How Green are Electric or Hydrogen-Powered Cars?: Assessing GHG Emissions of Traffic in Spain [1 ed.] 3319324330, 9783319324333

Citation preview

SPRINGER BRIEFS IN ENERGY

Jesús Montoya Sánchez de Pablo María Miravalles López Antoine Bret

How Green are Electric or Hydrogen-Powered Cars? Assessing GHG Emissions of Traffic in Spain 123

SpringerBriefs in Energy

More information about this series at http://www.springer.com/series/8903

Jesús Montoya Sánchez de Pablo María Miravalles López Antoine Bret •

How Green are Electric or Hydrogen-Powered Cars? Assessing GHG Emissions of Traffic in Spain

123

Antoine Bret ETSI Industriales Universidad de Castilla-La Mancha Ciudad Real Spain

Jesús Montoya Sánchez de Pablo Universidad de Castilla-La Mancha Ciudad Real Spain María Miravalles López Universidad de Castilla-La Mancha Ciudad Real Spain

ISSN 2191-5520 SpringerBriefs in Energy ISBN 978-3-319-32433-3 DOI 10.1007/978-3-319-32434-0

ISSN 2191-5539

(electronic)

ISBN 978-3-319-32434-0

(eBook)

Library of Congress Control Number: 2016936581 © The Author(s) 2016 This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, express or implied, with respect to the material contained herein or for any errors or omissions that may have been made. Printed on acid-free paper This Springer imprint is published by Springer Nature The registered company is Springer International Publishing AG Switzerland

Contents

1 Introduction . . . . . 1.1 Introduction . . 1.2 What We Do. . 1.3 Why We Do It 1.4 How We Do It References . . . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

1 1 2 4 4 5

2 Current Figures . . . . . . . . 2.1 Energy to Wheels . . . 2.2 Current Emissions . . . 2.3 Emission Coefficients . References . . . . . . . . . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

7 7 9 10 12

3 The Hydrogen Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1 Potential and Hazards of the Hydrogen Model. . . . . . . . . . . . 3.1.1 Hydrogen Production . . . . . . . . . . . . . . . . . . . . . . . . 3.1.2 Hydrogen Storage . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1.3 The Fuel Cell . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1.4 Hydrogen in Spain . . . . . . . . . . . . . . . . . . . . . . . . . 3.2 The Hydrogen Vehicle . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2.1 The Fuel Cell Electric Vehicle . . . . . . . . . . . . . . . . . 3.2.2 Elements of the Fuel Cell Electric Vehicle . . . . . . . . . 3.3 Hydrogen Required . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.3.1 Hydrogen Versus Internal Combustion Engine Vehicle. 3.4 Energy Scenarios for CO2 Equivalent Calculation . . . . . . . . . 3.4.1 Definition of the Scenarios . . . . . . . . . . . . . . . . . . . . 3.4.2 Electricity Generation. . . . . . . . . . . . . . . . . . . . . . . . 3.4.3 Emissions Model Simplifications . . . . . . . . . . . . . . . . 3.4.4 Scenario 1. Current Hydrogen Generation Mix . . . . . . 3.4.5 Scenario 2. Home Generation of Hydrogen. . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

13 13 14 18 21 21 25 25 26 28 29 30 30 32 33 35 40

v

vi

Contents

3.4.6 Scenario 3. IEA Scenario for Short-Term Implementation . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.4.7 Scenario 4. IEA Scenario for Mid-Term Implantation. . 3.4.8 Scenario 5. IEA Scenario for Long-Term Implantation . 3.4.9 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.5 Sensitivity Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . .

. . . . . .

. . . . . .

43 45 47 49 51 53

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

55 55 56 56 58 58 60 60 62 65 65 65

........ ........ ........

67 68 70

5 Conclusion. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

71 74

Appendix A . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

75

4 The Electrical Model . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.1 The Electrical Vehicle . . . . . . . . . . . . . . . . . . . . . . . 4.1.1 Pros and Cons of Electrical Vehicles . . . . . . . . 4.1.2 Batteries for EV . . . . . . . . . . . . . . . . . . . . . . 4.2 Electrical Energy Required . . . . . . . . . . . . . . . . . . . . 4.2.1 Electrical Versus Internal Combustion Vehicles. 4.3 GHG Emissions. . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.3.1 Electricity Production Scenario . . . . . . . . . . . . 4.4 Fossil Fuels Threshold: Scenario “x” . . . . . . . . . . . . . 4.5 Impact on the Spanish Electricity System . . . . . . . . . . 4.5.1 Electricity Demand . . . . . . . . . . . . . . . . . . . . 4.5.2 Daily Load Curve . . . . . . . . . . . . . . . . . . . . . 4.5.3 Installed Capacity in Spain and “Future Mix” Scenario . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.6 Sensitivity Analysis . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

Chapter 1

Introduction

1.1 Introduction A multitude of studies about climate and resources during the last decades have concluded that the current energetic model, based on fossil fuels, is unsustainable in the long run. The most immediate reason is the climate change induced by the massive use of fossil fuels, and the less immediate reason (but certainly close), the future exhaustion of fossil resources [1, 2]. In this context, several alternatives have been proposed to replace the current energetic model. These alternatives are based on the direct or indirect employment of renewable resources, such as biomass, solar or wind energy, wave power or hydropower. The integration of alternative energies in the energetic model of a country, and a better usage of resources, is usually known as sustainable development. It constitutes a goal for every company or government. Regarding road traffic, a sector with an important contribution to greenhouse gas emissions and primary energy consumption, the transition to cleaner models typically contemplates the use of new engines running with liquefied petroleum gas or syngas,1 electric engines powered by batteries, or hydrogen fuel cells. In the case of battery-powered vehicles, we will talk about the “Electrical Model.” Regarding hydrogen fuel cells car, we will talk about the “Hydrogen Model.” As the main energy institutions contemplate, the main advantage of these models is the possibility of designing a completely clean energetic scenario. Because the energy content per unit volume of oil is larger than that of batteries or fuel cells, one could wonder whether such vehicles could efficiently substitute current ones. Indeed, with nowadays technology, an electrical or hydrogen powered family car with nearly 1,000 km autonomy remains out of reach. Yet, one does not need to cover 1,000 km every day. Figure 1.1 shows a 2009 statistical analysis of the distances traveled per day in USA. Noteworthily, 99 % are shorter than 200 km (and the mean distance is certainly shorter in Spain than in the US). In other words, one really needs a car autonomy longer than 200 km only 1 out of 100 days. This is the reason why current electric 1 Syngas

is the gas obtained after gasification of coal, biomass or wastes.

© The Author(s) 2016 J. Montoya Sánchez de Pablo et al., How Green are Electric or Hydrogen-Powered Cars?, SpringerBriefs in Energy, DOI 10.1007/978-3-319-32434-0_1

1

2

1 Introduction

Fig. 1.1 Statistical analysis of the distances traveled per day in USA. About 21 % of the daily rides are between 0 and 10 miles. 23 % between 20 and 10. 17 % between 30 and 20. And so on. 99 % are shorter than 200 km (125 miles) [3]

or hydrogen cars, with about 200 km of autonomy, could very well replace current ones. By charging it at night, an electric or hydrogen car could almost seamlessly replace a current one.

1.2 What We Do This study aims at analyzing the effect of a global change in the energetic model associated to Spanish traffic, from a conventional fossil model to the electrical or hydrogen models. This study is mainly focused on the impact on greenhouse gas (GHG) emissions. Therefore, the question contemplated is the following: If all private vehicles in Spain were replaced by electrical batteries, or hydrogen fuel cells, vehicles, how lower would greenhouse gas emissions be ? Is the big effort required for the change really worth it, or would environmental concerns still remain?

The main objective of this project is to provide a technical overview of the elements involved in both energetic models, and estimate the rise or reduction in GHG emissions depending on how these models are configured. In order to define the “equivalent” electric, or hydrogen-driven traffic, our strategy consists in determining the energy currently provided at wheels level. The equivalent traffic is then defined as the electric, or hydrogen-driven one, providing the same energy at wheels. The knowledge of the energy at wheels allows to determine the amount of electricity and/or hydrogen needed to power it. Several scenarios are considered in this respect. The GHG emissions of each scenarios are finally compared to the current ones. The resulting algorithm is pictured on Fig. 1.2, and can be described as follow: • Evaluation of the energy currently provided at wheels level in the current Spanish private cars traffic (Chap. 2).

1.2 What We Do

3

• Evaluation of the GHG emissions from the current Spanish private cars traffic (Chap. 2). • Determination of the extra electricity, or hydrogen, needed to provide the same energy to wheel (Chap. 4 for electricity, 3 for hydrogen). • Determination of the GHG emissions resulting from the various electricity, or hydrogen, production scenarios (Chap. 4 for electricity, Chap. 3 for hydrogen). • Our results rely on many parameters (efficiencies, emissions coefficients, etc.). In order to assess their sensitivity to some small variations of these parameters, a sensitivity analysis is carried out in Sect. 3.5 for the hydrogen model, and in Sect. 4.6 for the electrical one. These are the most technical sections of this book. Because the analysis of the electrical model is relatively simpler than that of the hydrogen one (there are far more ways to generate hydrogen than electricity), we could perform two additional steps for the former: • Estimate of the critical amount of non-GHG emitting sources in the electrical mix, for the switch to the electrical model to bring less emissions (Sect. 4.4). • Potential impact of traffic electrification on the Spanish electrical system (Sect. 4.5).

Fig. 1.2 Strategy used to assess the profitability of the electrical and hydrogen traffic models, with respect to GHG emissions. The additional analysis performed in Sects. 4.4 and 4.5 for the electrical model are not represented

4

1 Introduction

1.3 Why We Do It There are two main motivations to carry out the evaluation of the environmental impact of both models. On the one hand, we can check the information provided by some organizations and mass media about them, as possible solutions to climate change mitigation for traffic: how really green are electric of hydrogen cars? On the other hand, we can determine if it is actually possible to implement this model in Spain and which would be the most proper configuration. The hydrogen economy, for example, is considered by many as the solution to uncontrolled GHG emissions to the atmosphere and the end of the energetic dependence. This idea was introduced by Jeremy Rifkin’s The Hydrogen Economy [4]. Mass media and automotive manufacturers usually consider fuel cell vehicles are clean vehicles (or “zero emissions” vehicles), because in a fuel cell, hydrogen is converted into heat, electricity, and water, without any emissions but water vapor. But there are no natural sources of hydrogen on earth. There are oil wells, gas fields, coal mines, but no hydrogen wells. Therefore, the hydrogen powering the cars would have to be generated from primary energy sources, and part of these sources could very well emit GHG. In theory, the hydrogen economy can be sustained by completely clean, renewable sources. If this model is properly configured, GHG emissions and the energetic dependence could be significantly reduced. Nonetheless, the traffic needs could prove too large to remain independent from fossil fuels (a question which is out of the scope of this book). The electrical model could equally offer a fully non-GHG emitting solution for traffic, again provided only nonemitting energy sources are used to produce the required electricity. One of its main advantages is that most of the infrastructures are already in place. While an hydrogen grid would have to get started from scratch, the electricity grid is already there, even if could would need some adjustment [5].

1.4 How We Do It This study has been conducted from data related to the Spanish energy demand, greenhouse gas emissions, driving trends, and energy use. Therefore, it could also be valid for countries with similar operative conditions or where the status of the hydrogen economy, or that of the electrical model, is similar, that is, has not started yet. In those cases, the data could be adapted and conclusions studied afterwards (more on this in the conclusion). To be as realistic as possible, the data used in this study have been extracted from recent institutional databases. The environmental impact is studied with data from the United Nations Framework Convention on Climate Change (UNFCC). Energy requirements are studied with data from the Spanish Instituto para la Diversificación y Ahorro de la Energía (IDAE). Data related to electricity infrastructure, production, and demand are provided by the Red Eléctrica de España database (REE).

1.4 How We Do It

5

This study is also limited to private cars, as other kind of vehicles are defined by different driving trends. Nonetheless, further studies could use the present methodology to integrate all types of road vehicles, as well as other means of transportation, to provide an overall assessment for the transport sector. Besides Sects. 3.5 and 4.6, devoted to a sensitivity analysis of our results, the maths implemented in this book do not go beyond simple arithmetic. Calculations are explained in detail so that the reader may straightforwardly reproduce them. By virtue of their common objective, the two main parts of this work share many common features. Most of them have been gathered in Chap. 2. Regardless of the topic the reader is interested in, this chapter is therefore a must-read. Having said that, Chap. 3 on the hydrogen model, or Chap. 4 on the electrical one, can be read separately. Note that the latter is significantly simpler than the former, mainly because the hydrogen production scenario contemplated are more heterogenous than the electricity generation scenario. Some details about the notations and units used throughout this book are given in Appendix A.1. Though we do not use abbreviations in a systematic manner, the reader lost in notation can always refer to the listing on the same Appendix, p. 75.

References 1. IPCC. Climate Change 2013 - The Physical Science Basis (Cambridge University Press, Cambridge, 2014) 2. A. Bret, The Energy-Climate Continuum: Lessons from Basic Science and History (Springer, Heidelberg, 2014) 3. J. Krumm, How people use their vehicles: Statistics from the 2009 national household travel survey. Technical report, SAE Technical Paper 2012-01-0489, 2012 4. J. Rifkin, The Hydrogen Economy (Polity Press, Cambridge, 2002) 5. S. Kabisch, A. Schmitt, M. Winter, J. Heuer, Interconnections and communications of electric vehicles and smart grids. In First IEEE International Conference on Smart Grid Communications (SmartGridComm), pp. 161–166, Oct 2010

Chapter 2

Current Figures

In order to assess the benefits of our two models in terms of GHG emissions, we need to determine the emissions of the current, fossil fuels based, model. Similarly, the energy to wheel of the current model is required if we are to compute the needed electricity or hydrogen to power the equivalent traffic. The GHG emissions of the electrical or hydrogen models comes in part, if not all, from electricity generation. Assessing the emissions of both models requires therefore the knowledge of the so-called “emissions coefficients.” These numbers are simply the amount of GHG emitted when generating 1 Watt-hour of electricity from any given primary energy sources. They are common to both models and are therefore discussed in this chapter.

2.1 Energy to Wheels According to the “Instituto Nacional de la Estadistica” (ine.es), there were 10,510,112 gasoline cars in Spain in 2011, and 11,763,255 diesel ones. The same source also gives the average amount of kilometers covered by each kind of cars: 10,486 for gasoline and 14,466 for diesel cars. The “Instituto para la Diversificación de Ahorro y Energía” runs a database1 gathering technical information on private cars sold in Spain. From there, we could compute their average consumption. The result is 6.55L/100 km for gasoline, and 5.53L/100 km for diesel cars. These numbers allow to derive the total amount of energy provided to Spanish private cars in 2011. Considering an energy content of 32.18 MJ/L for gasoline and 35.86 MJ/L for diesel fuels [1], the total amount of energy provided to the tanks was,

1 http://coches.idae.es/portal/BaseDatos/MarcaModelo.aspx.

© The Author(s) 2016 J. Montoya Sánchez de Pablo et al., How Green are Electric or Hydrogen-Powered Cars?, SpringerBriefs in Energy, DOI 10.1007/978-3-319-32434-0_2

7

8

2 Current Figures

6.55 × 32.18 = 2.32 × 1011 MJ, 100 5.53 = 11,763,255 × 14,466 × × 35.86 = 3.37 × 1011 MJ. 100

E t,gas = 10,510,112 × 10,486 × E t,die

We now need to derive from these numbers the amount of energy at wheels level. To do so, we simply account for the various efficiencies from the tank to the wheels. These stem from the engine losses, and then from the mechanical losses between the engine and the wheels. For gasoline engines, we consider an efficiency of 35 %. For diesel engines, an efficiency of 44 % [2]. In each case, the mechanical efficiency is taken as 80 % [3]. The chains of efficiencies are pictured in Figs. 2.1 and 2.2, for gasoline and diesel vehicles respectively, resulting eventually in a 28 % Tank-to-Wheels efficiency for gasoline, and 35.2 % for diesel. Hence, the total amount of energy to wheels is, E w,gas = 0.35 × 0.8 E t,gas = 6.5 × 1010 MJ, E w,die = 0.44 × 0.8 E t,die = 1.19 × 1011 MJ, E w = E w,gas + E w,die = 1.83 × 1011 MJ = 51.06 TWh.

(2.1)

Therefore, if the Spanish private car traffic was to be electrified or “hydrogenized,” the energy provided to wheels would have to be 51.06 TWh in order for the traffic to be equivalent to the 2011 one. This number represents 20 % of the 260 TWh consumed in the country in 2013 [4]. Table 2.1 summarizes the results of this section.

Fig. 2.1 Tank-to-Wheels efficiency for gasoline vehicles

Fig. 2.2 Tank-to-Wheels efficiency for diesel vehicles

2.2 Current Emissions

9

Table 2.1 Current state of the Spanish private cars traffic Number km/year L/100 km Gasoline Diesel

10,510,112 11,763,255

10,486 14,466

6.55 5.53

E tank (MJ)

E wheel (MJ)

2.32 × 1011

6.5 × 1010 1.19 × 1011

3.37 × 1011

2.2 Current Emissions Since we want to assess how lower, or not, GHG emissions would be within some alternative models of traffic, we need to determine the emissions of the current, fossil fuel-driven traffic. For Spain in the year 2011, the United Nations Framework Convention on Climate Change (UNFCCC, unfccc.int—Flexible GHG data queries) provides the following numbers, • Total GHG emissions, road transportation, diesel oil: 63.52 Tg CO2 eq. • Total GHG emissions, road transportation, gasoline: 15.99 Tg CO2 eq. Unfortunately, data are not split by kind of transportation. The numbers above include motorbikes’ and trucks’ emissions as well. We shall thus compute private cars emissions, and contrast with the aggregated UNFCCC numbers for a coherence check. For gasoline cars, GHG emissions amount to 2.36 kg CO2 eq/L. For diesel fuel, it is almost the same, namely, 2.63 kg CO2 eq/L [1]. Given the number of cars and kilometers reported in Table 2.1 for each kind of vehicle, we find, • Total GHG emissions, private cars, diesel oil: 17.05 Tg CO2 eq, • Total GHG emissions, private cars, gasoline: 24.72 Tg CO2 eq. Hence, Total GHG emissions, private cars = 41.77 Tg CO2 eq.

(2.2)

This figure is coherent with the all-road transportation emissions given by the UNFCCC. It will be our reference for comparison with the coming hydrogen and electrical scenarios. Noteworthily, the previous amount of GHG emissions per liter of diesel, or gasoline, makes perfect sense. Let us compute them from first principles. Considering roughly that 1 L of gasoline holds 1 kg of carbon, we find its combustion should release 1 kg of carbon atoms. And since combustion pairs up two oxygen atoms with one carbon atom, the resulting molecule of CO2 weights 3.66 times as much as its carbon atom.2 We therefore come to the rule of thumb that burning 1 L of gasoline should emit 3.66 kg of CO2 , which is not far from the actual figures used above (for an even more accurate result, one can consider that 1 L on gasoline holds only 0.7 kg of carbon). 21

mole of C is 12 grams. 1 mole of O is 16 grams. The 3.66 factor comes from (12 + 2 × 16)/12.

10

2 Current Figures

2.3 Emission Coefficients Within the electrical model, GHG are emitted when generating the extra electricity needed to run the electric cars. Within the hydrogen model, GHG are emitted when generating the hydrogen required. As we shall see in Chap. 3, there are many ways to do so, and many of them involve electricity. One common point of our two models is therefore that both contemplate GHG emissions through electricity generation. In turn, electricity can be generated from fossil fuels, nuclear, wind, etc., each option emitting more or less GHG. The amount of GHG emitted per Watt-hour generated is called the “emission coefficient,” typically given in grams of CO2 equivalent per kWh, or Teragrams per TWh. The emission coefficients ei are pictured in Fig. 2.3 and given in Table 2.2. They are the outcome of a large-scale literature review performed on the topic in Ref. [5] and are expressed in terms of their distribution with minimum, 25th percentile value, average, 75th percentile value and maximum. Note that although they do not directly emit CO2 when generating electricity, sun, hydro, biomass, wind, or nuclear energy are not strictly attributed zero emissions per TWh. The main reason why is that these numbers account for the full life cycle of the technologies considered. For example, it takes concrete and energy to build a

Fig. 2.3 Estimated GHG emissions coefficients throughout the life cycle (g CO2 eq/kWh) for different power generation technologies. “CCS = Carbon capture and storage” (Figure SPM.8 from [5], p. 19)

2.3 Emission Coefficients

11

Table 2.2 GHG emissions coefficients ei in Tg CO2 eq/TWh ([5], p. 982) Min 25th Average 75th percentile percentile Coal Oil Gas Sun Hydro Biomass Wind Nuclear

0.67 0.51 0.29 1.2 × 10−2 0 0 2 × 10−3 1 × 10−3

0.87 0.722 0.42 4.3 × 10−2 3 × 10−3 0 8 × 10−2 8 × 10−2

1 0.84 0.47 6.8 × 10−2 4 × 10−3 6.8 × 10−3 1.2 × 10−2 1.6 × 10−2

1.13 0.9 0.54 0.11 7 × 10−3 6 × 10−3 2 × 10−2 4.5 × 10−2

Max 1.69 1.17 0.93 0.3 4.3 × 10−2 6.08 × 10−2 8.1 × 10−2 0.22

dam or a nuclear power plant, and concrete production is a notorious GHG emitter.3 It also takes energy to build solar cells or wind mills, etc. We therefore find that once all operations related to electricity production have been accounted for, even “non directly GHG-emitting technologies,” do emit some. It is nevertheless a fact that looking at the average values of the emissions coefficients in Table 2.2, one finds that such non directly GHG-emitting technologies, emit between 15 and 250 less GHG than coal. Besides the non-zero value of the emissions coefficients for sun, hydro, biomass, wind and nuclear, their spread may also intrigue the reader. For example, we find a factor 25 for sun between the “max” and the “min” columns. Why isn’t there one single column in Table 2.2? Mainly because when it comes to accounting for every possible emissions during the life cycle of a dam, for example, all authors don’t always consider the same boundaries for their calculations. For example, it takes concrete to build a dam, but should we consider the emissions of the trucks carrying the concrete, on the top of the emissions generated by the concrete manufacturing itself? If yes, should we then account for the GHG emitted when building these trucks? And so on. The evaluation of the emissions resulting from the generation of the same Watt-hour may thus vary with the number of secondary processes accounted for. In spite of such disparities, the coefficients extracted from many studies (296!) cluster around some average numbers. Yet, we chose to translate their spread into our calculations by presenting our scenario’ emissions for each column of Table 2.2. Finally, it is interesting to recognize that on a purely physical ground, the numbers featured in Table 2.2 for fossil fuels make sense. Let us compute, starting with oil, their order of magnitude:

3 Nearly

0.8 tons of CO2 per ton of cement produced. See [6], p. 758, 1345.

12

2 Current Figures

• 1 L of oil contains about 42 MJ, that is, 11.6 kWh. • If the energy content of 1 L of oil could be perfectly translated into electricity, then 1 L of oil would generate 11.6 kWh of electricity. Assuming this is done with a 50 % efficiency, we find 1 L of oil generates 5.8 kWh of electricity. • How much carbon do we emit in the process? Oil is mainly carbon. Let us simply consider that burning 1 L of oil releases 1 kg of carbon. • We should therefore emit 1 kg of carbon to generate 5.8 kWh of electricity from oil. That is, 1000/5.8 = 171 grams of carbon per kWh. • Finally, and as noted above, each atom of carbon catches 2 oxygen atoms to form a CO2 during the combustion. The molecular weight of CO2 is 3.66 times that of carbon. We thus need to multiply our 171 grams of carbon by 3.66 to find the amount of CO2 emitted. • We finally get to 171 × 3.66 = 627 g CO2 /kWh, or 0.63 Tg CO2 /TWh, perfectly in line with the numbers presented. As expected, emission coefficients for gas or coal are similar, order of magnitude wise, since their energy content is comparable. But when it comes to the other coefficients, like nuclear, solar, or wind, such back-of-the-envelope calculations are not performed so easily, because now, emissions only come from indirect processes such as power plant building, solar cells manufacturing, and so on.

References 1. R. Edwards, J.-F. Larivé, J.-C. Beziat, Well-to-wheels analysis of future automotive fuels and powertrains in the european context. Technical Report Version 3c, European Commission Joint Research Centre, July 2011 2. F. Payri, J.M. Desantes, Motores de combustión interna alternativos (Reverté, 2011) 3. E.J. Domínguez, J. Ferrer, Sistemas de transmisión y frenado Ciclos Formativos (Editorial Editex, Pozuelo De Alarcon, 2012) 4. REE. The spanish electricity system 2013. Technical report, Red Eléctrica de España, 2014 5. O. Edenhofer et al., Renewable Energy Sources and Climate Change Mitigation: Special Report of the Intergovernmental Panel on Climate Change (Cambridge University Press, (Cambridge, 2011) 6. IPCC. Climate Change 2014: Mitigation of Climate Change (Cambridge University Press, Cambridge, 2015)

Chapter 3

The Hydrogen Model

3.1 Potential and Hazards of the Hydrogen Model About two-thirds of the primary energy today is used directly for transportation or as heating fuels [1]. As discussed in the introduction, a change in the global energetic model is necessary to stop climate change. The usage of alternative energy sources and nonfossil-based fuels is the key to the achievement of a sustainable model with reduced emissions of pollutants and greenhouse gases. In this environment of changes, several options have been suggested depending on the energy sector considered: alternative energies (wind power, solar power and photovoltaic, geothermal, hydropower, biomass and wave power) and nuclear power for electricity generation. Electric, biofuel, and fuel cell vehicles for transportation. In this context, a hydrogen-based model, or “Hydrogen Economy,” offers potential benefits in terms of emissions and diversification of primary energy sources, as it can encompass all the primary energy generation methods in hydrogen production and allows designing a near-zero emissions energetic model, provided hydrogen production is carbon free. Hydrogen is not an energy source but an energy carrier, and it needs to be obtained, processed, and delivered before usage. The feasibility of the hydrogen economy has been technically demonstrated as hydrogen vehicles, heating, and power systems are technically available. But it needs further development for being viable. As will be checked in this chapter, indirect emissions could be as high as current ones if the energetic model is not properly designed. Hydrogen production for automotive could compete for alternative resources with conventional electricity needs, and fresh water resources could be overused [2]. The main challenges of the hydrogen model are the development of cheaper materials and manufacturing processes for fuel cell systems (about $30/kW for automotive and $1000/kW for stationary power generation systems), the improvement of the reliability and durability of the cells, the improvement of the power density

© The Author(s) 2016 J. Montoya Sánchez de Pablo et al., How Green are Electric or Hydrogen-Powered Cars?, SpringerBriefs in Energy, DOI 10.1007/978-3-319-32434-0_3

13

14

3 The Hydrogen Model

and of the specific power of the storage options, and the large-scale implementation of generation and delivery technologies for pure hydrogen. Secondary elements, like sensors and online control systems, also need to be improved [2]. Current requirements are fuel cells which withstand load cycling and strong environmental swings—even through the freezing point—with acceptable levels of degradation. This implies lifetimes over 5500 h for automotive applications (equivalent to 425,000 miles at 50 km/h) and over 40,000 h of steady operation under changing external temperature conditions for stationary fuel cells. Furthermore, for automotive applications, power densities higher than 650 W/kg are required to avoid overweight. A minimum power density is also needed for auxiliary and portable applications (150 and 100 W/kg, respectively) [2].

3.1.1 Hydrogen Production As an energy carrier, hydrogen needs to be generated from natural resources, renewable or not. As will be discussed soon, hydrogen vehicle emissions only consist in water vapor. Therefore, carbon dioxide and other pollutant emissions only depend on hydrogen production, delivery, and storage methods. Chemical Hydrogen Production Chemical processes use liquid- or gas-phase carbon-based fuels to obtain a hydrogencontaining stream through chemical reactions. Two important parameters are involved in the design of the process, and define the chemical reaction: the air–fuel ratio and the hydrocarbon used as hydrogen source. A major advantage of using carbon-based fuels to obtain hydrogen is the cost effectiveness and the compactness of the plants [2], as well as an high-energetic conversion yield, even higher than any other hydrogen production methods. However, the finite nature of fossil resources determine the main limitation of this kind of processes. In most chemical processes, carbon monoxide (CO) is obtained as a byproduct, together with the hydrogen in the output stream. CO is undesired due to its toxicity and the performance reduction of the subsequent fuel cell (it can seriously harm low-temperature fuel cells [2]). Therefore, in most cases, an additional process is performed to oxidize carbon monoxide into carbon dioxide in a exothermic reaction. Carbon dioxide can be sequestered from the product stream using specific equipments to reduce the environmental impact of the process. This process is known as carbon capture and storage (from now on, CCS) and in some cases it can reduce CO2 emissions down to 80–90 %.1 In order to avoid undesirable corrosion problems and efficiency losses, equipments such as desulfurization, washing or heat recovery units, should be installed with the reformer and the water vaporization unit. 1 See

http://www3.epa.gov/climatechange/ccs/.

3.1 Potential and Hazards of the Hydrogen Model

15

The most important chemical processes for obtaining hydrogen are steam reformation, partial oxidation, autothermal reforming and gasification [2]. Steam Reformation In this process, water and fuel are mixed to produce hydrogen, carbon dioxide, and other minor species such as CO, in an endothermic reaction as follows:  y H2 C x Hy + (2x)H2 O + heat → xC O2 + 2x + 2 This is the chemical process producing the highest hydrogen mole fraction in the output stream, reaching 80 % of H2 on a dry basis. Among all the reformable fuels, methanol is the simplest liquid-based one, due to the lack of carbon–carbon bond and low heat input required (reformation temperature is around 200 ◦ C [2]). The possibility of obtaining methanol from biomass and other biological processes allows for the designing of carbon-free life cycles for methanol fuel cells. However, methanol cannot be found free in nature, so the hydrogen industry is currently using fossil fuels in reforming processes. These fuels are mainly natural gas (48 % of the current global hydrogen production) and oil (30 %) [3]. A larger heat input is required to reach temperatures between 760 ◦ C and 980 ◦ C and induce the reaction. The energetic efficiency of steam reformation processes is relatively high, approaching 80 % [4]. In the case of natural gas, the process is called steam methane reforming (SMR)2 because the reformed gas is methane. It represents a well-known technology. The CO2 emissions of a typical hydrogen production plant by steam reformation, or Natural Gas, are included in Table 3.1. Oil reformation requires its previous cracking, as reformed gases in that case are the Liquefied Petroleum Gases (LPG). The CO2 emissions factor between hydrogen production by LPG reforming emissions, and the internal combustion engine emissions is 0.52 [7]. Partial Oxidation Partial Oxidation (POX) is an exothermic reformation where the input fuel stream is mixed with oxygen or air, to partially oxidize the fuel and produce an hydrogen-rich stream. The process is a rich combustion3 where the lack of oxygen produces a higher CO fraction than in catalyzed steam reformation. The typical POX reaction for a mixture with air is: 1 1 C x Hy + x(O2 + 3.76N2 ) → xC O + y H2 + (1.88x)N2 + heat. 2 2

(3.1)

Due to the high CO fraction in the output stream, the installation of a water gas shift reactor, to convert CO into CO2 , is usually required. Natural gas is also usually used as a fuel for these kinds of processes. 2 Methane molar fraction in imported Natural Gas in Spain typically approaches 90 %, but it depends

on the origin [5]. rich combustion is an oxidation process where the fuel/air ratio is higher than 1, that is, there is less air (or oxygen) in the environment than the necessary to achieve a complete oxidation of the fuel.

3A

16

3 The Hydrogen Model

Table 3.1 CO2 Emissions of a typical hydrogen production plant by SMR [6] Plant Size (kg H2 per stream day) 1,200,000 24,000 Value without CCS (kgCO2 /kgH2 9.22 9.83 Value with CCS (kgCO2 /kgH2 ) 1.53 1.71

480 12.1 –

Autothermal Reformation This process is based on using the heat obtained from partial oxidation processes to induce a steam reformation process and obtain a stream with a high hydrogen molar fraction. The mixed configuration is known as autothermal reforming (ATR). It can achieve hydrogen molar fractions around 40–50 %, intermediate between the one achieved by steam reformation, and the fraction obtained by partial oxidation. Gasification Gasification converts organic or fossil-carbon-rich fuels into carbon monoxide, hydrogen, and carbon dioxide, in a gas stream usually called syngas. This reaction occurs at high temperatures (>700 ◦ C) without combustion, with a limited amount of oxygen, air, and/or steam. High temperatures reduce the formation of char, tars, and phenols and favor carbon conversion to gas. A typical reaction for coal gasification is, Carbon atom (solid) + H2 O + heat → CO + H2

(3.2)

The CO produced in the reaction is usually converted into CO2 in a water–gas shift reactor. This reaction is endothermic, and additional heat is required. The weaknesses of the process, in comparison with natural gas reformation, are a limitation of the large-scale implementation for hydrogen production, and a rise of the cost of the resulting hydrogen. Hydrogen production from coal is commercially mature and coal is abundant in many parts in the world, but environmental advantages would favor the use of biomass in the gasifiers. The efficiency of gasification process depends on the raw material. It reaches 50 % for coal and 60 % for biomass. The GHG emissions of a typical hydrogen production plant from coal can be found in Table 3.2. In the case of biomass, life cycle assessment of emissions should consider both CO2 produced during the gasification process, and the CO2 required for the biomass to grow up through its lifetime. During biomass growing, 219.8 tons of CO2 eq (per TJ of H2 produced) are absorbed by the plants, whereas 221.4 tons of CO2 eq are vented during the biomass-to-hydrogen conversion [9].

Table 3.2 Characteristics of a hydrogen from coal production plant [8] Coal requirements Hydrogen production CO2 eq (tons/TJ H2 (tons/day) (tons/day) produced) 2,700

345

135

CO2 eq w/CCS (tons/TJ H2 produced) 41

3.1 Potential and Hazards of the Hydrogen Model

17

Electrochemical Hydrogen Production This method is based on electrolysis of water using electricity previously generated by another energy source. The main advantage of electrolysis is that the hydrogen produced can be carbon-free, provided the required electricity is nuclear or renewable. Moreover, the hydrogen and the oxygen produced are extremely pure and can therefore be used in low-temperature applications, where minute impurities can drastically reduce performance. Electrolysis is an established method for small-scale local generation of highly pure hydrogen and oxygen. This process is also used, if needed, as local support in larger hydrogen production infrastructures. For large-scale hydrogen generation, electrolysis faces several limitations such as the low efficiency of the conversion process (between 65 and 75 %, based on the Lower Heating Value4 ). This raises the production cost in comparison with natural gas reformation, unless electricity comes from an inexpensive source. Moreover, electrolysis requires a source of freshwater, a globally scarce resource. As many power grids worldwide do not currently have the capacity to supply power for hydrogen production and usual demand, additional electricity infrastructure would be needed [2]. One way to bypass the limitations of electrolysis is to raise the temperature of the process. At 1000 ◦ C, the electrical energy required to split water is considerably less than at 100 ◦ C because the heat released by the Joule effect during the electrolysis process is also used in the H2 electrolysis5 [10]. This enables high-temperature electrolyzers to operate at higher efficiencies and produce hydrogen at higher rates than low-temperature electrolyzers. At this temperature, the reactions in the electrodes are also more reversible and in this conditions, a solid oxide fuel cell (SOFC) can produce an electrolysis reaction. Current attempts are underway to reduce the electricity consumed in the electrolysis process through heat input from geothermal, solar, or natural gas sources [11]. Thermolysis Thermolysis or thermochemical water splitting is the conversion of water into hydrogen and oxygen by a series of thermally driven chemical reactions. When temperature approaches 3000 ◦ C, the thermal agitation of water molecules causes the splitting of 10 % of the water stream. Efficiency can exceed 50 % for this kind of processes and can reduce hydrogen production costs. The main limitation relates to corrosion resistance of the materials at high temperatures, to the complex design of heat exchangers and heat storage equipment, and to the risk associated with high temperatures. Nowadays, this technology is still being developed to achieve commercial feasibility and reduce the operating temperatures 4 The Lower Heating Value (LHV) is the heat released by a determined compound after a combustion

process, without accounting for the heat absorbed by the water during its vaporization (latent heat of vaporization). 5 The balance between the downward in electricity demand by the electrolyser at 1000 ◦ C and the heat required for heating water to this temperature, is positive. It represents an energy saving with respect to electrolysis at 100 ◦ C.

18

3 The Hydrogen Model

of the process [11]. The high temperature required for thermolysis cycles is typically reached using high-temperature solar systems or nuclear reactors. Water Photolysis Photoelectrolysis is the process by which water is directly split into hydrogen and oxygen due to the incidence of light, which typically comes from the sun. Hydrogen production by photolysis is a developing technology included in R&D programs of several countries worldwide. Once developed, it could reduce the cost of electrolytic hydrogen. Laboratory-scale devices have been developed over the last years with solar-to-hydrogen conversion efficiencies up to 16 %. The availability of this technology depends on the development of photoelectrode materials and of high-efficient, corrosion-resistant processing systems [11]. Biological Processes Biological methods are based on biological processes controlled by fermentive or photosynthetic organisms. Urban wastes and other biomass sources are used as inputs to produce hydrogen-containing fuels and reduce garbage storage. The main limitation is the large size of the production plants required to produce a certain amount of hydrogen, compared to other hydrogen production methods, as well as the longer duration of the process. The main biological processes to produce hydrogen are biophotosynthesis of water using algae and cyanobacteria, photodecomposition of organic compounds by photosynthetic bacteria or fermentative bacteria, and hybrid systems using organic compounds and both types of bacteria [2].

3.1.2 Hydrogen Storage Fuel storage is critical when the power source is mobile. The best example may be found in automotive applications, where the fuel tank constitutes the energy reservoir allowing the correct performance of the system. The main storage requirement for the automotive sector is high energy density, since the vehicle driving range (distance traveled by a vehicle with a full tank) needs to be as high as possible. Ideally, hydrogen storage would be performed at atmospheric pressure and room temperature. But in such conditions, hydrogen is under gas phase and 1 kg of it occupies some 12 m3 . Hydrogen does enjoy a high-energy content per weight (hydrogen holds 120 MJ/kg, oil “only” 42 MJ/kg), but its volumetric energy density is much lower than that of any other fuel (at room temperature and pressure, hydrogen density is over 10,000 times smaller than that of gasoline [2]). In the case of fuel cell vehicles, about 10.4 kg of hydrogen would be needed to replace, on an energy basis, a 15 gallons (56 L) gasoline tank (the usual volume tank for current Internal Combustion Vehicles) [2]. Implementing properly the hydrogen economy requires a careful analysis of the storage and delivery systems.

3.1 Potential and Hazards of the Hydrogen Model

19

Compressed Gas Storing hydrogen as pressurized gas is currently the most developed storage technology. As compressed gas, hydrogen is stored in vessels at 700 bar for vehicular applications. This pressure exceeds 4 times the hydrogen pressure currently used in helium balloons, welding equipment, and other everyday applications in industry. Low volumetric and gravimetric storage density, compared to gasoline storage and other hydrogen storage technologies (even at as high pressure), constitutes one of the main drawbacks of compressed gas as a storage method. A 1-meter long, 30-cm wide cylindrical vessel can store 2 kg of hydrogen at 700 bar. Three to four vessels would therefore be needed to achieve a normal cruising range for a typical automobile [2]. Hydrogen compression process is energy intensive. The external energy input required for compression can exceed 20 % of the specific energy content of the compressed hydrogen itself [2]. Nonetheless, energy requirements to prepare hydrogen for usage and delivery are lower than for cryogenic storage systems. Hydrogen flammability and diffusivity, together with the intrinsic risk of high-pressure storage, demands special safety requirements for public handling. For safety concerns and better performance, hydrogen dispensing in delivery grids and hydrogen stations, from high-pressure sources to low-pressure containers, cannot be too quick. Temperature would quickly increase otherwise, and the volume of hydrogen transferred would be smaller. Cryogenic Liquid Hydrogen can also be stored under liquid phase. At 1 bar, hydrogen condenses into liquid around −252 ◦ C and can be kept in storage vessels. Adaptable vessel shape, together with faster refill capability, suppose an advantage in comparison with compressed gas storage technology. Another advantage is the high density achieved in cryogenic conditions, allowing to reduce storage pressure and volume, and to improve transport efficiency from the generation stations. As its main drawbacks, liquid storage faces the high-energetic cost of liquefaction (up to 30 % of the total specific energy content), as well as boil-off phenomena [2]. Boil-off occurs when the temperature of the liquid is raised over the boiling point by heat leakage into the storing vessel. As a result, pressure inside the vessel could exceed the mechanical stress limits of the walls. The hydrogen gas in excess must then be vented or consumed. If it cannot be consumed, venting results in energy losses and safety issues (hydrogen is highly inflammable). This is why technical and safety measures have to be taken for public handling, as is the case for compressed storage. This technology is relatively well developed, but it still needs improvement for its implementation in the mid and long term. Future production costs of liquid hydrogen are predicted to be significatively higher than for compressed hydrogen, or hydrogen from methanol, or other conventional oil resources. Technical achievements will determine storage methodology. Hydrogen refilling needs to be automated at hydrogen stations to avoid losses.

20

3 The Hydrogen Model

Currently, R&D initiatives have developed a mixed storage system that combines the advantages of compressed and liquid storage. This storage technology is known as cryo-compressed storage, and it is also performed in vessels. Solid Storage Solid storage of hydrogen consists in mixing hydrogen with other materials to generate solid-phase compounds or heterogeneous solid mixtures. This kind of technologies is still being developed with the goal of achieving light materials with a high storage density. Solid storage avoids most of the disadvantages of compressed and liquefied hydrogen and provides safer operating conditions (hydrogen is not carried as an element, but as a chemical compound). The main technologies for solid storage of hydrogen are hydrides and carbon storage. For automotive applications, a minimum of 6 % of hydrogen is required in the final compound mass, whether hydrides or carbon, to avoid excessively heavy storage systems [2]. Hydride storage is based on the storage of hydrogen as a part of a solid material matrix. It is to be released afterward by chemical or thermal reactions. As said before, this technology is under development and has proved potential advantages compared to other storage technologies (higher safety and easier control, for example). However, solid storage will not be implemented in vehicles unless better storage densities and cost improvements are achieved. For hydride storage technologies, heat transfer and flow systems are required to adjust the rate of storage and release of hydrogen. Hydride storage can be performed in chemical hydrides and metal hydrides. Hydrogen filling process is not reversible in the case of chemical hydrides, but it can be so if the element which is combined with the hydrogen is a metal. In this case, hydrogen combination and release can be performed in the storage module itself, working like a battery. Otherwise, for chemical hydrides, the fuel must be recycled at chemical plants to recover hydrogen and create new chemical hydrides. Currently, the 6 % goal has not been achieved yet and cycle durability for hydrides also needs improvement. Moreover, as hydrides are very sensitive to impurities, hydrides technology requires highly pure hydrogen for chemical reactions. Hydrogen can also be stored if it is introduced into specialized carbon nano configurations such as activated carbon, exfoliated graphite, fullerenes, nanotubes, nanofibers, or nanohorns. Storage densities up to 6 % of hydrogen in the compound mass have been observed for some carbon storage technologies at low temperatures (−196 ◦ C) or high pressures (>10 MPa), but still not at room temperature or atmospheric pressure, which would be needed for automotive applications [2]. Liquid Fuel Storage Hydrogen carrier fuels are also an option to reduce storage volume in portable applications. These fuels are not liquid hydrogen, but hydrogen-containing fuels, like methanol or ammonia. Due to the reduced performance of the fuel cell, it is an acceptable option for small, portable applications because of the reduced complexity and size of the overall system, but not for large applications. Hydrogen carriers can be internally reformed in high-temperature solid oxide or molten carbonate fuel cells,

3.1 Potential and Hazards of the Hydrogen Model

21

but due to the characteristics of these kind of fuel cells, they are more appropriate for stationary applications than for mobile ones [2].

3.1.3 The Fuel Cell The fuel cell is a system which uses hydrogen, or a hydrogen-containing fuel, to generate electricity through an electrochemical reaction. In a fuel cell, liquid- or gas-phase fuel and oxidizer streams enter separately through the electrodes, causing a chemical reaction which generates an electrical current. At the anode, the electrochemical oxidation of the fuel produces two separate flows: electrons are conducted to an external electrical circuit through the bipolar plate connection, generating useful energy, and ions are driven through the electrolyte to complete the circuit. Once used by the external circuit load, the electrons return to the cathode catalyst and are combined there with the oxidizer through the cathodic oxidizer reduction reaction (ORR), resulting in some chemical products. In the electricity generation process, fuel cells also dissipate heat as a consequence of the conversion efficiency from enthalpy of reaction to useful work. Yet, heat losses in fuel cells are lower than those in internal combustion engines (energy conversion efficiency is higher in electrochemical reactions than in combustion processes). The operating principles are similar to those in conventional batteries. Refilling rates for fuel and oxidizer are faster than charging and discharging processes in batteries [2]. Although, theoretically, a single fuel cell can achieve any power or current required, its voltage is limited by the electrochemical potential of the reactants and is always under 1 V. Higher voltages are achievable using fuel cells associations, whether in series or in series-parallel, configuring what is known as fuel cell stack. For automotive applications, where a high voltage is required, over 200 fuel cells can be used in a single stack. However, many auxiliary equipments are required for the correct performance of any type of fuel cell. For example, all of them need a recovery system to reuse nonreacted fuel after the electrochemical reaction [2]. Figure 3.1 pictures the way a fuel cell operates, and Table 3.3 shows the various types of fuel cells currently under development. Together with a brief indication of their main characteristics, one can find potential applications of each, as well as their advantages and disadvantages.

3.1.4 Hydrogen in Spain Hydrogen already has its market niche. It is currently used in physics and engineering. It has also many applications in chemical industries such as ammonia production or oil treatment at refineries. Before the conception of hydrogen economy, hydrogen technologies have been regulated by several standards, such as UNE-ISO 14687:2006 for hydrogen usage as fuel (although it was revised in 2012 considering its utiliza-

22

3 The Hydrogen Model

Fig. 3.1 Fuel cell operation process [12]

tion in Proton Exchange Membrane Fuel Cell road vehicle systems), UNE-ISO/TR 15916:2007 IN for safety concerns, or UNE-EN ISO 11114-4:2006 for compressed hydrogen vessels. However, a hydrogen-based model would require the development of new specific regulations considering hydrogen handling. In this respect, the US Department of Energy has designed the Hydrogen Codes & Standards Coordinating Committee (HCSCC). So did the European Union with the European Integrated Hydrogen Project (EIHP). In Spain, the Spanish Fuel Cell Association provides information about current standards related to hydrogen technologies applicable in Spain and Europe, mostly developed by International Standard Organization (ISO) [14]. Regarding the current status of Spain, a specific program for hydrogen technologies has not been developed yet, and thus the Hydrogen Economy is far from being a reality in the Spanish context. There are very few hydrogen refilling stations—in 2015, only four are operative6 and three are out of operation—and although Spain

6 Located

in Sevilla, Zaragoza, Huesca and Albacete. The map of operative, nonoperative, and planned hydrogen stations can be found at http://www.netinform.net/H2/H2Stations/Default.aspx.

600–800

160–220

Nafion polymer membrane

Molten alkali metal (Li/K or Li/Na) carbonates in porous matrix

Direct Methanol Fuel Cell (DMFC)

Molten Carbonate Fuel Cell (MCFC)

Phosphoric Solution of Acid Fuel phosphoric Cell (PAFC) acid in porous silicon carbide matrix

20–110

60–250

Solution of potassium hydroxide in water

Alkaline Fuel Cell (AFC)

Operating temperature (◦ +C)

Electrolyte material

Fuel cell type

H2 /reformed HC

H2 /CO/reformed HC

CH3 OH

H2

Fuel used

O2 /Air

CO2 /O2 /Air

O2

O2

Oxidant

Sulfur, high levels of CO

Sulfur

CO, sulfur, metal ions, peroxide

CO2

Major poison

Table 3.3 Fuel cell types, descriptions, and characteristics [2, 13]

200–1000

250–3000

0.001–10

1–100

Power (kW)

40–50

50–60

45–55

70

Efficiency

1–2 % CO tolerant, good-quality waste heat, durability

CO tolerant, fuel flexible, high-quality waste heat, inexpensive catalyst

Low-temperature operation, high efficiency, high H2 power density, rapid start-up

High efficiency, low oxygen readuction reaction losses

Pros

Low power density, expensive, platinum catalyst used, slow start-up, loss of electrolyte

Electrolyte dissolves cathode catalyst, long start-up time, CO2 must be injected to cathode, electrolyte maintenance

Expensive catalyst, low durability, poor-quality waste heat. Intolerance to CO, thermal and water management

Pure oxygen without CO2 contaminant required for running

Cons

(continued)

Premium stationary power

Stationary power applications including cogeneration

Portable applications, electronic devices

Space applications with pure O2 /H2 available

Applications

3.1 Potential and Hazards of the Hydrogen Model 23

Electrolyte material

Flexible solid perfluorosulfonic acid polymer

Yttria (Y2 O2 ), zirconia (ZrO2 )

Fuel cell type

Polymer Electrolyte Fuel Cell (PEFC)

Solid Oxide Fuel Cell (SOFC)

Table 3.3 (continued)

600–1000

30–100

Operating temperature (◦ +C)

H2 /CO/reformed HC

H2 /reformed HC

Fuel used

O2 /Air

O2 /Air

Oxidant

Sulfur

CO, sulfur, metal ions, peroxide

Major poison

40–50

45–55

300–3105

Efficiency

20–250

Power (kW)

CO tolerant, fuel flexible, high-quality waste heat, inexpensive catalyst

Low-temperature operation, high efficiency, high H2 power density, rapid start-up

Pros

Long start-up time, durability under thermal cycling, inactivity of electrolyte below ∼600 ◦ C

Expensive catalyst, low durability, poor-quality waste heat. Intolerance to CO, thermal and water management

Cons

Stationary power applications, including cogeneration

Portable, automotive, and stationary applications

Applications

24 3 The Hydrogen Model

3.1 Potential and Hazards of the Hydrogen Model

25

plays an important role in Europe as road vehicle manufacturer, most of the technical advances in hydrogen technologies are being performed abroad. At the European level, Germany, Switzerland, and the union of Denmark, Sweden, and Norway, have already developed and implemented a mobility plan for hydrogen. France and the United Kingdom are in the preparation and assessment phases of their plans [15, 16]. On October 22, 2014, the Directive 2014/94/EU on the deployment of alternative fuels infrastructure was approved in Europe [17]. This directive requires EU Member States to adopt national policy frameworks where implementation timings are fixed for each alternative technology and notify them to the European Commission two years before the approval date. As a Member State, Spain has the chance of including hydrogen mobility plans in its alternative fuel policy framework.

3.2 The Hydrogen Vehicle 3.2.1 The Fuel Cell Electric Vehicle As an alternative to conventional internal combustion engine vehicles, several vehicle systems have been developed, powered by more environmental-friendly options. The fuel cell electric vehicle is only one type of electric vehicle. Among the various kinds of electric vehicles, the most important configurations are: • Electric Vehicle (EV). The vehicle powertrain7 runs only with electricity stored in a battery. The model based on such vehicles is the focus of Chap. 4. • Hybrid Electric Vehicle (HEV). HEVs combine a conventional internal combustion powertrain with an electric propulsion system. Depending on the capacity of the electric powertrain to propel the vehicle, there are mainly two degrees of hybridization. Full hybrid vehicles can run independently from the combustion engine. Mild hybrids alway need the conventional engine switched on, even when the electric engine is running. A special type of HEV is the plug-in hybrid electric vehicle (PHEV). This kind of vehicles are equipped with batteries that can be charged connecting them to an external electric powersource like the electric grid. They are not considered in this book. • Fuel Cell Electric Vehicle (FCEV). Such vehicles are the focus of this chapter. They have a substantially electric powertrain, where electricity is provided by a fuel cell running with hydrogen. The functioning and main characteristics of this kind of electric vehicles will be detailed later.

7 The

word powertrain refers to the group of elements in a vehicle that transforms the energy contained in the fuel into useful power and delivers it to the surroundings. Powertrain includes the engine, transmission, drive shafts, differentials and wheels.

26

3 The Hydrogen Model

In general, vehicles equipped with electrical systems, whether pure or hybrid, are more silent and emit less-pollutant gases to the atmosphere at vehicle level than conventional ICEs. Nonetheless, emissions during electricity generation for the vehicle can be higher than emissions released during ICEs usual driving (see Chap. 4). Due to the reduced driving range of the electrical vehicles, a generalized hydrogen economy may well see EV’s and FCEV’s coexisting, each covering different driving ranges. Electric vehicles would be used for driving ranges lower than 200 km (city run-arounds, second cars) while fuel cell electric vehicles would be used for larger rides (heavy-duty trucks, transport, passenger cars, etc.) [18]. Such differentiation also stems from the differences in charging, or refueling, times: batteries are slow to charge while hydrogen refueling from a dispensing point is very fast. Yet, the development of faster charging points, or of some batteries exchange system, could tip the balance toward EVs. In a FCEVs, several fuels can be used to feed the fuel cell. Such a vehicle can incorporate a small reformer to produce hydrogen from gasoline, naphtha, methanol, or diesel. This could enable the design of an energetic model based on hydrogen without the complex infrastructure required for pure hydrogen. Handling would also be safer and more effective. Nevertheless, this kind of vehicles is very complex, emits greenhouse gases8 and bring the complexity of an additional system on board (additional weight, wide dimensions, complex control). Moreover, reforming processes are too slow to be efficiently implemented at the vehicle level, and meet the expected power. With the current state-of-the-art hydrogen technologies, pure hydrogen vehicles are more likely to be preferred in a hypothetical model switch.

3.2.2 Elements of the Fuel Cell Electric Vehicle There are fundamentally two types of fuel cell electric vehicles: (1) simple fuel cell electric vehicles only use hydrogen to power the vehicle. (2) hybrid fuel cell electric vehicles are powered by other elements apart from hydrogen-containing fuel (for example, batteries charged from the grid) [13]. Both of them integrate at least the following systems: • Fuel processor. Prepares the hydrogen-containing fuel (gaseous hydrogen, methanol, etc.) before it is introduced into the fuel cell. It is important to eliminate impurities in fuel cell systems where purity requirements are very high, like proton exchange membrane fuel cells (PEMFC). This system can also include a small reformer in the case of methanol. • Fuel cell. Produces electricity from air and fuel. In automotive applications, the fuel cells used are typically PEMFCs. Fuel cell outputs are electricity (used to move the vehicle), heat, and water vapor which constitute the single emission of FCEVs. 8 CO 2

and other pollutants, depending on the fuel. See Sect. 3.1.3.

3.2 The Hydrogen Vehicle

27

• Electric engine. Transforms electrical energy into mechanical energy at wheels. Usually uses alternating current. • Electronic Control Unit. Controls all the units and systems commanded by electrical signals (brakes, sensors, valves, gearbox, etc.) and synchronizes them to achieve a proper vehicle running. Its function is the same as for internal combustion vehicles. • Accumulators (batteries). Accumulate energy under the form of electricity (batteries) or mechanical energy (compressed air, springs) to be used later by the vehicle. In this kind of vehicles, the main accumulators are batteries. • Fuel tank. Stores the hydrogen to be used by the fuel cell. Apart from the previous systems, hybrid fuel cell electric vehicles also integrates some power generation systems to supply energy to the vehicle when the fuel cell is not working. In both HFCEVs and FCEVs, there are also other auxiliary systems necessary for the correct operation of the vehicle (e.g., cooling systems to extract heat from the fuel cell, compressors for air conditioning, lightning systems, etc.). As in other kind of vehicles, regenerative braking allows to recover part of the kinetic energy when the vehicle slows down (see Appendix A.2). In this study, only simple fuel cell electric vehicles, with PEM fuel cells running with hydrogen, will be considered for the assessment of GHG emissions. The elements of a fuel cell vehicle are sketched on Fig. 3.2.

Fig. 3.2 Elements of a fuel cell electric vehicle [13]

28

3 The Hydrogen Model

3.3 Hydrogen Required Which amount of hydrogen would be required to seamlessly replace all conventional vehicles by fuel cell electric vehicles with PEM fuel cells? Answering this question requires the knowledge of the amount of energy which should get to the wheel of the FCEV to mimic the current traffic. The answer has been provided in Chap. 2. We then need to estimate the Tank-To-Wheel ηT T W efficiency of the hydrogen vehicles. Among others, such efficiency depends on the way hydrogen is stored in the car. As was commented in Sect. 3.1.2, there are fundamentally three methods to store hydrogen: as compressed gas in pressurized vessels, as liquid in thermally isolated tanks, or forming a solid compound with other elements. The last one is not commercially available yet and needs more R&D before being implemented in a vehicle. We will therefore focus on the first two options: pressurized vessels and liquid storage in thermally isolated tanks. Each storage technology determines a different vehicle configuration and the efficiency from the tank to the wheel slightly varies from one configuration to the other. Figures 3.3 and 3.4 show the different processes performed that incur in energy losses for gaseous and liquid storage. Each process defines an energetic efficiency that enable the assessment of a Tank-To-Wheel (TTW) efficiency. All the elements involved in the FCEV power train are collected next to their energetic efficiencies in Table 3.4. Note that for vehicles equipped with gaseous hydrogen vessels, vehicle performance is slightly better than for those equipped with liquid tanks. Nonetheless, as liquid hydrogen provides a higher driving range than gaseous hydrogen for the same volume unit, liquid storage is pointed out as the best candidate for the future.

Fig. 3.3 Tank-To-Wheel efficiency of the FCEV for storage as gas

Fig. 3.4 Tank-To-Wheel efficiency of the FCEV for storage as liquid

3.3 Hydrogen Required

29

Table 3.4 Breakdown of FCEV efficiencies ηT T W Storage as gas Storage as liquid Element Efficiency Element Gaseous storage [2] Fuel cell consumption [2] Battery [20] Electric Motor [21] Transmission [22] Total

1 0.5 0.8 0.9 0.8 0.288

Efficiency

Liquid storage [19] Fuel cell consumption Battery Electric Motor Transmission Total

0.979 0.5 0.8 0.9 0.8 0.282

The energy to wheels required by the ICE vehicle fleet has been calculated in Sect. 2.1 and is approximately 57 TWh.9 If a fleet of FCEVs had to deploy the same energy at wheels, the total amount of energy which would be required under the form of hydrogen would be, Energy required =

Energy at wheels ICE ηT T W

(3.3)

Knowing 1 kg of hydrogen holds 120 MJ of energy [23], we can then compute the required mass of hydrogen. The results can be found in Table 3.5. Considering hydrogen losses of 3.3 % in the delivery and storage phases for gaseous hydrogen, and losses of 1.2 % for liquid hydrogen (these figures will be justified in Sect. 3.4), one can estimate the total hydrogen mass which is necessary to produce to supply the FCEV fleet. The results can also be found in Table 3.5.

3.3.1 Hydrogen Versus Internal Combustion Engine Vehicle We now conduct a comparative energy survey between gasoline, diesel, and electric vehicles. We distinguish the two cases for internal combustion because, as we saw earlier, the diesel vehicle is more efficient than the gasoline one (because of the engine). Table 3.5 H2 mass required by the FCEV fleet

9 This

Storage

Gaseous

Liquid

H2 to dispense (Mtons/year) H2 to generate (Mtons/year)

5.93 6.18

6.06 6.14

number slightly differs from the 51 TWh mentioned in Chap. 2. This is because it was computed with slightly different fuel consumption statistics for diesel and gasoline cars.

30

3 The Hydrogen Model

The consumption of an internal combustion engine car, gasoline, or diesel, is expressed in liters/100 km, and the one of an electric car in kWh/100 km. To compare them, we need to know the energy content of the fuel, namely [24], • One liter of gasoline holds about 32.18 MJ. • One liter of diesel fuel holds about 35.86 MJ. For internal combustion vehicles, the average consumption calculated from the official figures published in the Instituto para la Diversificacion y Ahorro de la Energía and already used in Chap. 2 are [25], • Average actual consumption of gasoline cars is 6.55 L/100 km. • Average actual consumption of diesel cars is 5.53 L/100 km. We therefore find the following energy consumption for internal combustion vehicles, • Gasoline: 6.55 × 32.18 = 210.8 MJ/100 km. • Diesel: 5.53 × 35.86 = 198.3 MJ/100 km. Regarding the electric cars, we just saw that reproducing the current traffic would require about 6 Mt of hydrogen per year, that is, about 7.2 × 1011 MJ. We also checked that with such an amount of energy, our electric cars would be able to cover 1.1 × 1011 (gasoline) + 1.7 × 1011 (diesel) = 2.8 × 1011 km. As a result, the average energetic consumption for these cars is, • Hydrogen: 7.2 × 1011 /2.8 × 1011 km = 257 MJ/100 km. This figure is twice the one declared, for example, by Honda for its FCX Clarity (1 kg H2 /100 km, with 120 MJ/kg of H2 ). Note however that the EPA10 consumption of the Honda vehicle is measured through driving cycles under controlled conditions of one single vehicle. Meanwhile, the figure above considers the mean values of distance traveled, gasoline and diesel consumption, and vehicle efficiency for an entire fleet.

3.4 Energy Scenarios for CO2 Equivalent Calculation If, then, the current Spanish traffic was to be “hydrogenized,” what would happen to GHG emissions? Several scenarios have been defined depending on the way the hydrogen economy is implemented.

3.4.1 Definition of the Scenarios Production, delivery, and storage technologies, even at vehicle level, vary from one scenario to the other according to technical considerations and the projected devel10 Environmental

Protection Agency, see http://www3.epa.gov/otaq/sftp.htm.

3.4 Energy Scenarios for CO2 Equivalent Calculation

31

opment of hydrogen technologies. As the hydrogen model has not been implemented in Spain yet, and current technologies are not definitely mature, the list of possible, valid combinations, is relatively large. The absence of completely implemented models in the world also complicates the elaboration of a best practices list based on experience. However, in this assessment, five scenarios will be considered following the guidelines provided by the International Energy Agency and by other experts: Scenario 1 Hydrogen is produced through the current hydrogen generation mix. Electricity required in the process also follows the current generation mix for electricity in Spain. Scenario 2 considers the self-generation of hydrogen at every household by an electrolysis process. Electricity required is generated using photovoltaic cells. There is no other input of electricity. Scenario 3 Hydrogen is completely generated from natural gas. Electricity for hydrogen processing equipment is generated by the current generation mix. This is the IEA short-term scenario. Scenario 4 contemplates a combination of 50 % coal and 50 % biomass for hydrogen production. The current electricity generation mix is considered for the generation of the electricity required in the delivery phase for hydrogen compression and liquefaction. This is the International Energy Agency (IEA) mid-term scenario. Scenario 5 contemplates an entirely renewable energetic model. Hydrogen is generated exclusively from renewable electricity, and electricity for the delivery phase also comes from a 100 % renewable mix. This is the IEA long-term scenario. The three last scenarios, in this order, are considered by the International Energy Agency (IEA) to be the logical order of implementation of a hydrogen-based energetic model [11]. Table 3.6 collects each scenario definition for an easy comparison. The starting point for all scenarios is the current energy demand of the road traffic in Spain. Every scenario must supply as much energy at wheels level as the current energetic model; otherwise, consumer’s driving trends and habits would be different

Table 3.6 Energetic scenarios for the hydrogen model Scenario Feedstock Generation Delivery 1 2 3 4 5

Gas, oil, coal, grid electricity Electricity from solar PV Gas, grid electricity Coal, biomass, grid electricity Renewable electricity

Storage at vehicle

Centralized

Truck

Gaseous

Distributed

–

Gaseous

Centralized Centralized

Truck Truck

Gaseous Gaseous

Centralized

Pipelines

Liquid

32

3 The Hydrogen Model

and models would not be comparable. The assessment of energy demanded by the current vehicle fleet in Spain has been done in Chap. 2. The analysis of GHG emissions in each scenario is developed in several steps, as follows: 1. Determination of the technologies involved in the production, delivery, and storage processes, and if applicable, participation percentage in the whole phase. Raw material extraction and processing are also considered. 2. Partial efficiency calculation and determination of the energy required for every phase of the scenario. Estimation of the global scenario efficiency and energy requirements. 3. Estimation of the hydrogen losses during the delivery phases, and of the total hydrogen mass necessary to produce and deliver. GHG emissions directly depend on the amount of hydrogen required by the fleet. 4. Determination of the electricity required to compress or liquefy hydrogen, and estimation of the electricity that a power plant would have to produce (the difference between the two is basically the grid loss). 5. Assignment of CO2 eq values to each process and estimation of the total amount of CO2 eq emitted. Where possible, a statistical distribution of the emissions is considered (quartile distribution). 6. Comparison between scenario and reference emissions. Once developed, the emissions assessment algorithm is identical for every scenarios. Calculations can be performed with a Microsoft Excel spreadsheet, making it easy to change any parameters in case of technical advances or institutional decisions. The model is also flexible enough to accommodate technical combinations not considered in this study. Note that emissions assessment algorithms have also been implemented into specific software by recognized institutions. One example is the GREET Model, developed by the US Department of Energy.11

3.4.2 Electricity Generation In all scenarios but number 2, it is necessary to bring in certain amount of electricity not directly related to the production and delivery processes. It powers compression and liquefaction equipments, and it is not negligible. As commented in Sect. 3.1, hydrogen has to be compressed before being dispensed (storage in the vehicle tank would be completely ineffective otherwise, and the autonomy very low). Hydrogen also needs to be liquefied if it is to be delivered by trucks or pipelines, and these processes spend large amounts of energy (Fig. 3.5). Electricity typically comes from the electrical grid and is generated according to a certain generation mix. The generation mix is the percentage distribution of each 11 Pathways,

assumptions, and procedure can be found at https://greet.es.anl.gov/.

3.4 Energy Scenarios for CO2 Equivalent Calculation

33

Fig. 3.5 Delivery diagram including electricity input

electricity generation method (renewable sources, nuclear power plants and fossil fuel power plants running with natural gas, oil, coal, etc.) in the total amount of electricity produced. GHG emissions from electricity production contribute significantly to the overall emissions of any given scenario. As the vehicle does not directly emit any pollutant, GHG emissions arise from the production and delivery processes. GHG emissions due to electricity generation can be estimated from Chap. 2, using Fig. 2.3 for every generation method, resulting in Table 2.2, p. 11. The numbers reported in Fig. 2.3 and Table 2.2 account for every emissions throughout the electricity generation life cycle, including construction of the power plant, raw material extraction and distribution, and power generation. This is the reason why renewable energies and nuclear power, which do not emit CO2 during the electricity generation process, get nonzero values for their life cycle GHG emissions. For the same reason, electricity generation from some types of biomass can have negative values of CO2 eq in their life cycle assessment. It is important to note that crops, woods, and other biomass sources absorb CO2 in the growing phase of the plant, so that the net balance can be zero. It is also remarkable that among the renewable energies, photovoltaic and biomass have the highest maximum value of emissions. The main emissions due to electricity generation are caused by fossil fuels, which explain the high levels of CO2 emissions in scenarios where those processes are present. The Spanish current electricity generation mix is shown in Fig. 3.6. Such is the mix we will consider anytime electricity is to be obtained from the grid, except in Scenario 5.

3.4.3 Emissions Model Simplifications In order to facilitate the emissions assessment, several assumptions have been made for every scenario: • Assessment is made only for private cars. Light, medium, and heavy-duty vehicles, as well as motorcycles, buses, etc. are not considered in the study.

34

3 The Hydrogen Model

Nuclear 21%

Biomass 2%

Coal 17% Oil 12%

Wind 19% Hydroelectric 16%

Natural gas 9% Solar 5%

Fig. 3.6 Electricity generation mix in 2014 [26]

• Reference emissions are also estimated for private cars only. • Internal combustion vehicles are replaced by simple fuel cell electric vehicles only, running with pure hydrogen. • In every scenario (except scenario 2)v there are at least two energy inputs during the life cycle chain. One of them is typically the electricity used for hydrogen compression and liquefaction. The rest of the energy inputs are energy sources used as raw material in the hydrogen production process. • Energy inputs in hydrogen production processes are considered single resource inputs (natural gas, coal, sunlight, etc.). This resource is, moreover, the raw material of the production process. The rest of inputs with energy content (steam, electricity, fuel for heating) will be considered as an equivalent part of the main energy source. • Water vapor released to the atmosphere is not considered when computing the amount of CO2 eq emitted by a scenario (water vapor is the main atmospheric GHG). In this respect, the amount of water in the atmosphere only depends on the temperature at the Earth surface (pressure can be considered constant and equal to 1 atmosphere), and water in excess comes back to the surface as rain within 10 days or so [27]. • GHG from hydrogen delivery in trucks, as well as the hydrogen vented to the atmosphere during the process, are very small when compared to the emissions of other processes. As such, they are not considered in the assessment. This last assumption can be justified by a quick estimation of the GHG emissions associated with trucks deliveries of hydrogen. The calculation is detailed in Appendix A.3 and shows that they only represent a small fraction of the reference emissions estimated in Chap. 2.

3.4 Energy Scenarios for CO2 Equivalent Calculation

35

3.4.4 Scenario 1. Current Hydrogen Generation Mix This pathway contemplates the effects of implementing the hydrogen economy if hydrogen is simply produced at large scale from the current hydrogen commercialization conditions. In other words, this scenario assesses the effects of resizing the current hydrogen market to adapt it to a higher demand. This could be a hypothetical starting point for a hydrogen economy implementation. Production In this pathway, hydrogen is obtained following the methods used in the current hydrogen market. Therefore, 48 % of hydrogen is produced from natural gas, 30 % from oil, 18 % from coal, and the 4 % remaining using grid electricity generated by the current electricity mix of Spain. Every raw material involves its own production process, with several stages and technologies implied. Figures 3.7, 3.8, 3.9, and 3.10 describe the different phases of the productive process of hydrogen Tables 3.7, 3.8, 3.9 and 3.10 indicate the corresponding efficiencies. Hydrogen from natural gas Natural gas is extracted from land or sea deposits and is then treated to eliminate hydrogen sulfide (HS is corrosive and reduces the heat capacity of the natural gas) and carbon dioxide. Natural gas is carried to an hydrogen plant by truck or pipeline. The hydrogen plant transforms the natural gas stream into a hydrogen-containing stream through a steam methane reforming process. The output stream presents some impurities, which are later removed by a process known as pressure swing adsorption (PSA). The result of the process is gaseous hydrogen which needs to be liquefied or compressed before being delivered.

Fig. 3.7 Hydrogen production process from natural gas

Fig. 3.8 Hydrogen production process from oil

36

3 The Hydrogen Model

Fig. 3.9 Hydrogen production process from coal

Fig. 3.10 Hydrogen production through an electrolysis process, employing electricity from the grid

Table 3.7 Breakdown efficiencies for hydrogen production from natural gas

Operation

Efficiency

Natural gas extraction [28] Processing and desulphurization [28] Transport to reforming point [13] Steam methane reforming (SMR) [4] Pressure swing adsorption [28] Total

0.957 0.972 0.95 0.8 0.85 0.601

Table 3.8 Breakdown efficiencies for hydrogen production from oil

Operation

Efficiency

Oil extraction [29] Oil transportation [13] Cracking [13] Steam reforming [4] Pressure swing adsorption [28] Total

0.8 0.95 0.85 0.8 0.85 0.439

3.4 Energy Scenarios for CO2 Equivalent Calculation Table 3.9 Breakdown efficiencies for hydrogen production from coal and biomass

Table 3.10 Breakdown efficiencies for hydrogen production from electrolysis

37

Operation

Efficiency

Coal mining/biomass growth and processing [29] Coal/biomass transportation [28] Gasification [30] Pressure swing adsorption [28] Total

0.8 0.99 0.575 0.85 0.387

Operation

Efficiency

Renewable electricity generation [31] REE transport [32] Electrolysis [2] Total

0.4 0.985 0.7 0.276

Hydrogen from oil In parallel to the previous production chain, oil is extracted from land or sea deposits and is cracked and reformed to obtain the liquefied petroleum gases (LPG), which are reformed to afterward obtain hydrogen. Impurities, sulfur compounds, and other useless and dangerous substances are removed before transporting the LPG to a hydrogen plant. Gases are reformed there following a steam reforming process and hydrogen stream with impurities is obtained as product. A pressure swing adsorption process allows the accomplishment of pure-enough gaseous hydrogen for delivering. Hydrogen from coal Coal can be relatively easy to find and is obtained in mines with a low extracting cost. After mining, coal is transported to the hydrogen plant (usually by train, but also by ship or road transport) and gasified to obtain an hydrogen-containing stream. This stream has more impurities than in the case of the other production methods, and once gasified it needs to be purified to allow for an efficient delivery. Hydrogen from water Hydrogen is produced in this case through an electrolysis process. First, electricity is generated in conventional power plants or in renewable plants following the 2014 generation mix in Spain. This electricity is conducted by the Spanish grid (Red Eléctrica de España) to the hydrogen production plant. Hydrogen is produced there from water through a electrolysis process obtaining highly pure gaseous hydrogen that can be delivered after compression or liquefaction. Delivery and Storage at Vehicle As commented before, this scenario would be the first step toward the switch to the hydrogen model. It corresponds to the idea of a quick implementation of the hydrogen economy. Therefore, the delivery and storage technologies also attend to an immediate need, and must be technically and commercially available.

38 Table 3.11 Breakdown efficiencies for hydrogen delivery by truck

3 The Hydrogen Model Operation

Efficiency

Gas liquefaction [33] Transport by truck [28] Vaporization and compression [2] Gaseous storage at fueling station [28] Dispensing [28] Total

0.65 0.972 0.8 1 0.989 0.5

For this reason, delivery is here performed by trucks, from the different hydrogen production plants (distributed generation).12 Before being introduced in the tank of the truck, hydrogen needs to be compressed or liquefied using electricity from the grid to raise its energetic density. In this scenario, the transport is considered under liquid phase. After the transport by truck, hydrogen is stored as a gas at the hydrogen station to refill the car tanks (see Table 3.11 and Fig. 3.11). FCEVs are supposed to be the current commercially available models, equipped with gaseous hydrogen tanks. Tank-To-Wheel efficiency for these vehicles is 28.8 %, as calculated in Sect. 3.3. For this scenario, where hydrogen handling mainly occurs in liquid phase, losses of 2.8 % are considered during the transport by truck and 1.1 % during hydrogen dispensing [28] (as a very diffusive gas, hydrogen dispensing is not as clean as diesel or gasoline). The electricity required to compress and liquefy hydrogen in this scenario is 114.95 TWh/year. Energy Efficiency and GHG Emissions Once the hydrogen production process is deployed and the energetic efficiency of their stages quantified, a global scenario efficiency can be calculated to estimate the yield of raw material supplying road transport energetic needs. Therefore, ηglobal =

Energy at wheels Energy input

(3.4)

The equation above considers both the energy employed in generating electricity for compression and liquefaction and the energy inverted as raw material.13 For this scenario, the global energy efficiency (Well-To-Wheel) is 10.27 %. The energy required by the scenario is 554.61 TWh, equivalent to 41 % of the total primary energy consumed in 2013 (1,348 TWh, [34]). It also represents about twice the energy required by the current ICV fleet.

12 Setting

up a pipeline grid is a slow and expensive process. every process consumes a certain amount of energy that is not related to the raw material used as input of the process. See model simplifications at the beginning of the section. 13 This is only an assumption. During its performance,

3.4 Energy Scenarios for CO2 Equivalent Calculation

39

Fig. 3.11 Diagram of hydrogen delivery process by truck

Such a large amount of energy required is basically due to the low efficiency of the extraction and delivery processes. For the conventional fossil models, delivery has an efficiency close to 100 % due to the high energy density of the fuel and its easy handling. Moreover, fuel does not need additional treatment to be delivered. The energy cost of hydrogen compression and liquefaction, together with a lower generation efficiency, raises the primary energy required by the model. This scenario eventually results in the GHG emissions shown in Fig. 3.12. GHG emissions are in this case higher than for a conventional fossil model (3 to 4 times higher than the reference emissions, and the highest of the five scenarios considered). Why is it so?

Fig. 3.12 GHG emissions for Scenario 1

40

3 The Hydrogen Model

One reason is the use of fossil fuels to generate hydrogen in highly emitting processes. Among all these processes, coal gasification is the one with the highest pollution figure per kilogram of hydrogen generated. Also, producing hydrogen from natural gas results in high emissions, which happens to be the main hydrogen production method in this scenario. Another reason is the high level of emissions associated to electricity generation for the delivery phase. Furthermore, the large amount of energy required by the model, whether for hydrogen generation or processing, is in itself another reason for such high emissions.

3.4.5 Scenario 2. Home Generation of Hydrogen We here consider that the consumer generates at home the hydrogen she/he needs. This could be done in small electrolyzers running with solar energy. An additional grid of slightly larger electrolyzers, also running with solar energy, could supply energy to streets, roads, or hydrogen stations. Currently, there are several initiatives and projects with these characteristics to impulse the hydrogen model implementation. This model is especially designed for isolated areas where hydrogen from the grid cannot reach all users, but it has also been suggested as a global model supplying hydrogen to all FCEV. Production As previously commented, hydrogen would be generated from water by an electrolysis process (Fig. 3.13). To achieve the electricity required, solar panels would be installed on the roofs. Hydrogen obtained by electrolyzer would be already compressed and ready to fill the tank of the vehicle. As hydrogen is generated in the very place where it is dispensed, this production system also corresponds to a distributed generation configuration. Delivery and Storage at Vehicle Onsite hydrogen production allows for a direct refueling from the electrolyzer. This scenario contemplates car refueling from the solar electrolyzer itself, making pipes and transport trucks unnecessary.

Fig. 3.13 Home production of hydrogen from photovoltaic electricity

3.4 Energy Scenarios for CO2 Equivalent Calculation

41

Fig. 3.14 Diagram of hydrogen pathway from home photovoltaic generation

The possibility of not setting up an additional delivery system for hydrogen would result in lower costs of implementation for the hydrogen economy (pipelines and trucks would not be needed), and lower consumption of resources due to the increase of the global efficiency. These gains would therefore result in lower GHG emissions. As this hydrogen model can be integrated with generation through local, renewable sources, it is possible to design a clean model with near-to-zero emissions. In this case, nonetheless, the energy saved through the lack of delivery processes is lost through the low efficiency of hydrogen generation by renewable sources.14 For the energy and GHG assessment in this scenario, hydrogen losses at dispensing points (about 1.1 %) will not be considered because of the scarce influence in the final result. As a consequence of the lack of delivery phase, there is no extra electricity to generate either. Energy Efficiency and GHG Emissions Since there is no external input of electricity in this scenario, it is possible to estimate the global scenario efficiency as the product of efficiencies of the linear life cycle model (Fig. 3.14). Global energy efficiency (Well-To-Wheel) is then ηW T W = 2.72 %. Such a low efficiency is mainly due to the limited efficiency of electricity generation from sunlight (commercial PV cells typically have efficiencies around the 15 %). This way, the energetic cost at the beginning of the cycle is very high, the highest of the five scenarios considered in this study (2,093 TWh, which supposes the 150 % of the primary energy consumption in Spain and about 7.5 times the current model energy requirements). Nonetheless, as this energy naturally comes from the Sun, economic concerns are not focused on primary energy production, but on the amount and on the cost of the solar panels required to generate the final electricity needed by the electrolysis processes. Assuming an average solar irradiation of I = 1.5 MWh/m2 /year in Spain,15 and a roof coverage factor16 R = 0.9, the panel surface required to supply enough electricity for hydrogen generation for one vehicle, and for one year, is given by, 14 The most efficient configuration would be based on hydrogen production from fossil fuels in the same place where hydrogen is consumed. This is not always possible and can incur in unacceptably high GHG emissions in the long run. 15 See http://solargis.info/doc/free-solar-radiation-maps-GHI. 16 Solar panels usually cannot cover the totality of the surface where they are located.

42

3 The Hydrogen Model

Fig. 3.15 GHG emissions for Scenario 2

S=

E wheels (one vehicle) = 62.75 m2 . ηW T W R I

(3.5)

This value is small enough to contemplate an installation on the roof of a private house. In the case of a flat, where roof surface is shared by several households, additional panels (located elsewhere) would be needed. Generalizing the figure, the size of the panel surface required for the whole fleet is 1,396 km2 , considering the number of vehicles in Chap. 2. This represents 0.3 % of the country territory, so that the implementation of the model is physically possible.17 The overall GHG emissions of the present scenario are shown in Fig. 3.15. As observed, the GHG emissions of the model can vary between zero and nearly the current GHG emissions, evidencing several features. First, the absence of electricity requirements at delivery phase reduces substantially the GHG emissions with respect to models in which delivery is included. Second, in the worst case (max emissions), this model does not suppose an advantage in environmental terms with respect to the current fossil model. But in the best case, GHG emissions would completely cease. Note that the electrical energy required by the model is very high. Therefore, electricity generation, although renewable, has a big impact during the construction phase of the equipments. Furthermore, solar PV is the renewable source with the highest emissions coefficient, as observed on Chap. 2, Fig. 2.3. 17 Note

that Spain surface is 504,645 km2 , but 300,422 km2 are already occupied by land exploitations [35].

3.4 Energy Scenarios for CO2 Equivalent Calculation

43

Consequently, this model can constitute a solution, with GHG emissions probably dramatically reduced. At any rate, the efficiency of the configuration would be determined by the GHG emitted during the construction phase of the required infrastructures.

3.4.6 Scenario 3. IEA Scenario for Short-Term Implementation This configuration is considered by the IEA as the most probable short-term implementation [11]. The reason is the high efficiency of natural gas reforming processes, and the low cost of both the raw material and the production process, in comparison with other hydrogen production methods. In this assessment, GHG emissions per year are estimated independently from the duration of the scenario. Production The production process is the same as commented in Scenario 1 for production from natural gas. Hydrogen production system is distributed for this scenario, as the deployment of a pipeline grid supposes a high inversion of capital and time. Therefore, production can be expected from small hydrogen production plants, from where hydrogen is eventually delivered by truck (Fig. 3.16). Delivery and Storage at Vehicle As commented before, the production model is distributed in this scenario. To address the demand from different geographical areas, hydrogen would be dispatched by truck. Once the demand is established, larger production plants would be installed and the construction of a pipeline network for hydrogen would get started to improve the efficiency of the model. Similarly to Scenario 1, hydrogen is transported under liquid phase to improve energetic density during delivery, and dispensed under gas phase at hydrogen stations. The delivery diagram is the same as for Scenario 1 (see Fig. 3.11 and Table 3.11).

Fig. 3.16 Hydrogen production process from natural gas

44

3 The Hydrogen Model

Electricity used for hydrogen processing is taken from the grid, and is consequently generated following the current generation mixture. The amount of electricity required for hydrogen compression and liquefaction, as a consequence of the delivery stage, is 114.95 TWh/year. As a short-term scenario, the model would use the currently available technical resources and generation configuration. For this pathway, as in scenario 1, hydrogen losses of 2.8 % are considered for truck transportation, and 1.1 % during hydrogen dispensing. The FCEVs considered in this scenario are also the commercially available vehicles at present time. Consequently, storage at tank is performed under gas phase. Energy Efficiency and GHG Emissions Considering the processes described above, the global scenario efficiency (Well-ToWheel) is then 12.45 %, requiring 457.53 TWh of primary energy at the beginning of the chain. This amount represents 32.3 % of the primary energy consumption in Spain, and about 1.5 times the current energy requirements of the fossil model. The resulting GHG emissions are shown in Fig. 3.17. The use of natural gas for hydrogen generation entails high emissions. This, together with the emissions released during hydrogen compression and liquefaction, sets this scenario emissions between 2.5 and 3.5 times the reference emissions. Therefore, this model cannot replace the current model to reduce GHG emissions. However, it can be used as a transition to less-polluting configurations, although figures and time scales must be considered in government anti-polluting planning.

Fig. 3.17 GHG emissions for Scenario 3

3.4 Energy Scenarios for CO2 Equivalent Calculation

45

Fig. 3.18 Hydrogen production equally from biomass and coal

3.4.7 Scenario 4. IEA Scenario for Mid-Term Implantation The IEA mid-term hydrogen model mainly relies on coal and biomass hydrogen generation. This configuration supposes another step in the transition to a completely renewable energetic model, as biomass is slowly introduced in hydrogen generation. Coal is employed to replace natural gas as a cheap fossil resource to produce hydrogen, once natural gas becomes a scarce resource.18 Current electricity generation mix is considered for the electricity required in the delivery phase for hydrogen compression and liquefaction. Production The production process from biomass and coal is detailed in Fig. 3.18. The production mix considers gasification processes for a combination of 50 % of coal and a 50 % of biomass. Hydrogen production system is still distributed for this scenario. To simplify the assessment of GHG gases, the efficiency of each production stage will be considered equal for both options (see Sect. 3.4.3). Delivery and Storage at Vehicle It is assumed that an hypothetical transition to this scenario would occur when the pipeline network is not operative yet, and the centralized production plants start being built. Therefore, hydrogen distribution is performed under liquid phase by truck and distributed under gas phase at hydrogen stations, as in Scenario 1 (see Fig. 3.11 and Table 3.11). The electricity used for hydrogen processing is taken from the grid, and is consequently generated following the current generation mix. The amount of electricity required for this scenario is 114.95 TWh/year. As for scenarios 1 and 3, hydrogen losses are 2.8 % during the transport by truck and 1.1 % during hydrogen dispensing. 18 The

peak of world gas production is forecasted around 2020. That of coal, around 2050 [36].

46

3 The Hydrogen Model

Fig. 3.19 GHG emissions for Scenario 4

The FCEVs considered are also the commercially available ones at present time, so that hydrogen is stored as gas in the vehicle tank. Energy Efficiency and GHG Emissions Considering the production, delivery and vehicle efficiencies in the assessment, global scenario efficiency (Well-To-Wheel) is 8.81 %. The energy requirements at the beginning of the cycle for this scenario are 646.77 TWh, supposing 45.9 % of the primary energy consumption in Spain and almost 2.5 times the current requirements for the road transport model. Such a small efficiency and large energy consumption are mainly due to the high-energy requirements of hydrogen production and processing stages. Furthermore, gasification of biomass and coal is less efficient a process than natural gas reforming or conventional electricity generation. The GHG emissions resulting from this scenario are pictured in Fig. 3.19. As observed, emissions are between 2 and 3 times the reference emissions considered in this study. Emissions remains above the reference emissions, but this scenario supposes an advance with respect to a model based on hydrogen exclusively produced from natural gas. This is mainly due to the usage of biomass, with low or even negative emissions ratio,19 in hydrogen production. The usage of coal, the most pollutant hydrogen generation technology (except if hydrogen is generated by nonrenewable electrolysis, where coal combustion is also involved), constitutes the main contributor in GHG emissions during generation stages. 19 CO 2

emissions during the hydrogen generation may be lower than the CO2 absorbed by the plant during its growth.

3.4 Energy Scenarios for CO2 Equivalent Calculation

47

Fig. 3.20 100 % renewable production

Consequently, this model does not suppose either a solution to the environmental concerns associated to road transport. But as was the case with the previous scenario, it constitutes a useful step toward a clean road transport model. As commented before, it pertains to a mid-term IEA scenario.

3.4.8 Scenario 5. IEA Scenario for Long-Term Implantation This pathway contemplates an entirely renewable energetic model. It is considered by the IEA as the last stage of the hydrogen economy implementation, achieved in the long run [11]. In this scenario, hydrogen is generated exclusively from renewable electricity, and electricity for the delivery phase also comes from a 100 % renewable mix. Production As commented above, hydrogen is completely produced from renewable sources in this scenario. Figure 3.20 shows a possible electricity generation mix for both hydrogen production and processing. It is important to note that hydropower is not incremented from the current generation mix since the number of dams constructed in Spain has now nearly saturated.20 Once generated, whether by centralized or distributed configurations, electricity is sent to the hydrogen production plants through the Spanish electrical grid. In the long run, hydrogen demand would be stable and the large-scale hydrogen production plants would be already operative, as well as the pipeline network. Therefore, the production configuration for this scenario is assumed centralized. Delivery and Storage at Vehicle Hydrogen delivery is performed under liquid phase through a pipeline network (see Fig. 3.21 and Table 3.12). It is also assumed that technical advances in liquid storage allows for the commercialization of FCEVs with tanks for liquid hydrogen instead 20 Maximum

potential about 60 TWh ([37], p. 462).

48

3 The Hydrogen Model

Fig. 3.21 Diagram of hydrogen delivery process through pipelines Table 3.12 Breakdown efficiencies for hydrogen delivery by pipelines

Operation

Efficiency

Gas liquefaction [33] Transport through pipelines [19] Gaseous storage at fueling station [28] Dispensing [28] Total

0.65 0.991 0.997 0.989 0.635

of gaseous vessels. Tank-To-Wheel efficiency of the FCEV would then be 28.2 % (see Table 3.4, p. 29). The electricity required in this pathway for hydrogen liquefaction is 72,68 TWh/year. The generation mix considered can be observed in Table 3.13. For liquid hydrogen pipelines and vehicle tanks, usual hydrogen loss rate due to boil-off is 0.3 % per day [19]. Assuming hydrogen remains three days in the pipeline and seven days in the vehicle tank, total hydrogen diffusivity losses to the atmosphere are therefore 1.2 %. Energy Efficiency and GHG Emissions As in the previous pathways, the global scenario efficiency (Well-To-Wheel) is 7 %. Energy requirements for this scenario are 814.3 TWh, which amounts to 57.8 % of

Table 3.13 100 % Renewable generation mix for Scenario 5 Coal Oil Gas Solar Hydro Biomass 0%

0%

0%

25 %

15 %

25 %

Wind

Nuclear

35 %

0%

3.4 Energy Scenarios for CO2 Equivalent Calculation

49

Fig. 3.22 GHG emissions for Scenario 5

the total primary energy consumption in Spain, and to three times more energy than the one currently used by the Spanish vehicles. The high energy consumption is associated to the low efficiency of electricity generation by renewable sources, close to 40 % on average. This scenario results in the GHG emissions shown in Fig. 3.22. Here, emissions can vary from 0 to 90 % of the current ones, as a consequence of the use of renewable sources. In the best case, this model completely removes GHG emissions. In the worst one, they are only cut by 10 %. The average result, which, let us remind it, is also the most probable, gives an emission cut of 83 %. The efficiency of the scenario depends on the GHG released during the construction of the renewable power plants, and the best practices during infrastructure building should reduce emissions. A potential improvement in electricity generation from renewable sources could allow for a reduction of the number of hydrogen generation power plants, hence GHG emissions.

3.4.9 Summary Each scenario, without exception, implies certain energy requirements, together with the release of a specific amount of greenhouse gases, according to the way this energy is handled.

50

3 The Hydrogen Model

Figure 3.23 pictures altogether the energy requirements and the GHG emissions for all five scenarios. Emissions decrease within the succession Short Term → Mid Term → Long Term envisioned by the IEA, while energy needs are incremented in parallel. The only scenario capable of replacing the current energetic model without increasing emissions, are number 2 and 5, both based on an entirely renewable hydrogen generation. Note also that the worst option consists is choosing the current generation mix to supply an entire vehicle fleet, as it incurs the highest GHG emissions of all scenarios.

Fig. 3.23 Energy requirements and GHG emissions for the five scenarios. The reference for energy requirements is the 2013 electricity production. For GHG emissions, it is the emissions of the current traffic, as evaluated in Chap. 2. The percentages represent the relative gains for the average emissions

3.5 Sensitivity Analysis

51

3.5 Sensitivity Analysis The parameters used in this study are not known with certainty, and they are also susceptible to change in time. To which extent are the previous results altered by a modification of such and such parameter? Sensitivity analyses allow knowing the worth of some parameters in the global cycle. Some of them have a big influence on the final result, meanwhile other parameters can widely fluctuate with no notable consequences. They also quantify the potential effects of the lack of solid knowledge and test the validity of the model assumptions. Rough and unrealistic simplifications may come close to the real results, and further improvements may not bring significantly better accuracy. Furthermore, information about the stability of the model can also be provided by this studies. Technology is being constantly developed and market demand changes every day, so a high sensitivity to changes could suppose a high risk for the energetic model. In return, highly sensitive models easily improve their energetic efficiency and environmental advantages when technical advances are performed. If a parameter P is a function of a variable v, the variation of P resulting from small variations of v can be calculated as follows: P =

∂P v, ∂v

(3.6)

where ∂ P/∂v is the partial derivative of P with respect to v, and v the variation of the variable v. Regarding GHG emissions, the main variables are the GHG emitted during hydrogen generation, and the hydrogen mass required by the fleet (electricity required for hydrogen compression and liquefaction is expressed as a percentage of the energy content of the hydrogen). This mass is the sum of the hydrogen dispensed at hydrogen stations, plus losses during delivery phases. It directly depends on the energy required at wheels by the fleet and on the Tank-To-Wheel efficiency. Taking these points into account, the sensitivity of the model with respect to relative variations of the number of vehicles, efficiencies, and GHG emissions related to hydrogen generation, is reflected in Table 3.14 in terms of the parameters defined in Table 3.15. The full calculations are detailed in Appendix A.4. For example, if the TTW efficiency η FC E V is increased by 5 %, the GHG emissions G H G T O T AL and the energy requirements E Scenario will both change by −1 × 5 %. Analogously, if the efficiency of hydrogen generation ηG E N goes up 5 %, while the number of cars N is also increased by 5 %, the total GHG emissions of the model G H G T O T AL will change by −0.68 × 5 + 1 × 5 = +1.6 %. These figures reveal the importance of the technology involved in the model. For example, a 5 % improvement of the fuel cell efficiency, increasing their performance from 50 to 55 %, would suppose an improvement of the Tank-to-Whells efficiency η FC E V of (0.31 − 0.28)/0.28 = 10 % (see Table 3.4). Consequently, the energy requirements E Scenario , for example, would change by −1 × 10 = −10 %.

52

3 The Hydrogen Model

Table 3.14 Sensitivity analysis A↓ B→ M H2 ηW T W N ηGEN ∗ ηDEL ηFCEV GHGGEN /T J H2

1 0 0 −1 0

0 0.82 0.82 1 0

E Scenario

GHGGEN

GHGDEL

GHGTOTAL

1 −0.82 −0.82 −1 0

1 −1 0 −1 1

1 0 0 −1 1

1 −0.68 0 −1 1

If a parameter A varies by x %, then parameter B varies by y × x %, where y is given by the table. For example, if the number of vehicles N is increased by 5 %, the total GHG emissions GHGTOTAL are increased by 1 × 5 %. All parameters are defined in Table 3.15 Table 3.15 Definition of the parameters used in Table 3.14 N Number of vehicles ηGEN Efficiency of the hydrogen generation process in generation plants ∗ ηDEL Efficiency of the hydrogen delivery process ηFCEV Efficiency of the hydrogen vehicle (Tank-to-Wheel) GHGG E N /T J H2 Tg of CO2 eq when producing 1TJ of hydrogen M H2 Hydrogen mass required by the vehicle fleet ηWTW Well-to-wheels efficiency E Scenario Energy needed by the scenario GHGGEN GHG emissions by hydrogen generation GHGDEL GHG emissions by electricity generation GHGTOTAL Total GHG emissions from the scenario

It can be observed that incrementing the efficiency of hydrogen generation, or that of the whole delivery processes, has less impact than incrementing the TTW efficiency associated to the FCEV. Also, incrementing the efficiency of the delivery process, without accounting for the electricity generation, does not vary the total amount of GHG released to the atmosphere. Why? Because when assessing emissions associated to transport by truck or pipelines, dispensing or storage have been considered negligible (see model simplifications in Sect. 3.4.3). It is important to indicate that this analysis is only valid for small variations of the variable due to intrinsic assumptions of the model (variations are deduced from a partial derivation). This analysis would not be valid for high increments of the variables, and further analysis would be needed. For example, a possible running configuration for the FCEV would be a direct connection between the fuel cell and the electric engine, with an energy saving of 20 % with respect to the configuration considered in this study. In that case, this analysis could estimate neither the variations in the energy requirements nor the GHG emissions.

References

53

References 1. J.M. Ogden, Prospects for building a hydrogen energy infrastructure. Ann. Rev. Energy Environ. 24(1), 227–279 (1999) 2. M.M. Mench, Fuel Cell Engines (Wiley, Chichester, 2008) 3. José Ignacio Linares Hurtado and Beatriz Yolanda Moratilla Soria. El hidrógeno y la energía (2007) 4. J. Jechura, Hydrogen from natural gas via steam methane reforming (SMR) (2015) 5. Información básica de los sectores de la energía 2012. Technical report, Comisión Nacional de la Energía (2012) 6. Board on Energy and Environmental Systems, Comittee on Alternativesand Strategies for Future Hydrogen Production and Use, Division on Engineering and Physical Sciences, National Research Council, and National Academy of Engineering. The Hydrogen Economy: Opportunities, Costs, Barriers, and R&D Needs (National Academies Press, 2004) 7. Intergovernmental Panel on Climate Change (IPCC). Climate Change 2007 - Mitigation of Climate Change: Working Group III contribution to the Fourth Assessment Report of the IPCC. Assessment report (Intergovernmental Panel on Climate Change): Working Group (Cambridge UniversityPress, 2007) 8. P. Jaramillo, C. Samaras, H. Wakeley, K. Meisterling, Greenhouse gas implications of using coal for transportation: life cycle assessment of coal-to-liquids, plug-in hybrids, and hydrogen pathways. Energy Policy 37(7), 2689–2695 (2009) 9. M. Ruth, M. Laffen, T.A. Timbario, Hydrogen pathways: cost, well-to-wheels energy use, and emissions for the current technology status of seven hydrogen production, delivery, and distribution scenarios. Technical report, National Renewable Energy Laboratory (NREL), Golden, CO (2009) 10. M.A. Laguna-Bercero, Recent advances in high temperature electrolysis using solid oxide fuel cells: a review. J. Power Sources 203, 4–16 (2012) 11. T. Riis, E.F. Hagen, P.J.S. Vie, Ø. Ulleberg, Hydrogen production and storage. R&D priorities and gasps. Technical report, International Energy Agency (IEA) (2006) 12. Wikipedia. Public domain—wikipedia, the free encyclopedia (2016) [Online]. Accessed 1 Mar 2016 13. M.A. Hortal, A.L.M. Barrera, El Hidrógeno: Fundamento de un futuro equilibrado (Editorial Díaz de Santos, S.A., 2012) 14. Asociación Española de Pilas de Combustible APPICE. Safety and regulations (2015), http:// www.appice.es/app.php?x=3&x2=5_4. Accessed 05 May 2015 15. R. Luque, Oportunidades de la economía del hidrógeno. Presentation for GENERA 2015, AEH2 (2015) 16. E. Girón, FCH JU under h2020: projects/success stories. Presentation for GENERA 2015, AEH2 (2015) 17. The European Parliament and the Council of the European Union. Directive 2014/94/ue (2014) 18. R. Edwards, H. Hass, L. Lonza, H. Maas, D. Rickeard et al., in WELL-TO-WHEELS Report version 4. a: JEC WELL-TO-WHEELS ANALYSIS (Publications Office of the European Union, 2014) 19. W.A. Amos et al., Costs of storing and transporting hydrogen (National Renewable Energy Laboratory Golden, CO, USA, 1998) 20. L. Ole Valøen, M.I. Shoesmith, The effect of phev and hev duty cycles on battery and battery pack performance, in PHEV 2007 Conference (2007), pp. 4–5 21. Wikipedia. Maximum power transfer theorem, https://es.wikipedia.org/wiki/Teorema_de_ maxima_potencia 22. E.J. Domínguez, J. Ferrer, Sistemas de transmisión y frenado, Ciclos Formativos, Editorial Editex (2012) 23. U.S. Department of Energy. Lower and higher heating values of fuels. Web page: http:// hydrogen.pnl.gov/tools/lower-and-higher-heating-values-fuels. Accessed 08 July 2015

54

3 The Hydrogen Model

24. R. Edwards, J.-F. Larivé, J-C. Beziat, Well-to-wheels analysis of future automotive fuels and powertrains in the european context. Technical Report Version 3c, European Commission Joint Research Centre (2011) 25. Instituto para la Diviersificación y Ahorro de la Energía (IDAE). Consumo de carburante por tipo de motorización, marca y modelo, http://coches.idae.es/portal/BaseDatos/MarcaModelo. aspx 26. Red Eléctrica de España. The spanish electricity system. preliminary report 2014. Technical report, Red Eléctrica de España (2015) 27. National Aeronautics and Space Administration of the United States (NASA). Global water budget, http://neptune.gsfc.nasa.gov/index.php?section=26. Accessed 08 July 2015 28. T. Ramsden, M. Ruth, V. Diakov, M. Laffen, T.A. Timbario, Hydrogen pathways: Updated cost, well-to-wheels energy use, and emissions for the current technology status of ten hydrogen production, delivery, and distribution scenarios. Technical report, National Renewable Energy Laboratory (NREL), Golden, CO. (2013) 29. A.B. Chhetri, R. Islam, Inherently-Sustainable Technology Development (Nova Science, New York, 2008) 30. C. Fernández-Bolaños Badía, R. Velázquez Villa. Energética del hidrógeno. contexto, estado actual y perspectivas de futuro. Master’s thesis, Escuela Técnica Superior de Ingenieros Industriales. Universidad de Sevilla (2005) 31. Eurelectric. Electricity industry trends & figures (2015), http://www.eurelectric.org/electricityindustry-trends-figures/. Accessed 08 July 2015 32. Fundación para estudios sobre la energía. Red alta tensión (2015), http://www. fundacionenergia.es/contenidos.htm. Accessed 09 July 2015 33. Iea energy technology essentials. Hydrogen production & distribution. Technical report, International Energy Agency (IEA) (2007) 34. International Energy Agency. Key World Energy Statistics 2015 (2015) 35. Instituto Nacional de Estadística (INE). Encuesta sobre la estructura de las explotaciones agrícolas - año (2013), http://www.ine.es/dyngs/INEbase/es/operacion.htm?c=Estadistica_C&cid= 1254736176854&menu=ultiDatos&idp=1254735727106. Accessed 010 July 2015 36. G. Maggio, G. Cacciola, When will oil, natural gas, and coal peak? Fuel 98, 111–123 (2012) 37. Instituto para la Diversificación y Ahorro de la Energía. Plan de energías renovables (per) 2011-2020. Technical report (2011)

Chapter 4

The Electrical Model

Besides the possibility of hydrogen-driven vehicles, purely electrical vehicles, where electricity is stored in batteries, can equally be contemplated. Such an option is the focus of this chapter, where the question answered is now: If all current private vehicles in Spain where replaced by electrical ones, how would GHG emissions be changed?

4.1 The Electrical Vehicle When it comes to discussing electrical vehicles, three groups can be singled out. Cable vehicles like trams, which follow fixed routes. They are not the subject of this work. Hybrid vehicles These are a combination of conventional internal combustion vehicles (ICV) and pure electric vehicles (EV). The electric engine functions twofold: on the one hand, it replaces the combustion engine in short sections of urban traffic, and on the other hand, it reinforces the conventional engine. Fuel savings can also be achieved by taking advantage of the vehicle kinetic energy when it slows down (regenerative braking). The batteries of these vehicles can be recharged by a generator driven by the combustion engine or directly from the grid. Since we will focus on 100 % electrical vehicles, we will not detail hybrid ones any further. Electrical vehicles Some of them are powered by hydrogen. They were the focus of Chap. 3. For now, we target 100 % electric vehicles powered by a storage device (battery or capacitor). From the mechanical point of view, they are simpler than internal combustion vehicles (ICV). They can be charged from a charging point (charging station) or even from home. In addition, they incorporate regenerative braking system which, as we shall see, could only amount for a few percent of the recharge (see Appendix A.2). © The Author(s) 2016 J. Montoya Sánchez de Pablo et al., How Green are Electric or Hydrogen-Powered Cars?, SpringerBriefs in Energy, DOI 10.1007/978-3-319-32434-0_4

55

56

4 The Electrical Model

4.1.1 Pros and Cons of Electrical Vehicles The most notable benefits of 100 % electrical vehicles are: • A quick and silent start-up, as an electrical engine provides maximum power from the start. • The absence of a gearbox ensures smooth power delivery and easier driving. • Being simple vehicle with fewer components than conventional cars (clutch, oil, spark plugs, exhaust, timing belt, valves, etc.), they require less maintenance. • Electrical engine are silent and vibrationless. • Electricity generation notwithstanding, emissions of exhaust gases disappear. This reduces urban air pollution compared to internal combustion engines, with some associated public health benefits [1]. • Regenerative braking makes it possible to exploit the vehicle kinetic energy when it slows down. • Better instant response at low speeds, i.e., better throttle response. • High autonomy is not always a critical parameter, because 99 % of daily traveled distances remain below 200 km, that is, the autonomy of an electric vehicle (see Fig. 1.1 p. 2). The main drawbacks of 100 % electrical vehicles are, Lower autonomy than VCI, leading to more frequent recharging. Lower maximum speed than VCI for comparable categories. High cost of the battery, which increases the final price of the vehicles. Eight to ten hours needed to recharge from home, depending on the car and the country (since a lower voltage requires more time). • Limited access to charging points away from home, due to the scarcity of charging station today. • The vehicle itself does not emit GHG, but when accounting for the extra electricity generation, GHG emissions can even be higher than with conventional vehicles. The point of this chapter is precisely to assess this effect.

• • • •

4.1.2 Batteries for EV Batteries are responsible for storing and supplying electric power to the vehicle. Depending on the type of battery, the characteristics of the vehicle vary with parameters such as: useful life (maximum number of complete loading/unloading cycles), voltage, specific energy (amount of usable energy per kg, in Wh/kg), energy density (amount of energy stored per unit volume), etc. The reader will find below a generic classification of the different families of batteries, allowing to understand which kind we focus on, and why.

4.1 The Electrical Vehicle

57

Lead-Acid Batteries They are the most widespread and common in this type of vehicle, due to their low cost. Among their main advantages are their high voltage and power (very useful when accelerating), their long-known technology, easiness of implementation and recycling. Among their disadvantages one can find their low specific energy, moderate useful time, gases escapes, and a strong environmental impact. Nickel-Metal Hydride This second family provides nearly twice as much specific energy as lead–acid batteries. They also allow for faster reloads, good specific energy, and less environmental impact. Their main drawback is their cost, which can quadruple that of lead–acid. Lithium-Ion Batteries They are the most advanced devices and the ones we will consider in this study. They are characterized by high efficiency, high voltage, and high specific energy, both per mass and volume. Also, these batteries are capable of storing twice as much electrical energy as nickel-metal hydrides, and more than four times as much as lead-acid ones. Their useful life is also higher, allowing for a lower environmental impact. However, one cannot ignore their high cost, or their loss of performance at high temperatures (50◦ C) (Table 4.1). In summary, we can conclude that while lead-acid batteries are older and have lower performance, their low cost motivates their use in VE with reduced autonomy. Nickel-metal hydride batteries, although affected by the memory effect, offer better performances and are improving in terms of energy density. Lithium-ion technology is very recent, primarily driven by the mobile phone industry. These batteries offer the best energy density. Yet, they overheat easily and are expensive. Note that the price issue should be mitigated as electric vehicles market expands. Due to their high efficiency and loading/unloading performances, we will consider such devices in the sequel.

Table 4.1 Typical performances of lead-acid, nickel-metal hydride and lithium-ion batteries [2, 3]. Lead-acid NiHM Li-ion Voltage (V) Specific energy (Wh/kg) Volumetric energy (Wh/L) Number of cycles Environmental impact Efficiency charge/discharge (%)

2 10–40 50–100 400–800 High 50–92

1.2 60–80 250 300–600 Low 66

3–4.5 80–170 170–450 500–3000 Moderate–low 80–90

58

4 The Electrical Model

4.2 Electrical Energy Required We now consider the 2011 traffic is replaced by 100 % electrical vehicles, without hybrid or hydrogen cars. How much extra electricity would we have to produce in order to power our electric traffic? In other words, how many Watt-hours of electricity do we need to produce, if we want the wheels of the electrical vehicles to receive the 51.06 TWh computed in Chap. 2? Before the energy makes its way to the wheels of the car, we need to produce it, transport it, charge the car battery, discharge the car battery, run the electrical engine, and transmit the corresponding mechanical energy to the wheels. Each step brings its own losses. We here consider: • Production efficiency: ηgene = 92.7 % (iea.org). Because it takes energy to produce electricity. • Transport efficiency: ηnet = 98.5 % (ree.es). These are the national network losses. • Battery charge efficiency: 90 %, for Lithium-Ion batteries [4, 5]. • Battery discharge efficiency: 90 %, for Lithium-Ion batteries [4, 5]. • Electrical engine efficiency: 90 % [6]. • Mechanical efficiency: as in Chap. 2, we consider 80 % [7]. (The transmission of an electrical slightly differs from that of a conventional one. This is therefore an approximation aiming at simplifying the calculations.) Note that the first two numbers pertain to Spain, while the others are general. The total efficiency, from the production to the wheels, is therefore, η = 0.927 × 0.98 × 0.9 × 0.9 × 0.9 × 0.8 = 53.25 %.

(4.1)

For better clarity, Fig. 4.1 pictures the chain of efficiencies. According to Eq. (2.1) p. 8, the energy provided to the wheels of the current internal combustion vehicles is 51.06 TWh. Hence, we directly derive the required amount of extra electrical energy to power the electrical traffic, Ee,x =

Ew = 95.89 TWh. η

(4.2)

Note that this is already a significant fraction of the 2011 Spanish electricity production, namely 293 TWh (iea.org).

4.2.1 Electrical Versus Internal Combustion Vehicles We here conduct a comparative energy survey between gasoline, diesel, and electric vehicles. We distinguish gasoline from diesel because, as we saw earlier, the diesel vehicle is more efficient than the gasoline one (mainly due to the engine).

4.2 Electrical Energy Required

59

92.7%

Production Losses 7.3%

91.3%

Network Losses 1.5%

Li-Ion

Battery Discharge 90%

Li-Ion

82.2%

Battery Charge 90%

74%

66.6%

Electric Engine 90%

53.25%

Mechanical Efficiency 80%

53.25% to wheels

Fig. 4.1 Efficiency of an electric vehicle (considering losses in power generation and network)

We assume that different engines power the same car, that is, cars with the same design or aerodynamic, and the same weight. The consumption of an internal combustion engine car, gasoline, or diesel, is expressed in liters/100 km, and the one of an electric car in kWh/100 km. To compare them, we need to know the energy content of the fuel, namely, • One liter of gasoline holds about 32.18 MJ. • One liter of diesel fuel holds about 35.86 MJ. For internal combustion vehicles, the average consumption calculated from the official figures published in the Instituto para la Diversificacion y Ahorro de la Energía, are [8], • Average actual consumption of gasoline cars is 7 L/100 km. • Average actual consumption of diesel cars is 6 L/100 km. We therefore find the following energy consumption for internal combustion vehicles, • Gasoline: 7 × 32.18 = 225.26 MJ/100 km. • Diesel: 6 × 35.86 = 215.16 MJ/100 km. Regarding the electric cars, we just saw that reproducing the current traffic would require to charge them with 103.5 × ηgene × ηnet = 94.5 TWh.1 We also checked that with such an amount of energy, our electric cars would be able to cover 1.1 × 1011 (gasoline) + 1.7 × 1011 (diesel) = 2.8 × 1011 km. As a result, the average energetic consumption for these cars is, 1 In

order to compare with the figure we just obtained for internal combustion cars, we need to consider only the amount of electrical energy that would enter the car battery, not the amount that would have to be produced. This means generation and network losses can be bypassed.

60

4 The Electrical Model

• Electrical: 94.5 TWh/2.8 × 1011 km = 121 MJ/100 km. Hence, an electric car consumes around 45 % less than a gasoline car, and 42 % less than a diesel one. The main reason for this difference is the higher efficiency of the electrical engine with respect to internal combustion.

4.3 GHG Emissions In order to gauge the GHG emissions from the production of the extra electricity required by our electric cars, we need to compare the would-be emissions to the current ones. Because we could not find direct data on private cars emissions in Spain, we performed the calculations reported in Chap. 2. The result is 41.77 Tg CO2 eq. How does this number compares to the emissions produced when generating the 95.89 TWh required by the electrical traffic? It clearly depends on the way this electricity is produced. This is why we chose to consider four generation scenarios: Scenario 1 Current mix. Namely, producing the extra electricity according to the current Spanish mix. Scenario 2 100 % fossil fuels. Scenario 3 100 % nonfossil fuels. Scenario 4 Future mix, that is, a mix resulting from the maximum exploitation of the installed power available in Spain. Let us now explain them thoroughly.

4.3.1 Electricity Production Scenario A given scenario is defined by the share αi of each primary energy “i.” For each scenario, the share coefficients αi are given in Table 4.2. They entirely define the electricity generation scenarios. The first scenario or “current mix,” reflects the 2013 Spanish electricity mix, as reported by Red Eléctrica de España [9]. What if the extra 95 TWh required were produced the way electricity was produced in Spain in 2013? As can be seen from the first column of Table 4.2, this model is mainly based on nuclear and wind energy, the rest being devoted to hydro and fossil fuels. The second scenario contemplates a 100 % fossil fuel generation. We arbitrarily choose to simply devote 1/3 of the production to each fossil fuel, although, clearly, many more combinations are possible. Our main goal here is to answer the question: what if the excess electricity was entirely produced with fossil fuels?

4.3 GHG Emissions

61

Table 4.2 Share coefficients αi in % for the 4 scenarios Current mix 100 % fossil Coal Oil Gas Sun Hydro Biomass Wind Nuclear

15 14 10 5 15 2 19 20

100 % nonfossil

Future mix

20 20 20 20 20

17.7 12.41 31.42 1.65 11.45 0.93 12.78 11.67

33 33 33

The third scenario represents the extreme opposite to the previous one. What if the excess electricity was entirely produced by non-GHG emitting technologies? Here also, many mixes are possible, and we simply split 1/5 of the would-be production between sun, hydro, biomass, wind and nuclear. The fourth scenario or “future mix,” pictures the mix resulting from the maximum exploitation of the installed power and will be explained in more details in Sect. 4.5.3. It relies on the installed power and on the capacity factor of each technology. The data required to compute the future mix are reported in Table 4.3. The installed power has been extracted from Ref. [9] and the capacity factors from https://es.wikipedia.org/wiki/Factor_de_planta. Multiplying the former by the latter, and taking the average of the extreme values, we obtain the “future mix” percentages reported in Table 4.2. Note that the maximum production, estimated from 414 to 505 TWh, would be enough to cover the 2011 Spanish electricity production, namely 293 TWh, plus the 95.89 TWh required by the electrical vehicles (see Eq. 4.2). Each of the primary source releases ei grams of CO2 per TWh produced. Because the source “i” has to produce the energy αi × 95.89 TWh, it emits ei × αi × 95.89 grams of CO2 when doing so. The total amount of GHG emitted is then obtained summing over all the sources “i,” that is, GHG emissions =



ei αi 95.89 grams of CO2 .

(4.3)

i

The emission coefficients ei are reported in Chap. 2, on Fig. 2.3 and Table 2.2, p. 11. They represent the amount of GHG emitted by each electricity generation technologies. Together with Eq. (4.3) and Table 4.2, they allow for the calculation of the GHG emissions for each scenario. The results for the four scenarios appear in Table 4.4 and are illustrated on Fig. 4.2. Unless electricity generation is 100 % nonfossil, emissions can be larger than the current ones, for the maximum values of the emissions coefficients. Only the 100 %

62

4 The Electrical Model

Table 4.3 Calculation of the “future mix” Inst. Power (MW) Cap. factor (%) Coal Oil Gas Sun Hydro Biomass Wind Nuclear Total

11,641 10,746 27,206 6,981 19,824 984 22,900 7,866

70–90 60 60 10–15 30 40–60 20–40 60–98

Max. prod. (TWh)

Future mix (%)

71.38–91.78 56.48 142.99 6.12–9.17 52.1 3.45–5.17 40.12–80.24 41.34–67.53 414–505

17.24–18.16 13.64–11.17 34.54–28.29 1.48–1.81 12.58–10,31 0.83–1.02 9.69–15.87 9.99–13.36

The percentages eventually considered are the averages of the numbers in the last column. See also Sect. 4.5.3 Table 4.4 GHG emissions for each scenario, in Tg CO2 eq Scenario Min 25 % Average 1 2 3 4

Current mix 100 % fossil 100 % nonfossil Future mix

19.45 46.68 0.29 26.32

26.9 63.95 1.19 36.48

31.09 73.1 2.03 41.6

75 %

Max

35.36 81.8 3.65 47.49

56.71 119.9 12.58 75.02

Bold numbers are the ones larger than the current private cars emissions, 41.77 Tg CO2 eq (see Chap. 2, Eq. 2.2)

fossil scenario always yields higher emissions. The “current mix” scenario results in less emissions on average, but its upper value exceeds the current ones. Finally, the “future mix” scenario hardly cuts emissions, and could also result in more.

4.4 Fossil Fuels Threshold: Scenario “x” It then appears than our 100 % fossil fuel mix would increase GHG emissions by 75 % (average value). We also find that our 100 % nonemitting mix results in 95 % less emissions. It, then, all-fossil means more emissions, and zero-fossil fuels means less emissions, where is the threshold for even emissions? To analyze this question, we can devise a “shade” of scenarios bridging between the number 2 (all-fossil) and the number 3 (all non-fossil). Suppose scenario “x” is defined as follow: within scenario “x,” the corresponding share αix of each primary energy “i” goes like, (4.4) αix = (1 − x)αi2 + x αi3 ,

4.4 Fossil Fuels Threshold: Scenario “x”

63

140 120

Tg CO2 eq

100 80 +75% 60 -0.41%

40 -25% 20 -95%

0

Current mix

100% non-fossil

100% fossil

Future mix

Fig. 4.2 Box plot of the values reported in Table 4.4. The red line pictures current cars emissions, 41.77 Tg CO2 eq. The percentages represent the relative gains for the average emissions (“average” column in Table 4.4)

where x goes from 0 to 1, and αi2 , αi3 are the shares of the primary energy “i” in scenarios 2 and 3 respectively. For x = 0, we simply have αix = αi2 for all i, so that scenario x reduces to scenario 2 when x = 0. Conversely, x = 1 gives αix = αi3 for all i: scenario x reduces to scenario 3 when x = 1.  Furthermore, we can check scenario x is a “real” scenario, that is, fulfilling i αix = 1. This comes directly from, 

αix =



[(1 − x)αi2 + xαi3 ]

i

i

=



(1 − x) αi2 +

i

= (1 − x)





xαi3

i

αi2 + x

i

= (1 − x) + x = 1,



αi3

i

(4.5)

  since scenarios 2 and 3 are real scenarios, implying i αi2 = i αi3 = 1. Simply put, scenario x is a blend made of (1 − x)% of scenario 2, and x% of scenario 3. How could we compute the GHG emissions resulting from scenario x? Since our calculations are just linear, the resulting emissions can be obtained from the same recipe than the shares. For example, we read from Table 4.4 that the average

64

4 The Electrical Model

emissions of scenario 2 are 73.1 Tg CO2 eq. For scenario 3, they are 2.03. The average emissions of scenario x will then be, GHGav = (1 − x)73.1 + x 2.03 Tg CO2 eq.

(4.6)

The same equation can obviously be written for the minimum, 25th percentile, 75th percentile, and maximum emissions. Each of them will drop linearly with x, from their value in scenario 2 when x = 0, to their value at scenario 3 when x = 1. Figure 4.3 pictures this evolution in terms of x together with the emissions of reference. As expected, average emissions, as well as the others, pass below the reference ones beyond a certain value of x which, therefore, is the threshold we were looking for. Using Eq. (4.6), solving GHGav = 41.77 is straightforward and gives the threshold value for the average emissions, xT ,a = 0.44

(4.7)

The very same method gives the threshold values for the minimum, 25th percentile, 75th percentile, and maximum emissions. They are, xT ,min = 0.1 xT ,25 = 0.35 xT ,75 = 0.51 xT ,max = 0.73

(4.8)

It is tempting to conclude from Eq. (4.7), that the switch to the electrical model becomes profitable beyond an electrical mix containing at least 44 % of non-GHG emitting sources. One must however not forget that our 100 % fossil mix has been arbitrarily attributed 1/3 of each fossil fuel, so that our scenario x does not explore every possible blends. For example, our first scenario, which considers the current mix, brings about less emissions, even with only 40 % (not even 44) of nonemitting sources. But it cannot be written as a linear combination of scenarios 2 and 3. Also, if

100 75th percentile Average

80

25th percentile

60

Minimum

Reference

40 20

Emissions (Tg CO2 eq.)

120

Maximum

0 0

All fossil

0.1

0.2

0.3

0.4

0.5

0.6

x

0.7

0.8

0.9

1

All non-fossil

Fig. 4.3 Minimum, 25th percentile, average, 75th percentile, and maximum emissions for scenario x in terms of x

4.4 Fossil Fuels Threshold: Scenario “x”

65

our scenario 2 had been 100 % gas, we would find a threshold lower than xT ,a = 0.44, since emissions would be lower at x = 0. And so on. A full analysis of this point would imply, among other, the prediction of the most probable fossil mix, which is beyond the scope of this book.2 For now, let us therefore record this figure of a 40–50 % threshold, beyond which the electrical switch is likely to bring less GHG emissions.

4.5 Impact on the Spanish Electricity System So far, we have been assessing the impact of a hypothetical private traffic switch to 100 % electric vehicles. We now look at the consequences it could bear on the electric system. Could the current installations meet the new demand, whether under its current form or within a “future mix”-like version? This section answers this question.

4.5.1 Electricity Demand Electricity consumption in Spain has been declining since 2010, as shown in Fig. 4.4. From the 260 TWh consumed in 2013, a complete electrification of the traffic would demand a 40 % increase of the electricity production. Would the current system be able to cope? An examination of a typical daily load curve can help figuring this out.

4.5.2 Daily Load Curve The electricity demand is usually pictured by the so-called load curve like the one on Fig. 4.5, which shows how consumption evolves during a 24-h time window. It typically displays a peak in the morning, and another one, more pronounced, in the afternoon. The whole system is therefore designed to be able to cope with the maximum daily load, nearly 40 GW on the day displayed on the figure. The all-time record was achieved on December 17, 2007 at 6:50 pm, with 45.3 GW [9]. While households’ consumption nearly drops to zero during the night, most of the industry needs electricity 24/7. This is why the load curve never falls to zero. On December 1, 2015 displayed on Fig. 4.5, it reached a minimum at 4:30 am, with only 23.3 GW required.

2 And

probably bound to fail anyway, like Vaclav Smil wrote in 2005 “for more than 100 years long-term forecasts of energy affairs…have, save for a few proverbial exceptions confirming the rule, a manifest record of failure” ([10], p. 121).

66

4 The Electrical Model

Fig. 4.4 Electricity consumption in Spain, in TWh. Source Red Eléctrica de España 38.000 36.000

Load (MW)

34.000 32.000 30.000 28.000 26.000 24.000 22.000 20.000

Fig. 4.5 Typical daily load curve in Spain. This one is from December 1, 2015 8:00 pm, to December 2, 2015 8:00 pm. The red triangle can be used to estimate the energy that could be produced in addition. Source Red Eléctrica de España

By subtracting the value of the load curve at night to, say, 40 GW, one could get an estimate of the power that could be produced in excess. Integrating the result over a year would then give the amount of additional electrical energy that could be generated by the current system during a year. Let us instead perform a crude estimate of this quantity. We start picturing, schematically, the extra electrical energy that could have been produced by the area A of the red triangle. Considering it is 15 h wide and 14 GW high, we find A = 15 × 14/2 = 105 GWh.

(4.9)

4.5 Impact on the Spanish Electricity System

67

Multiplying this result by 365, we find an annual potential for an additional production of, 105 × 365 = 38 TWh. (4.10) This falls short of the 95.89 TWh required by Eq. (4.2) to power the electrical traffic. Because we miss a factor 2.5, an accurate calculation of the potential additional production, would obviously lead to a similar conclusion (a more accurate calculation may differ by ±20 %, not 250 %). We therefore come to the conclusion that supplying the extra electricity needed would require more than the current system. Yet, in its current mode of operation, this current system does not run at its maximal capacity. Even if doing so would pose serious security issues (which is why it is not happening), let us try to evaluate the capability of current installations, if they were pushed to their limit.

4.5.3 Installed Capacity in Spain and “Future Mix” Scenario The total installed power in Spain in 2013 was 108,148 MW, as shown on the first column of Table 4.5. Figure 4.6 pictures the 2013 mix of installed power, evidencing the important role of natural gas, wind, and hydroelectricity. How is it then that according to Table 4.2, Nuclear Energy, for example, was responsible for 20 % of the 2013 production, while it is far from having the largest installed power? It turns out that an installed capacity of 1 MW in wind energy is useless when there is no wind, whereas a nuclear power plant operates almost all year long. The difference between the potential and the actual production of a given technology, is measured by the capacity factor. Given an installed power P delivering the energy E per year, the capacity factor C reads, E . (4.11) C= P × 3600 × 24 × 365

Table 4.5 Installed power, annual generation, and capacity factor in Spain in 2013 [9] Installed power (MW) Generation (TWh) Capacity factor (%) Coal Oil Natural gas Solar Hydroelectric Biomass Wind Nuclear

11,641 10,746 27,206 6,981 19,824 984 22,900 7,866

42.38 39.29 28.98 12.95 41.30 5.02 54.30 56.38

41.56 41.74 12.16 21.18 23.78 58.24 27.07 81.82

68

4 The Electrical Model

Fig. 4.6 2013 mix of installed power. Source Red Eléctrica de España Table 4.6 Maximum capacity factors Coal Oil Gas 70–90 % 60 % 60 %

Solar 10–15 %

Hydro 30 %

Wind 20–40 %

Nuclear 60–98 %

Source https://es.wikipedia.org/wiki/Factor_de_planta

It is therefore a measure of the fraction of the time spent producing energy. For the Spanish network in 2013, the installed power, the annual generation and the resulting capacity factors, are shown for each production technology in Table 4.5 (last column). One can see how wind energy came just behind nuclear in terms of annual production, in spite of nearly three times more installed power. This is simply one of the many aspects of wind intermittency. Now comes a question closely related to our problem: what if the same installed capacity was pushed to the limit? Could the resulting generation provide enough electricity to power our electric cars? In other words, which are the maximum production for each technology? The answers appears in Table 4.6. From there, one can straightforwardly compute that the maximum generation ranges from 414 to 505 TWh, already displayed in Table 4.3. The amount of TWh generated by each technology in this regime, defines our “Future Mix” scenario.

4.6 Sensitivity Analysis In view of the large amount of coefficients involved in the calculation, we now conduct a sensitivity analysis in order to check how the final result varies in term of the data involved.

4.6 Sensitivity Analysis

69

We consider the share coefficients in Table 4.2 are fixed, since they have been either chosen arbitrarily, or simply extracted from the statistics of Red Eléctrica de España. Also, the sensitivity analysis with respect to the emission coefficients reported in Table 2.2 is built-in in the present work. For the remaining coefficients, the question answered here is the following: given a relative variation of one parameter, all others being kept constant, what is the corresponding relative variation of the GHG emissions? The starting point of the calculation is the energy to wheels Ew , for gasoline and diesel cars, first computed in Sect. 2.1, Eq. (2.1). The required electrical energy Ee,x is then computed in Sect. 4.2 through, Ew Ee,x =  , j ηj

(4.12)

where the ηj ’s are the efficiencies involved in the calculation of η in Eq. (4.1), and  η denotes the product of the ηj ’s. Finally, GHG emissions are given by Eq. (4.3). j j Gathering these equations, we can write, Ew  GHG emissions =  αi ei . j ηj i

(4.13)

Here, we get from Sect. 2.1, Ew = ηg ηm Et,gas + ηd ηm Et,die = ηm (ηg Et,gas + ηd Et,die ),

(4.14)

where ηm is the mechanical efficiency, and ηg and ηd are the gasoline and diesel engines efficiencies respectively. Noteworthily, because one of the ηj ’s in Eq. (4.13) is the mechanical efficiency ηm , from the engine to the wheels, we find the emissions are eventually independent of this parameter. For the other coefficients ηj ’s, namely, the electricity production and transport efficiencies, the battery charge and discharge efficiencies, and the electrical engine efficiency, differentiating Eq. (4.13) with any of the ηj ’s gives, δηj0 δGHG =− GHG ηj0

(4.15)

where ηj0 is the efficiency ηj under scrutiny. We thus find that any relative variation of these parameters translates to the same relative variation of the emissions (opposite in sign). It turns out that besides the mechanical efficiency ηm which cancels out in the calculation, the other parameters are all larger than 90 %. As a consequence, their relative variation cannot exceed 10 %. For example, a technological improvement

70

4 The Electrical Model

shifting battery charge efficiency from 90 % to 99 % would result in a relative variation of this coefficient of 10 %. We can thus conclude that our final result is stable, within a 10 % accuracy, with respect to any variation of the parameters defined in Sect. 4.2. The same can be said for the variations of Ew , since Eq. (4.13) allows to write directly, δEw δGHG = . GHG Ew

(4.16)

References 1. J. Woodcock, P. Edwards, C. Tonne, B.G. Armstrong, O. Ashiru, D. Banister, S. Beevers, Z. Chalabi, Z. Chowdhury, A. Cohen, O.H. Franco, A. Haines, R. Hickman, G. Lindsay, I. Mittal, D. Mohan, G. Tiwari, A. Woodward, I. Roberts, Public Health Benefits of Strategies to Reduce Greenhouse-gas Emissions: Urban Land Transport. The Lancet 374(9705), 1930–1943 (2009) 2. Fundación de la Energía de la Comunidad de Madrid. Guía del Vehículo Eléctrico, Capítulo 6 (2009) 3. J. Sun, Car Battery Efficiencies, Submitted as coursework for Physics 240 (Stanford University, Cambridge, 2010) 4. Y. Idota, T. Kubota, A. Matsufuji, Y. Maekawa, T. Miyasaka, Tin-based amorphous oxide: a high-capacity lithium-ion-storage material. Science 276(5317), 1395–1397 (1997) 5. S.J. Gerssen-Gondelach, A.P.C. Faaij, Performance of batteries for electric vehicles on short and longer term. J. Power Sour. 212, 111–129 (2012) 6. R. Bargalló, J. Llaverías, H. Martín, El vehículo eléctrico y la eficiencia energética global, in 11th Portuguese-Spanish International Conference of Electrical Engineering (2011) 7. E.J. Domínguez, J. Ferrer, Sistemas de transmisión y frenado (Ciclos Formativos, Editorial Editex, 2012) 8. Instituto para la Diviersificación y Ahorro de la Energía (IDAE). Consumo de carburante por tipo de motorización, marca y modelo, http://coches.idae.es/portal/BaseDatos/MarcaModelo. aspx 9. REE. The spanish electricity system 2013. Technical report, Red Eléctrica de España (2014) 10. V. Smil, Energy at the Crossroads: Global Perspectives and Uncertainties (MIT Press, Cambridge, 2005)

Chapter 5

Conclusion

How should we then consider the potential benefits of an electric or hydrogen switch? A serious answer, pondering most of the energetic issues involved, is “neither always better nor worse. It depends”. The law of energy conservation is merciless. Neither electricity nor hydrogen are energy sources, because there are no electricity or hydrogen wells on Earth. Hydrogen and electricity are energy vectors. So if our cars were to run on electricity or hydrogen, the energy content of these vectors would have to come from somewhere else. That is, from a true primary energy source, namely, fossils fuels, nuclear, wind, solar, hydro, geothermal energy, etc. In a world like the current one, where more than 80 % of the 2013 primary energy, and nearly 70 % of the electricity, comes from fossil fuels [1], the sources filling the vectors would most likely be fossils. In this event, Figs. 3.23 p. 50 and 4.2 p. 63, for hydrogen and electricity respectively, show that the switch to the new models would probably not result in lower GHG emissions. In Spain, emissions would be slightly lower producing the electricity required by the electrical model with the current Spanish mix. But the Spanish mix is far cleaner than the World mix, where 41 % of the electricity still comes from coal [1]. Regarding the hydrogen model, making it the way it is currently made would definitely result is drastically higher emissions. Yet, we need to remind the potential public health benefits of the switch [2], as pollution would be removed from the cities, home of 7 out of 10 Spaniards [3]. Why are these results so disappointing, especially when it comes to hydrogen? One factor could be summarized this way: “hydrogen doesn’t like to move”. Moving hydrogen means compressing it, or liquefying it, each option taking its own energy toll. Leakage during transportation, nearly impossible to cancel with hydrogen, also result in losses. Besides Scenario 2 of our hydrogen model, all of them require transportation. Another factor is simply the way hydrogen is generated. And besides Scenarios 2 and 5, all of them require fossil fuels (see Table 3.6, p. 31). Finally, hydrogen also needs storage plus conversion to electricity inside the vehicle, which results in even more losses. © The Author(s) 2016 J. Montoya Sánchez de Pablo et al., How Green are Electric or Hydrogen-Powered Cars?, SpringerBriefs in Energy, DOI 10.1007/978-3-319-32434-0_5

71

72

5 Conclusion

In contrast, there are no oil losses when storing it, or transporting it. This is the reason why in spite of the far better efficiency of the electrical engine when compared to the internal combustion one (90 % vs. 35 or 45 %), the overall energy needed to power our alternatives models is not necessarily smaller. To which extent are our conclusions generalizable to other countries? Clearly, one can simply adapt our calculations to Germany, France or Brazil without any technical hurdles. However, some conclusions can be drawn immediately by simply looking at the electricity production mix of a given country. This is true in particular for the electrical model. The difficulty of the hydrogen model comes from the fact that there are many ways to produce hydrogen. Hydrogen can be produced directly from fossil fuels, renewable sources, or from electricity. And electricity, in turn, can be produced in many ways. But even in India or China, where more than 70 % of the electricity comes from coal, a switch to Scenario 2 of the hydrogen model (home generation from solar PV) would be profitable. A straightforward adaptation of our calculations is therefore intricate. In contrast, the electrical model only needs more electricity. Here, the tendency obviously highlighted by our analysis, is that switching to an entire electrical traffic is all the more profitable than the country’s electrical mix is non-emitting. China, for example, with 75 % of its electricity from coal, is quite close to our 100 % fossil scenario. Here, it is nearly certain that any switch to the electrical model, without a reform of the electrical mix, would give the same outcome than our electrical scenario number 2: far more GHG emissions than before. The idea is therefore to analyze the current mix of a country, in terms of its share of nonemitting technologies, namely sun, hydro, biomass, wind, and nuclear. What is the percentage of these 5 within the production mix? Figure 5.1 answers this question for a number of countries. We can here recall the result of Sect. 4.4 on Scenario “x”: a switch to the electrical model in Spain would bring less emissions provided the mix contains more than 40– 50% of non-emitting sources. Such a conclusion is not directly applicable to any country, because we also need to know about the emissions of the current traffic in the same country. But a quick look at Fig. 5.1 suggests that if the current mixes could be maintained under the electric model, the switch would likely bring less emissions in Spain, Portugal, France, Denmark, Canada, or Brazil. And more emissions in Greece, Japan, India, China, and South Africa. Of course, some detailed calculations adapted to the country would be needed to confirm, or not, these hints. It turns out that the main lesson we can draw from this study, is that the electric or hydrogen cars only keep their promises in a world which is yet to come. For the hydrogen model, only scenarios 2 and 5 result in drastically reduced emissions. In scenario 2, because everyone charges up his/her car at home, from solar energy. And in scenario 5, because the electricity needed to centrally produce hydrogen is 100 % emissions-free. Regarding the electric model, the only scenario tremendously cutting emissions is number 3, where the required electricity is also 100 % emissions-free.

Gas

Coal

Non-emitting

Non-emitting Coal Gas Oil

Fig. 5.1 Electricity mix in 2013 for various countries, in terms of share between fossil fuels and non-emitting sources (sun, hydro, biomass, wind and nuclear). Source International Energy Agency

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Oil

5 Conclusion 73

74

5 Conclusion

The success of the implementation of either model is therefore tied to the “decarbonization” of our primary energy sources. On the one hand, the global shift to non-emitting energies that is bound to occur within the next decades, could drag the implantation of electric and hydrogen cars. But on the other hand, the importance and the dynamism of the transport sector could equally drag the de-carbonization of the other energy sectors. Let us hope such a virtuous circle will promptly be initiated.

References 1. International Energy Agency. Key World Energy Statistics 2015. 2015 2. J. Woodcock, P. Edwards, C. Tonne, B.G. Armstrong, O. Ashiru, D. Banister, S. Beevers, Z. Chalabi, Z. Chowdhury, A. Cohen, O.H. Franco, A. Haines, R. Hickman, G. Lindsay, I. Mittal, D. Mohan, G. Tiwari, A. Woodward, I. Roberts, Public Health Benefits of Strategies to Reduce Greenhouse-Gas Emissions: Urban Land Transport, vol 374(9705) (The Lancet, London, 2009), pp. 1930–1943 3. Fundación BBVA. La población en españa: 1900-2009. Technical report, 2010

Appendix A

A.1

Notations, Acronyms, and Units

Powers of 10 We use the usual notations for the powers of 10. “M” for Mega, like in Mega Joules (MJ): 106 . “G” for Giga, like in Giga Joules (GJ): 109 . “T” for Tera, like in Tera Watt-hour (TWh): 1012 . Energy Energetic content of oil, gas, or coal are usually given in Joules, or Mega Joules. Electricity production is usually given in Watt-hour (Wh). One Watt-hour is the energy produced by a source of power 1 W during 1 h, that is, 3,600 J. Emissions Carbon dioxide, CO2 , is the main greenhouse gases (GHG) emitted by the transportation sector. However, it is not the only one. In order to keep measuring emissions by one single number, climate scientists have devised a way to compare the global warming effect of CO2 , with that of, say, methane.1 As a result, a certain amount of emissions of carbon dioxide, plus methane, plus any other GHG, can be expressed by a single number of emissions in “grams CO2 equivalent”. In this work, emissions are interchangeably expressed in “grams CO2 equivalent” or “grams CO2 ” because most of the emissions contemplated are nearly pure CO2 . Acronyms Many acronyms appear in these pages. Though we do not use them in a systematic manner, here is a dictionary explaining their meaning:

1 See

Sect. 8.7 of Ref. [1].

© The Author(s) 2016 J. Montoya Sánchez de Pablo et al., How Green are Electric or Hydrogen-Powered Cars?, SpringerBriefs in Energy, DOI 10.1007/978-3-319-32434-0

75

76

Appendix A ATR CSS EV FCEV GHG HEV HFCEV ICV IDEA IEA INE LPG ORR PEM PEMFC POX PSA REE SMR TTW UNFCC

A.2

Autothermal reforming Carbon capture and storage Electrical vehicle Fuel cell electrical vehicle Greenhouse gases Hybrid Electric Vehicle Hybrid Fuel Cell Electric Vehicles Internal combustion vehicle Instituto para la Diversificación y Ahorro de la Energía International Energy Agency Instituto Nacional de la Estadistica Liquefied petroleum gas Oxidizer reduction reaction Proton Exchange Membrane Proton Exchange Membrane Fuel Cell Partial Oxidation Pressure Swing Adsorption Red Eléctrica de España Steam methane reforming Tank to wheels United Nations Framework Convention on Climate Change

Regenerative Braking

Regenerative braking in electric vehicle is the conversion of the kinetic energy of a slowing down vehicle, into chemical energy. This energy is stored in batteries to be used later in propulsion. The kinetic energy of a vehicle in motion is given Ec = 21 mv 2 , where m and v are its mass and velocity, respectively. Let us here conduct a back-of-the-envelope assessment of the amount of energy recoverable this way, in order to decide whether or not we should take it into account in our study. Suppose an electric vehicle with a mass 1,000 kg circulates at 100 km/h (28 m/s). Its kinetic energy is Ec = 0.4 MJ. Suppose it stops. Its kinetic energy goes down to 0 MJ. Assuming such deceleration occurs some 20 times over a 100 km journey, the recoverable kinetic energy is 20 × (0.4 − 0) = 8 MJ. This number represents only 7 % of the energy consumed by the electric vehicle over 100 km (121 MJ/100 km—see Sect. 4.2.1). For this percentage to reach 50 %, the car would have to stop 20 × (50/7) = 142 times in 100 km. In addition, this calculation has been performed assuming 100 % of the kinetic can be tapped, and neglecting aerodynamic losses, losses of heating pads or brake disc, emergency braking where the kinetic energy is so great that the system cannot absorb the energy in so little time, and so on.

Appendix A

A.3

77

Estimation of CO2 Emissions Associated to Hydrogen Transport by Truck

One of the assumptions of our assessment in Chap. 3 is that GHG emissions associated to hydrogen delivery by trucks are negligible. In this appendix, this assumption is justified with an estimation of the GHG emissions released to the atmosphere during the transport phase by trucks. The delivery is considered to be performed in conventional fossil fuel trucks. As truck engines typically run with diesel, the total emissions correspond to the product of the truck fleet consumption, by the CO2 emissions of diesel fuel. Let us consider that the tank of a truck carries 3,370 kg of hydrogen [2]. We determined in Sect. 3.3.1, p. 29, that the entire Spanish vehicle fleet would require 6 × 109 kg of hydrogen per year. The number of one-way trips needed to deliver such an amount, from the production to the distribution sites, is: N = 6 × 109 /3370 = 1.8 × 106 .

(A.1)

Assuming each trip is 100 km long, the total amount of kilometers travelled by the trucks would be, (A.2) D = 2N × 100 = 3.56 × 108 km, where the factor 2 accounts for the fact that the trucks need to return to the production site after delivering. Assuming a truck consumption of 15 L/100 km, then the total amount of diesel fuel burnt a year to deliver the hydrogen is, A = D × 15/100 = 5.34 × 107 L.

(A.3)

Finally, we know from Chap. 2 that burning 1 L of diesel fuel releases 2.63 kg of CO2 eq. Burning the amount of diesel fuel mentioned above would therefore result in the emission of, 5.34 × 107 × 2.63 = 1.4 × 108 kg = 0.14 Tg of CO2 eq.

(A.4)

We thus find that the emissions resulting from the delivering of the hydrogen represents only 0.3 % of the 41 Tg emitted by the current traffic model (see Chap. 2). We can therefore safely neglect them. A similar calculation could be conducted for the electric model. Within the current model, trucks have to deliver the fuel at the gas stations. But in our electric model, such deliveries would not occur, since we assumed the cars would be charged at home. One could therefore argue that in order to compare GHG emissions from both scenarios, the emissions from the deliveries that would not occur in the electric scenario have to be accounted for. Yet, here also, a brief estimation shows they are negligible anyway.

78

A.4

Appendix A

Sensitivity Analysis for the Hydrogen model

We here study how parameter variations are related between them in the hydrogen model. The equations determining the effect of relative variations are derived from Eq. (3.6 p. 51).

A.4.1

Energy at Wheels

Applying Eq. (3.6) to Eq. (3.3) p. 29, we get Ewheels = kmtravelled mfuel/km ηTT Wf ρfuel Hcfuel N

(A.5)

where kmtravelled is the distance travelled by one vehicle in one year, mfuel/km the consumption of fossil fuel per km, ηTT Wf the Tank-To-Wheel efficiency of the conventional engine, ρfuel the density of the fossil fuel, Hcfuel the heat of combustion of the fuel (LHV) and N the number of vehicles circulating in Spain. Then, dividing by Eq. (3.3): N Ewheels (A.6) = Ewheels N

A.4.2

Hydrogen Mass Required

We apply Eq. (3.6) to Eq. (3.3): kmtravelled m ˙ fuel/km ηTT Wf ρfuel Hcfuel HcH2      N 1   × − 2 N , ηFCEV  + η η

MH2 =

FCEV

(A.7)

FCEV

where HcH2 is the heat of combustion of hydrogen (LHV) and ηFCEV the Tank-ToWheel efficiency of the FCEV. Then, dividing by (3.3), we obtain,    ηFCEV  N MH2 +  . =− MH2 ηFCEV  N

(A.8)

Appendix A

A.4.3

79

Well-To-Wheel Efficiency

The expression for the WTW efficiency is: ηW T W = ηGEN

Ewheels Ewheels + ∗ ηDEL ηFCEV

Eelec

(A.9)

,

∗ denotes the efficiency of the delivery process without taking where ηGEN and ηDEL into account the energy required for hydrogen compression and liquefaction, and Eelec the energy in term of electricity required for hydrogen compression and liquefaction. This last term can be calculated as follows:

Eelec = ψ

MH2 , HcH2

(A.10)

where ψ is an energetic coefficient that determines the percentage of the equivalent energy content necessary to compress or liquefy hydrogen. For example, for ψ = 0.3, the energy needed to liquefy or compress hydrogen is 30 % of the hydrogen energy content. In this study, for the whole delivery process ψ is equal to 0.55 if hydrogen is both liquefied and compressed, or to 0.35 if hydrogen is only liquefied. Sensitivity analysis is only performed for ψ = 0.55. Hydrogen losses to the atmosphere are not considered. Applying Eq. (3.6) to (A.9), we obtain,

ηW T W =

1 1 ∗ ηGEN ηDEL

+ψ



ηFCEV + ηFCEV

+

1 2 ∗ ηGEN ηDEL ∗ ηGEN ηDEL +

1 ∗2 ηGEN ηDEL ηFCEV ∗ ηGEN ηDEL +

ψ

ηGEN ψ 

∗ ηDEL .

(A.11)

Dividing by (A.9), the relative variation of the WTW efficiency is, ηFCEV ηW T W ηW T W = + ∗ ηW T W ηFCEV ηGEN ηDEL ηFCEV



ηGEN η ∗ + ∗DEL ηGEN ηDEL



(A.12)

∗ The factor ηGEN ηηWTW can be expressed as a function of ψ, ηDEL , and ηGEN ∗ DEL ηFCEV considering Eqs. (A.5, A.9, A.10). Therefore,

ηW T W 1 = ∗ ∗ ηGEN ηDEL ηFCEV 1 + ψ ηGEN ηDEL

(A.13)

80

Appendix A

Several calculations during the study estimated that ηGEN = 0.426 (on average), ∗ ηDEL = 0.969 (on average), and ψ = 0.55. Therefore, ηW T W 1 = = 0.82 ∗ ηGEN ηDEL ηFCEV 1 + 0.426 0.969 0.55

A.4.4

(A.14)

Energy Requirements

The expression for the energy required by scenario is, Escenario =

Ewheels ηW T W

(A.15)

Applying Eq. (3.6), we obtain, Escenario =

   Ewheels   ηW T W Ewheels + − 2 ηW T W ηW T W  1

(A.16)

Considering the results obtained in the calculations above, we get for the relative variation of the energy requirements of the scenario,    ∗ ηGEN ηFCEV Escenario N ηW T W ηDEL + = + + ∗ ∗ Escenario N ηFCEV ηGEN ηDEL ηFCEV ηGEN ηDEL (A.17)

A.4.5

Greenhouse Gas Emissions

Due to the complexity of the emissions estimation (there are two different sources: hydrogen production plants and electricity power plants), this estimation is performed considering every emissions source separately and integrating them later into a global figure. In the case of emissions released during hydrogen production: GHGGEN =

ghgGEN MH2 TJ H2 HcH2 ηGEN

(A.18)

GEN is a factor that denotes the amount of greenhouse gases released to where ghg TJ H2 the atmosphere during the production of 1 TJ of hydrogen. Applying Eq. (3.6) and dividing by (A.18), the relative variation of the GHG during hydrogen generation is,

Appendix A

81

    GEN  ghg N  ηFCEV   ηGEN  GHGGEN TJ H − + = ghgGEN2 + + − GHGGEN N ηFCEV   ηGEN  TJ H

(A.19)

2

Analogously, for power plants,

GHGDEL = ψ

ghgGEN MH2 %mix, ηFCEV HcH2 kW h

(A.20)

GEN where ghg is a factor that denotes the amount of greenhouse gases released to the kW h atmosphere during the production of 1 kWh of electricity, and %mix is the participation of a certain generation technology in the total amount of electricity produced. Applying Eq. (3.6) and dividing by (A.20), the relative variation of the GHG during hydrogen generation is,

ghgGEN   %mix N  ηFCEV   GHGDEL h + ghgkW + − = .  GEN GHGDEL N ηFCEV %mix kW h

(A.21)

To estimate the relative variation of the total GHG emissions, the simplest way is to calculate the variation with respect to emissions during delivery and hydrogen production: GHGTOTAL = GHGGEN + GHGDEL ⇒  GHGTOTAL =  GHGGEN +  GHGDEL .

(A.22)

The relative variation is therefore: GHGTOTAL GHGGEN GHGDEL =α +β GHGTOTAL GHGGEN GHGDEL

(A.23)

−1 −1   GHGGEN GHGDEL where α = 1 + GHG and β = 1 + . In this study, the average GHGDEL GEN results for GHG emissions during production and delivery phases are 57.3 MtCO2 eq and 27.5 MtCO2 eq, respectively. Therefore, the values of α and β are: α= β=

1 1+

GHGDEL GHGGEN

1+

GHGGEN GHGDEL

1

=

1 = 0.68 27.5 1 + 57.3

=

1 = 0.32 57.3 1 + 27.5

(A.24)

82

Appendix A

References 1. IPCC, Climate Change 2013 - The Physical Science Basis (Cambridge University Press, Cambridge, 2014) 2. M.M. Mench, Fuel Cell Engines (Wiley, New York, 2008)