.12 Conclusion The described technology for protection of the most sensitive and vulnerable power sector systems, such as communication, from HEMP is quite affordable and easily realizable if used along with the previously described methods of protection of sensitive electronic inputs. The market proposes a wide range of expensive and inexpensive elements and materials. Analysis of the situation showed that expensive devices promoted by their manufacturers fail to provide reliable protection of highly sensitive equipment. It is recommended to use simple, inexpensive, non-recyclable devices that can be ordered individually by a consumer. Proposed technical solutions and recommendations are mainly aimed at the retrofit of current systems but may be partly used for designing new systems built on optoelectronic and fiber-optic communication lines (FOCL).

Bibliography [1] Gurevich, V. Protection of Substation Critical Equipment against Intentional Electromagnetic Threats. Wiley, London, 2017, 240 p. [2] Gurevich, V. Facilities Ensuring Substation Direct Current Auxiliary Power System Survivability under Electromagnetic Pulse (HEMP). Part 1. Stationary Substations. Int. J. Electr. Electron. Res., 2017, Vol. 5, No 3, pp. 6–12. [3] Gurevich, V. Use of LC-Filters to Protect Equipment from Electromagnetic Pulse: Is It Real Necessity or “Business as Usual”. Int. J. Res. Stud. Electr. Electron. Eng., 2017, Vol. 3, No 1, pp. 1–7. [4] IEC 61643-311 Components for Low-Voltage Surge Protective Devices—Part 311: Performance Requirements and Test Circuits for Gas Discharge Tubes (GDT), 2nd ed., 2013. [5] Clark, M. Transient Voltage Protection across High Data Rate & RF Lines. MicroNote 122, Microsemi Corp. [6] Catalog HUBER+SUHNER “Lightning Protection”, Edition 2016, pp. 76–78. [7] Nalhorczyk, A. J. HEMP Filter Design to Meet MIL-STD-188-125 PCI Test Requirements. IEEE. 10th International Conference “Electromagnetic Interference & Compatibility”, 26–27 Nov., 2008, pp. 205–209. [8] Gurevich, V. Protection of Telecommunication Systems in Electric Power Facilities from Electromagnetic Pulse (EMP). Part. 1. Int. J. Adv. Comput. Technol., 2017, Vol. 6, No 9, pp. 2446–2450 (Vol. VI, Issue IX). [9] Gurevich V. The Issues of Electronic Equipment Grounding at the Power Facilities. Int. J. Res. Stud. Electr. Electron. Eng., 2017, Vol. 3, No 1, pp. 11–19.

15 Improvement of HEMP resilience of automatic fire-suppression systems 15.1 Introduction Detonation of nuclear ammunition at an altitude of more than 30 km results an electromagnetic pulse near the Earth’s surface with a field gradient of up to 50 kV/m. Due to the extensive and expanding use of low-voltage highly sensitive microelectronic and microprocessor-based equipment in the power industry, the electromagnetic pulses of high-altitude nuclear explosions (HEMP) are perceived by many armies worldwide as a prospective weapon to destroy the foundation of any country’s infrastructure, i. e., the power industry. Thus, many countries have recently started developing safety equipment to protect electric power equipment from HEMP [1, 2]. However, apart from electronic and electrical equipment, which is directly involved in generation, transmission and distribution of electric power, there is another type of highly-sensitive electronic equipment that is not directly involved in these processes, but can disrupt them. These are fire-suppression systems installed at any power plant and electrical substation.

15.2 Firefighting systems for power facilities There are many different systems: from fire alarm only to automatic fire suppression. Obviously, a faulty fire-alarm system will not cause any disaster to occur. However, this is not true for automatic fire-suppression systems as in case of their selfactuation, some of them will disconnect electric equipment first and then deliver a fire-extinguishing agent. As a rule, fire extinguishing of electric equipment in lowvoltage (up to 0.4 kV) electric cabinets is performed by delivering pressurized lowconductive powder, gas or a fine water spray, into the burning area (Figure 15.1) and does not require voltage disconnection. These systems are also very diverse: from local systems, containing one gas container and intended for a single cabinet or a group of cabinets (Figure 15.1) to centralized systems, which consist of rather sophisticated and branched systems (Figure 15.2). Local systems based on a single gas balloon and a simple temperature sensor do not consist of sophisticated electronic components, and thus it is unnecessary to protect them from HEMP. However, centralized fire suppression systems that employ microelectronic control equipment are susceptible to HEMP. Moreover, both types of damage are dangerous, i. e., failure of a fire suppression system to actuate and its faulty actuation. The latter is not just about the high cost of such wasted gas. In fact, gas wasted during faulty actuation will not be available to extinguish fire, which can start after breakdown of insulation of many types of electric equipment upon the impact of a highvoltage pulse. https://doi.org/10.1515/9783110639285-015

270 | 15 Improvement of HEMP resilience of automatic fire-suppression systems

Figure 15.1: Local gas and spray fire-suppression systems for low-voltage control cabinets.

Figure 15.2: Components of centralized gas fire-suppression system.

When extinguishing fire on large high-voltage equipment (where intensive cooling of the burning zone is necessary, in addition to deoxygenation), water is usually delivered through a previously mounted piping system surrounding the unit and sprayed by sprinklers, Figure 15.3. Disconnection of power before spraying is compulsory; it is performed upon a signal of an automatic fire-suppression system. Control panels (cabinets) of a fire alarm and automatic control are very sophisticated. They consist of microprocessors and many types of microchips (Figure 15.4), connected through dozens of cables that create a branched antenna system

15.2 Firefighting systems for power facilities | 271

Figure 15.3: Fire extinguishing for power transformers: above—water piping with sprinklers around a transformer; below—fire extinguishing in progress.

(hundreds-of-meter long), which absorbs the HEMP energy from a large area and delivers it directly onto highly sensitive electronic components. These systems are built based on network technologies at large facilities. Thus, firefighting control panels are equipped with RS422 or RS485 interfaces; they can also communicate via Ethernet or by means of modem transmission over a dial-up telephone channel, which makes it even more susceptible to HEMP and complicates their protection. Among other things, centralized systems of fire extinguishing control contain power electronic devices, such as soft-start systems of powerful electric motors of pumps that supply water into a hydraulic fire extinguishing system, Figure 15.5. These

272 | 15 Improvement of HEMP resilience of automatic fire-suppression systems

Figure 15.4: Control panel (cabinet) of fire alarm (a) with three control units (1, 2 and 3) and accumulators (4, 5); 6 (b)—external multi-core cables, connected to internal electronic circuit of the cabinet; c—microelectronic “internals” of one of three control units in the cabinet.

Figure 15.5: Power electronic devices to control electrical motors of a centralized fire-extinguishing system’s pumps.

15.3 Improvement of automatic firefighting system’s resilience to HEMP

|

273

power electronic devices are connected to a common power supply network of an electric energy facility, where HEMP-induced pulse overvoltage can reach as high as dozens of kilovolts. At the same time, my analysis of standards and other regulatory documents ([3–6] and others) that determine certain features, parameters and guidelines for testing of fire extinguishing systems and their use showed that they neither address, nor even consider, the HEMP impact at all, and thus, they do not require implementation of special measures that improve the invulnerability of these systems to HEMP.

15.3 Improvement of automatic firefighting system’s resilience to HEMP Is it possible to protect all these sophisticated systems from HEMP and how? The answer to this question is rather complicated. Indeed, tampering with the internal electronic circuits of fire suppression and an automated control system is prohibited, even for such a good reason as HEMP protection. For example, installation of any additional filters in RS422 or RS485, in order to establish Ethernet or modem communication, is an example of such tampering. Still, there is a feasible solution. For example, it is possible to briefly disconnect the electronic fire-extinguishing system using a remotely sent command upon obtaining immediate information about the danger of impact by a nuclear-tipped missile. Use of a high-altitude nuclear weapon (even with the aim to damage infrastructure and not the population) cannot be too spontaneous and unpredictable. As a rule, a number of collisions signifying disturbance of the environment occur prior to impact, in which case the state agencies become alert to the unfolding of an unfavorable situation, including the danger of HEMP impact. In this situation, it is not unusual that, while the nuclear-tipped missile approaches, advanced information about the HEMP danger will arrive, and this time will be sufficient for remote switch-off the critical fireextinguishing systems located at various remote power system facilities. De-energized systems are known to be much less susceptible to internal damage by HEMP. Moreover, a damaged de-energized system cannot adversely affect other systems. In practice, remote switch-off of automated fire-suppression systems can be achieved using various tools readily available on the market. These can be relatively expensive (200-300 USD) remote power switches installed in separate casings (Figure 15.6) and controlled via Ethernet. These switches are equipped with network interfaces and output’s electromagnetic relays, which respond to a special code, transmitted by network. The terminals of these relays can switch external devices either off or on. The so-called “network relays” (Figure 15.7) are much cheaper (20–30 USD), but are equivalent to those described above.

274 | 15 Improvement of HEMP resilience of automatic fire-suppression systems

Figure 15.6: Remote power switches.

Figure 15.7: Network relays with different number of channels and output electromechanical relays.

The GSM remote controller (Figure 15.8), which costs 100–150 USD, can be used as a device for remote switch-off of the fire-extinguishing systems. Use of these systems does not suppose intrusion into the internal circuit of fireextinguishing systems, since they are inserted into the external power supply circuit only and the breaks of this external circuit. However, they have some specific usage features. First of all, apart from the external power supply, the fire-suppression control systems contain an additional back-up power source. As a rule, this power source is designed as accumulator batteries (4 and 5 in Figure 15.4), so it is necessary to disconnect both power supplies. Secondly, in order to ensure remote return of

15.3 Improvement of automatic firefighting system’s resilience to HEMP

|

275

Figure 15.8: Remote switches controlled via Global System of Mobile Communications (GSM).

the fire-extinguishing control system to an operational state (after the attack is over), the power sources need to be disconnected in a certain order: first—the back-up battery, and then—the supply mains. Returning to an operational state is performed in a reverse order. Thus, the remote switch (relay) should have at least two independent channels with two independent output relays. It is noteworthy that the OFF/ON codes of the fire-suppression system, as well as the order of these procedures, should be clearly defined in the Emergency Guidelines, and the employees involved must be familiar with these guidelines. As for the remote return into the operational state, it is possible only in case there was no HEMP impact. Otherwise, the automatic firesuppression system should be thoroughly checked before returning to an operational state. Apart from relays in power supply circuits, another “permissible intervention” is installation of special filters into both the circuits of the external power supply of a fire-suppression control system and all other power control devices’ power circuits (such as those controlling the pumps’ motors, Figure 15.5). These filters need to consist of elements that limit the voltage amplitude and the current induced by HEMP. Metal-oxide varistors (MOV) and chokes are very convenient for low-voltage electronic control systems. There are standard units that protect single-phase AC supply circuits from over-voltages. These units are mounted in small casings that will be installed on a DIN-rail and contain two varistors and a gas discharge tube (Figure 15.9a). The majority of single-phase protective units are engineered according to this design, which ensures good lightning protection, but not HEMP protection. This is due to a long– duration response of the gas discharge tube (GDT) to pulses of this kind and a sharp increase of its breakdown voltage, as affected by a short pulse with high steepness of the leading edge (such as that of HEMP). Substitution of the gas discharge tube by a varistor (Figure 15.9b) could have improved the situation, but such units are very rarely produced for 230/400 VAC circuits. Instead, there are hundreds of designs of protective units that consist of two varistors and a jumper instead of a VR3 varistor (Figure 15.9c). The disadvantage of this circuit is double actuation voltage and double clamping voltage between the line (L) and

276 | 15 Improvement of HEMP resilience of automatic fire-suppression systems

Figure 15.9: Various protection units (a, b, c) and suggested connection of protection elements (d). VR1—VR3—metal-oxide varistors (MOV); GDT—gas discharge tube; L1–L2—current-limiting chokes.

the neutral (N) terminals, compared to voltage relative to the ground. However, the ground does not represent the zero-potential area (relative to which the over-voltage pulse is applied to the circuit under protection and to which this pulse should be diverted by the protection device) during HEMP (unlike lightning discharge). Thus, this layout is not suitable for HEMP protection because its high voltage pulse is very likely to be applied between the line and the neutral. At the same time, there are singlepole protective units with a single varistor inside, which can be successfully used to protect a single-phase AC circuit with nominal voltage 230 V (Figure 15.9d). In order to increase the efficiency of HEMP suppression, the layout is supplemented by the current-limiting chokes L1–L2. Single-pole protection varistor units DS71R-400 type manufactured by Citel and I2 R SA277-50 type units manufactured by Transtector (Figure 15.10a) can be recommended for certain circumstances. These units feature ideal specifications and in-

Figure 15.10: Single-pole protection units with varistors (a) and a unit with two current-limiting chokes (b).

15.3 Improvement of automatic firefighting system’s resilience to HEMP

|

277

clude an additional built-in element that disconnects the varistor in case of its damage, a visual indicator of damage and a contact for a remote alarm. A unit of chokes DSH 2x16 manufactured by Citel is recommended as the currentlimiting chokes L1–L2. This unit consists of two chokes rated 16A each, located in a small casing, which can be mounted on a standard DIN-rail (Figure 15.10b). Powerful varistor blocks R1–R6 (e. g., B40K460 type) and also powerful chokes L1– L4, manufactured by CWS (Figure 15.11) are more suitable for three-phase power circuits and other powerful loads. This type of varistor block can tolerate pulse currents rated up to 40 kA and is intended for long-time operation under 460 VAC. Upon the varistor’s actuation under HEMP impact, its clamping voltage will not exceed 1240 V. This is acceptable for power equipment that (according to IEC 61000 group of standards for EMC requirements) must tolerate pulse over-voltages of not less than two– four kV and features either built-in elements, protecting from this level of over-voltage, or corresponding levels of insulation. Small chokes (L1–L4), with a helix coil and special powdered cores encapsulated by epoxy resin, are manufactured by the American company CWS and designed for currents from 40 to 200 A.

Figure 15.11: Design and components of a filter for power circuits of powerful automatic fire extinguishing control systems. R1-R6—varistor blocks; L1-L4—special chokes.

Similar sets of varistors should be connected both where the three-phase power cable penetrate the automatic fire extinguishing control system, and at all the output cables’ from this system, plus on the other end of these cables, i. e., close to the motors of water pumps.

278 | 15 Improvement of HEMP resilience of automatic fire-suppression systems

15.4 Conclusion Critical automatic fire-suppression systems of critical types of electric equipment require adoption of special measures to improve their HEMP resilience. These measures should be based on the principle of non-interference into the internal circuit of firesuppression systems. Remote switching off of these systems, in case of danger of HEMP impact with their further return to an operational state (also remotely), as well as installation of HEMP filters described above, are examples of these measures.

Bibliography [1] Gurevich V. Protection of Substation Critical Equipment against Intentional Electromagnetic Threats. Wiley, 2017, 228 p. [2] Gurevich, V., Cyber and Electromagnetic Threats in Modern Relay Protection. Taylor & Francis Group, Boca Raton, 2015, 205 p. [3] BS EN 50130-4:2011 Alarm Systems. Electromagnetic Compatibility. Product Family Standard: Immunity Requirements for Components of Fire, Intruder, Hold up, CCTV, Access Control and Social Alarm Systems. [4] IEC 62599-2:2010 Alarm Systems. Part 2: Electromagnetic Compatibility: Immunity Requirements for Components of Fire and Security Alarm Systems. [5] NFPA 15, 2017 Standard for Water Spray Fixed Systems for Fire Protection. [6] NPB 88-2001 Fire-Extinguishing and Alarm Systems. Designing and Regulations Norms: Decree No. 31, Ministry of Interior of Russia, June 4, 2001.

16 Protection of diesel generators from HEMP 16.1 Introduction The problem of diesel-generator (DG) protection from HEMP becomes particularly relevant. First of all, DGs act as backup power sources and are designed to power up critical loads in emergency situations. Consequently, they need to be 100 % ready for use even after the HEMP. Secondly, DGs are often stored outdoors (outside of the buildings that can partially mitigate the HEMP impact). DGs stored outdoors may also become a target for Intentional Destructive Electromagnetic Interferences (IDEI), which can be produced by portable devices that generate pulse emissions of several gigawatts in the directional antenna [1]. Nowadays, there are thousands of kinds of DGs with a power rating from several kilowatts to dozens of megawatts. Some of them are small open-design portable devices that can be stored in a metal container protected from electromagnetic emissions. These can be used when necessary after removal from the container. Generally speaking, these low-capacity DGs have a simple design without sensitive electronics and are relatively inexpensive. Thus, it makes no sense to use any special measures to protect these DGs (except for a metal enclosure).

16.2 Increasing resilience of medium- and high-capacity DGs Medium-capacity industrial DGs (from dozens to hundreds of kilowatts) are large and heavy devices that are intended to be mobile. As a rule, they are confined in a casing with many sensors and microprocessor-based controllers that control the DG’s operation and measure and display various parameters, as well as protect them from overload and emergency modes. Protection from emergency modes in high capacity (1–50 MW) DGs is performed by digital protective relays (DPR) of the same type as those used in the electric industry in conventional power plants and substations. They are usually confined in standard relay protection cabinets, which are installed inside the DG’s casing. These cabinets are usually of the same type as those used in the electric energy industry in power plants and substations. Use of microprocessor-based controllers and DPR that are especially susceptible to HEMP and IDEI [2] in medium- and high-capacity DGs results in a dramatic drop in the DG’s efficiency as a backup power source for critical loads. Consequently, they need to be urgently addressed. It should be noted that there are two completely different modes of DG use (from the HEMP protective measures standpoint). One mode presupposes storage of de-energized DGs at warehouses, whereas in the other mode they are constantly connected to the local consumers’ electric network and can automatically start at any time should it become necessary to re-energize the power supply, or to flatten the load peaks. Let us look at the possible protective measures for mediumand high-capacity DGs in these two situations. https://doi.org/10.1515/9783110639285-016

280 | 16 Protection of diesel generators from HEMP

16.3 Protection of DGs stored and de-energized outdoors It should be immediately stressed that it is inappropriate to store DGs in centralized warehouses (as usually is the case). DGs are backup power sources that should be ready for use within the shortest possible time after an emergency’s occurrence (after HEMP in this situation). As HEMP impact is all encompassing and creates problems for transport, communication systems and computerized warehouse equipment, it becomes obvious that we need to aim at decentralization of backup DGs storage locations, moving them closer to potential consumers. The easiest solution to protect the internal equipment of de-energized DGs from HEMP is to put metal casing on top of the DGs. However, this approach has some serious drawbacks. First, the casing for medium capacity DGs (5–eight-m long; 1.5–twom wide; three-m high) should be equipped with special stiffeners. The metal casing should be made of sufficiently thick metal welded together to provide the necessary structural rigidity. Such casing will be so heavy that a user will need a crane to remove it from the DGs and prepare the DGs for a startup. However, this can hardly be a reasonable approach in a critical situation. Moreover, this casing will protect the DGs from five sides. But what about protection from the bottom? What about the inevitable multiple large gaps between the casing walls and the foundation of the DGs? On the other hand, medium- and high-capacity DGs are usually equipped with their own metal casings. Nevertheless, it should be noted that such casings have many cutouts, holes and blinds that drastically reduce its screening properties. Taking this mentioned into account, I suggest the following concept for mediumcapacity DG protection: 1. Improvement of screening capacity of the DG’s own casing by closing all the cutouts, holes and blinds with removable metal patches that can easily be removed when preparing the DGs for startup. 2. Disconnection of connectors of all the electronic appliances and sensors from internal wiring and cable harnesses. 3. Installation of the same type of connectors’ counterparts with short-circuited pins into connectors both on the side of electronic appliances and sensors and on the side of the cable harness. Points of common coupling of all the cable-harness wires should be connected to the DG’s chassis. 4. Short-circuiting of all the power leads of the generator’s rotor and stator at a point of common coupling and connection of this point with the DG’s chassis. 5. Removal of the electronic unit from the automatic power circuit breaker at the generator’s output and placing it into the screened casing. When deploying clause 1 of the suggested concept, special emphasis should be put on the window, which is cut out in the DG’s casing in front of the microprocessor-based controller’s screen. Such windows are present in most DG types, see Figure 16.1. They are intended for visual monitoring of the controller’s readings. However, they pose

16.3 Protection of DGs stored and de-energized outdoors | 281

Figure 16.1: Windows in front of the controllers’ screens cutout in casings of most DG types.

the greatest danger from the point of view of the DG’s susceptibility to HEMP. These windows should be tightly closed with a welded-on or bolted steel plate contoured with conductive rubber gaskets. The controller’s readings are not taken continuously. When the DG is started, it is enough to read its parameters by opening the door in the DG’s casing, which is located near the controller. Should it be absolutely necessary, it is possible to weld a small door opposite the controller’s screen (instead of the steel plate) or use conductive glass to cover the windows or glue a transparent conductive film [3] to the ordinary glass. However, one needs to understand that all of these alternative options will be less efficient than the first option. The second approach to improve the screening ability of the DG’s casing is to close the air intake and exhaust apertures and blinds, see Figure 16.1, with a solid-steel plate fixed with welded bolts and contoured with conductive rubber gaskets. These screening plates should be removed before the DG start-up. Another large opening in the DG’s casing is the cutout designed to connect external power cables to the DGs. This opening should also be closed by a removable bolted steel plate. When deploying other methods of the offered concept (related to disconnection of highly sensitive electronic equipment from internal electric circuit), it should be kept in mind that each point of intrusion into the internal arrangement of the DGs should

282 | 16 Protection of diesel generators from HEMP

Figure 16.2: Standard connectors of various types used to connect sensors to the DGs.

be registered in the check list, and each procedure of disconnection and restoration of circuits should be marked in this check list. In order to connect multiple DG sensors to the cable harness, standard connectors (Figure 16.2) are used. Subsequently, it is easy to buy mating parts for these standard connectors and use them as caps to short circuit the terminals of sensors and wires in harnesses. The various types of DGs use different types of controllers. Producers of these controllers often use their own, nonstandard connectors to connect external circuits. For example, one of these connectors, marked as 160-7689 in the documentation, is used in a well-known EMCP 4 controller, which is widely used in various types of DGs, see Figure 16.3. Some diesel units have another nonstandard connector labelled 9X-4391. However, nonstandard connectors are not really a problem as they (and many other types of connectors used in DGs) are readily available as spare parts and can be purchased both from a controllers’ manufacturer and on the web, say eBay (see Figure 16.4) at a relatively low price (50–60 USD).

16.3 Protection of DGs stored and de-energized outdoors | 283

Figure 16.3: Widely used DG controller, type EMCP 4.

Figure 16.4: Nonstandard 160-7689 and 9X-4391 connectors for DG controllers, which are readily available on the market.

284 | 16 Protection of diesel generators from HEMP

16.4 Protection of DGs connected to consumer network There are two options here: – immovable DGs located in a permanent place. These start up automatically whenever necessary; – transportable DGs that are arranged temporarily to power up a consumer. The latter are intended for frequent start-ups and for continuous operation during specific limited periods. In some circumstances, these DGs can be started in advance as a response to intelligence data about a danger of pending electromagnetic impact. Thus, they can be working during the HEMP impact. In the first case, the most efficient protection is achieved when locating the DGs in a closed container made of reinforced concrete with a fine-mesh reinforcement or a metal-sheet fabricated container. These containers should have no windows and their vents should be intended for cooling-air intake and release, and the exhaust gas holes also need to be closed with special honeycomb structure blocks. These blocks that close the vents are clearly seen in Figure 16.5.

Figure 16.5: Protective containers for immovable DGs. The vents are closed with special honeycomb structure blocks.

The honeycomb structure blocks (Figure 16.6) are made of round or rectangular (sometimes hex-shaped) section pipes welded alongside. The purpose of these blocks is to ensure cooling-air (or exhaust gases) circulation and prevent electromagnetic emissions from penetration into the protected area. Design features of such meshed structures (based on the so-called “below-cutoff waveguides”) and the method for calculation of their parameters have already been discussed. The vent panels constructed with such meshed structures are manufactured by numerous companies, such as: Holland Shielding Systems, MAJR Products Corp., Micro Tech Components GmbH, Parker Hannifin Corp, EMC EMI Ltd, Kemtron Ltd, Fos-

16.4 Protection of DGs connected to consumer network | 285

Figure 16.6: Honeycomb structure blocks designed to close the vents.

Figure 16.7: Vent panel sizes produced by Arrow Dragon Metal Products Co., Ltd.

han Huarui Honeycomb Technology Co., Ltd, Foshan Alucrown Building Materials Co., Ltd and Arrow Dragon Metal Products Co., Ltd, among many others. High- and medium-power diesel generators are equipped with large and powerful fans ensuring the high-speed movement of large volumes of cooling air within the diesel-generator enclosure. As a rule, there are two separate rather big cut-outs (0.5–2 m2 ) provided in the enclosure to intake and discharge the air. When the cage, grid or meshed structure covers the cut-outs, the air flow retardation inevitably rises and the effectiveness of the diesel-generator cooling is reduced. Too small meshes are prone to clogging quickly with dirt and dust, thus further worsening the situation. To avoid the significant reduction in cooling effectiveness, the mesh size must be sufficiently large. To ensure the below-cutoff waveguide function, the length of each elementary cell (mesh thickness) must be increased in proportion to the increase of the section area. Meshed structures with sufficiently large meshes, thickness and panel size are also manufactured by many companies, such as Arrow

286 | 16 Protection of diesel generators from HEMP Dragon Metal Products Co., Huaming Electronic Equipment Co., Holland Shielding Systems Bv., Parker Chomerics and Kemtron, see Figure 16.7. The retardation of increased airflow can be partially compensated by the increase in the total mesh-structure area covering the diesel generator enclosure cut-out, such as to make it bigger than the cut-out area. The entire enclosed casing, made of the same meshed structure used instead of the flat vent panel, can be helpful in such a case, see Figure 16.8.

Figure 16.8: Large-area meshed vent panel with an opening side wall.

In this case, the area is increased due to sidewalls allowing for additional air-flow circulation. To ensure the additional compensation, one or two side walls of the casing can be opened during the long runs of the diesel generator at high loads and high ambient temperatures, see Figure 16.8. In addition to the just-mentioned honeycomb blocks, the DGs located in a protective container should be equipped with special HEMP filters installed between the power leads of the DGs and the load located outside the protected area. These filters, which are designed for full-load current (Figure 16.9), are rather large and heavy. They need to be attached to the protective container in such a way that only the filters’ exit cables are free from pulse overloads, and powerful high-frequency signals can’t enter the protected area. The same is applicable to all control cables that also need to be run through corresponding filters before entering the protected area. All such filters need to be located in a separate container, see Figure 16.10. Actually, such protective containers will fit for not only immovable, but also movable DGs, but of relatively low capacity (up to 100–200 kW). This type of protected DG is produced by some companies, e. g., EMP Engineering. The price of a 60-kW DG in

16.4 Protection of DGs connected to consumer network | 287

Figure 16.9: Powerful HEMP filters for power circuits rated 800 and 1,200 A.

Figure 16.10: Protective container for immovable DGs. 1—filter block; 2—honeycomb structure block closing the opening for air bleeding and the exhaust aperture; 3—honeycomb structure block closing the cooling-air intake opening.

a protected container is 85,000 USD. Both such DGs and those located in immovable protective containers can work properly during the HEMP impact. Since HEMP impact is global and affects large regions and sometimes even entire countries, the approach to backup DGs use should be different from that employed for human-induced (technological) or natural disasters because the latter are: 1) limited in space; and 2) the space is not known in advance. Unlike local technological or natural disasters, locations for DGs installation in the case of global HEMP impact can be determined in advance. Consequently, one of the approaches to protect heavy and large movable, large capacity DGs (more than 0.5–one MW)—without protective containers intended for operation at different consumers’ during HEMP impact—is early location of fully equipped empty protective containers near critical loads, which will be powered from backup DGs during HEMP impact. Moreover, the DGs need to be delivered to the site and installed in the previously prepared protective containers.

288 | 16 Protection of diesel generators from HEMP Early transfer of critical loads to DG power and their disconnection from a centralized power supply in case of HEMP danger provides an additional benefit. This is due to the significant risk reduction of power system damage, when it is turned off (disconnected). Thus, this approach may become mandatory in practice. A more complicated and less reliable solution to ensure efficiency of large DGs that have no special protective casing under the possible HEMP impact is to use wellknown standard approaches to protect electric and electronic equipment in power plants and substations [1], in addition to installation of honeycomb structure blocks on vents, power filters and weld sealing of a window in front of the controller. The just-mentioned known protection measures include: – use of shielded control cables inside the DG’s casing; – use of metal (instead of plastic) cable trays; – use of filters embedded into control cables or ferrite filters put onto the control cable harness; – installation of excess-voltage suppressors that employ zinc-oxide varistors or powerful avalanche diodes in all the power and control circuits; – insertion of a high-frequency choke into the grounding circuit. Obviously, such a solution [1] is the most difficult to employ for a consumer having an unprotected DG. However, in some cases it can be the preferred approach, e. g., if the manufacturer of the DGs will initially adopt the just-mentioned protective measures at the order-processing stage.

16.5 Active protection method for diesel-generator controller Essentially, the electronics located inside the screened body (or inside the wellprotected cabinet) is virtually invulnerable to HEMP, provided that it is switched off and all its inputs and outputs are shorted. On the other hand, a complete set of special means, including special construction materials, screened rooms, screened cables, surge arresters, filters, chokes and special grounding systems [4], is required to ensure the HEMP protection of permanently operating electronics. Such means are deemed as passive. Each mean provides only the partial protection of equipment, therefore, to ensure complete protection, the full set of all such means must be applied. Thus, HEMP immunity of enabled and disabled electronics (with shortened inputs and outputs) differs dramatically. The only problem is that the disabled equipment with shortened inputs and outputs does not operate! Is this true? Certainly, the disabled equipment cannot operate. However, there are many types of online electronic equipment performing only certain functions within very short periods of time. For example, there are devices measuring the parameters and duty cycles on the equipment monitoring. Devices of this type can be active during one or more minutes of each hour

16.5 Active protection method for diesel-generator controller

| 289

or even once a day, or a month. During the remaining time, they are in standby mode. Some kinds of such devices are not critical and do not require HEMP protection since they are not able to cause a disaster if damaged upon HEMP impact. However, many diagnostic and monitoring devices of this kind use factory-installed main controllers of basic equipment to gather the data of such basic equipment. Therefore, such controllers must remain operational, even when the equipment is inactive so as to maintain the periodic operation of auxiliary monitoring electronics. This principle, for example, is used in systems consisting of widely separated 500–1,000-kW emergency diesel generators, monitored from the remote control center on a continuous basis. This signifies that, in order to ensure the periodical monitoring (once a month), the main controller of diesel generator must operate permanently, so it is not adequately protected against HEMP. After the main controller damage, the modern diesel generators cannot be started, so in an emergency (after the HEMP impact) the standby power sources will not be operational. The suggested method of active protection can help to solve this problem. Initially, the author suggested a method of active protection of electronics from external electromagnetic effects in 1996 [5]. Later, in [6, 7], it was amended and clarified. Active protection of electronics means that the devices designed for periodic operation are disabled remotely, and all their inputs and outputs are shortened for the standby periods. When needed, the devices are remotely enabled (by radio, cellular communication or computer network), and their inputs and outputs are unlocked. Therefore, during the standby periods (i. e., most of the time), the device is provided with the highest HEMP protection (disabled rarely and for a very short time period). It is quite clear that in order to prevent the protected electronics from penetrating HEMP during the standby phase, the unprotected remote-control device must be provided with high-voltage insulation from the protected electronics. As for the protecting device, including the remote control unit, there are no specific requirement for the HEMP invulnerability of such a device because it does not directly impact the operability of the main equipment (the main equipment can be activated when the protecting device is disabled) after HEMP impact, so such a protecting device must be dismounted and discarded. The proposed active protection device for electronics (APD), see Figure 16.11, is basically a compact unit consisting of three components: multicontact low-voltage electromechanical relay LVR, two high-voltage relays (HVR1 and HVR2) and a remote control unit RCU. Its principle of operation is simple and clear (Figure 16.11). In standby mode, all inputs and outputs of the protected device (such as the controller) are shorted (shunted) and its power supply is disconnected. In order to protect the internal circuits of such a device against high-voltage pulse penetrating through the permanently enabled remote control unit, HVR1 and HVR2 relays are provided with their own internal isolation. Alone, the remote-control unit does not require protection since it is designed to

290 | 16 Protection of diesel generators from HEMP

Figure 16.11: Active protection device for electronics (APD); LVR—multicontact low-voltage electromechanical relay; HVR—high-voltage relay; RCU—remote control unit; BB—black box (monitoring device); A, B, C—inputs and outputs of electronic equipment connected to a long cables run beyond the protected area; D, E, F—inputs/outputs used within the protected area.

operate only under the normal conditions, and its damage by HEMP does not influence the operability of the protected electronics. If there is a hazard of high-voltage pulse penetration through the long cables connected to the electronics and running beyond the protected area (such as the cabinet enclosing the equipment), such cables must be connected to electronics inputs/outputs through the contacts of additional high-voltage relays (HVR3–HVR5). Upon the remote instruction, the remote control unit activates the high-voltage relay, energizing the controller and feeding the multicontact relay coil which unblocks the controller’s inputs and outputs. At this time, the controller is ready for operation. After an operation, the device reverts to the original state upon the remote instruction. If some external circuits connected to the controller inputs cannot be shorted, the changeover contacts of the multicontact relay may be used instead of normally closed relay contacts, see Figure 16.12. These contacts will disconnect the external circuit before the controller inputs/outputs are shortened. The number of contacts on the LVR relay must correspond to the number of inputs and outputs on the protected electronics. Usually, the proposed electromechanical relays contain a maximum of eight or nine changeover contacts, see Figure 16.13, such as relay type R10 (Tyco Electronics); relay type C; and relay type D8-U200 (Mors Smitt), etc. Where applicable, several relays with its control coils connected in parallel can be used.

16.5 Active protection method for diesel-generator controller

| 291

Figure 16.12: Application of LVR changeover contacts when external circuits are not allowed to be shortened. OVP—over voltage protection components (such as varistors) ensuring relay contacts protection against the HEMP high-voltage pulse entering the device through the external cables A, B, C.

Figure 16.13: Multicontact electromechanical relays (eight or nine changeover contacts) type R10 (Tyco Electronics) and type C (Mors Smitt).

Another option is out-of-the-box relay modules consisting of 16 separate miniature relays, each with a single changeover contact, mounted on the compact printed boards provided with standard wire terminal strips, see Figure 16.14. Alternatively, the flat thin modules of miniature relays to be installed on the standard DIN-rail makes a difference, see Figure 16.15, and can be applied.

292 | 16 Protection of diesel generators from HEMP

Figure 16.14: Relay modules.

Figure 16.15: Flat thin relay modules to be installed on the standard DIN-rail.

In fact, any number of such modules mounted in line on DIN-rails located inside the cabinets can be installed. Today, the market proffers a wide choice of devices to be implemented as remotecontrol units RCUs. Such devices were described in Chapter 15. High-voltage relays (HVR) containing high-voltage contacts separated by the highvoltage isolation from the control coil (10–50 kV and above) are produced by such manufacturers as Meder Electronics, TPM Sanyu Electric, Pickering Electronics, SPS Electronic, Cynergy 3, Kilovac, Gigavac, Bright Toward Industrial, Gunther, Ross Engineering, Struthers-Dunn and Comus, among others. Basically, the relays of this type can be categorized as follows: low-power relays (50 W, 7.5–20-kV DC) and high-power relays (1–3 A switching current, 35–70-kV DC and greater). These two relay categories

16.5 Active protection method for diesel-generator controller

| 293

Figure 16.16: High-voltage reed switch relays selected for impulse insulation strength test. 1—type HA702-11824; 2—type HE24-2A69-BV548; 3—type 60-1-A-5/3; 4—type G81AB47. HV1 and HV2— outputs of high-voltage reed switch, LV—outputs of low-voltage control coil.

also significantly differ from each other in cost (60–100 USD for low-power relays and 400–1,000 USD for high-power relays). Since the proposed solution does not require the high-power switching ability under the high voltage, the high-voltage miniature relays based on a vacuum-sealed reed switch are the most suitable for this purpose. Sometimes, the rated breakdown strength of such relays (7.5–20 kV) is unable to ensure the reliable isolation upon HEMP, since its electric field strength can reach 50 kV/m. However, these parameters are rated by relay manufacturers of continuous DC voltage. Moreover, it is known that the dielectric-breakdown strength rises significantly under the pulse voltage, while HEMP is a very short pulse. Since neither manufacturer could inform us about the value of the impulse-insulation strength of its products, the author performed his own tests. For testing four types of high-voltage reed-switch relays were selected and kindly presented to author by the various manufacturers for the tests, see Table 16.1, Figure 16.16. All relays were tested with the pulse voltage of 1.2/50 µs and amplitude of 30 kV, see Figure 16.17, applied to the open contacts as well as between the contacts and the control coil.

294 | 16 Protection of diesel generators from HEMP Table 16.1: Basic parameters of high-voltage reed-switch relays tested at the pulse voltage. Type/Parameters Manufacturer Dielectric strength between contacts, kV peak Dielectric strength contact to coil, kV, peak Max. switching current, A Max. carry current, A Max. switching power, W Contact capacitance, pF – in data sheet – measured in whole device Cost, USD

60-1-A-5/3

HA702-11824

HE24-2A69-BV548

G81AB47

Pickering

TPM Sanyu Electric Inc.

Standex–Meder Electronics

Gigavac

15

18

20

11

–

–

25

–

3 3 50

3 4 50

3 5 50

– 5 (10) –

0.15 4.0

– 4.5

0.8 4.5

– 3.6

60

100

104

190

Figure 16.17: Test pulse and test bench. 1—memory oscilloscope type Hioki 8835; 2—high-voltage pulse generator type P35 (Haefely); 3—high-voltage divider with output ratio 1,000:1; 4—tested relay.

All relays successfully passed the test and confirmed our initial suggestion that the high-voltage relays with rated insulation strengths of 10, 15 20 kV were capable of withstanding much higher pulse voltages. It means that they can be used for ensuring effective protection of electronics against HEMP.

16.6 Conclusion | 295

Due to the significant difference in cost, the final choice of the relay may depend on the consumer budget and the required reliability. For example, the relatively expensive relays type HE24-2A69-BV548 produced by the German company Standex–Meder Electronics contain two in-series high-voltage reed-switches and ensure the largest rated isolation value. Accordingly, it can be suggested that they can withstand the larger amplitude of pulse voltage compared to the relays of other types, and therefore they have a significant advantage in terms of HEMP protection.

16.6 Conclusion Technical measures for DG protection from HEMP discussed in the chapter touch upon DGs of various typical sizes and purpose. Adoption of such measures is fairly easy for semiskilled technical staff and does not require large investments. In addition, it should be considered that DGs would not perform properly upon HEMP impact without such investments.

Bibliography [1] Gurevich V. Protection of Substation Critical Equipment against Intentional Electromagnetic Threats. Wiley, London, 2017, 240 pp. [2] Gurevich V. Cyber and Electromagnetic Threats in Modern Relay Protection. Taylor & Francis Group, Boca Raton, 2015, 205 p. [3] Gurevich V. I. Basic HEMP Protection Means for a Power Substation: A Quick Guide. Int. J. Res. Innov. Appl. Sci., 2017, Vol. II, No. IV. [4] Gurevich V. High Altitude Electromagnetic Pulse and Protection Means of Electric Equipment. Infra-Eng., 2018, 516 p. (by Russian). [5] Gurevich V. On Some Ways to Solve the Problem of Electromagnetic Compatibility of Relay Protection in the Power Industry. Ind. Power Eng., 1996, Vol. 3, pp. 25–27. [6] Gurevich V. Do We Need the Protection of Relay Protection?. Electr. Transm. Distrib., 2013, Vol. 2, pp. 94–97. [7] Gurevich V. Device for Protection of Relay Protection. Control Eng. Russ., 2013, Vol. 3, pp. 25–29.

17 Features of HEMP resilience-test methods for power system electronics 17.1 Introduction The ability of electronic equipment to operate efficiently under the external electromagnetic interferences is known as Electromagnetic Compatibility (EMC). The procedure for testing equipment for EMC is well proven and described in numerous standards. While HEMP, or rather its component E1, is just one form of electromagnetic interference disturbing power-system electronics operation [1, 2], it has several significant differences and features dictating the need for improvement and adjustment of known EMC-test procedures. There are dozens of reports and standards on HEMP aspects [3] but only a few dedicated to the procedure for testing the equipment resilience to such an impact [4, 5]. As these tests are highly sophisticated and power-engineering specialists have no experience in performing them in the field, two publications just mentioned do not provide enough background for effectively testing power system electronics. Several sporadic tests, such as testing the Digital Protective Relay SEL-311, can hardly be used as proper method due to the numerous errors, both in the test procedure and in the selected test pulse parameters [6]. Therefore, it is necessary to detail the features of HEMP resilience test methods for power system electronics.

17.2 Features of testing equipment on a HEMP simulator HEMP has the following features that are needed to be considered to develop the test procedure. The disturbance has a very short duration (a single pulse lasts for several nanoseconds); therefore, the failure of the equipment under test (EUT) must be registered within this period. This severely limits the number of EUT modes of operation to be controlled during the test. For example, during the HEMP immunity test, it is not possible to change the EUT modes of operation using the PC connected to the EUT in order to observe the EUT response to these changes (as is possible during the normal EMC test in the so-called anechoic chamber, where the EUT is exposed to electromagnetic radiation for a longer period). There is a danger of the appearance of the so-called soft faults, soft failures or soft errors, especially in electronic memory elements, which are difficult to discover instantly during the tests. The soft faults may show up in the tested apparatus only after a longer period, for example, upon access to the destroyed memory cells or program modules during certain operations. https://doi.org/10.1515/9783110639285-017

17.3 Test objectives |

297

These tests are expensive due to the limited number of test centers in the country and their affiliation to the Ministry of Defense. It is difficult to select the right EUT configuration, unlike the regular electromagnetic interferences: the HEMP has a rather more global nature than a local nature and affects both EUT and its feeding system, grounding system and channels of communication with other facilities. Therefore, instead of a single device, unit or module, a whole system of interconnected devices, units and modules, simulating the real-life environment, including a set of various and separated grounding points, needs to be tested.

17.3 Test objectives Due to the complexity and the high cost of the HEMP resilience tests, these tests should be applied only to a limited range of equipment types considered as critically important devices, the lack of which makes even partial operation of electric energy facilities impossible. The selection of the equipment should be the first stage of the test-plan development. The second stage should include the clear and transparent description of the test objective, as it will define both the test object and the test procedure. Potential test objectives can be defined as follows: 17.3.1 Test the resilience of operating equipment to the highest possible HEMP level without any protection. The objective of this test is to discover elements, units and systems exposed to HEMP, and thus requiring protection. 17.3.2 Test the effectiveness of the equipment protection against the maximum possible HEMP level using the minimum set of preinstalled protective means designed for operating a power system. These tests can be used to check the performance of the minimum set of protective means and to discover the types of equipment failures expected under HEMP. 17.3.3 Test the effectiveness of the equipment protection against the maximum possible HEMP level using the complete set of preinstalled protective means designed for newly constructed power systems. This test enables confirming the effectiveness of the most complex and expensive type of protection and justifying the costs of a protection system. 17.3.4 Test the operating equipment without special protective means under a series of pulses with the amplitude gradually increasing from 20 % to 100 % of the maximum possible level. The objective of this test is to: 1) find the equipment (or equipment types) most exposed to HEMP; 2) define the maximum HEMP amplitude that the unprotected equipment can safely withstand, in order to calculate the needed level of additional protection to improve the equipment’s own attenuation to the maximum level provided by the standard.

298 | 17 Features of HEMP resilience-test methods for power system electronics

17.4 Features of the test procedure As already mentioned, the number of controllable EUT operation modes during the test is very limited. Thus, these modes can include the unauthorized appearance or disappearance of signals at outputs of at least two EUTs interconnected via communication channels in the standby mode (waiting mode), or to high data-exchange mode (e. g., in emergency mode). In the latter case, the simulation of the emergency-mode start should be synchronized with the initiation of the HEMP pulse. After the test, because the complex microprocessor-based electronics may lack evident soft failures (even if no apparent damage or failures are registered during the test or the inspection), the equipment should further undergo a complete and thorough functional check. It means that the test designed to determine the required level of additional protection (see clause 17.3.4) should be followed by a functional check initiated after each level of HEMP impact. Clearly, this makes the test set much more complex because, after each higher amplitude pulse, the EUT should undergo the functional check. The EUT should be reconnected to the functional check systems after each HEMP test cycle. A programmable portable test system, preprogrammed to perform the particular functional checks, can simplify the test process. Many companies manufacture such systems (e. g., DOBLE, ISA, Omicron, Megger, etc.), and they are widely used in relay-protection testing. Since under the HEMP affect the grounding system acts as a huge antenna absorbing electromagnetic energy over the large area, delivering it directly to the grounded electronics, the test program must include the ground-system impact tests performed at two separated EUTs connected to the same grounding system at two remote points. As the HEMP simulator is not permitted to use the mesh bonded in the concrete base as the EUT grounding system, the separate grounding system, which is realized as a sufficiently large mesh, must be provided for the test purposes. The presence of both horizontal and vertical components of HEMP means that such a grounding mesh should be installed at 30–45° to the concrete base, rather than horizontally. The mesh can be assembled from separate sections interconnected with a suitable wire. Generally speaking, power-system electronic devices are connected to other electronics, sensors, power sources, electrical or electromechanical power units combined into the complex environment. For example, the relay protection system SCADA, a fire-protection system, is built on such a principle. Thus, the tests should cover the complete system rather than a single unit. For relay protection systems, such an arrangement can include two cabinets each equipped with a Digital Protective Relay (DPR), the battery acting as a power source, and a battery charger. These cabinets should be separated by the maximum allowable distance, connected to the grounding mesh and interconnected via a communication channel. DPR current and voltage inputs should be connected to the controlled current and voltage sources protected from the test-pulse impact. To arrange such protection, the source should be located in

17.4 Features of the test procedure

| 299

the isolated shielded section and connected to the EUT via a special filter (coupling– decoupling circuit), enabling the transmission of signals between the EUT and the protected equipment placed in the isolated section, but preventing the ingress of the HEMP test pulse. The current and voltage source must be equipped with remote controls to synchronize the startup with the HEMP test pulsing cycle. The state of DPR output circuits should also be considered during the test, see Figure 2. Usually, stationary test beds are pre-equipped with special shielded cables and filters designed for data transfer from the EUT, installed onto the test-bed to the isolated shielded section. As a rule, the power system electronic equipment is installed in the metal cabinets located in all-brick or concrete buildings significantly attenuating the HEMP impact, while other devices, such as grounding system, sensors, measuring current and voltage transformers and numerous outgoing cables, are located in an exposed area. Therefore, in real life, the various components of the common system are exposed to the electromagnetic impacts of different levels of energy. The traditional design of the test-bed radiating antenna, see Figure 17.1, includes the central section, where two parallel meshes: top and bottom are located at a fixed distance from each other, and two side sections, where the distance between the top and the bottom meshes is decreasing. Such design enables testing various components of the system under the varied

Figure 17.1: Arrangement of EUT in the test-bed. 1—Mobile battery 220 V, 2—Electrical cabinets separated from one another; 3—Tested electronics (such as Digital Protective Relays—DPR); 4—Communication devices; 5—Lockout relay controlled via DPR output circuits; 6—Battery charger; 7—Set of metal meshes comprising the ground-system model; 8—simulators of various modes of EUT operation synchronized with HEMP initiation system; 9—EUT status recorders; 10—Load with battery-charger output-voltage control unit.

300 | 17 Features of HEMP resilience-test methods for power system electronics impacts of the same test pulse, since the strength of the field between the top and the bottom meshes depends heavily on the distance between them. To obtain the conditions most closely relevant to real-life situations, the components of the tested system should be located on the different sections of the test bed. The presence of both vertical (directed from the high-altitude detonation point to the ground surface) and large horizontal component in the actual E1 component of HEMP is another aspect that needs to be considered during the HEMP test. To allow for the consideration of both HEMP components, the elements of the tested system located in-between the top and the bottom meshes should be arranged at a specific angle to the ground surface. The test procedure should include the registration of the EUT parameters under the HEMP with proper and adequately protected recorders. These could include external loggers, memory-impulse scopes with automatic triggers and also emergency event recorders built into the EUT and operating in parallel with other recording devices. The selection of protective means (such as special filters, overvoltage arresters, shielded cables, etc.) or choosing an unprotected mode, should depend on the defined test objective.

17.5 Test modes and test-pulse parameters According to IEC 61000-4-25, the test for electronics immunity to HEMP must be divided into two parts: radiated immunity (RI) and the conducted immunity (CI). In its turn, CI is divided into two types: a pulse voltage applied to apparatus inputs and pulse currents induced in long cables and wires. Determination of proper test concept (out of six available) is the first step for defining the rules for a certain test. Standards 61000-2-11 and 61000-5-3 describe these concepts. Concept 2b can be suitable for the EUT located inside the major reinforced concrete or all-brick building protected against lightning, but having no special protective filters. This concept allows for attenuation of EUI by 20 dB in the frequency band of 100 MHz–30 MHz due to the building structure. For this concept and for component E1, the strength of the radiation electrical field acting upon the tested equipment should be equal to kV/m (level R4) and the strength of the magnetic field should be equal to 13.3 A/m. Let us compare: for wooden buildings, providing zero EI attenuation, the strength of the electrical field equals to 50 kV/m (level R7). For the same concept and component E2, the strength of the electrical field should be equal to 10 V/m and the strength of magnetic field—0.08 A/m. EI pulse parameters are described in standards 61000-29, 61000-2-10, 61000-2-11, MIL-STD-461F [5]: pulse rise time (front edge) is 2.5 nanoseconds; pulse width is 25 nanoseconds; pulse shape is shown in Figure 17.2.

17.5 Test modes and test-pulse parameters | 301

Figure 17.2: The shape of the E1 component of HEMP according to standards: IEC 61000-2-9, IEC 61000-2-10, IEC 61000-2-11, and MIL-STD-461F.

The next step is to choose the CI test level according to IEC 61000-4-25. For the concept 2b, due to the presence of cables connected to the tested facility and not buried in the ground, the level of test impact should correspond to E8 (to allow for 50 % probability of the facility immunity) or E9 (to allow for 99 % probability). For E8 level, it is assumed that the tested facility is immune to the pulse voltage of 8 kV, and—for E9 level—to the pulse voltage of 16 kV. A probability of 50 % is deemed normal according to the standard and can be applied to the civil equipment. By “CI test voltage pulse” is meant the so-called “Electrical Fast Transient (EFT) pulse” with parameters (except the test voltage amplitude), and the test procedure is well described in IEC 61000-4-4, see Figure 17.3. The amplitude of the HEMP test voltage (designated as “special”) is marked as “X” in Table 1 of this standard and corresponds to E8 and E9 levels. Previously, EFT generators with a required level of output voltage 8 kV were manufactured by TESEQ, Kentech Instruments Ltd. and Thermo Electron Corp., and they were built on triggered vacuum spark-gaps to generate the test pulses. With the advent of solid-state switching elements, such as IGBT-transistors, triggered vacuum sparkgap generators were rendered obsolete by all three companies because the pulses generated by the transistors were much more stable and “correct” compared to pulses generated by the vacuum spark gaps. Unfortunately, the improved stability of generated pulses was accompanied by the decrease of their amplitude. Our analysis shows that EFT generators presently available on the market do not fully satisfy the requirements of pulse-amplitude standards (8 kV). Generator type PEFT 8010, manufactured by a Swedish company, Haefely EMC Technology, best matches the required pulse-amplitude value.

302 | 17 Features of HEMP resilience-test methods for power system electronics

Figure 17.3: Electrical Fast Transient (EFT)—fast pulse (IEC 61000-4-4).

17.6 Performance criteria The acceptable type of certain EUT response to electromagnetic interference during and after the particular test type is known as the performance criteria. The acceptable responses are listed here: – Graphical distortions on EUT display, display flickers or blinks off. – Display is showing incorrect data. – Signals or data distortion or loss. – Communication channel distortion or total loss. – Sensor malfunctions. – System false activation. – Sharp degradation of system’s ability to handle and transmit the information, and maloperation of the system.

17.7 Conclusion

– – – – –

| 303

Software failures. System hangs. Automatic restart of computerizing electronic system. Complete system failure due to the power-source fault or fuse blowing in feed circuits. Physical damage to the internal electronic components of EUT.

Basic standard IEC 61000-4-25 [4] describes only five types of performance criteria summarizing the EUT responses just listed: A) Normal performance within specified limits. B) Temporary degradation or loss of functions or performance that is self-recoverable. C) Temporary degradation or loss of functions or performance, which requires operator intervention or a system reset. D) Degradation or loss-of-function that is not recoverable, due to a loss of data or damage to equipment. E) Degradation that can lead to a safety problem (for example fire). For the test planning, the performance criteria should be predetermined separately for each type of test. These criteria should be used for assessing the EUT test results upon each test type. Apparently, only criteria A and B are relevant to power system electronics, so only these two are the acceptable choices.

17.7 Conclusion 1.

Due to the complexity and the high cost of the HEMP immunity test, such a test should be applied only to a limited range of equipment types considered as critically important devices, the lack of which makes even partial operation of electric energy facilities impossible. 2. The HEMP immunity test plan should start from the clear and transparent definition of the test objective or objectives. 3. The power system electronics should be tested as a whole system rather than as a set of individual devices. Such a whole system should include several electronic devices (two at least) interconnected by means of a communication channel, and connected to the common ground system, feeding source, control signal sources, etc. The plan-development phase should include the system flowcharting and the compiling of a list of necessary equipment to perform the test. 4. Depending on the EUT type, the following should be defined in advance: 1) the list of parameters to be controlled during the HEMP impact period, 2) the parameter checkout methods and 3) the types of apparatus needed to record the changes of the parameters during the test performance.

304 | 17 Features of HEMP resilience-test methods for power system electronics 5.

6.

7.

The results of the HEMP impact can become evident only after a certain period following the test. Thus, the control of the EUT state during the test should be supplemented by the full EUT functional check performed after the test on the HEMP simulator test-bed and then after feeding a high-voltage test pulse to the EUT using a contact method. In addition to the complete set of standard EMC tests, the test for immunity of the power system equipment to HEMP should consider the following two types of impacts: a) Electromagnetic pulse—two nanoseconds rise time, 25 nanoseconds pulse width and five to 50-kV/m field strength. b) Fast pulse (EFT) fed to EUT inputs using contact method—5/50 nanoseconds, 8 kV amplitude. Standard criteria A and B should be selected as the performance criterion for power system electronics.

Bibliography [1] Gurevich, V. (2015). Cyber and Electromagnetic Threats in Modern Relay Protection. Taylor & Francis Group, Boca Raton. [2] Gurevich, V. (2016). Protection of Substation Critical Equipment against Intentional Electromagnetic Threats. Wiley. [3] Gurevich, V. I. EMP and Its Impact on Electrical Power System: Standards and Reports. Problems of Power Engineering, 2016. [4] IEC 61000-4-25 Electromagnetic Compatibility (EMC) – Part 4-25: Testing and Measurement Techniques. HEMP Immunity Test Methods for Equipment and Systems. [5] MIL-STD-461F Requirements for the Control of Electromagnetic Interference Characteristics of Subsystems and Equipment, RS105, 2007. [6] Gurevich, V. Problems in Testing Digital Protective Relays for Immunity to Intentional Destructive Electromagnetic Impacts. Continuation of the Theme. Electr. Eng. Electromech., 2015, Vol. 6, pp. 66–69.

18 Methods and means of evaluation of the effectiveness of HEMP protection of the installed power-system 18.1 Introduction Recently, the challenge of protecting the equipment in the electric power system from high-altitude electromagnetic pulse (HEMP) has become very relevant due to the significant increase in power industry vulnerability and exposure to HEMP, and accounts for the ever-growing use of microelectronic and microprocessor systems in power generation, transmission and distribution processes. The military institutions of many developed countries have a tendency to intensively develop the means and methods for remote damage of microelectronic and microprocessor systems to have the ability to destroy the infrastructures of enemy countries during possible future conflicts [1]. Today, measures designed for protection of equipment in electric power systems from HEMP are also intensively developed [2]. However, since the power industry is in the civilian sector, it is hardly possible to use proven and tested military means of protection in this industry as they are rather expensive. Therefore, we need to compromise and find solutions to develop new and more affordable means of protection for this industry. In this context, the testing of the effectiveness of such new and previously unused means of protection is especially relevant. This chapter reviews the means designed for testing the effectiveness of protection of the electric equipment from HEMP that are available on the market and presents recommendations for choosing the right device.

18.2 Testing of equipment resilience to direct impact of the HEMP electrical field (E1-component) It is known that high-altitude explosions of nuclear devices (altitudes of 30–400 km) create a pulse electric field with a strength of up to 50 kV/m (the so-called E1 component of HEMP) near the ground surface which spans a large territory. There are special simulators designed for testing the sustainability of electric equipment when exposed to the E1 component. They can be categorized as either compact laboratory or large fixed simulators. The fixed HEMP test bed usually consists of the concrete base with bonded-in metal mesh acting as the first electrode, and the overhead metal mesh located 10– 15 m above the concrete base, which acts as the second electrode. A high-voltage pulse applied between these two electrodes is sent from the output of a special type of generator. Usually, it is a Marx Generator built on a set of high-voltage capacitors and https://doi.org/10.1515/9783110639285-018

306 | 18 Methods and means of evaluation of the effectiveness of HEMP protection switching spark gaps, controlled by compressed air and immersed in a large oil-bath or an SF6 gas reservoir. The size of such a test bed makes it suitable for testing very large items, such as tanks and aircraft. Many countries have large test beds generating HEMP with the required values. The common technical parameters of such beds are mentioned in IEC standard [3]. For example, there are several such test beds in the US (e. g., TORUS, ALECS, ARES, WSMR, ATLAS, VPBW, etc.) and two in Russia: 1) ALLUR COMPLEX in the High-Voltage Scientific-Research Center of Federal State Unitary Enterprise Russian National Electric Technical Institute (Istra, Moscow Region). 2) The research center in the Federal State Unitary Enterprise 12th Central ScientificResearch Institute of the Ministry of Defence of Russian Federation, Sergiyev Posad. Many European countries and also Israel (in the Rafael Advanced Defense Systems, Haifa) have a large test bed. Ukraine also has a similar test-bed, located at the “Molniya” Research and Design Institute, Kharkov. Compact laboratory simulators are manufactured by several companies and are available on the market, see Figure 18.1.

Figure 18.1: Compact test-beds for testing electronic apparatus sustainability to HEMP. a—Montena Technology, b—Applied Physical Electronics.

The parameters of pulses generated by these compact beds fully correspond to the requirements of MIL-STD-188-125-1 [4] and MIL-STD-461F [5] standards. However, such beds are rather expensive (over 100,000 USD). In addition, they are designed for testing smaller items, such as standalone digital protection relays (DPR). However, as we see from [6], since in the power industry the electronic devices (particularly DPRs) are used within the large distributed systems consisting of numerous long cables (acting as antennas absorbing the HEMP energy), the testing of individual devices, without

18.3 Equipment for HEMP filter testing

| 307

due regard to their numerous interconnections with other devices, is hardly reasonable. While such testing of individual devices could be interesting for developers, generally speaking, it is not useful for the power industry. Besides, such individual devices are mounted inside the metal cabinets. It means that the test should be combined (device + cabinet), instead of focusing on the single device only. All the just-mentioned facts lead to the conclusion that it is unreasonable to purchase very expensive compact laboratory test beds for testing individual types of electronic equipment for sustainability during and after HEMP. The testing of the set of equipment on fixed test beds can be much more effective [7].

18.3 Equipment for HEMP filter testing The feasibility of the application of special HEMP filters offered by numerous manufacturers is uncertain [7]. Components such as dismountable ferrite beads and chokes can be widely used for limiting current pulse amplitude within the electric circuits affected by HEMP impact. However, this refers to the very different operating currents with significantly different frequencies (from DC to high-frequency current) and amplitudes (from milliamps to hundreds of amperes). The electromagnetic properties of ferrite beads and chokes change significantly when exposed to the operating current frequency and amplitude change. Also, they essentially depend on analogous parameters of the noise to be eliminated by such beads [8]. Thus, the effectiveness of ferrite beads (more correctly, bead sets) and chokes must be tested under the conditions closest to the real-world operation, as well as to the real parameters of the impacting electromagnetic pulse. The effectiveness of ferrite beads is determined by their ability to attenuate the noise signal within a particular frequency range. While in this situation the noise is constituted by HEMP, the first stage should include the determination of the frequency range that the bead testing equipment must operate within. Then, this frequency range must be used to determine the test equipment corresponding to this parameter. Evaluation of various standards shows, see Figure 18.2, that the required frequency range for the test equipment can be selected in the range from several hundreds of kHz to one GHz. It is known that insertion loss (the degree of noise attenuation) is the basic characteristic of the electromagnetic filter. Within the system of so-called S-parameters (scattering parameters), the filters are characterized by direct and backward transmission coefficients, such as S21 and S12 . There are special devices used for measurement of these parameters, i. e., Vector Network Analyzers (VNAs). As a rule, VNAs have two ports (signal source and signal receiver) and are designed for measuring one set (onepatch) of S-parameters (S21 and S12 ), or two (two-patch) sets of them (S22 and S11 in

308 | 18 Methods and means of evaluation of the effectiveness of HEMP protection

Figure 18.2: Determination of HEMP frequency range in various standards.

addition to the previous set). Additional S-parameters (S22 and S11 ) determine the degree of signal reflection and are not peculiar to the filters. During the test, the filter is connected between two VNA ports, and the filter effectiveness is determined by the signal difference (in dB) on the VNA ports over the whole selected frequency range. Basically, the filter effectiveness is represented by a diagram demonstrating the filter inserted losses as a function of the input signal frequency. The dynamic range characterizing the signal attenuation ratio to be measured by the device is another very important parameter of VNA. As a rule, all VNAs have a dynamic range of 100– 120 dB, minimum. Therefore, the basic requirements for test devices for HEMP filters are as follows: – Implementable features—S21 and S12 . – Frequency range: 100–300 kHz up to one GHz. – Dynamic range—at least 100 dB.

18.3 Equipment for HEMP filter testing

| 309

All other numerous VNA parameters, usually enumerated by the manufacturers, are not critical for the desired goal. In reality, it is rather difficult to choose one particular device corresponding to all the mentioned requirements due to the wide breadth of such devices on the market, see Figures 18.3 and 18.4, with significantly diverse technical functions and costs, see Table 18.1.

Figure 18.3: Several types of standalone VNAs available on the market. Top—fixed devices with numerous additional features and large display, bottom—compact mobile devices.

Figure 18.4: VNAs w/o displays designed to be connected to PC.

310 | 18 Methods and means of evaluation of the effectiveness of HEMP protection Table 18.1: The cost of the most common types of VNAs. Type VNA (Manufacturer)

Frequency Range

Cost (USD)

ZNL3 (Rohde & Schwarz) C1209 (Cooper Mountain Techn.) LA19-13-03 (LA Techniques Ltd.) MS2024B (Anritsu) N9913A (Keysight Technology) TW4600A (Techwin Industry Co.) GSP-9330TG (GW Instek) TTR5003A (Tektronix) Planar TR5048 (Cooper Mountain Techn.) PicoVNA 106 (Picotech) VNAuhf (Array Solutions) USB-SA44B + USB-TG44A (Signal Hound) Planar TR1300/1 (Cooper Mountain Techn.) VNA 6050-5100 (AEA Technology)

5 kHz–3 GHz 100 kHz–9 GHz 300 kHz–8 GHz 50 kHz–4 GHz 30 kHz–4 GHz 100 kHz–3 GHz 9 kHz–3.3 GHz 100 kHz–6 GHz 20 kHz–4.8 GHz 300 kHz–6 GHz 5 kHz–1.2 GHz 10 Hz–4.4 GHz 300 kHz–1.3 GHz 100 kHz–1 GHz

20,000 22,000 10,600 10,450 15,000 9,300 7,000 9,000 7,800 6,000 1,250 1,750 2,900 2,300

Basically, the increase in upper-edge frequency leads to sharp increase in cost, so only the models with minimal upper-edge frequencies were chosen for the analysis. However, it transpired that there is even more to consider. Many manufacturers of spectrum analyzers equip their devices with integrated tracking generators, allowing the simple spectrum analyzer to be used as VNAs. Also, some of them offer sets consisting of two separate devices: a spectrum analyzer and a tracking generator, see Figure 18.5. Two such small boxes should be interconnected and then connected to a PC in order to use its display. Thus, we assemble a very compact, simple and inexpensive device.

Figure 18.5: Simple and inexpensive combined VNA–spectrum analyzer (USB-SA44B) c/w tracking generator (USB-TG44A).

18.4 Equipment for evaluation of the shielding effectiveness | 311

Figure 18.6: Two VNAs with the best value-for-money parameters for the particular application.

Based on the value-for-money parameter for this particular application, two VNAs, see Figure 18.6, can be selected (due to their satisfactory parameters and minimal cost) from the wide variety of offers available on the market. Therefore, it is the decision of the user which one to choose: the absolutely standalone compact PC-enabled unit with its own display or the device designed as a supplement to a PC. Also, along with a VNA, the set for VNA calibration must be purchased. In the simplest case, such a calibration set consists of three connectors: open (without additional inner components), short-circuited and 50-ohm (internal resistance) connector.

18.4 Equipment designed for evaluation of the effectiveness of building, room and cabinet shielding The devices of this type contain transmitters, receivers and a set of antennas for various frequency ranges. Before using such devices, they should be reset to zero. To do this, the receiver with the directed antenna should be installed near and opposite to the transmitter with the directed antenna. Then, the zero-value attenuation should be reached while the transmitter and the receiver remain active, see Figure 18.7. Subsequently, the transmitter should be placed outside the protected facility, and the receiver should be placed in the same facility at a distance equal to the distance used upon resetting to zero. Then, the difference between the sent and the received signal should be used for evaluating the facility’s shielding effectiveness (i. e., signal attenuation rate). As for such devices, the market situation is totally different from that of VNAs, already described. As it happened, the market offers only a very limited range of devices for the evaluation of shielding effectiveness, see Figure 18.8. Evaluation of the frequency range showed that only one of the devices (SEMS) more or less corresponds to the specific application. However, it is not 100 % suitable

312 | 18 Methods and means of evaluation of the effectiveness of HEMP protection

Figure 18.7: Adjustment (resetting to zero) and usage of device for evaluation of shielding effectiveness.

Figure 18.8: Devices designed for evaluation of shielding effectiveness available on the market.

Figure 18.9: Using VNA for evaluation of shielding effectiveness.

because its frequency range (10 kHz–300 MHz) does not reach one GHz. However, that was the best option that was available. The device costs 16,000 USD. Some publications state that VNAs can also be used for the shielding-effectiveness evaluation, see Figure 18.9. They suggest to connect the transmitter and receiver an-

18.5 Pulse voltage generators | 313

tennas to corresponding VNA ports and make the same adjustment (resetting to zero) procedure as for the special device. Probably it is really possible and efficient. If true, the market offers many different VNAs suitable for every application and less expensive ones as compared to the special standalone device. However, there is a minor problem related to the necessity to lead the receiver antenna wire out of the protected area according to the standard requirements. In any case, it should be noted that I did not test such a VNA application.

18.5 Pulse voltage generators The test voltage pulse applied directly to the inputs of the equipment under test means the so-called Electrical Fast Transient (EFT). It is a fast pulse whose parameters (except for the amplitude of the test voltage) and the test method are described in the standard IEC 61000-4-4, see Figure 18.10. The amplitude of HEMP-test voltage (designated as “special”) is denoted by “X” in Table 1 of this standard and corresponds to the E8 and E9 levels. Cited research [7] has demonstrated that, under the typical power industry conditions, the amplitude of this pulse should be equal to 8 kV. Previously, EFT generators with a required level of output voltage 8 kV were manufactured by TESEQ, Kentech Instruments Ltd., and Thermo Electron Corp. (see Table 18.2), and they were built based on triggered vacuum gaps generating test pulses. With the advent of solid-state switching elements, such as IGBT-transistors, triggered vacuum-gap generators were discontinued by all three companies because the pulses generated by the transistors were much more stable and “correct” compared to pulses generated by Table 18.2: Maximum amplitude of 5/50 nanoseconds output-voltage pulse of EFT generators (IEC 61000-4-4) available on the market. Type of EFT generator

Manufacturer

PEFT 8010 NSG 2025* J0101031/3* KeyTek ECAT E421* FNS-AX3-A16B EFT 500N8 TRA3000 EFT 6501 EFT-4060B EFT500 AXOS8

Haefely EMC Technology TESEQ Kentech Instruments Ltd. Thermo Electron Corp. NoiseKen Laboratory Co. EMTEST (Ametek) EMC Partner Schaffner Shanghai Yi Pai Electronmagnetic Techn. Suzhou 3Ctest Electronic Co. Hipotronics

* obsolete

Maximal pulse magnitude 7.3 8 8 8 4.8 7 5 4.4 6.6 5 5

314 | 18 Methods and means of evaluation of the effectiveness of HEMP protection

Figure 18.10: Electrical Fast Transient (EFT)—fast pulse (IEC 61000-4-4).

the vacuum gaps. Unfortunately, the improved stability of generated pulses was accompanied by a decrease of their amplitude. Our analysis shows that EFT generators presently available on the market do not fully satisfy the requirements of pulse amplitude standards (eight kV). Generator type PEFT 8010 (Haefely EMC Technology) and EFT 500N8 (Ametek) best match the required pulse-amplitude value. They consist of integral filters (Coupling-Decoupling Network—CDN) protecting the supply network from the penetration of pulses generated by the device, see Figure 18.11. They cost 25,000–30,000 USD. Similarly, for all these devices, the generators must be calibrated periodically. For this purpose, sets of special high voltage non-inductive dividers of 50 and 1,000 ohms, receiving the generator output pulse are provided, see Figure 18.12. The amplitude and length of such a pulse (with corresponding dividing ratio) are measured by the oscilloscope.

18.6 Conclusion | 315

Figure 18.11: EFT-generators with the parameters best matched to required pulse amplitude parameters.

Figure 18.12: Calibration sets for EFT-generators.

18.6 Conclusion This review of devices available on the market and recommendations for choosing a capable device can be helpful to the users of power systems. They can make both the evaluation of the vulnerability of existing unprotected equipment, as well as the assessment of the effectiveness of the protective means and methods for such equipment from HEMP, much easier and simpler.

Bibliography [1] Gurevich, V., Cyber and Electromagnetic Threats in Modern Relay Protection. Taylor & Francis Group, Boca Raton, 2015, 205 p. [2] Gurevich, V., Protection of Substation Critical Equipment against Intentional Electromagnetic Threats. Wiley, London, 2017, 240 p. [3] IEC 61000-4-32 Electromagnetic Compatibility (EMC) – Part 4-32: Testing and Measurement Techniques. High-Altitude Electromagnetic Pulse (HEMP) Simulator Compendium. [4] MIL-STD-188-125-1 High–Altitude Electromagnetic Pulse (HEMP) Protection for Ground Based C41 Facilities Performing Critical. Time-Urgent Mission. Part 1 Fixed Facilities, 2005.

316 | 18 Methods and means of evaluation of the effectiveness of HEMP protection

[5] MIL-STD-461F Requirements for the Control of Electromagnetic Interference Characteristics of Subsystems and Equipment, 2007. [6] Gurevich, V. Problems in Testing Digital Protective Relays for Immunity to Intentional Destructive Electromagnetic Impacts. Glob. J. Adv. Res., 2014, Vol. 1, No. 2, pp. 159–173. [7] Gurevich, V. Main Principles of Electromagnetic Pulse Immunity Test Methods for Power System Electronics. Int. J. Res. Stud. Electr. Electron. Eng., Vol. 2, No. 2, 2016, pp. 1–8. [8] Gurevich, V., The Problem of Correct Choice of Ferrite Beads. Electr. Eng. Electromech., 2016, Vol. 2, pp. 71–73.

19 Features of testing digital protective relays resilience to HEMP 19.1 Use of performance criterion during the electromagnetic compatibility (EMC) test of electronic equipment The response of an object under test (OUT) to electromagnetic impacts (EI) can be variable. For example, the OUT can be fully inoperative due to a breakdown of internal electronic components, while on the other hand, it can be temporarily non-operative only during the impact of either the impulse or electromagnetic field. Another possibility is a short-term fault in the software operation affected by the impulse voltage supplied to the OUT, which may require (or not) that the operator reset the internal program of the OUT. There are many types of responses of the OUT to EI. The acceptable response of this type of OUT to electromagnetic impact under some type of trial is called “performance criterion” (PC). The performance criterion is an extremely important indicator in the tests for EMC. When properly selected, it enables reaching a conclusion whether a specific device has passed the specific test or not. However, the EMC standards do not contain (and they are unlikely to contain) the methods of the correct selection of these criteria. As a rule, everything is limited by a sentence like: “Selection of the strictness degree for performance criterion is performed by people who develop and approve performance specifications and technical conditions” and a chart from which a specific PC can be selected out of three or four criteria offered by the specific standard. This is obvious, since the correct choice depends on the specific type of OUT and specific modes and conditions of its operation. Moreover, a different PC can be selected for the same type of OUT depending on its operation mode, connection diagram, purpose for use, working environment, etc. Thus, the understanding of specific features of each single OUT is very important since the choice of one or another performance criterion enables making the decision of whether this specific OUT fits (or doesn’t fit) a specific working environment based on the trial results.

19.2 Features of using performance criterion during the HEMP resilience test of digital protective relays (DPR) The North American Electric Reliability Council (NERC) established a list of equipment that needs to be tested for immunity to HEMP upon the request of the special commission: Congressional “Commission to Assess the Threat to the United States from Electromagnetic Pulse (EMP) Attack”. The list includes, in particular, digital protective relays (DPR) and SCADA (Supervisory Control and Data Acquisition—a general name for software and hardware systems of various types that provides real-time data collection from numerous sensors and detectors, and whereat processing, archiving, displaying and transferring information about objects under monitoring. The system https://doi.org/10.1515/9783110639285-019

318 | 19 Features of testing digital protective relays resilience to HEMP provides also transferring of operator’s commands to remote sites and is the basic of Substation Control System—SCS) system. Metatech Company conducted tests of the SEL-311L DPR (Differential Line Protection) and SEL-2032 controller of a SCADA system (Figure 19.1) under a shortened test program. The test was performed only for immunity to the E1 component of HEMP. The results of these tests are presented in the Meta-R-320 [2] report.

Figure 19.1: Digital protective relays SEL-311L type and SEL-2032 controller for a SCADA system produced by Schweitzer Engineering Laboratories (USA) subjected to testing for HEMP immunity.

As indicated in the report, evaluation of the correctness of operation and lack of damage after each test (and not during the test) was used as performance criterion in the DPR and SCADA controller tests. During the tests short-duration (5/50 ns) high-voltage impulses with an amplitude of up to 8 kV were applied to various terminals of the devices. The report also mentioned that application of impulses with an amplitude of up to 3.2 kV to the serial port resulted in spontaneous DPR switch-off, but then it returned to normal operation mode. Some other ports (e. g., IRIG—Inter-Range Instrumentation Group time code— time synchronization port) were damaged at as low as 600 V. The Ethernet communication module of the SCADA controller was damaged at 1.2 kV. The report also suggests that the record of oscillograph tests of current and voltage rates supplied to the relay’s terminals were selected as one of the additional parameters of the PC. It is mentioned in the report that abnormalities in the record were not revealed during testing.

19.3 Criticism of the DPR testing method used 1. In our opinion [1], it is incorrect to use the PC based on the DPR damages check after it has been subjected to interference [2]. This does not allow making a definite conclusion about the immunity of DPR to this interference. This is due to the fact that DPR possesses several specific features reviewed in [3, 4] as compared to the SCADA system. With all the importance and responsibility of the SCADA system, it is designed

19.3 Criticism of the DPR testing method used |

319

first and foremost for automatic collection, processing and displaying information. Despite the fact that the system includes the so-called Remote Terminal Units (RTU), remotely controlled actuating units, they cannot work in the automatic mode and are only intended for performance of operator commands from a remote control center. The majority of modern substations work in the automatic mode without any operator. Manual control of circuit breaker (CB) status on these substations (i. e., literally control of the power system configuration) is performed by an operator sitting in a remote control center through the SCADA system, which is susceptible to HEMP impacts. This is why, in case of HEMP impact, the remote control of a substation from the control center will likely be lost, and the configuration of the power system will only be determined by the relay protection system—the only system that can automatically control the CB status. At the same time, the DPRs, which comprise the foundation of the modern relay protection, are constantly sharing information and commands in the automatic mode via communication channels susceptible to HEMP (unlike the SCADA system, where critical control commands are delivered to CB spearheaded by a dispatcher). In the event of incorrect actions of the automatically operating relay protection, where the dispatcher cannot intrude, such as unnecessary operation caused by HEMP impact, the electric power system and then the whole energy system can fully collapse. This is one of the reasons why the digital protection relay should be tested for HEMP impacts during operation and not be checked for damages after the impact of interference. 2. There are various paths of the entrance of electromagnetic interference (represented by impulses supplied to protected terminals) and high-voltage electromagnetic waves (which enter directly into internal high-sensitive electronic components or through unprotected terminals/ports of electronic units, as well as through the multiple cables connected to DPR and functioning as antennas absorbing electromagnetic energy) to DPR. Moreover, the intentional destructive electromagnetic impacts (IDEI) is not limited to HEMP only. It includes directed ultra-broadband high-frequency emissions of special sources with power ratings of several gigawatts and is intended for remote destruction of electronic equipment [4]. Unfortunately, the danger is caused not only by special-purpose devices intended for affecting electronic equipment, but also by emissions of ordinary powerful radars. For instance, in 1999, there was an officially registered event of devastating failure of the SCADA system at the San Diego County Water Authority Company, which supplies water to San Diego, CA. The reason for the failure was an emission of marine radar located 25 miles away from the city. In 1980, a similar case was registered on a gas supply line located 1.5 km from Den Helder port in the Netherlands. The latter case of SCADA system damage by a marine radar resulted in a powerful gas explosion. This is why testing of DPR immunity to IDEI should not be limited by applying high-voltage impulses to certain terminals only. It should be accompanied by exposure of the OUT to electromagnetic emission from a directional antenna as stipulated by corresponding standards [1].

320 | 19 Features of testing digital protective relays resilience to HEMP 3. It should be taken into consideration that a HEMP occurrence will affect not only highly sensitive electronic equipment (DPR, hardware of SCADA system), but also power facilities of energy systems, such as linear insulators, transformers and power generators. It should be noted that this equipment will be affected by not only the E1 component of HEMP (modeled during trials [2]) under these circumstances, but also by the other two components, i. e., E2 and E3 [1]. Previous research [2] conducted in the former Soviet Union and the US shows that the effects on all components by HEMP can result in damage of high-voltage power equipment, such as the breakdown of linear insulators, saturation and burning of power transformers, punctures of power generator insulation, etc. In other words, the moment of impact of a powerful electromagnetic interference on DPR corresponds in time with the moment of the changing of the internal state of DPR elements, which is due to the appearance of emergency rates of controlled current and voltage at its terminals. How will DPR behave under this mode? Will the HEMP-affected relay protection be able to disconnect the saturating transformer or damaged part of the overhead line of damaged cable? Won’t the common directional operation of various DPR be a cause of full disintegration and collapse of the energy system? The research conducted in [2] does not provide answers to these questions. “We have produced designs so complicated that we cannot possibly anticipate all the possible interactions of the inevitable failures; we add safety devices that are deceived or avoided or defeated by hidden paths in the systems.”—wrote the famous specialist on reliability and susceptibility of complex systems Charles Perrow [5]. C. Perrow calls this problem “incomprehensibility” because even the ordinary incident can trigger interactions that are “not only unexpected, but also unpredictable for a certain critical period of time.” In most accidents, nobody could expect that certain “interaction algorithms” will trigger others. Thus, nobody could predict what happened. This is relevant to the modern, rather complicated and branched relay protection systems, the behavior of which is difficult to predict under the HEMP impact.

19.4 Analysis of the result of the second independent trial of the same type of DPR Another test of the same type of DPR (by an odd coincidence) is reported in a promotional presentation of the producer of these devices—Schweitzer Engineering Laboratories Company—SEL [6]. The presentation covers the results of testing of SEL-311 DPR samples for HEMP and electromagnetic impact, on the test-benches of US Army’s Picatinny Arsenal in New Jersey (see Figure 19.2). The promotional presentation suggests that all tests were successful. At the same time, deeper analysis of the material reveals several odd things. For instance, the advertisement depicted in Figure 19.3 suggests that the SEL-311 was tested “at field

19.4 Analysis of the result of the second independent trial of the same type of DPR

| 321

Figure 19.2: DPR SEL-311 type test for IDEI impact on Picatinny Arsenal test-benches in New Jersey [6].

Figure 19.3: The text from the promotional brochure of SEL Company [6].

strength varied from 25 to 1000 V/m whereas the military standard MIL-STD-461 requires only 50 V/m.” The specialists of the renowned SEL Company demonstrated a rather odd ignorance in their report. The fact is that, according to the MIL-STD-461 (and also MIL-STD-188-125-1 [13]), the units of measurement of the field strength under IDEI impact are kilovolts, while they report volts, and the figure “50” is presented in the standards as 50 kV/m instead of 50 V/m (Figure RS105-1, page 155 of MIL-STD461G). The bar diagram illustrated in Figure 19.4 is even odder. It shows that in reality the field strength of 1000 V/m was implemented for testing at 1000–1500 MHz frequencies, while at other frequencies it was almost two times lower. Another factor is that the dependence of amplitude on the frequency does not correspond to MIL-STD-461. It is obvious from the diagram that the levels of field strength are limited by the onset of instability of relay functioning (yellow areas on the tops of the bars). In other words, the diagram shows the area of steady operation of a separately installed (out-

322 | 19 Features of testing digital protective relays resilience to HEMP

Figure 19.4: Parameters of electromagnetic emission during SEL-311 DPR testing [6].

Figure 19.5: A diagram from page 155 (Figure RS105-1) of MIL-STD-461G to compare with the diagram illustrated in Figure 19.4 (10 ns corresponds to a frequency of 1 MHz).

side the relay protection system) SEL terminal. This implies that the relay doesn’t permit steady operation outside the area of values represented in this diagram with its extremely low levels of electromagnetic-field strength. When comparing it with the above mentioned MIL-STD-416, see Figure 19.5, one can see that the applied parameters of testing impacts are far away from the requirements of this standard. Considering the oddness of parameters selection to test SEL-311

19.5 Analysis of the result of the third independent trial of the same type of DPR

| 323

immunity to IDEI, who can take seriously the manufacturer’s statement that these relays are resistant to IDEI? Another problem is the selection of a single DPR terminal as an OUT. As a rule, these terminals are manufactured in metal housings that effectively weaken the electromagnetic emission. This is why the test results for electromagnetic impact on this single terminal are expected to be positive. In the field conditions, DPR is entangled by multiple cables acting as antennas and absorbing electromagnetic energy that deliver it to the internal elements of DPR; multiple terminals of DPR are interconnected through corresponding communication instruments susceptible to IDEI impact. Thus, the test should be performed on the whole relay protection system while in operation rather than on its single terminal. An example of the correct approach to testing of complex systems to which relay protection undoubtedly belongs is represented by the SCADA-system testing described in [7], see Figure 19.6.

Figure 19.6: Testing of SCADA system for immunity to HEMP [7]. An antenna system of EMI simulator is visible at the top. The elements of SCADA system are located in separate boxes and connected to each other via a standard communication system.

Thus, the results of two independent tests of the DPR conducted by different manufacturers do make possible coming to any conclusion on its real immunity to IDEI. So, who would need these results?

19.5 Analysis of the result of the third independent trial of the same type of DPR The most recent comprehensive work presented at the Relay Protection Conference in Washington (October 2017) [8], published under the auspices of IEEE, shows the

324 | 19 Features of testing digital protective relays resilience to HEMP results of a recent research conducted by SEL. The paper was expected to deal with lapses and mistakes of the previous paper and to provide correct and true data of DPR resilience to HEMP. Unfortunately, the expectations fell short. The pretentious 18-page paper written by one of SEL’s engineers (who has no previously HEMP-related publications) and co-authored by the Director of Quality, Director of Government Affairs, and Vice-President of R&D of the company, contains in our opinion, even more lapses than the previous article. Since the criticism should always be substantiated and supported by specific arguments and evidence, these arguments are made: 1. The authors do not distinguish between electromagnetic pulses of lightning (LEMP) and the high-altitude electromagnetic pulses of nuclear explosion (HEMP), and thus they transfer their experience on lightning protection gained in the electric power industry to protection from HEMP. In reality, these are completely different physical phenomena that have different specifications, impacting the equipment in different ways. For instance, the lightning EMP represents the localized impact, while HEMP covers a vast area. When lightning is an electric air disruption between two electrodes (a charged cloud with high potential relative to the earth and the earthing system with zero potential), there are no electrodes between which high voltage would have been applied, and air breakdown would have happened during HEMP. Moreover, the earth (grounding system) does not represent a zero-potential area for HEMP. Furthermore, HEMP features both vertical and horizontal components, whereas part of the electromagnetic energy falling on the earth is reflected from its surface. The parameters of a nuclear explosion’s pulse are significantly different from lightning’s EMP parameters, so some widely-used protecting elements (such as gas discharge tubes—GDT widely used for electronic equipment protection) are not acceptable as HEMP protection due to their long response time. Finally, it is worthy to cite an unambiguous statement, mentioned in Section 3 of IEC 61000-6-6 standard: “Lightning protection will not assure protection against HEMP… Lightning protection will not provide adequate HEMP protection…” 2. LC-filters recommended in the discussed paper as an efficient HEMP-protection measure are not efficient in practice because most of them (with some rare exceptions) are intended to divert the energy of high-voltage pulse (applied to the filter’s input) to the grounding system (i. e., to the zero-potential area), which actually does not represent zero potential for HEMP [9, 12]. It should also be noted that, in case of HEMP impact, the earthing system will become a huge antenna absorbing the energy from a large area and delivering it directly onto the internal sensitive components of DPR [10, 11]. Thus, recommendations given in the article about common methods employed for proper DPR grounding (approved for lightning protection or high voltages produced by short circuit currents) are not acceptable for HEMP protection. 3. The article continually claims that the reinforced concrete building, in which DPRs are usually located, can weaken the HEMP’s electric field by 20 dB (i. e.,

19.5 Analysis of the result of the third independent trial of the same type of DPR

| 325

ten times), and thus further on the electric field is taken as five kV/m instead of the initial value of 50 kV/m. Indeed, IEC 61000-2-11 [14] and IEC 61000-5-3 [15] standards really suggest that the concrete building with rebar can weaken the electromagnetic emission by 20 dB in the range from 100 kHz to 30 MHz. They also conclude that it can also weaken the HEMP’s electric field up to five kV/m. However, the problem is that these standards do not provide any definition of a “concrete building with rebar”. It is noteworthy that the shielding efficiency of reinforced concrete from electromagnetic emission is largely dependent on the mesh size of the internal steel net, the number of net layers, the thickness of reinforced concrete and the availability of windows and doors in the building, plus their size and number, etc. Another problem is related to the frequency range for which this 20 dB decrease has been determined and which is too far from that of HEMP. Thus, the field intensity of five kV/m can be assumed on certain conditions and very tentatively, and it does not allow to draw the far reaching conclusions about DPR’s resistance to HEMP. 4. The authors of the above mentioned article make a strange transition from the value of field intensity (five kV/m) to the voltage applied to DPR inputs (5 kV). But five kV/m and five kV are absolutely different values. For instance, theoretically the voltage at the ends of a ten-meter cable can reach as high as 50 kV at five kV/m field intensity. So, where did they take five kV from? 5. The article provides confusing information regarding testing of DPR’s resilience to so-called electrical fast transients (EFT). First of all, there are three tables in the article, which provide different values of EFT amplitudes (in kilovolts) for three different conditions, which have not been determined in the article, Table 19.1. What does this unusual distinguishing between conditions for each of the tables mentioned in [8] mean and what about the phrase: “The electrical transient immunity levels for DPR HEMP resiliency in legacy substation designs are higher than DPRs in new substation designs” (compare Table XIV and Table XV respectively). Table 19.1: EFT amplitude values for different DPR circuits and different conditions (according to [8]). DPR circuits to which test pulse voltage (EFT) is applied Signal (PTs, CTs, I/O) Signal (serial comm.) Power (DC) Power (AC) Telecommunications Antenna

Table XIII [8] DPR

Table XIV [8] DPR on new substation

Table XV [8] DPR on legacy substation

±4 ±2 ±4 ±4 ±2 ±2

±2 ±2 ±2 ±2 ±1 ±1

±8 – ±4–16 – – –

326 | 19 Features of testing digital protective relays resilience to HEMP What is the difference between a “legacy substation” and a “new substation”? Why should a new substation be less resistant than the legacy substation? Unfortunately, the reader will not find any answer to these questions in the article. It would be fair to note that some standards also provide such puzzling information regarding testing. For instance, ITU-T K.78 [16] suggests the same value of field strength for internal space of a reinforced concrete building, i. e., five kV/m (with a reference to the IEC standards previously mentioned), while Table 6 in [16] provides amplitude values of test EFT pulse—one kV for “signal ports” and Table 8 in [16]—eight kV for “signal ports (telecommunication)”. Additionally, both cases deal with telecommunication equipment (see the standard’s name) and deal with ten-meter long cables connected to its ports installed inside of the building. Moreover, clause 3.1.12 of this standard takes “telecommunication port” as an example of “signal port”. So, where does the eight-fold difference between test pulse amplitudes (intended for testing of the same equipment under the same conditions) come from? It is difficult to follow the logic of the authors of this standard.

Figure 19.7: Shape and time parameters of ESD test pulse according to IEC 61000-4-2 [17].

Secondly, the note under Table XV recommends to test resistance to electrostatic discharge (ESD) instead of EFT, and further in the text, the authors provide data regarding DPR’s testing for ESD resistance without even mentioning the EFT. But these are absolutely different treatments (Figures 19.7 and 19.8) that should be realized under different procedures and which have an absolutely different impact on DPR! The eight-kV electrostatic discharge applied to a surface of grounded metal casing of DPR is not equivalent to an eight-kV pulse voltage applied directly to DPR’s terminals!

19.5 Analysis of the result of the third independent trial of the same type of DPR

| 327

Figure 19.8: Shape and time parameters of EFT test pulse according to IEC 61000-4-4 [18].

6.

The authors’ statement that the shields of control cables need to be grounded on both sides to ensure higher shielding efficiency is not confirmed in practice. The reasons for that are covered in [19] in full detail. 7. For some unclear reasons, the authors start using the IEC 60255 standards group in the second half of the article to evaluate the DPR’s resilience to HEMP. However, these standards suggest the requirements for the protection of the relays’ resilience from ordinary electromagnetic interference, which has nothing to do with HEMP. 8. The authors easily discuss such a “simple” problem as the protection of a telecommunication system (which ensures communication in relay protection) from HEMP. They have no knowledge about the comprehensive aggregate of complex

328 | 19 Features of testing digital protective relays resilience to HEMP problems [20, 21] that persist in protecting communication systems (which are very susceptible to HEMP) and many other problems. 9. The article provides a rather strange representation of testing results of SEL’s protection relays. For example, the article says that, due to the 20-dB weakening achieved by the typical substation building, the field intensity inside this building will be five kV/m. Further on, the article says that DPRs were tested in a certified laboratory that mimics the environment of a typical substation, whereas in the next sentence, we read that based on those tests the SEL’s relays (obviously, they mean all types of SEL’s relays?) could resist a 50-kV/m pulse, and they refer to some inaccessible internal unpublished corporate document. What should be the reader’s conclusion based on this description? 10. The article refers to another internal unpublished document and claims that the test of an individual printed circuit board of the DPR “of the same SEL’s relay”, without casing and any metal shielding covers and exposed to an electric field pulse of 25 kV/m, confirmed resistance of SEL’s relays to HEMP. It should be noted that resilience of a passive set of electronic components (installed on an individual printed circuit board without any connection) to a pulse of an external electric field with the intensity of 25 kV/m does not prove anything to anybody. On the other hand, according to data presented at the symposium [22] organized by serious American organizations, such as North American Electric Reliability Corporation (NERC), Department of Energy (DOE), Department of Defense (DOD), Department of Homeland Security (DHS), Federal Energy Regulatory Commission (FERC), etc., the following facts and figures were repeatedly mentioned: – programmable controllers, PCs and communication ports of equipment can be damaged by applied voltage of 0.5–0.6 kV; – failures and operational disorders of DPR occur at 3.2–3.3 kV; – voltage occurring on the cable ends running inside typical buildings and connected to electronic equipment of power systems can reach as high as ten kV upon HEMP impact. Conclusions of [22]: – “The direct coupling of E1 HEMP fields inside the building is strongly influenced by construction type of the building”. – “Given that ordinary building protection level will typically allow up to 10 kV to be coupled to internal cables… leading to the electronics…, this indicated a potential problem”. – “…upsets on relays begin at 3.2 kV and damage to programmable logic controllers and personal computers begin at approximately 0.5 kV, indicates a serious concern for the continued reliable operation of substations.”

19.5 Analysis of the result of the third independent trial of the same type of DPR

| 329

How can we determine the actual level of DPR resistance to HEMP? Obviously, valid results can be obtained during real tests on HEMP simulators due to a broad range of DPR’s parameters and differences in the environments of their installations, as well as inaccuracies in standards and reports. Recently, SEL Company has published a new and an even more sensational related article. Analysis of this new presentation shows that all its revelations clearly point to the source of all of the just cited lunacies promoted by this company. Just take the titles alone, see Figure 19.9! Then, even more startling statements follow from these sensational revelations, see Figure 19.10.

Figure 19.9: This “magnum opus” and some of its sensational headlines are shown.

Figure 19.10: Conclusions made by SEL based on its “magnum opus” research.

330 | 19 Features of testing digital protective relays resilience to HEMP It is very interesting to know what breakthrough research or thrilling discovery made SEL arrive at the conclusion disproving all the previous findings of thousands of specialists over the last 50 years. As for the facts, the presentation contains the same allegations and misinterpretations as the previous SEL’s publications, such as the false thesis saying that lightning protection automatically ensures the HEMP sustainability. For some reason, the authors think that several lightning and HEMP parameters presented as a table, see Figure 19.11, substantiate their statements.

Figure 19.11: According to SEL engineers, lightning EMP protection automatically ensures the HEMP E1 sustainability.

It is equally interesting to understand what kind of “statistical” data on an amplitude of induced-HEMP voltage at the cables connected to the microprocessor-based terminals (allegedly below four kV) SEL is talking about, see Figure 19.12. Do they really have they own nuclear test site?

Figure 19.12: Against all available evidence, reports and standards, SEL’s HEMP specialists allege that the amplitude of the induced voltage at the cables connected to the relay terminal inputs does not exceed four kV.

19.6 Conclusions | 331

So, is there any way to determine the actual DPR immunity to HEMP? Clearly, considering the DPR parameter spread, variation in operating conditions and standard and report inaccuracies, the representative data can be taken only from the real-world tests performed using the HEMP simulators. Moreover, such tests must be carried out on the complete relay cabinets connected to the relevant communication, earth and auxiliary supply systems (see previous chapters and also [23] and [24]), rather than on the standalone terminals and the single printed boards.

19.6 Conclusions 1.

Due to methodology errors during the DPR tests conducted by independent organizations earlier, they cannot be considered as satisfactory or their results as meaningful. At the moment, there are no reliable data on the level of DPR immunity to HEMP, which suggests that the test should be conducted further. 2. The materials published by the American company SEL do not prove the advantages of the DPR of this company in comparison with the equipment of other companies and do not prove the stability of the microprocessor protection relays to HEMP, but are purely advertising materials containing myriad technical blunders and inaccuracies, capable of misleading readers. Reliable data on the stability of the relay protection system to HEMP can be obtained only on the basis of wellposed tests that require preliminary, rather complex and responsible work 3. The kinds and modes of DPR tests should be fully performed and correspond to the standards as described in [1]. 4. The performance criterion (PC) should be represented by a criterion that makes possible controlling DPR operation under normal and emergency modes of the object under protection when it is affected by an electromagnetic interference instead of a criterion that is based exclusively on checking the DPR condition after the test (when the impact of interference has ceased). 5. Testing should be performed on both the separate unit of DPR and the full relay protection system consisting of several DPR units connected to each other by several meters of cables via a corresponding communication device. At the same time, electromagnetic energy should affect the relay protection system, while impulse tests for applied voltage should be performed on both separate DPR units/communication devices and several DPR units connected together with communication devices simultaneously. 6. During the test, several steps of test-impulse amplitude and electric-field strength should be selected from a minimum to a maximum value within the ranges described in the standards. The obtained data can be used during the evaluation of immunity of the DPR installed in specific cabinets and buildings that possess a certain index of electric-field weakening. They can also be used in the process of

332 | 19 Features of testing digital protective relays resilience to HEMP elaboration of requirements to further weaken this field if it is revealed that the current conditions do not ensure the required immunity of DPR to IDEI.

Bibliography [1] [2]

[3] [4] [5] [6] [7] [8]

[9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21]

Gurevich V. Problems in Testing Digital Protective Relays for Immunity to Intentional Destructive Electromagnetic Impacts. Glob. J. Adv. Res., 2014, Vol. 1, No. 2, pp. 159–173. Savage E., Gilbert J., Radasky W. The Early-Time (E1) High-Altitude Electromagnetic Pulse (HEMP) and Its Impact on the U. S. Power Grid, Report Meta-R-320 for Oak Ridge National Laboratory, 2010. Gurevich V. Problems of Standardization in Relay Protection, St. Petersburg, DEAN, 2015, 168 p. Gurevich V. Vulnerabilities of Digital Protective relays. Problems and Solutions, Infra-Engineering, Moscow, 2014, 256 p. Perrow C. Normal Accidents. Living with High Risk Technologies, First ed. Princeton University Press, Princeton, 1984. EMP Effects on Protection and Control Systems, Schweitzer Engineering Laboratories, 2014, 31 p. Report of the Commission to Assess the Threat to the United States from Electromagnetic Pulse (EMP) Attack, April 2008. Minteer T., Mooney T., Artz S., Whitehead D. E. Understanding Design, Installation and Testing Methods That Promote Substation IED Resiliency for High-Altitude Electromagnetic Pulse Events. 44th Annual Western Protective Relay Conference, Washington, October 17–19, IEEE. Gurevich V. Use of LC-Filters to Protect Equipment from Electromagnetic Pulse: Is It Real Necessity or “Business as Usual”? Int. J. Sci. Res. Eng. Trends, 2017, Vol. 3, No. 4, pp. 85–89. Gurevich V. Functional Grounding of Digital Protective Relays. Transm. Distrib. (Australia), 2016, Vol. 12, pp. 32–35. Gurevich V. The Issues of Electronic Equipment Grounding at the Power Facilities. Int. J. Res. Stud. Electr. Electron. Eng., 2017, Vol. 3, No. 1, pp. 11–19. Gurevich V. The Problem of Grounding of Electrical Equipment as Main Protective Means against HEMP. Pro Elektrichestvo, 2017, Vol. 2. MIL-STD-188-125-1 High-Altitude Electromagnetic Pulse (HEMP) Protection for Ground Based C41 Facilities Performing Critical. Time-Urgent Mission. Part 1 Fixed Facilities, 2005. IEC 61000-2-11 Electromagnetic Compatibility (EMC)—Part 2-11: Environment—Classification of HEMP Environments. IEC/TR 61000-5-3 Electromagnetic Compatibility (EMC)—Part 5–3: Installation and Mitigation Guidelines—HEMP Protection Concepts. ITU-T K.78: High-Altitude Electromagnetic Pulse Immunity Guide for Telecommunication Centers. International Telecommunication Union, 2009. IEC 61000-4-2 Electromagnetic Compatibility (EMC)—Part 4-2: Testing and Measurement Techniques – Electrostatic Discharge Immunity Test. IEC 61000-4-4 Electromagnetic Compatibility (EMC)—Part 4: Testing and Measurement Techniques – Section 4: Electrical fast Transient/Burst Immunity Test. Gurevich V. Grounding of Control Cable Shields: Do We Have a Solution? Compusoft, 2017, Vol. VI, No. 5, pp. 2330–2334. Gurevich V. Protection of Telecommunication Systems in Electric Power Facilities from Electromagnetic Pulse (EMP). Part 1. Innov. Eng. Phys. Sci., 2017, Vol. 1, No. 1, pp. 7–12. Gurevich V. Protection of Telecommunication Systems in Electric Power Facilities from Electromagnetic Pulse (EMP). Part 2. Innov. Eng. Phys. Sci., 2017, Vol. 1, No. 1, pp. 13–18.

Bibliography | 333

[22] High-Impact, Low-Frequency Event Risk to the North American Bulk Power System. A Jointly-Commissioned Summary Report of the North American Electric Reliability Corporation and the U. S. Department of Energy’s, November 2009 Workshop, June 2010. [23] Gurevich V. Features of High-Altitude Electromagnetic Pulse (HEMP) Test Methods for Power System Electronics. Int. J. Res. Sci. Innov., 2017, Vol. III, No. IX, pp. 5–13. [24] Gurevich V. Devices for Testing Effectiveness of Electronic Equipment Protection against Electromagnetic Pulse (EMP). Int. J. Res. Stud. Electr. Electron. Eng., 2017, Vol. 3, No. 3, pp. 6–13.

20 Establishment of inventory of electronic equipment’s replacement modules as a way to improve survivability of the power system 20.1 Optimization of inventory of electronic equipment replacement modules One of the efficient ways to improve the survivability of the power system is to ensure quick restoration of damaged equipment by using spare parts, tools and accessories (SPTA). However, establishment of SPTA inventories requires significant investments, especially in case of extremely complex electronic digital systems for the protection, automation and management that are widely used in power systems [1]. That’s why all over the world people are looking for optimum inventory of SPTA that will make possible combining the required reliability at a minimum investment. Establishment of these inventories of SPTA is a common problem that is well known in numerous areas of engineering. As of today, the problem has been well studied theoretically, using various mathematical methods of optimization [2–9]. The known methods of optimization of SPTA inventories are based on statistical analysis of failures of elements, replacement modules and complex items. In other words, the quantity of necessary SPTA kits is calculated based on the fact that failures of electronic equipment are discrete accidental events that happen with a certain frequency governed by statistical laws of random variables distribution. The necessity of increasing the number of SPTA kits to ensure restoration of the functioning of equipment affected by HEMP is practically assured. But how should we increase this number? Obviously, HEMP impact on the energy system will result in simultaneous mass failures of electronic equipment that do not comply with any statistical law. Besides, the normal and rather lengthy process of ordering and obtaining a new SPTA kit to replenish the inventory after using up earlier prepared kits is not suitable under this situation. So, expansion of SPTA inventory by 1.5–two times (this is what short-sighted managers sometimes do) does not solve the problem, while expansion of inventory, which supposes that all types of electronic equipment will be backed by SPTA, is unrealistic due to economic reasons. That’s why, in order to calculate the optimum SPTA kit, completely different approach should be employed. The suggested method is based on three main principles: 1. Not all electronic devices should be equipped with SPTA. These should be only those that are considered critically important devices (CID), the lack of which makes impossible even partial operation of electric energy facilities, where only critically important objects (CIO) should be selected. 2. Full and not partial SPTA kits should be established for CID. https://doi.org/10.1515/9783110639285-020

20.2 The problem of the traditional mode of SPTA storage

3.

| 335

Inventories of SPTA kits for CID should be supplemented, regardless of the availability of SPTA kits that are already stored in warehouses.

Thus, optimization of SPTA inventories in this case is based on calculation of the number of CID necessary to supplement CIO in a specific energy system.

20.2 The problem of the traditional mode of SPTA storage The problem of SPTA storage requires addressing of two problems: where the SPTA should be stored and how. Today, SPTA kits of many energy systems are stored in warehouses, which are often located at a distance from energy facilities. When necessary, the kits are obtained by repair departments or directly by the operating team. Should urgent restoration of CID be necessary after HEMP impact, the problem of urgent delivery of critically important goods to critically important objects occurs because there is a probability that the impact of a powerful electromagnetic pulse will result in failure of the microcontrollers that manage the operation of modern transport vehicles. According to [3], electric communication systems use two levels of SPTA, such as “SPTA-0” and “SPTA-G” kits. An SPTA-0 kit is an integral part of a device (CID in this case) and should be stored on site (CIO in our case). SPTA-G kits (group kits) are used to replenish SPTA-0 kits, and they are stored in a large service center (or in a warehouse). Both SPTA-0 and SPTA-G should be checked and tested before placed in storage. This approach to SPTA storage for the necessary restoration of the energy system after HEMP impact is fully justified by many reasons. It is also efficient in the electric energy industry because it obviates the problem of urgent delivery of SPTA to CIO for CID restoration. Storing CID’s SPTA kits on site does not require addressing this problem. When organizing storage of CID’s SPTA kits in this way, another problem can be solved, i. e., the problem of their configuring and setup prior to installation into the equipment. This requires significant time investments and participation of highly skilled staff, as well as employment of special electronic equipment and computers (which can also be damaged). An example of this type of CID is represented by modern digital protective relays (DPR), the lack of which makes it impossible for a power system to operate. In the case of mass DPR failure as a result of HEMP impact, it will be extremely difficult to ensure simultaneous setup of dozens of DPR’s SPTA kits at many remotely located sites. That’s why the SPTA for DPR considered as CID should not only be stored near the operating DPR, but also should be preprogrammed, set up and configured for quick replacement of failed blocks of specific DPR that works with certain settings and adjustments. The second question that needs to be addressed is: How should we store CID’s SPTA kits? The problem is that HEMP creates the electric-field density of up to 50 kV/m at the ground surface. This density can create a difference of potentials at the outputs

336 | 20 Establishment of inventory of electronic equipment’s replacement modules of relatively small electronic components (within one printed circuit board). This difference of potentials is enough for electric disruption of p-n junctions (the thinnest layers of insulation in microprocessors) or erasure of information stored in memory cells. This means that CID’s SPTA kits should be stored in HEMP-resistant containers.

20.3 Requirements for protective containers What should be the features of these containers? Let’s consult the MIL-STD-188-125-1 standard [10], which stipulates the requirements to shielding performance of critically important objects from HEMP impact, Table 20.1. Table 20.1: Minimum requirements for the shielding performance of critically important objects from HEMP (according to Figure 1 from MIL-STD-188-125-1 [10]). Frequency

Shielding performance, dB

10 kHz 100 kHz 1 MHz 10 MHz 1 GHz

20 40 60 80 80

At the same time, [11] provides data about spectral density of radiation of various types of electromagnetic interferences, see Figure 20.1.

Figure 20.1: Spectral density of electromagnetic emission from various sources.

These data show that the density of radiation of E1 and E2 (the standard denotes E2 component as “lightning” because its parameters are close to those of lightning) components of HEMP remains the highest at frequencies lower than ten kHz and drops

20.4 Protective containers available on the market | 337

down dramatically at frequencies higher than 300 MHz. Other sources of IDEI (not HEMP) create a relatively high density of radiation in the range of higher frequencies up to ten GHz. That’s why efficient protection should be ensured in the frequency range from several kHz to ten GHz. It is known that the skin depth of metals is determined by skin effects and depends on the frequency: the higher the frequency (f ), the shallower the skin depth (Δ), i. e., the thinner the screen wall can be: Δ = 503 ⋅ √

ρ μ⋅f

where: Δ is the skin depth, ρ = electrical resistance of metal, μm = magnetic permeability of metal, f = emission frequency. The “skin depth” is defined as the surface layer of metal where the density of electromagnetic field is reduced in e = 2.718 times. According to [12], this layer will account for almost 86 % of energy coming from the surface. Table 20.2 shows the results of calculations based on the just-mentioned formula for the most widely used metal shield, i. e., aluminum. Table 20.2: The skin depth of an aluminum wall shield for various frequencies. Frequency skin depth, mm

1 kHz

10 kHz

100 kHz

1 MHz

10 MHz

100 MHz

1 GHz

2.6

0.83

0.26

0.083

0.026

0.0083

0.0026

It is obvious from the table that the aluminum container with a wall thickness of not less than three mm can ensure rather efficient weakening of radiation of all HEMP types.

20.4 Protective containers available on the market What does the market of protective containers offer today? First of all, the market offers large and heavy thick-walled metal containers, see Figure 20.2, equipped with protected ventilation systems and filters for input cables. These containers are widely used in the army and obviously ensure reliable protection of the equipment located inside. Unfortunately, these are very expensive means of protection and are unlikely to be used for SPTA storage in the electric energy industry. Another variety of protective container is represented by a room without windows the walls and doors and which are covered with copper sheets (these rooms are offered by Holland Shielding Systems). These protective containers also ensure perfect shielding (from 40 to 120 dB in the frequency range from ten kHz to ten GHz). However, they are also as expensive as those previously mentioned.

338 | 20 Establishment of inventory of electronic equipment’s replacement modules

Figure 20.2: Large metal containers protecting from HEMP and equipped with ventilation systems and filters to connect input cables.

Figure 20.3: Faraday bags with a metallized layer intended for protection of small electronic devices from HEMP.

Simple, reliable and very inexpensive containers for HEMP protection are represented by (according to their manufacturers) Faraday bags with a metallized layer of various sizes, see Figure 20.3. As a rule, manufacturers of these bags claim a high level of radiation weakening reaching as high as 40–45 dB. However, they conveniently conceal the frequency range over which these measurements were obtained. Can a metallized layer with a thickness of several microns efficiently weaken electromagnetic fields in the frequency range from several kHz to several Ghz? Table 20.2 provides a decisive answer to this question: No, it cannot! Another variety of protective container that is also widely offered in the market and promoted as a reliable means of protection from HEMP is exemplified by a tent,

20.4 Protective containers available on the market | 339

produced from the same (as Faraday bags) metallized plastic or at best woven from fiber containing metal yarn, Figure 20.4. Special portable metal containers with thick walls that ensure very efficient shielding are also widely available on the market, see Figure 20.5.

Figure 20.4: Protective tent produced from metallized fiber.

Figure 20.5: Protective container of Holland Shielding Systems BV, which possesses very high shielding performance.

340 | 20 Establishment of inventory of electronic equipment’s replacement modules Unfortunately, these containers are very expensive for SPTA storage, and their internal chambers are not large enough. The most suitable containers are those made of aluminum and welded from aluminum sheets in the shape of simple boxes with covers, see Figure 20.6. These containers with a wall thickness of 3/16 inch (4.8 mm) ensure a rather acceptable level of shielding: not less than 50 dB in the frequency range from 100 kHz to one GHz (76 dB at 300 MHz; 66 dB at one GHz). They are produced by several companies, including Montie Gear, EMP Engineering, etc., in standard or customized sizes.

Figure 20.6: Inexpensive protective containers for SPTA storage made of sheet aluminum.

It should be noted that these simple containers of sufficient sizes can be produced at any workshop that has welding equipment. At the same time, in order to prevent the impact of electromagnetic field of the upper part of frequency range on stored electronic devices (which can penetrate into the inner cavity of the container through vents created by poorly attached cover), it is recommended to locate extra-sensitive electronic devices (such as printed circuit boards with microprocessors and memory cells) into the metallized Faraday bags before putting them into the container.

20.5 Conclusion One of the ways to quickly restore an electric-energy system’ functionality after any type of HEMP impact is to create special SPTA kits for electronic equipment. The known optimization methods of SPTA inventories are not acceptable for the situation being discussed. In order to ensure quick restoration of an energy system’s electronic equipment, it is necessary to maintain complete kits of SPTA for critically important devices (CID) located at critically important objects (CIO) of the electric energy industry. Both CID and CIO should be determined in advance. The SPTA kits for CID should be independent of the total inventory of SPTA stored in warehouses. The CID’s SPTA kits should be preliminarily checked, set-up and con-

Bibliography | 341

figured; also, they should be stored in close proximity to the CID to which they are related. The CID’s SPTA should be stored in closed containers protecting from HEMP and other IDEI types. These containers can be produced by welding aluminum sheets with a thickness of about five mm. Extra-sensitive blocks, containing microprocessors and memory cells, should be previously put into metallized Faraday bags.

Bibliography [1]

Gurevich V. Cyber and Electromagnetic Threats in Modern Relay Protection, CRC Press (Taylor & Francis Group), Boca Raton–New York–London, 2014, 222 p. [2] Zhdanov V. Automation of Stock Estabishment in SPTA Kits: Methods and Techniques. Compon. Technol., 2010, Vol. 5, pp. 173–176. [3] Industry Standard 45.66-96 “Spare Parts, Tools and Accessories of Electrical Communication Means. Industry Standard. M.: Center of Scientific and Technological Information “Informsyaz”, 1997. [4] Zatsarinny A. A., Garanin A. I., Kozlov S. V. et al. Calculation of SPTA Kits in Protected Automated Information Systems. Syst. Means Inform., 2013, Vol. 23, No. 1, pp. 113–131. [5] Dopira R. V., Lysiuk A. P., Tsybenko D. V. et al. Method of Calculation of Spare Parts Supply System for Wide-Area Radio-Electronic Facilities. Soft. Prod. Syst., 2009, Vol. 1, pp. 128–130. [6] State Standard RV 20.39.303-98. A Comprehensive System of General Requirements. Apparatus, Instruments, Devices, and Equipment Designed for Military Application. Requirements to Reliability. Composition and Order of Task. M. IPK, Izdatelstvo Standartov, 1998. [7] Trimp M. E., Dekker R., Teunter R. H. Optimize Initial Spare Parts Inventories: An Analysis and Improvement of an Electronic Decision Tool. Report Econometric Institute E1 2004-52, Erasmus University Rotterdam, 2004, 70 p. [8] MIL-STD-1388-2. Department of Defense Requirements for a Logistic Support Analysis Record, 1993. [9] Love R. E., Stebbins B. F. An Analysis of Spare Parts Forecasting Methods Utilized in the United States Marine Corps. Thesis AD-A184 698, Naval Postgraduate School, Department of the Navy, 1987. [10] MIL-STD-188-125-1 High-Altitude Electromagnetic Pulse (HEMP) Protection for Ground-Based C4I Facilities Performing Critical, Time-Urgent Missions; Part 1: Fixed Facilities. [11] IEC 61000-2-13: 2005 Electromagnetic Compatibility (EMC)—Part 2-13: Environment. High-Power Electromagnetic (HPEM) Environments—Radiated and Conducted. [12] Nekrasova N. M., Katsevich L. S., Yevtyukova I. P. Industrial Electroheat Installations. M.: Gosenergoizdat, 1961, 415 p.

21 The problem of impact of geomagnetically induced currents on power transformers and it solution 21.1 Geomagnetically induced currents generated by solar storms The Earth’s ionosphere, located a couple hundred kilometers above the Earth’s surface, always contains the flowing electric currents generated by the Earth’s magnetic field and its axis-turning. They are maintained by the continuous generation of a large number of charged participles—ions and free electrons released from the molecules of the atmospheric gases split by solar radiation. Such currents have a great impact on the generation of Earth’s magnetic field. During solar storms, followed by the ejection of an enormous amount of ionized plasma towards the Earth (see Figure 21.1), the super high-energy streams of protons and electrons contained in the solar plasma provoke the sharp increase in strength of the electrical currents flowing in the ionosphere.

Figure 21.1: Distortion of the Earth’s magnetic field from the ejection of solar plasma.

Such a sharp current increase results both in a dramatic change in the Earth’s magnetic field and in the generation of the so-called geomagnetically induced currents affecting buried lengthy metal items, such as piping, railways and cables. Also, it induces strong currents in long overhead power lines. The currents induced in overhead power lines cause short circuits through the grounded neutrals of the power transformers, see Figure 21.2. Due to their very low frequency, such currents flowing through the transformer windings saturate the transformers’ magnetic cores and result in a sharp decrease in the transformers’ impedance. Since it is commonly known that the DC component of the current generhttps://doi.org/10.1515/9783110639285-021

21.1 Geomagnetically induced currents generated by solar storms | 343

Figure 21.2: Diagram of the inductance of ionospheric current in electric overhead power lines and ground.

ated by the power transformer also exists at its initiation (inrush current), the power transformer protection relays are usually tuned out from the DC component, so that one can ignore it. Moreover, DC current (or the current with a very low frequency— quasi DC) is normally not transferred through the transformer current. Consequently, the transformer should burn out since the traditional relay protection does not respond to the induced currents saturating the transformer. The problem of damage to power transformers by Geomagnetic Induced Currents (GIC) during solar storms is widely known [1, 2]. Quasi-direct currents of GIC flowing through grounded neutral terminals of power transformers located in the northern regions can reach up to 100–300 A. These currents result in deep saturation of magnetic cores, reduction of impedance and overheating of windings and magnetic cores, which can even lead to winding blow-out. The collapse of the Hydro-Québec (Canada) power grid is well known. This resulted in a situation, where six million people had no electric power for nine hours. Another case was the blowout of a powerful transformer of Public Service Electric and Gas Company in New Jersey (Northeastern US) during a solar storm in March of 1989, Figure 21.3 [3]. In 2005 and 2012, GIC (with an amplitude of up to 200 A) were registered in neutral terminals of power transformers in Finland, while in Sweden the amplitude of GIC reached as high as 300 A in 2000. The GIC impact on transformers in August 2003 resulted in a collapse of the power grid in the northeastern US and the bordering part of Canada. Until recently, all the registered cases of damaged supply transformers happened in the circumpolar regions of northern latitudes, Figure 21.4. There are also identified zones in the Southern Hemisphere [4], where GIC intensity is much lower than that in Northern Hemisphere, Figure 21.5. However, they are much more intensive than those in other, undistinguished areas.

344 | 21 The problem of impact of geomagnetically induced currents on power transformers

Figure 21.3: The blown-out power transformer of Public Service Electric and Gas Company in New Jersey during a solar storm in March of 1989.

Figure 21.4: Zones of intensive GIC in Northern Hemisphere. Left—according to Goddard Space Flight Center NASA (USA); right—according to Istituto Nazionale di Geofisica e Vulcanologia (Italy).

In other words, it was thought until recently that only regions close to the poles can be susceptible to significant impact of geomagnetic induced currents. However, in 2007 a sensational article of two authors from Cape Town University (South Africa) was published under the title: “Transformer failures in regions incorrectly considered to have low GIC risk” [5]. This article attracted the attention of researchers in various countries of the world. It is cited in dozens of articles of other authors; it is even referenced in official reports [6, 7]. Why? Because, if the data provided in the article are correct, this requires revising the existing approach and viewpoint regarding GIC and its influence on supply transformers. There are many other countries

21.1 Geomagnetically induced currents generated by solar storms | 345

Figure 21.5: The zones in Northern and Southern Hemispheres that are more susceptible to GIC impact (according to [4]). Red areas correspond to maximum possible intensity of GIC for the corresponding hemisphere.

on the same latitude with South Africa, where the danger of GIC was not considered previously. Moreover, some authors take the findings of the article and apply them to other regions, including the Middle East. For example, one of the reports issued on request of the Ministry of Energy of Israel states that Israel is located in the zone of dangerous impact of GIC on the power grid because there were registered cases of devastating impact of GIC on power transformers in South Africa. And now, we can find speculation about the GIC danger in Russia. One of four articles published in 2013 in the News of Electric Engineering magazine with a view to be sensational is titled: “Geomagnetic storms. A threat to national security of Russia” [8]. It should be noted that these articles do not provide any registered measurement results that would substantiate the thesis regarding the danger of high GIC values for Russia. What arguments are given by the authors from Cape Town University to substantiate their sensational claim? Let’s review key points of this article step by step in the sequence that they appear in the article. 1. There are references to several other articles, which were published earlier. There they conclude that the main type of damage to old, large transformers in South Africa is constituted by internal insulation damage. The authors put special emphasis on the fact that none of the previously published works mentions GIC as a reason for damage to power transformers in the South African power grid.

346 | 21 The problem of impact of geomagnetically induced currents on power transformers 2.

3.

There are references to known cases of power grid collapse in Canada (HydroQuebec) and the damage of a supply transformer at a nuclear power plant in New Jersey (Northeastern US) as a result of GIC impact (but exactly where in Canada and where in South Africa!). There is a reference to the Ph. D. thesis of one of the authors of the article, Dr. Koen J., where he applied a known method of GIC measurement to the South African power grid and showed that it is in agreement with experimental measurements of GIC in power transformers, see Figure 21.6. The figure shows that the maximum amplitude of actually measured GIC values in a powerful transformer does not exceed 6 A and flows for a very short period of time.

Figure 21.6: GIC currents in a neutral terminal of a 500-MVA power transformer during a solar storm in March 2001 (South Africa).

4. Among the references to articles of other authors, it is stated that there were multiple cases of saturation of three-phase, three-leg cores transformers (usually considered as significantly more resistant to GIC than single-phase or five-leg core three-phase transformers—“shell form” core transformers) with GIC currents reaching up to 2 A. However, there is no indication who obtained these data and where. 5. Then, the section called “Thermal damage by GICs during Nov 2003” includes photographs of transformers with blown-out windings (Figure 21.7), which resemble the damage to the transformer in New Jersey, where GIC currents reached hundreds of amps; also, a high concentration of oil-diluted gases was recorded.

21.1 Geomagnetically induced currents generated by solar storms | 347

Figure 21.7: A photo of damaged windings of power transformers mentioned in the article [5].

6.

It is also mentioned that those gases appeared after several geomagnetic storms, while some transformers experienced emergency shutdown many months later after the registered geomagnetic storms. It is noteworthy that regardless of continuous monitoring of GIC, the article doesn’t mention any actual current values of GIC registered in these specific transformers that could result in such serious damages of the transformers shown in Figure 21.7. The section “Other possible causes of damage” mentions that the damage to transformers due to GIC currents does not happen only during the geomagnetic storms. They can appear even a year after GIC impact as a result of accumulated stress. However, these conclusions are not justified or substantiated.

Thus, it is fair to say that the only experimentally confirmed fact is a case of occurrence of GIC with amplitude of up to 6 A in power transformers of the South African power grid. All other data are provided in a form of speculation and assumptions that are not confirmed by actually registered measurements. We now turn our discussion to how dangerous GIC with amplitude of up to six A are for large power transformers. Let’s check a new IEEE [9] standard, which summarizes the state-of-art experience in the area of GIC impact on power transformers. The standard discusses various aspects that weaken or strengthen the GIC impact, including structural features of transformers, the level of load, etc. So, this standard doesn’t consider or mention GIC currents lower than ten A due to their insignificant influence on transformers. Section 6.5 of this standard directly mentions that temperature increase of windings and other structural elements of supply transformers at ten-A GIC currents is negligible. Let’s review other publications that discuss the reasons for the faults in the South African power grid. Among multiple publications on this topic, it is worth mentioning a profound report [10] that contains analysis of 12,229 (from 1993 until the end of 2009) emergency cases in the South African power grid, i. e., it includes the period of 2003– 2004, which (according to the authors of the previous article) is notoriously known for

348 | 21 The problem of impact of geomagnetically induced currents on power transformers a mass damage of supply transformers due to geomagnetic-storm impact. This analysis is based on data provided to the authors by such authoritative organizations as Advanced Fire Information System, which implements special registering equipment installed on NASA satellites, as well as on data from other organizations. The distribution of reasons of faults is as follows: 38 % of all faults is caused by large birds; 26 %—by lightning; 22 %—due to fire in power equipment. Other reasons include vandalism, unskilled personnel, falling trees, etc. Furthermore, among more than 12,000 cases of faults, the faults caused by geomagnetic storms are not even mentioned. It is interesting that one of the authors of this publication is Professor C.T. Gaunt, who co-authored the sensational article five years ago about the fact that South Africa is located in the dangerous zone of GIC impact on the power grid. What happened with the viewpoints of this author within five years remains under wraps? What do other authors write about the power grid of South Africa? The article [11] provides the results of analysis of faults of 188 supply transformers (88–765 kV, 20–800 MVa) during five years. More than 80 % of the transformers are rated more than 400 MVA. The article concludes that the most frequent causes of faults for this group of transformers are related to internal insulation failures. The article doesn’t even mention damage caused by GIC. The report of Mitigation Action Plans & Scenarios (MAPs), prepared for the Ministry of Energy of South Africa [12], mentions the very poor condition of the electric energy industry in South Africa due to the lack of investment during the past 20 years. After thorough examination of the reasoning given in the mentioned sensational article and familiarization with the results of analysis of transformer damage in South Africa, as performed by various authors, there is a doubt in the creditworthiness of the statement made in the article and suspicion of an attempt to implicate solar storms in transformers damage to excuse real problems in the electric energy industry. What about countries located on similar latitudes as South Africa? The impact of GIC on power transformers in South Australia is thoroughly analyzed in [13] Real GIC current measurements in transformers during the geomagnetic storms return values that do not exceed 4–5 A, i. e., these are very close to the values obtained in South Africa. However, the authors mention that the expected GIC currents during other geomagnetic storms can theoretically be higher. There are no experimental data for transformers in the power grid of Uruguay [14]. Theoretical calculations return results similar to those for South Africa. The GIC currents experimentally registered in Southern Brazil 2009–2013 were very close to current values registered in South Africa, Figure 21.8, [15]. When analyzing the reasons of power transformers failures in India [16], Iran [17] and Pakistan [18], GIC impact is not even mentioned. In Japan, GIC values registered in power transformers during solar storms did not exceed 4 A [19]. In New Zealand, which is rather close to the area of southern maximum of GIC, GIC currents with an amplitude not exceeding six A and sometimes reaching as high

21.1 Geomagnetically induced currents generated by solar storms | 349

Figure 21.8: GIC currents in a neutral terminal of 500-kV transformer during a solar storm in October 2013 (Brazil).

Figure 21.9: GIC in the neutral terminal of power transformer during a solar storm in New Zealand in November 2001 [20].

as 22 A (Figure 21.9) were registered in several transformers during the strong solar storm in November 2001. It should be noted that individual emissions of GIC with an amplitude exceeding ten A are very short in duration, lasting 20 sec. according to [20], while the thermal time constant of power transformers’ heating according to the standard [9] amounts to 30–45 minutes, i. e., the temperature will not have enough time to rise significantly during such emissions. Even long-lasting GIC ranging 20–30 A (according to the same standard) are not enough to damage supply transformers.

350 | 21 The problem of impact of geomagnetically induced currents on power transformers Thus, the geomagnetically induced currents occurring during solar storms in all the cited regions located in the Southern Hemisphere do not reach values capable of damaging supply transformers. Alternatively, when passing through power transformers, these currents convert the transformers into powerful sources of harmonics, interfering with other types of electric equipment, primarily protection relays, in power grids. This interference explains faulty actuation of relay protection and tripping of transformers. However, this is another problem that can also be solved (e. g., through installation of additional filters in the relay protection circuits) and which has nothing to do with real damage to supply transformers. Is there any way to determine the values of geomagnetically induced currents attributable to certain regions and particular transformers? Actually, there are two methods: calculation using the special program module and experimental method requiring measurement of the geomagnetically induced currents within the power transformers’ neutral ground circuits. Naturally, these methods can be combined. A special program module was developed for the calculation of the geomagnetically induced currents PSS® E produced by Siemens [21] (commercially available). This program uses the data of the substation and the geographic location of the electric power lines, ground-system resistance, soil properties, resistance and group of transformer connections and the transformer magnetic-system structure as the design conditions for the calculation of the geomagnetically induced currents. Dynamic Ratings offers a simple current sensor type GIC-4 [22] enabling the direct measurement of the geomagnetically induced current values. It is designed like the current transformer equipped with the split core placed on the conductor connecting the transformer neutral to the ground system, see Figure 21.10.

Figure 21.10: A simple sensor of geomagnetically induced currents type GIC-4 produced by Dynamic Ratings.

21.1 Geomagnetically induced currents generated by solar storms | 351

This Hall element sensor contains the special filter blocking out the currents above three Hz. The sensors operate within the two ranges of maximum input geomagnetically induced currents: up to 45 A and up to 360 A. The sensors generate the standard output four–20 mA signal, acceptable for any recording and control system and ready for transmission to any remote control room. Additionally, Advanced Power Technologies offers the more sophisticated device ECLIPSE HECT [23], designed to measure the DC component of the transformer’s neutral current and the transformer’s current harmonics that appear upon the saturation of the transformer’s core with the geomagnetically induced currents, see Figure 21.11. The device contains the geomagnetically induced current Hall sensor connected to the power transformer’s neutral, the small measurement module, and two compact current transformers. Those additional current transformers should be connected to the secondary circuits of the standard current transformers, located in the bushings (walltube insulators) of the power transformers in order to analyze the current harmonics (up to seven harmonics). The measurement module may be used to convert the input signals to standard four–20-mA signals or to issue an alarm signal if the geomagnetically induced current or harmonic exceeds the set threshold.

Figure 21.11: ECLIPSE HECT—the device designed for the control of the geomagnetically induced currents (Advanced Power Technologies). The device measures the currents flowing in the power transformer neutral and the current harmonics.

352 | 21 The problem of impact of geomagnetically induced currents on power transformers

21.2 Geomagnetically induced currents generated by HEMP The HEMP magnetohydrodynamic effect (HEMP MHD) is one of the nuclear detonation HEMP components, referred to as E3. It is based on MHD effects of nuclear detonation products plasma interaction with heated ionized air and the Earth’s magnetic field. The MHD effect includes two stages: Blast Wave and Heave. The stages are generated differently and have different durations, see Figure 21.12. The first stage with a duration from 1 to 10 seconds is based on spreading (throwing out) of large plasma products generated upon the detonation in the thin air (at high altitude) under the action of the Earth’s magnetic field.

Figure 21.12: HEMP MHD consists of two stages [24]: a) Blast Wave and b) Heave.

Such physical effects cause significant displacement of the Earth’s magnetic field increasing in proportion to the detonation power and altitude. The second stage is accompanied with heave and rapid lift of the incandescent air due to the detonation, together with highly ionized air mass (that is, plasma). When ionized plasma crosses power lines of the Earth’s magnetic field, the air layer is polarized, and the powerful electric field is generated. The electric field in its turn forms high-altitude flowing currents in the ionosphere. These processes are rather slow. The duration of this detonation phase is from ten to 300 seconds. This provokes complex interaction between plasma ions, magnetic field and gamma- and X-rays accompanied by generation of electric curl field. In conjunction with such processes in thin air, a relatively slowly changing electric field (in strength from less than ten to tens of v/km, see Figure 21.12) is generated close to the surface of the Earth. Despite the low strength of the electric field generated by the E3 component of the nuclear detonation’s HEMP, it induces rather high electric currents of low frequency (less than one Hz), such as quasi-DC, in lengthy metal structures (such as pipes, rails, overhead transmission lines). The most dangerous are the currents induced in overhead transmission lines, see Figure 21.13.

21.3 The effect of the E3 component of HEMP on electric power equipment | 353

Figure 21.13: Changing horizontal component of electromagnetic field strength at the Earth’s surface under the E3 component of a nuclear detonation’s EMP [25].

21.3 The effect of the E3 component of HEMP on electric power equipment Due to the very low frequency (less than one Hz) of the geomagnetically induced current (GIC), its effect is similar to the effect of direct current, so it will basically influence the electric equipment containing electromagnetic systems, such as power transformers. Saturation of the power transformer yoke with quasi-DC in neutral conductors causes the excitation current rise and high distortion of the current curve in transformer winding, as well as a significant rise in loss on the transformer and increase of the winding and core temperature, see Figure 21.14. This mode is dangerous for the transformer since there is a high probability of making the transformer inoperative and a distortion of the entire electric power system. The steady-mode transformer is a powerful source of even and odd harmonics causing overloading of the capacitor banks of reactive power compensation systems and distorting normal operation of the relay protection. In the case of overloading, the capacitor banks are disconnected by the protection devices. Combined with a simultaneous jump in steady-mode transformers consumption of reactive power (if its power is high), it causes a significant lack of reactive power in the system. At the same time, the voltage is decreased, and the transformer on-load tap-changer is automatically triggered to restore the voltage level. Transformer on-load tap-changing device contractors are not designed for switching currents containing significant DC components and would likely be destroyed, thus damaging on-load tap-changing devices

354 | 21 The problem of impact of geomagnetically induced currents on power transformers

Figure 21.14: Steady-mode distortion of current shape in power transformer upon the flow of geomagnetic current of 50, 100 and 150 A [26] within the transformer neutral conductor.

and shorting the adjusting part of transformer winding. This mode should immediately trigger the circuit breakers (CB) and disconnect the damaged transformers. But are the high-voltage CB capable of disconnecting such short-circuit currents and load currents containing high DC components? The problem is that they are not designed for switching such currents. At the same time, what happens with capacitors shunting in-series poles of such a CB under the high-frequency harmonics? It is clear that there are more questions than answers. However, it is known that high-power solar storms generating effects similar to E3 of a nuclear detonation’s HEMP repeatedly have caused severe damage to power equipment and collapsed power systems in various countries of the world.

21.4 Protection of power equipment from geomagnetically induced currents It is quite clear that, in order to properly protect the power system equipment, what is needed is the prevention of the flow of GIC through the equipment. To do this, we need to prevent GIC flow through the overhead transmission lines (using capacitor banks for longitudinal capacitive compensation) or block the GIC penetration into the neutral conductor of power transformer (mounting capacitors in-series into the neutral ground circuit). Since a longitudinal capacitive compensation battery is very expensive, it is rarely used and only in long overhead transmission lines for balancing the line’s inductive

21.4 Protection of power equipment from geomagnetically induced currents | 355

reactance. Recently, capacitor-based units preventing GIC penetration into the power transformer neutral conductor have been intensively developed. However, there is a problem: If GIC is absent, such unit doesn’t have to affect normal operation of transformers and mains, i. e., it should not reduce effectiveness of neutral conductor PE while it must withstand the flow of high short-circuit currents and generate ferroresonance and overvoltage in transition modes. For this reason, the operation algorithm of all types of such units ensures constant shunting of the capacitor bank with bypass power CB, and the capacitor bank is made operable upon the de-shunting (opening this CB) only at the moment of GIC discovery, see Figure 21.15.

Figure 21.15: Typical design of the device blocking GIC in the neutral conductor of power transformer. T—transformer; S—switching apparatus designed for making the device inoperative; CB—special circuit breaker designed for disconnecting AC/DC; C—capacitors bank; R—current limiting resistor; F—special protection against overvoltage under emergency currents in neutral circuits.

The appearance of the emergency mode in the circuit with capacitors integrated in the neutral line can cause the appearance of very high voltages (above the neutral line isolation level) in the neutral circuit of the transformer and at the capacitors. Thus, such units should be equipped with special devices protecting against overvoltage F (regular power varistors are not suitable due their limited power dissipation for relatively prolonged short-circuit current). Figure 21.15 illustrates a conventional protection device containing a set of six powerful high-voltage thyristors and vacuum arresters; in practice, powerful controlled three-electrode arresters of special design (see Figure 21.16) are also used instead of the thyristor set. Moreover, the disconnection of this unit with special isolating switch-over apparatus (S) equipped with a discharge grounding switch designed for making the device inoperable (for technical maintenance) without transformer disconnection should be ensured. It is clear that such a unit is both complex and expensive (more than 300,000 USD), see Figure 21.17.

356 | 21 The problem of impact of geomagnetically induced currents on power transformers

Figure 21.16: High-power high-voltage controlled arresters by Advanced Fusion Systems types 4275 (35 kV, 100 kA); 3275 (500 kV, 250 kA); 4138 (75 kV, 250 kA).

Figure 21.17: GIC blocking unit integrated into the power transformer neutral conductor. Pictured at the top is the unit manufactured by ABB Company equipped with controller manufactured by SEL, and below that is the unit designed by Phoenix Electric.

There are other methods of protecting power transformers from GIC that require the modification of transformer design. Additional nonmagnetic gaps installed in the transformer core reduce the probability of its saturation but affect the basic technical indices of the transformer. Other known technical solution (see Figure 21.18) include the installation in the transformer core of additional winding balancing the DC effects and shunted with the special element (31) characterized by high-DC resistance and low-AC resistance.

21.4 Protection of power equipment from geomagnetically induced currents | 357

Figure 21.18: Power transformer with additional winding balancing GIC. 2–4—basic winding; 5–7— compensation winding; 31—element with high DC resistance and low AC resistance.

Apparently, this “special element” can be the capacitor bank while it is not clearly stated in US Patent 7432699. USA Patent 7489485 that describes the same technical solution. There are other technical solutions suggesting connecting this compensation winding to the external controlled DC source balancing GIC. Some other patents suggest connecting the transformer windings according to the Inverse Zig-Zag method allowing for cancellation of equal but opposite excitation currents at each phase, so the core is not saturated. All such technical solutions require modifying the power-transformer manufacturing processes, deteriorating the technical parameters and leading to the significant rise in price. So, any technical measures aimed to prevent/cancel GIC currents lead to significant material costs. Accordingly, there is the question if it is reasonable to invest significant means into electric equipment damage control to cope with such exceptional emergencies as high-altitude nuclear detonation. Some world-leading manufacturers of high-voltage transformers (such as Siemens, ABB, etc.) have announced the development of special transformers (“GIC Safe Power Transformers”) capable of withstanding GIC of up to 50 A within several hours and individual GIC pulses with amplitudes up to 200 A within several hours. Promotional materials do not contain any technical solutions enhancing the transformers immunity to GIC. However, it is clear that such solutions do not include measures blocking the penetration of quasi-DC current into the transformer neutral conductor, or balancing magnetic flows in the core since such solutions do not limit GIC exposure time, as well as not strictly limiting the GIC value. Probably, such transformers sustain high temperatures due to isolation of plates with special high-temperature lacquers and winding isolation materials. Improvement of the transformer immunity to the quasi-DC currents flowing through is not the complete solution since, as already mentioned, the transformer with a saturated core can significantly affect many other types of power equipment. This means that maintaining transformer operability doesn’t ensure operability of the power system.

358 | 21 The problem of impact of geomagnetically induced currents on power transformers A high investment in the protection of power system against individual and unlikely events is hardly reasonable. So, why is such protection kind developed and offered on the market? The point is that GIC generated by high-altitude nuclear detonations is also generated during the high-power solar storms repeatedly causing severe power system collapses. However, the danger of solar storms depends on the location. The most dangerous are the poles of the Earth and the least dangerous is the equator. Regions far away from the poles are not exposed to significant GIC capable of affecting power system operability. Nevertheless, the developers of protection usually mention that solar storm and high-altitude nuclear detonation GICs are very similar and recommend installing such devices in order to improve the power system immunity to high-altitude nuclear detonations despite the region. It sounds quite true. But there is one significant difference between GICs challenging this logic. It is GIC length. Upon high-altitude nuclear detonation, the GIC lasts only a few minutes. Within this period, power transformers of high heat-sink capacity do not have enough time to heat to dangerously high temperature, see Figure 21.19. For the first time, this very important conclusion was formulated by us in articles in Russian and English [1, 27], published as far back as 2011. Moreover, the article in English [27] was published on the internet since 2011, and direct links to it are given

Figure 21.19: Example of power transformer winding (left) and yoke (right) heating under the geomagnetic currents of 20, 30 and 50 A flowing in the neutral.

21.4 Protection of power equipment from geomagnetically induced currents | 359

by Google when searching for information on this topic. In this connection, the report of the world-famous EPRI research center [28], published in 2017, in which this conclusion is repeated, is somewhat confusing, but there are no references to our article. This is what the main conclusion made in the report states (page 3–9): … the hundreds to thousands of transformers that were then evaluated in more detail by performing a time domain thermal analysis, only a small number of them were found to be at potential risk of thermal damage. To provide some context, for the worst-case target location, only 14 of the tens of thousands of transformers that were included in the model were found to be at potential risk of thermal damage.

It is clear that, under the high GIC, the temperature of transformer parts will be high but not dangerous due to the very short exposure time. Currents distorted with asymmetrically saturated transformer are much more dangerous for other types of power equipment less inert than power transformers as already mentioned. Solar storm GICs may last for many hours, and during that period power supply should be ensured. High-altitude detonation GICs last several minutes, and during that period the power equipment can be disconnected to prevent damage and then put into operation again. Since GIC builds-up slowly, the transformers can be disconnected immediately after the discovery of the DC component in the neutral current before the core is saturated and the ensuing processes occur. Such protection of power system equipment from the EMP MHD of a nuclear detonation looks more reasonable compared to that discussed here since it is both highly effective and less expensive. All charges for the implementation of such protection relate to the installation of a special relay reacting to the appearance of the DC component in the neutral current and immediately triggering the transformer disconnection. Such a relay must have a special design different from the one used by ABB for controlling SolidGround™ device (industrial PLK of SEL-2240 type) because such a device provides protection from the E3 of a nuclear detonation’s EMP MHD, which appears after E1 and E2 and is likely to be capable of damaging microprocessor control device SEL-2240 before it is triggered. Figure 21.20a shows the operational principle of the relay sensitive to the DC component in the power transformer neutral and insensitive to the widely varying AC component. The relay consists of a reed switch, RS, with a coil placed on the cable (bus) that connects the transformer neutral to the grounding point perpendicular to the axis of the cable and a conventional toroidal current transformer, CT, installed on the same cable. If there is no DC current in the neutral the magnetic field of the cable (bus) acting directly on the reed switch, this is fully compensated by the magnetic field of the coil put on the reed switch and powered by the current transformer. AC current changes in the neutral lead proportionally to the changes in both magnetic fields acting on the reed switch, and to their compensation. Under high DC currents in the neutral (over

360 | 21 The problem of impact of geomagnetically induced currents on power transformers

Figure 21.20: Power-transformer protection relay protecting from low-frequency geomagnetic currents induced in the neutral circuit.

10–20 A), the balance of the magnetic fields acting on the reed switch is offset: The magnetic field of the cable (bus) still acts while the compensating magnetic field of the coil energized by the current transformer is disabled because the DC component of the current is not transformed by the current transformer. This leads to reed-switch activation. An actual relay circuit includes an additional output amplifier installed on a VS thyristor, varistor RU and the R1C1—all protecting the thyristor from interferences and voltage surges, see Figure 21.20b. The relay is equipped with a continuous electrostatic shield and a ferromagnetic shield with the only window on the cable side next to the reed switch and is connected to the circuit of the CB switch trip coil through a special twisted-pair control cable with the combined multi-layer shielding grounded at both ends and resistant to the electromagnetic pulses. The relay can be constructed on miniature high-voltage vacuum reed switches, for example, of type KSK-1A85 (manufactured by Meder Electronics), with the electric strength of insulation between the contacts of 4000 V and the bulb having a diameter of 2.75 mm and length of 21 mm. This reed switch is capable of switching loads up to 100 W (the maximum switching voltage is 1000 V; the maximum switching current is 1 A) with the response time of one ms and a maximum sensitivity of 20 A. Additional ferromagnetic elements (magneticfield concentrators) located next to the reed switch can be used to increase the sensitivity, see Figure 21.20c. To get a relay with lower sensitivity and a higher pickup, the longitudinal axis of the reed switch should form a non-perpendicular angle to the axis of the cable on which it is installed.

21.4 Protection of power equipment from geomagnetically induced currents | 361

The thyristor should also be miniature and of high-voltage, e. g., of type SKT50/18E (manufactured by Semicron), with a maximum voltage of 1800 V and maximum continuous current of 75 A, and must withstand high rates of voltage rise (1000 V/µs) under a wide operating temperature range (−40–130 °C). The power circuit of the trip coil is equipped with storage capacitor C3 enabling switch activation even under the loss of operating voltage. This section of the circuit can be slightly complicated and executed as a separate module, see Figure 21.21. The same modules can also be used to support the power supply of individual microprocessor-based relay protection and automation devices for the time necessary to carry out protective shutdown operations of the electrical equipment.

Figure 21.21: Unit backup feeding of circuit breaker trip coil and digital protection relays.

362 | 21 The problem of impact of geomagnetically induced currents on power transformers The R2C2 in series is designed to further enhance the immunity of the device. Capacitor C2 provides a certain delay of the thyristor switch-on, preventing it from unlocking under the powerful impulse noise. Application of the discrete high-voltage components instead of conventional microelectronics in the relay ensures its high reliability under powerful electromagnetic interference and surge voltages specific to solar storms and HEMP.

21.5 Conclusions 1.

The analysis makes possible concluding that there are no experimental data confirming damage to power transformers by geomagnetically induced currents during solar storms in South Africa, countries located on South African latitudes, Middle East countries, Russia, India and many others. 2. Nowadays, there are no experimentally confirmed data that would urge revisiting of the established acknowledged zones of higher GIC level that pose a danger for power transformers. 3. The references made elsewhere regarding damage to power transformers during solar storms (supposedly this happened in South Africa) are actually indefensible and should not be considered when studying the issue about the necessity to take special measures to protect such transformers in any region. 4. Nuclear detonation EMP MHD results in quasi-DC current flowing through the power transformer neutral conductors and affects both transformers and many other types of power equipment (particularly, capacitor banks and high-voltage circuit breakers). Thus, technical arrangements aimed to protect the systems from nuclear detonation EMP MHD should protect both transformers and other types of power system equipment. Technical arrangements aimed to improve only the transformer immunity to the flows of such currents without blocking or canceling GIC are not deemed as effective. 5. Available solutions preventing transformer core saturation can be subdivided as follows: – External units installed at the transformer neutral breaks and blocking the flow of quasi-DC current in the neutral circuit – Internal modification of the transformer (windings or core) preventing the saturation of the core during the flow of quasi-DC current in the neutral circuit All known solutions intended to maintain proper operation of power systems exposed to GIC are costly. 6. Any known technical solutions aimed at maintaining the normal functioning of the power system during the GIC impact on its elements are associated with significant material costs. 7. Despite of the similarity of parameters of different GICs (nuclear detonation’s E3 and solar storms), there is one significant difference, namely, the duration of

Bibliography | 363

the GIC. This difference dictates the employment of different methods of electric equipment protection from solar storms and E3 of nuclear detonation EMP MHD. 8. Known technical solutions providing protection of power system equipment from nuclear detonation EMP MHD cannot be deemed economically sound. In this case, it can be suggest that it is more reasonable to briefly (for several minutes) disconnect the power transformer triggered by special protection relay, and then put it into operation again automatically. 9. GIC-protection relay should have a special design ensuring its operation under the all components of nuclear detonation EMP.

Bibliography [1] [2] [3] [4]

[5]

[6] [7] [8] [9] [10] [11]

[12] [13] [14] [15]

[16]

Gurevich V. Power Transformers Too Are Subject to Effect of the Sun. Electr. Mark., 2011, Vol. 5, pp. 48–51, (by Russian). Gurevich V. Protection of Power Transformers against Geomagnetically Induced Currents. Serb. J. Electr. Eng., 2011, Vol. 8, No. 2, pp. 333–339. Marusek J. A. Solar Storm Threat Analysis. Impact, 2007, Bloomfield, Indiana 47424. Schulte in den Bäumen H., Moran D., Lenzen M., Cairns I., Steenge A. How Severe Space Weather Can Disrupt Global Supply Chains. Nat. Hazards Earth Syst. Sci., 2014, Vol. 14, No. 10, pp. 2749–2759. Gaunt C. T., Coetzee G. Transformer Failures in Regions Incorrectly Considered to Have Low GIC Risk, Mat Post 07, 3rd European Conference on MV & HV Substation Equipment, Nov 15–17, 2007, Lyon, France, Proceedings of Power Tech, July 15, 2007, Lausanne, Switzerland. Effects of Geomagnetically Induced Currents on Power Transformers and Power Systems. A2-304 CIGRE, 2012. E-threat Protection for Infrastructure Continuity. Final Report of Electrical Infrastructure Council (EIC), 2013. Sushko V. A., Kosych D. A. Geomagnetic Storms. The Threat to the National Security of Russia. News Electr. Eng., 2013, Vol. 4. IEEE Std. C57.163-2015: IEEE Guide for Establishing Power Transformer Capability while under Geomagnetic Disturbances, 2015. Minnaar U. J., Gaunt C. T., Nicolls F. Characterisation of Power System Events on South African Transmission Power Lines. Electr. Power Syst. Res., 2012, Vol. 82, No. 1, pp. 25–32. Minhas M. S. A., Reynders J. P., De Klerk P. J. Failures in Power System Transformers and Appropriate Monitoring Techniques. High-Voltage Engineering Symposium, 22–27 August 1999, Conference Publication No. 467, IEEE, 1999, Vol. 1, London. Energy Security in South Africa. Research paper, MAPS, 2014. Marshall R. A., Gorniak H., Van Der Walt T., et al. Observations of Geomagnetically Induced Currents in the Australian Power Network. Space Weather, Jan. 2013, Vol. 11, No. 1, pp. 6–16. Caraballo R., Sanchez Bettucci L., Tancredi G. Geomagnetically Induced Currents in the Uruguayan High-Voltage Power Grid. Geophys. J. Int., August 16, 2013. Barbosa C., Alves L., Caraballo R., Hartmann G. A., Papa A. R. R., Pirjola R. J. Analysis of Geomagnetically Induced Currents at a Low-Latitude Region Over the Solar Cycles 23 and 24: Comparison between Measurements and Calculations. J. Space Weather Space Clim., 2015, Vol. 5, A35. Rajurkar S. S., Nandapurkar J. G., Kulkarni A. R. Analysis of Power Transformer failure in Transmission utilities. 16th National Power Systems Conference, 15–17 December 2010.

364 | 21 The problem of impact of geomagnetically induced currents on power transformers

[17] Mirza M., Gholami A., Aminifar F. Failures Analysis and Reliability Calculation for Power Transformers. J. Electr. Syst., 2006, Vols. 1–2, pp. 1–12. [18] Shayan Tariq J., Raheel A., Akif Zia K. Transformer Failures, Causes & Impact. International Conference Data Mining, Civil and Mechanical Engineering (ICDMCME 2015) Feb. 1–2, 2015 Bali (Indonesia), pp. 49–52. [19] Watari S., Kunitake M., Kitamura K., et al. Effects of Geomagnetically Induced Current on Power Grids. J. Natl. Inst. Inf. Commun. Technol., 2009, Vol. 56, Nos. 1–4, pp. 125–133. [20] Marshall R. A., Dalzell M., Waters C. L., Goldthorpe P., Smith E. A. Geomagnetically Induced Currents in the New Zealand Power Network. Space Weather, 2012, Vol. 10, No. 8. [21] GIC Module to Analyze Geomagnetic Disturbances on the Grid PSS® E. Siemens Power Technologies International, 09.2013, Siemens AG and Siemens Industry, Inc. [22] GIC-4. Geomagnetic Induced Current Sensor. Dynamic Ratings, 2015. [23] Hoffman G. ECLIPSE GMD Induced Core Saturation Detection. Advanced Power Technologies, LLC, White Paper, 2013. [24] Study to Access the Effects of Magnetohydrodinamic Electromagnetic Pulse on Electric Power Systems. Report ORNL/sub-83/43374/1/v. Oak Ridge National Laboratory, 1985. [25] IEC 61000-2-9 Electromagnetic Compatibility (EMC) – Part 2: Environment – Section 9: Description of HEMP Environment. Radiated disturbance, 1996. [26] The Late-Time (E3) High-Altitude Electromagnetic Pulse (HEMP) and Its Impact on the U.S. Power Grid. Report Meta-R-321, Metatech Corp., 2010. [27] Gurevich V. Protection of Power Transformers against Geomagnetically Induced Currents. Serb. J. Electr. Eng., 2011, Vol. 8, No. 2, pp. 333–339. [28] Magnetohydrodynamic Electromagnetic Pulse Assessment of the Continental U.S. Electric Grid: Geomagnetically Induced Current and Transformer Thermal Analysis. Rep. No. 3002009001, EPRI, 2017, 70 p.

A Standards on HEMP A.1 Standards of International Electrotechnical Commission (IEC) 1.1 1.2 1.3 1.4 1.5 1.6 1.7

1.8

1.9

1.10

1.11

1.12 1.13 1.14 1.15

IEC TR 61000-1-3 Electromagnetic compatibility (EMC) – Part 1–3: General – The effects of high-altitude EMP (HEMP) on civil equipment and systems. IEC 61000-1-5 High power electromagnetic (HPEM) effects on civil systems. IEC 61000-2-9 Electromagnetic compatibility (EMC) – Part 2: Environment – Section 9: Description of HEMP environment – Radiated disturbance. IEC 61000-2-10 Electromagnetic compatibility (EMC) – Part 2-10: Environment – Description of HEMP environment – Conducted disturbance. IEC 61000-2-11 Electromagnetic compatibility (EMC) – Part 2-11: Environment – Classification of HEMP environments. IEC 61000-2-13 Electromagnetic compatibility (EMC) – Part 2-13: Environment – High power electromagnetic (HPEM) environments – Radiated and conducted. IEC 61000-4-23 Electromagnetic compatibility (EMC) – Part 4-23: Testing and measurement techniques – Test methods for protective devices for HEMP and other radiated disturbances. IEC 61000-4-24 Electromagnetic compatibility (EMC) – Part 4: Testing and measurement techniques – Section 24: Test methods for protective devices for HEMP conducted disturbance. IEC 61000-4-25 Electromagnetic compatibility (EMC) – Part 4-25: Testing and measurement techniques – HEMP immunity test methods for equipment and systems. IEC 61000-4-32 Electromagnetic compatibility (EMC) – Part 4-32: Testing and measurement techniques – High-altitude electromagnetic pulse (HEMP) simulator compendium. IEC61000-4-33 Electromagnetic compatibility (EMC) – Part 4-33: Testing and measurement techniques – Measurement methods for high-power transient parameters. IEC 61000-4-35 Electromagnetic compatibility (EMC) – Part 4-35: Testing and measurement techniques – HPEM simulator compendium. IEC 61000-4-36 Electromagnetic compatibility (EMC) – Testing and measurement techniques – IEMI Immunity Test Methods for Equipment and Systems. IEC/TR 61000-5-3 Electromagnetic compatibility (EMC) – Part 5-3: Installation and mitigation guidelines – HEMP protection concepts. IEC/TS 61000-5-4 Electromagnetic compatibility (EMC) – Part 5: Installation and mitigation guidelines – Section 4: Immunity to HEMP – Specifications for protective devices against HEMP radiated disturbance.

https://doi.org/10.1515/9783110639285-022

366 | A Standards on HEMP 1.16 IEC 61000-5-5 Electromagnetic compatibility (EMC) – Part 5: Installation and mitigation guidelines – Section 5: Specification of protective devices for HEM conducted disturbance. 1.17 IEC 61000-5-6 Electromagnetic compatibility (EMC) – Part 5-6: Installation and mitigation guidelines – Mitigation of external EM influences. 1.18 IEC 61000-5-7 Electromagnetic compatibility (EMC) – Part 5-7: Installation and mitigation guidelines – Degrees of protection provided by enclosures against electromagnetic disturbances (EM code). 1.19 IEC 61000-5-8 Electromagnetic compatibility (EMC) – Part 5-8: Installation and mitigation guidelines – HEMP protection methods for the distributed infrastructure. 1.20 IEC 61000-5-9 Electromagnetic compatibility (EMC) – Part 5-9: Installation and mitigation guidelines – System-level susceptibility assessments for HEMP and HPEM.

A.2 Standards of Institute of Electrical and Electronics Engineers (IEEE) 2.1 IEEE P1642 Recommended Practice for Protecting Public Accessible Computer Systems from Intentional EMI.

A.3 Standards of European Commission 3.1 Topic SEC-2011.2.2-2 Protection of Critical Infrastructure (structures, platforms and networks) against Electromagnetic (High Power Microwave (HPM)) Attacks.

A.4 Standards of International Telecommunication Union (ITU) 4.1 ITU-T K.87. Guide for the application of electromagnetic security requirements – Overview, 2016. 4.2. ITU-T K.78. High altitude electromagnetic pulse immunity guide for telecommunication centers, 2016. 4.3 ITU-T K.81. High-power electromagnetic immunity guide for telecommunication systems, 2014.

A.5 Military Standards (USA) 5.1. MIL-STD-2169B High-Altitude Electromagnetic Pulse (HEMP) Environmental, 2012 (Classified).

A.6 NATO Standards | 367

5.2 MIL-STD-188-125-1 High-Altitude Electromagnetic Pulse (HEMP) Protection for Ground Based C41 Facilities Performing Critical. Time-Urgent Mission. Part 1 Fixed Facilities, 2005. 5.3 MIL-STD-188-125-2 High-Altitude Electromagnetic Pulse (HEMP) Protection for Ground Based C41 Facilities Performing Critical. Time-Urgent Mission. Part 2 Transportable systems, 1999. 5.4 MIL-STD-461F Requirements for the Control of Electromagnetic Interference Characteristics of Subsystems and Equipment, 2007. 5.5 MIL-STD-464C Electromagnetic Environmental Effects. Requirements for Systems, 2010. Test Operations Procedure Report No. 01-2-620 High-Altitude Electromagnetic Pulse (HEMP) Testing. 5.6 MIL-STD-1377 Effectiveness of Cable, Connector, and Weapon Enclosure Shielding and Filters in Precluding Hazards of Electromagnetic Radiation to Ordnance (HERO), 1971. 5.7 MIL-HDBK-240 Hazards of Electromagnetic Radiation to Ordnance (HERO) Test Guide, 2002.

A.6 NATO Standards 6.1 NATO AECTP-500 Ed. 4. Electromagnetic Environmental Effects Test and Verification, 2011. 6.2 NATO AECTP-250 Ed. 2. Electrical and Electromagnetic Environmental Conditions, 2011.

B EMP and its Impact on Power System (List of Reports) B.1 EMP Theory 1.1 Report DTIC ADA955391: Capabilities of Nuclear Weapons. Part 2. Damage Criteria. Change 1. Chapter 7. Electromagnetic Pulse (EMP) Phenomena. NNA EM-1 / P. J. Dolan. – Defense Nuclear Agency, 1978. 1.2 Report DTIC ADA059914: Effect of Multiple Scattering on the Compton Recoil Current. – Mission Research Corp. for Defense Nuclear Agency, 1978. 1.3 Report AFWL-TR-80-402: EMP Interaction: Principles, Techniques and Reference Data / K. S. H. Lee. – Air Force Weapons Laboratory, 1981. 1.4 Interaction Note 435: Interaction of High-Altitude Electromagnetic Pulse (HEMP) with Transmission Lines. An Early-Time Consideration / K. S. H. Lee, et al. – LuTech Inc., 1983. 1.5 Report DTIC ADB094426: A Guide to Nuclear Weapons Phenomena and Effects Literature / Kenneth E. Gould. – Defense Nuclear Agency, 1984. Unclassified in 1986. 1.6 Interaction Note 458: A Study of Overhead Line Responses to High Altitude Electromagnetic Pulse Environments / F. M. Tesche. – LuTech Inc., 1986. 1.7 Report ORNL/Sub/86-18417/1: A Nominal Set of High-Altitude EMP Environments / C. L. Longmire, R. M. Hamilton, J. M. Hahn. – Oak Ridge National Laboratory, 1987. 1.8 Report ORNL/Sub/85-27461/1: The Effects of Corona on Current Surges Induced on Conducting Lines by EMP: A Comparison of Experimental Data with Results of Analytic Corona Models / J. P. Blanchard, F. M. Tesche, and B. W. McConnell, Oak Ridge National Laboratory, 1987. 1.9 Report DTIC ADA234306: Comparison of the Frequency Spectra of HEMP and Lightning / M. A. Uman. – Defense Nuclear Agency, 1991.

B.2 Geomagnetically Induced Currents and its Impact on Power System 2.1

2.2

2.3

Report ORML/Sub-83/43374/1/V3: Study to Assess the Effects of Magnetohydrodynamic Electromagnetic Pulse on Electric Power Systems / J. R. Legro, N. C. AbiSamra, and F. M. Tesche. – Oak Ridge National Laboratory, 1985. Report ORNL-6665: Electric Utility Industry Experience with Geomagnetic Disturbances / P. R. Barnes, D. T. Rizy, B. W. McDonell. – Oak Ridge National Laboratory, 1991. High-Impact, Low-Frequency Event Risk to the North American Bulk Power System. – A Jointly-Commissioned Summary Report of the North American Electric

https://doi.org/10.1515/9783110639285-023

370 | B EMP and its Impact on Power System (List of Reports) Reliability Corp. and the U. S. Department of Energy’s November 2009 Workshop. NERC, 2010. 2.4 Report Meta-R-322: Low-Frequency Protection Concepts for the Electric Power Grid: Geomagnetically Induced Current (GIC) and E3 HEMP Mitigation / J. Kappenman. – Metatech Corp., 2010. 2.5 Geo-Magnetic Disturbances (GMD): Monitoring, Mitigation and Next Steps. A Literature Review and Summary of the April 2011 NERC GMD Workshop. – North American Electric Reliability Corp. (NERC), 2011. 2.6 Effect of Geomagnetic Disturbances on the Bulk Power System and Electromagnetic Pulse Effect on the U. S. Power Grid / Thomas S. Popik. – Task Force on National and Homeland Security, 2012. 2.7 Report A2-304: Effects of Geomagnetically Induced Currents on Power Transformers and Power Systems / R.Girsis, V. Vedante, K. Gramm. – CIGRE, 2012. 2.8 Report A2-303: Behavior of Transformers Under DC/GIC Excitation: Phenomenon, Impact on Design / Design Evaluation Process and Modeling Aspects in Support of Design / T. Ngnegueu, F. Marketos, F. Devaux. – CIGRE, 2012. 2.9 Report LUTEDX/(TEIE-7242)/1-21/(2013): Geomagnetic disturbances and their impact on power systems / Olof Samuelsson. – Division of Industrial Electrical Engineering and Automation Faculty of Engineering, Lund University, 2013. 2.10 Report LUTEDX/(TEIE-5328)/1-062/(2014) GIC Distribution / Carlos David Fernández Barroso. – Division of Industrial Electrical Engineering and Automation Faculty of Engineering, Lund University, 2014.

B.3 EMP Impact on Power System 3.1 3.2

3.3 3.4 3.5 3.6

Report ORNL-4958: Power System EMP Protection, Final Report / H. Marable, P. R. Barnes, and D. B. Nelson. – Oak Ridge National Laboratory, 1975. Report ORNL-4919: Effect of Nuclear Electromagnetic Pulse (EMP) on Synchronous Stability of the Electric Power System / R. W. Manweiler. – Oak Ridge National Laboratory, 1975. Report DTIC ADA014489: Electromagnetic Pulse and Civil Preparedness. – Defense Civil Preparedness Agency, Washington, 1975. Report ADA009228: Electromagnetic-Pulse Handbook for Electric Power Systems / Edward F. Vance. – Stanford Research Institute, 1975. Report HCP/T5103-01: Impact Assessment of the 1977 New York City Blackout. – U. S. Department of Energy, Division of Electric Energy Systems, 1978. Report NUREG/CR-3069: Interaction of Electromagnetic Pulse with Commercial Nuclear Power Plant / D. M. Ericson, D. F. Strawe, S. J. Sandberg, etc. – Sandia National Laboratories, 1983.

B.3 EMP Impact on Power System

3.7

3.8

3.9

3.10

3.11

3.12

3.13

3.14

3.15

3.16

3.17

3.18

3.19

3.20

| 371

Report ORNL/Sub/82-47905/1: Electromagnetic Pulse (EMP) Interaction with Electrical Power Systems / H. W. Zaininger. – Oak Ridge National Laboratory, 1984. Report ORNL-6033: Nuclear Electromagnetic Pulse (EMP) and Electric Power Systems / P. R. Barnes, E. F. Vance, and H. W. Askins. – Oak Ridge National Laboratory, 1984. Report ORNL/Sub-84/73986/1: HEMP-Induced Transients in Transmission and Distribution (T and D) Lines / N. Engheta, K. S. H. Lee, F. C. Yang. – Oak Ridge National Laboratory, 1985. Report ORNL/Sub/84-89643/1: Study for a Facility to Simulate High Altitude EMP Coupled Through Overhead Transmission Lines / D. Smith. – Oak Ridge National Laboratory, 1985. Report ORNL/qub/84-89642/2: Design Concepts for a Pulse Power Test Facility to Simulate EMP Surges in Overhead Power Lines / R. Dethlefsen. – Oak Ridge National Laboratory, 1985. Report ORNL/Sub/84-89642/1: Design Concepts for a Pulse Power Test Facility to Simulate EMP Surges in Overhead Power Lines: Part 1, Fast Pulse / A. Ramrus – Oak Ridge National Laboratory, 1985. Report ORNL/qub/84-89642/2: Design Concepts for a Pulse Power Test Facility to Simulate EMP Surges in Overhead Power Lines: Part 2, Slow Pulses / R. Dethlefsen. – Oak Ridge National Laboratory, 1986. ORNL/Sub/83-43374/V1: Study to Assess the Effects of Electromagnetic Pulse on Electric Power Systems – Phase I Executive Summary / J. R. Legro, et al. – Oak Ridge National Laboratory, 1985. Report ORNL/Sub/83-43374/V4: Study to Assess the Effects of Nuclear Surface Burst Electromagnetic Pulse on Electric Power Systems- Phase I Final Report / J. R. Legro, et al. – Oak Ridge National Laboratory, 1985. Report ORNL/Sub/83-43374/1/VI: Study to Assess the Effects of High-Altitude Electromagnetic Pulse on Electric Power Systems- Phase I Final Report / J. R. Legro, et al. – Oak Ridge National Laboratory, 1986. Report No. AST 88-2081: HEMP-Type Impulse Transfer Tests on High-Voltage Bushing Current Transformers at the Maxwell Laboratory / E. R. Taylor. – Westinghouse AST, 1988. Report No. AST 88-7072: HEMP-Type Impulse Tests on High-Voltage Potential Transformers at the Maxwell Laboratory / E. R. Taylor. – Westinghouse AST, 1988. Report DTIC ADA206952: The Effects of High-Altitude Electromagnetic Pulse (HEMP) on Telecommunications Assets. – National Communication System, 1988. Report ORNL/Sub/88-02238/1: HEMP Test and Analysis of Selected RecloserControl Units / T. K. Liu, et al. – Oak Ridge National Laboratory, 1989.

372 | B EMP and its Impact on Power System (List of Reports) 3.21 Report EP 1110-3-2: Engineering and Design – Electromagnetic Pulse (EMP) and Tempest Protection for Facilities. – Department of the Army, U. S. Army Corps of Engineers, 1990. 3.22 Report ORNL/Sub/83-43374/2: Impacts of a Nominal Nuclear Electromagnetic Pulse on Electric Power Systems / V. J. Kruse, D. L. Nickel, J. J. Bonk. – Oak Ridge National Laboratory, 1991. 3.23 Report ORNL/Sub/85-28611/2: Impact of Steep-Front Short-Duration Impulses on Power System Insulation / L. M. Burrage, et al. – Oak Ridge National Laboratory, 1991. 3.24 Report ORNL/Sub/83-43374/2: Impact of a Nuclear Electromagnetic Pulse on Electric Power Systems (Phase III, Final Report). – Oak Ridge National Laboratory, 1991. 3.25 Report CD-90-0014: An Assessment of the Electromagnetic Pulse (EMP) Effects on the U. S. Civilian Infrastructure – Unclassified Summary and Recommendations / P. Chrzanowski, J. Futterman. – Lawrence Livermore National Laboratory, 1992. 3.26 Report ORNL/Sub-88-SC863: HEMP-Induced Transients in Electric Power Substations (final report) / C. V. Wiggins, D. E. Thomas, T. M. Salas. – Oak Ridge National Laboratory, 1992. 3.27 Report ORNL/Sub-91-SG913/1: Recommended Engineering Practice to Enhance the EMI/EMP Immunity of Electric Power Systems / C. W. Wagner, W. E. Feego. – Oak Ridge National Laboratory, 1992. 3.28 Report ORNL-6708: Electromagnetic Pulse Research on Electric Power Systems. Program Summary and Recommendations / P. R. Barnes, B. W. McConell, J. W. Van Dyke. – Oak Ridge National Laboratory, 1993. 3.29 Report OPNL/Sub/91-SG105/1: HEMP Emergency Planning and Operating Procedures for Electric Power Systems / T. W. Reddoch, L. C. Markel. – Oak Ridge National Laboratory, 1993. 3.30 Report ORNL/TM – 1999/93: Assessment and Testing of Long-Line Interface Devices / P. R. Barnes, B. W. McConnell. – Oak Ridge National Laboratory, 2000. 3.31 Technical Manual TM 5-690: Grounding and Bonding in Command, Control, Communications, Computer, Intelligence, Surveillance and Reconnaissance (C41SR). – Headquarters Department of the Army, 2002. 3.32 Report for Congress RL32544: High Altitude Electromagnetic Pulse (HEMP) and High Power Microwave (HPM) Devices: Threat Assessments / Clay Wilson. – Congressional Research Service, 2008. 3.33 Report of the Commission to Assess the Threat to the United States from Electromagnetic Pulse (EMP) Attack. Critical National Infrastructures. – 2008. 3.34 Critical Infrastructure Strategic Initiatives Coordinated Action Plan. – Technical Committee Report. – North American Electric Reliability Corp. (NERC), 2010.

B.3 EMP Impact on Power System

| 373

3.35 Report Meta-R-320: The Early-Time (E1) High-Altitude Electromagnetic Pulse (HEMP) and Its Impact on the U. S. Power Grid / E. Savage, J. Gilbert, W. Radasky. – Metatech Corp., 2010. 3.36 Report Meta-R-321: The Late-Time (E3) High-Altitude Electromagnetic Pulse (HEMP) and Its Impact on the U. S. Power Grid / J. Gilbert, J. Kappenman, W. Radasky, E. Savage. – Metatech Corp., 2010. 3.37 Report Meta-R-323: Intentional Electromagnetic Interference (IEMI) and Its Impact on the U. S. Power Grid / W. Radasky, E. Savage. – Metatech Corp., 2010. 3.38 Report Meta-R-324: High-Frequency Protection Concepts for the Electric Power Grid / W. Radasky, E. Savage. – Metatech Corp., 2010. 3.39 Strategy Research Project: Electromagnetic Pulse – A Catastrophic Threat to the Homeland / R. Oreskovic. – U. S. Army War College, 2011. 3.40 Report HC 1552: Developing Threats to Electronic Infrastructure. – House of Commons, Defence Committee, 2011. 3.41 Report HC 1552: Developing Threats Electro-Magnetic Pulses (EMP). – House of Commons, Defence Committee, 2012. 3.42 Report 5200.44: Protection of Mission Critical Functions to Achieve Trusted Systems and Networks (TSN). – U. S. Department of Defense, 2012. 3.43 Critical Infrastructure Protection Committee Strategic Plan. – North American Electric Reliability Corp. (NERC), 2012. 3.44 Report 3002000796: Electromagnetic Pulse and Intentional Electromagnetic Interference (EMI) Threats to the Power Grid: Characterization of the Threat, Available Countermeasures, and Opportunities for Technology Research. – Electric Power Research Institute (EPRI), 2013. 3.45 Report 113-85: Assured Microelectronics Policy. – U. S. Department of Defense, 2014. 3.46 E-Pro Report (International Electric Grid Protection) / Chris Beck. – Electric Infrastructure Security Council, 2013. 3.47 Report FM 3-38: Cyber Electromagnetic Activities. – Headquarters Department of the Army U. S., 2014. 3.48 Report of CIGRE C4.206 Working Group: Protection of High Power Network Control Electronics Against Intentional Electromagnetic Interference (IEMI). – CIGRE, 2014. 3.49 Report INL/EXT-15-35582: Strategies, Protections and Mitigations for the Electric Grid from Electromagnetic Pulse Effects. – Idaho National Laboratory (INL), 2016.

C European Projects related to Protection against HEMP European Union Seventh Framework Programme under grant agreement number FP7-SEC-2011-285257: “Protection of Critical Infrastructures (structures, platforms, and networks) against Electromagnetic (High Power Microwave (HPM)) Attacks” 1. HIPOW: Protection of Critical Infrastructures against High Power Microwave Threats 2. STRUCTURES: Strategies for the Improvement of Critical Infrastructures Resilience to Electromagnetic Attacks 3. SECRET: Security of Railways against Electromagnetic Attacks 4. SAVELEC: Safe Control of Non-Cooperative Vehicles through Electromagnetic Means 5. Switzerland Program: Impact of Intentional Electromagnetic Interferences on Swiss Electric Power System

https://doi.org/10.1515/9783110639285-024

Index accumulator 158, 249, 274 accuracy 39, 44, 98, 99, 176 adhesion 133, 134, 215, 216, 221 adhesive 139, 141, 229, 231 adjustable 218, 246 Admiral 49 admirals 49 Aeronautical 1 Aerospace 38, 39 Africa 344–346, 348, 362 African 345–347, 362 air 3–5, 10, 36, 54, 56, 60, 63, 64, 66, 67, 81, 85, 86, 96, 128, 139, 141, 155, 166, 227, 241, 281, 285, 287, 306, 324, 352 air-balloon 5, 7, 42 air-pressure 42 Alamos 1, 3, 19, 31, 52, 55 alloy 224, 225, 231 ALLUR 306 altitude 1, 5, 7, 9, 39, 41, 42, 46, 58, 63, 67, 68, 76, 155, 215, 222, 269, 352 aluminum 143, 171, 224, 228, 231, 242, 250, 267, 337, 340, 341 Aluzinc 224, 225 Amco 224 antenna 79, 81, 82, 137, 157, 159–161, 181, 270, 279, 299, 311, 313, 319, 323–325 anti-aircraft 36 anti-ballistic 42, 45 aperture 281, 287 apparatus 75, 123, 181, 263, 265–267, 296, 300, 303, 306, 355 army 5, 12, 55, 56, 88, 337 arrest 33, 34 arrester 62, 113, 114, 171, 288, 300, 355, 356 arrestor 115 Arsenal 320, 321 Atoll 7, 9, 33 atolls 5 atomic 13, 16, 20, 22, 23, 29, 31, 34, 36, 40, 42, 43 atoms 9, 10, 60, 62, 66, 67 auxiliary-power 238, 245, 251 backup 254, 274, 275, 279, 280, 287, 361 ballistic 40, 45, 57, 59 https://doi.org/10.1515/9783110639285-025

balloon 4, 5, 215, 241, 242, 269 battery 158, 207, 228, 238–240, 242, 244–246, 248–251, 267, 274, 275, 298, 299, 354 battery-bank 239, 240, 248 battery-charger 299 Baum 53, 71, 81, 82 bilateral 179, 182 billion 155 bipolar 98, 100 blast 31, 32, 43, 63, 190, 352 bomb 1, 3, 9, 12, 13, 15, 19, 20, 22–34, 45 bomber 5, 34, 41, 86 bombing 15, 27 bounding 152 Boxer 4 boxes 310, 323, 340 braids 174, 175 branched 95, 157, 160, 269, 270, 320 breakdown 51, 61, 79, 81, 96, 113–117, 119, 122, 123, 125, 155, 163, 181, 233–235, 256, 258–261, 269, 275, 293, 317, 320, 324 breaker 164, 165, 212, 253, 280, 319, 354, 355, 361, 362 bridge 239, 246 building 1, 24, 25, 27, 28, 34, 39, 40–42, 48, 50, 79, 88, 110, 128, 129, 135, 145–146, 154, 241–243, 267, 279, 285, 299, 300, 311, 324–326, 328, 331 burst 41, 43–45, 63 bypass 159, 258, 355 bypassed 260 cabined 167 cabinet 22, 23, 27, 30, 31, 34–36, 39, 40, 44, 47, 52, 90, 110, 113, 117, 118, 123–125, 128, 138, 142, 146, 150–152, 158, 160, 162–165, 167–169, 171, 177, 197, 216, 217, 222–236, 238–240, 246, 248, 150, 251, 254, 263–267, 269, 270, 272, 279, 288, 290, 292, 298, 299, 307, 311, 331 cabinet-enclosed 125 cabinet-installed 222 cable 3, 42, 46, 53, 75–77, 95, 100, 108, 109, 112, 118, 123, 125, 138, 142–144, 146, 150, 153, 154, 156, 158–160, 164, 168, 171–178, 180–185, 187, 189, 190, 195, 197, 198, 203,

378 | Index

205, 210, 213, 214, 216, 228–235, 238, 240, 243, 250, 254, 259, 264–267, 270, 272, 277, 280–282, 286, 288, 290, 291, 299–301, 306, 319, 320, 323, 325–328, 330, 331, 337, 338, 342, 359, 360 cable-shielding 177 capacitive 149, 150, 159, 169, 180, 181, 183, 354 capacitor 79, 80, 96, 187, 216, 305, 353–355, 357, 361, 362 carbon-containing 139–141 cavities 129 cell 240–242, 285, 296, 336, 340, 341 ceramic 51, 139–141 Ceramopen 140 Cerapen 140 changeover 166, 290, 291 channel 150, 153, 155, 157, 169, 212, 253, 254, 257, 263, 271, 274, 275, 297, 298, 302, 303, 319 cheap 249 chip 104, 149, 261 circuit-breaker 215, 216 classified 3, 7, 47, 53, 60, 77, 86 Club-K 58 CMOS 100, 102 coated 139–141, 146, 262 coating 140, 141, 144, 225 coaxial 174, 176 coaxially 164 coil 212, 215, 216, 219, 220, 277, 290, 292–294, 359–361 collectors 161 Commissar 12, 15, 16, 18, 19 Commissariat 12, 22 commission 20, 26, 47, 54–56, 60, 82, 317, 328 committee 13–18, 20, 21, 27, 30, 32, 44, 55, 82, 96 communication 26, 46, 50, 53, 54, 79, 95, 96, 108, 110, 135, 153, 154, 163, 169, 182, 194, 212–214, 216, 222, 233, 238, 246, 253–255, 263–268, 273, 289, 297–299, 302, 303, 319, 323, 327, 328, 331, 335 communication-equipment 265 Compton 9–11, 60–63, 66, 67 computer 53, 93, 95–97, 107–108, 203, 289, 328, 335 concluding 362 conclusion 1, 32, 42, 56, 108, 125, 146, 159, 178, 182, 184, 205, 210, 237, 268, 278, 295, 303,

307, 315, 317, 318, 323, 325, 328–331, 340, 347, 358, 359, 362 concrete 82, 129–140, 145–146, 284, 298, 299, 305, 324–326 connector 175, 177, 194, 267, 280, 282, 283, 311 consumption 100, 353 contact 19, 149, 151, 153, 164–167, 180, 181, 185, 186, 212, 213, 215, 218–220, 267, 277, 290, 291, 293, 294, 304, 360 contact-breaker 163, 165, 166 contact-closure 214, 216, 217 container 5, 30, 41–43, 56–59, 241, 243, 244, 250, 267, 269, 279, 284, 286, 287, 336–341 container-based 58 control 26, 42, 44, 46, 64, 79, 95, 96, 123, 125, 128, 141, 144, 158, 160, 171, 174, 177, 178, 180–182, 184, 185, 194, 196–198, 203, 205, 210–216, 219, 225, 228, 229, 232–234, 238, 242, 245, 248, 250, 253, 265, 269–275, 277–279, 286, 288–290, 292, 293, 299, 303, 304, 317–319, 327, 351, 357, 359, 360 control-cable 175, 179–182, 184–187 controller 128, 251, 274, 279–283, 288–290, 318, 328, 356 control-system 154 converter 90, 228, 242, 248–250, 254, 265 copper 135–138, 153, 163–165, 167, 169, 171, 174, 184, 213, 214, 216, 226, 231, 259, 265, 337 copper-conductor 254 coupling 110, 112, 159, 161, 181, 280, 328 coupling-decoupling 299, 314 criterion 42, 177, 210, 302–304, 317, 331 critical 49, 54, 95, 97, 142, 145, 146, 166, 177, 210, 235, 240, 244, 265–267, 273, 278–280, 287–289, 309, 319, 320 cryogenic 33 crystal 132 current 51, 60, 62–64, 75, 77, 90, 95, 96, 100–102, 111, 112, 119, 122, 123, 125, 152, 154–157, 159, 163, 165, 166, 169, 179, 181, 183, 186, 187, 189, 190, 196–198, 200, 203, 207, 209, 215, 218–220, 229, 230, 232, 234, 235, 238–241, 244, 247, 249, 250–252, 255–261, 267, 268, 275, 277, 286, 292, 294, 298–300, 307, 318, 320, 324, 332, 342–344, 346–355, 357–362

Index | 379

current-limiting 210, 240, 251, 261, 262, 276, 277 curve 60, 69, 203, 228, 353 cut 153, 280 cutoff 227, 228 cutout 228, 280, 281, 285, 286 damage 3, 26, 41–44, 46–48, 51, 53, 68, 96, 97, 99, 103, 104, 153, 159, 163, 164, 166, 178, 185, 240, 269, 273, 277, 288–290, 298, 303, 305, 318–320, 328, 343, 345–350, 354, 357, 359, 362 damaged 51, 96, 260, 273, 289, 318, 320, 328, 334, 335, 343, 347, 354 damage-effect 36 damaging 99, 104, 107, 155, 353, 359 danger 64, 109, 159, 163, 273, 281, 284, 288, 296, 319, 345, 358, 362 dangerous 64, 65, 155, 156, 161, 164, 181, 184–187, 192, 237, 269, 345, 347, 348, 352, 353, 358, 359 Darlington 247 data-acquisition 3, 5, 7, 43 data-collection 2 data-transfer 246, 254, 255 density 11, 61, 67, 71, 89, 96, 100, 104, 106, 138, 155, 156, 171, 335–337 department 11–15, 17, 19, 22, 24, 26, 54, 55, 76, 90, 110, 328, 335 depend 67, 99, 116, 117, 157, 169, 200, 254, 256, 295, 307 deposition 65 destroyed 296, 353 destruction 1, 97–101, 107, 259, 319 destructive 101, 154–156, 242, 279, 319 detect 44 detected 62, 73 detection 46 detectors 317 detonated 5, 27, 41, 43, 44, 70, 75 detonating 45 detonation 28, 32, 41, 76, 269, 300, 352, 357–359, 362, 363 develop 13, 49, 56, 58, 296, 305, 317 developed 11, 12, 20, 22, 24, 33, 36, 37, 44, 46, 56, 132, 139, 140, 212, 215, 305, 350, 355, 358 developer 133, 140, 194, 237, 257, 258, 358 developing 9, 38, 57, 269

development 14, 19, 23, 24, 27, 29, 31, 36, 40, 44, 45, 47, 49, 56, 57, 88, 89, 104, 114, 154, 189, 222, 245, 246, 357 diagnostic 289 diagram 68, 72, 80, 124, 153, 189, 214, 216, 256, 258, 261, 263, 308, 317, 321, 322, 343 dielectric-breakdown 293 diesel 53, 194, 195, 240, 242–244, 267, 279, 282, 285, 286, 289 diesel-generator 51, 241, 243, 279, 285, 288 differential 149, 154, 180, 181, 190, 212, 235, 261, 318 differently 26, 37, 183, 184, 352 diode 96, 97, 99, 100, 117, 123, 192, 196, 216, 217, 233, 235, 240, 246, 259, 261, 288 dip 74, 152 dipole 81, 83, 84, 92 discharge 3, 62, 79–81, 117, 149, 153, 155, 161, 164, 192, 255–260, 275, 276, 285, 324, 326, 355 discharged 43, 155, 249, 250 discharging 181 disconnect 269, 273, 274, 277, 290, 320, 354, 363 disconnected 198, 233, 238, 243, 267, 275, 288, 289, 353, 359 disconnecting 260, 354, 355 disconnection 166, 269, 270, 280–282, 288, 355, 359 disconnectors 212 discrete 96, 99, 149, 212, 213, 216, 334, 362 discrete-command 217 dismountable 197, 198, 307 dismounted 289 disrupted 97 disruption 9, 96, 106, 149, 324, 336 dissipated 239, 247, 248 dissipated-pulse 239 dissipates 247 dissipation 355 distance 1, 5, 43, 59, 67, 83, 131, 156, 298–300, 311, 335 distorting 353 distortion 257, 302, 342, 353, 354 distributed 155, 160, 181, 265, 306 distributing 161 distribution 70, 73, 74, 80, 103, 238, 239, 264, 265, 269, 334, 348 Doppler-radar 42

380 | Index

double-layer 173 DRIVER 101 duration 61, 63, 99–101, 103, 104, 156, 182–184, 258, 296, 349, 352, 362 dynamic 192, 308, 350 early-time 159 earth 10, 60, 61, 63–65, 67, 68, 74, 75, 149, 150, 155, 157, 159, 161, 162, 180, 181, 183, 218, 269, 324, 331, 342, 352, 353, 358 earthing 324 ECLIPSE 351 effect 7, 9, 10, 28, 44, 62, 63, 79, 95, 117, 123, 125, 137, 144, 149, 154, 155, 157–159, 181–184, 207, 226, 352, 353 elastic 9, 62 electrical-field 67, 68 electrically 95 electric-energy 340 electric-field 67–70, 72, 73, 76, 79, 96, 171, 242, 331, 335 electric-field-pulse 70 electric-power 64, 150, 222, 235 electrochemical 18, 241, 242 electrodynamic 9, 62 electro-magnetic 59 electromagnetic-field 11, 137, 322 electromagnetic-wave 179 electromechanical 154, 223, 274, 289–291 electromechanical-protection 222 electron 62, 66, 67, 95, 146, 301, 313 electronic-equipment 149, 222 Electronmagnetic 313 electron-source 67 electrophysical 215 electrostatic 149, 150, 180, 326 EMC 49, 142, 154, 189, 195–197, 206, 224, 227, 253, 277, 284, 296, 301, 304, 313, 314, 317 EMI 189, 284, 323 Emicore 199, 205, 206, 208 emission 3, 9–11, 41, 47, 61, 106, 128, 129, 135, 138, 142, 144, 145, 179, 195, 196, 225, 228, 254, 319, 322, 323, 325, 336, 337 EMPact 55 environment 75, 131, 141, 179, 273, 297, 317, 328, 329 environmental 142, 166 environmentally 241 epoxy 215, 216, 220, 221, 231, 277

epoxy-resin 221 EPRI 55, 359 explosion 1, 3–11, 18, 23, 26, 30, 36, 41–44, 46–49, 51–54, 56, 60, 61, 63–65, 67–69, 74, 75, 88, 108, 128, 155, 171, 181, 222, 241, 242, 253, 305, 319, 324 explosive 1, 2, 4, 5, 7, 9, 12, 24, 27, 28, 32–34, 37, 40, 41, 43, 44, 56, 76, 241 extraordinary 15, 26, 260 extra-sensitive 340, 341 extremely 13, 214, 317, 322, 334, 335 fabric 138, 229, 231 facility 18, 22, 23, 44, 47, 50, 128, 159, 161, 163, 229, 265, 267, 273, 301, 311 failure 5, 9, 26, 42, 44, 45, 51, 61, 96, 102, 108, 185, 238–240, 244, 248, 269, 296–298, 303, 319, 320, 328, 334, 335, 344, 348 Fair-Rate 204–206, 208 Fair-Rite 206, 208, 232, 233 Faraday-cage 149 fast-response 117 fault 42, 123, 149–151, 163, 182, 296, 303, 317, 347, 348 Federal 47, 55, 306, 328 FEMA 55 ferrite 123–126, 139, 141, 142, 146, 186, 187, 197–199, 201, 203–210, 230, 232, 233, 238, 267, 288, 307 ferrite beads 205 ferrite-filter 233 ferroresonance 355 Ferroxcube 199 fiber-optic 213, 214, 216, 254, 265, 268 fibers 212 Fishbowl 9 Fishman 35 flashes 6 flashover 163 float 162 floated 162 floating 162–164, 238, 239, 246, 248, 249, 266 floating-charging 245 flux 41, 60 foam 140, 141 foamed 140 Foamglass 140 foil 171, 173, 174, 183, 214 foiled 174

Index | 381

Fourier 73, 168 four-layer 173, 183 frequency 47, 62, 72–74, 82, 105–107, 123, 129–131, 133, 136, 138–142, 144, 145, 167–169, 171, 172, 182–184, 186, 187, 193, 194, 196, 198–200, 203–209, 225, 227, 228, 230, 232, 233, 235, 247, 255, 300, 307, 308, 310–312, 321, 322, 325, 334, 336–338, 340, 342, 343, 352, 353 galvanic 214–216, 218, 219, 255 galvanic-coupled 254 gamma 9, 42, 63, 352 gas 81, 141, 155, 161, 241, 255–260, 269, 270, 275, 276, 284, 306, 319, 324, 343, 344 gas-discharge 113, 155, 191 generator 49, 51, 53, 64, 79, 80, 85, 89, 91, 115, 117, 119, 160, 175, 194, 195, 203, 240, 242–244, 260, 267, 279, 280, 285, 286, 289, 294, 301, 305, 310, 313, 314, 320 geomagnetic 47, 51, 63, 75, 343–345, 347, 348, 354, 358, 360 geomagnetically 63, 342, 350–354, 362 geomagnetic-storm 348 gerkotron 215–217 Gigavac 292, 294 gigawatts 279, 319 Gilinski 11 Gilinskiy 10 glass 139–141, 145, 222, 225, 226, 265, 267, 281 glass-ceramic 140 government 3, 26, 53, 55, 64, 324 governmental 22, 27 graphite 132, 134, 141, 231 graphite-carbon 132 grid 55, 82, 227, 229, 253, 285, 343, 345–348, 350 ground 3, 5, 10, 39, 41–44, 46, 47, 52, 58, 61, 63, 67, 69, 70, 74–77, 82–85, 87, 128, 150, 151, 154–165, 167, 169, 171, 180, 190, 193, 215, 222, 238, 239, 242, 246, 248, 266, 276, 300, 301, 303, 305, 335, 343, 350, 354 grounded 150–154, 157–160, 163, 164, 168, 185, 192, 195, 214, 216, 298, 326, 327, 342, 343, 360 grounded-cable 184 grounding 128, 149–155, 157–165, 167, 179–182, 184–187, 236–239, 246, 248, 250, 266, 288, 297–299, 324, 355, 359

grounding-system 152, 157 ground-surface 3, 8, 11 ground-system 160, 298, 299, 350 ground-to-ground 42 Haefely 294, 301, 313, 314 Hardtack 5 harmonic 350, 351, 353, 354 Hawaii 9 heated 352 heating 134, 187, 349, 358 heat-insulating 140 heat-resistant 43 heat-sink 358 helical 231 helicopter 57, 58, 85 hemisphere 74–76, 343–345, 350 HEMP-filter 210 HEMP-protecting 258 HEMP-protection 253, 256, 324 HEMP-related 324 HEMP-resistance 228, 254 HEMP-resistant 336 HEMP-shielded 266 hexagonal 227 hex-shaped 284 high-amplitude 70, 157, 195, 196, 239 high-energy 64, 342 high-frequency 117, 119, 164, 165, 174, 175, 185–187, 195, 196, 198, 203–206, 209, 210, 227, 229, 232, 235, 242, 244, 246, 257, 259, 267, 286, 288, 307, 319, 354 high-sensitive 189, 319 high-sensitivity 151 high-speed 285 high-strength 86, 87 Hioki 294 hipotronics 313 Hiroshima 15, 16, 18, 20 hybrid 81, 84, 85, 93 hydrogen 31, 241, 242 hydrogen-storage 241, 244 Hydro-Quebec 346 IEEE 323, 347 IEMI 92 ignition 45

382 | Index

immunity 154, 195, 196, 213, 214, 296, 300, 301, 303, 304, 317–319, 323, 325, 331, 332, 357, 358, 362 impedance 119, 139, 159–161, 169, 198–202, 209, 232, 233, 239, 240, 249, 252, 255, 259, 342, 343 incandescent 352 indicator 277, 317 inrush 343 instrument 39, 41, 42, 301, 313, 323 insulated 119, 128, 146, 149, 157, 158, 160, 163, 165, 167, 173, 181, 190, 266 insulating 81, 219, 266 insulation 61, 81, 97, 104, 140, 141, 143, 144, 149, 154, 163, 164, 173, 178, 181, 212, 253, 266, 269, 277, 289, 293, 320, 336, 345, 348, 360 insulator 51, 163, 164, 167, 168, 194, 266, 320, 351 Intel 104, 105 interface 163, 212, 215, 218–221, 271 interference 159, 180, 279, 296, 297, 336, 360 inventions 49, 110 inventory 334, 335, 340 inverter 102, 241, 242 invisible 238 ionization 3, 46 ionized 46, 63, 156, 342, 352 ionosphere 1, 46, 64, 342, 352 ionospheric 46, 156, 343 isolated 118, 154, 162, 169, 218, 299 isolating 212, 221, 266, 355 isolation 18, 153, 154, 162, 167, 169, 214–219, 289, 292, 293, 295, 355, 357 isoline 75 ISOTOP-type 248 Israel 84, 92, 93, 306, 345 Istra 89, 306 Japan 92, 139, 348 Japanese 175 jump 353 jumper 275 junction 97, 233, 235, 261, 336 justified 222, 335, 347 Kapustin 40–43, 46 Kazakhstan 41, 43, 46, 48, 51–53 Kemtron 227, 284, 286

Keysight 310 KeyTek 313 Khariton 23–27, 32 Kharkov 11, 39, 44, 90, 91, 306 Khartron 44, 91 Khlopin 13, 17 Kikoin 17, 24, 25 Kirtland 81, 86 Laird 199, 233 laminate 141 Landau 11, 25, 34 lasers 212 lightning 3, 62–65, 71, 91, 150, 152–155, 157, 159, 160, 163, 165, 181, 184, 185, 187, 190, 191, 210, 233, 260, 275, 276, 324, 330, 336, 348 lightning-protection 155 liquid 89, 133, 140 liquid-state 221 Littelfuse 217, 218, 233, 234, 261 Loborev 47–49, 51–53 loop 159, 163 LORA 58, 59 loss-of-function 303 low-altitude 53, 67 low-amplitude 70 low-current 257 low-density 11 low-frequency 171, 187, 198, 203–206, 208, 230, 267, 360 low-impedance 258 low-power 110, 114, 115, 123, 186, 233, 235, 244, 292, 293 low-resistance 150, 207, 236, 259, 262 magnesite 133, 134 magnesium 133, 134 magnesium–shungite 132 magnetic-field 64, 360 magnetohydrodynamic 63, 352 magnetosphere 63 magnitude 8, 77, 95, 313 malfunction 105, 302 malfunctioning 96, 97, 235 maloperation 302 Metal-Oxide-Semiconductor 100 Metatech 54, 318 Meteolabor 258 Meteolabor-EMP 189

Index | 383

Microwave 142 military 12, 27, 38–40, 44, 48, 49, 53, 56, 60, 64, 85, 88, 90, 95, 160, 162, 166, 182, 189, 206, 207, 210, 215, 222, 225, 253, 305, 321 mineral 81, 133, 134 Minuteman 45 Moscow 24–27, 37, 40, 48, 50, 88, 89, 110, 114, 306 multichannel 234, 265 multicontact 289–291 multicore 123, 142, 153, 182, 197, 198, 203, 205, 210, 213, 214, 216, 259, 272 multi-layer 360 multilayered 133 multilayer-shield 216 multi-level 42, 56 multiplexer 214 Nagasaki 16, 18, 20, 24 NASA 55, 344, 348 Naval 88 Navy 55, 84, 88, 92 Nebraska-Lincoln 134 NEMP 92, 157, 160, 161 neutrons 64 nickel 141, 199 Nilsson 110, 111 noise 154, 163, 179, 181, 184, 186, 187, 194–201, 203, 204, 213, 214, 307, 362 noise-immunity 196 noise-signal 206 non-inductive 261 nonnuclear 222 nuclear-explosion 64, 67, 68 nuclear-weapon 45, 207 Ohmite 236, 262 Omicron 298 Omni-Threat 135 on-board 42, 44 Oppenheimer’s 19 optic 153, 169, 214 optical 47, 169, 212, 213, 254, 264 optimization 334, 335, 340 optocouplers 154 optoelectronic 264, 268 Orange 5, 7 oscillogram 115, 117, 257, 258 oscillograph 119–122, 318

oscilloscope 119, 203, 294, 314 overhead 51, 212, 305, 320, 342, 343, 352, 354 overheating 343 overload 259, 279, 286 overloading 353 overvoltage 53, 61, 62, 103, 110–112, 153, 154, 161, 191, 192, 194–197, 210, 233, 248, 256, 260, 273, 275–277, 300, 355 overvoltage-protection 238 oxygen 241 parachute 41 paradox 259 Parker 199, 227, 284, 286 partial 62, 75, 83, 113, 146, 288, 297, 303, 334 partially 8, 47, 52, 83, 146, 150, 160, 169, 279, 286 patent 132–134, 139, 140, 357 patented 132 penetrate 128, 142, 157, 159, 171, 179, 183, 185, 277, 340 penetrating 145, 179, 186, 221, 265, 289 penetration 161, 180, 184, 187, 217, 227, 267, 284, 290, 314, 354, 355, 357 Petersburg 49, 50, 88 Petrzhak 25 phenomenon 1, 3, 9, 46, 52, 62, 63, 154, 155, 159, 160, 324 Phoenix 356 Picatinny 320, 321 Pickering 292, 294 Pioneer 142 plasma 11, 63, 155, 342, 352 plaster–shungite 133 polarized 64, 83, 84, 156, 352 polymer 91, 133, 226 porous 139, 140 potential 27, 100, 107, 150–153, 155, 157–161, 163, 164, 169, 181, 212, 218, 239, 253, 261, 266, 280, 297, 324, 328, 359 powder 132–134, 139, 141, 143, 144, 146, 269 powdered 277 pulsed 215 pulse-decay 71 pulse-energy 73 pulse-full 69 pulse generated 106 pulse-generator 119 pulse-overvoltage 195, 196, 210

384 | Index

quality 125, 129, 174, 206, 210, 261, 324 quantum 9 quasi 343 quasi-DC 156, 352, 353, 357, 362 quasi-direct 343 quick-response 212

restored 97, 213 restricted-access 23, 24 Reynolds 215 Robert 5, 19 Robert Oppenheimer 3 Rosenberger 175, 177 Rosenbergs 12, 20 RRAFAEL 93 rubber 146, 226, 227, 267, 281

Rabinowitz 53 radar 39, 41, 42, 46, 47, 51, 319 Radasky 53 radiate 62 radiated 300 radiating 64, 84, 87, 299 radiation 32, 41, 62, 64, 110, 221, 267, 296, 300, 336–338, 342 radio-absorbing 138–142, 146 radio-absorption 140 radio-relay 212 Radium 13 Rafael 84, 92, 306 recombination 67 rectifier 239, 267 reed 212, 215, 216, 218–221, 293, 359, 360 reed-switch 219–221, 293–295, 360 reed-switch-based 212, 218, 219 relay 51, 53, 128, 150, 151, 153, 154, 161, 165, 167, 171, 184, 192, 196, 212, 217–219, 221–223, 229, 230, 232, 238, 253, 266, 273–275, 279, 289–296, 298, 299, 306, 317–323, 327, 328, 330, 331, 335, 343, 350, 353, 359–363 relay-protection 298 remote-control 289, 292 replacement 169, 334, 335 residential 88 residual 111, 255, 256, 258 resilience 100, 217, 222, 269, 273, 278, 279, 296, 297, 305, 317, 327, 328 resilience-test 296 resiliency 325 resin 221, 277 resist 138, 328 resistant 323, 326, 346, 360 resistivity 136, 224 resistor 98, 150, 164, 235, 260, 261, 355 restore 10, 96, 97, 240, 340, 353

SAAB 110 Sakharov 26, 29–32, 34 Samara 44 Sandia 54 SARA 55 Saratov 49 Sarov 22, 23 Sary-Shagan 43, 44, 46 satellite 44, 46, 348 saturate 342 saturated 357, 359 saturating 320, 343 saturation 207, 320, 343, 346, 351, 353, 356, 362 SAU 42 SCADA 95, 265, 317–320, 323 SCADA-system 323 scattered 9, 46, 62 scattering 9, 11, 60, 62, 307 Schneider 224, 225 Schottky 98 screen 110, 142–144, 149, 165, 171–176, 178, 214, 229, 231, 280, 281, 337 screened 174, 214, 238, 280, 288 screening 154, 164, 165, 238, 280, 281 sealed 218, 219, 229, 245 sealed-contact-reed 212 sealing 288 secrecy 13 secret 3, 12, 14, 15, 20, 22, 24, 41, 53, 64, 66 secretly 59 secrets 12, 38 SEL 320–322, 324, 329–331, 356 self-destruction 42 self-explanatory 192 self-propelled 57 self-recov-erable 303 semiconducting 75, 97 semiconductor 61, 97–100, 103, 105 Semicron 361

pulse-protective 119 pulse-suppression 234

Index | 385

Semipalatinsk 27, 33, 34 Shanghai 313 shield 55, 56, 107, 146, 160, 178–187, 212, 214, 251, 265, 267, 327, 337, 360 shielded 3, 108, 133, 138, 164, 171, 173, 174, 178, 183, 214, 229, 235, 267, 288, 299 shielded cables 300 shielding 73, 129, 131–137, 142, 145, 171–180, 183, 186, 189, 214, 224–227, 229, 284, 286, 311, 312, 325, 327, 328, 336, 339, 340, 360 short-circuit 51, 152, 354, 355 short-circuited 260, 280 short-circuiting 98, 280 short-current 51 shorted 288–290 shortened 288–291, 318 shunt 180, 219 shunting 354, 355 shutdown 96, 347, 361 Sidac 255 Siemens 350, 357 signal ports 326 signals 44, 46, 95, 119, 137, 153, 154, 175, 195, 198, 212, 213, 216, 219, 229, 230, 254, 255, 257, 265, 286, 298, 299, 302, 351 silica 134, 224 silver 151 simulate 260 simulated 160, 196 simulating 203, 297 simulation 84, 298 simulator 79, 80, 82–87, 89–93, 160, 296, 298, 304, 323 soil 157, 159, 160, 181, 350 Solany 227 solar 63, 342–344, 346, 348–350, 354, 358, 359, 362, 363 solid-state 233, 301, 313 spark 79, 162, 301, 306 spark-gaps 301 SPTA 334–337, 340, 341 SPTA-G 335 Stalin 3, 14, 24, 27–29, 33, 34, 36, 40 standard 14, 54, 56, 58–60, 64, 65, 69, 71, 73, 75–77, 81, 82, 96, 107, 111, 112, 117, 119, 123, 129, 136, 138, 141, 142, 145, 153, 160, 162, 163, 176, 178, 182, 189–193, 195–197, 203, 210, 215, 224, 234, 238, 240–242,

253, 255, 259, 260, 273, 275, 277–279, 282, 288, 291, 292, 296, 297, 300, 301, 303, 304, 306–308, 313, 314, 317, 319, 321, 323–327, 329–331, 336, 340, 347, 349, 351 Standex–Meder 294, 295 Starfish 9, 64 static 117, 162, 164, 181 substation 53, 110, 128, 142, 161, 171, 212, 222, 225, 238, 242–245, 250, 251, 269, 279, 288, 318, 319, 325, 326, 328, 350 super-emp 64, 222 super-fast 61 suppression 189, 192, 199, 203, 229, 233, 269, 273, 276 suppressor 117, 119, 121, 193, 196, 216, 233, 235, 236, 249, 257–259, 261 surge 116, 123, 210, 217, 219, 238, 248, 255, 256, 288, 360, 362 switches 165–167, 212, 215, 218, 220, 273–275 switching 51, 100, 181, 191, 215, 219, 220, 233, 250, 257, 258, 267, 292–294, 301, 306, 313, 353–355, 360 taconite 134 tactical 56, 57, 59 Tamm 29, 31, 32 tank 37, 306 tap-changer 353 tap-changing 353 Taran 44, 45 TDK 199 Teak 5 Tektronix 310 Telecom 92, 257 telemetry 41, 42, 253, 265 Teller 30, 32 TEMPS 93 terrorism 241 TESEQ 301, 313 tests 7, 9, 33, 42–44, 46, 47, 51, 53, 60, 79, 90, 118, 125, 189, 190, 192, 193, 204, 205, 218, 253, 259, 293, 296–298, 304, 317, 318, 320, 323, 328, 329, 331 thermonuclear 5, 9, 26, 29–35, 44, 45, 57 Thompson 9, 62 thyristor 215–219, 239, 255, 256, 355, 360, 362 thyristor-based 246 transducer 42, 241 transients 159, 161, 260, 325

386 | Index

transient-voltage 196, 217, 259 transistor 96–100, 104, 106, 247, 248, 313 transmitter 137, 212, 311, 312 transmitting 212, 213, 253 trigger 300, 320, 354 triggered 9, 51, 301, 313, 353, 359, 363 triggering 51, 359 Trinity 1, 2, 4, 27 trip 215, 216, 219, 360, 361 tubes 3, 79–81, 95, 155, 175, 177, 191, 227, 255–260, 324 TVS 119–121, 123, 192, 216, 217, 255, 258–262 TVS-diode 114–117, 119, 120, 123–125, 196, 197, 210, 216, 217, 233–235, 240, 255, 256, 258, 260, 261, 267 twisted 153, 174, 176, 183, 261–263, 267 twisted-pair 153, 214, 267, 360 Tyco 291 Tyco Electronics 290 U-bomb 13 Ukraine 88, 90, 92, 306 Ulam 32 Ulam–Teller 32–34 upset 97, 99–101, 106, 107, 328 Urals 35 uranium 12–17, 20, 24, 30 vacuum 3, 95, 219, 220, 301, 313, 355 vacuum-gap 313 vacuum-sealed 293 variability 108 variable 247

varistor 111–117, 119–126, 161, 191, 192, 196, 197, 207, 210, 215–217, 239, 240, 244, 246, 248, 249, 255, 256, 267, 275–277, 288, 291, 355, 360 vulnerability 95, 96, 99, 104, 107, 145, 253, 315 vulnerable 242, 268 war 1, 9, 18, 27 waveform 123, 159 waveguide 82, 227, 228, 284, 285 waveguide-below-cutoff 227, 228 wavelength 9, 46, 62, 63, 184, 227 weapon 1, 3, 9, 33–35, 40, 48, 49, 53, 54, 59, 65, 66, 81, 85, 253, 269, 273 weapon-grade 9 X-radiation 60, 63, 66, 67 X-ray 3, 9, 10, 60, 62, 64, 352 Yangel 44 Yeltsin 52 yield 60, 67–70, 75, 76 Zeldovich 9, 25, 32, 34 Zener 123 Zenith 49 zero-potential 150, 276, 324 zero-value 311 Zig-Zag 357 zinc 141, 199, 224, 226 zinc-oxide 191, 196, 216, 288 Zlatoust 89 Znamensk 40 Znamensky 48