FINANCIAL TRANSMISSION RIGHT FORMULATIONS Can Hybrid Models Work?

Size: px

Start display at page:

Download "FINANCIAL TRANSMISSION RIGHT FORMULATIONS Can Hybrid Models Work?"

Christiana Violet Lindsey
6 years ago
Views:

1 FINANCIAL TRANSMISSION RIGHT FORMULATIONS Can Hybrid Models Wor? William W. Hogan Center or Business and Government John F. Kennedy School o Government Harvard University Cambridge, Massachusetts Harvard Electricity Policy Group Atlanta, GA April 3, 2002

2 ELECTRICITY MARKET Transmission Management Deining and managing transmission usage is a principal challenge in electricity marets. Transmission Capacity Deinitions Contract Path Flow-Based Paths Point-to-Point Contract Path Fiction Parallel Flows Flows Implicit Physical OASIS Schedules Flowgate Rights Financial Transmission and TLR FGs Rights Financial PTP-FTRs FG-FTRs Obligations and Options 1

3 ELECTRICITY MARKET Flowgates I a single contract path is not good enough, perhaps many paths would be better. Since power lows along many parallel paths, there is a natural inclination to develop transmission services that would identiy the ey lins or lowgates over which the power may actually low, and to deine transmission rights according to the capacities at these lowgates. The assertion is that the commercially signiicant congestion can be represented by a system with: Few lowgates or constraints. Known capacity limits at the lowgates. Known power transer distribution actors (PTDF) or "shit" actors that decompose a transaction into the lows over the lowgates. An Evolution o Transmission Capacity Deinitions Contract Path Lin-Based Paths Point-to-Point Were these simpliying assumptions true, the decentralized model with physical rights might wor in practice. However, since these assumptions are not true, the physical interpretation o lowgate rights has been abandoned in avor o inancial transmission rights based on lowgates. 1 Contract Path Fiction Interace Capacity Varies with Conditions Parallel Flows Too Many Lins "Flowgates" To Acquire Capacity Flows Implicit Simultaneous Feasibility 1 William W. Hogan, "Flowgate Rights and Wrongs," Center or Business and Government, Harvard University, August

4 ELECTRICITY MARKET A Maret Framewor The Regional Transmission Organization (RTO) Millennium Order (Order 2000) contains a worable maret ramewor that is woring in places lie New Yor and the PJM Interconnection in the Mid- Atlantic Region. The RTO Order Contains a Consistent Framewor Bilateral Schedules at Dierence in Nodal Prices License Plate Access Charges Coordinated Spot Maret Bid-Based, Security-Constrained, Economic Dispatch with Nodal Prices Maret-Driven Investment Financial Transmission Rights (TCCs, FTRs, FCRs,...) 12/99 A spot maret with locational pricing gives rise to inancial transmission rights. 3

5 ELECTRICITY MARKET FTRs Point-to-point and lowgate inancial transmission rights could be obligations or options. The "hybrid" proposals include everything. Financial Transmission Rights Obligations Options Point-to-Point P Destination P Source Demonstrated in PJM, New Yor ( P P ) max 0, Destination Source More complicated to evaluate simultaneous easibility. Flowgate, η = shadow price, η 0, the lowgate amount. 0 Too complicated to use? o, η = shadow price, o η 0, the lowgate amount. Too complicated to use? 4

6 ELECTRICITY MARKET FTRs Point-to-point and lowgate FTRs could be obligations or options. Financial Transmission Rights FTR Deinitions -Point-to-point dierence in prices -Shadow price or lowgate amount -Obligations -Options DC-Load Power Flow Equilibrium $ $ PTP-FTR Obligation Right >7 Payment 3->7 51=79-28 PTP-FTR Option Right 3->7 Payment 3->7 51= FG-FTR Obligation 3->7 Line Shit Factor Shadow Price Payment 1->2 3/ >3-3/ >4-1/ >6 5/ >4 17/ >5 11/ >5 5/ >6 11/ >7 9/ >7 7/ Total FG-FTR Option 3->7 Line Shit Factor Shadow Price Payment $75.40 Constrained Line Shadow Price $ $ $ $ >2 3/ >3-3/ >4-1/ >6 5/ >4 17/ >5 11/ >5 5/ >6 11/ >7 9/ >7 7/ Total

7 ELECTRICITY MARKET FTRs Revenue adequacy is the inancial counterpart to the physical notion o "available transmission capacity." The physical ATC has no precise deinition, and it is not possible to guarantee physical delivery even when the ull grid is available. By contrast, it is possible to deine the requirement or revenue adequacy. 2 Revenue Adequacy: The revenue collected with locational prices in the spot maret dispatch is at least equal to the payment obligations or the FTRs in that same spot period. Simultaneous Feasibility: I the FTRs would be simultaneously easible under conditions available or the system operator, then the FTRs would be revenue adequate. We say that the set o FTRs is simultaneously easible i there is a u U such that: y = τ t L y, u + ι y = 0, K y, u 0. g,, the set o system controls, There is one loss balance and the security constraints. With thousands o monitored lines and hundreds o contingencies, there would be hundreds o thousands o constraints. 2 For details, see William W. Hogan, "Financial Transmission Right Formulations," Center or Business and Government, Harvard University, March 31,

8 ELECTRICITY MARKET FTRs Bid-Based, security-constrained, economic dispatch provides the maret equilibrium. yu, U Max B y t L y, u + ι y= 0, K y, u 0. Relaxation Solution Procedure 0 0 Step1: Select an initial candidate solution ( y, u ), ignore most (or all) o the constraints in the economic 0 dispatch using only the small subset K ( y, u ), and set the iteration count to m=0. Step 2: Construct the relaxed master problem as: Max B( y) yu, U m m m m m y y t L( y, u ) + L( y, u ) 0, m + ι y = u u m m m m m m m y y K ( y, u ) + K ( y, u ) 0. m u u m, u and update m=m+1. m Let a solution be ( y ) Step 3: Chec to see i the candidate solution ( m m, ) m K y u violates any o the constraints. I so, create a new ( y, u ) including some or all o these constraints and repeat Step 2. Else done. 7

9 ELECTRICITY MARKET PTP-FTR Obligations An FTR auction or point-to-point obligations is similar to an economic dispatch problem. Dispatch yu, U Max B y t L y, u + ι y= 0, K y, u 0. yu, Ut, 0, ρ 0 Max Auction t L y, u + ι y = 0, K y, u 0. β ( t, ρ ) y = t τ ρ g, Commercial sotware or economic dispatch can be and has been adapted with the relaxation strategy to solve or the auction awards and associated maret-clearing prices. PJM: Woring in a DC-Load approximation. New Yor: Woring in a ull AC Load Implementation. 8

10 ELECTRICITY MARKET PTP-FTR Options An FTR auction or point-to-point options is more complicated, but still similar to an economic dispatch problem. Dispatch yu, U Max B y t L y, u + ι y= 0, K y, u 0. Auction o Max β t, t, ρ o u U, t 0, t 0, ρ 0 0 o t LO tτ,{ t}, ρ g, u ι ρ g = 0, o w tτ,{ t}, ρ g, u 0. Evaluating each contingency in the dispatch requires solving a single load low. Evaluating a contingency and constraint combination in the auction requires solving an unconstrained optimal power low or the worst case impact o exercising options. I not too many constraints are binding, commercial sotware or economic dispatch might be adapted with the relaxation strategy to solve or the auction awards and associated maretclearing prices. Success with this auction model has not been demonstrated. Preliminary studies suggest the method might wor or the DC-Load implementation in PJM. The AC implementation in New Yor would be more o a challenge. 9

11 ELECTRICITY MARKET FG-FTRs An FTR auction or lowgate obligations or options would be more complicated. It is not lie an economic dispatch model. Max u U,0 φ,0 ρ,0 θ Auction (,, ) β φ θ ρ t L ρ g, u ι ρ g = 0, o K ρ g, u + φ + θ 0. Flowgate bids need not correspond to actual power lows, and the theory o the lowgate hedge suggests that all, or at least a very large number, o constraints would be ully subscribed in the auction. In principle, any constraint that could be binding in some dispatch (which is the deinition o a monitored constraint) would have some positive expected value or its uture shadow price. Hence, any such constraint should elicit a positive price in the auction. But then all the constraints would be binding. Further, the constraints are not separable. Hence, the ull model would have to be solved but the critical relaxation strategy or solving the auction model would not be available. 10

12 ELECTRICITY MARKET HYBRID MODELS An FTR auction or a hybrid model with point-to-point and lowgate obligations and options, is not lie an economic dispatch model. o u U, t 0, t 0,0 φ,0 ρ,0 θ Auction o (,,,, ) Max β t t φ θ ρ 0 o t LO tτ,{ t}, ρ g, u ι ρ g = 0, o o w tτ,{ t}, ρ g, u + φ + θ 0. The complications o the hybrid model produce a computational challenge. It is not clear that this auction model could be solved or a realistic grid. The constraints would be expensive to evaluate, and nearly every constraint would be binding. Available sotware or economic dispatch would not apply, and something else would be required. 11

13 ELECTRICITY MARKET HYBRID MODELS Can a hybrid model wor? And is one needed? When there is disagreement about elements o the maret design, an easy answer is to oer everything and let the maret decide. However, i oering everything is impossible or yields unintended consequences, it may be necessary to choose. Point-to-point Obligations: The PTP-FTR obligation wors in practice. By design ithe PTP-FTR obligation is closely matches the structure o economic dispatch. Point-to-point Options: Requested by most maret participants, but harder to implement. The design is close enough to the dispatch model, at least in the DC-Load case, to suggest that it might be possible to solve the auction problem. Flowgate Obligations and Options: These loo simple in small-scale examples, but are much more problematic in the real grid with security-constrained, economic dispatch. o o o There can easily be hundreds o thousands o lowgates in the real grid, greatly complicating any construction o hedges. The required lowgate amounts to hedge any transaction change requently with changing dispatch restrictions. The lowgate auction model may be too large to solve in its complete orm. Restrictions on the available lowgates may compromise the value as hedges or create more pressure to socialize the cost o lowgates. I all orms o rights are to be included in a hybrid model, the policy implication is to start with point-to-point obligations as the irst practical implementation. Later evaluate the introduction o point-to-point options or lowgate rights once these have been demonstrated to be worable in a real grid with complexity that does not appear in simpliied examples. 12

14 William W. Hogan is the Lucius N. Littauer Proessor o Public Policy and Administration, John F. Kennedy School o Government, Harvard University and a Director o LECG, LLC. This paper draws on wor or the Harvard Electricity Policy Group and the Harvard-Japan Project on Energy and the Environment. Ongun Alsaç, Art Desell, Fadil Dobraca, Scott Harvey, Guillermo Irisarri, Xingwang Ma, Jose Medina, Sasan Mohtari, Richard O'Neill, Andy Ott, Roberto Paliza, Susan Pope, Brian Stott, David Sun, and Greg Williams provided helpul discussion and comments. The author is or has been a consultant on electric maret reorm and transmission issues or American National Power, Brazil Power Exchange Administrator (ASMAE), British National Grid Company, Calpine Corporation, Comision Reguladora De Energia (CRE, Mexico), Commonwealth Edison Company, Conectiv, Detroit Edison Company, Duquesne Light Company, Dynegy, Edison Electric Institute, Electricity Corporation o New Zealand, Electric Power Supply Association, GPU Inc. (and the Supporting Companies o PJM), GPU PowerNet Pty Ltd., Mirant Corporation, National Independent Energy Producers, New England Independent System Operator, New England Power Company, New Yor Independent System Operator, New Yor Power Pool, New Yor Utilities Collaborative, Niagara Mohaw Corporation, Pepco, Public Service Electric & Gas Company, PJM Oice o Interconnection, San Diego Gas & Electric Corporation, Sempra Energy, TransÉnergie, Transpower o New Zealand, Westbroo Power, Williams Energy Group, and Wisconsin Electric Power Company. The views presented here are not necessarily attributable to any o those mentioned, and any remaining errors are solely the responsibility o the author. (Related papers can be ound on the web at 13

Loss Hedging Financial Transmission Rights

Loss Hedging Financial Transmission Rights Scott M. Harvey* and William W. Hogan** *LECG, LLC. Cambridge, MA **Center for Business and Government John F. Kennedy School of Government Harvard University