Capacity Expansion in Competitive Electricity Markets

Transcription

1 Capacity Expansion in Competitive Electricity Markets Efthymios Karangelos University of Liege November 2013

2 In the past... Vertical Integration A single utility owned & operated all power system infrastructure. Least cost capacity related decisions: 1. Investment & Retirement of Generating Units. 2. Network Reinforcement.

3 Nowadays... Competition Several actors own & operate parts of the system infrastucture. Decentralized, profit seeking capacity related decisions.

4 Overview Part I: Generation capacity expansion from the generators perspective 1. Discounted cash flow method. 2. Real options analysis. Part II: Generation capacity expansion from the consumers perspective 1. Long-term supply security in competitive markets. 2. Capacity reward mechanisms. Part III: Transmission capacity expansion 1. Cost-based model. 2. Merchant model.

5 Part I Generation Capacity Expansion The Generators Perspective

6 GenCo Long-term Perspective Generators need to invest in order to participate in the short-term electricity market.

7 Investment Life Cycle Construction Projected Lifetime Retirement 1. Construction Period (1 5yr) Planning & Permitting. Construction. 2. Projected Lifetime (1 30yr) Production & Sales. Maintenance. 3. Retirement Period Asset Liquidation.

8 Investment Cash Flows Decision Cost & Revenue Realization Construction Projected Lifetime Retirement Decision Making Problem The investor needs to compare in advance costs and revenues to be realized at different points in the future.

9 Discounted Cash Flow Method Principle Refer all cost and revenue cash flows to the decision making point by discounting the future according to: 1. The time value of money... Any amount of money has a greater worth the sooner it is collected (due to the potential to earn an interest). 2. The value of certainty... Any future amount of money has a greater worth if it is to be collected with a greater certainty.

10 Discounted Cash Flow Method Concepts Present Value (PV ) of a cash flow (CF) to be collected at a future year t is the discounted worth of this amount at the present point: PV (F, t) = CF (1 + ρ) t Risk-free Discount Rate (ρ), i.e. interest rate, is the annual percentage change in the value of money. Risk-adjusted Discount Rate (ˆρ) is the risk-free interest rate plus a security premium, increasing with the level of the cash flow uncertainty (ˆρ = ρ + ρ s ).

11 Discounted Cash Flow Method Concepts Net Present Value (NPV ) of an investment opportunity (i) is the difference between the present value of the associated revenues (R i,t ) and costs (C i,t ): t=t c+t L NPV (i, T c, T L ) = C i,0 t=tc+1 R i,t C i,t (1 + ρ) t Message V TL + (1 + ρ) T C+T L NPV shows the value of an opportunity over it s life cycle, referred at the decision making point.

12 Discounted Cash Flow Method Concepts Net Present Value (NPV ) of an investment opportunity (i) is the difference between the present value of the associated revenues (R i,t ) and costs (C i,t ): Rule t=t c+t L NPV (i, T c, T L ) = C i,0 t=tc+1 V TL + (1 + ρ) T C+T L R i,t C i,t (1 + ρ) t Investment opportunity i is profitable if NPV (i, T c, T L ) > 0

13 Discounted Cash Flow Method Example Investment in a new coal-fired power plant Technical Characteristics G i = 500MW ui,g = 0.8

14 Discounted Cash Flow Method Example Investment in a new coal-fired power plant Technical Characteristics G i = 500MW ui,g = 0.8 Construction Data (year 0) ci,0 = 1021$/kW Lifetime Data (years 1 30) c i,g = 11.77$/MWh π = 32$/MWh ρ = 8% Retirement Value Vi = 5 million $

15 Discounted Cash Flow Method Example Investment in a new coal-fired power plant Technical Characteristics G i = 500MW ui,g = 0.8 Construction Data (year 0) ci,0 = 1021$/kW Lifetime Data (years 1 30) c i,g = 11.77$/MWh π = 32$/MWh ρ = 8% Retirement Value Vi = 5 million $ Investment Cost C i,0 = c i,0 G i 1000 = $

16 Discounted Cash Flow Method Example Investment in a new coal-fired power plant Technical Characteristics G i = 500MW ui,g = 0.8 Construction Data (year 0) ci,0 = 1021$/kW Lifetime Data (years 1 30) c i,g = 11.77$/MWh π = 32$/MWh ρ = 8% Retirement Value Vi = 5 million $ Investment Cost C i,0 = c i,0 G i 1000 Annual Cost = $ C i,t = u i,g G i c i,g 8760h/y = $ /y

17 Discounted Cash Flow Method Example Investment in a new coal-fired power plant Technical Characteristics G i = 500MW ui,g = 0.8 Construction Data (year 0) ci,0 = 1021$/kW Lifetime Data (years 1 30) c i,g = 11.77$/MWh π = 32$/MWh ρ = 8% Investment Cost C i,0 = c i,0 G i 1000 Annual Cost = $ C i,t = u i,g G i c i,g 8760h/y = $ /y Annual Revenue Retirement Value Vi = 5 million $ R i,t = u i,g G i π 8760h/y = $ /y

18 Discounted Cash Flow Method Example Investment in a new coal-fired power plant t C i,t ($) R i,t ($) PV {R i,t C i,t } ($) NPV = millions of $

19 Discounted Cash Flow Method Example Investment in a new coal-fired power plant Should we invest? NPV = millions of $

20 Discounted Cash Flow Method Example Investment in a new coal-fired power plant Should we invest? NPV = millions of $ Or... keep our initial capital (510.5 millions of $) in the bank? invest on a different opportunity?

21 Discounted Cash Flow Method Internal Rate of Return (IRR) of an investment opportunity (i) is the discount rate that equates the present value of the future costs and revenues to the initial investment costs: IRR = ρ : NPV (i, T c, T L ) = 0 Message IRR shows the rate of growth the investment opportunity is expected to generate over it s lifecycle. IRR allows the relative evaluation of alternatives.

22 Discounted Cash Flow Method Example Investment in a new coal-fired power plant vs money in the bank ρ (%) NPV ( millions of $)

23 Discounted Cash Flow Method Example Investment in a new coal-fired power plant vs money in the bank ρ (%) NPV ( millions of $) Keeping the initial capital in the bank is more profitable if the banks offers an interest greater than %

24 Discounted Cash Flow Method Example Investment in a new coal-fired power plant vs money in the bank ρ (%) NPV ( millions of $) Keeping the initial capital in the bank is more profitable if the banks offers an interest greater than % Minimum Acceptable Rate of Return (MARR) A firm will decide to invest in a project if it yields an IRR greater than the firm s MARR (e.g. interest on bank deposit).

25 Discounted Cash Flow Method Example Investment in a new coal-fired power plant vs combined cycle gas turbine Technical Characteristics G1 = 500MW u 1,g = 0.8 Construction Data (year 0) c1,0 = 1021$/kW Lifetime Data (years 1 30) c1,g = 11.77$/MWh π = 32$/MWh ρ = 8% Technical Characteristics G2 = 500MW u 2,g = 0.8 Construction Data (year 0) c2,0 = 533$/kW Lifetime Data (years 1 30) c2,g = 20.78$/MWh π = 32$/MWh ρ = 8% Retirement Value V 1 = 5 million $ Retirement Value V 2 = 5 million $

26 Discounted Cash Flow Method Example Investment in a new coal-fired power plant vs combined cycle gas turbine ρ (%) NPV 1 ( millions of $) NPV 2 ( millions of $) < 0 0 At the assumed risk-free discount rate the coal-fired plant has a greater NPV! The combined cycle gas turbine has a greater IRR!

27 Effect of Cyclical Demand So far, we have been thinking that... The plant will run on a given utilization factor throughout it s lifetime. The investor will collect a constant price of electricity throughout the lifetime of the plant. What about seasonal variations in the demand and price of electricity?

28 Effect of Cyclical Demand Load Duration Curve (Annual) x-axis refers to a percentage of time. y-axis refers to a demand level. A point on the curve shows the percentage of periods in a year that the demand is greater than a certain level.

29 Effect of Cyclical Demand MWh 100 time (%) Wh /MW MWh

30 Effect of Cyclical Demand MWh MWh 100 time (%) MWh /MW Wh MWh

31 Effect of Cyclical Demand MWh MWh 100 time (%) MWh /MW Wh /MW Wh 100 time (%) MWh

32 Effect of Cyclical Demand MWh MWh 100 time (%) MWh /MW Wh /MW Wh 100 time (%) MWh

33 Effect of Cyclical Demand A point on the annual price duration curve shows the percentage of periods in a year that the demand is greater than a certain level. Projection of short-run cost determines the plant utilization factor. The integral of area R g is the annual net revenue from the operation of the plant.

34 Effect of Uncertainty So far, we have been thinking that... The plant will run on a constant marginal cost throughout it s lifetime. What if the fuel price is not known with certainty?

35 Effect of Uncertainty So far, we have been thinking that... The plant will run on a constant marginal cost throughout it s lifetime. What if the fuel price is not known with certainty? Sensitivity Analysis Calculation of NPV & IRR for different values of the marginal cost within the uncertainty range.

36 Sensitivity Analysis - Example Investment in a new coal-fired power plant for different fuel cost values Sensitivity plots for NPV & IRR. A steeper NPV curve shows a riskier project. A more variable IRR curve shows a riskier project.

37 Risk-adjusted Discount Rate Principle Risk-free interest rate plus a security premium: ˆρ = ρ + ρ s ρ s (%) NPV ( millions of $) Practical Challenge How to calculate the correct security premium for each uncertainty?

38 Effect of Uncertainty So far, we have been thinking that... The price of electricity throughout the lifetime of the plant is known with certainty. What if there is more than one future scenario?

39 Effect of Uncertainty So far, we have been thinking that... The price of electricity throughout the lifetime of the plant is known with certainty. What if there is more than one future scenario? Real Option Analysis Investment opportunity as an option which can be used now (under uncertainty) OR in the future (with better information).

40 Real Options Analysis Example Investment under uncertainty in a new coal-fired power plant Technical Characteristics G i = 500MW ui,g = 0.8 Construction Data (year 0) ci,0 = 1021$/kW Retirement Value Vi = 5 million $ Lifetime Data ci,g = 11.77$/MWh ρ = 8% (years 1 2) π = 32$/MWh (years 3 30) p 1 = 40% π 1 = 32$/MWh p 2 = 60% π 2 = 20$/MWh

41 Real Options Analysis Example Investment under uncertainty in a new coal-fired power plant Scenario 1 (p 1 = 40%) t C i,t ($) R i,t ($) PV {R i,t C i,t } ($) NPV = millions of $

42 Real Options Analysis Example Investment under uncertainty in a new coal-fired power plant Scenario 2 (p 2 = 60%) t C i,t ($) R i,t ($) PV {R i,t C i,t } ($) NPV = millions of $

43 Real Options Analysis Example Investment under uncertainty in a new coal-fired power plant R i,t = p 1 R 1 i,t + p2 R 2 i,t t C i,t ($) R i,t ($) PV {R i,t C i,t } ($) Expected NPV = 48.6 millions of $

44 Real Options Analysis Example Investment under uncertainty in a new coal-fired power plant Should we invest? Expected NPV = 48.6 millions of $

45 Real Options Analysis Example Investment under uncertainty in a new coal-fired power plant Should we invest? Expected NPV = 48.6 millions of $ Real Option Analysis What if we keep the opportunity to invest and decide after year 3? Is it better to invest after year 3, only if the electricity price stays at 32$/MWh?

46 Real Options Analysis Example Investment under uncertainty in a new coal-fired power plant Investment at year 3 if scenario 1 becomes true... t C i,t ($) R i,t ($) PV {R i,t C i,t } ($) Expected NPV = 86.5 millions of $

47 Real Options Analysis Example Investment under uncertainty in a new coal-fired power plant Investment NPV (millions of year 0 without uncertainty year 0 under uncertainty year 3 if scenario 1 is true 86.5 Option Value (The opportunity to invest is worth... ) V O =max{48.6, 86.5, 0} = 86.5 millions of $ Value of Waiting to Invest (Not investing now is worth... ) V W =max{ , 0} = 37.9 millions of $ Cost of Uncertainty (Not knowing the future costs... ) NPV p 1 =100% V O = millions of $

48 Real Options Analysis Extension Principle An investment opportunity can be evaluated as a financial call option. Financial Call Option The holder has the right to buy a commodity in the future at a given price. Option Fee: The cost of getting the right to buy. Exercise Fee: The price at which the commodity can be bought in the future. The option value is the difference between the expected future value of the commodity and the option + exercise fees.

49 Real Options Analysis Extension Financial Call Option Evaluation Model the dynamics of the commodity value in detail: 1. Optimal option capturing strategy (time, option & exercise fees). 2. Optimal time to exercise an option. Application to the GenCo s Problems Model the dynamics of the costs and revenues within the life cycle of a plant in detail: 1. Optimal investment strategy (type & capacity of plant, time, location.) 2. Optimal retirement strategy for an existing plant. Analytical focus on uncertainty origins and impacts.

50 Summary Decision Making Criteria Value of an investment over its life cycle. Risk on investment profitability. Net Present Value Analysis Projection of future cash flows at the decision making point. Qualitative consideration of investment risks. Real Options Analysis Detailed modeling of the cash flow evolution. Quantitative consideration of investment risks.

51 Part I Questions?

52 Part II Generation Capacity Expansion The Consumers Perspective

53 The Consumers Perspective The consumers demand capacity for a secure & reliable supply of electricity.

54 The Regulator s Perspective Objective The regulator must ensure that the market framework encourages the generators to invest & meet the consumers expectations. In theory... The operation a short-term competitive market should result in optimal investment decisions. In practice... Will the generators invest to secure the supply of electricity on a long-term basis?

55 A Theoretical Result When there is excess supply... The price of the short-term market is set at the marginal cost of the most expensive generating unit. Infra-marginal generators receive a premium in addition to their marginal costs (scarcity rent).

56 A Theoretical Result When there is excess demand... The price of the short-term market is set at the value of lost load. Every generator receives a scarcity rent.

57 A Theoretical Result In the long term... The marginal costs of every generating unit are completely recovered through the electricity market price. A rational investor should invest up to the point where the net revenues (area R g ) are equal to the investment and fixed operational costs.

58 A Theoretical Result In the long term... /MWh voll peaking mid merit base load 100 time (%) If an existing generator is not needed, it will be non-profitable and its owner should decide to retire it. If a new generator is needed, there will be a profitable investment opportunity. The system capacity should converge to an optimal mix.

59 What could go wrong? 1. Demand-side Participation Most electricity consumers do not participate in the short-term market. Voll is not common for different consumers, times of the day etc... The system capacity mix relies strongly on a statistical estimation of the average voll.

60 What could go wrong? 1. Demand-side Participation Most electricity consumers do not participate in the short-term market. Generators have opportunities for capacity withholding and strategic bidding (e.g. California Crisis).

61 Capacity Withholding If a generator reduces the capacity it makes available in the short-term market...

62 Capacity Withholding If a generator reduces the capacity it makes available in the market, it may increase its profit through pushing the price to the level of the voll.

63 Strategic Bidding The most expensive generator can only be outplaced by the voll.

64 Strategic Bidding The most expensive generator can only be outplaced by the voll. As long as it remains below the voll, it can increase its bid price above its marginal cost.

65 What could go wrong? 1. Demand-side Participation Most electricity consumers do not participate in the short-term market. Generators have opportunities for capacity withholding and strategic bidding (e.g. California Crisis). A regulator needs to apply a price cap on the short-term market to protect the consumers.

66 What could go wrong? 2. Regulatory Price Caps /MWh voll price cap cap peaking mid merit base load The price cap applies even when there is an actual capacity shortage. The market price never reaches the voll. 100 time (%) Investors face a missing money problem.

67 What could go wrong? 3. Ancillary & Balancing Services System operators procure ancillary & balancing services through out-of-market arrangements. System operators use these services according to technical rules. While the system is stressed the market price does not indicate the need for more capacity.

68 What could go wrong? 4. Investment Uncertainty /MWh Price variation due to wind - January 6 strategic firms Hour Maximum 90th percentile 75th percentile Median 25th percentile 10th percentile Minimum The profitability of generation capacity is uncertain as it depends on rare & extreme price levels. Missing an investment opportunity is better than risking to suffer a loss. Risk-averse investors tend to under-invest in generation capacity.

69 What could go wrong? 5. Power Plant Capacity Standardization Power plant can be built on standard capacity ratings. The optimal capacity expansion may not be a technically feasible option. Less than the optimal capacity risks the security of the supply. More than the optimal capacity risks the profitability of the investment.

70 The Regulator s Solution In the interest of the long-term supply security... Wh MW Security Margin A net generation capacity surplus should be maintained. Investment decisions should remain private but also become coordinated. 100 time (%) Generators should be encouraged to maintain a net capacity surplus through a capacity reward mechanism.

71 Capacity Reward Principle Electricity Consumers Energy Supply Risk Financial Security Security of Supply Investor Financial Risk A regulator should reward investors according to the value the consumers place on the firm commitment to develop a project.

72 Capacity Reward Mechanisms 1. Strategic Reserve e.g. Finland, Norway 2. Capacity Payment e.g. Argentina, Italy, Spain (past) 3. Capacity Market e.g. New England, PJM, Brazil, Colombia

73 Strategic Reserve Wh MW Strategic Reserve 100 time (%) Principle The system operator (on behalf of the regulator) reserves in advance & controls a part of the system s generation capacity.

74 Strategic Reserve In the long-term... The system operator centrally controls the system security margin (MW). The generators that engage in this scheme receive an additional certain revenue (euro/mw).

75 Strategic Reserve In the long-term... The system operator centrally controls the system security margin (MW). The generators that engage in this scheme receive an additional certain revenue (euro/mw). In the short-term... The system operator can make the reserved capacity available in the electricity market, or... The system operator can decide to use the reserved capacity in the case of an emergency. The consumers pay for the additional cost of the reserved capacity (euro/mwh).

76 Strategic Reserve Main Benefits The system operator can contract for the resources that are needed to protect the supply. The withdrawal of the reserved capacity from the market enhances investment profitability without endangering the supply. Criticism The use of the reserved capacity by the system operator may distort the operation of the short-term market. The capacity reward is only available to the generators that engage in the reserve procurement.

77 Capacity Payment Wh MW Security Margin 100 time (%) Principle The regulator offers a payment per MW of installed generation capacity to make investments more attractive and thus stimulate the development of a security margin.

78 Capacity Payment Cost-based Payment Method Payment level is defined as a portion of the investment & fixes operational costs of a power plant. Reliability-based Payment Method Payment level is defined hourly as the difference between the voll and the short-term market price, weighted by the outage probability. π r = lolp t (voll µ t )

79 Capacity Payment Main Benefit Profitability of generation capacity becomes controllable/ more certain when needed. Criticism No obligation on the generators to provide energy via the rewarded capacity. No association between the cost-based payment & the generator s contribution to the supply security. Generators have an opportunity to increase the level of the reliability-based payment through capacity withholding.

80 Capacity Market Wh MW Security Margin 100 time (%) Principle The regulator defines the desired security margin and creates a long-term forward market where generators compete to collect capacity revenues.

81 Capacity Market At the forward point... (e.g today, for ) The regulator awards a capacity credit per MW to every generator that is expected to be operational (existing & in construction today). Generators can sell these credits through a centralized market in exchange for: 1. a direct payment, 2. the responsibility to provide energy or reserve when instructed, 3. a cap on the future short-term market revenues.

82 Capacity Market The Capacity Demand Curve MW /M target Security Margin (%) The reward at the targeted margin is the expected net cost of a reference new power plant. The credit value gradually decreases (increases) above (below) the targeted security margin. Generators do not have an opportunity to manipulate the level of the capacity reward.

83 Capacity Market In the long-term... Generators receive the capacity value (euro/mw) through the forward auction. The consumers pay the regulator ex-post. In the short-term... The market revenue of the generators that receive a capacity reward is capped. The generators that receive a capacity reward face a penalty charge for failing to deliver energy or reserve when instructed.

84 Capacity Market /MWh voll price cap cap Capacity Value + Market Revenue Unavailability Penalty The GenCo s Perspective 100 time (%) In spite of the price cap cost recovery is guaranteed for a generator that is available when needed. A generator has an incentive to avoid the unavailability penalties.

85 Capacity Market Main Benefit Generators are rewarded for actually securing the supply. Consumers pay for the security level they receive. Criticism More suitable for centralized (pool) short-term markets. Complexity of availability measurement.

86 Summary Long-term Security of Supply Question The regulator should implicitly coordinate investment decisions through the electricity market rules. Capacity Reward Mechanisms Provide additional investment incentives to enhance the long-term security of the electricity supply. All mechanisms have different advantages & disadvantages. The forward capacity market appears more favorable.

87 Part II Questions?

88 Part III Transmission Capacity Expansion

89 In the past... Vertical Integration Transmission capacity expansion followed the evolution of the supply & demand.

90 Nowadays... Competition Transmission capacity expansion is an independent process.

91 The Transmission Business A service The delivery of electrical energy with certain quality & security from the point of production to the point of consumption. A natural monopoly No rationale for the existence of competing transmission lines along the same route. An impartial role Focus on the global economic welfare of the market actors (equitable treatment of generators and consumers).

92 Investing on the Transmission System Capital Intensive Assets Lines of long distance. Transformers & switchgear. Control & communication infrastructure. Life Time & Irreversibility Expected life time in the order of years. Insignificant value of retiring/reselling assets.

93 Investing on the Transmission System Standardized Assets Voltage & capacity of transmission equipment comes in standard ratings. Utilization should be expected to evolve throughout the asset life time. Economies of Scale Main cost component is line length rather than line rating. A line of greater capacity has a smaller investment cost per MW.

94 Models of Transmission Expansion Cost-based The transmission company invests under the approval of the regulator and receives a regulated, risk-free return on investment. Merchant (market-driven) The transmission company invests in expectation of the future market value of transmission capacity. The transmission company undertakes the risk of not recovering its investment costs.

95 Cost-based Expansion 1. Network expansion is planned by the transmission company according to the projected future needs.

96 Cost-based Expansion 1. Network expansion is planned by the transmission company according to the projected future needs. 2. The transmission company submits an expansion plan to the regulator.

97 Cost-based Expansion 1. Network expansion is planned by the transmission company according to the projected future needs. 2. The transmission company submits an expansion plan to the regulator. 3. The regulator reviews & possibly accepts the submitted plan.

98 Cost-based Expansion 1. Network expansion is planned by the transmission company according to the projected future needs. 2. The transmission company submits an expansion plan to the regulator. 3. The regulator reviews & possibly accepts the submitted plan. 4. If the plan is accepted, the regulator defines the net profit of the transmission company.

99 Cost-based Expansion 1. Network expansion is planned by the transmission company according to the projected future needs. 2. The transmission company submits an expansion plan to the regulator. 3. The regulator reviews & possibly accepts the submitted plan. 4. If the plan is accepted, the regulator defines the net profit of the transmission company. 5. The transmission company undertakes the accepted expansion plan.

100 Cost-based Expansion 1. Network expansion is planned by the transmission company according to the projected future needs. 2. The transmission company submits an expansion plan to the regulator. 3. The regulator reviews & possibly accepts the submitted plan. 4. If the plan is accepted, the regulator defines the net profit of the transmission company. 5. The transmission company undertakes the accepted expansion plan. 6. The transmission company collects its revenue via regulated use-of-network charges.

101 Use-of-network charges Scope Allocate the cost of the transmission infrastructure to the network users (producers & consumers). Ensure that the transmission company collects the regulated return on investment. Cost Allocation Alternatives 1. Postage Stamp Method 2. Contract Path Method 3. MW-mile Method

102 Postage Stamp Method Principle Charge network users according to their average use of the whole network. e.g. Allocate the regulated revenue of the transmission company to the generators (consumers) according to the annual energy produced (consumed). Charges do not depend on the location of connection.

103 Postage Stamp Method Advantages Simplicity. Disadvantages Allocation does not reflect differences in the actual network use. Remote transmission costs passed on to central network users. Allocation can only reflect transmission costs at a regional level.

104 Contract Path Method Principles Define an assumed route of the electricity flow between the points of production and consumption (contract path) for every transaction. Apply a per MW charge to every transaction for the use of a contract path. Charge generators and consumers according to the cost of the transmission infrastructure along the respective contract path.

105 Contract Path Method Advantages Allocation is more reflective of the use of the network. Locational differences are taken into consideration. Disadvantages Relevance between the contract path and the actual electrical path (according to Kirchoff s laws) is questionable. Allocation more suitable for long-term bilateral agreements.

106 MW-mile Method Principles Define a per MW-mile cost of the transmission network. Determine the actual electrical path of every transaction via power flow calculation methods. Charge generators and consumers according to the energy volume and the electrical path corresponding to every transaction.

107 MW-mile Method Advantages Allocation is based on Kirchoff s laws for the use of a network. Disadvantages Complexity (power flow is non-linear). Electrical paths are not independent, calculation relies on the adopted transaction sequence.

108 Merchant Expansion The transmission company... Invests in expectation of the future market value of transmission capacity. Undertakes the risk of not recovering its investment costs.

109 The Market Value of Transmission Capacity INVESTING IN TRANSMISSION 20 $/MWh A B 1000 MW 45 $/MWh G 1 G 2 Figure MW Simple example illustrating the value of transmission A transmission line in a competitive setting... on its own. Finally, we assume that the capacity of the transmission line is sufficient to support any power transmission that may be required. The consumers at bus B can either buy energy at 45 $/MWh from the local generator G 2 or buy energy at 20 $/MWh from the remote generator G 1 and pay for the transmission of generator this energy. If rather cost of than this transmissiona local, more is less than expensive 25 $/MWh, one. consumers will choose to buy their energy from G 1 because the overall cost would be less than Allows the 45 $/MWh a remote that theycheaper would havegenerator pay buy energy to compete from generator withg 2 a. local, It is thusmore not in expensive the best interestone. of the owner of the transmission line to charge more than 25 $/MWh because such a charge would discourage consumers from making use of the transmission system. In this example, the value of the transmission service is thus 25 $/MWh because, at that price, consumers are indifferent between using and not using transmission. The value of transmission is thus a function of the short-run marginal cost of generation. In this case, this function is very simple because there is Allows consumers to buy electricity from a remote, cheaper

110 Example (Borduria - Syldavia interconnection) 8.4 VALUE-BASED TRANSMISSION EXPANSION 235 Borduria P B Syldavia P S D B = 500 MW D S = 1500 MW Figure 8.2 Model of the Borduria/Syldavia interconnection Borduria The demands in Borduria and Syldavia are respectively 500 MW and 1500 MW. We continue to assume that these demands do not vary with time and are perfectly inelastic. In the absence of an interconnection, the two national electricity markets operate independently and the prices in Borduria and Syldavia are respectively 15 $/MWh and 43 $/MWh. The value of transporting the first megawatt-hour from Borduria to Syldavia is thus equal to the difference in price between the two countries, that is, 28 $/MWh. We saw in Chapter 6 that when the flow through the interconnection is 400 MW, generators in Borduria produce 900 MW. 500 MW of this production is for the local load, while the remaining 400 MW is sold to consumers in Syldavia. The remaining 1100 MW of Syldavian load is produced locally. Under these conditions, the prices in Borduria and Syldavia are 19 $/MWh and 35 $/MWh respectively. The value of transporting one additional megawatt-hour from Borduria to Syldavia is thus only 16 $/MWh. This is also the maximum price that consumers in Syldavia would agree to pay for the transport of a megawatt-hour that they have bought in Borduria for π 19 $/MWh. If the price of transmission S π B = 28 $/MWh were any higher, they would prefer to buy this megawatt-hour from local generators. When the flow on the interconnection reaches MW, the prices in Borduria D B = 500MW π B = P B $/MWh = 15 $/MWh Syldavia D S = 1500MW π S = P S $/MWh = 43 $/MWh The marginal value of transmission capacity is...

111 Example (Borduria - Syldavia interconnection) 8.4 VALUE-BASED TRANSMISSION EXPANSION 235 Borduria P B Syldavia P S D B = 500 MW D S = 1500 MW Figure 8.2 Model of the Borduria/Syldavia interconnection The demands in Borduria and Syldavia are respectively 500 MW and 1500 MW. We continue to assume that these demands do not vary with time and are perfectly inelastic. In the absence of an interconnection, the two national electricity markets operate independently and the prices in Borduria and Syldavia are respectively 15 $/MWh and 43 $/MWh. The value of transporting the first megawatt-hour from Borduria to π Syldavia is thus equal to the difference B = π in price S between the two countries, that is, 28 $/MWh. We saw in Chapter 6 that when the flow through the interconnection is 400 MW, generators in Borduria produce 900 MW. 500 MW of this production is for the local load, while the remaining 400 MW is sold to consumers in Syldavia. The remaining 1100 MW of Syldavian load is produced locally. Under these conditions, the prices in Borduria and Syldavia are 19 $/MWh and 35 $/MWh respectively. The value of transporting one additional megawatt-hour from Borduria to Syldavia is thus only F = MW 16 $/MWh. This is also the maximum price that consumers in Syldavia would agree to pay for the transport of a megawatt-hour that they have bought in Borduria for 19 $/MWh. If the price of transmission were any higher, they would prefer to buy this megawatt-hour from local generators. When the flow on the interconnection reaches MW, the prices in Borduria and Syldavia are equal: If we had a line of sufficiently large capacity (D B + F) = (D S F ) { πb = π S = 24.3$/MWh The marginal value of transmission capacity would be zero!

112 Example (Borduria - Syldavia interconnection) 8.4 VALUE-BASED TRANSMISSION EXPANSION 235 Borduria P B Syldavia P S D B = 500 MW D S = 1500 MW Figure 8.2 Model of the Borduria/Syldavia interconnection The demands in Borduria and Syldavia are respectively 500 MW and 1500 MW. We continue to assume that these demands do not vary with time and are perfectly inelastic. In the absence of an interconnection, the two national electricity markets operate independently and the prices in Borduria and Syldavia are respectively 15 $/MWh and 43 $/MWh. The value of transporting the first megawatt-hour from Borduria to Syldavia is thus equal to the difference in price between the two countries, that is, 28 $/MWh. We saw in Chapter 6 that when the flow through the interconnection is 400 MW, generators in Borduria produce 900 MW. 500 MW of this production is for the local load, while the remaining 400 MW is sold to consumers in Syldavia. The remaining 1100 MW of Syldavian load is produced locally. Under these conditions, the prices in Borduria and Syldavia are 19 $/MWh and 35 $/MWh respectively. The value of transporting one additional megawatt-hour from Borduria to Syldavia is thus only 16 $/MWh. This is also the maximum price that consumers in Syldavia would agree to pay for the transport of a megawatt-hour that they have bought in Borduria for 19 $/MWh. If the price of transmission were any higher, they would prefer to buy this megawatt-hour from localπgenerators. S π B = 16 $/MWh When the flow on the interconnection reaches MW, the prices in Borduria and Syldavia are equal: If we had a line of limited capacity (e.g. F max = 400MW ) π B = (D B + F max ) = 19$/MWh π S = (D S F max ) = 35$/MWh The marginal value of transmission capacity would be...

113 The Market Value of Transmission Capacity Transmission Demand Function π T (F) Expresses the marginal value of transmission as a function of the flow. e.g. Borduria - Syldavia interconnection π T (F) = π S (F) π B (F) = = (D S F) [ (D B + F)] = = F $/MWh

114 The Market Value of Transmission Capacity Transmission Revenue Function R T (F) Expresses the revenue of the transmission company as a function of the flow. e.g. Borduria - Syldavia interconnection R T (F ) = π T (F) F = = ( F) F $

115 The Market Value of Transmission Capacity Annuitized Transmission Cost Function C T (F) C T (F) = C 0 + k l F where C 0 : Annuitized fixed transmission capacity cost ($ per year) k: Annutized marginal transmission capacity cost ($/MW km year) l: Line length (km)

116 Optimal Transmission Capacity Annuitized Transmission Capacity Value V T (F) V T (F) = R T (F) 8760 (hr/year) C T (F) Optimal Transmission Capacity (F ) F {R T (F) 8760 C T (F)} : = 0 ( F πt (F ) ) 8760 F + π T (F ) = k l F F

117 Example (Borduria - Syldavia interconnection) Optimal Transmission Capacity { k = 35 } $/(MW km year) l = 100 km F = 800 πb = 23 πs = 27 MW $/MWh $/MWh Annual Cost & Revenue R T (F ) = (π S (F ) π B (F )) F = 28 millions of $/year C T (F ) = C 0 + k l F = C millions of $/year At the optimal capacity the investor would not recover its fixed transmission capacity costs!

118 The Regulator s Solution Transmission Capacity Uplift (π u ) R T (F ) = (π T (F) + π u ) F Principle Support Recover the fixed investment costs of merchant capacity expansion through an additional regulated charge. Transmission capacity expansion increases the net welfare of the electricity system stakeholders.

119 Transmission Constraints & Welfare Losses Merit Order Dispatch The lowest net electricity cost (maximum global welfare) solution to the electricity supply problem. e.g. Borduria - Syldavia interconnection {F MW P B = MW, P S = 566.6MW } C(P B ) = 10 P B P2 B $/h C(P S ) = 13 P S P2 S $/h C M = $/h Can only happen in a transmission network with infinite capacity!

120 Transmission Constraints & Welfare Losses Out of Merit Generation Cost (Cost of Constraints) The difference between the constrained optimal solution and the unconstrained merit order dispatch. e.g. Borduria - Syldavia interconnection {F = 800MW P B = 1300MW, P S = 700MW } C = $/h Hourly cost: C = C M C = 267 $/h

121 Example (Borduria - Syldavia interconnection) 8.4 VALUE-BASED TRANSMISSION EXPANSION Cost ($/h) 8000 Total cost 4000 Investment cost Cost of constraints Transmission capacity (MW) Figure 8.6 Evolution of the cost of constraints, the investment cost and the total transmission cost for the Borduria Syldavia interconnection At the optimal solution the total cost of transmission is minimized. The total cost of transmission is the sum of the cost of building the transmission system and the cost of constraints. As Figure 8.6 shows, the cost of building the transmission system increases with the transmission capacity while the cost of constraints

122 Capacity Expansion & Welfare Gain Gain of Capacity Expansion The difference between the constrained optimal solution and the case where the additional transmission capacity is not built. e.g. Borduria - Syldavia interconnection { F N = 0MW P B = 500MW, P S = 1500MW } Hourly gain: C N = $/h C N = C N C = $/h

123 Effect of Cyclical Demand MWh 100 time (%) Generalization Optimal transmission capacity must be calculated considering the duration of different load levels. Optimal transmission capacity would impose constraints only on a limited period of time.

124 Effect of Cyclical Demand Annuitized Transmission Capacity Value V T (F) N t V T (F) = RT t (F) Ht C T (F) t=1 Optimal Transmission Capacity (F ) F : { N t t=1 Rt T (F) Ht C T (F)} = 0 F N t ( π H t t T (F ) ) F + πt t F (F ) = k l F t=1 where: t: Type of period (e.g. peak/off-peak) H t : Duration of period t (hr/year)

125 Duration Example (BorduriaFigure - Syldavia 8.8 Simplified load-duration interconnection) curve 2700 D B (MW) 900 D S (MW) Figure Duration (h) Optimal Transmission Capacity { k = 140 $/(MW km yr) l = 100 km Duration (h) Load-duration curves for (a) Borduria and (b) Syldavia Table 8.1 Unconstrained economic dispatch for the peak and off-peak load conditions in the Borduria Syldavia system Load Generation } Generation FTotal hourly (MW) in Borduria in Syldavia = 400 generation π,p (MW) (MW) cost B ($/h) π ,p S π,o B = π,o S MW = 23 $/MWh = 27 $/MWh = 23 $/MWh Revenue only collected during peak hours!

126 optimization procedure has been applied to the IEEE 24 bus reliability test sys- RTS) Indepicted a network in Figure Forwith the detailsmore of this network, than see IEEE 2(1979). nodes... e 8.15 shows that, except for a small number of lines, line flows are well below BUS 17 BUS 18 BUS 21 BUS 22 BUS 23 BUS 16 BUS 19 BUS 20 BUS 14 BUS 15 BUS 13 BUS 11 BUS 24 BUS 12 BUS 3 BUS 9 BUS 10 BUS 6 BUS 5 BUS 4 BUS 8 Power flow accross all lines follows Kirchoff s current & voltage laws. Transmission demand function cannot be defined intuitively! Optimal transmission capacity expansion becomes a complex optimization problem! BUS 1 BUS 2 BUS 7 Figure 8.14 One-line diagram of the IEEE Reliability Test System

127 The Reference Network Concept A network with a set of given nodes (e.g. existing connection points) and transmission lines of optimal capacity. Defined through the solution of a transmission investment & operation cost minimization problem. Purpose Sets a basis for identifying transmission capacity value. Existing network suboptimality? Welfare gains of capacity expansion?

128 Optimal Transmission Capacity Problem Formulation Objective Function min P g(t),α j,f j N t N g N j H t c g P g (t) + t=1 g=1 N t : Set of periods in a year N g : Set of generators N j : Set of transmission lines H t : Duration of period t j=1 (hr/year) [ ] αj C 0 + k j l j F j

129 Optimal Transmission Capacity Problem Formulation Subject to... Power Flow Constraints ( t [1, N t ], j [1, N j ] ) ( h j Pg (t), P d (t) ) = 0 Line Flow Constraints ( t [1, N t ], j [1, N j ] ) f j ( Pg (t), P d (t), F j ) = 0 0 F j α j M Generation Constraints ( t [1, N t ], g [1, N g ] ) y g ( Pg (t) ) 0

130 Summary Investments on the Transmission Network Lumpy, capital intensive, economies of scale are critical. Cost-based Capacity Expansion Is the obvious solution for a natural monopoly providing an essential service. Merchant Capacity Expansion Driven by differences in short-term locational marginal prices. Cost recovery may need to be controlled by the regulator.

131 Part III Questions?