Energy Markets - Investment, Competition, and Regulation [1 ed.] 9783428526338, 9783428126330

Citation preview

Applied Economics Quarterly Supplement

2007

Beihefte der Konjunkturpolitik Issue 58 Energy Markets – Investment, Competition, and Regulation

Edited by Wolfgang Franz Rainer Winkelmann and Klaus F. Zimmermann

asdfghjk Duncker & Humblot · Berlin

Energy Markets – Investment, Competition, and Regulation

Applied Economics Quarterly Supplement Beihefte der Konjunkturpolitik Issue 58

Energy Markets – Investment, Competition, and Regulation

Edited by

Wolfgang Franz Rainer Winkelmann and Klaus F. Zimmermann

asdfghjk Duncker & Humblot · Berlin

Bibliographic information published by Die Deutsche Bibliothek Die Deutsche Bibliothek lists this publication in the Deutsche Nationalbibliografie; detailed bibliographic data is available in the Internet at .

Technical editor: Deborah Anne Bowen All rights reserved. No part of this book may be reproduced, translated, or utilized in any form or by any means, electronic or mechanical, without the expressed written consent of the publisher. # 2007 Duncker & Humblot GmbH, Berlin Typesetting and printing: Berliner Buchdruckerei Union GmbH, Berlin Printed in Germany ISSN 1612-2127 ISBN 978-3-428-12633-0 Printed on no aging resistant (non-acid) paper ∞ according to ISO 9706 *

Internet: http://www.duncker-humblot.de

Editorial This supplement to Applied Economics Quarterly reports on the 70th Annual Meeting of the Association of German Economic Research Institutes (ARGE-Institute), which took place in Berlin on April 19, 2007. The topic was “Energy Markets – Investment, Competition, and Regulation.” Energy markets pose a major challenge for the economy, politics, and society: the entire structure of power generation is on the verge of a fundamental renewal. By dealing with the important current question of how regulation will affect investments in generation and transmission capacities within liberalized electricity and gas markets, the conference offered useful contributions for policy advice to the federal government. Keynote speaker Carl Christian von Weizsäcker discussed investments in energy markets, while Jean Michel Glachant, the second keynote speaker, considered the choice between a “Vibrant European Market and Industry or a Bunch of National Energy Champions.” The other lectures focused on alternative strategies for renewable energies, biofuels as a climate policy strategy, as well as innovation, competition, and asymmetric strategic investment. We are very grateful to the German federal government – in particular the Federal Ministry of Economics and Technology – for their support and their constructive contributions to the discussion at the conference. We would like to thank Ulf Moslener and Tim Mennel (ZEW Mannheim) for the conceptual preparation of the conference, the organizers, Anke Winkler and Ralf Messer (ARGE-Institute), and all the participants and attendees for their contributions. Next year’s annual meeting is scheduled for April 9, 2008, in Berlin and will deal with “Regulation and Competition in Europe”. Berlin, June 2007

Wolfgang Franz Rainer Winkelmann Klaus F. Zimmermann

Contents Christoph Böhringer, Tim Hoffmann, and Thomas F. Rutherford Alternative Strategies for Promoting Renewable Energy in EU Electricity Markets . . .

9

Comment: Christian von Hirschhausen . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

27

Gernot Klepper Biofuels and Climate Policy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

31

Hans-Dieter Karl Investments of the German Electricity-Supply Industry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

47

Comment: Ingo Ellersdorfer, Stephan Kempe, Uwe Remme, Markus Blesl, Ulrich Fahl, and Alfred Voß . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

61

Reinhard Madlener and Pio Baake Asymmetric Strategic Investment Behavior in Network Industries: the Case of Natural Gas Distribution in Norway . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

69

Comment: Till Requate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

83

Franziska Holz How Dominant is Russia on the European Natural Gas Market? Results from Modeling Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

85

Comment: Ulf Moslener . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103 Participants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107

Alternative Strategies for Promoting Renewable Energy in EU Electricity Markets By Christoph Böhringer*, Tim Hoffmann**, and Thomas F. Rutherford***

Abstract We investigate the economic impacts of promoting electricity produced from renewable energy sources within the EU. We focus on two central regulatory instruments: Feed-in tariffs, i.e., direct subsidies to electricity production from renewable energy, and quota obligation systems with tradable green certificates. Based on a large-scale partial equilibrium model of the EU electricity market calibrated to empirical data, we find that differentiated feed-in tariffs incur substantial excess cost compared to an EU-wide tradable green quota.

1. Introduction Political support for renewable energy technologies has a history of over 30 years within the EU. Key motives as well as favored policies and measures to promote the market penetration of electricity from renewable energy sources have changed over time. The first major impetus for the promotion of renewable energies can be traced back to the oil crises in the early 70ies and 80ies: Renewable energy from EUinternal sources was seen as a long-term substitute for exhaustible and mainly imported fossil fuels in order to secure EU-wide energy supply. A second central push is linked to the negative environmental externalities associated with the combustion of fossil fuels. In the mid-80ies environmental concerns were related to local and regional problems of air quality and acidification. These problems were handled largely through end-of-pipe technologies for electricity production from coal but also provided political support to renewable energies. Much more substantial had been and still are the implications of anthropogenic carbon emissions from * University of Oldenburg and Centre for European Economic Research (ZEW), Mannheim, Germany; [email protected]. ** Centre for European Economic Research (ZEW), Mannheim, Germany; hoffmann@ zew.de. *** Ann Arbor, MI, U.S.A.; [email protected].

10

Christoph Böhringer, Tim Hoffmann, and Thomas F. Rutherford

fossil fuel combustion for global warming: Carbon-free energy supply technologies are considered as the central response to cope with the problem of global warming in the long run. More recently – and complementary to energy security as well as environmental objectives – “green” policy makers push renewable energy in order to create new jobs and strengthen competitiveness of the EU economy in terms of lead technologies that can be exported to world markets. As to policy measures for the promotion of renewable energies there had been a shift from command-and-control policies to market-based instruments such as taxes, subsidies, and tradable quotas: Energy taxation in many EU countries meanwhile comes along with tax breaks or tax exemptions to renewable energy working as implicit subsidies to correct relative prices with respect to concerns on energy security and environmental quality. In addition, direct subsidies for renewable energy are warranted – typically differentiated by the type of green energy, i.e., hydropower, wind, biomass, solar, etc. A relatively new strand of policy regulation is the use of tradable green quotas where energy suppliers are required to produce a certain share of energy services from renewable energy but are flexible to trade these shares between each other in order to exploit differences in marginal compliance costs. Against this background, we investigate the economic impacts of promoting electricity produced from renewable energy sources within the EU. Strategies for “greening” electricity production have been examined in several previous papers. Menanteau et al. (2003) highlight the potential efficiency gains of tradable green certificates vis-à-vis feed-in tariffs or tendering systems (see also Finon and Menanteau 2003 and Kühn 2000). Voogt et al. (2000) and Uyterlinde et al. (2003) discuss the cost-effectiveness of tradable green certificates with a focus on risk and transaction cost. Jensen and Skytte (2002) point out difficulties in the manageability of quota obligation schemes when a regulator has additional policy targets than just greening electricity. Morthorst (2000 and 2001) as well as Madlener and Stagl (2000) address the issue of regulatory overlap between renewable promotion schemes and emission allowance trading. In our applied analysis, we focus on the cross-comparison of two alternative policy instruments which are central to the contemporary EU strategy for the promotion of renewable energy in electricity production: Feed-in tariffs, i.e. direct subsidies to electricity production from renewable energy, and quota obligation systems with tradable green certificates (TGC). Based on a large-scale partial equilibrium model of the oligopolistic EU electricity market calibrated to empirical data, we find that differentiated feed-in tariffs incur substantial excess cost as compared to an EU-wide tradable green quota. In general terms, these excess cost can be interpreted as the price tag that policy makers have to attach to other objectives than the pure greening of electricity. Such objectives might include pursuits to reduce additional market failures associated with market barriers to specific infant renewable technologies, knowledge spillovers from private R&D, or strategic industrial policies.

Alternative Strategies for Promoting Renewable Energy in EU Electricity Markets

11

The remainder of this paper is as follows: Section 2 provides a brief summary of the EU policy initiatives for the promotion of renewable energy in electricity production. Section 3 describes the numerical framework and the database underlying our quantitative analysis. Section 4 presents illustrative policy scenarios and discusses the results. Section 5 concludes.

2. Policy Background: Promoting Renewables in Europe The EU Directive on the “promotion of electricity produced from renewable energy sources in the electricity market” (RES-E) is the main legislation for supporting renewable energy at the EU level (European Commission 2001). The objective of the Directive is a doubling of the share of renewable energy in Europe’s gross energy consumption from approximately 6 % in 1997 to 12 % in 2010. The latter has been translated into a specific share of 22.1 % renewable energy in final EU electricity consumption in 2010 (as compared to 14 % in 1997). This target has been already set in the 1997 White Paper on renewable energy sources (European Commission 1997) and endorsed by the Energy Council in May 1998. Beyond an aggregate policy target, the RES-E Directive specifies indicative targets for the penetration of RES-E for each EU Member State (see column “RES-E target” in Table 1). To date, EU Member States employ a myriad of support schemes (see Figure 1). Some of them directly stimulate the supply side of renewable electricity, while others directly affect the demand side. Promotion schemes can be furthermore distinguished according to the supported activity, i.e., either capacity installation or the generation of green electricity. A recent survey published by the European Commission (European Commission 2005e) reveals that feed-in tariffs are the most common promotion measure (in seven out of the EU-15 Member States) followed by quota obligation systems with tradable green certificates, so-called TGC (in four out of the EU-15 Member States). In contrast, tender schemes, investment subsidies, and fiscal measures only play a minor role (see also Table 1). Under pure economic efficiency considerations, the promotion of RES-E – alike other regulatory measures – has to be justified by market failures, i.e. the inability of markets to internalize all the social benefits and social costs associated with economic activities. For efficient internalization, each market failure typically goes along with one specific regulatory instrument (e.g. emission taxes, R&D subsidies, definition of property rights, or product liability rules). Regarding policy support for renewable energy, however, there is not a single clear-cut justification: Promotion of renewable energies may contribute to ameliorate simultaneously various externalities (such as market barriers to infant renewable technologies, undersupply of public knowledge spillovers from private R&D, or environmental externalities) while – at the same time – serving strategic interests (such as industry or competition policies).

12

Christoph Böhringer, Tim Hoffmann, and Thomas F. Rutherford Table 1 Indicative RES-E Targets in 2010 and National Support Schemes RES-E target [in %]

Feed-in tariffs

Austria Belgium Denmark Finland France Germany Greece Ireland Italy Luxembourg Netherlands Portugal Spain Sweden United Kingdom

78.1 yes 6.0 29.0 yes 31.3 21.0 yes 12.5 yes 20.1 yes 13.2 (announced) 25.0 yes (for PV) 5.7 yes 9.0 yes 39.0 yes 29.4 yes 60.0 10.0

European Union

21.7

National support schemes Quota obligaOther tions / TGC yes

Minimum price for renewables Tender schemes for offshore wind Tax exemptions and investment incentives

Investment incentives Tender schemes yes

Investment incentives yes yes

Source: EU 2005e.

Source: Uytelinde et al. 2003.

Figure 1: Classification of RES-E Policy Support Mechanisms

Alternative Strategies for Promoting Renewable Energy in EU Electricity Markets

13

Given exogenous RES-E targets, a tradable quota system will in principle assure cost-efficiency and targeted effectiveness without the need for central planner information. A feed-in tariff system, on the other hand, would require perfect information on all technologies, their costs and potentials, price developments on the electricity market, consumer preferences, etc.1 However, feed-in tariffs allow for a differentiated treatment of alternative renewable technologies taking into account other objectives than just the “greening” of electricity production. As a matter of fact, feed-in tariff systems in policy practice stand out for a pronounced discrimination across different green technologies: Usually, more costly technologies such as solar or geothermic energy receive higher subsidies than more competitive renewables such as hydro or wind power.2 If the policy objective was simply the greening of electricity production, such a differentiated feed-in tariff system is likely to create substantial excess costs.3 In contrast, quota obligation systems with tradable green certificates (TGCs) can meet overall national or supra-national targets in an efficient way. The quota system creates an explicit market for the “greenness” of electricity with an associated market price (or “green value”). The market will then sort out which type and quantity of renewable energy will serve most efficiently the RES-E targets. At a supra-national level, an EU-wide market for TGCs allows for “where”-flexibility in complying with national targets from an aggregate perspective, i.e. cost-efficiency across domestic borders. 3. Numerical Framework 3.1 Model Summary

Our quantitative assessment of different RES-E promotion strategies is based on a static large-scale partial equilibrium model of an oligopolistic European electricity market where firms compete strategically for market shares on regional electricity markets. Let R be the set of all regions (with index r 2 R), F the set of all firms (with index f 2 F), and I the set of all generation technologies (with index i 2 I). Within a region r a firm f disposes of a fixed generation capacity composed of specific generation technologies. Electricity is supplied to regional electricity markets which are distinguished by residential and industrial consumers. Industrial consumers face differentiated pricing over their load curve. In contrast, residential 1 Against this background it is not surprising, that – according to recent studies (e.g. Uyterlinde et al. 2003) – most of the Member States who employ feed-in tariff systems will not reach their indicative targets until 2010. 2 For example, the German Renewable Energy Sources Act sets a feed-in premium for solar electricity which is roughly four times higher than for electricity produced from biomass (Deutscher Bundestag 2001). 3 Again, the excess costs may be interpreted as an additional premium that policy makers have to attach to other objectives than the pure greening of electricity.

14

Christoph Böhringer, Tim Hoffmann, and Thomas F. Rutherford

consumers – who are usually supplied on the basis of long-term power purchase contracts – pay a flat fee for electricity delivered at any point of their load-curve. For the sake of simplicity, we only distinguish two load segments l – one for demand in base-load (l = base) and one for peak-load (l = peak) demand – in our present application; sInd f ;r;l then denotes supply of firm f in load segment l to industrial consumers in region r and sRes f ;r denotes the supply to residential consumers Res for the electricity prices to the indusrespectively. Adopting notations pInd r;l and pr Res trial and residential markets and notations DInd for the associated (isor;l and Dr elastic) demand functions, each firm f maximizes its profits f : max

sInd ; sRes f ;r;l f ;r

f ˆ

  XXh
Res Ind Res
sInd pInd DRes
sf ;r r;l Dr;l f ;r;l ‡ pr r r

 i Res : Cf ;r;l sInd f ;r;l ; sf ;r

l

Res Differentiating for sInd f ;r;l and sf ;r , we obtain the usual first-order conditions for a Cournot oligopoly where marginal values of supply to industrial and residential customers …wf ;r;l † equal marginal revenues:

wf ;r;l ˆ X l

Ll

pInd r;l


wf ;r;l ˆ

Ind f ;r;l


1

pRes r




! 8f ; r; l

Ind r;l

1

Res f ;r


P

Res r

L l l

! 8f ; r

Res where Ind denote the price elasticities of the demand segments, Ind r;l and r f ;r;l Ind Ind Res Res L (ˆ sf ;r;l =Dr;l ) and f ;r (ˆ sRes f ;r =Dr ) are the associated market shares, and l indicates the shares of base- and peak-load demand in total electricity demand (with Lbase ‡ Lpeak ˆ 1).

Domestic demand can be met by domestic electricity generation or by electricity imports from other regions. Transmission and distribution of electricity is priced with (exogenous) grid charges. Cross-border electricity trade is limited by capacity constraints of inter-regional exchange points. Costs for inter-regional electricity exchange thus account for the scarcity of exchange capacities on top of transmission and distribution margins. Plant-specific capacity limits impose an upper bound on the feasible amount of electricity production by each firm. Furthermore, suppliers are obliged to keep back a certain level of reserve capacity (determined as a fraction of total electricity supply in each region). With respect to our policy analysis, electricity producers face minimum targets for the deployment of RES-E capacities which are implemented by alternative regulatory regimes: Countries can promote RES-E production by subsidizing RES-E production on a per-unit basis (feed-in tariffs). RES-E subsidies are then paid by consumers through a tax on electricity consumption. Alternatively, regulatory authorities may impose specify domestic quotas for RES-E production with or without the option of inter-regional tradability.

Alternative Strategies for Promoting Renewable Energy in EU Electricity Markets

15

Numerically, our model is formulated as a mixed complementarity problem (MCP). The algebraic formulation is implemented in GAMS (Brooke, Kendrick and Meeraus 1987) using PATH (Dirkse and Ferris 1995) as a solver. A comprehensive algebraic model description is provided in the Appendix.

3.2 Parameterization

The model analysis in this paper is based on a parameterization of 15 EU regions.4 Benchmark electricity demand for each region is obtained from recent UCTE (2005) and NORDEL (2005) statistics. The disaggregation of aggregate regional demands into residential and industrial demands is based on detailed energy balance data from IEA / OECD (2002). In addition, we employ detailed statistics on hourly load values provided by international associations (UCTE, NORDEL) and national grid operators in order to determine the load-specific demand for both demand segments in each region. Regional electricity prices stem from the 4th Benchmarking Report of the European Commission (European Commission 2005a). The supply side of the model covers more than 1100 conventional thermal power plants. Each plant is owned by one of more than 220 explicit firms. Information on the installed capacity of each plant and on the ownership structure is taken from a comprehensive power plant database that covers all EU model regions (Meller et al. 2005). Technical as well as economic information on the power plants comes from the IKARUS data base (KFA 1994) providing detailed data on installation costs, operating and maintenance costs, and thermal efficiencies. We map the technologies described by Meller et al. to a set of 11 selected IKARUS technologies (covering fossil fuel-fired and nuclear power plants) and then apply dynamic investment calculus in order to obtain technology-specific electricity production costs. Fuel prices and data on labor costs for this calculation are obtained from recent EU statistics (European Commission 2005b, European Commission 2005c, European Commission 2000d, Eurostat 2005). Technology-specific carbon emission coefficients are again based on IKARUS. Our characterization of 16 RES-E technologies builds upon ADMIRE-REBUS data for renewable energy in Europe (Uyterlinde et al. 2004).5 We distribute the derived cost-potential curves for renewable energy technologies in each region across firms according to their shares of conventional capacity in regions’ total generation capacity. Information on inter-regional electricity trade and exchange capacities is based on recent statistics of system operators and associations (European Transmission 4 Austria, Belgium, Denmark, Finland, France, Germany, Greece, Ireland, Italy, Luxembourg, Portugal, Spain, Sweden, the Netherlands, and the United Kingdom. 5 Each renewable energy technology is thereby divided into “technology-segments” to account for different site qualities (e.g., differences in wind speed) and availability of specific renewable fuels (e.g., wood or other biomass).

16

Christoph Böhringer, Tim Hoffmann, and Thomas F. Rutherford

System Operators – ETSO 2001a, European Transmission System Operators – ETSO 2001b, NORDEL 2005a, NORDEL 2005b, UCTE 2005).

4. Policy Scenarios and Results 4.1 Policy Scenarios

We compare the economic impacts of promoting renewable energy in EU electricity production against a business-as-usual situation (scenario BaU) without political support to RES-E production. Regulation to foster electricity production from renewable energy sources is captured by four illustrative policy scenarios FEED_D, FEED_H, QUOTA_R and QUOTA_EU. Scenario FEED_D mimics the situation where Member States adopt technology-specific feed-in tariffs for RES-E at a level which assures compliance with the national RES-E targets. The feed-in tariffs are financed by an ad-valorem tax on the electricity sales to all consumers in the respective regions. This scenario reflects the present situation in most of the EU-15 Member States. Scenario FEED_H describes a regulation where Member States employ harmonized feed-in tariffs in their region, i.e., each renewables technology receives the same premium. As in scenario FEED_D, the feed-in tariff is endogenously determined to assure compliance with the national RES-E targets and gets financed by an electricity sales tax. In scenario QUOTA_R, EU Member States achieve their indicative RES-E targets through a national quota system imposed on electricity producers. This scenario implies a uniform national certificate price. The economic impacts under QUOTA_R can be expected to be rather similar to those of scenario FEED_H as the latter mainly differs with respect to the explicit cross-financing of feed-in tariffs. Scenario QUOTA_EU represents a completely harmonized situation where green quotas are traded across EU Member States. Regional targets can either be met by domestic production of RES-E or by importing TGCs. Likewise, TGCs may be sold to the international market if domestic costs for RES-E are lower than the international price. Hence, this scenario warrants a single EU-wide value for green certificates to meet the overall RES-E target of approximately 22 % in a cost-efficient way. Table 2 provides a brief summary of the four alternative green promotion schemes under consideration in our quantitative analysis.

Alternative Strategies for Promoting Renewable Energy in EU Electricity Markets

17

Table 2 Summary of Policy Scenarios Feed-in tariff schemes FEED_D FEED_H Promotion scheme

Financing of promotion

Harmonized green value Trade in TGC’s

Quota systems QUOTA_R QUOTA_EU

Regional and technologyspecific feed-in tariffs Through electricity tax

Harmonized feed-in tariff in all regions

Regional quota

Regional quotas and trade in TGCs

Through electricity tax

Through the electricity market

No

Yes (national / regional) No

Yes (national / regional) No

Through the electricity market and the TGC market Yes (EU-wide) Yes

No

4.2 Simulation Results

Figure 2 displays the technology mix in EU-wide electricity production for the four alternative RES-E promotion schemes as well as the reference situation (BaU). Administered additional RES-E production mainly displaces the use of coal: The share of coal (hard coal, soft coal and lignite) in total production of EU-15 countries decreases by approximately 10 % as compared to BaU. Yet, nuclear and coal-fired power plants still provide more than 50 % of the total electricity supplied to consumers. The differential impacts of the alternative promotion schemes are clearly visible in the deployment of RES-E technologies. Figure 3 shows the shares of green technologies in total green electricity production. In the BaU scenario, hydropower accounts for over 85 % of total RES-E production; only a small fraction of electricity is produced from wind, the remaining 13 % stem from biomass and waste. When technology-specific feed-in tariffs are adopted (scenario FEED_D), hydropower still accounts for the major share of green production (45 %); however, the RES-E mix exhibits much more technological diversity: Electricity from biomass and waste incineration now constitutes roughly 36 % of green production; even rather costly solar potentials are utilized (approximately 3 %). Under FEED_H and QUOTA_R the technological diversity prevails, but especially (onshore) wind power benefits from uniform regional green values at the expense of waste and biomass. Solar potentials are no longer exploited since regional green values are not high enough for solar electricity production to break even. Scenario QUOTA_EU implies an equalization of marginal costs of RES-E production across EU Member States thereby ensuring that the most profitable RES-E potentials are used: Compared to scenarios FEED_H and QUOTA_R wind potentials in France, Greece (onshore) and the Nordic region (onshore and offshore) will be used in addition – mostly at the expense of electricity production from waste incineration.

18

Christoph Böhringer, Tim Hoffmann, and Thomas F. Rutherford

Share in total production [in %]

100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% BaU

FEED_D

FEED_H

QUOTA_R

QUOTA_EU

Figure 2: Technology Mix of European Electricity Supply Share in total RES-E production

100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% FEED_D

FEED_H

QUOTA_R

QUOTA_EU

Figure 3: Technology Mix of European Green Electricity Supply

The administered increase of green production leads to higher electricity prices: Initially unprofitable capacities are phased-in, thereby substituting initially more profitable technologies. The increase in electricity depresses electricity demand. In the case of differentiated feed-in tariffs (scenario FEED_D) EU-wide electricity

Alternative Strategies for Promoting Renewable Energy in EU Electricity Markets

19

consumption decreases by 7.4 % as compared to the BaU level. The switch to uniform feed-in tariffs (scenario FEED_H) slightly ameliorates the increase in electricity prices and the associated decrease in electricity consumption at the EU level (–6.2 % from BaU levels). The upper part of Table 3 reports minimum and maximum values across EU Member States for the per-unit support level (green value), the direct costs of RESE promotion (indicating the total support level as the product of the green value and the subsidized RES-E production), and the efficiency costs (measured as the loss in consumer and producer surplus from BaU levels). The lower part of Table 3 provides the EU-wide aggregate of direct costs and efficiency costs. Table 3 Green Values and Induced Costs FEED_D

Green Values Direct costs (Variance)

A / MWh mill. A

Efficiency costs (Variance) mill. A

FEED_H

QUOTA_R

Min

Max

Min

Max

Min

37.1

108.4

28.9

104.0

28.9

Max 99.3

QUOTA_EU Min 44.5

Max 44.5

403.6 6,385.8 296.9 6,122.0 298.5 5,728.0 154.1 3,050.7 330.4 4,993.7

15.5 5,027.8

15.0 4,421.5

26.5 3,016.0

Total Direct costs EU-15

mill. A

19,537.9

16,936.4

16,487.1

14,057.0

Efficiency costs EU-15 mill. A

17,855.5

15,099.0

14,687.9

13,816.3

Obviously, green values differ across regions whenever inter-regional TGC trade is not possible. In scenario FEED_D, regional green values range between 37.1 A / MWh up to 108.4 A / MWh, total direct costs are 19.5 bn. A, and the loss in overall economic surplus amount to 17.9 bn.A vis-à-vis BaU. With uniform feed-in tariffs (scenario FEED_H) green values decrease at the regional level compared to FEED_D: Green values vary from 28.9 A / MWh up to 104.0 A / MWh across regions. EU-wide direct costs of RES-E promotion decrease by ca. 2.6 bn. A, and efficiency gains of FEED_H vis-à-vis FEED_D amount to 2.8 bn. A. Implementation of a regional tradable quota system – captured by QUOTA_R – leads to similar results as FEED_H. EU-wide trade in TGCs – facilitated under QUOTA_EU – reduces overall EU compliance costs significantly as the most profitable RES-E potentials are exploited. The direct costs of RES-E promotion are 14.0 bn. A, the loss in economic surplus amounts to 13.8 bn. A vis-à-vis BaU. Hence, QUOTA_EU reaches the overall EU target at approximately 20 % lower economic costs than FEED_D.

20

Christoph Böhringer, Tim Hoffmann, and Thomas F. Rutherford

5. Conclusions The political support for electricity produced from renewable energy sources has a long history within the European Union. At present, EU Member States employ a range of support schemes. Most common are feed-in tariff systems, i.e., direct subsidies to electricity production from renewable energy, and quota obligations with tradable green certificates. In this paper, we have compared the economic consequences of feed-in tariffs and tradable quota systems for the promotion of renewable energy sources in EU electricity production. Our quantitative simulations based on a large-scale partial equilibrium model of the EU electricity market indicates that differentiated feedin tariff schemes may incur substantial excess cost compared to regionally and EU-wide harmonized systems. If the “greening” of electricity was the only political objective, an EU-wide tradable green quota would reach the European RES-E target at 20 % lower costs than national feed-in tariff systems with technologyspecific premia. The higher costs of feed-in tariff systems can be interpreted as the price tag that policy makers have to attach to other objectives than the pure greening of electricity. As a consequence, policy makers should clearly lay out the multiple objectives and the respective weights that can justify discriminatory pricing across renewable energies. References Brooke, A. / Kendrick, D. / Meeraus, A. (1996): GAMS: A User’s Guide, GAMS Development Corp. Deutscher Bundestag (2001): “Gesetz für den Vorrang Erneuerbarer Energien,” Bundesgesetzblatt, Jg. 2004 Teil I Nr. 40, pp. 0033 – 0040. Dirkse, S. P. / Ferris, M. C. (1995): “The PATH solver: A non-monotone stabilization scheme for mixed complementarity problems,” Optimization Methods and Software, 5, 123 – 156. European Commission (1997): “Energy for the future: renewable sources of energy – White Paper for a Community strategy and action plan, COM(97) 0599 final,” Communication from the Commission. – (2001): “Directive 2001 / 77 / EC of the European Parliament and of the Council of 27 September 2001 on the promotion of electricity produced from renewable energy sources in the internal electricity market,” Official Journal of the European Communities. – (2003): “Directive 2003 / 54 / EC of the European Parliament and of the Council of 26 June 2003 concerning common rules for the internal market in electricity and repealing Directive 96 / 92 / EC,” Official Journal of the European Communities. – (2005a): “Annual Report on the Implementation of the Gas and Electricity Internal Market, COM(2004) 863 final,” Communication from the Commission. – (2005b): Gas Prices, Price Systems 2004, Luxembourg: Office for Official Publications of the European Communities.

Alternative Strategies for Promoting Renewable Energy in EU Electricity Markets

21

– (2005c): “The market for solid fuels in the Community in 2003 and 2004, SEC(2005) 105” Commission Staff Working Document. – (2005d): “The support for electricity from renewable energy sources – Impact Assessment, Annex to the Communication from the Commission COM(2005) 627 final, SEC(2005) 1571 final,” Communication from the Commission. – (2005e): “The support of electricity from renewable energy sources, COM(2005) 627 final,” Communication from the Commission. European Transmission System Operators – ETSO (2001a): “Definitions of Transfer Capacities in liberalised Electricity Markets, Final Report April 2001”. – (2001b): Procedures for Cross-Border Transmission Capacity Assessments, Oktober 2001”. Eurostat (2005): Online Data 2005: available at: http: // epp.eurostat.cec.eu.int / portal / page? _pageid=1996,45323734&_dad=portal&_schema=PORTAL&screen=welcomeref&open= / &pro-duct=EU_MAIN_TREE&depth=1. Finon, D. / Menanteau, P. (2003): “The static and dynamic efficiency of instruments of promotion of renewables,” Energy Studies Review, 12(1), 53 – 82. IEA / OECD (2002): Energy Balances of OECD Countries, 2002, Paris: OECD Publications. Jensen, S. G. / Skytte, K. (2002): “Interaction between the power and green certificate markets,” Energy Policy, 30, 425 – 435. KFA – Forschungszentrum Jülich (1994): IKARUS-Instrumente für Klimagas Reduktionsstrategien, Teilprojekt 4: Umwandlungssektor Strom- und Wärmeerzeugende Anlagen auf fossiler und nuklearer Grundlage. Teil 1 u. 2 . Kühn, I. (2000): “Ist die Quotenregelung mit Zertifikatshandel für Erneuerbare Energien effizient? – Eine Replik,” Zeitschrift für Energiewirtschaft, 24(4), 214 – 216. Madlener, R. / Stagl, S. (2000): “Promoting Renewable Electricity Generation through Guaranteed Feed-in Tariffs vs Tradable Certificates: an Ecological Economics Perspective,” in “Proceedings of the 3rd Biennial Conference of the European Society for Ecological Economics”. Meller, E. / Milojcic, G. / Reichel, W. / Schöning, G. (2005): Jahrbuch der europäischen Energie und Rohstoffwirtschaft 2004, Essen: Verlag Glückauf. Menanteau, P. / Finon, D. / Lamy, M.-L. (2003): “Prices versus quantities: Environmental policies for promoting the development of renewable energy,” Energy Policy, 31(8), 799 – 812. Morthorst, P. E. (2000): “Scenarios for the use of GHG-reduction instruments – how can policy-instruments as carbon emission trading and tradable green certificates be used simultaneously to reach a common GHG reduction target?,” Energy & Environment, 11(4). – (2001): “Interactions of tradable green certificate market with a tradable permits market,” Energy Policy, 29(5). NORDEL (2005): Annual Report 2004, Annual Statistics 2004. UCTE (2005): Oline data (year 2004).

22

Christoph Böhringer, Tim Hoffmann, and Thomas F. Rutherford

Uyterlinde, M. A. / Daniels, B. W. / de Noord, M. / de Vries, H. J. / de Zoeten-Dartenset, C. / Skytte, K. / Meibom, P. / Lescot, D. / Hoffmann, T. / Stronzik, M. / Gual, M. / del Rio, P. / Hernández, F. (2003): “Renewable electricity market developments in the European Union; Final Report of the ADMIRE REBUS project,”. Voogt, M. / Boots, M. G. / Schaeffer, G. J. / Martens, J. W. (2001): “Renewable electricity in a liberalised market – the concept of green certificates,” Energy & Environment, 11(1).

Appendix: Algebraic Model Description Our model is formulated as a mixed complementarity problem (MCP). Two classes of conditions characterize the market equilibrium: zero profit conditions and market clearance conditions. The former class determines activity levels (quantities) and the latter determines prices. The economic equilibrium features complementarity between equilibrium variables and equilibrium conditions: activities will be operated as long as they break even, positive market prices imply market clearance otherwise commodities are in excess supply and the respective prices fall to zero. Table 4 depicts the sets, parameters and variables of the model.

Table 4 Sets, Variables and Parameters Sets: R

Set of all regions (with index r 2 R where rs 2 R is an electricity exporting and rd 2 R an importing region)

F

Set of all firms (with index f 2 F)

I

Set of all generation technologies (with index i 2 I)

IR…R†

Subset of all RES-E technologies (with index i 2 IR)

P

Set of all power plants of technology type i in region r con-trolled by firm r (with index p 2 P)

Parameters: p0Ind r;l

Reference electricity price in region r on the industrial market for electricity demand in load segmet l

p0Res r

Reference electricity price in region r on the residential market

D0Ind r;l D0Res r;l

Reference electricity demand of industrial customers in region r and load segment l Reference electricity demand of residential customers in region r and load segment l

Ind r;l

Price elasticity of industrial demand in region r and load segment l

Res r;l

Price elasticity of residential demand in region r

Alternative Strategies for Promoting Renewable Energy in EU Electricity Markets

23

ci;r;l

Variable production costs of plant of technology type i in region r and load segment l

Kp

Generation capacity limit of plant p

T rs;rd

Capacity limit of all inter-regional exchange points between region rs and region rd

rmr:l

Regional reserve requirements in region r and load segment l

gcr

Regional charges for distribution of electricity in region r

tfrs;rd

Charges for inter-regional electricity transmission from region rs to region rd

dlrs;rd

Fraction of distribution losses of electricity exchange from region rs to region rd

cci

Specific carbon coefficient for electricity generation from technology i

CLr

Upper bound on carbon emissions in region r

ai

Adjustment factor for technology-specific feed-in tariff for RES-E technology i 2 IR

rqr

Minimum shares of renewable electricity in the total supply to region r

Price variables: pInd r;l

Price for electricity in region r on the industrial market in load segment l

pRes r

Price for electricity on the residential market in region r

wf ;r;l

Marginal value of electricity supply by firm f in region r and load segment l

r;l

Shadow value on reserve capacity constraint in region r and load segment l

p

Shadow price on capacity constraint of plant p in load segment l

r

Shadow value on the emissions constraint in region r

r

Shadow value on the renewables quota in region r



Price of tradable green certificates

rs;rd;l

Shadow price on transmission capacity between adjacent regions rs and rd Electricity tax in region r

r

Activity levels: SInd f ;r;l

Supply of firm f in load segment l to industrial customers in region r

SRes f ;r

Supply of firm f to residential customers in region r

Xp;l

Electricity production of plant p in load segment l

Zp;l

Set-aside capacity provision of plant p in load segment l

Ef ;rs;rd;l

Electricity trade by firm f from region rs to region rd

GEX r

Green certificates exports of region r to the international market

GIM r

Green certificates imports of region r from the international market

Zero-profit conditions:

  Zero-profit condition for industrial supply ? SInd f ;r;l : wf ;r;l ‡ gcr ‡ rmr;l
r;l 

pInd r;l




1

Ind f ;r;l Ind r;l

! :

24

Christoph Böhringer, Tim Hoffmann, and Thomas F. Rutherford


Zero-profit condition for residential supply ? SRes f ;r : X
 Ll
wf ;r;l ‡ rmr;l
r;l ‡ gcr  pr


P

Res f ;r


1

L l l

!

Res r

l

:


Zero-profit condition for reserve capacity provision ? Zp;l : p  r;l :


Zero-profit condition for inter-regional electricity trade ? Ef ;rs;rd;l : wf ;rs;l ‡ tfrs;rd;l ‡

X

rs;rd;l


rd;rs;l  wf ;rd;l 1


tlrs;rd :

rs;rd

Zero-profit conditions for electricity production in case of a quota obligation
System ? Xp;l : ci;r;l ‡ p;l ‡ cci
r ‡ r
rqr  wf ;l;r ‡ r ci;r;l ‡ p;l ‡ cci
r ‡ r
rqr  wf ;r;l

8i 2 IR ;

8i 2 = IR :

Zero-profit conditions for electricity production in case of a feed-in tariff system
? xp;l : ci;r;l ‡ p;l ‡ cci
r  wf ;r;l ‡ ai r ci;r;l ‡ p:l ‡ cci
r  wf ;r;l

8i 2 IR ;

8i 2 = IR :

Additional zero-profit conditions for international trade in TGCs:
Zero-profit condition for green certificates imports ? GIM : r   r :


: Zero-profit condition for green certificates exports ?GEX r r   :

Market-clearance conditions:

  Market-clearance condition for industrial supply ? pInd r;l : X f

Ind SInd f ;r;l ˆD0r;l


…1‡ 6 r †
pInd r;l p0Ind r;l

!Ind r;l :

Alternative Strategies for Promoting Renewable Energy in EU Electricity Markets

25

Market-clearance condition for residential supply ? pRr es †: X

SRes f ;r ˆ

X

2

…1‡ 6 r †
pRes r

4D0Res
r;l

f

p0Res r

l

3 !Res r;l 5:


Market-clearance condition for electricity trade ? rs;rd;l : T rs;rd 

X

X

Ef ;rs;rd;l

f ;rs

Ef ;rs;rd;l :

f ;rd


Market-clearance condition for reserve capacity ? r;l : X

Zf ;r;i;l  rmr;l


f ;i

 X L Res : SInd f ;r;l ‡ l
Sf ;r f


Market-clearance condition for electricity production ? wf ;r;l : X

Xp;l ‡

X 

X


 dlrs;r
Ef ;rs;r;l

1

rs…rs6ˆr†

p

rd…rd6ˆr†

L Res Ef ;r;rd;l  SInd f ;r;l ‡ l
Sf ;r :


Market-clearance condition for electricity production capacity ? p : KP 

X


Xp;l ‡ Zp;l :

l

Market-clearance condition for emission constraint …? r †: CLr 

X


cci
Xp;l :

p;l

Market-clearance condition for renewable quota …? r †: X

Xf ;r;i;l

f ;i2IR;r;l

IM GEX r ‡ Gr  rqr


 X L Res : SInd f ;r;l ‡ l
Sf ;r f ;l

Additional market-clearance condition for ad-valorem electricity tax in case of feed-in tariff systems …? r †: X l

pInd r;l

r




X

! SInd f ;r;l

‡ pRes r


r




X

f

SRes f ;r 

f

X


Xf ;r;ir;l
ai
r :

f ;ir;l

Additional market-clearance condition for international trade in TGCs …? †: X r

GEX r 

X r

GIM r :

26

Christoph Böhringer, Tim Hoffmann, and Thomas F. Rutherford

Market shares and load share:

  Market shares in industrial markets ? Ind f ;r;l : SInd f ;r;l Ind f ;r;l ˆ P Ind : f Sf ;r;l

  Market shares in residential markets ? Res f ;r;l : SRes f ;r;l res ˆ : P f ;r;l Res f Sf ;r;l

Load shares in residential demand left…? Ll †:  D0Res r;l
Ll ˆ

P

Res l D0r;l

…1‡




Res r †
pr

Res r;l

p0es r …1‡

Res r †
r

p0Res r

: Res r;l

Alternative Strategies for Promoting Renewable Energy in EU Electricity Markets Comment By Christian von Hirschhausen* 1. Main Arguments of the Paper What are the costs for different RES-E promotion strategies in an European context? The paper “Alternative Strategies for Promoting Renewable Energy” discusses alternative instruments to favor renewable energies within the European Union. In particular, it compares the efficiency of tradable green certificates vs. so called feed-in tariffs. Both have been used widely in the more advanced countries deploying renewable energy, in particular wind in the UK (quota system) and in Germany (feed-in tariffs). Feed-in tariffs have also dominated in other EU countries, such as France, Netherlands or Spain. After a descriptive introduction of renewables policies in the EU, the authors discuss various obstacles to the design and implementation of efficient renewable promotion policies. In Section 3, the main analytical framework for analysis is developed in a partial equilibrium model of the electricity markets. The model uses strategically acting firms operating in national electricity markets; it uses an iso-elastic demand function, taking 2005 regional demand volumes as reference demand. Supply is represented by 1100 conventional thermal power plants and 220 firms. The model is calibrated using real EU data for a large number of regions (23). The renewable sector is modeled in detail, amongst others including 16 renewable technologies, costs and other technical detail. Section 5 compares to policy scenarios and results: FEED_D: regional and technological specific (for example regionally different premiums per technology); FEED_H: harmonized premiums per technology; QUOTA_R: regionally differentiated quotas for RES-E; QUOTA_EU: EU-wide harmonized quota for RES-E (making trade possible). The section then discusses the economic consequences of promoting the increased market penetration of renewable energies: the “business as usual” (BAU) scenario is compared to four policy scenarios, combining feed-in tariffs and quotas with a national or a regional approach, respectively. A differentiated feed-in tariff generates substantial costs compared to a quota system. Generally, harmonization (FEED_H, QUOTA_EU) is more efficient than regional coor* Thanks to Florian Leuthold for discussions about the paper.

28

Christian von Hirschhausen

dination (FEED_D, QUOTA_R). Today, there are large inefficiencies due to co-existing overlapping policy strategies. In particular, an EU-wide tradable quota system would reach in the European RES-E target at 23% lower cost than independent diversified feed-in tariff systems with technology specific premiums. The authors conclude that under pure emissions regulation there is no need for complementary renewables energy policy and that technology – or country specific approaches are less efficient than EU-wide approaches.

2. Questions about the Paper The paper derives the standard theoretical argument against a specific renewables policy that one can expect: In a CO2-constrained world, emissions trading is the best instrument; when targeting energy in particular, quotas work at lower social costs then (differentiated) feed-in tariffs. I could stop the discussion at this point, because theoretically this is true, and I agree. However, let me add three specific questions to the paper, and more broad issues on renewables:  There is an important distinction between renewable energies that is not captured in the paper: different variable costs. Thus solar and wind have variable costs of 0 A / MWh, or perhaps 5 A / MWh when one counts balancing costs. In any case, they are by far lower than those modeled for conventional energies. In reality, the merit order starts with wind and solar, at negligible marginal costs, and so the marginal technology becomes the conventional (see Figures 2 and 3 in the paper). In fact, the very instrument is ill-designed: in order to promote wind and solar, one would need to support investments, i.e. the fixed costs component and not the variable one;  this is a highly stylized model. In particular, there is no transmission grid included. This might have an effect, in particular for countries that benefit substantially from the instruments. Take the case of Spain that has a high potential for wind and solar energy. And indeed, as Tables 3 and 4 show, the (genitive) effects of RES-support schemes are very high. However, if one assumes transmission constraints, free trade between the renewable electricity can not take place, and thus the national portfolios will need to adapt. This is particularly the case for Spain that has a limited domestic network, and almost no connections with France, so that exports are infeasible;  another empirical result that is somewhat surprising is that the costs of a unified feed-in-tariff are very close to those of a regional quota system which would imply that the two policies do not really differ from each other. This might indicate that feed-in is preferable, since it is much easier to implement.

Alternative Strategies for Promoting Renewable Energy in EU Electricity Markets

29

3. General Questions on Renewable Deployment Going beyond the specific details, one is tempted to observe that the theoretical results obtained in the paper do not always correspond to reality. As one says: “the proof of the pudding is in the eating!” In fact, if one accepts the political goal to promote renewables, the quota-system seems to work much worse than theory would indicate. This is particularly evident for the promotion of wind through quotas in the British system: less than 2 GW of capacity have been built. Here, I refer to research by Karsten Neuhoff and colleagues, doubting the conceived wisdom: “Experience in the UK and in other European countries suggests that the current UK policy instrument for the promotion of renewables deployment, the Renewables Obligation (RO), struggles to deliver on deployment effectiveness, cost-efficiency, and technological diversity” (Johnston, et al., 2007, p. 2). On the other hand, the “expensive” system that is prevalent in Germany has produced significant wind deployment, almost touching 20 GW in 2006. One may of course argue that the costs of this deployment are high; yet the fact of the matter is that fed-intariffs lead to investment, whereas the quota system does not seem to incite significant investment. Thus, we are directly back in the “prices or quantity” discussion initially proposed by Weitzman. There is a certain parallel to the CO2-markts that were discussed earlier: is it not the wish of most economists to replace the unstable quota system by a stable price also driven by the wish to facilitate investment? If this is the case, then, the difference between the feed-in tariff and the quota calculated in the paper looses importance. There is an additional aspect to be mentioned, which is currently making head waves in different sub disciplines of economics: the issue of stochasticity of outcomes, linked to the relation between risk and investment. In fact, the deployment of renewables is directly related to the issue of investment risk that has recently popped up in different infrastructure sectors (see Guthrie, 2006). The main argument is that higher investment risk leads to higher capital costs, and thus lower investments. This argument also holds for investments in renewables: a quota system generates considerable risks, whereas a feed-in tariff lowers risk. Even this is not per se an argument to always favor fixed prices, it is an important insight that one cannot ignore when the political objective is to deploy more renewables in electricity production. It has been a pleasure to read the paper and I look forward to further versions and discussions. Thank you! References Butler, Lucy / Neuhoff, Karsten (2005): Comparison of Feed in Tariff, Quota and Auction Mechanisms to Support Wind Power Development. University of Cambridge, EPRG Working Paper CWPE 0503 (~CMI Working Paper 70).

30

Christian von Hirschhausen

Guthrie, Graeme (2006): Regulating Infrastructure: The Impact of Risk and Investment. Journal of Economic Literature, Vol. XLIV (December), pp. 925 – 972. Johnston, Angus / Kavali, Amalia / Neuhoff, Karsten (2007): Take-or-pay Contracts for Renewables Deployment. University of Cambridge, EPRG Working Paper.

Biofuels and Climate Policy By Gernot Klepper* 1. Introduction There is a growing consensus that the emissions of greenhouse gases (GHG), especially CO2 should be reduced as soon as possible. This essentially means that the consumption of fossil fuels needs to be reduced thus helping to mitigate climate change and to preserve scarce resources such as crude oil. The Fourth Assessment Report (FAR) of the IPCC shows growing evidence that activities to significantly reduce GHG emissions need to be taken within the next ten years if climate change is to be limited to an increase of average temperatures of about 2 degrees Celsius (IPCC 2007a). This reduction in the use of fossil fuels requires changes in energy use. Many studies undertaken (e.g. IEA 2005) indicate that a shift in the composition of energy consumption towards renewable energy sources together with improvements in the efficiency of end-energy use and advanced technologies in electricity production can achieve reductions of CO2-emissions by 2050 in the order of 50 percent relative to the business as usual scenarios. The “Leitstudie” of the German Minister for the Environment (BMU 2007) investigates options that lead to even stronger reductions of Greenhouse gases (GHGs). Among the different forms of renewable energy biofuels are one option that is particularly suited to replace fossil energy needs in the transport sector. They are therefore seen as a special option since the other renewable energy sources usually replace fossil fuels in the electricity generation or the provision of heat but not in the transport sector. The European Commission has decided to set specific targets for the consumption of biofuels in its “Renewable Energy Road Map” (Commission of the European Union 2007). Given the important role that biofuels seem to play in the policy arena and in the general public it is therefore interesting to take a closer look at the pros and cons of following the particular biofuel strategy that is developing in Germany and the European Union. The paper addresses first the context in which the promotion of biofuels needs to be looked at. It then describes the European and international market development for biofuels and the government support that enables biofuel producers to compete with traditional fossil fuels. The potential GHG-savings from the use of biofuels are assessed and cost estimates for this particular climate policy are given. * The Kiel Institute for the World Economy, Kiel, Germany.

32

Gernot Klepper

The paper concludes with a discussion of policy options for the future support of the use of biomass for climate policies.

2. The Context for Assessing Biofuels as a Part of Climate Policy There is general agreement in almost the whole research community concerned with energy and climate issues that the substitution of fossil fuels through renewable energy sources is inevitable in the long run. As the IPCC report (IPCC 2007a) has made clear a stabilization of the concentration GHGs in the atmosphere at around 500 ppmv requires a drastic reduction of emissions in the order of 80 percent by the end of the century. The reference scenario of the IEA (2005) for the primary energy use world wide as shown in Figure 1 clearly shows that the world is still far away from this goal and and achieving it would – among other things – require a much larger share of renewable energy sources in the energy mix than is envisaged in this scenario.

Source: IEA (2005) World Energy Outlook 2005 (reference scenario).

Figure 1: Primary Energy Use by Energy Source Reference Scenario IEA World Energy Outlook 2005

The share of renewable energy in the German energy use is 6.4 percent of final energy use. Roughly two third comes from biomass, the rest from water and wind energy and to a small degree from solar and geothermal energy.1 The production of energy from biomass is limited by the amount of land available within a country or – if international trade is included – by the world wide agricultural area. Currently 1

See BMU 2006.

Biofuels and Climate Policy

33

the consumption of biofuels is concentrated on domestically produced fuels. This is to a significant degree determined by regulation and only partially by market forces.2 Regardless whether biofuel consumption relies only on domestic production or on global sources the limiting factor for the expansion of its consumption is how much land should and will be dedicated for the production of biomass. The promotion of biofuels needs to be seen in a competitive framework that goes beyond the fuel sector. The feedstock for all current biofuel technologies and for most so called second generation biofuels is biomass produced on agricultural soils. Figure 2 illustrates the many different uses of biomass. Whereas historically biomass has served nutritional needs and provided energy, its role as a provider of energy has diminished with the introduction of first coal, then oil and subsequently gas and nuclear energy. In this process agricultural activities have been reduced mainly to food production. This process was beneficial as the growing world population required increasing amounts of agricultural land to come under cultivation for food production. Now this process is moving back to a system with multiple demands for biomass. There is, first of all, the competition between producing bioenergy or producing food, the well-known debate on “fuels versus food”. There are also other biomass uses such as the non-energetic and non-food use of biomass as a raw material in industry.

Figure 2: Competing Uses of Biomass

Even the bioenergy production from biomass is subject to competing uses. There is the question as to whether agricultural products c should go into the production of biofuels or into other energy uses such as the production of electricity or heat. 2

More on this in section 4.

34

Gernot Klepper

Even within the biofuel uses the question is whether to promote Biodiesel or rather Bioethanol. And this is still only a part of the multidimensional uses into which land can go. The preservation of biodiversity through protected areas, the set aside of natural land for tourist and recreational uses, and finally the destruction of fertile land for commercial non-agricultural uses such as roads, buildings etc. are all competing uses of scarce land resources. It is therefore indispensable that an evaluation of policy options in climate and energy policies involving biomass needs to take into account that these uses are competing with other valuable activities on scarce land resources. 3. The Market for Biofuels There are different types of biofuels with rather different production technologies and different agricultural raw materials, usually called feedstocks. Yet, they are all part of a world market that is becoming increasingly globalized. The share of alternative fuel sources in the fuel market is still limited world wide to about 4 percent (Figure 3). Of those 4 percent less than a third actually consists of biofuels. The rest is composed of fuels based on natural gas which is not a renewable resource but is essentially another fossil fuel.

Source: Hart’s World Refiring and Fuels Service, 2006.

Figure 3: The Composition of World Fuel Consumption (2005)

There are currently two types of biofuels dominating the market: Bioethanol and Biodiesel. Bioethanol is produced from feedstocks such as sugar cane, corn, other grains, and sugar beet. Biodiesel is produced from vegetable oils, mostly rapeseed oil, palm and soy bean oil. The market for bioethanol with 45 mio. m3 output is dominated by the Americas where the USA and Brazil dominate production. The EU contributes only 6 percent of the bioethanol (2.7 mil. m3) to world markets of which France and Germany produce about 50 percent. The rest is more or less evenly divided among the other EU member states.

Biofuels and Climate Policy

35

World

South America 35%

EU Other EU 10%

Asia 14% Oceania 0%

France 30%

UK 9% Sweden 4%

EU 6% Other Europe 3%

Spain 15%

Africa 1%

Germany 18%

Poland 7%

N&C America 41%

Italy 5%

Hungary 2%

Figure 4: Ethanol Production by Region (2005)

Biodiesel production amounts to only 4.3 mil. m3 and is dominated by the European Union, especially by Germany as Figure 5 shows. This is mainly due to the fact that Germany is one of the few countries that have for many years given tax incentives for diesel fuel resulting, compared to international averages, in a large share of diesel engines in the automobile fleet in Germany and some other European countries. Consequently, Germany’s biodiesel production on the basis of rapeseed already requires almost the complete land area suitable for rapeseed production to be used. Others EU-25 10% Slovakia 2% Czech Rep. 4%

RoW 8% USA 7%

Poland 3%

EU 85%

Italy 12%

Germany 53%

France 16%

Firgure 5: Biodiesel Production by Region (2005)

With very few exceptions the different biofuels are all not competitive against fossil fuels at current market prices. The only sidnificant exception is Brazil where bioethanol based on sugar cane as a feedstock is less costly that gasoline even without any price support or subsidies. European and North American bioethanol as well as biodiesel from Asia can not compete with fossil gasoline or diesel.

36

Gernot Klepper

Therefore the current biofuel production in Europe, North America, and Asia is mainly driven by a whole set of government policies in support of biofuels.

4. Government Support for Biofuels The largest markets for biofuels are Brazil, the USA, and the EU. Brazil that has promoted biofuels after the oil crises in the seventies, now has a completely open market without any governmental support measures. In the USA bioethanol is supported through a whole set of policies in the Farm Act such as tax refunds and through compulsory blending of bioethanol with fossil gasoline. In the EU, many governments support biofuels through a full or partial exemption from mineral oil taxes. These incentives are accompanied by compulsory blending requirements. Germany has started its political support for biofuels with a complete tax exemption for all biofuels. Given biofuel production costs of between A 0.50 and A 0.70 per litre, biofuels immediately became a very competitive product with its price advantage increasing even further the more oil prices were to increase in the future. Two reasons led to the policy change of reducing the tax exemption for some uses of biofuels and of abandoning it altogether for other uses. The European Union ruled overcompensation illegal, i.e., tax exemptions can not lead to a price of biofuels that is below fossil fuel prices at the gas station. Secondly, already in 2005 the tax exemption led to a fall in revenues from the mineral oil tax of about 1.3 bn A with the prospect of rising to more than 5 bn A if the objectives of the German biofuel initiative were to be met and the already planned investments in ethanol plants were to be realized. Therefore the tax exemption was strongly modified. For the overall share of biofuels in the fuel consumption a quota system as proposed by the Commission of the EU has been introduced. The EU requests that member states reach a share of 5.75 percent of biofuels by the year 2010. In addition, a differentiated system of tax rates for the different types of biofuels was introduced. Table 1 shows that essentially those biofuels that are mixed to fossil fuels with a share of up to 5 percent, i.e., Ethanol (E5) and Biodiesel (B5), are now fully taxed. In contrast, second generation biofuels such as BtL or Lignocellulose and E85 or Biogas – which at the moment are only niche products – are still tax exempt. It has already been mentioned that all biodiesel and bioethanol from European sources can not compete with fossil fuels at current market prices. It is only the tax exemption and the compulsory blending that bridges the gap between production costs for biofuels and fossil fuels. In addition, both for biodiesel and for bioethanol non-European suppliers can offer lower prices than European producers. Bioethanol from Brazil has production costs of 0.20 to 0.25 A / l, corn based bioethanol from the USA costs about 0.29 A / l whereas European producers have cost ranging from – in the best case – 0.35 A / l to 0.80 A / l. For biodiesel, the situation is not that clear since currently 85 percent of the biodiesel production takes place in Eur-

Biofuels and Climate Policy

37

ope. However, subsidized imports from the USA enter the EU and the production of biodiesel on the basis of palm oil in Malaysia and Indonesia is expected to increase significantly in the next years. These potential exports will most likely have lower production costs than biodiesel based on rapeseed in Europe. This lack of competitiveness of European biofuels has not led to significant imports, simply because European biofuel markets are protected through several trade barriers. Table 1 German Tax rates for Biofuels in A / litre 2006

2007

2008

2009

2010

E85

0

E5 Ligno

0 0

0.65

0.65

0.65

0.65

B100 B5 SVO BtL

0.09 0.15 0 0

0.09 0.47 0

0.15 0.47 0.10

0.21 0.47 0.18

0.27 0.47 0.26

Biogas

0

2011 Tax Free 0.65 Tax Free 0.33 0.47 0.33 Tax Free Tax Free

2012

2013

2014

2015

0.65

0.65

0.65

0.65

0.45 0.47 0.45

0.45 0.47 0.45

0.45 0.47 0.45

0.45 0.47 0.45

E85 = Fuel with 85% Ethanol, E5 = Fuel with 5 % Ethanol, Ligno = Biofuel from lignocellulose, B100 = pure Biodiesel, B5 = Diesel with 5 % Biodiesel, SVO = Straight Vegetable Oils, BtL = “Biomass to Liquid”. Source: Own compilation.

Imports of Ethanol face an import tariff with three different rates. 0.24 A / l for undenaturated Ethanol, 0.13 A / l for denaturated and a value tariff of 6.5 percent for Ethanol as a “chemical product” are the different tariff rates. Most European countries require the use of undenaturated Ethanol in the biofuel sector although there is no reason for this requirement except for the fact that it has the highest tariff rate and makes sure that most imports are hardly competitive in European markets.3 In the biodiesel market non-tariff barriers provide most of the protection to foreign competition. The fuel quality regulations for biodiesel prohibit the use of biodiesel that has used palm oil or soybean oil as a feedstock. There are some not very convincing reasons to ban American and Asian biodiesel imports because of problems with using these biodiesel blends at low temperatures. In any case, such 3 This system has led to innovative activities such as “splash blending” where some gasoline is added to the Ethanol in order to turn it from an “undenaturated” into a “denaturated” product.

38

Gernot Klepper

quality regulations – whether they are a non-tariff barrier intentionally introduced for protecting local markets or for reasons purely related to technical aspects – are appreciated by European biodiesel suppliers. The combination of partial or complete tax exemptions with mixing requirements and trade barriers have created a European market that is essentially isolated from the rest of the world and provides a profitable business for European farmers and biofuel producers. The price difference between foreign and European biofuels already makes clear that there is a welfare loss due to the impossibility of exploiting obvious comparative advantage from international trade in biofuels. In addition, there is the climate policy perspective of consuming biofuels. The GHGs saved from substituting gasoline and diesel through their biomass based counterparts differ significantly. The prospects for protecting the climate through the different biofuels therefore need to be considered.

5. How Much Contribute Biofuels to Climate Protection? The “Renewable Energy Roadmap” (Commission of the European Communities 2007) has set specific targets for biofuels. However, the argument for promoting the substitution of fossil fuels by the biofuels is not based on climate protection but only on energy security. Their contribution to climate protection did not seem to be worth mentioning. The German policies towards biofuels mention three objectives: energy security, mitigation of climate change, and support to rural areas. In fact, the question as to how much biofuels can contribute to the reduction of emissions of GHGs is discussed quite controversially. Proponents see it as a cornerstone of the move away from fossil energy sources towards renewable sources. Critics argue that biofuels are not as effective in substituting fossil fuels than other uses of renewable energy (Henke / Klepper / Schmitz 2004). They argue that biofuels are by no means free of fossil carbon since the production of the feedstock and the conversion of the agricultural feedstock into biofuels requires substantial inputs from fossil sources. Most important are the energy inputs for planting, harvesting, and transport as well as the use of chemical fertilizers. The conversion processes in most cases also use fossil energy sources such as oil, gas or even brown coal. There are several ways in which the contribution of biofuels to climate protection can be assessed. A direct approach is to compare the GHG emissions from consuming fossil fuels with those of biofuels. As a result one can compute the amount of GHGs saved per litre of gasoline. A broader approach for assessing the role of biofuels takes into account the fact that biomass production is limited by the availability of land suitable for growing specific crops. The economically most valuable use of this limiting factor would need to be determined by evaluating not only the market value of different types of land uses but also the external effects. These externalities refer to climate protection but also to the preservation of biodi-

Biofuels and Climate Policy

39

versity or the use of land for recreational purposes that may depend on intangible aspects such as landscape. However, such a comprehensive assessment of the social value of different types of land uses has so far not been done, not the least because such an evaluation would require a substantial interdisciplinary research effort. Nevertheless, a forward looking land use policy should be based on such a comprehensive assessment of the social value of land. Such assessments are not available but partial analyses of some of the externalities have been done and attempts to use those in regulating the uses of biomass are currently investigated. Until such assessments are implementable a direct and simple indicator for the contribution of biofuels to climate policy with respect to competing land uses consists of the amount of GHG emissions saved per unit of land in different uses. The physical amount of GHG savings from biofuels – either per unit of land or per unit of fuel produced – also needs to be compared to the production costs of the different biofuels. With these production costs and the price of fossil fuels it is possible to compute the GHG abatement costs of biofuels. The GHG abatement costs are defined as the difference between the marginal costs of producing a certain biofuel and the corresponding costs of the comparable fossil fuel measured in energy equivalents.4 Hence, GHG abatement costs are both depending on oil prices and on prices for agricultural products and the conversion costs. Rising oil prices will lower GHG abatement costs and rising feedstock prices for biofuels will increase them. This means that a move towards the “Second Generation biofuels” such as BtL that rely on byproducts from agricultural production or waste materials would automatically lower GHG abatement costs. Table 2 Indicators for GHG Savings from Biofuels Biodiesel Bioethanol Bioethanol Bioethanol (Grains) (Sugar Beet) (Sugar Cane) Gross Energy Savingsa) Net Energy Savingsa) GHG Savingsb) GHG Abatement Costsc) a)

BtL

Biogas

51

54

132

137

135

178

38 3.4

30 2.9

88 7.2

116 15.5

118 10

113 8

154

252

290

–27

272

273

GJ / ha – b) tCO2e / ha – c) A / tCO2e. Source: FNR (2006).

4 The conversion into energy equivalents is necessary since biofuels have a lower energy density such that the same amount of biofuel delivers a lower energy service than the corresponding fossil fuel. In fact, one litre of Biodiesel replaces 0.9 l of fossil diesel and one litre of Bioethanol replaces 0.65 l of gasoline.

40

Gernot Klepper

Table 2 summarizes different indicators for the energy and GHG balances of different biofuels. The first column refers to Biodiesel as RME (rapeseed methyle ester) produced in Germany, the second to Bioethanol from grain, the third to Bioethanol with sugar beet as a feedstock, and the fourth one to Bioethanol from Brazil which uses sugar cane as a feedstock. Biomass to Liquid (BtL) is a second generation fuel still in the research face which uses as feedstock all kinds of biomass including biowaste. Finally, Biogas refers to the production of gas with energy maize as a feedstock. Biogas is usually not turned into a car fuel but is used for heat and electricity production. It illustrates a different route of land use that also replaces fossil energy source although not for transportation. The largest energy yield that one can get from a hectare of agricultural land is not through biofuels but in the form of Biogas as the first row of Table 2 shows. Sugar beet and sugar cane have high hectare yields resulting in a high gross energy output of Bioethanol whereas grains and rapeseed by far do not provide a comparable energy content. Since gross energy yields do not take into account that the production of the biofuels requires inputs of fossil fuels the appropriate measure for the fossil energy savings from biofuels is the net energy yield where the fossil energy inputs are subtracted from the energy content of the biofuel. The highest net energy yield of existing production activities comes from Brazilian Bioethenol.5 European Biodiesel and Bioethanol net energy yields are much lower since they require substantive inputs of fossil energy. A comparison of the Biodiesel and the German Bioethanol gross and net yields indicates that between 25 and 40 percent of the energy content in the biofuels has been used as fossil fuels in the production of the biofuels. These fossil energy inputs occur both in the production of the feedstock as well as in the conversion process. They are more pronounced in the production of rapeseed, grains and sugar beet than in the production of sugar cane or biogas. In the latter, the conversion processes use very little to almost no fossil energy and the energy requirements of the feedstocks are also of less importance. Under a global perspective in which there is an increasing shortage of land available for food and bioenergy production it would not be reasonable to produce biofuels in Germany if they lack competitiveness when compared to, for example, Brazilian Bioethanol. If the policy objective is to substitute fossil fuels by energy sources from biomass in general the best choice in Europe would be biogas as it brings about the highest net energy yield per hectare. Finally, under a purely national perspective of supporting the production of biofuels the choice between Diesel, Bioethanol from grain and Bioethanol based on sugar beet goes clearly in favour of sugar beet based Bioethanol. Such a national or European orientation is debateable, yet the revealed preference of European decision makers as it is evi5 BtL and Biogas are in a similar range but the former is currently not available on an industrial scale and the latter is not a substitute for fossil transport fuels.

Biofuels and Climate Policy

41

dent from the high tariff barriers for biofuels is clear. They prefer national or at most European but not global solutions. Turning from energy savings to the savings of GHGs which is the main issue when one is concerned with climate policy, the GHG savings on a hectare of land are again largest in Brazil. Bioethanol from sugar cane saves 15.5 tCO2e / ha compared to 2.9 tCO2e / ha for grain based Ethanol and up to 7.2 tCO2e / ha for Ethanol from sugar beet with Biodiesel being in the middle. Even biogas is not competitive with respect to Brazilian Ethanol. Even though land as a scarce resource is an important determinant for the availability of biofuels and the competition between agricultural products for food and fuels, abatement costs are an equally important factor for an efficient use of alternative renewable energy resources. The different biofuels have quite divergent abatement costs. GHG-Abatement costs are defined as the additional costs that are incurred if a certain amount of fossil fuel is replaced by a biofuel. In this calculation included is the fact that biofuels have a different energy density, i.e., a specific quantity of biofuel does not replace the same quantity of fossil fuel. One litre of Bioethanol, for example, replaces roughly 0.65 litres of fossil fuel. In other words the fossil fuels have a higher energy density and thus a higher energy service per unit of fuel. The GHG-Abatement costs for the different fossil fuels vary drastically. Biodiesel costs somewhat more than 150 A / tCO2e. Reducing GHG-emissions through the substitution of Bioethanol from grain and sugar beet already cost 250 and 290 A / tCO2e. Biogas and BtL are similarly expensive. In contrast, Brazilian Bioethanol has negative abatement costs. This is due to the fact that Bioethanol in Brazil today is – even without subsidies – less expensive than fossil gasoline such that it is profitable to use Bioethanol instead of traditional gasoline. The abatement costs of biofuels heavily depend on the price of gasoline. Whereas the figures in Table 2 are based on simulations with a crude oil price of 50 $ / barrel, at todays prices the abatement costs would also be lower. At an oilprice rising above 80 $ / barrel even European biofuels can become competitive vis-a-vis gasoline. However, the relative competitive position of the different biofuels in terms of their abatement costs will not change under different oil prices. European biofuels will remain much more costly than foreign alternative products such as Bioethanol from Brazil or Biodiesel produced from Palmoil. Efficient abatement of greenhouse gases is achieved if the least costly options for reducing these gases are chosen. The ranges of abatement costs of European biofuels are significantly higher than many other abatement options. The carbon prices in the European Emission Trading Scheme (ETS) are around 25 A / tCO2 for the second commitment period. Emissions reduction certificates under the KyotoProtocol sell at even lower prices. Alternative measures such as energy savings in energy end use activities are estimated to be almost costless or in some cases even negative. Other technologies such as photovoltaics are so expensive that they are only efficiently applicable in specific circumstances.

42

Gernot Klepper

The quantities of fossil fuels that can be replaced by biofuels is strongly limited if the biofuels are to be produced within the EU. The “Renewable Energy Roadmap” of the EU sets a target of 10 percent for biofuels in the overall fuel consumption. This would according to a report (DLG 2007) require to devote about 50 percent of the agricultural land in the EU-15 to the production of feedstock for biofuels. The situation in the USA is similar where about 30 percent would be needed to meet a 10 percent target. Only Brazil for Ethanol and in part Malaysia and Indonesia for Biodiesel have sufficient land areas suitable for biofuel production. Given the world wide demand even these areas would not be able to satisfy more than a small proportion of fuel demand world-wide. In addition, there is doubt whether these supplies can really be supplied on a sustainable basis without threatening other environmental resources. Especially the supplies from Asia and from Brazil have been criticized for indirectly threatening climate protection and for destroying scarce environmental resources. The illegal logging in primary rainforests in Indonesia and the draining of peatlands for Palmoil production may lead to such high emissions of GHGs that the savings from biofuels are more than eaten up. The expansion of areas devoted to sugar cane in Brazil are believed to push the previous agricultural activities such as cattle raising and soy bean production more towards the tropical rainforests. It is argued that – although indirectly – the Ethanol production is responsible for the destruction of the Amazon forest. If biofuels are produced under such conditions they compromise the attempts to use biofuels for climate protection. Hence appropriate measures need to be introduced to preserve the credibility of the GHG-savings that are achieved under proper and sustainable agricultural practices. In addition, expanding the production of biofuels beyond already existing land areas which are used so far for food production needs to be monitored carefully in order not to undermine other land use objectives such as the preservation of biological diversity.

6. Policy Options for Biofuel and Bioenergy Policies As described above, biofuels in Europe rely on a sophisticated system of support ranging from trade protection through tariffs to tax incentives and mixing requirements. The current policy mix encourages national solutions which are more costly than global solutions. It also does not take into account criteria for an efficient use of scarce land resources and the sustainability of biofuel strategies. The efficient allocation of scarce land resources would require to put those activities in place that provide the highest social benefit, i.e., not the highest profits based on market prices but also the benefits from protecting the climate or from protecting biological diversity. In terms of climate protection those options of bioenergy production should then be chosen that have the lowest CO2-abatement costs. If one looks only at the biofuel production locally it would favour Biodiesel over Bioethanol. If instead other

Biofuels and Climate Policy

43

uses of Biomass for energy production were also included options such as the direct conversion of biomass into heat and power may yield higher fossil energy savings at lower cost.6 In the case of Biogas the abatement costs could be significantly lowered if the biogas plants could produce biogas together with heat that can be used in the neighbourhood. The current system of mixing requirements for biofuels and of price support for Biogas and other biomass based energy production does not provide the appropriate incentives for a socially efficient allocation of land resources. The objective of reducing GHG-emissions through the use of Biomass should use mechanisms that provide signals to farmers such that the best climate policy is automatically chosen. This would require a rethinking of support measures by moving away from the ad hoc regulation of different Bioenergy activities towards an integrated climate policy with respect to Bioenergy. Such a unifying system should be based on the GHG-savings that a particular activity provides. If the GHG-savings are the measuring rod for support, those activities with the highest savings would receive the highest support whereas those with low savings would be supported less. A possible solution would be a certification system in which the GHG-balance for a particular plant all along the supply chain is assessed. The GHG-savings could then be traded in an emission trading scheme such as the ETS or even the ETS itself. This would, however, require the inclusion of all energy related activities such as transport and small incineration activities into the ETS. Such a system in effect internalizes in an efficient way the GHG-savings that arise from the use of biomass for energy production and thus avoids the differential treatment of different land use activities with respect Bioenergy. Such a carbon accounting logic for determining the best energy production from Biomass would first of all mean that many of the adhoc support policies for different biofuels could be abandoned. This would also include a revision of the mixing directive of the EU. But since it is in direct conflict with another EU requirement, namely the reduction of the carbon content of fuels by one percent per year, a reform of the biofuel policies of the EU will be necessary in any case. This proposed reform would not constitute a policy directly geared at the fuel sector but would provide equal opportunities for all energy sources from Biomass. This implicitly assumes that fossil fuels, i.e., oil, are good to perfect substitutes with other oil uses such as heating7, an assumption that is difficult to refute. It would therefore make no difference whether crude oil is saved in the transport sector or in the heating activities. In terms of climate protection those land uses should be preferred that provide GHG-savings at the lowest abatement costs. The current focus on local supplies of biofuels also poses a serious efficiency problem and thus a policy problem. The local biofuels are more expensive and less 6 7

See for example Henke / Klepper / Schmitz (2004) Oil is practically irrelevant for electricity production in Germany and most of Europe.

44

Gernot Klepper

climate friendly than foreign supplies, e.g. from Brazil. The comparative advantage for biofuels is not in Europe for mainly climatic reasons that lead to lower yields per hectare, but also because of the higher cost of production and, not the least, because of the lower CO2-savings of the biofuels. This poses a direct conflict between an efficient climate policy and the support of agriculture in Europe. From a climate policy perspective the biofuel as well as the bioenergy strategy should be internationally oriented thus exploiting the best options for GHG-savings. An opening of biofuel markets to international competition would result in an immediate loss of competitiveness of European biofuels, but it would also pose a threat of uncontrolled imports of biofuels that may not be as climate friendly as other supplies that are properly produced. Controlling the imports of biofuels that provide no real GHG-savings can be overcome by an appropriate certification scheme that assess both the sustainability and the GHG-savings for biofuels from a particular foreign or local supplier. Such certification schemes are in planning in several European countries and their introduction could strongly ease the opening of European biofuel markets. The conflict between agricultural support and climate policies may seem to look unsolvable in the short run. In the long run this problem is likely to be resolved by market forces. The demand for food products, especially meat, is growing with rising incomes in many countries with large populations. This is accompanied by an increasing demand for feedstocks in the biofuel production as well as an increased use of biomass for electricity and heat. World markets have already responded with prices for grain and oil products that are increasing fast and that have broken a trend of falling prices which has persisted for many years. On the one hand, this provides higher income opportunities for farmers. On the other hand it also raises feedstock prices for biofuels thus making biofuel production relatively less profitable than food production. Hence, the world-wide trend towards higher grain and vegetable oil prices will change the relative profitability of European farming activities. In other words, even in a market without trade barriers Europeans farmers are unlikely to loose. Instead they might move back towards food production and let the comparative advantage of other countries in the production of biofuels be realized. In summary, a move towards a system that provides support to energy from biomass based on GHG-savings would provide more rationality to land use control. A particularly efficient way would be to link bioenergy production with the European emission Trading Scheme (ETS) through an international certification of biomass. The opening of biofuel markets could not only exploit unused comparative advantage in cost terms, it could also exploit higher GHG-savings from foreign supplies. The increasing scarcity of agricultural areas that will become even more pronounced with climate change will raise agricultural prices such that income from agricultural activities is likely to increase and the profitability of bioenergy production in Europe will diminish.

Biofuels and Climate Policy

45

References BP (2006): “Statistical Review of World Energy”. London Bundesminster für Umwelt, Naturschutz und Reaktorsicherheit (BMU) (2007): “Leitstudie 2007 – Ausbaustrategie Erneuerbare Energien.” Berlin. Commission of the European Communities (2007): “Renewable Energy Roadmap – Renewable Energies in the 21st Century: Building a More Sustainable Future,” COM(2006) 848 final, Brussels. Fachagentur für nachwachsende Rohstoffe (FNR) ed. (2006): “Biokraftstoffe – Eine vergleichende Analyse,” Gülzow. Henke, J. M. / Klepper, G. / Schmitz, N. (2004): Tax exemption for biofuels in Germany: Is bio-ethanol really an option for climate policy? Energy. International Energy Agency (2005): “World Energy Outlook 2005,” OECD / IEA Paris. – (2006): “Energy Technology Perspectives 2006,” OECD / IEA Paris. IPCC (2007a): “Climate Change 2007 – Impacts, Adaptation, and Vulnerability – Summary for Policy Makers,” http: // www.ipcc.ch / SPM13apr07.pdf. – (2007b): “Climate Change 2007 – Mitigation of Climate Change – Summary for Policy Makers,” http: // www.ipcc.ch / SPM040507.pdf.

Investments of the German Electricity-Supply Industry By Hans-Dieter Karl* Abstract This article presents, on the basis of the fundamental conditions of investments in the electricity supply, the development of the electricity sector in Germany and the structure of investments up to 2005. On the basis of the existing power plants and informed by the further development of gross power consumption as well as the planned discontinuation of nuclear power plants, investments of the electricity-supply industry are estimated up to 2025. They include the costs for conventional power plants, the power grid as well as for miscellaneous investments, and amount to ca. A 180 billion, including investments in plants for electricity generation from renewable energy sources. Finally the effects of changes in the production structure on fuel input and on carbon dioxide emissions were determined. In spite of the assumed efficiency gains in electricity generation, an increasing use of fossil energy sources results. JEL Classification: Q41, Q42, Q43, Q50 Keywords: Investment, electricity supply, gross power consumption, fuel use.

1. Features of Electricity Investments Electricity is energy in a pure form and as such has an exceptionally broad application spectrum, such as powering machines, driving computers, the transmission of information or the generation of heat. This versatility is also the reason for its economic importance and for the great efforts taken for the provision of electrical energy. These strong advantages of electricity have one major drawback: electrical energy is not storable in large quantities and at justifiable expenditure. Another feature is that electrical power must be generated when demand arises. The generation and provision of electrical energy requires considerable investments in power plants, transmission and distribution networks. The costs for electrical energy are unusually high also in comparison to the supply of other energy sources. This is documented, for example, in the 2006 World Energy Outlook of * Ifo Institute for Economic Research at the University of Munich, Poschingerstaße 5, 81679 Munich, Germany. E-Mail: [email protected].

48

Hans-Dieter Karl

the International Energy Agency (IEA 2006, p. 77). This study estimates total investments worldwide for the energy supply infrastructure for the period 2005 to 2030 to be $ 20 trillion, in 2005 prices. These investments are for expanding energy supply capacities as well as for replacing available plants. The prominence of the electricity sector is underscored by the fact that electricity generation, transmission and distribution alone accounts for more than $ 11 trillion, or 56 % of total energy supply investments. Including also investment costs for the supply of fuels for power plants would add another ten percentage points to the total. The capital intensity of the electricity sector – measured as gross plant investments per provided energy unit – is of an order of magnitude of four to five times the corresponding values for oil or natural gas. These worldwide investments are high not least because many of the developing and threshold economies are building up comprehensive energy supply systems in the forecast period. The developed countries will also need to spend considerably to safeguard and expand their energy supply. This also applies to Germany, especially in terms of the energy- and environmental-policy regulations. It should be noted that of all investments for the energy supply in Germany a much greater share of overall investments is in the electricity sector since natural gas, mineral oil and anthracite coal are mainly imported and thus only a portion of total investments for the provision of these energy sources are domestic. Of the domestic energy sources, it is mainly lignite coal that provides a major contribution to the energy supply, and this lies exclusively in the generation of electricity.

2. The Development of the Electricity Sector in Germany The economic recovery of Germany after the Second World War was linked with considerable additional need for energy, in particular electrical energy. Between 1950 and 1960, total power consumption increased by 167 percent in West Germany, which corresponded to an average annual increase of approximately 10%. Real economic output in this period increased by 113 percent. This yields an elasticity coefficient between the rise of power consumption and economic growth of 1.48. In the following decades the growth weakened continuously. Between 1960 and 1970 power consumption rose a little more than two-fold and economic output increased by about 60 percent. Growth was more electricity intense, since the elasticity coefficient in this decade amounted to 1.78. Not least because of the oil crisis in the 1970s, the dynamic economic development decreased clearly, which also led to reduced demand for electricity, but still for one unit of growth in economic output 1.5 units of electrical energy were employed. In the following decade, 1980 to 1990, this changed fundamentally: The economy grew faster than power consumption, yielding an elasticity coefficient of 0.85 for this decade. In reunified Germany this trend continued from 1991 to 2006; total power consumption was clearly behind economic growth (see Table 1).

Investments of the German Electricity-Supply Industry

49

Table 1 Electricity Consumption and Economic Development in Germany from 1950 to 2006 Perioda) 1950 1950 – 1960 1960 1960 – 1970 1970 1970 – 1980 1980 1980 – 1990 1990 1991 – 2006 2006 a)

Increase in electricit consumption in %

Economic growth in %

Elasticity coefficient

167.0

113.0

1.48

Electricity productivity, 1950 = 100 100 79.8

105.2

59.2

1.78

49.6

32.2

1.54

19.7

23.3

0.85

61.9 54.7 56.3 14.1

24.7

0.57 61.5

All Germany as of 1991.

Source: AG Energiebilanzen; VDEW; Ifo Institute.

For the years in question this development can be shown, conversely, in terms of total electricity productivity, defined as a quotient of GDP and gross power consumption. Whereas in the initial decades electricity productivity clearly decreased, in the mid-1980s a trend reversal occurred. Since then, electricity productivity has increased again and in 2006 was on the level of 1970 (see Table 1). The main reason for this development can be seen in the reaction to increasing electricity prices, which made it increasingly worthwhile to use more efficient and energysaving equipment. In addition, as a result of structural changes in the German economy, industries with energy-intensive production showed small growth in comparison with other industries or increasingly shifted such production abroad. Since, however, the energy-savings potential for electrical energy is evidently less than for combustible fuel, electricity productivity increased more slowly than the productivity of primary energy (DIW-Wochenbericht Nr. 8 / 2007, p. 107). The strong growth of power consumption up to the 1970s was the main determinant for investments in the general electricity supply. In the first post-war years it was mainly a question of creating a secure electricity supply and of providing sufficient supply in light of increasing electrification. Only in later years did the determining reason for the investments change. With increasing power infrastructure, the need for replacements increased automatically, and after the strong price increases for fuel oil and natural gas from the early 1970s, substitution by more economical fuel became increasingly more important. But also continuous modernisa-

50

Hans-Dieter Karl

tion and an increase in the reliability of the power grid required high investments. Finally the deregulations of wide parts of the electricity supply that began in the 1990s confronted the enterprises with completely new challenges. Overall, the changes in investment motives can characterised over time as follows:  growth stage,  consolidation stage,  competition stage.

In the course of this development the investment volumes tended to decrease relative to the rise in power consumption. Apart from the construction stage in the 1950s, the reason is that the plants became more reliable and that their technical and economic life spans increased. After substantial investments particularly in power plants were still carried out in the 1980s, in the following years a clear decrease in capital spending occurred due to the reduced growth of power consumption. Figure 1 shows the development from 1950 to 1999, displaying for individual decades the cumulated power consumption and the cumulated nominal investments – converted into approximate real values. 1000 900 800 700 600 500 400 300 200 100 0 1960-1969 a)

1970-1979

1980-1989

1990-1999

Germany.

Source: VDEW; Ifo Institute.

Figure 1: Electricity Consumption and Gross Fixed Capital Formation of the General Electricity Supply in past Decades – West Germany, 1950 till 1959 = 100

The increased investments in the 1980s is striking. In spite of an already evident flattening in the growth of power consumption, investments reached record values. The reason for this increase was investments for electrical power production. Several large nuclear power plants, with particularly high specific investments, were

Investments of the German Electricity-Supply Industry

51

constructed, and existing power plants fired by anthracite and lignite coal were retrofitted with flue gas cleaning equipment. The investment spike in the 1970s is also due to the construction of several nuclear power plants. With the construction of nuclear power plants, the cost-required share in the base-load power plants was increased. These power plants provide base-load electricity year round and have low variable production costs. In addition to nuclear power plants, the base load is supplied by run-of-river power stations and lignite-coal power plants. The development for the period 1970 to 1991 is shown in Figure 2.

Net electricity consumption

350 Investments (nominal) Investments (real)

300

250

200

150

100 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991

Source: VDEW; Ifo Institute.

Figure 2: Electricity Consumption and Gross Fixed Formation of General Electricity Supply from 1970 to 1991 – West Germany, 1970 = 100

In the 1990s German reunification posed new challenges for the German electricity sector. The overhaul and modernisation of transmission and distribution facilities in eastern Germany stood in the foreground. A high priority was the expansion of the power grid between eastern and western Germany, with enhanced reliability of the electricity supply in eastern German. In addition to the closing of numerous power plants, many other plants were modernised and new ones were constructed. High investment expenditure went especially for the construction of several large lignite-coal power plants. In western Germany investments were relatively low in the 1990s. Not until recent years have investment expenditures been increased again, a trend that is likely to continue in light of the high replacement needs (see Figure 3). The development of gross fixed capital formation in the general electricity supply shown in Figure 3 is based on surveys of the association of the electricity-supply industry among its members and on projections based on these surveys. Conse-

52

Hans-Dieter Karl

quently neither the investments of the industrial electricity producers nor the spending of operators of wind power plants that are not members of the association are included. The current expansion of investments in the general electricity supply sector result from both an increase in expenditures for power plants as well as increasing investments in transmission and distribution plants. The other investments remain almost unchanged. If these investment plans are implemented, the entire electricity investments in 2007 would achieve a level last reached in 1998. Since the newly constructed power plants are almost exclusively as substitution for existing plants, the total power capacity of all power plants will change only little. It is to be assumed that the new power plants will be more reliable than the old plants; for this reason and only in comparison to the conventional power plants, they will need lower backup requirements.

10000 9000 8000 7000 6000 5000 4000 3000 2000 1000 0 80

82

84

86

88

90

92

94

96

98

00

02

04

06

Source: VDEW; Ifo Investment Survey; Calculations of the Ifo Institute.

Figure 3: Gross Fixed Capital Formation in the General Electricity Supply in Germany (nominal) – after 1991 incl. Eastern Germany

3. Conditions for Electricity-Sector Investments In the coming years up to 2025, determining factors for the investments of the electricity sector will be both restructuring as well as the replacement and modernisation of the plants. That is particularly the case for the power plants and to a lesser extent for the power grid. The investments must satisfy the energy-policy goals of efficiency, security of supply and environmental compatibility. The determinants for the investments of the electricity industry do not differ fundamen-

Investments of the German Electricity-Supply Industry

53

tally from that of other industries. The specific determinants for the electricity industry are:  Estimated turnover: The growth in power consumption, which was the main motive for investments in past years of strong power consumption, has lost importance.  Prices: The fluctuation of electricity prices in the light of increasing competition and regulatory intervention as well as fuel price developments remain determinants.  Extent of the capital stock: Renewal of existing plants is gaining importance as an investment motive.  Regulatory framework: Rules of competition and regulations increase the investment risks; environmental regulations influence the spectrum of investment alternatives; energy-policy measures necessitate specific measures on the part of enterprises.

Even if the importance of growth in power consumption has weakened as an investment motive in recent decades, it should retain its strong influence in the coming years. Whereas from 1991 to 2000, not least because of German reunification, it appeared that the growth of GDP and the increase in power consumption had been clearly decoupled compared with the preceding decade, the situation now is quite different. The elasticity coefficient of the relationship between change in power consumption and change in GDP was 0.43 between 1991 and 2000 and 0.96 in the period 2000 to 2006. Assuming that the annual growth rate of the economy up to 2010 will average 2 percent, at an elasticity coefficient of 0.5 a gross power consumption of 640 TWh (terawatt-hours) would result; in 2006 gross power consumption already amounted to 615.8 TWh. Up to 2025, as a result of the declining population, average economic growth will be lower, at a 1.4 percent annual rate. With the relative decline in the importance of the manufacturing sector, economic growth will probably be less power intensive and the elasticity coefficient will fall to ca. 0.2. In 2025 gross power consumption, based on these assumptions, should amount to 670 TWh. That would be an annual average increase 2006 and 2025 of 0.4 percent. This is assuming that domestic power demand is completely met by domestic power plants; gross power consumption thus corresponds to gross electrical power production. For the growth in power consumption of 54 TWh, an additional power plant of 12,000 MW (megawatt) would be necessary, under the assumption of an average utilisation of plant of 4,500 hours a year. Since, however, the German electricity sector recorded an electricity export surplus in 2006 of 20 TWh, additional plants necessitated by increasing power consumption could be reduced to about 8,000 MW by reducing the export surpluses. It should be pointed out that our projection of the development of gross power consumption / gross power production is at the upper limits in comparison with other studies. In their reference forecast for the Energy Report IV, EWI / Prognos

54

Hans-Dieter Karl

(Bundesministerium für Wirtschaft 2005, p. 9) forecast a gross electrical power production of 617 TWh for 2010, 594 TWh for 2020 and 584 TWh for 2030. The German Energy Agency, Dena, arrives at far lower values in their scenario. With the implementation of measures for the further exploitation of energy-efficiency and energy-savings potential, a reduction of gross electricity generation is to be reduced to ca. 590 TWh already in 2010; in 2020 gross electricity production would amount to 570 TWh. According to this scenario the conventional condensation power plants (fossil, nuclear) would have a production share of 59 percent in 2020; the remaining production would come from renewable energy and from combined heat and power units (Bundesministerium für Wirtschaft, Bundesministerium für Umwelt, Naturschutz und Reaktorsicherheit 2006, p. 59 f.). In light of the above considerations, however, it is questionable whether the possibilities of the exploitation of electricity-savings potential exhaustion have not been assessed too optimistically, under the existing conditions. Since the prices of power-plant fuel are, in general, subject to price formation on the international market and are thus difficult to forecast for the long term, they only have a comparably small effect on changes in investment strategy. However, there are ways to reduce this dependency on world market prices for fuels. Lignite coal, which for all practical purposes can only be used domestically, can be used increasingly for electrical power production. Furthermore, the use of renewable energy can help counter increasing fuel prices. Also electrical power production from atomic energy, for which fuel costs play only a minor role, would be suitable to counteract higher fuel costs; with the renunciation of the further use of atomic energy in Germany, a stabilizing element would be lost. In the power station area, a number of challenges result in connection with the updating of present facilities. In the period under review, nuclear power plants with a capacity of 21,350 MW are to be replaced; this is also the case for numerous coal-fired power plants. A structural change results, on the one hand, in that an increasing number of natural gas power plants are being built for environmental reasons and, on the other hand, from the continuous construction of power generators based on renewable energy, particularly from wind power plants. Because of the higher prices for natural gas in comparison to coal, the use of such plants will be primarily limited to the medium-load range. With the gradual renewal of the stock of plants with power plants that are considerably more efficient than the plants that they replace, the efficiency of the transformation of energy will be clearly increased. The addition of plants that use renewable energy will also contribute to an increase in the average energy efficiency in electricity production. This will also mean a perceptibly lower environmental impact from air pollutants and climate-endangering emissions. The framework conditions for energy and environmental policies have a decisive influence on investments in the electricity supply sector. With the deregulation of the electricity supply and the increasing introduction of competition in this market, the risks for investments have clearly increased in comparison to the former

Investments of the German Electricity-Supply Industry

55

situation of a state-protected monopoly position. In this new situation, the interest rate demands of the capital providers may also increase. Moreover, the environmental protection regulations also have a special influence on the spectrum of investment alternatives. The use of lignite coal for electricity generation, which would be clearly attractive from an economic viewpoint, is limited because of its environmental effects. Immediate effects on the relevant energy sources for electricity generation result from the development of the prices for CO2 certificates. With increasing certificate prices, the competitiveness of natural gas, for example, improves over other fuels. However, here too the problem of estimating the longerterm development of certificate prices exists. Finally effects on investment behaviour emanate from specific energy-policy measures. The policy-driven expansion of renewable energy confronts the operators of power-grids with new challenges. In order to ensure the integration of electricity from these plants, specific investments in the power-grid must be made.

4. Electricity Supply Investments to 2025 The investments discussed in this section must be seen in the light of the above considerations. On the one hand, they are based on the present plans of the supply enterprises, on the expectations and forecasts of different institutions as well as on our own appraisals. In the area of power generation, German and foreign investors have announced the construction of conventional power plant capacities of ca. 20,000 MW by 2012; in addition about 10,000 MW power will come from renewable energy (Bundesministerium für Wirtschaft und Technologie, Bundesministerium für Umwelt, Naturschutz und Reaktorsicherheit 2006, pp. 49 ff. and Bantle 2006, pp. 22 ff.). By 2025 an estimated 24,000 MW of additional conventional power plant capacity will be built. The investments in power plants using renewable energy should increase strongly under current conditions; by 2025 the additional capacity will be increased to 35,000 MW. Of the power provided by conventional power plants in 2025, about half will be accounted for by coal-fired power plants, and about a third by gas-fired plants and the rest of the power plants; of the power generation from renewable energy, wind energy (land- and sea-based) will have a share of approximately two thirds. Here, to safeguard the operational efficiency of electricity production, a continuingly high share in electricity production is assumed for lignite coal in the basic load. In total, due to the greater efficiency of the added power plants, the average degree of efficiency of electricity production, which was ca. 41 percent in 2005, will be increased by about 10 percentage points by 2025. The power plant investments are calculated on the basis of specific investments per power unit. We have assumed that the values for the individual power plant types remain constant in real terms for the forecast period. Calculations are average values for the individual plant types. The following specific investments were applied (Bartels, M. / Peek, M. / Schulz, W. 2005, p. 233):

56

Hans-Dieter Karl

 A 1,200 / kWel for lignite coal plants,  A 1,050 / kWel for anthracite coal plants,  A 550 / kWel for combined cycle gas turbine (CCGT) plants.

For the other power plants a value of A 1,500 / kWel (kilowatt electric) was used as a basis. For electricity generation from renewable energy, an average value of A 2,500 / kWel was used for the range of plants. It was assumed here, that wind power generation will increasingly be build offshore. Under these assumptions, investments in prices of 2005 for the construction of conventional power plants with a capacity of 44,000 MW (15,000 MW natural gas, 7,000 MW lignite coal, 16,000 MW anthracite coal and 6,000 MW other power plants) in the period 2005 to 2025 amounted to A 42 billion; for the electricity generation from renewable energy considerably more must be invested: A 87.5 billion. Investments in transmission and distribution plants should remain at the level of 2005 in the forecast period. This means that in 2005 prices, ca. A 2 billion per year will be spent on the power grid or ca. A 40 billion by 2025. In addition to the obligatory measures for replacement, modernization and enlargement investments, also included is the spending that must be done for the integration of electricity generated from wind power and for strengthening the interfaces in the international co-ordination of electricity transmission. Finally, investments in the miscellaneous category must be considered. We assumed that an average annual amount of A 0.5 billion or a total of A 10 billion up to 2025 should be applied. In total, the investments amount to approximately A 180 billion for the electricity supply between 2005 and 2025; this corresponds to an average annual amount of about A 9 billion. The investment volume without the plants for electricity generation from renewable energy sources would amount to A 92 billion. However, without the use of renewable energy, conventional power plants would tend to be larger and differently structured.

5. Effects of the Alterations of the Electricity Supply Because of changes in the institutional framework conditions, particularly the introduction of competition into the electricity market and the massive promotion of renewable energy for the generation of electricity, the structure of the electricity supply will continue to change. Especially of the addition of electricity-production plants based on renewable energy sources will mean that production plants will be preferred that enable flexible operation and that have comparatively low load factors. They incur lower capital costs and tendentially higher fuel costs that power plants that have been built without the feed-in of electricity from renewable energy (Verband der Elektrizitätswirtschaft 2006, p. 8). This will both reduce the shortterm investment risks as well as facilitate the integration, for example, of electrical power from wind energy, which is subject to strong fluctuations. In addition, these

Investments of the German Electricity-Supply Industry

57

power plants can assume, in a more cost efficient way, the reserve functions, which are more plentiful in electricity generation with renewable energy. An example of the problem is that only approximately 6 percent of present wind energy capacity can be seen as sufficiently assured (Deutsche Energie-Agentur 2005, p. 12). The need for the so-called standard and reserve power has clearly increased with the expansion of renewable energy for electrical power production. In particular, the strong increase in some regions in standard power requirements for maintaining grid stability may have led to a lower supply security in some areas. Because of the phasing out of the use of atomic energy in Germany, which we assume will happen, and the obligation Germany has assumed internationally for lowering CO2 emissions, the fuel structure of electrical power production will change radically in the coming decades. In Table 2 the development is shown up to 2025. Because of the perceptible increase in power plant efficiency and the enhanced use of high-efficiency wind power plants, fuel input will decline despite increasing electrical power production. Of the fossil energy sources, the use of lignite coal will remain largely unchanged, and the combustion of anthracite coal will be expanded slightly and that of natural gas strongly. With these changes in the fuel structure, import dependency will doubtlessly increase since both natural gas and anthracite coal are increasingly imported. By abandoning atomic energy, an important element of the diversification of the energy supply is lost that can only be partly compensated by the use of renewable energy. The basic assumption is that the supply security will tend to worsen. Table 2 Fuel Use for Electricity Generation Energy source Anthracite coal Lignite coal Other energy sources Heating oil Natural gas Hydropower / Aeropower Nuclear energy Total a)

2005a) 40.4 50.9 3.2 3.4 15.6 5.8 60.7 185.1

2010

2020

2025

Energy use in Mill. t SKE 40.5 41.9 46.5 50.2 47.4 50.4 9.5 10.8 12.3 3.7 4.2 4.6 20.5 23.3 27.1 9.8 11.9 14.4 47.0 35.0 8.2

48.1 50.0 13.5 4.8 26.7 15.5 0.0

181.1

2015

174.6

163.5

158.6

Preliminary.

Source: AG Energiebilanzen; Ifo Institute.

Electrical power production based to a great extent on fossil fuels automatically has high emissions of particles that can affect the climate. Here we only

58

Hans-Dieter Karl

examine the carbon dioxide emissions. Based on the forecast assumptions, their output will increase in the forecast period by about 13 percent to 343 million tonnes due to the stronger use of fossil fuels (see Table 3). The higher output of carbon dioxide is particularly due to the increased combustion of anthracite coal and natural gas. Table 3 CO2-Emissions from Electricity Generation – Nuclear energy use will be terminated – Energy source

2005

2010

2015

2020

2025

125.1 163.8 4.3 2.1 44.4

129.6 162.4 4.7 2.2 43.7

339.8

342.6

Anthracite coal Lignite coal Other energy sources Heating oil Natural gas

108.8 165.4 1.1 1.6 25.6

109.1 163.0 3.3 1.7 33.6

millions of t CO2 112.9 154.0 3.8 1.9 38.2

Total

302.4

310.7

310.9

Source: Umweltbundesamt; Ifo Institute.

It must be noted that with this change in the structural of electricity generation no contribution can be made to the lowering of carbon dioxide emissions. Although it cannot be ruled out that incentives can be created to reduce emissions by way of additional political measures such as a massive shortage of CO2 certificates, such measures would have negative macroeconomic effects. Finally energy policy measures can be taken to slow down the increase of electricity consumption to a greater extent than assumed here or even lead to a reduction and thus contribute to the prevention of emissions; but such measures often cause non-negligible additional costs. As a comparison, the CO2 output of the electricity supply was determined in a status-quo calculation assuming the further use of atomic energy. It was assumed that the output of nuclear power plants decreases because of the closure of several older plants from 163 TWh in 2005 to 140 TWh in 2025. In this calculation it was also assumed that renewable energy will continue to be enlarged. The use of natural gas for electrical power production will also be expanded, if not to the same extent as in the above scenario, whereas the use of anthracite coal will remain largely constant and the use of lignite coal for electricity production will be clearly reduced. Under these conditions, carbon dioxide emissions from 2005 to 2025 will fall by around 77 million tonnes under the levels in 2005 or by one quarter. In comparison to the result in Table 3, the emissions for 2025 are about 117 million tonnes lower (see Table 4). At the same time, investments in power generation plants up to 2025 could be about A 17 billion lower; this spending would then be postponed until later years.

Investments of the German Electricity-Supply Industry

59

Table 4 CO2-Emissions from electricity generation – Nuclear energy use will be continued – Energy source

2005

2010

2015

2020

2025

Anthracite coal Lignite coal Other energy sources Heating oil Natural gas

108.8 165.4 1.1 1.6 25.6

101.3 153.7 3.3 1.7 31.8

millions of t CO2 105.4 118.0 3.8 1.9 36.5

110.4 87.8 4.3 2.1 37.6

108.0 72.0 4.7 2.2 38.6

Total

302.4

291.9

265.6

242.3

225.6

Source: Umweltbundesamt; Ifo Institute.

Faced with the extensive investments in plants for the electrical power production from renewable energy, which largely have a much lower exploitation capacity over the year, the use of all power plant capacities decreases. According to Energiereport IV the average full-utility hours per year fall between 2000 and 2030 from 4,490 to 4,024 hours (Bartels, M. / Peek, M. / Schulz, W. 2005, p. 237). The utilisation of power plants in our forecast should play out in the same way. Since by way of investments particularly in renewable energy a comparably high capital stock is built up, a clear drop of capital productivity is to be expected vis-à-vis the initial year. In return, however, in accordance with the political goal of advancing the use of renewable energy, as a result of its increased use for electricity generation there is an increase in productivity with relation to the use of fossil fuel; with renewable energy a substitution takes place of capital for combustible energy. The question arises, however, of whether this investment strategy will lead to higher overall welfare in light of the total additional costs that are involved. Ultimately this restructuring of the electricity-supply industry in Germany advanced by the state authorities can only be assessed in terms of economic-, energy- and environmental-policy goals.

References Bantle, Ch. (2006): “Investitionserhebung 2005: Investitionen der Stromversorger sind stabil,” in: ew 23 – 24. Bartels, M. / Peek, M. / Schulz, W. (2005): “Energiereport IV: Entwicklung des deutschen Elektrizitätsmarktes bis 2030,” in: Zeitschrift für Energiewirtschaft, 29 (2005) 3. Bundesministerium für Wirtschaft und Technologie, Bundesministerium für Umwelt, Naturschutz und Reaktorsicherheit (2006): Energieversorgung für Deutschland, Statusbericht für den Energiegipfel am 3. April 2006.

60

Hans-Dieter Karl

Bundesministerium für Wirtschaft (2005): EWI / Prognos – Studie, Die Entwicklung der Energiemärkte bis zum Jahr 2030, Energiereport IV – Kurzfassung. Deutsche Energie-Agentur (2005): Energiewirtschaftliche Planung für die Netzintegration von Windenergie in Deutschland an Land und Offshore bis zum Jahr 2020, Zusammenfassung, DIW-Wochenbericht 8 / 2007: “Primärenergieverbrauch in Deutschland nur wenig gestiegen”. International Energy Agency – IEA – (2006): World Energy Outlook. Verband der Elektrizitätswirtachaft – VDEW – (2006): Fakten, Berlin, September 2006.

Investments of the German Electricity-Supply Industry By Ingo Ellersdorfer*, Stephan Kempe, Uwe Remme, Markus Blesl, Ulrich Fahl, and Alfred Voß

Abstract The paper analyzes the long-term development of the electricity generation system in Germany under different environmental- and technology-related policy frameworks. Adopting the CO2-reduction targets that have been agreed on in the Kyoto protocol, the perspectives of various electricity production technologies have been analyzed within five contrary scenarios. Focusing on the contribution of renewable energy sources, fossil fuel technologies with carbon capture and storage and nuclear-based electricity production, it has been shown that clean coal technologies and nuclear-based electricity generation are projected to be the most costeffective technology options for electricity supply and CO2 mitigation in Germany. JEL Classification: Q48, Q55, D92 Keywords: Model-based energy scenarios, investment decisions, electricity generation

1. Background Regarding the investment needs of more than 70 GW until 2030 for replacing and adding electricity generation capacity in Germany,1 energy policy has a strong impact on the structural-, economical- and environmental-related development of the electricity generation system. Given the CO2 reduction targets, that have been adopted from the Kyoto protocol and the further efforts that have to be made to realize a sustainable energy supply, various technological options are available that have to be analyzed concerning their contribution to a secure, cost effective, and environmentally friendly future energy and electricity-supply system, respectively. As political measures strongly influence the investment decisions of energy companies, the policy has to take these impacts into consideration. Therefore, * Corresponding author and co-authors. Institut für Energiewirtschaft und Rationelle Energieanwendung (IER), Universität Stuttgart, Germany. E-mail Adresse: ingo.ellersdorfer@ier. uni-stuttgart.de. 1 See Karl (2007).

62

I. Ellersdorfer, S. Kempe, U. Remme, M. Blesl, U. Fahl, and A. Voß

environmentally oriented policies like the committed CO2-reduction targets lead to diverse structural changes in the energy system if they are combined with different technology-related policy measures like the promotion of renewable energy supply, the nuclear phase-out or the use of carbon capture and storage technologies. To quantify the technological and economical implications of different policy frameworks, energy system models can be applied to simulate alternative scenarios to project the range of possible energy system developments. Model-based scenario analyses are therefore well-suited tools for projecting long-term investment decisions. 2. Model The German bottom-up energy system model TIMES-D is based on the TIMES2 model generator, which follows a partial equilibrium approach for representing, optimizing and analyzing energy systems on local, regional, national, or global scales. It is demand-driven and technology-oriented. GDP, population, heated floor space requirement per person, freight-kilometer demand, etc., are the driving forces to determine the energy demand by sectors. Given a set of technological options to meet supply demand along with assumptions on energy prices and resource availabilities, the model optimizes the total system costs to meet the sectoral energy demand. Demand sectors considered are industry, service / commercial, residential, and transport, which are further disaggregated. The German model includes more than 380 end-use technologies encompassing several vintage classes divided into the four above-mentioned demand sectors and represented by techno-economic parameters such as utilization factor, energy efficiency, lifetime, capital costs, operating and maintaining costs, etc. The supply side of the model covers energy-conversion processes including petroleum refining, coke production, heat and electricity generation, etc. It includes over 120 conversion technologies for central electricity and district heating based on fossil (coal, oil, gas), nuclear and renewable (hydro, wind, solar, biomass, geothermal) resources. The supply-side technologies are divided into different groups of technology generations, depending on the plant size and operation time. New technologies that are available in the medium or long term for power generation include Integrated Gasification Combined Cycle (IGCC), ultra super critical coal, IGCC with CO2 sequestration, fuel cells, etc. In addition, assumptions on energy prices, resource availabilities and energy trade, etc., are major input parameters for the model.3 2 TIMES has been developed within the Energy Analysis Programme (ETSAP) of the International Energy Agency (IEA). Developments and refinements still continue and replace the widely used MARKAL approach. 3 See Remme (2006) and Blesl (2007) for further details.

Investments of the German Electricity-Supply Industry

63

3. Scenarios The TIMES-D energy system model is used to analyze different scenarios regarding the perspectives of various electricity generation technologies given alternative policy measures for climate protection in Germany. Besides one reference, four different policy scenarios have been analyzed.4 Table 1 gives an overview of the scenarios. Table 1 Description of Scenarios Analyzed Scenario

Description

GHG reduction target

REF

Reference Scenario

– Extrapolation of present energy policy – Nuclear phase-out – No specific target for GHG reduction

PEE

Preference for Renewable Energy Sources

– 2020: 20 % of electricity generation – 2030: 30 % of electricity generation Nuclear phase-out No clean coal technologies allowed

2010: –21 % 2020: –35 % 2030: –50 %

CCT

Clean Coal Technologies

– Clean coal technologies allowed – Nuclear phase-out – No promotion of renewables

2010: –21 % 2020: –35 % 2030: –50 %

ERL

Efficient Resource Utilization, Lifetime Extension

– Nuclear lifetime extension to 60 years – Clean coal technologies allowed – No promotion of renewables

2010: –21 % 2020: –35 % 2030: –50 %

ERN

Efficient Resource Utilization, New Nuclear Capacities

– New nuclear capacity investment – Clean Coal Technologies allowed – No promotion of renewables

2010: –21 % 2020: –35 % 2030: –50 %

For the reference scenario an extrapolation of the present energy policy, including the agreement on a nuclear phase-out has been implemented to analyze the development of the energy system under a business-as-usual scenario. Moreover, no explicit GHG reduction target constraints the deployment of fossil fuels for energy service supply. As a consequence, mainly lignite and hard coal with a minor share of natural gas dominate the electricity generation in Germany in the year 2030, holding approximately 82 % of overall electricity production (cf. Figure 1). Due to the strong use of fossil fuels, the energy-related CO2 emissions (830 million tons) are projected to be 61 % higher than the CO2 limits within the four other scenarios. Regarding the use of renewable energy sources for electricity generation, the increase of wind energy supply is ongoing, resulting in generation of 50 TWh in 2030. Compared to biomass (35 TWh) and hydro energy (25 TWh), wind stays the 4

The analyses are based on Voß (2006).

64

I. Ellersdorfer, S. Kempe, U. Remme, M. Blesl, U. Fahl, and A. Voß

most important renewable energy carrier after 2010. This strong increase in wind and biomass-based electricity production is mainly driven by the policy induced promotion of renewables by the Erneuerbaren Energien Gesetz (EEG)5. The dominant position of hard coal and lignite is also reflected by the capacity investments in the German electricity supply system until 2030. The capacities of fossil-fuel-based power plants increase by approximately 48 GW between 2011 and 2020 and additionally by 24 GW between 2021 and 2030, followed by 17 GW of new wind converter capacities until 2030 (cf. Figure 2). Given the subsidy-induced investment in new solar PV capacity of 9 GW until 2030, photovoltaics are projected to hold a low proportion of overall electricity generation (0.9 %) in Germany.

Net electricity generation [TWh]

700 600 500 400 300 200 100

statistics

2010

2020

ERL

ERN

PEE

CCT

REF

ERL

ERN

PEE

CCT

REF

ERL

ERN

PEE

CCT

REF

2005

1990

2000

0

2030

Figure 1: Net Electricity Generation in Germany

In the second scenario (PEE) the impact of a strong promotion of renewable energy sources for electricity production has been analyzed. As the emissions of CO2 are constrained to 515 million tons in 2030, i.e. -50 % compared to 1990, wind (65 TWh), biomass (53 TWh) and the import of renewable electricity (49 TWh) become the most important sources for achieving the renewable targets (cf. Table 1) in Germany. Due to geographical and technical restrictions in Germany, hydro-based production remains constant at approximately 25 TWh per year. Another interesting result of the PEE scenario is the strong increase in natural-gas-based electricity generation, which increases to 290 TWh in 2030 and leads to 59 GW of newly installed power plant capacity between 2011 and 2030. This 5

See EEG (2004).

Investments of the German Electricity-Supply Industry

65

development can be explained firstly by the shift of coal-based electricity production to natural-gas-based production to reach the CO2 targets and secondly by the need of more flexible generation capacities for providing reserve power due to the stochastic nature of wind energy supply. As the use of clean coal technologies with carbon capture and storage is not allowed in the reference and the PEE scenario the contribution of these technologies have been analyzed within the CCT scenario. One of the major results that can be observed is the significant contribution of lignite and hard coal to overall electricity generation, which keeps coal the most important energy carrier for electricity generation in Germany until 2030. Almost 76 % of the electricity production is generated by power plants with carbon capture and storage technologies resulting in approximately 223 million tones of CO2 stored in 2030. Moreover, clean coal technologies are more cost effective to reduce CO2 emissions in the German electricity supply system than a further increase of renewable energy use.

Net capacity investments [Gw el]

80 70 60 50 40 30 20 10 0 REF PEE CCT ERL ERN 2006-2010

REF PEE CCT ERL ERN

REF PEE CCT ERL ERN

2011-2020

2021-2030

Figure 2: Net Capacity Investments in Germany

Allowing for an optimization of the electricity supply system by making all technological options available without restricting or preferring investment decisions to selected generation technologies, structural development of the German generation portfolio is different from the REF, PEE, and CCT scenarios. Within the two scenarios ERL and ERN efficient resource utilization including the possibility of lifetime extension (ERL) as well as new investments in nuclear power plant capacity (ERN) is simulated. As all technological options are allowed to be invested in the energy system, nuclear is projected to be the most cost effective option, first for electricity generation and second for CO2 mitigation in Germany.

66

I. Ellersdorfer, S. Kempe, U. Remme, M. Blesl, U. Fahl, and A. Voß

Given the option to extend the lifetime for nuclear power plants from 40 years to 60 years at additional costs of 110 A / kW per year in the ERL scenario it can be observed that nuclear generation capacities increase by 20 GW until 2030. Moreover, allowing for new investment in nuclear power plants in the ERN scenario, capacities increase by 59 GW, resulting in 78 % of overall electricity generation by nuclear. Beside the strong deployment of nuclear energy, clean coal technologies contribute to CO2-emission reduction by 127 million tons (ERL) and 15 million tons (ERN) stored in the year 2030, respectively. Regarding the overall installed generation capacities it can be seen that the higher utilization rates of thermal generation capacities like nuclear and lignite lead to lower overall capacity needs compared to the REF and PEE scenarios, which are characterized by higher shares of fluctuating capacities and therefore more capacity requirements. Furthermore, the ERN and ERL scenarios show higher overall electricity generation, mainly driven by a stronger use of electricity devices for heat supply, e.g. heat pumps. This result is in line with the diverse development of average electricity generation costs, which are lower in the ERL and ERN scenario compared to the REF, PEE, and CCT scenarios (cf. Table 2). Due to the high average generation costs, especially in the PEE scenario, options for electricity savings become more cost effective, leading to the lowest overall electricity generation level in 2030.

Installed net capacity [GW el]

180 160 140 120 100 80 60 40 20

2020

ERL

ERN

PEE

CCT

REF

ERL

ERN

PEE

CCT

REF

ERL

2010

ERN

CCT

REF

PEE

2005

1990

2000

0

2030

Figure 3: Installed Generation Capacity in Germany

In line with the differences in average electricity generation costs, the cumulative CO2-mitigation costs vary significantly among the scenarios. Compared to the baseline, the PEE supply results in the highest CO2-reduction costs of approxi-

Investments of the German Electricity-Supply Industry

67

mately 90 billion A2000 in 2030, whereas the option for investment in new nuclear generation capacities results in cumulative cost savings of 142 billion A2000. This result is mainly driven by the nuclear phase out in the reference scenario, which leads to an overall cost increase for electricity generation. Allowing for efficient resource utilization including the options for clean coal technologies and the investment in new nuclear capacities in the ERL and ERN scenarios, therefore allows reaching the CO2-reduction targets at negative costs compared to the business-as-usual scenario as it is defined by present energy policies. Table 2 Scenario Description Scenario REF PRE CCT ERL ERN

Cumulative CO2 mitigation costs until 2030 [billion A2000]

Average electricity generation costs in 2030 [Act2000 / kWh]

+24 +90 –73 –142

4.4 5.5 7.5 4.3 2.7

4. Conclusions To realize a secure, cost-effective and environmentally friendly future energy system, policy has to make all possible investment options available. The modelbased scenario analyses have shown that the significant CO2-emission reductions can be reached without taking a strong burden on the energy system. Mainly two technological options that can contribute to a sustainable energy supply system have been identified. First it has to be made sure, that the political framework for the deployment of clean coal technologies with carbon capture and storage for electricity generation will be developed. Due to the technological development of this type of power plants, it is necessary to provide investment security for the generating companies which plan to replace much of their existing capacities until 2030. Second it has been shown, that especially nuclear-based electricity generation can contribute to the emission reduction targets at low costs. As the analyses of possible lifetime extension of the existing nuclear power plants and the possibility for new investments have shown, this technology can be a cost effective option for emission reduction and electricity supply.

68

I. Ellersdorfer, S. Kempe, U. Remme, M. Blesl, U. Fahl, and A. Voß

References Blesl, Markus / Das, Anjana / Fahl, Ulrich / Remme, Uwe (2007): “Role of Energy efficiency standards in reducing CO2 emissions in Germany: An assessment with TIMES.” Energy Policy, 35, 2007, 772 – 785. EEG (2004): BGBl. Teil I Nr. 40 v. 31. 07. 2004, 1918 ff. Karl, Hans-Dieter (2007): “Investitionen der deutschen Elektrizitätswirtschaft”, Paper presented at ARGE Konferenz: Energiemärkte – Investieren, Konkurrieren, Regulieren, April 19th, 2007, Berlin. Remme, Uwe (2006): “Zukünftige Rolle erneuerbarer Energien in Deutschland: Sensitivitätsanalysen mit einem linearen Optimierungsmodell”, Dissertation an der Universität Stuttgart, IER-Forschungsbericht, Band 99, 2006, Stuttgart. Voß, Alfred (2006): “Wege zu einer nachhaltigen Energieversorgung in Deutschland”, In: Bernd Hillemeier (ed.), Proceedings of archatech Symposium I: Die Zukunft der Energieversorgung in Deutschland: Herausforderungen – Perspektiven – Lösungswege, November 21th, 2006, Berlin.

Asymmetric Strategic Investment Behavior in Network Industries: the Case of Natural Gas Distribution in Norway By Reinhard Madlener* and Pio Baake** Abstract In this paper we study the optimal strategic investment in network infrastructure when competing firms use different and exclusive supply technologies. The case is for natural gas distribution by pipeline vs. distribution of liquefied natural gas (LNG) by special trucks or small vessels. For the analysis we propose a simplified dynamic game-theoretic model. We assume that the more flexible LNG supplier can enter geographically dispersed markets simultaneously, whereas the pipeline supplier can enter new markets only by sequentially expanding the network. Furthermore, the LNG supplier as the leader has some sunk investment from which the pipeline company can benefit. Characterizing the open-loop Nash equilibrium and using a simple numerical example we find that the LNG supplier invests more in markets in which the pipeline supplier enters late. Increasing pipeline costs spur investment by the LNG supplier, whereas an increase in the investment costs of the LNG supplier has only a small impact on the investment decisions of the pipeline supplier. JEL Classification: C72, D21, D43, L13, L95, O33 Keywords: strategic investment, network industry, natural gas, LNG, second-mover advantage

1. Introduction In the economics, management science, and corporate finance literature valuations of alternative investment options are typically carried out by calculating the net present value (NPV) of the expected total payoff of a project that can be realized by a single investor (e.g., Brealey et al., 2005). If the expected NPV turns out to be positive, then the firm should invest, otherwise it should refrain from doing so. In the real world, the value of an investment is also influenced by the behavior of competing firms, if any, and the flexibility regarding the timing and size of an * Corresponding author. Email: [email protected], Tel: +49 241 80 97 162, Fax: +49 241 80 92 206. RWTH Aachen University, Faculty of Business and Economics / E.ON Energy Research Center, Templergraben 55, 52056 Aachen, Germany. ** DIW Berlin (German Institute for Economic Research), Department of Information Society and Competition, Mohrenstrasse 58, 10117 Berlin, Germany.

70

Reinhard Madlener and Pio Baake

investment, especially if at least part of the project cost is sunk. Both aspects can be expected to be critical components of optimal investment behavior, and should therefore be explicitly taken into account in an optimal investment analysis. Furthermore, if there is more than one investor, and if these investors face different cost structures, then strategic considerations also become an issue. Finally, if an investor invests in a physical network technology, then path-dependence (i.e., in the present context the need to sequentially invest into an interconnected grid) is an important restriction, too. In other words, some investors may be unable to freely choose in which market to invest at a certain point in time. Instead, they may have to account for the sequential nature of entering the different markets of interest over time, making strategic investment decisions more complex than otherwise. In the game-theoretic model proposed, we address the above-mentioned aspects, thus combining and contributing to different bodies of literature, including strategic investment, energy economics, and applied game theory. Our aim is to study asymmetric strategic investment behavior in network industries in the presence of an innovative technology that allows for a higher investment flexibility than the well-established technology, but suffers from the disadvantage of having lower economies of scale. The analysis is undertaken for a finite number of markets, and the dynamic deterministic model serves for investigating the optimal investment plan of two asymmetric firms. In recent years, research interest in strategic games applied to imperfectly competitive markets has grown considerably. Early deterministic game-theoretical investment models were introduced by Reinganum (1981) and Fudenberg and Tirole (1985). These inspired later work by Grenadier (1996) (real estate market), Smets (1995) (multi-national setup), Lambrecht and Perraudin (1997), and Grenadier (1999), among others. Slightly more recent contributions include Huisman (2000, 2001) (technology adoption), Weeds (2002) (R&D competition), Mason and Weeds (2005) (first-mover advantage, externalities), Décamps and Mariotti (2004) (impact of asymmetric information), Perotti and Rossetto (200, 2001) (internet portals as portfolios of entry options), and Shakleton et al. (2000) (market leadership). For a useful survey of this literature see Boyer et al. (2004b). In our research we investigate asymmetric strategic investment in a duopolistic network industry, and seek optimal investment rules within a continuous time framework. The assumption is that one of the firms (the ‘follower’) is slightly dominant, in that beyond a certain market size it outperforms the other firm (the ‘leader’) in terms of lower cost. The leader, in turn, has an incentive to grow in each market at a pace that optimizes revenues over the duration of time it can succeed in the market (i.e., from the time of its own investment until it is driven out of the market by the follower). Build-up of capacity is assumed to start at a level sufficiently low to make sure that the non-dominant firm has a competitive edge at the beginning. While the leader expands capacity and the opening up of new markets, the dominance of the follower gains momentum.

Asymmetric Strategic Investment Behavior in Network Industries

71

In principle, two types of equilibria may exist: one that implies tacit collusion (the follower agrees to postpone investment such that both firms are able to maximize returns), and one that implies that the follower uses its dominant position (the follower optimizes its second-mover advantage).1 In our case the secondmover advantage does not arise because of quality-improving technological progress, but mainly due to economies of scale and reduction in market entry costs due to sunk investments by the leader.2 In other words, the game considered here is not a preemption game, but a waiting game (cf. Moretto, 2003), and it is not of the winner-takes-all type, since both players are enabled to make profits, at least until the market is completely saturated and the leader driven out of all markets. The remainder of the paper is organized as follows: Section describes the emerging natural gas retail distribution market in Norway as an interesting case for applying our model. Section 3 introduces the dynamic model formulation developed and section 4 concludes.

2. Competing Natural Gas Distribution Systems as a Case Study 2.1 General Considerations About Pipeline vs. LNG Supply Systems

Traditional pipeline and the more innovative liquefied natural gas (LNG) systems are two alternative ways of transporting natural gas over long distances (wholesale market) or distributing natural gas locally (retail market). Cost functions of the two technologies are very different, and affected by factors such as steel prices, engineering capacity, and possibilities of joint project execution (synergies). LNG, as an alternative supply chain with considerable flexibility, has the potential to enhance competition in natural gas markets, and also to lower market entry barriers for natural gas. In recent years, the LNG market has seen rapid growth, not least due to major technological progress in LNG production and transportation. This has attracted more and more interest among energy market analysts and researchers alike, especially concerning market development at the international level. A useful and comprehensive survey on the functioning of the global LNG market has been provided by Jensen (2003). The author finds that due to the enormous capital expenditures required for complete LNG chains, risk diversification by a diversified portfolio of LNG supplies, tanker transportation capacity, and downstream commitments is costly, making such internal diversification only feasible for the largest market players (the so-called “Five Sisters‚). Smaller players may have to take refuge to 1 Note that collusive strategies between firms of unequal size aiming at the maximization of combined firms do not lead to an equilibrium (cf. Boyer et al., 2004a). 2 We assume that the leader paves the way for the follower by (i) inducing the final consumers to switch from other energy sources to natural gas (e.g., by advertising) and (ii) by investing in retail supply infrastructure components that can be used by both firms.

72

Reinhard Madlener and Pio Baake

regional and / or vertical integration instead. Jensen concludes that the LNG business remains highly complex, and that niche opportunities still exist both for marketers and upstream companies owning economically attractive gas discoveries (“LNG is, after all, a growth business‚; Jensen, 2003, p. 44). 2.2 Retail LNG Distribution in Norway3

Prospects for the further market diffusion of natural gas, apart from other factors, depend strongly on the limitations set by the supply infrastructure. One of these limitations is determined by the economic viability of extending the distribution infrastructure to the end-user, which can vary widely depending on the technology chosen and the prevailing local circumstances. While large-scale pipelinebased systems benefit considerably from economies of scale, they require huge initial investments and may in certain cases not be economical, especially when enduse energy densities are low and the topography difficult. In Norway a niche market has been developing in recent years where LNG is distributed either by special (super-insulated) tanker trucks or small vessels (Madlener and Jarsby, 2007).4 An important reason for the emergence of this new market is that up to now only few large (onshore) natural gas pipelines exist in Norway. Existing natural gas pipelines have often been built for the sole purpose of bringing natural gas from offshore fields onshore, in order to process and / or re-export it to larger markets in continental Europe and elsewhere. Moreover, the existing large onshore pipelines only range over relatively short distances near the Western coast, and do not form an integrated grid. However, they offer an opportunity for tapping some gas for local distribution. Small-scale LNG distribution systems have essentially four advantages: (1) relatively low investment costs for small capacities, (2) modularity, (3) locational flexibility, and (4) storability. This kind of small-scale natural gas distribution takes place in a saturated market for heat energy that has traditionally been supplied with a number of alternative energy commodities (essentially fuel oil, gasoil, LPG (propane and butane), district heating, biomass, and electricity). In general, consumers will only switch to natural gas if they can be convinced about the relative merits of natural gas for heating and other purposes in comparison with competing options. In recent years, the debate on the extension of natural gas supply in Norway has become controversial and politicized. On the one hand, certain stakeholders lobby 3 This section is largely based on Jarlsby (2004) and Madlener and Jarlsby (2005, 2007), respectively, which contain a more detailed account of the development and the economic aspects of of the inland LNG market in Norway. 4 Superinsulation is necessary in order to keep the gas in liquefied state at below –162 degrees Celsius. Liquefaction, while being expensive, has the advantage that the density of the natural gas is raised by a factor of about 600.

Asymmetric Strategic Investment Behavior in Network Industries

73

for heavy, state-sponsored investment into one or more pipelines, which would bring large quantities of natural gas to industrial centers, at least some of which have struggled to maintain their market position in the past. On the other hand, there are stakeholders that have argued for the promotion of modular and more flexible small-scale LNG technology systems that could enhance competition and provide end-users with natural gas that would otherwise not be supplied. So far the Norwegian Government has favored the LNG approach, which under present regulatory and market conditions is often cost-effective already today. One reason for this is the indirect benefit that accrues from the fact that competing products are taxed, whereas LNG is not. To push the LNG market further, a state-owned body has lately provided modest amounts of financial support for erecting several LNG receiving terminals that will be accessible by ship. Two interesting side issues of our study are related to competition. First, traditional semi-public regional electricity companies, currently in the process of redefining and reshaping themselves as energy and multi-utility firms, very much dislike the prospect of having to cope with new players intruding the regional gas markets. Second, several large oil companies, including Statoil and Norsk Hydro, apparently want to play a role in this market as well. The example of the burgeoning Norwegian LNG retail market can be used as a testbed for modeling asymmetric strategic investment behavior of network industries. The Norwegian LNG market could in fact emerge in many other countries in the world that are lacking extensive gas pipeline distribution systems today. Hence the model market that forms the basis for our discussion is more attractive and less exotic than it may seem at first sight.

3. Dynamic Strategic Optimization Model In this section we present a dynamic optimization model, which is inspired by the emerging competition between retail distribution of natural gas, either by pipeline system or in the form of LNG that is transported by truck or special vessel and re-gasified at the point of final consumption.

3.1 Basic Model Set-up

We consider two firms L and P and a discrete number of markets m ˆ 1; 2; . . . M. While firm L offers liquefied natural gas by truck / ship and is assumed to be able to serve each market independently from its activity on other markets, firm P relies on a pipeline to deliver natural gas. We assume that firm P will expand its pipeline only from market to market, that is, in order to serve market m it has to build a pipeline going from markets 1; 2; . . . ; …m 1† to market M. Firms’ costs to serve the markets are given by

74 …1†

Reinhard Madlener and Pio Baake cLm …tmL †

with

dcLm …tmL † 2 …xm ; m† : @xm @xm

We also assume that firms have to invest once in consumer-related infrastructure for the provision of access to the grid (local pipeline grid, connection to the buildings, metering devices etc.). These investment costs are sunk, and once the investment is made consumers can switch between the two firms at no cost. Let xLm and xPm denote the number of consumers firms L and P want to serve on market m and let denote the fraction of consumers that switch from firm j to firm i (i; j ˆ L; P and j 6ˆ i). Firms have the same cost function cc …xim ; xjm † for consumer-related infrastructure investments which, for simplicity reasons, is assumed to be linear: …3†

 cc …xim ; xjm † ˆ k max xim

 xjm ; 0 :

Finally, considering competition between firms we assume that firms compete in prices and that firm L has higher marginal costs to supply each consumer with natural gas. Firms’ marginal costs are constant; specifically, firm L incurs marginal costs c, while the marginal costs of firm P are normalized to zero. To simplify the analysis further, we assume that firm L leaves (is forced to leave) the market as soon as firm P has entered the market.5 P P Defining tmP :ˆ m iˆ1
ti , firms’ profits L and P on market m at time t can be written as (

…4†

L …xLm ; m†

ˆ

P …xPm ; m† ˆ

cxLm

0 

…5†

…xLm ; m† 0 …xPm ; m†

if

tmL  t < tmP

else , if else .

Collecting terms and taking into account the time firms need to reach the markets, firm L0 s discounted profit in period t ˆ 0 is given by 5 Note that considering the possibility of re-entry by firm L, and assuming immediate price adjustments, it is also an equilibrium if only firm P is active and chooses the monopoly price.

Asymmetric Strategic Investment Behavior in Network Industries …6†

L ˆ

" M Z X

min‰tmP ;1Š

tmL

mˆ1

75

# L …
†e

rt

dt

cc …xLm †e

rtmL

cLm …tmL † :

Similarly, firm P0 s discounted profit in period t ˆ 0 can be written as …7†

P ˆ

" M Z X mˆ1

1

tmP

# P …
†e

rt

dt

P cc …xPm ; xLm †e rtm

P cPm …
tmP †e rtm

1

:

In the following we analyze the firms’ optimal investment decisions. We assume that firms simultaneously decide on their investments in period t ˆ 0, that is, we focus on an open-loop Nash equilibrium.

3.2 Optimal Investment and Network Expansion

3.2.1 Consumer-related Infrastructure To solve this game we start with the firms’ investment decisions with respect to consumer-related (access) infrastructure. Employing (6), maximization with respect to xLm shows that for firm L the optimal number of consumers to supply, xLm …tmL ; tmP ; m†, is implicitly given by 1 P L @L …
† 0  …1 e r…tm tm † k and r @xLm   1 P L @L …
† k xLm 0 ˆ …1 e r…tm tm † r @xLm

…8†

for tmL < tmP and by xLm …tmL ; tmP ; m† ˆ 0 for tmL  tmP . Similarly, the optimal number of consumers xPm …xLm ; tmL ; tmP ; m† firm P wants to serve on market m can be characterized by (assuming xPm > xLm ) …9†

0

1 @P …
† r @xPm

k

and

0ˆ

 1 @…
† r @xPm

 k …xPm

xLm † :

Since L …
† and P …
† are strictly concave, equations (8) and (9) lead to unique L P P L P solutions, xL m …tm ; tm ; m† and xm …tm ; tm ; m†. Note further that …10† …11†

sign

@xL m …
† ˆ sign…e @tmP

r…tmP tmL †

†>0

and

L xP m …
† > xm …
† :

That is, firm L0 s tends to invest more in markets where its monopoly position lasts longer (i.e., where firm P arrives later), while firm P (sequentially) invests in all markets in order to serve additional consumers.

76

Reinhard Madlener and Pio Baake

3.2.2 Distribution Infrastructure L P We now turn to firm L0 s investments to reach the markets. Using xL m …tm ; tm ; m†, L P let L …tmL ; tmP ; m† :ˆ 1r L …
† kxL …
†. Then, the optimal t …t ; m† – if it exists – can m m m be characterized by

…12†

 min tmL jrL …tmL ; tmP ; m†e

rtmL

‡

 dcLm …tmL † ˆ 0 : dtmL

Note that if (12) is empty, then firm L will never enter marketm. Employing (10) and assuming that tmL …tmP ; m† exists, (12) implies that @tmL …tmP ; m† @tmP < 0. That is, firm L invests the more to reach market m the longer it takes firm P to build its P pipeline to that market. Note further that tmL …tmP ; m† does not depend on tm‡i with i ˆ 1; 2; . . . ; ‰M …m ‡ 1†Š. To solve firm P0 s investment problem we can rely on the optimality principle of dynamic programming. That is, given any expansion path to the first m markets, the expansion to the rest of the …M m† markets must be optimal. Hence, defining P …tmP ; tmL ; m† as the equilibrium profit firm P earns on market m, i.e., P …
† : ˆ 1 0 k…xP xL m m †, firm P s discounted profit from expanding to the rest of the r P …
† …M m† markets can be written as …13†

Pm‡1 …
† ˆ

M h X

P …
; j†e

r…tjP tmP †

i cPm …
tjP † e

rtmP

:

jˆm‡1

L Using (13) the optimal
tmP …tmp 1 ; tL †, with tL :ˆ …t1L ; t2L ; :::tM † taken as given, is therefore determined by

…14†

   min
tmP r P …
; j† ‡ Pm‡1 …
† e

r
tmP

‡

 dcPm …
tmP † ˆ 0 : d
tmP

Likewise, if (14) is empty, then firm P will not expand its pipeline grid to markets m; …m ‡ 1†; . . . ; M. Furthermore, condition (14) shows that firm P’s investment incentives are driven by two potentially countervailing effects. On the one hand,
tmP tends to be the lower the higher m is. This effect is based on (10) and the fact that firm P’s profits on each market are the higher the more firm L has already invested in consumer-related infrastructure. On the other hand,
tmP tends to increase with m since the profits that can be earned by reaching the remaining …M m† markets earlier in time decrease with m. This second, countervailing effect implies that firm P0 s expansion path tends to be slowed down over time. Considering the Nash equilibrium with respect to tmP and tmL note that
tmP …tmp 1 ; tL † depends on all investment decisions of firm L. Eqs. (12) and (14) for all M markets imply that a rather complex non-linear system of 2M equations has to be solved. In the following subsection, we therefore present a rather simple example with only four markets to illustrate the working of our model.

Asymmetric Strategic Investment Behavior in Network Industries

77

3.3 Illustrative Example

In order to illustrate the qualitative properties implied by (12) and (14) we analyze the following numerical example. There are M ˆ 4 identical markets with linear inverse demand functions p…x; m† ˆ 2

x:

Firms’ investment costs to reach the markets are given by cLm …tmL † ˆ

kmL tmL

and

cPm …
tmP † ˆ

kmP ;
tmP

where we assume kmP > kmL > 0. The other cost parameters are c ˆ 0:1

and

kˆ

1 : 2

Using these specifications and starting with kmP ˆ 3 and kmL ˆ 0:1 we get the following equilibrium decisions: m xLm tmL tmP

1

2

3

4

0.6 0.371 1.115

0.812 0.35 2.351

0.864 0.349 3.797

0.89 0.348 5.707

The comparative static effects of changing the firms’ costs are as expected. More specifically, increasing kmP leads to the graphs for the firms’ investment timing decisions shown in figures 1 and 2. Figure 3 shows the corresponding quantity decisions xLm for different kmP . t

tP4

6

tP3

4

tP2

2

tP1 3. 5

4

4. 5

5

kPm

Figure 1: Firm P’s timing decisions

Moreover, increasing kmL shows that with higher investment costs firm L refrains from serving the first markets. More specifically, for kmP ˆ 3 and increasing kmL firm

78

Reinhard Madlener and Pio Baake

L stops to serve market 1 with kmL  0:23. Furthermore, firm L will serve market 2 only if kmL  1:8 holds. Although an increase in kmL leads to an increase in tmL and thus to a decrease in xLm , firm P does not reduce its investments very strongly. Since firm P has to invest in markets 1 and 2 in order to reach markets 3 and 4, its investments do not react very much to reductions in xL1 and xL2 either. Figure 4 shows t2L and t2P for 0:23 < kmL  1:8. t 0. 37 0. 365 0. 36

tL1 tL2 tL3 tL4 kPm

0. 355 0. 35 3. 5

4

4. 5

5

Figure 2: Firm L’s timing decisions xL4 xL3 xL2

0. 9 0. 85 0. 8 0. 75

xL1

0. 7 0. 65 3. 5

4

4. 5

5

kPm

Figure 3: Firm L’s quantity decisions t

tP2

2. 3 2

tL2

1. 5 1 0. 5

0. 25

0. 5

0. 75

1

1. 25

1. 5

1. 75

kLm

Figure 4: Firms’ timing decisions for xL1 ˆ 0

Asymmetric Strategic Investment Behavior in Network Industries

79

4. Discussion and Conclusions In this paper we introduce a dynamic deterministic optimization model for a duopolistic market set-up with strategic investment of two asymmetric firms. The ‘leader’ (LNG distributor) is characterized by a modular and largely reversible investment structure (divisible investment). It is forced to invest immediately in all markets, in order to gain some positive market share, and expected to maximize its payoff stream by optimally choosing the investment level in line with the expected time of market exit (i.e. the arrival time of the competitor). The ‘follower’ (pipeline distributor), in contrast, has an irreversible cost structure, supplying its product via a physically interconnected network and enjoying both a second-mover advantage and lower marginal costs in supplying each consumer with natural gas. The follower, in the duopolistic game setting, chooses its speed of sequential arrival in the finite number of markets considered in a profit-maximizing way that also takes into account the second-mover benefit. Hence the optimal time for the follower to invest (i.e. to enter the market) is, among other issues, also influenced by the activity level of the leader. Although our model is relatively simple, as we focus on open-loop equilibria and the deterministic case only, the main strategic effects shown can also be expected to arise under more general assumptions. Clearly, analyzing closed-loop equilibria, taking into account more complicated pipeline networks, and considering different forms of competition on final gas markets might lead to more realistic models. However, we do not expect the basic structure of the induced leader / follower game to change. The possibility to supply LNG without huge infrastructure investments, and more intense competition due to market entry, are the two main determinants for the strategic interaction between LNG suppliers and pipeline operators. Hence, we argue that our model is potentially useful for analyzing the dynamic development of the natural gas market. Regarding potential policy implications for the optimal design of measures aimed at fostering the market diffusion of natural gas, the model points out that subsidizing the spread of LNG may have two positive effects. First, it increases investments of LNG suppliers and may thus promote the use of natural gas. Second, it can help to overcome potential market failures due to positive external effects. Policy measures that increase the dissemination and consumption of LNG can also stimulate investments in pipeline networks and may thus lead to potential efficiency gains due to a better exploitation of economies of scale. Overall, the lessons that can be learned from the Norwegian experience with small-scale LNG distribution are of potential interest for a large number of other countries throughout the world. Likewise, the model developed in this paper might provide useful for the study of similar situations where asymmetric strategic investments in other network industries are concerned.

80

Reinhard Madlener and Pio Baake

5. Acknowledgements The authors gratefully acknowledge helpful comments received from Till Requate (discussant) and other participants of the 70th ARGE Conference, as well as fruitful discussions with Erik Jarlsby on the small-scale LNG business in Norway.

References Boyer, M. / Gravel, E. / Lasserre, P. (2004a): Real options and strategic competition: a survey. mimeo, CIRANO, Montreal, Canada. Boyer, M. / Lasserre, P. / Mariotti, T. / Moreaux, M. (2004b): Preemption and rent dissipation under price competition. International Journal of Industrial Organization, 22(3): 309– 328. Brealey, R. A. / Myers, S. C. / Allen, F. (2005): Principles of Corporate Finance. McGrawGill, New York, 8th edition. Décamps, J.-P. / Mariotti, T. (2004): Investment timing and learning externalities. Journal of Economic Theory, 118(1): 80 – 102. Fudenberg, D. / Tirole, J. (1985): Preemption and rent equalization in the adoption of new technology. The Review of Economic Studies, 52(3): 383 – 401. Grenadier, S. R. (1996): The strategic exercise of options: Development cascades and overbuilding in real estate markets. Journal of Finance, 51(5): 1653 – 1679. – (1999): Information revelation through option exercise. Review of Financial Studies, 12(1): 95 – 130. Huisman, K. (2000). Strategic technology adoption taking into account future technological improvements: a real options approach. CentER Discussion Paper No. 2000 – 52, Tilburg University, Tilburg, The Netherlands. – (2001): Technology Investment: A Game Theoretic Real Option Approach. Kluwer Academic Publishers, Norwell. Jarlsby, E. (2004): Lowering downstream entry barriers for natural gas: small scale LNG distribution in Norway. Proceedings of the IRAEE International Conference “Energy and Security in the Changing World”, Tehran, 25 – 27 May 2004. Jensen, J. T. (2003): The LNG revolution. The Energy Journal, 24(2): 1 – 45. Lambrecht, B. / Perraudin, W. (1997): Real options and preemption. JIMS Working Paper No. 15 / 97, University of Cambridge, Cambridge, U.K. Madlener, R. / Jarlsby, E. (2005): Technological competition in the downstream natural gas market: Economics of small-scale LNG distribution systems in Norway. Paper prepared for the 7th IAEE European Conference “European Energy Markets in Transition”, 28 – 30 August 2005, Bergen, Norway. – (2007): Economics and rationale of LNG distribution in Norway. IAEE Newsletter, International Association for Energy Economics, Baltimore. (forthcoming). Mason, R. / Weeds, H. (2005): Can greater uncertainty hasten investment? Unpublished mimeo, version as of 4 January 2005. (available at: www.soton.ac.uk / ~ram2 / ).

Asymmetric Strategic Investment Behavior in Network Industries

81

Moretto, M. (2003): Competition and irreversible investments under uncertainty. FEEM Working Paper No. 32.2003, Fondazione Eni Enrico Mattei, Milan, March. Perotti, E. C. / Rossetto, S. (2000): Internet portals as portfolios of entry options. Tinbergen Institute Discussion Paper No. TI 2000 – 105 / 2. – (2001): Strategic advantage and the optimal exercise of entry options. CEPR Discussion Paper No. 3061, Centre for Economic Policy Research, London, November. Reinganum, J. F. (1981): On the diffusion of new technology: A game theoretic approach. The Review of Economic Studies, 48(3): 395 – 405. Shakleton, M. / Tsekrekos, A. / Wojakowski, R. (2004): Strategic entry and market leadership in a two-player real option game. Journal of Banking and Finance, 28(1): 179 – 201. Smets, F. (1995): Exporting versus foreign direct investment: The effect of uncertainty, irreversibilities, and strategic interaction. Bank for International Settlements Working Paper, Basel, Switzerland. Weeds, H. F. (2002): Strategic delay in a real options model of R&D competition. Review of Economic Studies, 69(3): 729 – 747.

Asymmetric Strategic Investment Behavior in Network Industries: the Case of Natural Gas Distribution in Norway Comment By Till Requate

The authors model strategic investment in markets with asymmetric physical networks. The main application they have in mind is the Norwegian gas market where LNG competes with natural gas being supplied through a pipeline system. After giving a short description of pipeline versus LNG supply systems in general, and the Norwegian gas market in particular, the authors set up a simple theoretical model in which LNG suppliers can build up infrastructure simultaneously in m different market. By contrast the natural gas suppliers via a pipeline system can only conquer the markets subsequently one after the other since market m ‡ 1 can only be connected to the pipeline network if market m is connected to it. The authors then analyse the open-loop Nash equilibrium of both, the timing of investment and the number of consumers to be served by each of the two firms. They set up the equilibrium conditions and draw some conclusions, basically through carrying out some numerical simulations. The main finding of the simulations is that with higher cost of setting up LNG infrastructure, the LNG firm refrains to serve those markets that can be reached rather quickly by the natural gas supplier via the pipeline system. Although higher investment cost for the LNG supplier also shifts out the point of time when that firm enters market m, this has little effect on the natural gas supplier’s volume of investment into the markets which can be reached first. The authors draw the policy conclusion that LNG gas should be subsidized in order to spread out LNG gas and to “overcome potential market failures due to positive external effects”. The paper offers an interesting attempt to model a gas market which exhibits an asymmetric structure due to special geological features that in particular can be found in Norway, where remote areas can more easily served by LNG gas rather than by a system of gas pipelines. Unfortunately the authors do not go very far in their analysis. The headlines of section 3 and subsection 3.2 “Dynamic strategic optimization model” and “Optimal investment and network expansion” are a bit misleading since the authors neither investigate the socially optimal investment structure nor the socially optimal timing of investment. A comparison of the first best to equilibrium investment would be necessary to study the scope and direction

84

Till Requate

of market failure. Therefore I also do not buy the policy conclusions: By assumption natural gas supply is more efficient than LNG, once the infrastructure has been set up. It is well known that in a case like this it is not efficient to subsidize the less efficient technology. To verify the authors’ statement it would be necessary to set up the optimality conditions for a first-best allocation and to compare a subsidized or otherwise regulated market to the first-best. It is also not so clear where we find the network effects in the LNG market since those can be served independently and no economics of scale exist. Also consumers do not enjoy network effects in the sense that their utility rises with more consumers using LNG. In the introduction of their paper the authors claim that two different kinds of equilibria may exist: one that implies tacit collusion and another one where the follower exploits its dominant position. However, in the theoretical analysis this does not show. I am also not convinced that a tacit collusion open-loop equilibrium exists since those kind of equilibria are usually sustained by some threat of punishment. Since the authors consider open-loop equilibria only, I doubt that tacit collusion constitutes an equilibrium under this solution concept. To sum up, the model set-up might be appropriate to model competition between LNG suppliers and natural gas suppliers via pipeline networks. The set-up is less appropriate to model that kind of competition for countries like Germany where a gas pipeline network is already established. Even if establishing a pipeline system takes time to build, I doubt that LNG gas should be subsidized. The contrary might be optimal: in order to exploit lower operation costs and increasing returns it might be optimal to subsidize the construction of gas pipelines. We also saw in the simulations that a variation of the LNG-supplier’s set-up cost hardly affects the natural gas-via-pipeline suppliers’ pipeline decision. But a definite policy conclusion should be subject to further careful investigation.

How Dominant is Russia on the European Natural Gas Market? Results from Modeling Exercises By Franziska Holz* Abstract The importance of Russia for the European natural gas supply tends to be over-estimated. This paper uses simulation results with the GASMOD model to analyze the Russian role in a two-fold perspective. First, it is shown that in a globalizing natural gas market alternative suppliers can ensure the European natural gas supply, namely via LNG. However, some countries, especially in Eastern Europe, will need pipeline infrastructure expansions to benefit from alternative supplies. Second, it is shown that Russia has no incentive to withhold exports to liberalizing European markets as it would suffer from profit reduction. Conversely, the European countries should concentrate their efforts an liberalizing the natural gas market. JEL-Codes: L95, F14, C69 Keywords: natural gas markets, Europe, Russia, simulation analysis

1. Introduction Natural gas imports from Russia have recently come into the focus of politics and the general public, after several incidents questioning the reliability of Russian supplies, and with the Nordstream pipeline project that is receiving prominent support in Germany. Natural gas imports from Russia to Central Europe and Western Germany today must be viewed in a two-fold perspective. On the one hand, the European natural gas market is integrating more and more into the globalizing natural gas market, where liquefied natural gas (LNG) shipped in vessels connects regional markets around the world. The traditional pipeline market in Europe is thus complemented by LNG imports from regions outside Europe which changes * DIW Berlin, Germany. Email address: [email protected]. The author thanks Prof. Christian von Hirschhausen for his suggestions and support, and Prof. Franz Hubert for his insightful comments. This paper also benefited from comments by the participants of the Stanford University Energy Modeling Forum on Natural Gas Models, the 9. Symposium Energieinnovation Graz (Austria) 2006, the Annual Meeting of the Verein für Socialpolitik 2006 Bayreuth (Germany), and the annual ARGE meeting 2007 Berlin (Germany). All remaining errors are those of the author.

86

Franziska Holz

the European supply situation fundamentally. On the other hand, the European Commission is increasing its efforts in liberalizing the European natural gas markets in order to create a competitive market structure and a truly functioning internal market. This new market situation in Europe would also create a new framework for the players on the export market because it will fundamentally change the supply-demand equilibrium. In this paper, the Russian exports to Europe will be analyzed within these two perspectives. After having presented the past and current export figures from Russia (Section 2), I will argue that the European dependency on Russia is lower than perceived, and that the Russian market importance will remain limited in the next decades. This is due to the possibility for Europe to diversify its imports with pipeline and LNG imports from other suppliers. It is also due to the market structure of the international natural gas markets where the few suppliers that have natural gas reserves are competing in an imperfect market. The imperfect market structure gives the suppliers the possibility to sell less gas at higher prices, which allows market entrance of more and of higher cost suppliers (Section 4). Moreover, the perspectives of an oligopolistic exporter to a liberalized European market will be discussed. While the oligopoly in the export market ensures high prices for the exporter, perfect competition in the downstream markets leads to high demand. The latter effect (a quantity effect) gives an incentive to the exporter to not too strongly withhold its exports, which in turn ensures security of European supplies (Section 5). I will use results obtained with the numerical simulation model GASMOD, presented in Section 3, to support our argumentation. In the conclusions, some recommendations for the European energy policy will be given, highlighting the necessity to focus on market liberalization in Europe.

2. Where Do We Stand – Quantifying Russian Exports to Europe Russia, and its predecessor, the Soviet Union, has been an important supplier for natural gas to Western and especially to Eastern Europe since the beginnings of international natural gas trade in Europe in the 1960s. Today, Russia supplies the major part of natural gas imports of most Eastern European countries. Several West European countries like Germany, Austria and France have had long-standing import relations with Russia. Table 1 gives an overview of natural gas imports to some selected European countries over the last decade. While it shows the strong dependency of East European countries on Russian supplies, it also reveals a trend towards diversification of imports. In recent years, the East European countries could reduce the Russian market share by importing from alternative suppliers such as Turkmenistan (Hungary, Poland) or Norway (Czech Republic). The Russian natural gas reserves are significant and are equaled only by the region around the Persian Gulf. According to recent estimates (BP 2006), Russia

How Dominant is Russia on the European Natural Gas Market?

87

has reserves of 47,82 trillion cubic meters, which can guarantee production for the next 80 years with the current technologies and at current prices.1 The large reserves combined with its geographic proximity to Europe “naturally” make Russia a privileged supplier of natural gas. Currently, Russia exports about 25 % of its production (BP 2006), all of it westwards by pipeline to Europe. Russian natural gas production is concentrated in Western Siberia, although there are large reserves in East Siberia, too. Only recently, Russia has started to develop pipeline or LNG export projects in the East, to China (via pipeline) or the Pacific LNG market (Sakhalin projects). Table 1 Imports from Russia in Billion Cubic Meters (bcm) per Year and in Percentage of Total Imports of some Selected European Countries 1996 Germany Austria France Poland Hungary Czech and Slovak Republic

32.4 5.7 11.6 7 8.9 15.6

41 % 87 % 33 % 90 % 100 % 100 %

2000 34.7 5 12 6.6 7.9 14.3

2005 46 % 80 % 29 % 81 % 87 % 88 %

38 6.8* 9.2 6.9 8.8 13.8

42 % 78 %* 20 % 66 % 73 % 86 %

Source: IEA 2000, IEA 2004, and IEA 2006a, sowie *BP 2006.

3. A Market Power Model of the European Natural Gas Market 3.1 GASMOD Model Description

I use a large-scale numerical simulation model of the European natural gas market to assess the role of Russian exports. This model, called GASMOD, is described in detail in Holz et al. (forthcoming). Contrary to many simulation models of the natural gas sector that implicitly assume perfect competition, this model takes into account the concentrated market structure of the European export and wholesale markets. In fact, the limited number of natural gas producers as well as the traditional market organization with State-owned monopolies in many European countries are potential sources for uncompetitive behavior. In Holz et al. (forthcoming) we show that the current status quo of the European natural gas mar1 We must distinguish between reserves and resources. Reserves are defined as the resources that can economically be exploited with current technologies and at current prices. The rise of prices over the last years has therefore led to an improvement of the reserve situation. Resources in general comprise all proven natural gas deposits, but the costs to recover these resources may exceed current prices (e.g., remote fields, unconventional resources like methane in ocean water).

88

Franziska Holz

ket can be characterized as a Cournot market on the export and on the European wholesale stage, with the notable exception of the liberalized British market. GASMOD is structured as a two-stage game of successive exports to Europe (upstream market) and subsequent wholesale trade within Europe (downstream market). We concentrate on the elements “Production / Exports” and “Wholesale Trade” of the value chain of the international natural gas industry (the columns highlighted in grey in Figure 1) because we want to analyze the international trade of natural gas. However, the natural gas sector is a network industry and as such it is dependent on the availability of transport infrastructure to deliver natural gas from one place to another. This transport infrastructure can be natural gas pipelines or LNG facilities (LNG tankers, liquefaction and regasification terminals). We hence include the transport infrastructure into the model as an element that limits the trade between trading partners. The availability of transport possibilities into a national market can be essential for the market structure in a country. If this country is not connected to the international market, then no other company besides the local incumbent can access that market and exert competitive pressure.

Figure 1: The Value Chain of International Natural Gas Sector

In GASMOD, market power can be exerted at both stages, i.e., the producers and the wholesale traders can influence the market price in such a way that it is above the marginal cost level. We assume the players in each market to behave à la Cournot, i.e., to compete with each other in a non-cooperative game by setting quantities. We assume that the market stages do not influence each other, i.e., there is no bargaining or coordination between exports and wholesales. However, the upstream players (the exporters) know the final demand that the wholesale traders are facing in each European country and the exporters know the behavior of the wholesale traders. From a mathematical point of view, GASMOD is an “equilibrium model” (see Gabriel / Smeers 2006). The model results reflect the equilibrium solution of the Cournot game(s) and do not take into account the adjustment process to that equili-

How Dominant is Russia on the European Natural Gas Market?

89

brium state. GASMOD is programmed in the mixed complementarity format, using the software GAMS and the solver PATH.

3.2 Data

We aim at giving an exhaustive representation of the European market, and therefore include all continental European countries in the data set (see Table 2). We incorporate Turkey because it has traditionally been an importer of natural gas and it is becoming a transit country for (potential) gas exports from Russia (via the Black Sea), Iran, and the Caspian Region to Europe. As exporters, we include all regions that are currently able to export to Europe, via pipeline or LNG, or that are expected to do so in the next 20 years. Our list of exporters therefore takes into account the large natural gas producers on the Eurasian continent2, and LNG producers adjacent to the Atlantic basin. Table 2 Regions in the GASMOD Model Exporting Regions Russia Europe: Norway Netherlands United Kingdom North Africa: Algeria Libya Egypt Middle East: Iraq Iran Middle East (Qatar, UAE, Oman, Yemen) Overseas LNG: Nigeria Trinidad Venezuela

Importing Regions United Kingdom Netherlands Spain / Portugal France Italy / Switzerland Belgium / Luxemburg Germany Denmark Sweden / Finland Austria Poland Czech Rep. / Slovak Rep. / Hungary Former Yugoslavia / Albania Romania / Bulgaria Baltic States (Estonia, Latvia, Lithuania) Greece Turkey

Source: Holz et al. forthcoming.

2 Exports from the non-Russian Caspian regions (Turkmenistan, Uzbekistan, Azerbaijan, and Kazakhstan) are treated as exports of Russia because they have to pass via Russia.

90

Franziska Holz

Due to the concentrated nature of the national natural gas markets and the limited data availability, we assume that each country is one player in the international and the European natural gas market. Examples of previous State monopolies and national champions like Gaz de France (GDF) in France and ENI in Italy on the import (wholesale) side, and of Gazprom in Russia and Sonatrach in Algeria justify this choice. The cost data are from (OME 2001) for the base year (2003) simulations, and from (OME 2005) for the simulations up to 2025. These are long-run marginal cost data, i.e., data that include average capacity investment and maintenance costs in addition to operational costs. They combine production costs and transport costs to the European borders (“border prices”). Remarkably, the cost data between 2001 and 2005 mainly differ for Russian and LNG supplies (Table 3). Table 3 Long Run Marginal Production and Transport Costs to Europe in US-Dollar per Million British Thermal Units of Russia and Alternative LNG Suppliers Country and region (field) of origin Russia Nadym Pur Taz (Western Siberia)

Transport via / to via Ukraine (Druzhba pipeline system) via Belarus (Yamal-Pipeline) via Nordstream pipeline to Germany (to be built)

Costs in USD per MMbtu (OME 2001)

Costs in USD per MMbtu (OME 2005)

2.49 2.01

2.49 2.01

2.53

2.53

Russia Volga-Ural-Basin

via Ukraine via Blue Stream and Turkey

1.62 1.58*

1.53 2.10**

Russia Yamal field (to be developed)

via Belarus via Nordstream pipeline to Germany (to be built)

2.59

2.59

3.10

3.11

Algeria LNG

to European Mediterranean

2.42 – 2.50

2.12

Egypt LNG

to European Mediterranean

2.40 – 2.60

2.15

Libya LNG

to European Mediterranean

2.50 – 2.60

2.27

Trinidad and Tobago LNG

to Western Europe (Atlantic coast)

3.00

2.45

Nigeria LNG

to Western Europe (Atlantic coast)

3.00

2.45

to European Mediterranean

2.80

2.48

Quatar LNG

* Costs for delivery to Turkey. ** Costs for delivery via Turkey to Bulgaria. Source: OME 2001 and OME 2005.

How Dominant is Russia on the European Natural Gas Market?

91

While LNG supplies have become cheaper due to technological advances, the Russian supplies to Europe are becoming relatively more expensive. Despite the abundance of natural gas in Russia, it is costly to produce that gas because the reserves are located in remote areas with a difficult climate, and it is costly to transport to Europe because of the very long distances.

4. Which Market Share for Russia Today and in the Future? The GASMOD model is used to assess the role of the Russian natural gas exports for the European market given economic drivers. In particular, we are assuming a profit-maximizing behavior by the market players, which leads them to seek minimization of their costs and price-setting such that their sold quantity optimizes their profit. The inclusion of these economic variables gives a long-run equilibrium result which can be understood as a steady state of the market. Long-term contracts, which are one characteristic of the European market, are not included as a constraint into the model because they should be reflected in the long-term equilibrium relations in the model results. I will first compare the hypothetical outcomes of different market structures for the base year 2003, and then run simulation runs until 2025. 4.1 Alternative Market Scenarios

In (Holz et al. forthcoming) we showed that the current European natural gas market is well characterized by a Cournot market in the export (upstream) as well as in the wholesale (downstream) stage. I will therefore continue using the market scenario of a Cournot-Cournot market (“Cournot Competition”) as a benchmark. The results for European imports from its major natural gas exporters are shown in Table 4, where the Cournot competition results are compared to three alternative market scenarios: “Perfect Competition” with perfectly competitive upstream and downstream markets; “EU Liberalization” with a Cournot market upstream and a perfectly competitive downstream market; and “Cournot-Competitive Fringe” where the smaller players on the export market are assumed to be price-takers (competitive fringe) of the Cournot market formed by the large exporters and the downstream market is a Cournot market. Two striking results can be found in the benchmark scenario of Cournot competition: a diversity of imports from all supply sources, and smaller Russian exports than in the reference year. Comparing the results of all scenarios, total exports are the smallest in the scenarios with imperfect competition, and are higher the higher the degree of competition is, as a suggested by economic theory. Total export results are close to the reference values in the Cournot Competition scenario and especially in the Cournot-Competitive Fringe scenario, which is evidence that these scenarios are representing the real-world market.

92

Franziska Holz

The major European natural gas producers (Norway, Netherlands, UK) are the biggest suppliers in all scenarios and deliver about 50 % or more of European imports. Russia is the second largest supplier with a market share between 20 and 30 % of European imports. While the model results of North African exports generally are about the same as in the reference year, the supplies from the Middle East and from overseas could be higher. The values of Russian exports that can be explained by the economic model are below today’s levels with 59 and 86 bcm per year, respectively, in the Cournot and EU Liberalization scenario, compared to 132 bcm reported in 2003. Russian exports are expensive because of their high production costs and transport costs to Europe, and this is a major driver of the economic model. Export levels in the 2003 reference year are to a large extent determined by long-term contracts concluded in the past decades when the economics of the European natural gas market were different. Especially the cost reductions of LNG exports over the last years have led to a change of the European import options. Table 4 Static Model Results for Exports to Europe in bcm per Year (Market Share of Exporter in Percentage) for the Base Year 2003 Exporter Russia Europe North Africa Middle East Overseas LNG Total Exports

CournotCournot Perfect EU LiberaliReference Competitive Competition Competition zation Exports 2003 Fringe 59 (18 %) 263 (68 %) 24 (6 %) 13 (3 %) 25 (6 %) 384

196 (30 %) 296 (45 %) 92 (14 %) 37 (6 %) 41 (6 %) 662

134 (22 %) 296 (49 %) 92 (15 %) 37 (6 %) 41 (7 %) 601

86 (18 %) 281 (59 %) 53 (11 %) 14 (3 %) 41 (9 %) 476

132 (30 %) 234 (53 %) 59 (13 %) 6 (1 %) 10 (2 %) 441

Source: Holz et al. forthcoming for the first three market scenarios, BP 2004 for the reference exports.

Other suppliers than Russia export more natural gas than in the reference year, and there is a rich diversity of imports from all possible exporters to Europe. This diversity is a result of the model assumption of a Cournot market which results in high prices and small quantities. The Cournot players, including Russia, are exerting market power and influencing the price on the European import market by withholding quantities to their advantage such that their profits are maximal. But while the possibility of the exporters to exert market power may be considered as negative

How Dominant is Russia on the European Natural Gas Market?

93

for the European market because it results in high prices for the European consumers, it also has a positive effect. Namely, the high prices attract more producers and a diversity of producers, including high cost producers. Parts of the alternative supplies come as LNG from overseas, the Middle East, and North Africa. This diversity of supply sources ensures security of the European natural gas supplies. In a more restricted oligopoly (scenario Cournot-Competitive Fringe), the Cournot players, and among them Russia, benefit from the smaller number of strategic players with increased exports and higher prices3. However, the existence of pricetaking exporters also benefits the European importers as the competitive fringe exporters can increase their exports until their marginal costs meet the price. The closer proximity of the results of the Cournot-Competitive Fringe scenario to the reference values can be considered as evidence that this scenario reflects the current market status quo better4, and I will hence use it for the simulation runs until 2025.

4.2 Russian Exports to Europe until 2025

Besides assessing the current role of Russia, I also want to investigate the perspectives for Russian exports to Europe until 2025. I use the GASMOD model with updated data input in five-year increments.5 I assume an upstream Cournot market with a competitive fringe, and downstream Cournot markets in Europe with the exception of a liberalized UK market and perfect competition in small markets. For the demand for natural gas, a continuous demand increase of 1.64 % p.a. for Western Europe and of 2.2 % p.a. for Eastern Europe is assumed, based on the demand forecast by (EIA 2004). Infrastructure capacities are updated based on current knowledge of extension and new-built plans, adopting a rather conservative approach as to which projects will be realized and when. Due to uncertainty of the infrastructure projects realized beyond 2015, most of the included infrastructure extensions are for the decade until 2015. Total capacities of exports to Europe are assumed to grow from 619 bcm per year in 2003 to 832 bcm in 2015 and 959 bcm in 2025 with major increases in liquefaction capacity in the Middle East and overseas (primarily Western Africa), and in pipeline capacity from Algeria to Europe. On the other hand, only few infrastructure expansion projects for intra-European transport are currently discussed and could therefore be included in the model. The outlook to the medium-term future until 2025 in Table 5 confirms the results of the current market simulations: Europe will benefit from a diversified import Prices not reported here. The specification with a competitive fringe is also found to be the best representation of the European market in other models, e.g., in (Egging and Gabriel 2006). 5 These model runs were i.a. prepared for the Stanford University Energy Modeling Forum Nº 23 (http: // www.stanford.edu / group / EMF / projects / projectemf23.htm), and they were presented in Holz et al. 2007. 3 4

94

Franziska Holz

portfolio, and Russia will play an important but not a dominant role. The growing demand for natural gas will be satisfied by increasing imports, given that domestic production in Europe will be declining in many countries. LNG imports from overseas (West Africa, Caribbean) show the largest increase, followed by North African LNG supplies. The Middle East cannot expand its LNG exports but increases its pipeline exports (in 2025 from Iran and Iraq to Turkey). Natural gas production in Europe, especially in the UK, is declining and hence is the import share of European suppliers (from 59 % in 2003 to 37 % in 2025). LNG from Norway can find a market in Europe in the next decades, as can LNG from Russia (Shtokman) that is included in the model starting in 2020. Table 5 Model Results 2003 – 2025 of Exports to Europe in bcm per Year and Market Share of each Exporter in Percentage (in Parentheses)

Russia -pipe -LNG Europe -pipe -LNG (Norway) North Africa -pipe -LNG Middle East -pipe -LNG Overseas Total exports

2003

2010

2015

2020

2025

104 (24 %) 0 (0 %)

116 (23 %) 0 (0 %)

140 (26 %) 0 (0 %)

151 (27 %) 6 (1 %)

172 (29 %) 10 (2 %)

259 (59 %) 0 (0 %)

252 (50 %) 6 (1 %)

229 (42 %) 8 (1 %)

215 (38 %) 7 (1 %)

218 (36 %) 3 (1 %)

22 (5 %) 21 (5 %)

22 (4 %) 35 (7 %)

22 (4 %) 38 (7 %)

30 (5 %) 43 (8 %)

33 (5 %) 48 (8 %)

10 (2 %) 9 (2 %)

10 (2 %) 8 (2 %)

14 (3 %) 20 (4 %)

30 (5 %) 17 (3 %)

40 (7 %) 11 (2 %)

11 (3 %)

57 (11 %)

74 (14 %)

62 (11 %)

63 (11 %)

437

507

546

560

600

Source: Holz et al. 2007.

The Russian market share will slightly increase to satisfy the growing demand for natural gas in Europe. Eastern Europe (and especially landlocked countries like Hungary, the Slovak and the Czech Republic, Romania and Bulgaria) can not benefit from an expansion of the international LNG market, but is dependent on Russian gas by pipeline. No alternative pipeline project from the Caspian and Middle

How Dominant is Russia on the European Natural Gas Market?

95

East (Iran) region into Central Europe can currently be assumed to materialize by 2025. In particular, I do not include the Nabucco project in the model as its realization seems to be uncertain. The Eastern European countries can also not benefit from the increase of pipeline export capacity from other exporters, such as Algeria, because of limited intra-European transport infrastructure. Algerian pipeline exports, for example, are locked in the border countries Spain and Italy because of very restricted pipeline capacity out of Spain (to France) and Italy (to Austria). Russia hence continues to play an important role for Eastern Europe. Western Europe, on the other hand, will be able to satisfy its growing demand for natural gas with increased pipeline imports from North Africa and Norway and with LNG from various sources. While Mediterranean countries such as Spain, Italy, Greece, Turkey, and France have been long-standing LNG importers, some Northern European countries recently started importing LNG (Belgium already in 1987, the UK in 2005). Other countries have more or less well-advanced projects of first LNG imports (Netherlands, Germany, and Slovenia) that were included with different start-up times in the model. All LNG importers are planning expansions of their regasification capacities in order to keep up with the expansions on the global liquefaction market. Figure 2 shows the increasing imports of LNG in Europe until 2025 when they will reach a market share of approx. 25 %. Most of the increase of LNG imports will already happen until 2015 when the major regasification capacity expansions will take place.

Exports to Europe in bcm per year

700

600

500

400

300

Russia

200

Others LNG

100

Others Pipe 0 2003

2010

2015

2020

2025

Source: Holz et al. 2007.

Figure 2: European Imports of Russian Gas, and Pipeline Gas and LNG from other Sources 2003 – 2025, in bcm per Year

Among the LNG exporters, the suppliers from overseas will experience the largest increase in volume and market share. This is due to the relatively advanta-

96

Franziska Holz

geous cost structure of these exporters from the Caribbean and West Africa (see Table 3). North African and European LNG exports will also contribute more than proportionally to the satisfaction of rising demand and can increase their market share. The LNG exports from the Middle East, however, will stagnate, due to their relatively high border prices. Given the limit on regasification capacity in Europe, the European demand for LNG is primarily satisfied with lower-cost supplies from all other LNG exporters, and the Middle East only contributes marginally.

5. Russian Exports to Liberalized European Markets The objective of the reform process initiated by the European Commission in the 1990s is the liberalization of the European natural gas markets that should be organized competitively without allowing the European wholesale companies to exert market power. This situation is reflected in our simulation scenario “EU Liberalization” with Cournot upstream markets (the export market cannot be influenced by European regulation) and competitive European downstream markets. In (Holz et al. forthcoming), we showed that the European economies would benefit from the EU liberalization with increased welfare. Given this perspective of a possible market development in Europe, it is necessary to analyze the Russian role in a liberalized EU market, and specifically the incentives for Russia to withhold exports to the European market. In Figure 3, the European imports for the Cournot and EU liberalization scenario are depicted.6 Due to the lower prices in the liberalized EU market, total demand and exports are considerably higher than in the Cournot Scenario (35 %). While European production is bound by capacity restrictions, all other exporters can benefit from the liberalization with higher exports to Europe. Especially North Africa and Russia can significantly expand their exports and increase their market shares (27 % compared to 19 % for Russia). After the January 2006 disruption of Russian natural gas supplies to Ukraine and later to Belarus and Georgia, fears in Europe resurged that Russia might use the possibility of withholding its natural gas exports as an instrument of foreign policy, also with regard to Europe. It is therefore interesting to study the effects of withholding by Russia of its natural gas supplies to Europe. I propose two cases of export limitation: a maximum capacity of 150 bcm per year (approx. the current export level), and a reduction of its export capacity to 75 bcm per year. These cases can also be interpreted with respect to the maintenance of the Russian export infra6 For these simulations, we adapted the model calibrations to take account of an increased demand for natural gas and of a larger elasticity of demand in a liberalized market, i.e., more flexibility of the consumers’ demand for natural gas. These calibrations explain the differences of the results compared to Table 4. They are necessary to ensure the equilibrium model results in the restricted capacity scenarios. The little available transport capacity, especially into Eastern Europe, would otherwise not allow the model to find equilibrium results.

How Dominant is Russia on the European Natural Gas Market?

97

structure to Europe. Indeed, much of the Russian export pipeline system is outdated today and would need considerable investment in maintenance or even replacement (IEA 2003). At a different level in the value chain, the Russian production is in need of investment in the development of new natural gas fields as the current fields already passed their peak production capacity (IEA 2006). However, the Russian investment efforts so far have been limited.

Exports to Europe in bcm per Year

700 600 500 400 300 200 100 0 Cournot Competition (470 bcm)

EU Liberalization (635 bcm)

Figure 3: Exports to Europe in bcm per Year from all Exporting Regions in the Cournot Market and the EU Liberalization Scenario (Market share for each scenario, respectively, in parentheses)

Figure 4 shows that the restriction of Russian exports to 150 bcm per year does not have any impact in the Cournot scenario and only a small reduction effect in the EU Liberalization scenario (–3 %). There are two reasons for the limited impact of this restriction: first, the Russian exports in the unrestricted case were below or close to 150 bcm per year (90 bcm in the Cournot scenario and 169 bcm in the EU Liberalization scenario); second, the European countries have the possibility to substitute failing Russian supplies by natural gas imports from other sources. In the EU Liberalization scenario with the 150 bcm capacity limit, the LNG suppliers from overseas and the European exporters are primarily increasing their market share. The restriction of Russian exports to 75 bcm per year does have a stronger impact on European imports with a reduction of imports by 3 % (Cournot scenario) and 15 % (EU Liberalization scenario). This limitation cannot be entirely compensated by imports from other sources, mainly due to transport capacity bottlenecks in and into Eastern Europe. Increased natural gas imports again mainly come from European exporters and overseas LNG, but the Middle East and North Africa also increase their market share in the EU Liberalization scenario.

Exports to Europe in bcm per Year

98

Franziska Holz 700 600 500 400 300 200 100 0 Cournot Competition (470 bcm)

EU Liberalization (616 bcm)

Capacity Limit Russia 150 bcm

Cournot Competition (460 bcm)

EU Liberalization (541 bcm)

Capacity Limit Russia 75 bcm

Figure 4: Exports to Europe in Case of Capacity Limitation of Russian Exports in bcm perYear and with Market Shares of each Exporter for each Scenario, Respectively, in Percentage (in Parentheses)

Having assessed these effects of a Russian export constraint, let us examine whether Russia has an incentive to implement such a constraint. Figure 5 reports the profits for the unrestricted case and the two cases of export capacity restriction. Russia clearly benefits from the liberalization of the EU markets in all cases. This is due to the fact that the imperfect upstream market can ensure relatively high prices for exports (price effect for the Russian profits), while at the same time benefiting from an increased downstream demand where prices are relatively low in the competitive markets (quantity effect). Figure 5 also shows that Russia has no incentive to reduce its export capacity to Europe. The limit of Russian exports to 75 bcm per year to a liberalized EU market results in a reduction of profits by 37 % compared to the unrestricted case. The reduction of profits, and hence the incentive for Russia not to restrict its exports, is smaller in the Cournot scenario with only about 6 %. However, as discussed in Section 4, Cournot markets do ensure a diversity of supply to Europe and hence a “second best” security of supply in case of reduced Russian exports. Again, the cases of capacity restriction can also be interpreted with respect to transport and production infrastructure maintenance. In this regard, the liberalization of EU markets will provide an additional incentive to Russia to maintain its export capacity as the opportunity cost of not doing so will increase compared to imperfect European markets. Hence, the European countries themselves have an incentive to increase their efforts for domestic market liberalization if they want to ensure their natural gas supplies.

How Dominant is Russia on the European Natural Gas Market?

99

Russian profits in million US-$

16000 14000 12000 10000 8000 6000 4000 2000 0 Cournot Com petition

EU Liberalization

No capacity lim itation

Cournot Competition

EU Liberalization

Capacity 150 bcm

Cournot Competition

EU Liberalization

Capacity 75 bc m

Figure 5: Russian Profits for Different Market Scenarios and Capacity Limitation Cases (in million US-Dollar)

6. Conclusions Russia is an important supplier of natural gas to Europe, but its role is less dominant than generally assumed. In this paper, a market power simulation model (GASMOD) is used to assess the role of Russian natural gas exports for the European market. In different market and simulation scenarios, it is shown that the importance of Russian supplies is smaller than sometimes perceived, especially for the Western European market, and that Russia is far from being a dominant player on the European market. Let us detail our conclusions with respect to the following aspects: Russian production development, Russian export and transit infrastructure, European transport infrastructure, and European market structure. First, with respect to the Russian production capacity I showed that Russia has an incentive to maintain its current level of export capacity, including production capacity for exports, to benefit from a liberalized European market with high demand. Given that all currently exploited fields in Western Siberia have reached their peak production level, Russia must start developing new fields (IEA 2006). This is even more necessary in the absence of an efficient price signal to its domestic markets, which leads to an increasing competition for Russian gas between the domestic market and the export market (Sagen / Tsygankova 2006). Second, the incentive to Russia to maintain its export capacity also encompasses its export and transit pipeline system which today is outdated and in dear need of overhauling (IEA 2003). Third, given the existing export pipeline capacities of more than 180 bcm per year, there is no need to build an additional pipeline such as the Nordstream pipeline, even in the future until 2025. The simulation results give Russian export levels below its current export capacities of more than 180 bcm. The current transit

100

Franziska Holz

system consists of the pipelines through Ukraine, generally referred to as the “Druzhba pipeline system” (approx. 150 bcm of yearly capacity), and the pipelines through Belarus in the Yamal-Europe pipeline system (approx. 30 bcm of yearly capacity), plus some minor pipelines to the Baltic countries and Finland. Russian exports in the current market power environment are not bound by the export capacity.7 The high construction costs of the Nordstream pipeline that are specified to be at least 5 billion Euro (Nord Stream AG 2006) also shed doubts on the necessity of this pipeline. The applied game-theoretic literature gives some explanation to the Nordstream pipeline project. It considers Russia as a player in the transit game with Ukraine and Belarus where it has an incentive to threat with the construction of the Nordstream pipeline in order to improve its bargaining position (Hubert / Ikonnikova 2004, Hubert / Suleymanova 2006). Previous transit options such as the construction of the Yamal Europe pipeline through Belarus can be seen in the same perspective (Hirschhausen et al. 2005). Fourth, the supply security for Europe can be enhanced by increasing the transport capacities to and inside Europe. Increasing regasification and import pipeline capacities improve the possibility to compensate failing imports from one supplier. Increased intra-European pipeline capacity would also enable the transit of additional imports to non-border countries (e.g., to Eastern Europe). Moreover, an increased access capacity to a country can improve its market structure towards more competitive markets by increasing the number of players in that country. A substitute or a complement to intra-European pipeline expansion can be the development of liquid markets with the possibility to swap or hedge quantities between marketplaces instead of physical transport. This aspect of market development needs further research. Fifth, this paper showed that the European countries have a strong incentive to push for liberalization of their markets to ensure supply diversification and security. EU liberalization would come along with increased profits for the exporters and hence increase their opportunity costs of failing. Moreover, the diversity of supplies that is becoming reality with the development of the LNG market shows that worries of supply security are becoming less important. The improvement of European welfare and consumer surplus by liberalizing the European markets should instead be the priority of natural gas politics.

References BP (2006): Quantifying Energy – BP Statistical Review of World Energy, London. Egging, R. G. / Gabriel, S. A. (2006): “Examining Market Power in the European Natural Gas Market”, Energy Policy 34, 2762 – 2778. 7 The results obtained with GASMOD are confirmed by simulation results of the GASTALE model (Lise et al. 2005) and the European Gas Model (Egging et al. 2006).

How Dominant is Russia on the European Natural Gas Market?

101

Egging, R. / Gabriel, S. A. / Holz, F. / Zhuang, J. (2006): “A Complementarity Model for the European Natural Gas Market”, College Park, University of Maryland. EIA (2004): International Energy Outlook 2004, Washington, D.C.. Gabriel, S. A. / Smeers, Y. (2006): “Complementarity Problems in Restructured Natural Gas Markets”, in “Recent Advances in Optimization. Lecture Notes in Economics and Mathematical Systems” edited by A. Seeger, Volume 563, Berlin / Heidelberg, Springer, 343 – 373. Hirschhausen, C. v. / Meinhart, B. / Pavel, F. (2005): “Transporting Russian Gas to Western Europe – A Simulation Analysis”, Energy Journal 26, 49 – 68. Holz, F. / Hirschhausen, C. v. / Kemfert, C. (2007): “GASMOD Model – Simulation Results for EMF 23”, Presentation, Energy Modeling Forum Stanford University, http: // www. stanford.edu / group / EMF / projects / emf23 / JanDocs / DIW-GASMOD.pdf. – (forthcoming): “A Strategic Model of European Gas Supply (GASMOD)”, Energy Economics. Hubert, F. / Ikonnikova, S. (2004): “Hold-Up, Multilateral Bargaining, and Strategic Investment: The Eurasian Supply Chain for Natural Gas”, Berlin. Hubert, F. / Suleymanova, I. (2006): “Strategic Investment in International Gas-Transport Systems: A Dynamic Analysis of the Hold-up Problem”, Berlin. IEA (2003): World Energy Investment Outlook, Paris, OECD / IEA. – (2006): Optimising Russian Natural Gas – Reform and Climate Policy, Paris, IEA / OECD. Lise, W. / Hobbs, B. F. / Oostvoorn, F. v. (2005): “Security of Supply in the Liberalised European Gas Market, Simulation Results with the Dynamic GASTALE Model”, Bergen (Norway). Nord Stream AG (2006): Nord Stream: The New Gas Supply Route to Europe, Zug / Switzerland, Press Release, 22 November 2006. OME (2001): Assessment of Internal and External Gas Supply Options for the EU – Evaluation of the Supply Costs of New Natural Gas Supply Projects to the EU and an Investigation of Related Financial Requirements and Tools, Sophia-Antipolis, Observatoire Méditerranéen de l’Energie. – (2005): Assessment of Internal and External Gas Supply Options for the EU, Evaluation of the Supply Costs of the New Natural Gas Supply Projects to the EU and an Investigation of Related Financial Requirements and Tools, Sophia-Antipolis, Observatoire Méditerranéen de l’Energie. Sagen, E. L. / Tsygankova, M. (2006): “Russian Natural Gas Exports to Europe: Effects of Russian Gas Market Reforms and the Rising Market Power of Gazprom”, DIW Discussion Paper 597 / 2006.

How Dominant is Russia on the European Natural Gas Market? Comment By Ulf Moslener* 1. A Topical Issue What are we to make of the relationship between Russia and the European Union with around 40 % of the old continent’s natural gas imports coming from the former Soviet Union? While Europe’s imports are rising and its own reserves are dwindling, the question of Russian dominance on European natural gas markets and the consequences thereof is at the center of many debates in public as well as in the science community. Holz provides a timely and relevant contribution to this contentious topic. The paper analyzes the question whether Russia currently is or can be expected to be dominant on European natural gas markets in the near future. The analysis is based on a numerical two stage game theoretic model of European natural gas markets. In a first stage (production), upstream competition between the gas exporters to Europe and the domestic gas producers is assumed, while in a second stage downstream wholesale trade between European companies is captured. In both stages the model can be run either in a perfect competition or in a Cournot mode. The results of the analysis are – in essence – that Europe’s natural gas market can be described as an imperfect market. As a consequence of increased gas imports, European natural gas supplies will be diversified and lead to a significantly higher fraction of liquefied natural gas (LNG) in European imports. Nevertheless, Russia can be expected to be capable of exerting market power by strategically withholding quantities from the European market. The paper argues that, even in the medium term future, Russia has no economic incentive to do so. While Russia will stay a very important player on European gas markets, it will not – the author concludes – dominate the market in the sense that the European dependence on Russian supplies will remain limited.

* Centre for European Economic Research, ZEW, Mannheim.

104

Ulf Moslener

2. The Modeling Framework In addition to the thematic contribution the work also presents a conceptual one, namely, the Simulation Model of European Natural Gas Supply (GASMOD). Modeling natural gas markets has always been one of the bigger challenges in the field of energy economics. This is partly caused by characteristics of the gas markets leading to a situation far away from what economists would typically consider a perfect market. Examples are the grid-bound nature of the gas markets and a small but growing fraction of non-grid-bound gas in the form of liquefied natural gas (LNG), which can be transported with ships over large distances and requires a different but expensive infrastructure to liquefy and re-gasify the natural gas. Other reasons are the limited number of market participants on an international scale as consequence of an uneven distribution of the natural resource, and the amount of necessary infrastructure investment resulting in high fix costs and very low marginal costs of production. GASMOD is a static model and thereby abstracts from the consequences that investments in additional infrastructure capacity may have on the agents’ quantitydecisions. Its strength can be seen in explicitly modeling the upstream and downstream markets separately while at the same time keeping the flexibility to assume perfect or oligopolistic market structures on either stage. Also reflecting the pipeline network within Europe and gas transport capacity to Europe makes the modeling framework suitable for the analysis of a plethora of problems ranging high on the current policy agenda in the European Union. A look at the “Presidency Conclusions” of the European Council in March 2007 offers a variety of relevant areas where useful applications of the model seem possible. Among these are strengthening the EU internal market for energy as well as extending intra-EU transmission capacity or strategically extending pipeline capacity into Europe. Already in this analysis the author strongly argues in favor of liberalizing the EU gas market and furthermore suggests that subsidies to the Nord Stream pipeline do not seem necessary. 3. Useful Insights As to the thematic contributions of the work, the first result relates to the structure of the European market for natural gas. The gas import shares are calculated in the model for four different assumptions regarding the market structure: (i) Imperfect (Cournot) competition on both stages, (ii) Cournot competition on both stages where the smaller players on the export markets are assumed to be pricetakers (Cournot-fringe), (iii) perfect competition on both stages, and (iv) a scenario called EU-Liberalization, which assumes Cournot competition in the producer (upstream) market but perfect competition in the EU wholesale (downstream) market. It turns out that the results of an oligopolistic up- and downstream market best correspond to actual import shares. While it seems a bit strong to interpret this as

How Dominant is Russia on the European Natural Gas Market?

105

evidence for an oligopolistic market structure, it can at least be used as a plausibility check for the assumption of a two-sided oligopolistic market. It is worth to realize that Russia’s role in the four computed scenarios is the smallest in the case of Cournot competition. The paper argues, that the best response to Russian upstream market power would be to liberalize the European downstream market thus being essentially along the lines of the European Commission. This conclusion is based on the separate two-stage game implemented in the model. A relevant follow-up question could be whether this still holds for moving from two separate stages to some kind of an integrated bargaining problem between upstream and downstream competition. As noted in the paper, while Russia’s actual fraction of EU-25-gas-imports was around 40 % in 2003, the model shows about 20 % for the case of oligopolistic structure on both stages and an increased fraction in all other scenarios. One reason for the underestimation of the Russian market share in the model could be the price formation in the simulation. It is governed by the Cournot assumptions and transport costs are expressed per unit gas per kilometer. In typical contracts on the international gas market, the price formation is based on a substitute for gas in the receiving country (usually oil), and the price at the border of the exporting country is determined by this base-price less the transport costs to the receiving country. Since the distance between Russia and some important European gas consuming nations (e.g., Germany) is relatively large, transport costs in the model will turn out relatively high leading to less imports of Russian gas as compared to less distant producers. Therefore, the price formation in the simulation model does not lead to the export levels which are observed in reality. However, the behavior of the model provides information on the export structure if trade is governed only by the economic drivers as they are assumed in the model, and is therefore very useful for economic analysis. One example is that in today’s gas trade, scarcity rents are hardly visible. Producing natural gas involves extremely high up-front investment followed by low marginal costs of production later on. In order to avoid being forced to sell at marginal production costs, the producer often does not start building the infrastructure to exploit a new gas field until the larger part of the expected natural gas is sold by long-term contracts at prices indexed to oil. This shows the important role of the investment decision in determining the quantities of natural gas reaching the market, which is not reflected in the modeling framework. Nevertheless, as it is calibrated to real world data, GASMOD can provide some information on scarcity rents and welfare effects, as they can be expected for different assumptions regarding the market structure. An extra section of the paper focuses on Russian exports to liberalized European markets. After recalibrating the model for technical reasons – in order to account for substantially increased natural gas demand and a higher anticipated demand elasticity – the author is able to establish two additional important results.

106

Ulf Moslener

First, Russia gains substantially from EU liberalization (due to lower end-use prices and thereby higher European demand) and second, as a consequence Russia is likely to suffer from potential capacity restrictions (regardless of whether they are intended or not) since they hinder Russia to meet the demand. At this point the vastly increased Russian profits provoke further thoughts. One interesting question from a European perspective is which part of the additional Russian profit can be attributed to efficiency gains due to liberalization, and which part can be attributed to shifting rents from the down-stream (European) actors to the up-stream (mostly non-European) actors. In other words, to what extend may Russia reap the European welfare gains from gas market liberalization?

4. A Difficult Question While the author acknowledges Russia’s ability to exert market power, it remains a bit unclear what it means to not be a “dominant” player. As the title “How Dominant. . .” suggests, the answer might be rather complicated than a simple “yes” or “no”. The model simulations show that increased imports to Europe lead to some increases in Russian exports but most of all to diversification of supply in favor of LNG. But in order to come to a more comprehensive answer to the question, additional considerations might be helpful. One step may be to turn the attention to the mechanisms Russia may use to exert market power. Beyond simply producing less gas – be it by leaving capacities idle or under-investing beforehand – Russia may attempt to use its influence on the pipeline network or even to maximize its impact on the downstream market by strategic purchases in Europe. These aspects – although possibly relevant to the question of Russian dominance – tend to go beyond the scope of the paper. It is assumed that no vertical integration exists in European natural gas markets. This assumption seems justified for Europe in general, but when analyzing the case of Russia there are hints that the major producing company, Gazprom, shows a tendency towards vertical integration. It tries to buy into European downstream natural gas companies, already possesses 35 % of Wingas, a German distribution company, 10 % of the important “Interconnector” pipeline between Belgium and Britain, and expressed plans to buy more. At the same time the Kremlin makes it difficult for foreign investors to invest in Russia. Another relevant topic could be the potential of a strategic co-operation between Russia and Algeria. These two make up for more than half of the total exports to European gas markets, and the recently signed co-operation agreement between Gazprom and the Algerian state monopoly Sonatrach sounds worrying to some. By and large the question of Russian dominance is not yet fully answered, but the analysis by Holz is a valuable and useful contribution to disentangle the complex issue from an economic point of view.

Participants 1. Friedrich Aumann

FATM, University of Münster

2. Thomas Bänsch

Berliner Volksbank

3. Hubertus Bardt

Institut der deutschen Wirtschaft Köln

4. Johann Behrends

IHK Berlin

5. Ulrich Blum

IWH, Halle

6. Heiner Bruhn

BMWi

7. Rudolf Bürkl

GfK, Nuremberg

8. Matthias Creutzig

BMWi

9. Steffen Dagger

Office of Dr. Maria Flachsbarth, MdB

10. Michael Dauderstädt

Friedrich Ebert Foundation

11. Ingo Ellersdorfer

IER, University of Stuttgart

12. Christian Flachsland

PIK Potsdam

13. Birgit Frank-Bosch

Federal Statistical Office

14. Wolfgang Franz

ZEW

15. Jean Michel Glachant

Université de Paris 11

16. Enno Harks

German Institute for International and Security Affairs

17. Christian von Hirschhausen

TU Dresden / DIW Berlin

18. Tim Hoffmann

ZEW, Mannheim

19. Franziska Holz

DIW Berlin

20. Franz Hubert

HU Berlin

21. Hans-Dieter Karl

ifo München

22. Michael Kilpper

BMWi

23. Carl Dominik Klepper

Stiftung Marktwirtschaft

24. Gernot Klepper

Institute for World Economics, Kiel

25. Thomas Knaus

BMWi

26. Rolf Kroker

IW, Cologne

27. Johannes Lambertz

RWE Power AG

28. Harald Legler

NIW

29. Reinhard Madlener

DIW Berlin / ETH Zurich

108

Participants

30. Tim Mennel

ZEW Mannheim

31. Ralf Messer

ARGE

32. Ulf Moslener

ZEW Mannheim

33. Alexander Müller

BMF

34. Walter Otremba

BMWi und FATM, University of Münster

35. Michael Pahle

PIK Potsdam

36. Anna-Marleen Plume

BMWi

37. Walter Radermacher

BMWi

38. Joachim Reichert

BMWi

39. Heinrich Reitz

BASF

40. Till Requate

University of Kiel

41. Gerd Rosenkranz

Deutsche Umwelthilfe e.V.

42. Tim M. Rusche

European Commission Brussels

43. Christoph M. Schmidt

RWI Essen

44. Hans-Werner Sinn

ifo Munich

45. Dennis Snower

IFW Kiel

46. Hildegard Stahmer

ARGE (HWWI), Hamburg

47. Thure Traber

DIW Berlin

48. Regina Vogel

German-British Foundation

49. Matthias Vollert

Foreign Office

50. Werner Voß

FH Gelsenkirchen

51. Dirk Weber

Berliner Volksbank

52. Carl-Christian von Weizsäcker

Max Planck Institute

53. Anke Winkler

ARGE

54. Klaus F. Zimmermann

DIW Berlin