에너지경제연구 제11권 제2호 - PDF 무료 다운로드

Korean Energy Economic Review Volume 11, Number 2 KOREAN RESOURCE ECONOMICS ASSOCIATION KOREA ENERGY ECONOMICS INSTITUTE September 2012 Economic Evaluation of Offshore Wind Power Demonstration Project by the Real Option Method / Dongsoo Lee and Seokjun Yun and Sangmin Kim and Kiho Jeong Forecasting the Demand for Local District-Heating Systems / Byeongseon Seo and Sang Yul Shim An Empirical Analysis on the Comovement between Coal Price and Other Commodity Prices / Won-Cheol Yun An Analysis of the Effect of Carbon Tax on CO2 Reduction Using Interfuel Substitution Relations in Korean Manufacturing Sectors / Dongheon Yoo and Changsuh Park and Youn Jai Lee Capacity Constrained Supply Function Equilibria: Modeling and Simulations / Soohyun Yum Regulating rents from renewable energy policies: focusing on RPS in South Korea / Tae-hyeong Kwon Analysis on the Residential Demand for the Smart Meter and the Inhome-display / Daeyoung Koh and Hyunseung Cho and Moonsoo Park The Economy of Nuclear Power : Review of the Issue and Suggestion / Kyungmin Ko and Bum-Jin Chung KOREAN ENERGY ECONOMIC REVIEW Korean Energy Economic Review

KOREAN ENERGY ECONOMIC REVIEW Korean Energy Economic Review

차 례 에너지경제연구제 11 권제 2 호 이동수, 윤석준, 김상민, 정기호 서병선, 심상렬 윤원철 유동헌, 박창수, 이윤재 Soohyun Yum 권태형 고대영, 조현승, 박문수 고경민, 정범진

에너지경제연구 Korean Energy Economic Review Volume 11, Number 2, September 2012 : pp. 1~26 실물옵션을이용한해상풍력실증단지 사업의경제성평가 1

ln ln 7

[ 그림 1] 해상풍력단지건설로드맵 10

< 표 1> 회귀분석결과 14

< 표 2> 미래현금흐름추정결과 < 표 3> NPV 계산을위한위한변수값 15

< 표 4> NPV법을활용한경제성분석결과 16

< 표 5> SMP 단위근검정결과 18

< 표 6> SMP 확률과정 (GBM) 추정결과 19

< 표 7> 옵션가격도출을위한변수들의값 < 표 8> 실물옵션을이용한해상풍력실증단지사업의가치 20

접수일 (2012 년 7 월 10 일 ), 게재확정일 (2012 년 8 월 14 일 ) 23

. 2011... 2. 1. pp.37~43.. 2011.. KBR. 15 2.. 2010.... 2000. KOSPI200. 8. pp.1~26.. 2006. R&D... 2009.... 2005... 54. 4. pp.173~199. 2007... 9. 4. pp 763-779. 2011... 2003... 12 2. pp.217~224.. 2011. CO : CCS RPS.. 20. 1. pp.61~98.. 2009.. 5. 2010-490. 2003... 24

. 2000... 7 1.. 2001. Real Option IT. Black, F. and Scholes, M. J. 1973. The Pricing of Options and Corporate Liabilities. Journal of Political Economy 81: pp.637-654. Chan, K., Karolyi, G.A., Longstaff, F.A. and Sanders, A.B. 1992, An Empirical Comparison of Alternative Models of the Short-term Interest Rate, Journal of Finance 47: pp. 1209-1227. Dahlquist, M. 1996. On alternative Interest Rate Process. Journal of Banking and Finance 20: 1093-1119. Duffee, G. R. 1996. Idiosyncratic Variation of Treasury Bill Yields. Journal of Finance 51: pp.527-551 Eom, Y. H. 1998. On Efficient GMM Estimation of Continuous -Time Asset Dynamics: Implications of the Term Structure of Interest Rates. working paper : Yonsei University. Frode, K. 2007. A real option analysis of investment in hydropower : The case of Norway. Energy Policy 35: pp 5901-5908. Larsen B. M. 2000. Investments and Option Values in a Deregulated Electricity Sector. Proceeding of IAEE Conference Metcalf, G. E., and Hassett, K. A. 1995. Investment under Alternative Return Assumptions Comparing Random Walks and Mean Reversion. Journal of Economic Dynamics and Control 19: pp.1471-1488. Myers, S.C. 1977. Determinants of Corporate Borrowing. Journal of Financial Economics 5: pp. 147-176. Myers, S. C. 1987. Finance Theory and Financial Strategy. Midland Corporate Finance Journal 5: 6-13. Schulmerich, M. 2005. Real Options Valuation. Berlin : Springer. 25

ABSTRACT Although wind power has economic dominance over other renewable energy sources, onshore wind farm has some problems such as noise, land lease and unstability of both wind direction and intensity. Offshore wind farm can reduce these problems, leading to recently attracting attentions both domestically and internationally. This study conducts economic evaluation for the Step 1 Test Bed Construction Project outlined in the Southwestern Offshore Wind Plan published in November 2011. Considering the high uncertainty surrounding the offshore wind power, we adopt the real option method as evaluating method. The project is evaluated as having about the profit of 1,000 billion won including the growth option value of about 830 billion won. This result of high option value implies a cautious decision process is required for the project. Key Words : Offshore wind power, Option to grow, Real option JEL Codes : Q21, Q42 26

에너지경제연구 Korean Energy Economic Review Volume 11, Number 2, September 2012 : pp. 27~55 지역난방열에너지수요예측 27

~ 28

~ 32

v.s. or RSS r RSS u RSS RSS F RSS 34

SupF Max F n AveF Mean F n SupF Ave F Min Min S n 35

[ 그림 1] 분계점추정 k y t i x t i d t i i 36

~ ~ ~ ~ ~ 37

< 표 1> 열사용량의기술통계 ~ 38

[ 그림 2] 분포함수추정 HEAT107.008.007.006.005.004.003.002.001.000 0 40 80 120 160 200 240 280 320 360 400 440 480 520 HEAT207.009.008.007.006.005.004.003.002.001.000-100 0 100 200 300 400 500 600 39

< 표 2> 상태의존성검정 41

~ 42

[ 그림 4] 열수요의계절적요인 ( 월 ) 2008 년 A 지역 2008 년 B 지역 400 400 350 1 2 350 1 2 300 250 200 150 100 3 4 5 10 11 12 Monthly Seasonal Factors 300 250 200 150 100 3 4 5 10 11 12 50 6 7 8 9 50 6 7 8 9 0 1 2 3 4 5 6 7 8 9 10 11 12 0 1 2 3 4 5 6 7 8 9 10 11 12 Month Month 2007 년 A 지역 2007 년 B 지역 400 400 350 1 350 300 250 200 150 100 50 2 3 4 5 6 7 8 9 10 11 12 Monthly Seasonal Factors 300 250 200 150 100 50 1 2 3 4 5 6 7 8 9 10 11 12 0 1 2 3 4 5 6 7 8 9 10 11 12 0 1 2 3 4 5 6 7 8 9 10 11 12 Month Month 43

[ 그림 5] 열수요의계절적요인 ( 시간 ) 2007 년 A 지역 2007 년 B 지역 240 240 220 220 200 180 160 140 120 22 23 24 8 9 21 1 10 20 7 2 3 4 5 6 11 19 12 18 13 14 15 17 16 Hourly Seasonal Factors 200 180 160 140 120 22 23 24 21 8 1 9 20 2 10 7 19 3 11 4 5 6 12 18 13 14 17 15 16 100 2 4 6 8 10 12 14 16 18 20 22 24 100 2 4 6 8 10 12 14 16 18 20 22 24 Hour 2008년 A지역 Hour 2008년 B지역 240 240 220 220 200 180 160 140 120 22 23 24 8 21 9 1 20 7 10 2 19 3 4 5 6 11 12 18 13 14 17 15 16 Hourly Seasonal Factors 200 180 160 140 120 22 23 24 1 8 21 9 20 2 7 10 19 3 4 5 6 11 18 12 13 14 17 15 16 100 2 4 6 8 10 12 14 16 18 20 22 24 100 2 4 6 8 10 12 14 16 18 20 22 24 Hour Hour 44

< 표 3> 추정결과 (2007년지역 A) 46

< 표 4> 추정결과 (2007년지역 B) 48

< 표 5> 지역난방수요모형의예측력 50

< 표 6> 지역난방수요모형의표본외예측력 51

~ 52

접수일 (2012 년 2 월 7 일 ), 수정일 (2012 년 3 월 18 일 ) 게재확정일 (2012 년 4 월 30 일 ), 2001,,, 10/2., 2008,,, 17/3, 91~134., 2002,,, 11/3, 447~463., 2007, :,, 16/4, 763~781., 2001,,, 10/3., 2009,., 2000,., 1999,., 2000, ()., 2005, (). Amjady N., 2001, "Short-term hourly load forecasting using time-series modeling with peak load estimation capability," IEEE Trans Power Systems 16, 798805. Andrews, D.W.K., 1993, Tests for parameter instability and structural change with unknown change point, Econometrica 61, 821856. Arvastson, L., 2001, Stochastic modelling and operational optimization in 53

district-heating systems, Ph. D. thesis, Mathematical Statistics, Lund University, Lund, Sweden. Bertotti, G., 1998, Hysteresis in magnetism: For physicists, materials scientists, and engineers, Academic Press. Davies, R.B., 1987, Hypothesis testing when a nuisance parameter is present only under the alternative, Biometrika 74, 3343. Dotzauer, E., 2002, Simple model for prediction of loads in district-heating systems, Applied Energy 73, 277-284. Hansen, B.E., 1996, Inference when a nuisance parameter is not identified under the null hypothesis, Econometrica 64, 413430. Härdle, W. and O. Linton, 1994, Applied nonparametric methods, Handbook of Econometrics, Vol., Elsevier science B.V., 2295~2339. Hippert S, Pedreira CE, Souza RC., 2001, Neural networks for short-term forecasting: a review and evaluation, IEEE Trans Power Systems 16, 44-55. Infield DG, Hill DC., 1998, Optimal smoothing for trend removal in short-term electricity demand forecasting, IEEE Trans Power Systems 13, 1115-1120. Olofsson, T, Andersson, S., and R. Ostin, 1998, A method for predicting the annual building heating demand based on limited performance data, Energy and Buildings 28, 101-108. Wiklund H., 1991, Short term forecasting of the heat load in a DH-system, Fernwärmwirtschaft International 19, 286294. 54

ABSTRACT The demand for local district-heating systems is closely related to the temperature and socioeconomic variables. The climate change and unpredictable climate variations, technological progress, and preference changes generate wide and prolonged prediction errors from the forecasting model. Based on the stylized facts of the district-heat data, this paper proposes a real-time forecasting model of district-heat demand. The proposed model is predicated on the state-dependent relationship between the temperature and the district heat demand. The socioeconomic seasonal variables and dynamic properties are incorporated to explain the district heat demand in two metropolitan areas for the period 2007~2008. A sizable amount of improvement in prediction accuracy is observed. Keywords : climate change, district-heating, hysteresis, prediction, threshold model JEL Classification : Q41, Q47 55

에너지경제연구 Korean Energy Economic Review Volume 11, Number 2, September 2012 : pp. 57~83 발전용유연탄가격과여타상품가격의 동조화현상에대한실증분석 57

exp exp < 표 1> 일치도계수의유의수준별반응표면회귀식추정치 61

± ± 62

< 표 2> 분석대상상품및지수내역 65

< 표 3> 표본자료의기초통계량 : 상호연계성이없는상품 < 표 4> 표본자료의기초통계량 : 상호연계성이있는상품 66

[ 그림 1] 상승국면및하락국면분석 : 유연탄 < 표 5> 상승국면및하락국면요약 : 상호연계성이없는상품 < 표 6> 상승국면및하락국면요약 : 상호연계성이있는상품 68

< 표 7> 상관계수 : 상호연계성이없는상품 < 표 8> 상관계수 : 상호연계성이있는상품 69

< 표 9> 상관계수의유의수준 : 상호연계성이없는상품 70

< 표 10> 상관계수의유의수준 : 상호연계성이있는상품 < 표 11> 일치도계수 : 상호연계성이없는상품 71

< 표 12> 일치도계수 : 상호연계성이있는상품 72

< 표 13> 일치도계수의유의수준 : 상호연계성이없는상품 < 표 14> 일치도계수의유의수준 : 상호연계성이있는상품 73

접수일 (2012 년 3 월 12 일 ), 게재확정일 (2012 년 4 월 30 일 ) Bry, G. and C. Boschan, 1971, Cyclical Analysis of Time Series: Selected Procedures and Computer Programs, New York: National Bureau of Economic Research. Burns, A. and W. Mitchell, 1946, Measuring Business Cycles, NBER Studies in Business Cycles No.2, (New York: National Bureau of Economic Research). Cashin, P., J. McDermott and A. Scott, 1999, "The myth of comoving commodity prices," IMF Working Paper No. 169. Cashin, P., J. McDermott and A. Scott, 2002, "Booms and slumps in world commodity prices," Journal of Development Economics, 69, 277-296. Deb, P., P. Trivedi, and P. Varangis, 1996, "The excess co-movement of commodity prices reconsidered," Journal of Applied Econometrics 11, 275-91. Harding, D. and A. Pagan, 1999, "Dissecting the cycle," Melbourne Institute Working Paper No. 13/99, Melbourne: University of Melbourne. McDermott, C. and A. Scott, 2000, "Concordance in business cycles," IMF Working Paper No. 37. Pagan, A., 1999, "A framework for understanding bull and bear markets," mimeo, Canberra: Australian National University. Palaskas, T., 1993, "Market commodity prices: the implications of the co-movement and excess co-movement issue," in M. Nissanke and A. Hewitt (eds.), 75

Economic Crisis in Developing Countries: New Perspectives on Commodities, Trade and Finance, New York: Pinter, 89-103. Palaskas, T. and P. Varangis, 1991, "Is there excess co-movement of primary commodity prices?: A co-integration test," Working Paper Series No. 758, International Economics Department, Washington: World Bank. Pindyck, R. and J. Rotemberg, 1990, "The excess co-movement of commodity prices," Economic Journal 100, 1173-89. Trivedi, P., 1995, Tests of some hypotheses about the time series behavior of commodity prices, in G. Maddala, P. Phillips, and T. Srinivasan (eds.), Advances in Econometrics and Quantitative Economics: Essays in Honor of C. Rao, Oxford: Blackwell, 382-412. 76

[ 부록그림 1] 상승국면및하락국면분석 : 코코아 [ 부록그림 2] 상승국면및하락국면분석 : 구리 77

[ 부록그림 3] 상승국면및하락국면분석 : 면화 [ 부록그림 4] 상승국면및하락국면분석 : 목재 78

[ 부록그림 5] 상승국면및하락국면분석 : 소맥 [ 부록그림 6] 상승국면및하락국면분석 : 아연 79

[ 부록그림 7] 상승국면및하락국면분석 : 호주산유연탄 [ 부록그림 8] 상승국면및하락국면분석 : 인도네시아산천연가스 80

[ 부록그림 9] 상승국면및하락국면분석 : Brent 원유 [ 부록그림 10] 상승국면및하락국면분석 : Dubai 원유 81

[ 부록그림 11] 상승국면및하락국면분석 : WTI 원유 [ 부록그림 12] 상승국면및하락국면분석 : 우라늄 82

ABSTRACT Recently, international commodity markets are showing severe price fluctuations and comovement. This study empirically analyzes the comovement phenomenon using steam coal price and other major commodity prices. For this purpose, it tries to compare the empirical results from widely used correlation analysis and concordance analysis. An interesting finding is that the degrees of comovement among unrelated commodities vary depending on the periods analyzed. Key Words : Bituminous Coal Price, Correlation, Comovement JEL Codes : E32, Q41, Q43 83

에너지경제연구 Korean Energy Economic Review Volume 11, Number 2, September 2012 : pp. 85~114 산업별에너지원간대체관계추정을통한탄소세의 CO 2 감축효과분석 * 85

~ ~ 87

~ 88

ln ln ln ln ln ln ln ln 90

ln ln ln ln ln ln ln ln ln ln ln ln 91

ln ln ln ln 92

< 표 1> 비용비중함수추정식의통계적유의도및설명력 94

100

< 표 4> 탄소세부과방법및탄소세율산정흐름 ( 음식담배업종예 ) 101

< 표 5> 이산화탄소배출계수 102

< 표 6> 업종별에너지원별탄소세 ( 전력포함 ) 율 103

< 표 7> 탄소세부과에따른에너지소비량및이산화탄소배출량변화 ( 제조업, 2008 기준 ) 104

105

< 표 8> 탄소세부과에따른이산화탄소배출량변화율비교 106

< 표 9> 탄소세부과에따른업종별에너지소비및 CO 2 배출량변화율 107

< 표 10> 탄소세부과에따른원별에너지소비및 CO 2 배출량 변화율 ( 제조업전체 ) < 표 11> 탄소세부과에따른원별에너지소비및 CO 2 배출량 변화율 ( 에너지다소비 3개업종 ) 108

109

110

접수일 (2012 년 5 월 16 일 ), 게재확정일 (2012 년 7 월 2 일 ). 2006.. 06-11... 2009.... 2005.. 05-13... 2003.. 03-15... 2010...,,, 2011.9.16. (http://www.ekn.kr/ne ws/articleview.html?id xno=72745), 2012a. 12-22(2012.6.15) p.16., 2012b. 12-24(2012.6.29) p.31. Barker, T., P. Ekins, N. Johnstone. 1995. Global Warming and Energy Demand, Routledge. Taylor & Francis Group. Berndt, E. R. and D. O. Wood. 1975. "Technology, Prices, and the Derived Demand for Energy". The Review of Economics and Statistics, Vol. 57, pp. 259-268. and. 1979. "Engineering and Econometric Interpretations of 111

Energy-Capital Complementarity". The American Economic Review, Vol. 69 No. 3, pp. 342-354. Bjørner, T. B. and H. H. Jensen. 2002. "Interfuel Substitution within Industrial Companies: An Analysis Based on Panel Data at Company Level". The Energy Journal, Vol. 23 No. 2, pp. 2750. Brännlund, R. and T. Lundgren. 2004. "A Dynamic Analysis of Interfuel Substitution for Swedish Heating Plants". Energy Economics, Vol. 26, No. 6, pp. 961-976. Cho, W. G., K. Nam, and J. A. Pagán. 2004. "Economic Growth and Interfactor/Interfuel Substitution in Korea". Energy Economics, Vol. 26, pp. 31-50. Considine, T. J. 1989a. "Estimating the Demand for Energy and Natural Resource Inputs: Trade-offs in Global Properties". Applied Economics, Vol. 21, pp. 931-945.. 1989b. "Separability, Functional Form and Regulatory Policy in Models of Interfuel Substitution". Energy Economics, Vol. 11 No. 2, pp. 82-94.. 1990. "Symmetry Constraints and Variable Returns to Scale in Logit Models". Journal of Business and Economic Statistics, Vol. 8 No. 3, pp. 347-353. Considine, T. J. and T. D. Mount. 1984. "The Use of Linear Logit Models for Dynamic Input Demand Systems". Review of Economics and Statistics, Vol. 66, pp. 434-443. Field, B. C. and C. Grebenstein. 1980. "Capital-Energy Substitution in U.S. Manufacturing". The Review of Economics and Statistics, Vol. 62 No. 2, pp. 207-212. Floros, N. and A. Vlachou. 2005. "Energy Demand and Energy-related CO 2 Emissions in Greek Manufacturing: Assessing the Impact of a Carbon Tax". Energy Economics, Vol. 27, pp. 387-413. Fuss, M. A. 1977. "The Demand for Energy in Canadian Manufacturing: An Example of the Estimation of Production Structures with Many Inputs". Journal of Econometrics, Vol. 5 No. 1, pp. 89-116. Greene, W. H. 2003. Econometric Analysis. 5th ed. Prentice Hall. Griffin, J. M. and P. R. Gregory. 1976. "An Intercountry Translog Model of Energy Substitution Responses". The American Economic Review, Vol. 66 No. 5, pp. 112

845-857. IEA. 2010. World Energy Outlook 2010. OECD/IEA. Jones, C. T. 1995. "A Dynamic Analysis of Interfuel Substitution in U.S. Industrial Energy Demand". Journal of Business & Economic Statistics, Vol. 13 No. 4, pp. 459-465. Kim, B. C. and W. C. Labys. 1988. "Application of the translog model of energy substitution to developing countries: The case of Korea". Energy Economics, Vol. 10 No. 4, pp. 313-323. Kratena, K. and M. Wüger. 2003. "The Role of Technology in Interfuel substitution: A Combined Cross-section and Time Series Approach". WIFO Working Papers, No. 204. Ma, H., L. Oxley, J. Gibson and B. Kim. 2008. "China s energy economy: Tecnical change, factor demand and interfactor/interfuel substitution". Energy Economics, Vol. 30, pp. 2167-2183. Pindyck, R. S. 1979. "Interfuel Substitution and the Industrial Demand for Energy-An International Comparison". The Review of Economics and Statistics, Vol. 61, pp. 169-179. Steinbuks, J. 2010. "Interfuel Substitution and Energy use in the UK Manufacturing Sector". EPRG Working Paper 1015. University of Cambridge. Turnovsky, M., M. Folie and A. Ulph. 1982. "Factor Substitutability in Australian Manufacturing with Emphasis on Energy Inputs". The Economic Record, Vol. 58 No. 1, pp. 61-72. Urga, G. and C. Walters. 2003. "Dynamic translog and linear logit models: a factor demand analysis of interfuel substitution in US industrial energy demand". Energy Economics, Vol. 25, pp. 1-21. 113

ABSTRACT The purpose of this study is to analyze the impact of carbon taxes on the reduction of energy consumption and CO 2 emissions by estimating the elasticities of interfuel substitution in 9 Korean manufacturing sectors and to draw its policy implications. For the purpose, this study estimates cost share functions of each energy source throughout static linear logit models, and calculates own-price elasticities and cross-price ones of fuel resources. Furthermore, using estimated elasticities of interfuel substitution for each manufacturing sector, this study analyzes the effect of carbon tax on CO 2 emissions reduction. In order to find the differences in impacts of imposing carbon tax on both primary energy and final energy, this study analyzes both of them and shows that imposing carbon tax on primary energy is more cost effectiveness than imposing carbon tax on final energy in reducing carbon dioxide emissions. That is, if carbon taxes are to be introduced, imposing carbon tax on primary energy like in other nations will make it easier to achieve policy goals. Key Words : logit model, carbon tax, elasticity of substitution, manufacturing JEL Codes : D24, Q41, Q48 114

에너지경제연구 Korean Energy Economic Review Volume 11, Number 2, September 2012 : pp. 115~139 Capacity Constrained Supply Function Equilibria: Modeling and Simulations 115

. Introduction The proliferation of the wholesale electricity market in various jurisdictions throughout the world is a salient feature in the restructuring of the electricity business 1). Various electricity markets have been implemented in which the principal competitive mechanism is an auction. The most promising approach for the analysis of these new electricity markets is based on the idea of supply function equilibria introduced by Klemperer and Meyer(1989). This approach was first applied to the electricity market by Green and Newbery(1992), who viewed that the set-up of supply function equilibrium fit well with the structure of the electricity markets. Klemperer and Meyer show that the supply function equilibrium is characterized by differential equations. In general, the supply function equilibrium approach yields multiple equilibria; for any given set of supply and demand conditions, the market price and vector of firm outputs are not uniquely specified, and many researches have focused on the range and uniqueness of the supply function equilibrium. The range of equilibria can be limited by capacity (Green and Newbery, 1992; Baldick and Hogan, 2002). Genc and Reynolds(2004) analyze how pivotal producers reduce the range of equilibria. From an analytical perspective, Holmberg(2008) shows that there is a unique equilibrium when there are symmetric producers with strictly convex cost functions, and Holmberg(2007) shows that there is a unique equilibrium when producers have identical constant marginal costs. In reality, however, the marginal cost function for firms is usually not constant and their production capacities are 1) See Gross, G(1994) 116

asymmetric. Because Klemperer & Meyer s first-order conditions constitute a system of non-autonomous ordinary differential equations, solving this system analytically is difficult. Baldick and Hogan(2002) posit that solutions that meet the requirement that supply functions must be non-decreasing are difficult to find. There are three exceptions to this assertion, including symmetric firms with identical cost functions, cases with affine marginal costs and no capacity constraints, and cases in which variations in demand are small. However, these situations are too limited to represent real market structure. Due to the multiplicity of supply function equilibrium(sfe) and the difficulty in handling non-autonomous ordinary differential equality, predicting outcomes and generating comparative statics from the SFE model is difficult, especially when realistic features of the electricity market structure are taken account such as asymmetric firm capacities and increasing marginal cost and price dependent demand. In practice, this problem may be avoided at the cost of limiting the strategic spaces of firms; for example, certain applications limit firms to linear supply functions 2). This paper adopts an approach that is used more often in operations research than by academic economists. Instead of simplifying the model to make the computation of optimal actions possible, more details may be incorporated into the model with a simplified solution concept by limiting the set of admissible decision alternatives. Here a restriction is imposed to permit only piecewise linear bidding functions instead of allowing suppliers in a power exchange to submit any type of bid function. Such a restriction is not as limiting as it may first appear because (i) in the real world many bidders use rules of thumb, such as markup or linear bidding strategies, instead of maximally optimal bidding strategies and (ii) there may be institutional restrictions that prevent bidders from adopting an optimal bidding rule even if they desire to do so. For example, the rules of the England and Wales Power Pool limit the bid functions a supplier can submit to 2) Recent applications of this methodology include Baldick et al.(2004). 117

piecewise linear functions with up to three break-points. This approach is particularly effective in incorporating the unique features and operational aspects of electricity generation, consumption, and exchange mechanisms. This paper s model has more realistic representations, including: (i) the explicit incorporation of the firms capacity constraints, and (ii) price-sensitive demand described by a random variable. The principal thrust of this work involves the application of the model to evaluate the performance of the electricity exchange as a function of several important factors, such as the number of firms, the difference in firm sizes and the variability of demand. The paper suggests algorithms to find equilibrium strategies for firms, and it implements simulations and compares the performance of various market structures.. Model Let us consider a market with firms. Each firm owns and operates a single generation unit with capacity, and we use to denote the output of the unit of firm. The output is a function of the price and firms compete for the right to serve the demand through a sealed-bid, uniform price auction. Let the demand be a monotonically nonincreasing function of price. Such a relationship captures the reduction in customer demand as prices increase. The incorporation of the price sensitivity of demand makes the equilibrium price and quantity interdependent. Demand is also considered as a random variable determined by a random factor. It follows that the equilibrium price and the allocation of the equilibrium quantity are also random variables. 118

Each firm submits a supply curve and the aggregated N supply curves form the market supply curve. The market equilibrium is, therefore, determined as follows: (1) The equilibrium price is determined by the supply demand condition in eq.(1). The production costs of a unit are expressed as a polynomial, or in certain cases, as piecewise polynomial function of the unit output. In this exposition, a piecewise quadratic representation is adopted. is used to denote the costs of unit to produce where and is the index of intervals of the piecewise polynomial function. (2) (3) The polynomial coefficients, are selected to reflect the so-called no load costs of the unit and to ensure that the marginal cost is a nondecreasing function of output. Marginal cost has a simple piecewise linear form, expressed as follows: (4), (5) The form in (4) and (5) is used to specify the form of the bid function for each unit. For simplicity, piecewise linear bid functions are considered. Let the bid function in a unit's output range be expressed as follows: 119

, (6), (7) Here, is the bid price of the unit for supplying. Because price and cost are independent, does not need to have any relationship with the marginal cost. The parameters and are selected to ensure that the inverse function exists, and specifies the bidding strategy of firm. The supply function is formally defined to be a function of the price and the bidding strategy and represented by the following: max (8) The term max is a specified maximum price and is the maximum output of unit. Next, the market supply curve is constructed. Let be the collection of firm s supply functions, resulting in the market supply function when firms supply functions are aggregated: (9) The equilibrium condition in eq (1) is set as follows: (10) and determines the equilibrium price. Clearly, is a function of and the collection of strategies. 120

The profit of firm is given in the following (11) Each firm can determine its strategy to maximize its profits. However, the equilibrium price is a function of and, where, is the complement of in the set. The strategy set is made up of the strategies that are selected to maximize the expected value of the profits. Thus, each firm would choose to maximize the following: (12). Description of algorithms This section describes algorithms to find equilibrium supply function for each firm through simulation. By definition of supply function equilibrium, each firm must select the best supply function under the restrictions imposed by this paper, given the other firms supply functions. Thus, the supply function of a firm must make the biggest expected profit among other alternatives that satisfy our restrictions. The actual simulation process repeats to find every firm s supply function that maximizes expected profit and to update the firm s strategy with the supply function of each firm until all firms supply function converges. When searching for a firm s optimal supply function, the strategies of the other firms are fixed. Because 121

demand is uncertain, profits expected from random draws must be calculated. Due to the restrictions imposed on supply functions, the optimal supply function of a firm is a vector of an intercept and slopes which maximize expected profit. Fmin function in MATLAB is used to find the optimal supply functions. Algorithm 1 is for the calculation of the expected profit with a strategy of a specific firm, given the strategies of the other firms. The function obtained from Algorithm 1 is an objective function that aims to maximize at Algorithm 2 with fmin function. Algorithm 2 is for the finding of the supply function equilibrium. Algorithm 1: Given the other firms' (supply function) strategies, calculate the expected profit with a strategy of a specific firm: Initialization For 1:number of iteration Draw demand from a specific probability distribution Initialize price While demand is not equal to supply Calculate demand and supply at the price If demand is greater than aggregate supply, increase the price If aggregate supply is greater than demand, decrease the price End Find quantity corresponding to equilibrium price Calculate profit End Calculate the expected profit Algorithm 2 : find an equilibrium strategy Initialization While at least one firm s strategy not converged 122

For 1:number of firms Compute a firm s optimal response based on the other firms strategies (find the strategy which maximize the expected profit from Algorithm 1 using fmin function in MATLAB) Update the firm s strategy End End. Simulation Results To compare the performance of various market structures, simulations are implemented that consist of various combinations of firms. Marginal cost(mc) curves are considered; these are piecewise linear and continuous functions, with one kink point. Each firm has its own capacity so that marginal cost beyond each firm's capacity is infinite. This paper supposes market capacity to be 30, and in symmetric market structures, the capacity of each firm is assumed to be 30 divided by the number of firms. In asymmetric market structures, a large firm is either two or three times larger than the small firm(s), but the summation of all firms' capacity remains at 30. The marginal cost curves of symmetric firms are identical. The MC curve of a large firm that is twice as large as that of a small firm is exactly the horizontal summation of two identical small firms, and the MC curve of a large firm three times as large as a small firm is exactly the horizontal summation of three identical small firms. If the horizontal summation of MC curves of all firms in the market is determined, then the MC curve of the market is expressed as follows: 123

, (13), (14), (15) To facilitate computations, piecewise linear and continuous bid(or supply) functions are considered, and restrictions are imposed to allow for only one knot. Another restriction imposed by this paper is to place the kink point of two slopes at the same kink point as that of the marginal cost curve, meaning that the optimal strategy of each firm consists of a intercept and two slopes. Let demand be, where is a random factor drawn from a certain distribution. When using a uniform distribution for, is drawn from an uniform distribution where the domain is [0, 40]. First, to compare the performance of market structures with different numbers of firms, 5 simulations are performed with the equal-sized firm structures of 2, 3, 4, 5 and 10 firms, respectively, where the random factor of the demand is drawn from a uniform distribution. Table 1 provides the simulation results for the optimal strategies of firms with the aforementioned restrictions. Table 1. Equilibrium supply curve of each firm of uniform distribution, symmetric firms number of firms intercept the first slope the second slope 2 1.8324 (2)* 1.0888 (0.50) 1.8741 (1.00) 3 1.8702 (2) 1.2044 (0.75) 2.2336 (1.50) 4 1.9006 (2) 1.3985 (1.00) 2.6556 (2.00) 5 1.9211 (2) 1.6180 (1.25) 3.1126 (2.50) 10 1.9629 (2) 2.8139 (2.50) 5.5364 (5.00) * intercepts and slopes of MC curves are in parentheses. To compare market performance, these supply functions must be aggregated. Table 2 offers aggregated supply curves corresponding to each market structure. 124

The results of the simulation are intuitive, as aggregated supply functions are steeper with fewer firms. In the market composed of 10 firms, the firm's optimal strategy is very close to the marginal cost pricing. Table 2. Aggregate supply curve of uniform distribution, symmetric firms case number of firms intercept the first slope the second slope marginal cost pricing 2 0.25 0.5 2 1.83 0.545 0.935 3 1.87 0.4016 0.743 4 1.90 0.35 0.665 5 1.92 0.324 0.622 10 1.96 0.28 0.55374 Second, to see the effects of demand variances on the optimal strategies of firms, simulations are carried out with different demand distributions, beta distributions where the mean and range are the same as the uniform distribution considered earlier but where the variances differ from the uniform distribution. The parameters of beta distribution 1 are(2.5, 2.5) and the parameters of beta distribution 2 are(5.5, 5.5). The mean of the uniform distribution and the two beta distributions are the same, but the variance of beta distribution 2 is one fourth of the uniform distribution and one half of that of beta distribution 1. The variance of beta distribution 1 is smaller than the uniform distribution but greater than that of beta distribution 2. Table 3 provides the simulation results for the optimal strategies of firms and the aggregate supply functions of beta distribution 1 and Table 4 provides the simulation results for the optimal strategies of firms and the aggregate supply functions of beta distribution 2. 125

Table 3. Equilibrium supply curve of each firm and aggregate supply function of beta distribution 1, symmetric firms case Equilibrium supply curve of each firm number of firms intercept the first slope the second slope 2 1.6860(2)* 1.1110(0.50) 2.0735(1.00) 3 1.7307(2) 1.2406(0.75) 2.3368(1.50) 4 1.7988(2) 1.4301(1.00) 2.7460(2.00) 5 1.8425(2) 1.6467(1.25) 3.1951(2.50) 10 1.9237(2) 2.8419(2.50) 5.5967(5.00) Aggregate supply curve number of firms intercept the first slope the second slope marginal cost pricing 2 0.25 0.5 2 1.6860 0.5555 1.0368 3 1.7307 0.4135 0.7789 4 1.7988 0.3575 0.6865 5 1.8425 0.3293 0.6390 10 1.9237 0.2842 0.5597 * intercepts and slopes of MC curves are in parentheses. Table 4. Equilibrium supply curve of each firm and aggregate supply curve of beta distribution 2, symmetric firms Equilibrium supply curve number of firms intercept the first slope the second slope 2 1.5353(2) 1.1303(0.50) 2.8714(1.00) 3 1.5280(2) 1.2911(0.75) 2.5441(1.50) 4 1.6396(2) 1.4813(1.00) 2.8865(2.00) 5 1.7138(2) 1.6974(1.25) 3.3088(2.50) 10 1.8554(2) 2.8943(2.5) 5.6690(5.00) Aggregate supply curve number of firms intercept the first slope the second slope marginal cost pricing 2 0.25 0.5 2 1.5353 0.5652 1.4357 3 1.5280 0.4304 0.8480 4 1.6396 0.3703 0.7216 5 1.7138 0.3395 0.6618 10 1.8554 0.2894 0.5669 * intercepts and slopes of MC curves are in parenthesis. 126

Table 5 shows a comparison of the supply functions that results from the simulations conducted with different demand distributions. The supply curves are steeper when demand is drawn from the beta distributions than when demand is drawn from the uniform distribution. Additionally, the supply curve is steeper when demand is drawn from beta distribution 1 than when it is drawn from beta distribution 2. The variance of beta distribution 1 is smaller than that of the uniform distribution but greater than that of beta distribution 2, and the equilibrium supply functions of beta distribution 1 is steeper than those of the uniform distribution, but it is gentler than those of beta distribution 2. Table 5. Comparison of supply functions equilibrium of different demand distributions demand distribution intercept the first slope the second slope uniform distribution 2 firms beta distribution1 2 firms beta distribution2 2 firms uniform distribution 3 firms beta distribution1 3 firms beta distribution2 3 firms 1.8324 1.0888 1.8741 1.6860 1.1110 2.0735 1.5353 1.1303 2.8714 1.8702 1.2044 2.2336 1.7307 1.2406 2.3368 1.5280 1.2911 2.5441 So far, we have addressed a market structure that, is composed of symmetrically sized firms with identical marginal cost curves. Next, we address market structures composed of different sized of firms. Third, to compare the performance of the market structure composed of the different sizes of firms, 3 simulations are conducted with 2, 3 and 4 different sized firms, respectively, where the random factor of the demand is drawn from a uniform distribution. In such asymmetric 127

market structures, there is one large firm and one or more small firm of the same size. Table 6 shows the simulation results regarding a market structure in which there is one large firm which is twice as large as the small firms. Table 7 is a presentation of markets where the large firm is three times the size of the small firms. Table 6 and 7 show that small firms submit their supply functions more aggressively than do large firms. The difference between the supply function curves and the MC curves of small firms is smaller than such difference in one large firm. For example, in a market structure of two firms in which a large firm is twice as large as the small one, the second slope of the small firm is approximately 1.5 times the size of the MC curve while the second slope of the large firm is approximately 2 times the size of the MC curve. Table 6. Supply functions where the large firm is twice as large as other firms Market structure: one large firm and one small firm intercept the first slope the second slope SF of small firm 2.0332(2)* 1.2527(0.75) 2.2049(1.5) SF of large firm 1.5808(2) 1.0967(0.375) 1.4689(0.75) Market structure: one large firm and two small firms intercept the first slope the second slope SF of small firms 1.9565(2) 1.4185(1) 2.6522(2) SF of large firm 1.6013(2) 1.1016(0.5) 1.6324(1) Market structure: one large firm and three small firms intercept the first slope the second slope SF of small firms 1.9477(2) 1.6267(1.25) 3.1207(2.5) SF of large firm 1.6793(2) 1.1525(0.625) 1.8502(1.25) * intercepts and slopes of MC curves are in parentheses 128

Table 7. Supply functions where the large firm is three times as large as other firms Market structure: one large firm and one small firm intercept the first slope the second slope SF of small firm 2.1116(2)* 1.4334(1) 2.9157(2) SF of large firm 1.6595(2)* 1.0926(0.33) 2.4996(0.67) Market structure: one large firm and two small firms intercept the first slope the second slope SF of small firms 1.9928(2) 1.6480(1.25) 3.2272(2.5) SF of large firm 1.6350(2) 1.0492(0.4167) 1.8976(0.834) structure: one large firm and three small firms intercept the first slope the second slope SF of small firms 1.9680(2) 1.8735(1.5) 3.5981(3) SF of large firm 1.6129(2) 1.0808(0.5) 1.6230(1) * intercepts and slopes of MC curves are in parentheses We next calculate the expected quantities and expected prices with these equilibrium supply functions. The expected price decreases as the number of firms rises and increases as the asymmetry between firm sizes grows. The expected quantity increases as the number of firms rises and decreases as the asymmetry between firm sizes grows. Table 8. Expected price and quantity with supply function equilibrium market structure number of firms expected quantity expected price symmetric firms 2 12.6357 9.4486 symmetric firms 3 13.7758 8.3085 symmetric firms 4 14.2572 7.8271 symmetric firms 5 14.5132 7.5711 symmetric firms 10 14.9530 7.1313 twice large 2 12.3693 9.7151 twice large 3 13.5659 8.5184 twice large 4 14.1113 7.9730 three times large 2 11.7898 10.2945 three times large 3 13.1485 8.9359 three times large 4 13.8451 8.2392 129

Fourth, to compare the market powers of various market structures, Lerner indexes 3) are calculated based on the previous simulation results. For purposes of comparison, two additional benchmark cases are also considered. First, on the assumption that electricity is storable, the Cournot equilibrium is calculated using the uniform distribution for the demand uncertainty. The power system is a prototype of a just-in-time-manufacturing systems in which all output must be consumed exactly at the time it is manufactured. If electricity is storable, firms produce electricity in off-peak times in an amount that is greater than immediate demand and save the remaining amount and to meet demand at peak times. In this case, the problem of firms is the same as Cournot competition, and firms aggregate all variant demands and optimize their production by maximizing their expected profits. Table 9 shows the results for Cournot quantities and Cournot Table 9. Price and quantity of the Cournot equilibrium with storable goods market structure number of firms Cournot quantity Cournot price symmetric firms 2 11.4768 10.6075 symmetric firms 3 12.6849 9.3994 symmetric firms 4 13.3896 8.6947 symmetric firms 5 13.8515 8.2328 symmetric firms 10 14.8770 7.2073 twice large 2 11.3470 10.7373 twice large 3 12.5482 9.5361 twice large 4 13.2612 8.8231 three times large 2 11.0439 11.0404 three times large 3 12.2059 9.8784 three times large 4 12.9373 9.1470 3) The Lerner index, named after the economist Abba Lerner, describes a monopoly's market power. Mathematically, it is measured with the following formula: L=(P-MC)/P, where L is the Lerner index, P is the selling price and MC is the marginal cost. For a perfectly competitive firm (where P=MC), L=0. It has no market power. 130

prices. The pattern of expected prices and expected quantities is the same as in the case of supply function equilibrium. The expected price decreases when the number of firms rises and increases as firms become more asymmetrical. The expected quantity increases as the number of firms rises and decreases as firms become more asymmetrical. Next, the Cournot points for every demand realization are calculated, with the results shown in Table 10. If firms can adjust their production after demand is realized, then firms will produce at the point of the Cournot equilibrium for each demand. Because firms must submit one supply function for all variant demands, firms cannot attain Cournot points of variant demands. Table 10. Expected price and quantity of the Cournot points market structure number of firms expected quantity expected price symmetric firms 2 11.3092 10.7750 symmetric firms 3 12.3993 9.6851 symmetric firms 4 13.4229 8.6613 symmetric firms 5 13.4239 8.6605 symmetric firms 10 14.2943 7.7899 twice large 2 11.1320 10.9522 twice large 3 12.2297 9.8545 twice large 4 12.8699 9.2144 three times large 2 10.8912 11.1931 three times large 3 11.9680 10.1163 three times large 4 12.6392 9.4451 Table 11, compares the market powers of the supply function equilibrium obtained from previous simulations to the Cournot equilibrium for storable goods and Cournot points in various market structures. The means of the Lerner indexes of the supply functions are far smaller than those of the Cournot points. The means of Lerner indexes of the Cournot points are far smaller than those of the 131

Cournot equilibrium, assuming that electricity is storable. These difference stems from the production of electricity being a just-in-time-manufacturing system. Table 11. Lerner indexes Lerner index of the market composed of symmetric firms number of firms supply function Eq Cournot Eq** Cournot points 2 0.3585*(0.1169) 0.5410** 0.4795*(0.1234) 3 0.2312 (0.0850) 0.4498 0.3906(0.1030) 4 0.1674 (0.0648) 0.3850 0.3292(0.0883) 5 0.1305 (0.0519) 0.3365 0.2844(0.0770) 10 0.0614 (0.0255) 0.2065 0.1712(0.0490) Lerner index of the market composed of asymmetric firms when the large firm is twice the size of small firms number of firms supply function Eq Cournot Eq ** Cournot points 2 0.3834* 0.5495 0.4912* 3 0.2568 0.4613 0.4051 4 0.1874 0.3976 0.3444 Lerner index of the market composed of asymmetric firms when the large firm is three times the size of small firms number of firms supply function Eq Cournot Eq ** Cournot points 2 0.4293* 0.5688 0.5066* 3 0.3021 0.4886 0.4264 4 0.2221 0.4278 0.3663 * Lerner indexes are means for supply function and Cournot points for each demand realization ** Cournot equilibrium is calculated assuming storable goods The Lerner index decreases as the number of firms increases, and it increases as firm size becomes more asymmetrical. For example, the Lerner index of the market composed of three symmetric firms is almost the same as that of the market in which there are four firms, but one large firm is three times the size of the three small firms. So far, the Lerner indexes have been calculated at the market level. Table 12, however, shows market power at the firm level. In an asymmetric market structure, the Lerner index of a large firm is larger than that of small firms. 132

Because the unit price paid is the same for all production supplied by firms in the uniform price auction, the proportion of the production of a small firm is larger than the proportion of the size of the firm. Because price is the same for every firms, the MC of the large firm is smaller. Large firms produce less in proportion to the size. The difference between the Lerner indexes of a large firm and those of small firms increases with the asymmetry of firm sizes. Table 12. The Lerner index of each firms in the market with symmetric and asymmetric firms Lerner indexes symmetric firms: 2 firms (0.3585, 0.3585) symmetric firms: 3 firms (0.2312, 0.2312, 0.2312) symmetric firms: 4 firms (0.1674, 0.1674, 0.1674, 0.1674) large firm is 2 times larger: 2 firms (0.2700,0.4343) large firm is 2 times larger: 3 firms (0.1842,0.1842, 0.3246) large firm is 2 times larger: 4 firms (0.1390, 0.1390, 0.1390, 0.2563) large firm is 3 times larger: 2 firms (0.2329 0.4856) large firm is 3 times larger: 3 firms (0.1611 0.1611 0.3869) large firm is 3 times larger: 4 firms (0.1210, 0.1210, 0.1210, 0.3150) Table 13 shows consumer surplus, profits and deadweight loss. Deadweight loss is calculated from a comparison with competitive equilibrium satisfying p = MC. Consumer surplus depends on equilibrium quantities and prices. As expected, consumer surplus increases as the number of firms of the market rises, and total profits decrease as the number of firms increases. Deadweight loss is larger in a less competitive market so that it decreases as the number of firms rises. 133

Table 13. Welfare analysis of markets composed of different numbers of symmetric firms number of firms consumer surplus profit deadweightloss 2 103.7535 90.5738 6.8982 3 123.1216 84.0705 2.3382 4 131.9601 76.4573 1.1130 5 136.8241 72.0659 0.6405 10 145.4539 63.9480 0.1285 Table 14. Comparison of consumer surplus # symmetric firms twice as large as* 3 times as large as ** 2 103.7535 99.7138 89.8344 3 123.1216 119.5683 111.9763 4 131.9601 129.3343 124.5264 * large firm is twice as large as small firms ** large firm is 3 times as large as small firms Total profits of symmetric firms are smaller than those of the same number of asymmetric firms. This paper intentionally structures the market so that production units of a large firm are obtained by multiplying the production units of a small firm. The results indicate that the profit of a large firm is smaller than the proportionate profit of a smaller firm. In asymmetric market structures, small firms tend to bid more aggressively, and consequently, the amount of demand a small firm serves is large in proportion to its size. Because the unit price paid is the same for all production supplied by firms in the uniform price auction, a small firm's proportionate profit is larger than the proportionate size of the firm. 134

Table 15. Profits profits of firms symmetric firms: 2 (45.2869, 45.2869) symmetric firms: 3 (28.0235, 28.0235, 28.0235) symmetric firms: 4 (19.1143, 19.1143, 19.1143, 19.1143) large firm is 2 times larger: 2 firms (38.5848, 61.7516) large firm is 2 times larger: 3 firms (22.9364, 22.9363, 40.3368) large firm is 2 times larger: 4 firms (16.1696, 16.1695, 16.1695, 29.8040) large firm is 3 times larger: 2 firms (33.7527, 71.1358) large firm is 3 times larger: 3 firms (20.7014, 20.7014, 49.4103) large firm is 3 times larger: 4 firms (14.6063, 14.6063, 14.6063, 37.7091) Table 16 compares the deadweight loss in different market structures. As expected, dead-weight loss is larger when the market consists of a smaller number of firms. Severe asymmetry between firms results in large deadweight loss. For example, the deadweight loss in the market which consists of two small firms and one large firm 3 times as large as each of the small firms is almost the same as the market composed of two symmetric firms. Table 16. Comparison of deadweight loss # symmetric firms large firm is 2 times larger large firm is 3 times larger 2 6.8982 9.4801 14.8074 3 2.3382 3.7525 6.7408 4 1.1130 1.8834 3.4759 Even though asymmetric equilibria for symmetric market structures are not excluded in the computation process finding the optimized supply function equilibrium, only symmetric equilibria are obtained. Symmetric firms submit the same bid functions so that production facilities are operated in equilibrium up to the same marginal cost. However, asymmetric firms submit different bid functions, 135

and as a result, the higher cost production facility of one firm is operated rather than the lower cost production facility(ies) of the other firm(s). Table 17 shows the efficiency loss from the operation of less efficient production facilities. A more severe asymmetry results in a larger efficiency loss. Table 17. Comparison of efficiency loss # symmetric firms large firm is 2 times larger large firm is 3 times larger 2 0 1.2761 2.6407 3 0 0.7920 2.1115 4 0 0.4625 1.3744. Conclusion This paper models the restructured wholesale electricity markets as oligopolies facing uncertain demand and capacity constraints where each firm chooses as its strategy a supply function, as described by Klemperer and Meyer. It intends to provide comparative statics for the SFE model, which incorporates the more realistic market structure features. To avoid the difficulties of solving the SFE model analytically, especially considering the more realistic features of the electricity market structure and to facilitate the computation of optimal strategies, this paper considers piecewise linear bid(or supply) functions, imposing restrictions that allow for a limited number of knots. Simulations using various combinations of firms are implemented to compare the performance of various market structures. The results of these simulation show that firms equilibrium supply functions are steeper with fewer firms and with a more severe asymmetry between the small firms and a large firm. We also compare the Cournot equilibrium, assuming 136

storability, and the Cournot points of every demand realization. The results show that the Lerner index of supply function equilibrium with capacity constraints is far smaller than that of the Cournot equilibrium, assuming storability, and the Cournot points. The Lerner index decreases as the number of firms rises and it increases as the asymmetry between firm sizes grows. Because we can get the similar results with Cournot equilibrium, these comparative statics of SPE accord well with our predictions. The results showing that the firms equilibrium supply function are steeper and the Lerner indexes are larger with a smaller variance of demand correlate to the characteristic of the supply function equilibrium where each agent s best strategy considers all demand uncertainty or variation. From the supplier s point of view, they can better optimize their strategies and earn greater profits with a smaller variance of demand. We can infer from these simulation results that the more firms in the restructured wholesale electricity market should result in lower price and smaller deadweight loss. A market composed of similarly scaled firms should have a lower market price and smaller deadweight loss than a market in which one firm dominates the other firms in scale. However, firms in a power exchange market in which they submit price-quantity schedules cannot attain Cournot profits. Because the firm's best strategies must consider the variance of demand, its profits are limited. This paper s modeling framework and simulation method can serve for the study of various policy issues, including investigating the effects of market design and alternative bidding rules on expected price, price variability, and economic efficiency. In addition to this modeling framework and simulation method, more meaningful results may be produced using more realistic cost and demand functions from real electricity market data. 접수일 (2012 년 2 월 28 일 ), 수정일 (2012 년 7 월 17 일 ), 게재확정일 (2012 년 8 월 1 일 ) 137

Baldick, R., and W. Hogan. 2002. Capacity constrained supply function equilibrium models for electricity markets: Stability, non-decreasing constraints, and function space iterations. POWER Paper PWP-089, University of California Energy Institute. Baldick, R., R. Grant, and E. Kahn. 2004. Theory and Application of Linear Supply Function Equilibrium in Electricity Markets. Journal of Regulatory Economics 25 (2): pp143-67 Genc, T. and Reynolds S. 2004. Supply Function Equilibria with Pivotal Electricity Suppliers. Eller College Working Paper No.1001-04, University of Arizona. Green, Richard J. and David M. Newbery. 1992. Competition in British Electricity Spot Market. Journal of Political Economy 100 : pp929-953. Gross, G. 1994. Electric Utility Industry Restructuring, Proceedings of the NSF Workshop on Electric Power Systems Infrastructure, Washington State University, Pullman, pp17-40, October 27-28, Holmberg, P. 2007. Supply Function Equilibrium with Asymmetric Capacities and Constant Marginal Costs. The Energy Journal, 28 (2): pp55-82. Holmberg, P. 2008. Unique Supply Function Equilibrium with Capacity Constraints. Energy Economics, vol. 30: pp148-172. Klemperer, Paul D. and Margaret A. Meyer. 1989. Supply Function Equilibria in Oligopoly Under Uncertainty. Econometrica 57: pp1243-1277 138

염수현 139

에너지경제연구 Korean Energy Economic Review Volume 11, Number 2, September 2012 : pp. 141~163 신재생에너지지원정책의지대발생효과와규제 : 신재생에너지공급의무화제도 (RPS) 를중심으로 141

142

143

144

2) 145

[ 그림 1] 공급의무화제도와지대발생원리 146

147

< 표 1> 신재생에너지발전기업의총수입액비교 ( 벨기에사례 ) 148

149

< 표 2> 에너지원별신재생에너지공급인증서 (REC) 가중치 150

151

[ 그림 2] 에너지원별발전비용에따른가중치의결정 8) 152

α 11) 153

< 표 3> 에너지원별발전차액지원제도기준가격 154

155

< 표 4> 공급의무화제도와발전차액지원제도의에너지원별가중치비교 156

157

158

접수일 (2012 년 6 월 25 일 ), 게재확정일 (2012 년 8 월 2 일 ) 159

. 2012. :,, 18 1, 217-238.. 2008.,, 20 3, 107-133.. 2011. : FIT RPS,, 45 3, 305-333.. 2002. 21,, 36 3, 147-166.. 2011. : RPS,, 19 4, 79-111.. 2008. : EU, 7 4, 1-37.. 2009. (RPS), :. 2005., :.. 2011. (RPS), 15 4, 1-8.. 2010. -, :.. 2010., :.. 2010., :. 160

. 2009., :.. 2009.,, 18 1, 187-209. Bergek, A., Jacobsson, S. 2010. "Are tradable green certificates a cost-efficient policy driving technical change or a rent-generating machine? Lessons from Sweden 2003-2008" Energy Policy 38: pp.1255-1271. Buckman, G. 2011. "The effectiveness of Renewable Portfolio Standard banding and carve-outs in supporting high-cost types of renewable electricity" Energy Policy 39: pp.4105-4114. Hass, R., Eichhammer, W., Huber, C., Langniss, O., Lorenzoni, A., Madlener, R., Menanteau, P., Morthorst, P. E., Martins, A., Oniszk, A., Schleich, J., Smith, A., Vass, Z. and Verbruggen, A. 2004. "How to promote renewable energy systems successfully and effectively" Energy policy 32: pp.833-839. Hass, R., Resch, G., Panzer, C., Busch, S., Ragwitz, M., Held, A. 2011. "Efficiency and effectiveness of promotion systems for electricity generation from renewable energy sources Lessons from EU countries" Energy Policy 36: pp.2186-2193. Jappe, A. B., Newell, R. G., Stavins, R. N. 2005. "A tale of two market failures: Technology and environmental policy" Ecological Economics 54: pp.164-174. Komor, P. 2004. Renewable Energy Policy, New York: iuniverse. Langniss O. and Wiser, R. 2003. "The renewables portfolio standard in Texas: an early assessment" Energy Policy 31: pp.527-535. Menanteau, P. Finon, D. and Lamy, M. L. 2003. "Prices versus quantities: choosing policies for promoting the development of renewable energy" Energy policy 31: pp.799-812. Mitchell, C. and Connor, P. 2004. "Renewable energy policy in the UK 1990-2003" Energy Policy 32: pp.1935-1947. Ricardo, D. 1996. Principles of Political Economy and Taxation, Amherst, N.Y.: 161

Prometheus Books. (Reprinted version of the original title in 1817) Tullock, G. 2005. The Rent-Seeking Society. Liberty Fund, Indianapolis. Verbruggen, A. 2009. "Performance evaluation of renewable energy support policies, applied on Flanders tradable certificates system" Energy Policy 37: pp.1385-1394. Verbruggen, A. 2008. "Windfalls and other profits" Energy Policy 36: pp.3249-3251. Wiser, R., Barbose, G., Holt, E. 2011. "Supporting solar power in renewable portfolio standards: Experience from the United States" Energy Policy 39: pp.3984-3905. Woodman, B., Mitchell, C. 2011. "Learning from experience? The development of the Renewables Obligation in England and Wales 2002-2010" Energy Policy 39: pp.2914-3921. 162

ABSTRACT Electricity from renewable energy sources requires support to correct the negative externality of electricity generation based on fossil fuels as well as the positive externality of technology innovation effects of renewable energy technologies. However, poorly designed market regulations can lead to windfall profit for the renewable energy sector, increasing policy costs. In addition, those who benefit from such regulations may actively seek rents. Therefore, any regulatory intervention should be designed to reduce rents and rent-seeking behavior. This study investigates the rent generated through Renewable Portfolio Standard(RPS) and examines how it can be reduced through effective policy designs by focusing on the banding system of RPS in Korea. Multipliers in the banding system are examined and compared to those of Feed-in Tariff(FIT). This study concludes by summarizing its policy implication to improve the effectiveness of the banding system of RPS. Key Words : Renewable energy, Renewable Portfolio Standard, JEL Codes : Q48, L51 Feed-in-Tariff, Rent, Rent-seeking 163

에너지경제연구 Korean Energy Economic Review Volume 11, Number 2, September 2012 : pp. 165~190 스마트미터 인홈디스플레이일반가구수요분석 165

166

~ 167

168

169

170

~ ~ ~ 171

~ 172

173

174

i f i f Pr exp 175

exp Pr exp Pr P r Pr P r 기존전력량계 스마트 기존전력량계 기존전력량계 176

스마트 스마트 스마트 스마트 스마트 스마트 < 표 1> 인구통계학적특징및유틸리티서비스이용특징 ( ) 177

스마트 스마트 178

스마트 스마트 스마트 스마트 ~ 179

< 표 2> 추정결과 스마트 스마트 180

스마트 ~ ~ 181

182

183

184

접수일 (2012 년 2 월 20 일 ), 게재확정일 (2012 년 4 월 9 일 ) 185

. 2009.. Issue Paper,. 2009. AMI(Automatic Metering Infrastructure). 27(11) : 93-97. 2010. Global Smart Metering. 2010 8 : 100-107. 2009. ().. 2010a.,.. 2010b.. 2010 1. 2009.. 2009 7 : 535-536. 2010. 2010 (IHD).. 2011. 2011 : (IHD) (2). Adamowicz, W., P. Boxall, M. Williams and J. Louviere. 1998. Stated Preference Approaches for Measuring Passive Use Values: Choice Experiments and Contingent Valuation. American Journal of Agricultural Economics 80 : 64-75. Ahn, J., J. Lee, J. D. Lee and T. Y. Kim. 2006. An Analysis of Consumer Preferences among Wireless LAN and Mobile Internet services. ETRI Journal 28 : 205-215. Alvarez-Farizo, B. and N. Hanley. 2002. Using Conjoint Analysis to Quantify Public Preferences over the Environmental Impacts of Wind Farms: an Example from 186

Spain. Energy Policy 30 : 107-116. Carlsson, F. 2003. The Demand for Intercity Public Transport: the Case of Business Passengers. Applied Economics 35 : 41-50. Darby S. 2006. The Effectiveness of Feedback on Energy Consumption. Oxford, United Kingdom, http://www.defra.gov.uk/environment/climatechange/uk/energy/researc h/ pdf/energyconsump-feedback.pdf. Darby S. 2010. Smartmetering: What Potential for Householder Engagement?. Building Research and Information 38(5) : 442457. Department of Energy and Climate Change (DECC). 2009. Energy Metering. A Consultation on Smart Metering for Electricity and Gas, DECC/The Stationery Office, London, http://www.decc.gov.uk/en/content/cms/consultations/smart_metering/smart_meter ing.aspx. Elburg, H. 2009. Smart metering and In-home Energy Feedback; Enabling a Low Carbon Life Style. ECEEE 2009 Summer Study : 1745-1750. Faruqui A., S. Sergici and A. Sharif. 2010. The Impact of Informational Feedback on Energy Consumption: A Survey of the Experimental Evidence. Energy 35 : 1598 1608. Green, P. E. and V. Srinivasan. 1978. Conjoint Analysis in Consumer Research: Issues and Outlooks. Journal of Marketing Research 5 : 103-123. Jeong, G., D. Koh and J. Lee. 2008. Analysis of the Competitiveness of Broadband over Power Line Communication in Korea. ETRI Journal 30 : 469-479. Kim, Y. 2005. Estimation of Consumer Preferences on New Telecommunication Service: IMT 2000 Service in Korea. Information Economics and Policy 17 : 73-84. Kim, Y., J. Lee and D. Koh. 2005. Effects of Consumer Preferences on the Convergence of Mobile Telecommunications Devices. Applied Economics 37 : 817-826. Layton, D. F. 2000. Random Coefficient Models for Stated Preference Surveys. Journal of Environmental Economics and Management 40 : 21-36. 187

Logica CMG. 2007. Turning concern into Action: Energy Efficiency and the European consumer. London. Mackenzie, J. 1993. A Comparison of Contingent Preference Models. American Journal of Agricultural Economics 75 : 593~603. Office of the Gas and Electricity Markets (OFGEM). 2010a.. Smart Metering Implementation Programme: In-Home Display. Supporting Document, 27 July 2010, http://www.decc.gov.uk/assets/decc/consultations/smart-meter-imp-prospectus/233-sma rt- metering-imp-in-home.pdf Office of the Gas and Electricity Markets (OFGEM). 2010b. Consumers views of Smart Metering. Report by FDS International, http://www.ofgem.gov.uk/e-serve/sm/documenta tion/documents1/smart%20metering%20-%20consumer%20fds%20report.pdf Office of the Gas and Electricity Markets (OFGEM). 2011a. Smart Metering - What It M eans for Britain Homes. Factsheet 101, http://www.ofgem.gov.uk/media/factsheets/ Documents1/consumersmartmeteringfs.pdf Office of the Gas and Electricity Markets (OFGEM). 2011b. OFGEM Consumer First Pa nel Year 3. Opinion Leader, March 2011, http://www.ofgem.gov.uk/markets/retmkt s /rmr/documents1/ofgem_opinionleader_tariff_report_final.pdf Roe, B., K. J. Boyle and M. F. Teisl. 1996. Using Conjoint Analysis to Derive Estimates of Compensating Variation. Journal of Environmental Economics and Management 31 : 145-159. Rossini, G. 2009. Hydro One: In-home Real Time Display. Customer Feedback from a 30,000 Unit Deployment. Presentation given at the Home Energy Displays Conference. Roth K. and J. Brodrick. 2008. Emerging Technologies: Home Energy Displays. ASHRAE Journal July 2008 : 136-138 San Miguel, F., M. Ryan and E. McIntosh. 2000. Applying Conjoint Analysis in Economic Evaluations: an Application to Menorrhagia. Applied Economics 32 : 832833. 188

Train, K. 2003. Discrete Choice Methods with Simulation. Cambridge University Press. Vasconcelos, J. 2008. Survey of Regulatory and Technological Developments Concerning Smart metering in the European Union Electricity Market. Robert Schuman Centre for Advanced Studies, Florence, Italy. Walters, N. 2008. Can Advanced Metering Help Reduce Electricity Costs for Residential Consumers?. AARP Public Policy Institute, Insight on the Issues 18, November 2008 : 1-8 Wooldridge, J. 2010. Econometric Analysis of Cross Section and Panel Data. The MIT Press, Cambridge 189

ABSTRACT This study attempts to estimate the residential demand and willingness to pay for smart meter and In-home-display (IHD) using conjoint survey data in Korea. Further, policy implications regarding measures for constructing smart meter and IHD were derived based on the empirical analysis. From the estimation result, it was found that consumers valued the benefits of smart meter and IHD considerably, while willingness to pay for smart meter and IHD being estimated to be very low, 674 won per month on average. Key Words : Smart Meter, In-home-display, Residential demand, Conjoint Analysis, Binary logit model JEL Codes : L94, Q48, Q49 190

에너지경제연구 Korean Energy Economic Review Volume 11, Number 2, September 2012: pp. 191~219 원자력의경제성 : 쟁점검토와해결과제 * 191

192

193

194

[ 그림 1] 주요발전원별평준화발전원가비교 할인율 5% 의경우 할인율 10% 경우 195

196