# Amicus Curiae Brief — Fed. Energy Regulatory Comm'n v. Elec. Power Supply Ass'n, 135 S. Ct. 2049 (2015) (No. 14-840) > Briefs, arguments, decisions, and more. URL: https://www.frixlaw.com/law-library/documents/brief%3Amicro_IA40385016_0317%3A28 ## Record - **Collection:** Supreme Court brief - **Document type:** Amicus Curiae Brief - **Published:** January 1, 2015 ## Text ee | FILED Nos. 14-840 & 14-841 ll 6 2015 7 OF i _ HE CLERK Supreme Court of the United States FEDERAL ENERGY REGULATORY COMMISSION, Petitioner, v. ELECTRIC POWER SUPPLY ASSOCIATION, ET AL., Respondents. ENERNOC, INC. ET AL., Petitioners, v. ELECTRIC POWER SUPPLY ASSOCIATION, ET AL., Respondents. ON WRITS OF CERTIORARI TO THE UNTTED STATES COURT OF APPEALS FOR THE DISTRICT OF COLUMBIA CIRCUIT BRIEF FOR ELECTRICITY CONSUMERS AND DEMAND RESPONSE PROVIDERS AS AMICI CURIAE IN SUPPORT OF PETITIONERS SETH P. WAXMAN Counsel of Record MARK C. KALPIN DAVID M. LEHN SAURABH H. SANGHVI WILMER CUTLER PICKERING HALE AND DORR LLP 1875 Pennsylvania Ave., NW Washington, DC 20006 (202) 663-6000 seth.waxman@wilmerhale.com TABLE OF CONTENTS Page TABLE OF AUTHORITIES ........:cscsscsssscssssecssnessssesessees iii INTEREST OF AMICI CURIAE..u....0-cccssssssssessseseseseeen 1 SUMMARY OF ARGUMENT ..0....:cccccscscssecsssesscsessssseereve 2 SE AT 6 I. THE STATES’ INABILITY TO EFFECTIVELY REPLICATE FERC’S EFFORTS CONFIRMS THAT FERC HAS JURISDICTION To ISSUE EEN Er eer a eS eee 6 A. The Federal Power Act Provides For Comprehensive And Effective Federal Authority Except Where It Reserves Authority To The States ..0..........ccccccsccsseeceseerseees 7 B. States Cannot Regulate Wholesale Demand Response...............csscsssssssssserseeseeseseeessers 8 1. Economic demand response ......................-.. 9 2. Reliability-based demand response .......... 15 Il. FERC REASONABLY DECIDED THAT DE- MAND RESPONSE SHOULD BE PAID THE LOCATIONAL MARGINAL PRICE ..........-ccccccescceeeeeeeee 19 A. The LMP System Pays Resources Based On Value Provided And Does B. Demand Response Provides Equal Value To Generation In Balancing The Markets And Thus FERC Reasonably Decided Not To Discriminate Against iP csesdaensessenssemnnsansennscenianserenesnrananicentenenantiinmenienenedl 22 rH TABLE OF CONTENTS—Continued Page C. FERC Reasonably Recognized That Uniform Pricing Would Remove An Improper Barrier To Demand Re- sponse Participation .....................-csecseseseeseeeeees 29 D. FERC Properly Rejected EPSA’s Ar- guments About Potential Abuse..................00+ 31 ee taitcieitseehivsciionshinenisminiiansbbeieninaiasnaisociinieiniciii 32 ‘ii TABLE OF AUTHORITIES CASES Page(s) Arkansas Louisiana Gas Co. v. Hall, 453 U.S. aT ara iatnctinnrcninciinnsnneccstiincanieanninicnnniantanemninniaiatibeeimnetine 6 FPC v. Louisiana Power & Light Co., 406 US. Ga ei iteeecncareseseresennsemcosnnnegiesonenesnmissenmeenes 7,8 FPC v. Transcontinental Gas Pipe Line Corp., ee I dtiestnreeninsetenritindicniionnininteittinnsescenniinins 8 Morgan Stanley Capital Group, Inc. v. Public Utility District No. 1 of Smohomish County, 554 U.S. 527 (2008) .........ccceccscceccsessccsecceseess 27 New York v. FERC, 535 U.S. 1 (2002) .........csccseceseeserees 10 Northwest Central Pipeline Corp. v. State Corporation Commission, 489 U.S. 493 SET ticnedesiceniceneneeneitsieuituesiniatidianiaateiatiiasedinintaiatadiaeneel 6 Oneok v. Learjet, No. 13-271, slip op. (U.S. Apr. Si TIITTI si enschoctadieaiinentnrtinicnaeuniatanmainianinaitiiaemasdeabiinpiamastinais 7 Panhandle Eastern Pipe Line Co. v. Public Service Commission, 332 U.S. 507 (1947)............-+- 7 Permian Basin Area Rate Cases, 390 U.S. 747 oe Wctcscssinsinceniniecnnnceiensinssirncenetninaianmaeninsiinmmainesenneesies 26 ADMINISTRATIVE PROCEEDINGS Comments of the Electric Power Supply Association, FERC Docket No. EL15-21- Fe es St settrctereensmessecesnnemenenemansnmnses 15 iv TABLE OF AUTHORITIES—Continued Page(s) Delaware Public Service Commission Comments to FERC Regarding D.C. Circuit May 23, 2014 Opinion Vacating FERC Order No. 745, FERC Docket No. RM10-17-000 (July 2, 2014) .20.......eecscsseseseeeereees 14, 17 Notice of Proposed Rulemaking, FERC Docket Nos. RM10-17-000, EL09-68-000 (Mar. 18, SRE This Court has an “established practice of citing inter- changeably decisions interpreting the pertinent sections” of the Natural Gas Act and the Federal Power Act. Arkansas La. Gas Co. v. Hall, 453 U.S. 571, 577 n.7 (1981). 7 to exist at the wholesale level, a level at which the States have no authority or ability to regulate, the de- cision below would create the very kind of regulatory “no man’s land” that Congress intended to avert. FERC, therefore, must have authority under the Act to regulate demand response at the wholesale level. At bottom, EPSA is arguing that demand response, in all its aspects in all its forms, is categorically a mat- ter of exclusive State concern. This Court has taken a much more nuanced approach to determining in whose regulatory domain a particular activity falls. Cf. Oneok v. Learjet, No. 13-271, slip op. 13 (Apr. 21, 2015). Be- cause States are necessarily limited in their ability to fully regulate demand response, it is entirely consistent with congressional intent that certain aspects of de- mand response, including those at issue here, be regu- lated by FERC while others are regulated by the States. A. The Federal Power Act Provides For Com- prehensive And Effective Federal Authority Except Where It Reserves Authority To The States In enacting the Federal Power Act, Congress ““meant to create a comprehensive and effective regula- tory scheme’ of dual state and federal authority.” F'PC v. Louisiana Power & Light Co., 406 U.S. 621, 631 (1972) (quoting Panhandle E. Pipe Line Co. v. Public Serv. Comm’n, 332 U.S. 507, 520 (1947) (citation omit- ted)). Under this scheme, Congress intended federal authority “to be broadly complementary to that re- served to the States, so that there would be no ‘gaps.”” Id. Thus, “in a borderline case where congressional au- thority is not explicit [the Court] must ask whether 8 state authority can practicably regulate a given area and, if ... it cannot, then [the Court is] impelled to de- cide that federal authority governs,” lest “a “no man’s land” ... be created.” Louisiana Power, 406 U.S. at 631 (quoting FPC v. Transcontinental Gas Pipe Line Corp., 365 U.S. 1, 19-20 (1961)). “This congressional blueprint has guided judicial interpretation of the broad language defining [FERC] jurisdiction.” /d. B. States Cannot Regulate Wholesale Demand Response A dual approach to demand response not only was prescribed by Congress but also makes good sense. Both FERC and the States bring their own relative advantages to bear on the regulation of demand re- sponse. On the one hand, States are properly posi- tioned to make policy judgments regarding the retail pricing schemes that allow demand response programs to suit their local needs. On the other hand, demand response is a critical tool in wholesale markets that only FERC can regulate. Consistent with congressional policy, over the past two decades electricity markets have became more ef- fective and reliable through the creation of FERC- regulated interstate ISOs.’ These ISOs are charged by FERC with balancing supply and demand within the markets they administer—known as “organized mar- kets”—on both a real-time and prospective basis, and with ensuring efficient reliability of the electric grid. Demand response is a market resource for address- ing two fundamental problems that have impaired ’ For ease of reference, this brief includes within the category of ISOs the closely related regional transmission organizations (“RTOs”). 9 ISOs’ ability to achieve these objectives: the relative inelasticity of retail demand for electric power, and the inefficiency of relying solely on generation to ensure system reliability and resource adequacy. The basic challenge in solving these problems is to create market structures for consumers that achieve efficient levels of demand. In this arena, the States, as discussed below, are legally and practically unable to function as an ef- fective substitute for FERC. 1. Economic demand response Energy markets are characterized by relative ine- lasticity of electricity demand because consumers gen- erally pay fixed rates, which respond to wholesale price changes only long after the fact. As a result, instead of consumers naturally modulating their consumption in response to retail price changes to maintain an efficient balance of supply and demand, ISOs have had to call on increasingly expensive marginal supply in times of high demand, which eventually raises retail electricity pric- es. To remedy this problem, FERC-regulated ISOs have established economic demand response programs, in which demand response providers bid into wholesale energy markets, typically on a day-ahead basis, to re- duce demand if the market clears at or above their par- ticular bid price. Such programs thus increase elastici- ty of demand by enabling the provider’s customers to experience wholesale price changes and compensating them at the clearing price for reducing electricity con- sumption. States are free to implement, and do implement, their own demand response programs at the retail lev- 10 el.‘ Such programs generally involve some limited var- iation in retail pricing that creates a price incentive for customers to reduce demand in peak hours.’ These programs are often salutary, but they cannot supplant wholesale demand response programs in the interstate markets operated by ISOs. The States are constitu- tionally incapable of regulating interstate electricity markets. Indeed, Congress’ motivation for enacting the Federal Power Act was to address that limitation and the “gap” in the regulation of electricity markets it cre- ated. See New York v. FERC, 535 U.S. 1, 5-6 (2002) (explaining the “Attleboro gap”). All that States could muster in the absence of fed- eral regulation of demand response would be a patch- work of potentially inconsistent regulations and local demand response programs that could not replicate the benefits of wholesale economic demand response pro- * Moreover, FERC has been respectful of the States’ efforts to regulate retail demand response when approving demand re- sponse programs at the wholesale level. See 18 C.F.R. § 35.28(g 1 iA) (limiting requirement that [SOs accept demand response bids when “not permitted by the laws or regulations of the relevant electric retail regulatory authority”); Order Accept- ing Tariff Sheets As Modified, 95 FERC 61,306, 62,043 (May 30, 2001) (same in approving [SO tariff). > State-administered demand response programs at the retail level could include, for example: “critical-peak pricing,” in which customers pay lower prices in most hours in exchange for paying very high prices during peak events; “peak time rebates” where customers are given rebates for reducing demand in certain peak hours; “time-of-use pricing,” which charges customers different fixed rates depending on the time of day that the power is con- sumed; and “real-time pricing,” which exposes the consumer to changes in wholesale prices directly. See generally FERC, Energy Primer: A Handbook of Energy Market Basics 47-48 (July 2012), at http//www.ferc.gov/market-oversight/guide/energy-primer.pdf. 11 grams regulated by FERC and implemented by [SOs at the interstate level. The critical question in facilitating economic de- mand response is determining how to pay for the com- pensation to customers that is necessary to reduce de- mand throughout the wholesale market. Demand re- sponse in a given regional organized market benefits all of its ratepayers. An ISO, therefore, is ideally posi- tioned to account for the total costs and benefits of such compensation. EPSA argues that this compensation could come from state-regulated load-serving entities (“LSEs”)— entities that supply power to end-use consumers— which could set up demand response programs with their retail customers. See EPSA Br. in Opp. 31 (argu- ing that “traditional public utilities and competitive power suppliers that purchase power in the wholesale market and then re-sell it to their retail customers can contract with retail customers and offer them incen- tives to reduce their demand consistent with state reg- ulation, which will in turn, allow those load-serving en- tities to reduce their purchases in the wholesale mar- ket”). But this simplistic assertion ignores key practi- cal realities of the markets. As other amici have recog- nized, the scheme EPSA lays out likely could not be dispatched by the ISO or directly integrated into the market-clearing price, limiting the ability of such a de- mand response program to compete with generation effectively. See NRG Cert. Amicus Br. 9; EnerNOC Br. 40. Moroever, FERC has recognized the “long- standing” and “difficult to address” problem that utili- ties are unlikely to encourage their customers to reduce demand. FERC, Assessment of Demand Response and 12 Advanced Metering 72 (Aug. 2006, rev. Dec. 2008) (“2006 FERC Report”), at http//www-ferc.gov/legal/ staff-reports/demand-response.pdf; see also FERC, 2010 Assessment of Demand Response and Advanced Metering 47 (Feb. 2011), at http//www_-ferc.gov/legal/ staff-reports/2010-dr-report.pdf (reiterating concern). Utilities are generally paid as a function of the total power they distribute; accordingly, reducing power sales will either reduce their revenue, which they obvi- ously do not want, or cause them to petition for rate in- creases. Furthermore, even LSEs that are not utilities may not have the necessary price incentives. LSEs are often indifferent to short-term wholesale price increas- es because they either own their generation (if vertical- ly integrated), enter into fixed-price contracts with generators, or have otherwise largely hedged their power purchases. Thus, LSEs may not immediately see the necessary price signals for their customers to reduce demand. As even a major generator and LSE has noted, in- dependent demand response providers create essential competition to overcome the traditional reluctance of utilities and LSEs to facilitate demand response. See NRG Cert. Amicus Br. 10. In wholesale markets, third- party demand response providers have enabled cus- tomers to manage their electricity usage and coordinat- ed their participation in demand response programs in order to provide significant benefits to their customers and all ratepayers in those markets. However, LSEs have a financial disincentive to allow third-party provi- sion of services in their service territory in addition to the disincentive to reduce demand for electricity. It is 13 not clear how EPSA’s scheme would permit any com- petition in the aggregation of demand response.° Nor would the States within organized interstate markets be positioned to provide this compensation themselves. The benefits of such compensation would be diffused across all States within the same organized market but the cost would be concentrated within the single State implementing the program. States are generally reluctant to impose the full costs of demand response programs on their own ratepayers when those ratepayers will not capture all the benefits of the in- vestments. See NERA Economic Consulting, Distrib- uted Resources: Incentives 11 (May 2006), at http:// www.nera.com/content/dam/nera/publications/archive 1/ PUB_distributed_resources_8.2006.pdf (discussing in- efficiencies due to the misalignment of costs and bene- fits of demand response).’ State-level dynamic pricing programs, whereby re- tail prices vary somewhat based on the time of day or market conditions, would not solve the problem, either. Most of these dynamic pricing programs are generally economically equivalent to a compensation-based pro- gram, in that a utility or LSE pays the consumer a re- bate for reducing demand during a peak period. Thus, these programs would suffer from the same problems ° And even if somehow, against their economic interest, LSEs did permit participation by third-party demand response provid- ers, a diverse multitude of LSE programs with potentially conflict- ing requirements would severely hinder any such participation. "A State served by a vertically integrated utility could pro- vide such an incentive, as the State’s ratepayers would also fully internalize the benefits of lower prices and reduced congestion resulting from the demand reduction. But these States lose out on the efficiencies that sparked creation of the [SOs in the first place. 14 just discussed. Indeed, to the extent that these pro- grams exist today in States that are part of organized markets, the relevant LSE or utility often acquires the compensation for the consumer by offering that con- sumer’s demand response into the ISO’s wholesale en- ergy markets." More generally, fully dynamic pricing (that is, where consumers are exposed to the full volatility in wholesale prices) is neither feasible nor desirable from a policy perspective. Large and small consumers alike tend to prefer stable pricing because of their aversion to price spikes, and States support that preference. In addition, consumers are always free to contract around a dynamic pricing scheme by purchasing their power at a fixed price. Such hedging limits the potential effec- tiveness of dynamic pricing as a long-term solution. Moreover it demonstrates that inelasticity of retail de- mand is not a temporary phenomenon, but rather a * EPSA disingenuously claims that somehow the elimination of wholesale demand response programs “may benefit State de- mand response initiatives.” EPSA Br. in Opp. 29-30. EPSA sup- ports this assertion with concerns by certain States regarding FERC’s proposed compensation scheme. Jd. But whether a State may express concerns with a particular wholesale compensation scheme is entirely distinct from whether the State would prefer to see the wholesale demand response programs eliminated altogeth- er. In fact, some of the very States on which EPSA relies express- ly confirm the importance of wholesale demand response to their initiatives. See C.A.J.A. 386 (recognition by Public Utilities Com- mission of Ohio of numerous benefits of “demand response re- sources participating in the real-time energy market”); Delaware Public Service Commission Comments to FERC Regarding D.C. Circuit May 23, 2014 Opinion Vacating FERC Order No. 745, FERC Docket No. RM10-17-000 (July 2, 2014) (urging FERC “to pursue vigorously all available means for seeking reversal” of the decision below). 15 fixed feature of the electricity market that must be ad- dressed by wholesale-level demand response programs. 2. Reliability-based demand response Electricity supply must be available when needed. To ensure short-term reliability and long-term adequa- cy of supply, many ISOs manage not only energy mar- kets, but also capacity markets.’ In capacity markets, LSEs traditionally pay generators, through an auction, for the commitment to produce power on the ISO’s re- quest. The auctions generally occur long before the ca- pacity is needed, to encourage the maintenance or building of sufficient generation capacity. Fulfilling these commitments can be expensive, and the resulting generation is often inefficient. In some cases, genera- tors are paid substantial sums to maintain or build power plants that might run only a few hours in an en- tire year. ISOs have created other demand response pro- grams to help ensure short-term reliability and long- term resource adequacy in their electricity markets. In these programs, demand response providers commit, months or years in advance, to reduce demand when called. These commitments can then be offered into the capacity markets for compensation, just like genera- 7 Although Order 745 directly addressed only the energy markets, neither the decision below nor the generators before this Court expressly limited their reasoning to demand response com- pensation in those markets. Indeed, following the Court of Ap- peals’ decision, generators rushed to FERC to argue that that de- cision also deprived FERC of authority to allow demand response resources to participate in capacity markets. See, e.g., Comments of the Electric Power Supply Association, FERC Docket No. EL15-21-000 (Dec. 4, 2014). It is therefore appropriate for this Court to consider the impact its decision will have on demand re- sponse participation in the capacity markets. 16 tion. ISOs appropriately have recognized that having demand response providers on stand-by to reduce de- mand during periods of high usage is more efficient than building excess power plants that may run only a few hours in a year.’° As with economic demand response, States are le- gally incapable of regulating reliability-based demand response in organized markets. See supra p. 10. And again, States are unable to replicate the bene- fits of reliability-based demand response programs through programs of their own. Currently, ISOs decide whether to ensure system reliability by dispatching marginal generation or demand response, and they do so on an interstate basis. For example, PJM can com- pensate for a temporarily downed generator in Penn- sylvania by calling on commitments by customers in Maryland to lower their consumption. Similarly, '° This benefit of demand response is most visible in its effect on capacity market prices. One independent analysis calculated that, without demand response, capacity prices for the 2017/2018 delivery year could have tripled in certain PJM zones. See Boshart, Navigant: Absent DR Participation, PJM Capacity Prices Could Jump up to $-fold, SNL Financial (Oct. 6, 2014), at https://www.snl.com/Interactive X/Article.aspx ?edid=A-29411194- 12081. Moreover, removing demand response in PJM would in- crease system-wide capacity costs by up to $9 billion for that same year. See FERC Pet. 32; EnerNOC Pet. 29-30. Having to redo capacity auctions in PJM for 2016/17 and 2015/16 would potentially increase costs by $10 billion and $14 billion, respectively. See Mon- itoring Analytics, Analysis of the 2016/2017 RPM Base Residual Auction 37 (Apr. 18, 2014), at http://www.monitoringanalytics.com/ reports/Reports/2014/IMM_Analysis_of_the_20162017_RPM_Base _Residual_Auction_20140418.pdf; Monitoring Analytics, Analysis of the 2015/2016 RPM Base Residual Auction 5 (Sept. 24, 2013), at http//www.monitoringanalytics.com/reports/Reports/2013/Analysis _of_2015_2016_RPM_Base_Residual_Auction_20130924.pdf. And these figures are just for a single organized market. 17 through the capacity market mechanism, an interstate ISO can ensure long-term resource adequacy by balanc- ing a lack of generation capacity in one State with de- mand response resources in another State. Thus, to be effective in ensuring reliability in an organized market, demand response resources must participate at the ISO-level. For this reason, while States in organized markets may have their own reliability-based programs, those programs likely could not exist if demand response could not participate in FERC-regulated capacity mar- kets. Those programs, generally operated by state util- ities, function much as private demand response aggre- gators at the wholesale level, permitting their custom- ers to act as demand response resources and then offer- ing the aggregated resources into ISO capacity auc- tions in return for compensation. Such programs are cost-effective largely because this capacity market par- ticipation offsets the state utilities’ capacity obligations in ISO auctions. Indeed, recognizing that the decision below has en- dangered States’ ability to fund their own programs, States in organized markets have voiced strong opposi- tion to the decision. See, e.g., Joint States Br. 23-24 (explaining that Maryland’s and Pennsylvania’s demand response programs have relied in significant part on PJM capacity market revenues for funding); Delaware Public Service Commission Comments to FERC Re- garding D.C. Circuit May 23, 2014 Opinion Vacating FERC Order No. 745, FERC Docket No. RM10-17-000 (July 2, 2014) (noting that Delaware’s $26 million Del- marva program and statutory efficiency goals are now 18 in jeopardy because of potential inability to access de- mand response in wholesale markets).'' + * * Before the decision below, the implementation of various kinds of demand response programs reflected the approach Congress envisioned of complementary state and federal regulation. Those programs also built on the tremendous advances of the past decades, in which interstate ISOs have realized substantial effi- ciencies across state lines. States cannot regulate these ISO demand response programs, nor can they replicate them with programs of their own. Thus, the decision below creates a regulatory no-man’s land and threatens to unravel the progress that these ISOs have achieved. The Federal Power Act does not require this result. '' After the Court of Appeals’ ruling, PJM proposed allowing LSEs to use demand response to offset their ISO capacity obliga- tions, while barring independent demand response aggregators from bidding into the market. See PJM Interconnection, L.L.C., Revisions to the Reliability Pricing Market and Related Rules in the PJM Open Access Transmission Tariff and Reliability Assur- ance Agreement Among Load Serving Entities, FERC Docket No. ER15-852-000 (Jan. 14, 2015). FERC rejected this proposal, which PJM conceded would have eliminated the providers that “have historically accounted for a majority of the demand response registered in PJM.” Jd. at 3. The proposal was both legally and practically defective. There is no basis in the Federal Power Act to conclude that FERC has authority to regulate demand response programs provided at the wholesale level by LSEs but not by oth- er entities. Moreover, as discussed above, demand response pro- viders create essential competition to overcome the reluctance of LSEs and utilities to offer demand response. PJM itself described its proposal as only a “stop-gap” program that is inferior to the status quo, and recognized that demand response participation “could be substantially lower under this proposal than it has been historically.” Jd. at 2-3. 19 I. FERC REASONABLY DECIDED THAT DEMAND RE- SPONSE SHOULD BE PAID THE LOCATIONAL MARGINAL PRICE Order 745 is reasonable. FERC properly deter- mined that when demand response is as capable as gen- eration of balancing supply and demand and is cost- effective, an ISO should value it the same as generation and thus it should be paid the same amount that gener- ation is paid: the LMP, or locational marginal price. EPSA instead argues that, even when the above conditions are met, FERC was required to value de- mand response less than generation, namely, as a LMP minus the avoided retail cost of electricity, G. But EPSA’s argument rests on a cramped characterization of economic efficiency—which includes the factors it wants but ignores the rest—that is not required by the Federal Power Act. Particularly in light of the com- plexities of the electricity markets, FERC has ample discretion in weighing different considerations in its ratemaking. FERC’s judgment was based on its rele- vant expertise and experience; it merits substantial deference. A. The LMP System Pays Resources Based On Value Provided And Does Not Inquire Into Costs EPSA insists that the price paid to demand re- sponse participants in wholesale energy markets should reflect those participants’ costs. But that is not how pricing in wholesale energy markets works. The stand- ard wholesale price paid to power generators reflects the value of the electricity they provide, not their costs. Under the locational marginal pricing system in de- regulated markets, generators typically provide offers 20 that indicate the minimum price they would need to be paid to generate a particular quantity of load at a par- ticular location and at a particular moment in time.” Based on these signals and other constraints like transmission availability, the ISO identifies the lowest cost method of serving the load throughout the inter- state electricity grid it manages. It does so by calculat- ing, at every point in the system, the instantaneous marginal cost of serving the last unit of load required at that location, even if some proportion of the load (be- fore the last unit) could be served at a lower cost. This is referred to as the locational marginal price, or “LMP.” The LMP then serves both as a dispatch signal and the market-clearing price: any generator whose of- fer at a particular location is at or below the LMP is dispatched by the ISO and paid LMP by the LSE for the electricity they receive from the generator. Because costs of generation vary significantly across generators, the marginal cost of supply can be steeply increasing, particularly during times of peak load. In these circumstances, most generators in the system—certainly those that bid below the market- clearing price—will be paid more than their original of- fers, and thus more than their cost, upon dispatch.” As FERC explains, in determining appropriate compensa- tion it generally does not “inquire into the costs or ben- efits of production for the individual resources partici- pating as supply resources.” Order 745 62 (EnerNOC '2 See generally Synapse Energy Economics, LMP Electrici- ty Markets: Market Operations, Market Power, and Value for Consumers (Feb. 5, 2006), at http//www.publicpower.org/files/ PDF s/SynapseL MP Electricity Markets013107.pdf. '3 Thus, even though [SOs may generally require generators to base their offers on their actual costs, that does not mean that generators are paid on the basis of their actual costs. 21 Pet. App. 186a). As a result, if the LMP is $150/MW, an efficient baseload plant that produces power for $15/MW will receive the same compensation as the inef- ficient plant producing it for $150/MW. The more- efficient generators get to pocket the difference be- tween their cost and the LMP as profit. Therefore, the LMP system in deregulated mar- kets compensates resources based on the value they provide, not the costs they incur. LMP is a measure of that value because it represents the amount the system would have to pay the next resource to offer the same generation if the cheaper one were to disappear. EPSA is wrong to focus on whether demand re- sponse customers in essence “do better” than genera- tors because they also avoid the retail cost of electricity not bought, “G.” The avoided electricity cost for de- mand response participants is just one part of the in- ternal calculus that a customer will consider when as- sessing whether to offer demand response, alongside numerous other costs such as comfort and convenience (especially for residential customers), business inter- ruptions, investment in metering technology, and over- time. Indeed, everyone in the market has costs and benefits associated with their participation, and so more-efficient generators “do better” than less-efficient ones since all generators are paid LMP. Just as FERC generally “does not inquire into the costs or benefits of production” of generation resources, it should not “sin- gle out demand response resources for adjustments to compensation,” so long as the value of the resource to the ISO is the same. Order 745 ¢ 62 (EnerNOC Pet. App. 186a). 22 B. Demand Response Provides Equal Value To Generation In Balancing The Markets And Thus FERC Reasonably Decided Not To Dis- criminate Against It The question FERC grappled with, then, was whether demand response offers equal value to genera- tion to the ISO in the relevant sense. FERC rightly and reasonably determined that it does, and so pay- ment of LMP is appropriate. 1. Demand response provides equal value to the ISO because an incremental megawatt of load can equally be supplied by one megawatt of generation or one megawatt of demand response. See Order 745 4 55 (EnerNOC Pet. App. 181a) (“Generation and load must be balanced by the RTOs and ISOs when clearing the day-ahead and real-time energy markets, and such bal- ancing can be accomplished by changes in either supply or demand.”). Or, as even an expert supporting EPSA, Dr. Robert Borlick, put it before FERC, “economic de- mand response and generation are equivalent balancing resources.” Response of Robert L. Borlick to Professor Alfred E. Kahn 4, FERC Docket No. RM10-17-000 (Oct. 9, 2010) (capitalization omitted). As discussed above, FERC has encouraged the de- velopment of ISOs in deregulated markets because their ability to balance supply and demand across state lines improves efficiency and thus helps ensure just and reasonable rates. So FERC reasonably can focus on the resource’s ability to balance supply and demand in assessing its value. Providing for demand response to be compensated at less than generation would improp- erly force ISOs to discriminate against demand re- sponse, requiring ISOs to choose a more expensive way 23 to balance supply and demand. The Federal Power Act does not require that result. An example helps to demonstrate how an LMP-G approach improperly would force the ISO to discrimi- nate against demand response, ultimately to the detri- ment of consumers. Suppose that after using genera- tion to balance 1000 MW of load, an ISO must procure resources for an additional 10 MW of load. Incremental generation offers are $110/MW for the next 5 MW and $120/MW for the next 5 MW after that. There is also a demand response resource that, after considering the costs of an interruption to its operations, is willing to provide 10 MW of demand reduction in return for $100/MW. G is $30/MW. Under Order 745, the ISO would see that the $100/MW demand response offer provides the cheapest resource to balance the load. But before dispatching it, the ISO would confirm, pursuant to Order 745, that the demand response offer passes the net benefits test. FERC required this test to address the “billing unit effect,” in which in certain circumstances, it may be in- efficient to dispatch demand response, even if it is the cheapest resource, because it would reduce the load base. See Order 745 4¥ 51-53 (EnerNOC Pet. App. 178a-179a) (explaining billing unit effect). If the test is passed, the demand response will be dispatched, and LMP will be $100/MW, whereas LMP would be $120/MW without demand response. The 1000 MW of generation and 10 MW of demand response will be paid this $100/MW, while the 1000 MW of load receiving the generation will pay $100/MW plus an additional $1/MW “uplift” to account for the reduction in load base by de- 24 mand response.’ The net benefits test ensures that this uplift will be lower than the reduction in LMP caused by the demand response—here, the $1/MW up- lift would be less than the $20/MW increase in LMP ab- sent demand response—and thus that demand response is dispatched only when it results in lower costs for the system overall. EPSA suggests that under its LMP-G approach, the demand response resource would be paid $70/MW in this example, rather than $100/MW. But that’s not what would actually happen. Rather, the demand re- sponse resource had determined that it was willing to provide demand response only if it is paid at least $100/MW—that was the basis of its prior hypothetical offer—and that position would not be any different un- der EPSA’s approach. So, under EPSA’s approach, the demand response resource would have to increase its offer from $100/MW to $130/MW, ensuring that it will '* The reason for the uplift payment is, in short, that supply counts both generation and demand response, but load counts only the demand for generation. Without demand response, there would be 1010 MW of supply (all generation) paid for by 1010 MW of load, and so a perfect match between the two sides. If demand response is dispatched, however, the supply side remains 1010 MW (now 1000 MW of generation plus 10 MW of demand response), but the load side is reduced by the amount of demand response and is thus just 1000 MW—since, by definition, demand response re- sources are not using the supply of power. Because demand re- sponse resources cannot be charged for the power they are not using, but all supply resources must still be paid LMP even if de- mand response is dispatched, the load base has to cover the differ- ence. This difference is the uplift, and it is calculated by evenly allocating the amount of the payment owed to the demand re- sponse providers across the load base. In this example, the total payment to demand response would be $1000 (10 MW x $100/MW). Allocating that payment over 1000 MW of load yields an uplift payment of $1/MW. 25 be dispatched only when its compensation would at least match its requirement of netting $100/MW. The result is that the two incremental generation offers will be dispatched instead of demand response, LMP will be $120/MW, and the load receiving the generation will pay $19/MW more than under Order 745. Properly understood, then, the issue is not whether demand response should be paid more as a general mat- ter; as the example shows, demand response customers will invariably adjust their offers so that they are paid at least what they need to justify the interruption. In- stead, the question is whether an ISO should discrimi- nate between two equivalent ways of balancing supply and demand, valuing demand response less and thus increasing system costs by dispatching demand re- sponse at the price it is willing to be paid only when the next unit of generation is more expensive by at least G. There is no reason for such discrimination. As dis- cussed, demand response is as capable of balancing supply anc demand as generation is. Moreover, alt- hough demand response lowers the load base, the net benefits test ensures that the uplift would be more than offset by the reduction in LMP, and the dispatch of de- mand response is thus efficient. 2. In arguing that FERC unreasonably deter- mined that ISOs should not discriminate against de- mand response, EPSA and its supporters fixate on one particular sense in which the dispatch of demand re- sponse may differ from the dispatch of generation: when demand response is dispatched, society may in some cases lose the value of whatever productive use the demand response provider (or more precisely, its customers) would otherwise have made of that electric- ity. Thus, using highly stylized examples, amici sup- 26 porting EPSA below speculated that Order 745 would cause customers to “forgo socially valuable activity.” Borlick et al. C.A. Amicus Br. 18. Under EPSA’s ap- proach, this activity is valued at G because that is the amount that the demand response provider otherwise would have paid to engage in that activity. This argument is flawed. FERC’s mandate is not to maximize social welfare generally, nor is FERC obli- gated to adhere to EPSA’s interpretation of economics textbooks. See Permian Basin Area Rate Cases, 390 U.S. 747, 776-777 (1968) (“rate-making agencies are not bound to the service of any single regulatory formula; they are permitted, unless their statutory authority otherwise plainly indicates, to make the pragmatic ad- justments which may be called for by particular cir- cumstances” (quotation marks omitted)); see also Order 745 ¥ 46 (EnerNOC Pet. App. 175a) (noting that FERC ‘is not limited to textbook economic analysis of the markets subject to our jurisdiction, but also may ac- count for the practical realities of how those markets operate”). Rather, in addition to ensuring that rates are not unduly discriminatory or preferential (which as discussed, Order 745 does), FERC must ensure that wholesale electricity rates are just and reasonable. See 16 U.S.C. § 824d. This is no easy feat: unlike any other market in the world, electricity markets require continuous and in- stantaneous balancing of supply and demand, feature a steeply increasing marginal supply curve, and demand the utmost reliability in operation in light of electrici- ty’s importance for basic human necessities. See Order 745 { 56 (EnerNOC Pet. App. 181la) (recognizing “dis- tinctive” characteristics of electricity markets in ex- plaining decision). In these complex circumstances, FERC ensures just and reasonable rates in significant 27 part by relying on ISOs in organized markets to bal- ance supply and demand effectively and éfficiently across state lines. Thus, it is entirely appropriate for FERC to focus on the value demand response provides in balancing supply and demand when assessing its val- ue as a supply resource. This determination is owed “great deference.” Morgan Stanley Capital Grp., Inc. v. Public Util. Dist. No. 1 of Snohomish Cty., 554 U.S. 527, 532 (2008) (the “statutory requirement that rates be ‘just and reasonable’ is obviously incapable of pre- cise judicial definition”).’° In any event, EPSA is wrong in thinking that dis- patch of demand response causes the systematic loss of social value. In many cases, such loss would not occur or would be entirely speculative. For example, many businesses have supplies or inventory to accommodate just such curtailment of electricity consumption. And if the demand response merely shifts an industrial pro- duction process from one particularly hot summer weekday to a cooler weekend, then the social benefits of that process will still be realized while sparing the system the costs of highly expensive peak-time power. In other cases, the social value lost when the demand response customer reduces consumption may be entire- ly intangible, such as students sitting in classrooms that are 1 degree warmer in the summer or an office build- ing’s decorative water fountain being off for an hour. 'S For the same reason, EPSA missed the point below when it argued that the LMP-G approach is necessary to ensure that de- mand response customers see the same price signal as if the elec- tricity markets had full real-time retail pricing. See EPSA C.A. Br. 52. Nothing in the Federal Power Act requires FERC to rep- licate retail pricing signals. FERC is free to determine that an- other signal better comports with its mandate to ensure just and reasonable wholesale rates. 28 Nothing in the Federal Power Act requires FERC to discriminate against demand response because of these unlikely, speculative, or intangible effects. EPSA also ignores the countless positive externali- ties of dispatching demand response and negative ex- ternalities of dispatching generation that FERC recog- nized in Order 745. Demand response lowers wholesale price volatility by reducing the hourly variation in con- sumption patterns.'© By reducing peak load, demand response limits strain on shared resources like the transmission system. It mitigates generators’ market power by developing an alternative means to supply load. See Order 745 J 10 (EnerNOC Pet. App. 147a). It increases reliability by giving ISOs another tool to bal- ance supply and demand in emergencies that is qualita- tively different from generation. See id. (EnerNOC Pet. App. 147a-148a).'’ To the extent it reduces total load, demand response benefits the environment, and it also helps promotes the use of renewables like wind and 1 See Order 745 | 10 n.16 (EnerNOC Pet. App. 147a) (“De- mand response tends to flatten an area’s load profile, which in turn may reduce the need to construct and use more costly resources during »eriods of high demand; the overall effect is to lower the average cost of producing energy.” (quoting ISO-RTO Council Re- port, Harnessing the Power of Demand How RTOs and [SOs Are Integrating Demand Response into Wholesale Electricity Mar- kets)). '7 FERC has recognized that demand response can be more effective than generation in responding to emergencies. First, it can be much quicker. it is often easier to turn off consumption than to start a new generator. Second, it can be more reliable, because it is a “statistical resource.” See 2006 FERC Report 119-120. That is, whereas the failure of a few consumers to fulfill their commit- ments to reduce consumption when requested has a amall effect on the relationship between supply and demand, the failure of a gen- erator to provide power on request has a substantial effect. 29 solar power because it can efficiently plug gaps in those sources’ otherwise intermittent generation. '® ° Thus, if FERC were to account for the full range of costs and benefits of demand response, it could well conclude that LMP undercompensates demand re- sponse participation in organized markets. Instead, FERC has decided to focus simply on the value that demand response provides in balancing supply and de- mand. That focus, though more specific than it could be, is reasonable. C. FERC Reasonably Recognized That Uniform Pricing Would Remove An Improper Barrier To Demand Response Participation FERC’s experience with demand response also shows that valuing demand response equally to genera- tion is essential to achieving successful demand re- sponse programs and is therefore reasonable. In the notice of proposed rulemaking, FERC found that despite the benefits they offer, at that point “de- mand response providers collectively play[ed] a small role in wholesale markets.” Notice of Proposed Rule- making { 9, FERC Docket Nos. RM10-17-000, ELO09- 68-000 (Mar. 18, 2010) (“NPRM”) (J.A. 32). With the experience of “several years of observing demand re- sponse participation in ISO and RTO markets with dif- ferent, and often evolving, demand response struc- tures,” FERC expressed concern that “some existing[] 18 Por example, a recent Navigant study concluded that de- mand response could reduce carbon dioxide emissions by 2%, which is 10% of the Environmental Protection Agency’s target of 20% reductions by 2030. See Navigant, Carbon Dioxide Reduc- tions from Demand Response 1 (Nov. 25, 2014), at http//www. ieca-us.com/wp-content/uploads/Carbon-Dioxide-Reductions-from- Demand-Response_Navigant_11.25.14.pdf. 30 inadequate compensation structures have hindered the development of demand response” Jd. For example, FERC noted that when PJM went from essentially an LMP mechanism,’’ demand response dropped by 36.8%. Id. f 10 (J.A. 33). With Order 745, then, FERC reason- ably concluded that uniform compensation at LMP was necessary to spur development of this valuable re- source. See Order 745 ¥ 57 (EnerNOC Pet. App. 181a- 183a); Order 745-A 4 74 (EnerNOC Pet. App. 92a) (ex- plaining need for uniform approach).”° This does not mean, as amici for EPSA claimed be- low, that FERC “assume[d] that more demand re- sponse is always better, regardless of the amount.” Borlick et al. C.A. Amicus Br. 18 (quotation marks omitted). FERC simply recognized that demand re- sponse had (at least) equal value to generation in bal- ancing supply and demand, that existing compensation schemes discriminating against demand response were underutilizing this value, and that adjusting the com- pensation for demand response to be equal to the com- pensation for generation would thereby ensure more just and reasonable rates. This is precisely the kind of policy judgment Congress entrusted FERC to make, particularly in light of its policy directive that “unnec- '9 DIM provided for payment of LMP when LMP was at least $765/MWh and paid LMP minus the generation and transmission components of the retail rate when LMP was less than $75/MWh. NPRM f 10 n.33 (J.A. 33). - Experience has proven FERC prescient. In a report ana- lyzing the first several months of implementation of Order 745, PJM found that demand response participation had significantly increased and that the performance of demand response had “dra- matically improved.” PJM, 2012 Economic Demand Response Performance Report, at http://www.pjm.com/~/media/markets- ops/dsr/20160701-order-745-impact-on-economic-dr.ashx. 31 essary barriers to demand response participation in en- ergy, capacity and ancillary service market’ shall be eliminated.” Energy Policy Act of 2005, Pub. L. No. 109-58, § 1252(f), 119 Stat. 594, 965-966 (16 U.S.C. § 2642 note). D. FERC Properly Rejected EPSA’s Arguments About Potential Abuse Finally, some of EPSA’s arguments amount to a claim that Order 745 could be abused, by encouraging parties to take steps to profit from the compensation approach in a way that does not lead to actual demand reductions. This is a red herring. Whenever a regulato- ry scheme compensates beneficial activity, there will be possibilities of abuse, whether that activity is demand response, generation, or something else. It is unre- markable that compensating demand response provid- ers substantially less, as EPSA’s proposed LMP-G pric- ing would do, might reduce the incentive for abuse. But that is no reason to undercompensate demand response. Rather, the proper way to address such concerns is verification and enforcement. As FERC recognized in Order 745, “demand reductions that are not genuine may be violations of the Commission’s anti- manipulation rules.” Order 745 ¢ 95 (EnerNOC Pet. App. 208a); see 17 C.F.R. § 1.c.2 (prohibiting fraud, false statements, and manipulation in connection with pur- chase or sale of electric energy). Moreover, every ISO has specific measurement and verification protocols de- signed to ensure that demand response reductions are genuine. For example, PJM will only compensate those demand reductions that are “executed in response to the real-time and/or day-ahead LMP or as dispatched by PJM and that are not implemented as part of normal operations.” PJM Manual 11 Energy & Ancillary Ser- 32 vices Operations (Apr. 9, 2015), ch. 10.1, at http//www. pjm.com/~/media/documents/manuals/m11l.ashx. PJM further enumerates examples of ineligible demand re- duction, which include “[slettlements based on On-Site Generator data if the On-Site Generation is not sup- porting demand reductions executed in response to ... LMP,” and “[cjonsecutive daily settlements that are the result of a change in normal demand patterns.” /d. Repeat offenders can be suspended and referred to FERC’s Office of Enforcement. /d. EPSA in essence argues that Order 745 will in- crease the likelihood of violations of these rules, and now wants FERC to throw out the baby with the bathwater. But FERC is well within its power to be more exacting, as it is in allowing other beneficial prac- tices that nonetheless could be abused. Indeed, genera- tion likewise is vulnerable to market manipulation, and appropriately subject to compliance monitoring and en- forcement. Thus, Order 745 properly directed ISOs to review their protocols in light of the change in compen- sation, and either explain how their protocols are ade- quate or propose any necessary changes in their com- pliance filings. See Order 745 94 (EnerNOC Pet. App. 207a). FERC’s approach is reasonable.” CONCLUSION For the foregoing reasons, the decision of the Court of Appeals should be reversed. 7! EPSA is free to raise any concerns it has with an ISO’s measurement and verification protocols in those separate proceed- ings. JULY 2015 33 Respectfully submitted. SETH P. WAXMAN Counsel of Record MARK C. KALPIN DAVID M. LEHN SAURABH H. SANGHVI WILMER CUTLER PICKERING HALE AND DORR LLP 1875 Pennsylvania Ave., NW Washington, DC 20006 (202) 663-6000 seth.waxman@wilmerhale.com --- Source: Frix Law Library, https://www.frixlaw.com/law-library/documents/brief%3Amicro_IA40385016_0317%3A28. Public record. Not legal advice.