Ex post payoffs of a tolling agreement for natural-gas-fired generation in Texas

Transcription

1 Ex post payoffs of a tolling agreement for natural-gas-fired generation in Texas The 5th IAEE Asian Conference, University of Western Australia, Spring, 2016 Yun LIU, Ph.D. Candidate Department of Economics, Hong Kong Baptist University WLB 906, 34 Renfrew Road, KLN, HK Tel.: ; @life.hkbu.edu.hk

2 Prologue Acknowledgement is given to co-authors Prof. Chi-Keung Woo and Dr. Jay Zarnikau. The revised version will appear in the Journal of Energy Markets, March 2016.

3 ey Questions How the investment incentive for natural-gas-fired generation may move with the drivers? Will the investment incentive vanish with rising wind generation? If yes, what can we do to integrate vast wind energy into the grid? An energy-and-capacity market?

4 genda + Overview + Background + Tolling Agreement + Data and Descriptive Analysis + Regression + Key Findings + What s Next? + References + Appendix

5 Over view

6 OVERVIEW A regression based on a large sample of over 134, minute observations in a 46-month period of 01/01/ /31/2014 from Electric Reliability Council of Texas (ERCOT) shows: An increase in ERCOT s non-west regional loads tends to increase the ex post payoffs. A natural-gas price increase enhances the ex post payoffs. Declining nuclear generation tends to increase the ex post payoffs. However, these payoff increases may vanish because of rising wind generation. These findings lend support to a suggestion of ERCOT s eventual transition from an energy-only market to an energy-and-capacity market, so as to mitigate the missing money problem.

7 Back ground

8 BACKGROUND: MARKET REFROM AND INVESTMENT INCENTIVE Market reform has led to competitive wholesale electricity markets with volatile prices. Market restructure has taken place in North America (US: CA, TX, NE, NY, PJM; Canada: ON, AL), South America (Argentina, Brazil, Chile) Europe, Australia, New Zealand, Asia (Singapore). Volatile wholesale market prices with occasional spikes caused by: + daily fuel cost variations, + whether-dependent hourly demands, + planned & forced outages, + carbon price variations, + transmission constraints and + contingencies.

9 BACKGROUND: AREAS OF RESEARCH Electricity finance Spot price behavior and dynamics Pricing of derivatives: forward, futures, and options (call, put, spark spread) Risk management: efficient frontier, value-at-risk, procurement prudence Electricity economics Market design: zonal vs. nodal pricing; energy-only vs. energy & capacity Market power abuse and price manipulation Capacity adequacy and system reliability + Generation investment incentive in an energy-only market mainly comes from high spot prices + What if wind generation suppresses spot prices?

10 BACKGROUND: THE ERCOT MARKET The electricity features 1. Around 85% of electricity demands of Texas are transacted via ERCOT market with annual peak demand of 66,454 MW in 2014.

11 BACKGROUND: THE ERCOT MARKET The electricity features 2. Total installed capacity in 2014 is around 74,000 MW, which is dominated by natural-gas-fired generation (55%).

12 BACKGROUND: THE ERCOT MARKET The electricity features 3. Natural gas is the state s marginal fuel. 4. The state s installed wind capacity in 2014 is more than 12,000 MW, the largest among all states in the U.S. 5. An energy-only market which is claimed to be the most competitive market in the U.S.

13 BACKGROUND: THE ERCOT MARKET Concerns of insufficient generation investment + No major new generation under construction + Postponed projects + Substantial retirement of generation units + Impact of federal environmental regulations upon coal plants planning reserve margin of 9.8%, well below the reserve margin target of 13.75% of peak load

14 BACKGROUND: THE ERCOT MARKET What does the ERCOT data show? + Substantial zero values appear in the payoff of a hypothetical tolling agreement in Houston. + Rising wind generation tends to moderately reduce the payoffs.

15 Tolling Agreement

16 TOLLING AGREEMENT + After making a lease payment to a generator, a buyer obtains the right, but not the obligation, to use the generator s plant to generate electricity. (Similar to a cab driver entering into a longterm lease for a cab) + The buyer may procure its own fuel or pay for the fuel cost based on a market index (e.g., Henry Hub). (Similar to the cab driver paying for gasoline) + A tolling agreement offers a steady known cash inflow to the generator, critical for obtaining low-cost financing. (similar to the cab owner getting a loan from a bank) + The agreement s expected payoff measures the buyer s willingness-to-pay for the lease payment, and hence the incentive for generation investment. (Similar to what the cab driver expects to earn daily)

17 TOLLING AGREEMENT: HOW TO MODEL THE EXPECTED PAYOFF? + A capacity call option gives the buyer the right, but not obligation, to buy electricity at the strike price S (e.g., $100/MWH) The option s payoff = max (P S, 0) + At P = $150/MWH, the payoff is $150/MWH - $100/MWH = $50/MWH; + At P = $90/MWH, the payoff is $0/MWH. + There are simple formulae for computing the option s expected payoff (e.g., Black-Scholes).

18 TOLLING AGREEMENT: HOW TO MODEL THE EXPECTED PAYOFF? + Spark-spread option s payoff = max(p C, 0) + P = spot market price ($/MWH); C = per MWH fuel cost = H G, H = heat rate (MMBTU/MWH), G = spot gas price ($/MMBTU). + Call option with daily-varying strike price since G varies daily. Spark-spread option s payoff + Assuming zero O&M cost, a tolling agreement is a series of 15- minute spark-spread options that continuously expire over the agreement s contract length. + No simple formula to compute the agreement s expected payoff + No formula to compute effect of the payoff drivers + Hence, a regression approach 45 o 45 o 0 Low C High C Market price P

19 TOLLING AGREEMENT: PAYOFF FORMULA + We define the ex post payoff of a hypothetical 1-MW tolling agreement as V = max(p C, 0), C = per MWH fuel cost = H G. + We assume the heat rates (H) of 7 MMBTU/MWH for a new combined cycle gas turbine (CCGT), 9 MMBTU/MWH for a new combustion turbine (CT), and 11 MMBTU/MWH for an old CT. + An OLS regression analysis of the V data likely yields biased estimates because of the series many zero values.

20 TOLLING AGREEMENT: PAYOFF FORMULA + We use an alternative approach proposed by Woo et al. (2016). + Define the per MWH procurement cost of a LDC as: Y = min(p, C). (1) + We can compute the per MWH payoff as: V = P Y. (2) + We can find the marginal payoff effect of a fundamental driver X k : V/ X k = P/ X k - Y/ X k. (3)

21 Data &Descriptive Analysis

22 DATA AND DESCRIPTIVE ANALYSIS Electricity price. We use real time market price (RTM) as it is more volatile. Regional loads are the four major regions 15-minute total loads (MWH). Nuclear generation is the total 15-minute nuclear generation (MWH). Natural-gas price ($/MMBTU) is the daily Henry Hub natural-gas price. $ Wind generation is the total 15-minute wind generation (MWH).

23 DATA AND DESCRIPTIVE ANALYSIS + Regional prices are + volatile, can spike, and can be negative + Wind generation is highly volatile + Nuclear generation is at full capacity when available + Henry Hub prices are volatile and can spike + Loads are volatile and spiky

24 Re gression

25 REGRESSION: SPECIFICATION For region j in ERCOT, equation (4) is the market price regression. Equation (5) is the per MWH procurement cost regression. (4) (5) We use the ITSUR results to estimate the kth driver s marginal payoff effect h k and its variance Var(h k ). (6) (7)

26 REGRESSION: RESULTS A $1/MMBTU increase in natural-gas price tends to increase the payoffs by at least $3.72/MWH in the non-west regions. A 1-MWH increase in the non-west loads tends to raise the payoffs in all regions by at least $0.016/MWH. The same 1-MWH increase in the West regional load tends to reduce the non-west regional payoffs. A 1-MWH increase in nuclear generation tends to reduce the regional payoffs, but its estimated effects are insignificant. A 1-MWH increase in wind generation tends to cut the payoffs by about $0.01/MWH for the non-west regions, approximately half of the estimated effects for the West region.

27 Key Findings

28 KEY FINDINGS All non-west regions are projected to see payoff increases that are over 45% of their average payoff levels, by a $1/MMBTU increase in natural-gas price. We see statistically significant payoff increases in non-west regions caused by a 100-MWH regional load increase. The state s loss of nuclear generation of 1,112 MW can be as much as 88.14% for a new CCGT in the West region (statistically insignificant). The completion of the 5000-MW Mariah project may substantially reduce the payoff levels, ranging from -$3.13/MWH [-44.41%] for an old CT in the Houston region to -$8.3005/MWH [-53.93%] for a new CCGT in the West region.

29 KEY FINDINGS: POLICY IMPLICATIONS Possible remedies are: + A capacity market operated by the independent system operator + Bilateral contracts between wind energy buyers and CCGT and (or) CT owners + Direct ownership of CCGT and CT by LDCs + Back-up capacity requirement imposed on wind generators

30 What s Next?

31 WHAT S NEXT? Our data sample has not seen large changes (e.g., mass retirement of old generation units and large development of wind generation) that can render our regression results inaccurate. + That said, our regression-based approach remains valid and useful, as additional data can be collected in the next few years to update our analysis. Is the missing money problem universally true for all types of renewable generations? + We analyze the impact of rising wind generation on the investment incentive for energy storage device (e.g., compressed air, pumped hydro) a working paper forthcoming in the near future.

32 Re ferences

33 REFERENCES + Alagappan, L., R. Orans, and C.K. Woo (2011). What drives renewable energy development?. Energy Policy, 39(9): Joskow, P. L. (2013). Symposium on Capacity Markets. Economics of Energy and Environmental Policy, 2(2). + Spees, K., Newell, S. A., and Pfeifenberger, J. P. (2013). Capacity Markets-Lessons Learned from the First Decade. Economics of Energy and Environmental Policy, 2(2). + Steggals, W., Gross, R., and Heptonstall, P. (2011). Winds of change: How high wind penetrations will affect investment incentives in the GB electricity sector. Energy Policy, 39(3), Traber, T., and Kemfert, C. (2011). Gone with the wind? Electricity market prices and incentives to invest in thermal power plants under increasing wind energy supply. Energy Economics, 33(2), Traber, T., and Kemfert, C. (2012). German nuclear phase-out policy: Effects on European electricity wholesale prices, emission prices, conventional power plant investments and electricity trade (No. 1219). Discussion Papers, German Institute for Economic Research, DIW Berlin. Available from: + Woo, C. K., Horowitz, I., Horii, B., Orans, R., and Zarnikau, J. (2012). Blowing in the wind: Vanishing payoffs of a tolling agreement for natural-gas-fired generation of electricity in Texas. The Energy Journal, 33(1), Woo, C.K., Horowitz, I., Zarnikau, J., Moore, J., Schneiderman, B., Ho, T., and Leung, E. (2016). What Moves the Ex Post Variable Profit of Natural-Gas-Fired Generation in California? The Energy Journal, 37(3),

34 App endix

35 APPENDIX Proof for equation (2) When P> C, Y = min(p, C) = C, implying P Y = V = max(p C, 0) = P C > 0. When P C, Y = min(p, C) = P, implying P - Y = V = max(p C, 0) = P - P = 0. Computation for the state s 15-minute total wind generation increase caused by the completion of the Mariah project 406-MWH = 5000 MW 0.33 capacity factor (15 minutes / 60 minutes) 0.33 capacity factor : the state s average annual wind generation (MWH) in the sample period divided by [8760 hours the state s average annual wind capacity (MW) in the sample period].

36 APPENDIX: REGRESSION OUTPUT

37 APPENDIX: REGRESSION OUTPUT

38 APPENDIX: MARGINAL PAYOFF EFFECT

39 APPENDIX: PAYOFF CHANGES

40 Yun LIU, Ph.D. Candidate, HKBU With over 10 published articles in professional journals/newspapers such as the Financial News, he was: (a) a part-time external analyst of the China Elite Risk Management and Actuarial Company; (b) a co-founder of the Green Orange Culture Company Limited; (c) a council member of (i) the Hong Kong Jiangsu Youth Association and (ii) the New Hong Kong Youth Association and (iii) the Nantong Natives (HK) Association; and (e) a Deputy Secretary General of the Rugao Natives (HK) Association. The Author

41 Q&A Thanks for your attention