PSC REF#:215945 ______________________________________________________________________________ Joint Application of Wisconsin Electric Power Company and Wisconsin Gas LLC, both d/b/a WE Energies, to Conduct a Biennial Review of Costs and Rates - Test Year 2015 Rates Docket No. 05-UR-107 ______________________________________________________________________________ REBUTTAL TESTIMONY OF ASHLEY C. BROWN ON BEHALF OF WISCONSIN ELECTRIC POWER COMPANY ______________________________________________________________________________ 1 Q. Please state your name, occupation, and address 2 A. My name is Ashley C. Brown. I am Executive Director of the Harvard Electricity Policy 3 Group (HEPG) at the John F. Kennedy School of Government, at Harvard University. 4 HEPG is a “think tank” on electricity policy, including pricing, market rules, regulation, 5 environmental and social considerations. HEPG, as an institution, never takes a position 6 on policy matters, so my testimony today represents solely my opinion, and not that of 7 the HEPG or any other organization with which I may be affiliated. 8 Q. Please describe your professional qualifications. 9 A. I am an attorney licensed to practice law in Massachusetts, Ohio, and the District of 10 Columbia. 11 I served 10 years as a Commissioner of the Public Utilities Commission of Ohio (1983- 12 1993), where I was appointed and re-appointed by Democratic Governor Richard Celeste. 13 I also served as a member of the NARUC Executive Committee and as Chair of the 14 NARUC Committee on Electricity. I was a member of the Advisory Board of the 15 Electric Power Research Institute. I was also appointed by the U.S. Environmental REBUTTAL-WEPCO/WG-BROWN-1 Public Service Commission of Wisconsin RECEIVED: 09/12/14, 11:43:05 AM BEFORE THE PUBLIC SERVICE COMMISSION OF WISCONSIN 1 Protection Agency as a member of the Advisory Committee on Implementation of the 2 Clean Air Act Amendments of 1990. I am also a past member of the Boards of Directors 3 of the National Regulatory Research Institute and the Center for Clean Air Policy. 4 I have served on the Boards of Oglethorpe Power Corporation, Entegra Power Group, 5 and e-Curve, and as Chair of the Municipal Light Advisory Board in Belmont, MA. 6 I serve on the Editorial Advisory Board of the Electricity Journal. 7 I have been at Harvard continuously since 1993. During that time I have also been 8 Senior Consultant at the firm of RCG/Hagler, Bailly, Inc. and have been Of Counsel to 9 the law firms of Dewey & LeBouef and Greenberg Traurig. I have also taught in training 10 programs for regulators at Michigan State University, University of Florida, and New 11 Mexico State University (the three NARUC sanctioned training programs for regulators), 12 as well as at Harvard, the European Union School of Regulation, and a number of other 13 universities throughout the world. I have advised the World Bank and the Inter- 14 American Development Banks on energy regulation, and have advised governments and 15 regulators in more than 25 countries around the world, including Brazil, Argentina, South 16 Africa, Costa Rica, Zambia, Tanzania, Namibia, Mozambique, Hungary, Ukraine, Russia, 17 India, Bangladesh, Saudi Arabia, Indonesia, and The Philippines. I have written 18 numerous journal articles and chapters in books on electricity markets and regulation, and 19 am co-author of the World Bank’s Handbook for Evaluating Infrastructure Regulation. 20 I hold a B.S. from Bowling Green State University, an M.A. from the University of 21 Cincinnati, and a J.D. from the University of Dayton. I have also completed all work, 22 except for the dissertation, on a Ph.D. from New York University. My current CV is 23 provided as Ex.-WEPCO/WG-Brown-1. REBUTTAL-WEPCO/WG-BROWN-2 1 Q. Have you previously testified before the Wisconsin Public Service Commission? 2 A. No. I have testified, however, before FERC and various state commissions, as well as 3 before Congressional and state legislative committees. 4 Q. On whose behalf do you offer your testimony? 5 A. On behalf of Wisconsin Electric Power Company (WE). 6 Q. What is the purpose of your testimony? 7 A. The purpose of my testimony is to assess the value of distributed (DG) solar 8 photovoltaics (PV) and appropriate pricing for its value and output. In the course of my 9 testimony I will address various points made in the direct testimony of RENEW/ELPC 10 witnesses Gilliam, Rabago and Vickerman; MMSD witness Cicchetti; TASC witnesses 11 Hornby and Friedman; and Commission Staff witness Singletary. 12 In particular, I will address the argument advanced explicitly or implicitly by witnesses 13 Gilliam, Vickerman, Cicchetti, Hornby and Singletary that retail net metering is an 14 equitable way to compensate customers who own DG. (Direct-RENEW-Gilliam-15-16; 15 Direct-RENEW-Vickerman-3-4, 6-7, 10-18, 20-31, 33-34; Direct-MMSD-Cicchetti-23- 16 24; Direct-TASC-Hornby-3-4, 8-10, 14-28; and Direct-PSC-Singletary-13-24, 36.) I will 17 also address the argument advanced by witnesses Gilliam, Rabago and Hornby in favor 18 of a “value of solar” approach to compensating solar PV DG customers. (Direct- 19 RENEW-Gilliam-6-10, 15-16; Direct-RENEW-Rabago-44-50; Direct-TASC-Hornby-10- 20 13.) 21 Q. What conclusions do you reach in your testimony? 22 A. My conclusions are that retail net metering and “value of solar” are severely flawed 23 schemes for pricing solar PV DG. Net metering overvalues the energy produced as well REBUTTAL-WEPCO/WG-BROWN-3 1 as the installed capacity, is very heavily cross-subsidized by non-solar customers, and is 2 socially regressive in that it transfers wealth from less affluent to more affluent 3 consumers. The "value of solar" approach promulgated by solar advocates like witnesses 4 Gilliam, Rabago, and Hornby artificially inflate the benefits of solar PV DG and discount 5 the costs. I also conclude that WE’s proposal is a very reasonable, market-based system 6 for pricing solar PV DG, and that the proposals for demand and fixed charges as applied 7 to solar PV DG hosts are reasonable ways to rectify the cross-subsidies in net metering. 8 Net metering and the value of energy and other attributes of distributed solar PV. 9 Q. Do you see value in distributed solar PV? 10 A. Yes, I do. Solar PV has some very real benefits and long-term potential. The marginal 11 costs of producing this energy are zero. If one looks at environmental externalities, then 12 the carbon emissions from the actual process of producing this energy itself, without 13 taking the secondary effects into consideration, are also zero. Significantly, the costs of 14 installing solar PV have declined in recent years, adding to the potential long term 15 attractiveness of solar. Those are very real benefits that would be valuable to capture. In 16 its current, most common configuration, however, it has some drawbacks that inhibit it 17 from capturing its full value. Solar PV is intermittent and thus requires backup from 18 other generators and cannot be relied on to be available when called upon to produce 19 energy. Thus, its energy value is entirely dependent on when it is produced and its 20 capacity value is, at best, marginal. To fully develop the resource, therefore, it is 21 imperative to provide pricing that will allow solar PV DG to fulfill its potential, by 22 linking itself to storage, more efficient forms of catching the sun’s energy, or other types 23 of generation (e.g. wind) that complement its availability. Thus, it is critical that prices REBUTTAL-WEPCO/WG-BROWN-4 1 be set in such a fashion as to provide incentives for productivity and reliability and not to 2 subsidize it at a decidedly low degree of optimization. Solar has huge potential, but to 3 attain it, it needs to receive the price signals to fully exploit its capabilities. 4 To fully assess the value of solar, one would need to look at the resource in all of its 5 dimensions, not simply the costs and environmental effects derived from the energy- 6 producing process itself. It is also critical to think of pricing in the context of 7 establishing incentives for technological improvements that would increase its efficiency. 8 Solar PV's present value, on a stand-alone basis, is very limited because of its intermittent 9 nature and its uncertain availability. 10 Q. 11 12 Are you saying that the prices paid for distributed solar PV should be the result of an analysis of all the criteria and ascribing empirical value to them to establish prices? A. No, I am not saying that at all. What I am saying is that many in the solar industry have 13 come to recognize that retail net metering (NEM) is, in the age of smart grid and more 14 sophisticated technology, no longer a defensible method for pricing solar DG. Having 15 recognized the inevitable demise of a pricing system that favors their resource through 16 cross-subsidization by other customers, many solar advocates have shifted to an argument 17 that pricing should be based on consideration of the “value of solar.” While I do not 18 subscribe to that point of view, I recognize that the argument is being made and a 19 regulatory forum, such as this, has an obligation to provide a fair hearing for those who 20 call for a “value of solar” based pricing system, so it seems appropriate to respond to the 21 arguments being advanced. I will, however, focus heavily on net metering (NEM) 22 because that is the historical status quo in Wisconsin which WE has proposed to change, REBUTTAL-WEPCO/WG-BROWN-5 1 and which many of the solar interveners, by attacking WE’s proposal and offering no 2 alternative of their own, are effectively trying to defend. 3 Q. How do you define net metering (NEM)? 4 A. To begin, I would note that I use the terms "net metering," "retail net metering" and 5 "NEM" interchangeably and synonymously. What I mean is that the meters run forward 6 when solar PV DG customers are purchasing energy from the grid. When those 7 customers produce energy and consume it on premises, the meter simply stops, and when 8 the customer produces more energy than is consumed on premises, the meter runs 9 backwards. Thus the solar PV DG customer pays full retail value for all energy taken off 10 the grid, pays nothing for energy or distribution when self-consuming energy produced 11 on premises, and is paid the fully delivered retail price for all energy exported into the 12 system. At the end of whatever period is specified, the meter is read and the customer 13 either pays the net balance due, or the utility pays the customer for excess energy 14 delivered. The reconciliation is made without regard to when energy is produced or 15 consumed. This is how transactions between owners of DG and utilities have 16 traditionally been handled, and is the scheme which the intervenors in this case would 17 like to maintain. 18 Some call this arrangement retail net metering, and that is what I am referring to when I 19 say net metering or NEM. There are other forms of net metering such as wholesale net 20 metering, where exports into the system are compensated at the wholesale price -- often 21 LMP -- and certain of WE's tariffs have been so structured. There are other variations as 22 well, but for purposes of my testimony, when I use the terms NEM or net metering I am 23 referring to the retail variety. REBUTTAL-WEPCO/WG-BROWN-6 1 Value of Solar 2 Q. What is the "value of solar"? 3 A. There are, conceptually, four possible approaches to pricing solar PV DG. One is to set 4 the price to reflect the market clearing price in the wholesale market at the time the 5 energy is produced. That is a market approach. A second approach would be a cost 6 based approach, where the price is set based on a review of the costs or according to 7 standard costing methodology. A third approach would be, as I already discussed, net 8 metering. Finally, a fourth approach would be to administratively derive a “value of 9 solar” based on analysis of avoided costs and whatever else the evaluators believe to be 10 worthy of measure. As you will see, while I do not believe this fourth approach to be 11 appropriate, I do think that analysis of the criteria its advocates believe are important 12 should be conducted and evaluated -- not to set the price, but simply to establish the 13 context for evaluating the reasonableness of the pricing methodology approved. 14 Q. should be established by assessment of the “value of solar,” what exactly do you mean? 15 16 When you say that you do not subscribe to the point of view that distributed solar prices A. Optimally, prices for electricity are determined by a competitive market or derived from 17 cost-based regulation, and thus are essentially subjected to an external discipline that 18 should result in efficient resource decisions devoid of arbitrary or “official” preferences. 19 Subjective consideration of the “value” of particular technologies and where they may 20 rank in the merit order of “social desirability,” as proposed by certain of the intervenors' 21 witnesses, effectively removes the discipline that is more likely to produce efficient 22 results. (See, e.g., Direct-RENEW-Gilliam-6-10; 15-16; Direct-RENEW-Rabago-44-50). 23 Whereas both the marketplace and transparent cost-based regulation are likely to produce REBUTTAL-WEPCO/WG-BROWN-7 1 coherent pricing that allows us to enjoy a degree of comfort knowing that efficient 2 performance will likely lead to productivity, subjective consideration of soft criteria, like 3 “value of solar,” are a step away from coherence and reasonable predictability. 4 Q. Are you saying that economic efficiency should be the sole basis for resource selections? 5 A. I am saying that economics is critical, and that efficiency is of vital importance. I am 6 also aware and certainly agree that there are other economic values, besides efficiency, 7 especially those that go beyond short term efficiency. It is also obvious that many people 8 believe that other, non-economic factors need to be considered. Certainly the fairness of 9 the impact on customers also needs to be factored into any decisions. There has, for 10 many years, been a running debate in electricity regulation as to whether externalities 11 ought to be factored into regulatory decisions. I do not, for purposes of this testimony, 12 intend to join in that debate. Similarly, I do not want to express any point of view as to 13 what is permissible or not permissible under applicable law. 14 I would also suggest two things about consideration of externalities. If they are to be 15 considered, then all relevant ones deserve attention, as opposed to “cherry picking” the 16 issues to best protect a particular interest. Secondly, if the Commission believes that non- 17 economic objectives should be factored into its decision, then it would be wise to 18 prescribe the ways of attaining them that are the most efficient from an economic point of 19 view. 20 Q. If one were to assess all of the criteria that need to be considered, what would they be? 21 A. There are a number of them that are quite important in the full valuation of solar PV. 22 One would begin, of course, by looking at the cost of producing energy. Beyond that, the 23 criteria would include availability/capacity, reliability, energy value, impact on system REBUTTAL-WEPCO/WG-BROWN-8 1 operations and dispatch, transmission costs and effects, distribution costs and effects and 2 hedge value. Solar proponents often phrase these issues in terms of avoided costs. 3 While, as you will see in my testimony, I do not agree that many of the costs identified 4 are actually avoided, I have no problem examining the claims in the context of whether 5 costs are avoided. In addition to those dimensions, there are also the following: degree 6 of subsidization and cross-subsidization, efficiency considerations, impact on alternative 7 technologies, market price impact, reliability, and social effects including the 8 environmental and customer class impacts. There is also the issue of whether distributed 9 solar PV enhances the level of competition in the industry, or whether anything WE has 10 proposed would deny the solar industry its right to compete. 11 In the interests of full disclosure I draw many of the criteria from the writings of RENEW 12 witness Rabago. While I do not agree with the value of solar approach to setting rates for 13 solar PV DG, I generally believe that the criteria he has identified, and, in fact, testified to 14 in this proceeding are among the ones that one should contemplate if one follows that 15 approach. I was, however, disappointed that in his testimony, Mr. Rabago simply lists 16 the criteria and does absolutely no analysis to demonstrate that the actual facts justify his 17 assertions. He provides a laundry list of “benefits” he asserts WE fails to consider, but 18 offers absolutely no insight into why anyone would think solar PV DG offers any of the 19 benefits he enumerates. (Direct-RENEW-Rabago-46). He also claims that solar PV DG 20 offsets “expensive peak generation,” but provides absolutely no data to back up his claim, 21 and, as my testimony will demonstrate, that assertion is generally wrong. (Direct- 22 RENEW-Rabago-13). In fact, solar PV DG production is, as a general rule, not 23 coincident with peak. As my testimony will point out, the evidence is compelling that REBUTTAL-WEPCO/WG-BROWN-9 1 solar PV DG offers few, if any, of the benefits Mr. Rabago claims, and in some cases, the 2 purported benefits are in fact detriments. I would also note that The Alliance for Solar 3 Choice witness Hornby also calls for an assessment of the cost and benefits of what he 4 calls “customer owned generation” (Direct-TASC-Hornby-10-13), a sometime accurate, 5 but often inaccurate description of facilities that are often, in many jurisdictions, owned 6 by large solar companies that “lease” them to property owners. 7 Q. Does net energy metering (NEM) capture all of those values? 8 A. It most decidedly does not. NEM, as practiced, significantly over-values distributed solar 9 10 generation. In terms of the values enumerated earlier in my testimony, solar PV DG does the following: 11 1. Creates a cross-subsidy from non-solar to solar customers; 12 2. Fails to reflect the inefficiency of small scale solar PV relative to other forms of 13 generation, including alternative renewable resources; 14 3. Constitutes price discrimination in favor of an inefficient resource; 15 4. Significantly overvalues both the capacity and reliability value of solar PV DG; 16 5. Adversely impacts the degree of competitiveness in the industry; 17 6. Artificially inflates the transmission value of solar PV DG; 18 7. Fails to account for the fact that the value of energy varies widely depending on when it 19 is actually produced; 20 8. Distorts price signals for energy efficiency; 21 9. Causes socially regressive economic impact; 22 10. Assumes system benefits from solar PV DG that, in fact, may not exist; 23 11. Overvalues its contribution to carbon reduction; and REBUTTAL-WEPCO/WG-BROWN-10 1 2 12. Vastly inflates its value as a fuel hedge. Q. 3 4 If NEM fails to capture those values, why has it become the prevalent form of tariff for solar DG in the U.S. today? A. It was never developed as part of a fully and deliberatively reasoned pricing policy. 5 NEM was simply never a conscious policy decision. It is basically a default product of 6 two no longer relevant considerations, one practical and the other technological. The 7 practical reason is that distributed generation had such an insignificant presence in the 8 market that its economic impact was marginal at best. Thus, no one was seriously 9 concerned about “getting the prices right.” The second, technological reason is that the 10 meters most commonly deployed, especially at residential premises, until recently have 11 had very little capability other than to run forward, backward, and stop. Thus, for 12 technical reasons, NEM was simple to implement and administer, and, as a practical 13 matter, given the paucity of DG, there was no compelling reason to go to the trouble of 14 remedying a clearly defective pricing regime. Many states, however -- some of which 15 Mr. Vickerman notes at page 9 in his testimony -- have recognized the problems with 16 NEM, but seeing no alternatives to it, put in place production caps to limit any harm 17 caused by a clearly deficient pricing regime. 18 Retail net metering sets up unfair and counterproductive cross-subsidies. 19 Q. Beyond failing to capture the values you mentioned, are there other problems with NEM? 20 A. Yes. Under NEM, when DG providers export energy into the system, consumers are 21 required to pay them full retail rates for a wholesale product. What everyone agrees upon 22 is that solar PV DG provides an energy value, although there is considerable 23 disagreement about what that value is. Solar proponents argue that it has a capacity value REBUTTAL-WEPCO/WG-BROWN-11 1 as well. I believe that value, if it exists at all, is minimal, but solely for purposes of 2 answering this question I will assume some level of capacity value. Both energy and 3 capacity are wholesale products and should be compensated as such. While, as I will 4 discuss later in my testimony, there may well be reasons to treat DG differently from 5 wholesale generation for transmission purposes, there is, absent a solar host leaving the 6 grid, absolutely no reason to discriminate between wholesale and DG producers in regard 7 to the fixed costs of the distribution system and its operations. Indeed, no solar industry 8 witness in this proceeding even makes any claim about solar providers providing 9 distribution services. Under NEM, however, solar PV DG providers are compensated at 10 full retail prices for what they provide. That includes the not-insignificant cost of 11 services that they, indisputably, do not provide, including distribution costs, 12 administrative and back office operations. There simply can be no justification for 13 forcing consumers to pay a provider for service they not only do not provide, but, in fact, 14 have no capability to provide. 15 Solar PV DG producers remain connected to the grid, and are fully reliant upon it in the 16 many hours of the day when solar energy is not available. Under NEM, that solar PV DG 17 producer is excused from paying his/her share of the costs of the distribution system 18 when energy is being produced on premises. If the costs of the distribution system were 19 variable with energy production, that would be sensible, but they are not. Distribution 20 costs are fixed, and do not vary with energy production or consumption. Thus, excusing 21 solar PV DG customers from paying for their own distribution costs at the time their solar 22 units are functioning has no justification in either policy or economics. Making matters 23 worse, the costs solar PV DG providers do not pay under NEM are either reallocated to REBUTTAL-WEPCO/WG-BROWN-12 1 non-solar customers or have to be absorbed by the utility. Both outcomes are 2 unacceptable and unjustifiable. There simply is no reason why solar PV DG customers 3 should receive free backup service compliments of either their neighbors or the utility. 4 Q. Is that the cross-subsidy you referred to in your testimony earlier? 5 A. Yes, that is part of it. 6 Q. Beyond what you have already testified to, what other cross-subsidies are embodied in 7 8 NEM? A. Another cross-subsidy relates to the intermittent nature of solar energy. No utility with 9 an obligation to serve, such as WE, can be fully reliant on the availability of solar when it 10 is needed. While I will explain more about that in subsequent testimony, suffice it to say 11 that this gives rise to two types of cross-subsidy. The first arises when the distributor 12 relies on the availability of solar for making day-ahead purchases, and the other arises 13 when it does not do so. When it does rely on the availability of solar and it turns out that 14 solar energy is not available when called upon, the utility is compelled to purchase 15 replacement energy in the spot market at the marginal cost, which is almost certainly 16 higher than the price of the solar energy on whose availability it had relied. In notable 17 contrast to what happens in the wholesale market when a supplier who is relied upon fails 18 to deliver, those incremental costs have to be borne by the utility, which passes them on 19 to all customers, as opposed to being borne by the specific solar PV DG customer whose 20 failure to deliver caused the costs to be incurred. 21 If the distributor, in recognition of solar’s intermittency, instead chooses to hedge against 22 the risk of solar’s unavailability, the cost of the hedge is likewise passed on to all REBUTTAL-WEPCO/WG-BROWN-13 1 customers rather than simply those whose supply unpredictability caused the cost to be 2 incurred. 3 Both of these forms of cross-subsidy violate a bedrock principle of regulation -- that costs 4 should be allocated to the cost causer. The function of that principle, of course, is to 5 provide price signals to improve performance, but NEM fails to provide such signals and 6 essentially holds solar PV DG providers harmless for their own very low capacity factors 7 and inefficient performance. 8 Q. Are NEM cross-subsidies helpful to the solar PV DG industry? 9 A. Yes. In the short term they constitute a wealth transfer from WE's non-solar customers to 10 the solar industry. In the long term, however, they are actually harmful to solar energy 11 because NEM provides absolutely no incentive to improve the performance of a 12 generating resource that, among renewables, already ranks last in efficiency and in cost 13 effectiveness for reducing carbon emissions. In effect, the solar industry is putting its 14 short-term profits ahead of the long-term value of solar energy. If they prevail in seeking 15 to maintain NEM, that victory will be short-lived, because markets, both regulated and 16 unregulated, do not prop up inefficient resources over the long term. 17 Q. Do the WE proposals address the cross-subsidy issues? 18 A. Yes they do, in two ways. 19 First, the fixed charges being proposed for solar PV DG customers will compel them to 20 pay more of the distribution costs they currently avoid under NEM whenever they are 21 producing energy. By requiring solar customers to pay all of their share of customer 22 related fixed costs, the non-solar ratepayers will no longer have those costs re-allocated to 23 them. REBUTTAL-WEPCO/WG-BROWN-14 1 Second, the demand charge will end the cross-subsidy inherent in socializing the costs of 2 hedging against the unreliability of solar PV, or the alternative of incurring the costs 3 associated with buying energy at the marginal cost when solar panels are not producing 4 what is required. 5 It is also noteworthy that, if adopted, WE’s proposal on these two issues will adhere to 6 the regulatory principle that the cost causer pays for those costs he/she causes to be 7 incurred. 8 Q. How effective is NEM for providing appropriate price signals? 9 A. It is woefully ineffective. Electricity prices can be quite volatile over the course of every 10 day, and, of course, vary seasonally as well. Rather than reflecting those prices, NEM 11 simply treats all energy the same regardless of the time during which it is produced. For 12 example, it fails to differentiate between energy produced on peak and off peak. It pays 13 off-peak solar PV DG a price that is averaged with on-peak prices, thus effectively over- 14 valuing the energy. Conversely, if solar PV DG were actually produced on peak, NEM 15 would average that price with off-peak prices, thus undervaluing the energy. Any form 16 of dynamic pricing, ranging from time of use to real time, could address this issue with 17 more precision than flat, averaged prices. Interestingly, under the first scenario, cross- 18 subsidies would be paid to solar producers, while, in the second scenario, solar producers 19 would be cross-subsidizing the other ratepayers. In short, the price signal, and the 20 efficiency that would flow from that, is completely lost. 21 22 Q. Do you agree with the solar witnesses that, if adopted, WE’s proposals will harm the profitability of the solar PV DG industry? REBUTTAL-WEPCO/WG-BROWN-15 1 A. They may, but if solar producers improve their efficiency, these changes may actually 2 improve the industry's profitability. After all, under the WE proposal, if solar units 3 produce on peak, they will be better compensated than if they are off peak, so their 4 financial performance will be driven in large part by their efficiency as producers. 5 As a matter of public policy, however, I believe that the solar witnesses' core argument is 6 very flawed. What they are seeking is a pricing regime that has, at its core, the assurance 7 of profitability for them, without regard to productivity or efficiency. NEM accomplishes 8 that for them. Their self-serving reasoning reminds me of Charles Wilson, the former 9 Chairman of General Motors, later Secretary of Defense, who once famously stated: 10 “What’s good for General Motors is good for the country.” 11 The argument being made about the industry’s profitability is flawed on several counts. 12 Indeed, it is directly contrary to some bedrock principles of regulation and ratemaking. 13 First, seeking a pricing regime that locks in guaranteed profits, as NEM does, violates a 14 fundamental principle of regulation, namely that profits are not guaranteed but are earned 15 by performance. It also violates another key regulatory principle that pricing should 16 include balanced and symmetrical incentives for improvements in productivity and 17 efficiency, which NEM decidedly lacks. Finally, regulation needs to seek a balance 18 between the consumer interest and the industry interest, and NEM is very heavily skewed 19 against non-solar customers and in favor of the solar industry. 20 Q. 21 22 23 What about the argument that what WE is proposing would be unfair because it changes the rules on solar PV DG in the middle of the game? A. My view is that the solar PV DG industry is not only not regulated; it does not want to be. Indeed, it seeks the best of both worlds: low risk and high profits. It seeks that status by REBUTTAL-WEPCO/WG-BROWN-16 1 attempting to shield its profits from regulatory oversight while manipulating the 2 regulatory process to extract those profits from a largely captive customer base (which 3 includes both solar and non-solar customers who are compelled to pay NEM prices). 4 Given the perch of the industry in both regulated and unregulated aspects of business, it 5 should come as no surprise that the laws, rules, and tariffs do sometimes change. Indeed, 6 economic conditions and technology change, so it is simply unreasonable to expect rules 7 that last for eternity. 8 In the specific context of NEM, however, it is simply disingenuous to suggest that 9 proposals such as the ones made by WE in this proceeding are unexpected and 10 unanticipated. NEM was, as noted, never a thoroughly reasoned policy derived from a 11 public deliberation, but was simply a default policy that was convenient and 12 inconsequential at the time of implementation. It should surprise no one that once 13 dynamic or time-sensitive pricing came into existence, and once solar PV DG achieved 14 substantial levels of market penetration, changes regarding NEM would be on the table. 15 Moreover, Wisconsin is hardly the only state where this issue is on the table. It is being 16 debated in many regulatory forums, not only in the U.S., but in various jurisdictions 17 around the world. Most importantly, for reasons already noted, subsidies and cross- 18 subsidies such as NEM are almost always designed for short-term purposes and rarely 19 last forever. Finally, as the chart from Lazard found on page 23 of my testimony 20 demonstrates, the market is already aware of the high cost of solar DG PV, so investors 21 have almost certainly already factored the risk associated with high costs into their 22 expectations from the industry. The solar industry's surprise at WE's proposal is 23 reminiscent of Claude Raines in the film Casablanca, expressing complete shock that REBUTTAL-WEPCO/WG-BROWN-17 1 gambling was occurring at Rick's Place. No one can ever have any reasonable 2 expectation that a pricing regime providing an above-market price for a less efficient 3 product is sustainable. 4 Q. 5 6 Do you agree with Alliance for Solar witness Hornby that imposing demand and facilities charges on solar PV DG hosts is unfairly discriminatory (Direct-TASC-Hornby-14)? A. No. It is not unfairly discriminatory at all. Solar hosts, and I use that term to generically 7 encompass both owners and lessees of solar panels, have very unique characteristics. 8 They are users of the distribution system, either as active users or passive users (i.e., 9 reliability backup), depending on when their panels are generating energy. They also 10 depend on the system to deliver their excess output for sale. Under NEM they do not, 11 however, pay for the service the grid provides them when they are self-consuming their 12 own energy production. Moreover, under NEM, they are paid as if they were providing 13 distribution services even though they provide no such service. Those are a very unique 14 set of characteristics that I do not believe are shared by any other class of customers, and 15 that justify the imposition of a facilities charge on them that may not be imposed on a 16 different class of customer. Indeed, the failure to recognize that solar hosts are a different 17 type of customer, as is implicit in NEM, imposes more costs on non-solar customers and 18 is, in fact, discriminatory as to those customers because it forces them to pay costs which 19 they did not cause to be incurred. 20 In regard to the demand charge, again solar PV DG customers are unique. They are 21 intermittent and somewhat unpredictable generators of energy, but the utility has an open 22 ended obligation to serve them. The utility will have to buy energy at the marginal cost 23 to serve solar hosts when they fail to generate their own supply, or it will have to hedge REBUTTAL-WEPCO/WG-BROWN-18 1 against that contingency. Either way, those costs are caused entirely by the nature of 2 solar PV DG and are entirely attributable to the hosts of those facilities. It is hardly 3 discrimination to compel a customer to pay for costs he/she caused to be incurred. 4 Q. Aren't the cross-subsidies needed to promote the growth of renewable energy? 5 A. One can debate this point, but nothing in what I am saying, or in what WE is proposing, 6 would remove the tax credits and other government-sanctioned or sponsored subsidies. 7 Thus, nothing in this case will reduce the amount of direct subsidy that flows into solar 8 energy. The issue in this case is whether the value of a highly inefficient cross-subsidy 9 inherent in NEM is outweighed by the misguided price signals, regressive social 10 implications, unfair cost-shifting, transgressions against some bedrock principles of 11 regulation, and other adverse effects of NEM. As my testimony progresses, it will 12 become evident that I believe that NEM as currently designed has many more negative 13 than positive effects. Thus, even if there were a case for cross-subsidizing solar PV DG, 14 and I am not convinced that there is, NEM is not the proper vehicle for it. 15 Q. Why do you say that? 16 A. The fact that conscious subsidies and/or cross-subsidies are designed to promote a 17 particular technology raises two key issues. Many would argue that the government, 18 including regulators, should not be picking winners and losers in the marketplace. While 19 I generally agree with that premise, I can see circumstances where, for policy reasons, 20 government might want to provide support for a socially and economically desirable 21 technology and/or assist it to get past the commercialization hump. That leads inexorably 22 to the second, and, more relevant issue, concerning solar PV DG: namely, that subsidies 23 and cross-subsidies for those purposes, especially the latter, need to be designed as near- REBUTTAL-WEPCO/WG-BROWN-19 1 term boosts rather than a permanent crutch. In other words, the subsidy/cross-subsidy 2 should be designed to serve as both a stimulus for the designated technology and an 3 incentive to the producers and vendors of the technology to become more efficient. In 4 the case of solar PV DG, that means to attain grid parity with other resources. The 5 objective is to assist a technology to achieve commercial viability. The problem with 6 NEM, of course, is that it is effectively an indefinite crutch, with absolutely no built-in 7 incentive to increase efficiency and/or to achieve grid parity. In effect it requires non– 8 solar customers to pay more for the least efficient renewable resource in common use and 9 provide the solar industry with no economic incentive to improve its productivity or 10 availability or wean itself off dependence on the cross-subsidy. It is particularly telling 11 that most of the solar witnesses in this case treat NEM as a form of entitlement, since it 12 existed at the time some of the solar investments were made. In essence, they are 13 demanding a permanent, assured revenue stream from captive customers to support their 14 market position without having to improve their productivity or efficiency. 15 The reason why solar PV DG vendors and providers cling to cross-subsidies is because 16 they find more comfort in receiving substantial cross-subsidies than they do in the 17 prospect of becoming competitive. Solar PV DG is the most expensive form of 18 generation widely used today. The chart that follows illustrates that point: REBUTTAL-WEPCO/WG-BROWN-20 1 Q. Why should NEM be revisited now? 2 A. We now have pricing methods that are more capable of measuring DG production as well 3 as consumption on a more dynamic basis. In addition, solar DG market penetration has 4 dramatically increased to the point that it can no longer be dismissed as marginal, so 5 appropriate pricing is now a non-trivial issue. In addition, we now have very precise, 6 location-specific energy and transmission price signals that provide a very transparent 7 market price by which one can measure the economic value of distributed generation. 8 These new developments, plus the fact that NEM was put in place on a default basis, 9 mean that it is now time for a full-blown policy consideration of the most appropriate 10 11 12 pricing policy for distributed generation. Q. Are there reasons to discard retail NEM as you have defined it above in favor of the sorts of policies WE has proposed? REBUTTAL-WEPCO/WG-BROWN-21 1 A. Yes. There are very compelling reasons to do so. 2 Q. Please enumerate the reasons why NEM should be discarded. 3 A. The answer to this is quite simple. For all of the reasons I have noted, NEM pricing 4 results in large cross-subsidies, offers no incentives for efficiency -- indeed, may even 5 provide disincentives to invest in efficiency improvements -- and results in consumers 6 paying energy prices for solar PV DG that are far in excess of its market value and not 7 even subject to cost-based oversight. Moreover, its raison d'être -- inability to more 8 accurately price solar PV DG facilities and low market penetration by solar energy -- no 9 longer exists. Solar energy is penetrating the market in greater numbers and is likely to 10 continue to do so. Secondly, more sophisticated pricing enables us to measure solar 11 energy and customer behavior on a much more efficient, dynamic basis. The 12 fundamental reality is that NEM utterly and completely fails to capture the value of the 13 product being priced. 14 At the risk of being a bit repetitive, it is worth emphasizing just how imperfect NEM 15 actually is. The price of electric energy is not constant. Wholesale markets reflect that 16 reality. Net metering and many forms of incentives do not reflect the values established 17 by the market. Rather, a net metering regime relieves the solar panel host of any 18 obligation to pay for the costs of the distribution system when energy is being produced, 19 even though he/she remains reliant on it and, when the meter runs backwards, is 20 effectively paid the full retail price for energy exported from the customer’s premises. 21 As a point of illustration, see below for a funding mechanism for residential customers 22 presented by DTE Energy to the Michigan Public Service Commission. According to REBUTTAL-WEPCO/WG-BROWN-22 1 DTE, the 3 cent per kWh difference between the net metering credit and the unrecovered 2 fixed utilities cost represents a differential that non-participating customers must pay. 3 Under NEM, compensation at retail rates is not cost-reflective because net metering 4 means that solar DG energy exported into the distribution network is compensated at the 5 full bundled retail rate on an averaged basis rather than at a price based on the unbundled 6 cost of producing the energy at the actual time of production. Thus, it does not reflect the 7 obvious fact that the energy has greater value at peak demand than it does off peak. It is 8 a deeply flawed value proposition from an economic point of view. The fact is that the 9 wholesale market produces hour-by-hour prices that provide generators (renewable and 10 non-renewable alike) and consumers with important price signals that reflect real-time 11 values. Both generators and demand responders are compensated according to those real- 12 time prices. Solar DG-produced energy, by contrast, is compensated on a basis that lacks 13 a foundation in either market or cost. The compensation is out of market because it is a REBUTTAL-WEPCO/WG-BROWN-23 1 flat price regardless of when it is produced. It is hard to avoid the conclusion that on an 2 economic basis, the net meter-derived price paid for solar DG energy completely misses 3 the value of solar during most hours of the day. In fact, that failure is famously 4 illustrated by the following California ISO Duck Curve: 5 As is dramatically illustrated in the graph, enticed by a number of factors, not the least of 6 which is net metering, substantial investment in the growth of solar capacity in the 7 Golden State has enormously magnified the need for additional fossil plants, operating on 8 a ramping basis, to compensate for the drop off in solar production at peak. In that 9 context, the absence of any meaningful signal to make solar more efficient (e.g. linking it REBUTTAL-WEPCO/WG-BROWN-24 1 with storage) is simply something that can no longer be tolerated.1 While Wisconsin’s 2 situation is not identical to California’s, it would be pure folly for the state not to learn 3 the lesson of what has gone wrong in other jurisdictions and adopt a remedy before 4 finding itself in a similar dilemma. 5 Not coincidentally, the charts from both the California and New England ISOs (found 6 further below in my testimony), as well as that from DTE, which is in the MISO, 7 illustrate the wisdom of compensating Solar PV DG at LMP, so its price accurately 8 reflects its value at the time of actual production and avoids requiring non-solar 9 customers to pay prices for energy that far exceed its value. 10 Contrary to the intervenors' claims, the solar industry's positions are anti-competitive. 11 Q. Does distributed solar PV increase the level of competition in the industry? 12 A. This is one of the most interesting arguments put forward by many of the proponents of 13 solar PV. In theory, new actors and new technology in the marketplace should increase 14 efficiency and thereby boost competition. That, however, can only occur when market 15 prices prevail, or, where market imperfections require regulatory oversight, when cost- 16 based regulated prices are designed to provide appropriate incentives for productivity 17 gains. Were such a pricing regime in place, my answer would be yes, but, in the context 18 of NEM, the answer is absolutely not. In fact, NEM reduces the competitiveness of the 19 market. 20 What the solar industry is seeking in this proceeding, and in others like it around the 21 country, is not the right to compete, but rather the right to be compensated at arbitrarily 22 high levels justified not by market or costs, but by the highly inflated price derived from 1 For further discussion of the implications of the duck curve, see What the duck curve tells us about managing a green grid, Ex.-WEPCO/WG-Brown-2. REBUTTAL-WEPCO/WG-BROWN-25 1 net metering, a primitive and unsophisticated relic of the age of dumb prices. The effect 2 of the status quo, net metering, is to require consumers to pay a higher price for a 3 designated technology, distributed solar PV, and provide less market opportunity for 4 other, more efficient producers. Stated succinctly, the solar PV industry could play a 5 constructive role in increasing the level of competition, but to do that they would have to 6 increase their productivity and give up their demands for prices substantially higher than 7 those earned by their more efficient competitors. 8 Q. How have the other parties to this proceeding addressed this competitiveness issue? 9 A. This perverse view of competition is well articulated by Sunvest Solar’s witness, 10 Neumann, as well as RENEW Wisconsin’s witness, Vickerman, who suggests that the 11 Commission’s primary criteria for evaluating the merits of WE’s proposal should be how 12 profitable it will be for their industry. They contend that the proper price is not the result 13 of market forces, or even of cost-based analysis, but rather is whatever price protects their 14 desired margin -- a conclusion they reach by focusing on what is lucrative for the subset 15 of market participants they represent. Indeed, Vickerman actually complains that WE’s 16 proposal does not include recognition of the fact that older solar installations may be less 17 efficient, as if a competitive market would give special dispensation to a market 18 participant for its own inefficiency. (Direct-RENEW-Vickerman-14). The lack of 19 competitive balance in the arguments is also well illustrated by RENEW witness Rabago, 20 who extols the virtues of the economies of scale in solar panel production even as he 21 denies that central station generation shares such virtues. (Direct-RENEW-Rabago-13). 22 Curiously, Mr. Rabago also fails to mention that the economies of scale in energy REBUTTAL-WEPCO/WG-BROWN-26 1 production at central stations (including large-scale solar stations) are completely absent 2 in the small-scale solar PV DG installations he is promoting. 3 One other aspect of the anti-competitive nature of net metering is seen in the fact that 4 wholesale generators do not obtain any guarantee of purchase of either their capacity or 5 their energy without actually delivering energy when called upon to do so. If they fail to 6 deliver, under most capacity arrangements, they must cover the marginal costs of 7 replacing what they failed to deliver. They obviously receive no energy payments unless 8 they actually produce energy, and all payments to them are disciplined by competition, or 9 in the absence of a competitive market, by a cost-based price. While it is also true that 10 solar PV DG providers are not paid unless they deliver energy, in notable contrast to the 11 wholesale market, solar PV DG producers on NEM tariffs are compensated at levels that 12 include capacity, energy, distribution costs, and all other aspects of retail service, even 13 though the product they deliver guarantees no level of capacity and provides absolutely 14 no distribution service whatsoever. Moreover, it is not subject to competitive market 15 forces or cost-based regulatory oversight. Indeed, it is not even subject to the “avoided 16 cost” standard articulated for renewable energy under PURPA. In short, the risk- reward 17 symmetry applied to solar PV DG is low-risk, high-reward, as opposed to the more 18 symmetrical arrangements to which wholesale generators are subject. 19 Q. 20 21 Has WE, in this proceeding, proposed anything that would deny the solar industry the right to compete? A. No. In fact, the WE proposal would make the market far more competitive and would 22 produce far more value for consumers. Nothing being proposed by WE in this 23 proceeding would deprive solar generators of the right to compete. In fact, quite the REBUTTAL-WEPCO/WG-BROWN-27 1 opposite is the case. What WE has proposed would enable solar PV to participate in the 2 market on a fair, relatively level, and open basis. First, WE will continue to purchase any 3 energy produced by distributed solar that is in excess of what is required on the premises 4 of the solar host. In fact, this would provide some competitive advantage to solar PV DG 5 because it would provide an assurance of purchase without asking for a reciprocal 6 commitment of guaranteed delivery. Second, it proposes to pay all energy producers, 7 large or small, central or distributed, the market price, namely the locational marginal 8 cost. In other words, it allows the market-clearing locational price for all generators, 9 which has the advantage of bundling both transmission costs or savings and energy costs. 10 It is a rather level playing field for all generators, with a slight advantage to solar PV DG 11 because, again, it assures purchase without assured delivery. 12 Placing a Value on Solar PV DG 13 Q. 14 15 How do you respond to the intervenors' testimony on pricing, particularly Mr. Rabago’s testimony (Direct-RENEW-Rabago-44-50)? A. Needless to say, pricing is of critical importance. I intend to address that, but in the 16 context of tangible, enumerated values. I say that because there are efforts in some 17 places to attach a subjective value to solar and then derive prices from that value. I do 18 not share that point of view. I think it is best to derive prices from the values established 19 by either costs or market, not ephemeral and subjective considerations. For that reason it 20 is useful to analyze all of those considerations, so that all of the claims being made can be 21 fairly analyzed. As I have noted, however, my view on NEM is clear: it produces very 22 poor price signals, forces consumers to pay prices that exceed the value they receive, and 23 has been overtaken by events and technology. REBUTTAL-WEPCO/WG-BROWN-28 1 Q. How do you assess the availability/capacity value of distributed solar PV? 2 A. The capacity value of a generating asset is derived from its availability to produce energy 3 when called upon to do so. If a generator is not available when needed, it has little or no 4 capacity value. By its very nature, solar DG on its own, without its own backup capacity 5 (e.g. storage), can only produce energy intermittently. It is completely dependent on 6 sunshine in good atmospheric conditions. Unless sunshine is guaranteed at all times at 7 which solar DG is called upon to produce, it cannot be relied upon to be available when 8 needed. Moreover, even if all days were reliably sunny, the energy derived from the sun 9 is only accessible at certain times of the day. In many jurisdictions, the presence and 10 potency of sunshine is not coincident with peak demand. Frequently, for example, solar 11 DG capacity is greatest in the early afternoon, while peak demand occurs later in the 12 afternoon or in early evening). Two charts illustrating the lack of coincidence of solar 13 production and peak demand in New England may be found below. They are derived 14 from a 2013 Update on Solar PV and Other DSG in New England prepared by ISO New 15 England, and illustrate that the facile assumptions made by RENEW witness Rabago 16 (Direct-RENEW-Rabago-13, 46) (i.e., that solar benefits include near-term reductions in 17 peak generation) are precisely that. REBUTTAL-WEPCO/WG-BROWN-29 1 These two charts dramatically demonstrate that on the days chosen as representative of 2 summer and winter in New England, solar PV is completely absent during the winter 3 peak, reaches its peak production as peak demand is rising in the summertime, and drops 4 off dramatically during almost the entire plateau period when demand is at peak. It 5 should also be noted that on the days chosen, the sun was shining. The graph, of course, 6 would look very different on cloudy days when solar production is virtually nil. Given 7 this reality, it is very difficult to attach any credibility to Mr. Rabago’s assertion of 8 reductions at peak, or to attach much capacity value to solar PV. 9 Another graph, from EPRI, reveals the same thing on a national level: REBUTTAL-WEPCO/WG-BROWN-30 1 As noted earlier in my testimony, providers of capacity in the wholesale market may also 2 have availability issues. In their case, however, if they are not available when called 3 upon to produce, they are typically obligated to either provide replacement energy or to 4 pay the marginal cost of energy that they failed to deliver. Unless a similar obligation is 5 imposed on solar DG providers, the capacity value of solar DG is reduced even further. 6 Good pricing policy would suggest that DG prices should be fully reflective of the value 7 of the type of capacity that is actually provided. As currently implemented, net metering 8 does not adequately reflect how the capacity availability measures up to demand. 9 Additionally, solar DG remains connected to the grid, and the value of the grid should not REBUTTAL-WEPCO/WG-BROWN-31 1 be underestimated. An unplanned system of rooftop solar photovoltaics is more difficult 2 to administer than the grid that is in place. Further, solar DG may be installed in places 3 that are inefficient, generating electricity at particularly unhelpful times of day. 4 Determining the actual capacity value of solar DG is a fact-specific question that should 5 consider capacity availability resulting from timing of generation and less than optimal 6 placement of photovoltaics 7 Q. What about the availability and reliability value of solar PV? 8 A. Many solar advocates assert that solar DG enhances overall reliability because the units 9 are small and widely distributed, but close to load, and not reliant on the high voltage 10 transmission system. It is argued that they are less impacted by disasters and weather 11 disturbances. These claims are highly speculative and, for the reasons I will explain, not 12 necessarily accurate. It would be a mistake to simply assume that solar DG improves 13 reliability. 14 Solar DG is subject to disaster as much as any other installation. Strong winds, for 15 example, can harm rooftop solar as much as any other facility connected or not connected 16 to the grid. Cloudy conditions can disrupt solar output while not affecting anything else 17 on the grid. 18 Solar DG has more reliability benefit in some places than others. In Brazil, for instance, 19 a system that relies on large hydro plants with large storage reservoirs, solar has 20 considerable long-term reliability value because whenever it generates energy it 21 conserves water in the reservoirs, thereby adding to the reliability of the system. 22 However, in a thermal dominated system like MISO, where there is little or no storage, 23 reliability has to be measured on more of a real-time basis. Therefore, solar’s REBUTTAL-WEPCO/WG-BROWN-32 1 intermittency makes it unable to assure its availability when called upon to deliver 2 energy. Indeed, it is far more likely that a thermal unit will have to provide reliability to 3 back up a solar unit than the other way around. 4 It is also important to examine rooftop solar reliability issues in two contexts: that of the 5 individual customer and that of the system as a whole. Solar vendors, as part of their 6 sales pitch, claim that reliability is increased for a customer with a rooftop solar unit 7 because on-site generation provides the possibility of maintaining electric power when 8 the surrounding grid is down. When the sun is shining, this claim is likely to be true. 9 Conversely, without the sun, the claim has no validity. That argument, however, only 10 applies to the solar host. During a system outage the power inverter, an electronic device 11 or circuitry that converts direct current to alternating current, is automatically switched 12 off to prevent the backflow of live energy onto the system. That is a universal protocol to 13 prevent line workers from encountering live voltage they do not anticipate. Thus, if a 14 solar DG unit is functioning properly, when the grid goes down, the solar DG customer’s 15 inverter will also go down, making it impossible to export energy. If the solar DG unit is 16 not functioning properly, then the unit may be exporting, but will do so at a considerable 17 risk to public safety and to workers trying to restore service. The result, of course, is that 18 the solar panel provides virtually no reliability to anyone other than perhaps to the solar 19 host. There are virtually no reliability benefits for the system in Wisconsin, and therefore 20 no basis for calculating a payment for such service. 21 Q. Please continue. 22 A. Attributing reliability benefits to an intermittent resource is a stretch. By definition, 23 intermittent resources are supplemental to baseload units. The only possible exceptions REBUTTAL-WEPCO/WG-BROWN-33 1 to that are, as noted above, where there are individual reliability benefits or where the 2 availability of the unit is coincident with peak demand. Absent those circumstances, and 3 absent storage, it is almost certainly the case that the system provides reliability for solar 4 DG, rather than the other way around. That is particularly ironic given that in the context 5 of net metering, solar DG hosts do not pay for that service while generating electric 6 energy, and collect payments for distribution service they rely upon rather than provide. 7 Indeed, from a reliability perspective, net metering is truly perverse because non-solar 8 customers pay solar DG providers for reliability benefits that solar DG does not provide 9 them, while solar DG customers do not pay for the reliability benefits they actually do 10 receive. 11 From an investment perspective, solar DG pricing methods like NEM, which redirects 12 distribution revenues from utilities to solar PV providers who offer no distribution 13 services, are detrimental to reliability because they deprive utilities of the revenue needed 14 to maintain high levels of service. For utilities, the diversion of funds leaves them with 15 the Hobson’s choice of either delaying maintenance and/or needed investment, or seeking 16 additional funds -- in effect, a cross-subsidy from non-solar DG users. It is also relevant 17 to reliability to note that the prevalence of intermittent resources on the grid, including 18 solar DG, may well cause new, cleaner, and more efficient generation to appear less 19 attractive to investors. Over the long term, that effect could lead to reliability problems 20 associated with inadequate generating capacity, especially at times of peak demand. 21 Q. Does distributed solar PV avoid transmission costs? 22 A. DG advocates assert that there are real transmission savings are achieved through the 23 deployment of DG, especially in systems that use locational marginal cost pricing REBUTTAL-WEPCO/WG-BROWN-34 1 (“LMP”). (See, e.g., Direct-RENEW-Vickerman-28). The argument is that by producing 2 energy at the distribution level, less transmission service will be required, thereby 3 reducing or deferring the need for new transmission facilities. It is also often contended 4 that DG will reduce congestion costs, and perhaps even provide some ancillary services. 5 All of that is theoretically possible, but certainly not uniformly or even inevitably true. 6 Of course it is true that DG, absent any adverse, indirect effect DG might have on the 7 operations of the high voltage grid, does not incur any transmission costs in bringing its 8 energy to market. However, that is quite different than asserting that DG provides actual 9 transmission savings. In fact, it would be incorrect to simply conclude that solar DG will 10 achieve transmission savings. It is possible that there could be transmission savings 11 associated with solar DG deployment, but that can only be ascertained on a fact- and 12 location-specific basis. Such savings would most likely be derived from reducing 13 congestion or providing ancillary service of some kind. It is also theoretically possible, 14 but highly unlikely, that massive deployment of solar DG will eliminate (or, more likely, 15 defer) the need to build new transmission facilities. However, for a variety of reasons 16 including the complexities of transmission planning, the time horizons involved, the 17 complex interactions of multiple parties, and economies of scale in building transmission, 18 it is improbable that solar PV actually saves any investment in transmission capacity. 19 Indeed, a mere glance at the California ISO duck graph showing the need for ramping 20 capacity to make up for the intermittent availability of solar PV DG is almost a prima 21 facie case for believing that the opposite is true and that solar PV DG may cause a need 22 for more transmission to be built. Regardless, it is virtually impossible to demonstrate REBUTTAL-WEPCO/WG-BROWN-35 1 that solar DG will obviate the need for transmission, much less quantify the cost savings 2 associated with this purported benefit. 3 Of course, there is a simple way to calculate any actual transmission savings, and that is 4 by compensating solar DG providers in the organized markets at the locational marginal 5 cost of electricity at their location. That compensation model would have the benefit of 6 capturing both the energy value and the demonstrable transmission value of solar DG. 7 Absent that formulation, efforts to calculate actual transmission savings would be a 8 difficult task indeed. It is, of course, important to note that WE has proposed to do 9 exactly that. 10 Q. Does distributed solar PV DG avoid distribution costs? 11 A. No. It is more likely that solar PV DG will cause more costs than it saves. That is 12 because these generation sources could change voltage flows in ways that will require 13 adjustments and maintenance. It will also inevitably increase transaction costs for the 14 utility to execute interconnection agreements and do the billing for an inherently more 15 complicated transaction than simply supplying energy to a customer. It is impossible, 16 unless perhaps when a solar PV DG host leaves the grid, to envision a circumstance 17 where solar PV DG would effectuate distribution savings. 18 Regarding distribution line losses, DG offers value only to DG providers when they 19 consume what they produce because any DG output exported to the system is subject to 20 the same line loss calculations that any other generator experiences. If there were 21 locational prices on the distribution system, there might be line loss benefits that could be 22 captured by DG but, since those price signals do not exist, the argument is purely 23 academic. REBUTTAL-WEPCO/WG-BROWN-36 1 Q. Is there a fuel hedge value associated with distributed solar PV? 2 A. The theory advanced by some solar proponents is that because the marginal cost of solar 3 is zero, it serves as a hedge against price volatility. In theory that might make sense. In 4 reality, however, solar is an intermittent resource that cannot serve as a meaningful hedge 5 unless such zero-cost energy is both sufficiently and timely produced. Thus, solar PV is 6 the equivalent of a risky counter party whose financial position renders him incapable of 7 assuring payment when required. Moreover, the value of a hedge depends on the amount 8 of money the purchaser of the hedge is obliged to pay and the size and probability of the 9 price he/she seeks to avoid paying. With a NEM system (or the high-priced “value of 10 solar” approach that solar advocates seek), the price paid is highly likely to exceed the 11 fuel price most utilities would hedge against. In short, the argument ventures into the 12 realm of the absurd. It amounts to: Pay me a fixed price that is higher than the price you 13 want to avoid, in order to avoid price volatility. 14 Effects of solar PV DG on other renewable resources. 15 Q. 16 17 How does the efficiency of distributed solar PV compare with other renewables resources? A. Not well. Since 2008, as the chart below from the Energy Information Agency points 18 out, solar PV has had the lowest capacity factor of any commonly used renewable energy 19 resource in the U.S. It is also worth noting that while the overall costs of installing solar 20 panels has declined (as noted above), the productivity of solar PV has remained constant 21 at consistently low levels. REBUTTAL-WEPCO/WG-BROWN-37 1 Q. What about the claim that solar PV DG reduces carbon emissions? 2 A. The stark reality of solar PV's combination of high prices and poor capacity factor carries 3 over into the cost of reducing carbon emissions. An interesting dialogue occurred 4 recently between Charles Frank, an economist at Brookings, and Amory Lovins of the 5 Rocky Mountain Institute. Their dialogue, while contentious on many points, reflects 6 similar views on the realities depicted in the EIA chart. (Exs.-WEPCO/WG-Brown-3, 4, 7 5). Frank analyzed five generation resources by their cost effectiveness in reducing 8 carbon and concluded that nuclear and natural gas, followed by hydro, wind, and solar 9 were, in that order, the most cost-effective types of generators for reducing carbon. 10 Lovins took issue with Frank for using outdated data and for not looking at energy 11 efficiency. He also argued that nuclear ranked last in cost effectiveness, and expressed 12 some reservations about the ranking of natural gas. What is significant, however, is that 13 among renewable resources, Lovins concurred with Frank that solar is the least efficient REBUTTAL-WEPCO/WG-BROWN-38 1 renewable resource for reducing carbon.2 Thus, in the view of both men -- who hold 2 quite divergent views on how best to reduce carbon emissions -- not only is solar PV DG 3 expensive, it is the least cost-effective renewable resource for reducing carbon emissions. 4 Q. 5 6 What is the impact on other renewable resources of the higher-than-market prices paid for distributed solar PV? A. Net metering is completely inconsistent with pricing policy applied to other forms of 7 electricity generation, including non-DG renewables. All wholesale generators, 8 renewable and otherwise, have to incorporate transmission and distribution costs into the 9 price of energy delivered to customers. As I discuss elsewhere in my testimony, it is true 10 that transmission issues play out differently for distributed generation than for wholesale 11 generation. Since DG, by definition, does not rely on transmission capacity, although 12 DG might impact congestion costs in various ways, wholesale energy’s delivered cost 13 reflects transmission capacity costs while DG's does not. Thus, any competitive 14 advantage for DG on that score is quite natural. Under the net metering scheme, DG 15 providers also do not have to incorporate distribution costs into their end product, and 16 that results in a serious economic distortion of the generation market. In fact, as noted 17 elsewhere in my testimony, solar PV DG providers under NEM are actually paid for 18 delivering their energy even though they provide no such service. Wholesale generators, 19 unlike their DG counterparts, enjoy no such comparable enrichment for service they do 20 not provide. The effect of NEM’s highly inefficient and non-cost-reflective rates is to 21 distort market prices in ways that reward inefficiency and will likely deprive the 22 wholesale markets of price signals that enable construction of new, highly efficient 2 For the sake of completeness, I've included two blog posts by Frank in which he addresses some of the points made by Lovins. (Ex.-WEPCO/WG-Brown-6, 7). As Frank puts it, though, even after addressing Lovins' criticisms, "Wind continues to rank number four and solar ranks number five by a large margin." (Ex.-WEPCO/WG-Brown-6). REBUTTAL-WEPCO/WG-BROWN-39 1 generation. In addition, at a critical mass, artificially elevated solar DG prices are highly 2 likely to create distortions and inefficiencies in the capacity and energy prices found 3 within organized markets. 4 An environment with two parallel pricing regimes, one market- or cost-based, and the 5 other an arbitrary one neither market- nor cost-based, is simply economically incoherent 6 and unsustainable. WE has recognized that and proposes a more coherent and sustainable 7 pricing regime that effectively-- indeed, elegantly -- bridges the gap between DG and 8 wholesale generation by using LMP as the common denominator that provides 9 universally coherent price signals. 10 The overall effect of net metering is to increase the prices consumers pay for energy 11 overall, without any assurance of any long-term benefit in the form of increased 12 efficiency. Solar DG is artificially elevated to a preferential position above more- 13 efficient, larger scale generation, including renewables. The disparity in treatment 14 between solar DG and other forms of energy suggests that net metering is not only 15 preemption bait (as further discussed below); it is fundamentally anti-competitive as well. 16 Indeed, it compels consumers to both cross-subsidize less efficient producers and to pay 17 higher prices than necessary for energy. 18 Q. Please expand on your concerns about the effect of NEM pricing on other renewables. 19 A. Large-scale bulk power renewables (e.g. large-scale wind and solar farms, geothermal), 20 are put at a particular disadvantage by NEM pricing of solar PV DG independent of costs 21 or market for three basic reasons. First, large-scale renewables are more efficient and 22 more cost-effective than DG, yet net metering provides a subsidy to the less efficient 23 form of generation. In fact, solar DG providers are compensated for the energy they REBUTTAL-WEPCO/WG-BROWN-40 1 export at a price that can range from two to six times the market price for energy. 2 Second, in those states with renewable portfolio standards (“RPS”), the entry of a critical 3 mass of non-cost-justified solar DG units into the market could have the effect of driving 4 more efficient, large-scale renewables out of a fair share of the RPS market. Third, as 5 noted above, for renewables purchased in the wholesale market, the price paid by 6 consumers reflects all of the transmission and distribution network costs incurred in 7 delivering the energy. It is assumed for DG that there are no transmission costs. The 8 effect, in a competitive market, is to bias the market to incentivize highly inefficient 9 small-scale solar to the detriment of less costly larger-scale solar. Because locational 10 costs are not factored into the price the solar DG user is paid, such compensation is likely 11 to lead to poor locational decisions of the sort that we see in Germany. The ultimate 12 result is higher costs than necessary to keep the system reliable and stable. 13 In a similar vein, even if renewables have a salutary impact environmentally, one must 14 look at DG PV in the context of how its attributes compare with those of other 15 renewables (e.g. utility-scale solar, wind, small hydro, geothermal). Capacity factor is 16 one attribute that may be used for such a comparison; it is the ratio of actual output to the 17 maximum potential output of a particular energy generator over a period of time. 18 Q. How would you compare solar PV DG with other renewables? 19 A. An environmental analysis should include an examination of the least-cost, most efficient 20 ways to get to the desired results. Problematically, the preferential pricing of DG 21 generally, as opposed to non-DG forms of renewables, may lead to distortions that favor 22 DG over larger-scale, usually more efficient, and less costly forms of non-emitting 23 generation that will achieve more environmental benefits at lower cost. Results such as REBUTTAL-WEPCO/WG-BROWN-41 1 that cannot be justified on the basis of externalities, which are no different between DG 2 and larger-scale renewables. Indeed, it could well be argued that overpayments for DG 3 have the effect of squeezing more efficient forms of renewable energy out of RPS 4 markets by using preferential pricing to grab a disproportionate share of the RPS market. 5 In the long run, of course, the inherent favoritism in pricing DG over other renewable 6 energy sources does not bode well for the future of renewables. Discrimination in favor 7 of inefficient resources on a long-term basis is almost never sustainable. The inevitable 8 backlash in the marketplace has the potential to sweep away public support for renewable 9 energy, an outcome no one concerned about the environment would want. One of the 10 most notable ironies emanating from the use of net metering to price solar DG is that it 11 will almost certainly lead to changes in retail pricing that will undermine the promotion 12 of energy efficiency. The reason for this is that as solar DG becomes more widely 13 deployed, utilities and their regulators will likely become increasingly concerned with the 14 diminution of revenues required to support the distribution system that is caused by the 15 use of net metering. 16 Additionally, when solar DG is being self-consumed at the host premises, no revenues 17 are being paid by that host to the utility for providing what essentially amounts to a 18 battery to supplement their self-generation. Since the costs of the distribution are 19 essentially fixed and not variable with the use of “behind the meter” generation, net 20 metering results in a delta of revenue that is either made up for by non-solar customers or 21 constitutes a loss for the utility. Neither outcome is likely to be satisfactory to either the 22 utility or the regulators. Inevitably there will be ratemaking consequences. REBUTTAL-WEPCO/WG-BROWN-42 1 Q. 2 3 Is there a possibility of the states being preempted by the federal government because of inconsistent pricing regimes for DG and wholesale generation pricing? A. Yes, unfortunately, there is. Because of the economic distortions caused by NEM, there 4 are some who are calling for DG to be under the control of the Federal Energy 5 Regulatory Commission (“FERC”) rather than state public utilities commissions' 6 jurisdiction (see, for example: David B. Raskin, The Regulatory Challenge of Distributed 7 Generation, 4 Harv. Bus. L. Rev. Online 38 (2013). (Ex.-WEPCO/WG-Brown-8). 8 Unless states begin to remedy the price distortions inherent in net metering, it would be 9 surprising if many aggrieved wholesale generators did not seek relief from FERC. In a 10 somewhat analogous situation where New Jersey and Maryland sought to use state 11 subsidies/mandates to support the construction of new power plants in order to 12 manipulate and/or bypass the PJM capacity market, the FERC, in a decision (135 FERC 13 61,022, April 12, 2011) which was later affirmed by the Third Circuit Court of Appeals 14 (New Jersey Board of Public Utilities et al. v. FERC, 744 F.3d 74 (2014)), struck down 15 the state program by preemption. State Commissions that continue to prop up a net 16 metering regime with no basis in either market-based pricing or cost-of-service regulation 17 may well discover the prospect of preemption hanging over them. As a former state 18 regulator myself, I am not generally well disposed toward preemption, but while still 19 serving on the Ohio Commission, I came to recognize the likelihood of that outcome if 20 we continued in a direction that put us on a collision course with federal regulators. 21 Further foreshadowing preemption are several other examples of state net metering 22 programs running contrary to federal pricing regimes. First, the Public Utility Regulatory 23 Policies Act (“PURPA”) places an avoided cost ceiling on power purchases; net metering REBUTTAL-WEPCO/WG-BROWN-43 1 evades that ceiling. Under net metering arrangements, not only are purchases of excess 2 power mandated at levels well in excess of avoided costs, but they also include a cross- 3 subsidy from non-solar customers for the distribution costs of solar DG providers. Bulk 4 power renewables are subject to all of the rules of the wholesale market, which may 5 include such costs as congestion costs, ancillary services, penalties for no availability, 6 and others. Under net metering, solar DG providers are subject to none of these 7 disciplines. In addition, I have heard wholesale renewable generators complain angrily 8 that the arbitrarily high prices paid under net metering have the effect of attracting 9 enough solar DG providers to fill up the RPS market, so that they are being effectively 10 squeezed out of the portfolio entirely. What is particularly ironic about this is that, as 11 noted elsewhere in my testimony, distributed, small-scale solar is the least efficient form 12 of commonly used renewable energy sources in the U.S. All of these factors indicate that 13 an increasing number of parties are likely to be motivated to ask FERC to preempt net 14 metering and other state-mandated regimes that allow for unreasonably discriminatory 15 and anti-competitive pricing. 16 Q. Does distributed solar PV DG have the effect of driving down other generating costs? 17 A. This argument, which is made by RENEW witness Rabago, is a variation of the flawed 18 fuel hedge argument. (Direct-RENEW-Rabago-46). It is in fact reminiscent of a 19 vegetarian who demands to be paid by carnivores for not eating meat, because by 20 reducing demand for meat, he is providing a service to them by driving down meat's 21 price. 22 Given that solar PV DG is, as previously demonstrated, the least efficient and most 23 expensive form of commonly used generation in the U.S., this argument is, like the fuel REBUTTAL-WEPCO/WG-BROWN-44 1 hedge argument, premised on the assumption that one should pay a very high price to 2 avoid having to pay for a less expensive product. 3 On both a theoretical and practical level, this is not a serious argument. 4 Q. Does distributed solar PV have a positive environmental footprint? 5 A. Despite the widespread notion that solar energy is green, this is actually quite a 6 complicated question. Expectations of environmental externality benefits may be the 7 biggest motivator for supporting and subsidizing solar DG. Proponents of solar DG note 8 that solar has zero carbon or other harmful emissions from the process of producing 9 energy. Additionally, solar proponents contend, to the extent that wide deployment of 10 solar PV avoids the need to invest in technologies that do have carbon and other 11 undesirable emissions, there is an environmental benefit that avoids the social costs 12 associated with pollution. In the absence of legal limits on relevant emissions such costs, 13 solar advocates correctly point out, are not captured in the internalized costs of the 14 competing technologies. Therefore, solar advocates suggest that regulators and policy 15 makers should take these external social costs into consideration in setting prices for 16 various forms of energy. 17 The use of external social costs, as opposed to solely the internalized economics of 18 various forms of energy, is a controversial subject. Many oppose the use of externalities 19 as a factor in pricing because it distorts the market and necessitates social judgments 20 economic regulators may not be empowered to make. In the views of such opponents, 21 the only externalities that ought to be incorporated into pricing are those that are 22 internalized by legal mandate. Proponents of incorporating externalities into rates, on the 23 other hand, contend that doing so is the only way to accurately reflect all social costs. REBUTTAL-WEPCO/WG-BROWN-45 1 They also contend that factoring in environmental externalities is a form of insurance 2 against future regulatory requirements. My purpose in testifying is not to debate the 3 inclusion of non-mandated externalities in pricing, but rather to acknowledge that the 4 debate exists, and address the positions advanced by those who favor inclusion of 5 externalities. While this testimony takes no position as to the merits of incorporating 6 externalities into ratemaking, on the assumption that at least some regulators and 7 policymakers will look at externalities for purposes of measuring the value of solar PV, it 8 will address how externalities ought to be factored in and what externalities should be 9 considered. 10 Q. Please continue. 11 A. Before delving into this issue any further, it is important to note that the U.S. E.P.A., 12 whose jurisdiction over carbon emissions has been affirmed by the U.S. Supreme Court 13 (Massachusetts v. Environmental Protection Agency, 549 US 497 (2007)), has proposed 14 new rules under Section 111(d) of the Environmental Protection Act that would, if 15 promulgated, internalize the costs of carbon into electricity ratemaking. If this were to 16 occur, the issue of whether or not to consider the costs of carbon would no longer be 17 debatable. Thus, in the short term, there is a great deal of uncertainty, which effectively 18 strengthens the hand of those who contend consideration of carbon emissions would be a 19 form of insurance against future regulation. In the longer run, however, if carbon 20 limitations are imposed, then the cost of carbon will be fully internalized in all energy 21 resources. The effect of that, of course, is almost inevitably that carbon-free resources 22 would be of greater value, while those that emit carbon would have less value. Should 23 that happen, therefore, the market itself should produce the right price signals for all REBUTTAL-WEPCO/WG-BROWN-46 1 forms of generation, including renewable ones such as solar. At that point, special 2 programs such as RPS and NEM could actually serve to impede the most efficient ways 3 of reducing carbon, by diluting price signals that are formed with carbon control in mind. 4 Regardless, the environmental issue, in terms of solar PV DG, is how cost-effective such 5 installations are for reducing carbon. 6 I have already testified to some of the key factors in evaluating the cost-effectiveness of 7 solar PV DG for reducing carbon. There is little dispute that small scale solar PV is the 8 least efficient of all renewable energy resources in common use in this country. As 9 noted, there is even a consensus, which includes Amory Lovins, that solar PV DG is the 10 least efficient renewable resource for reducing carbon. That view is fully supported by 11 the facts in the California duck graph, as well as the ISO-New England and EPRI Value 12 of the Grid data, which demonstrate conclusively that solar PV is consistently off peak. 13 As I mentioned earlier, that means it has less value as energy, but is also powerful 14 evidence that it also has less environmental value. This is because, as a general rule, the 15 least efficient and “dirtiest” plants are dispatched at times of peak demand. Thus, when 16 solar PV DG is producing energy, it is not displacing the most carbon emitting plants. 17 Instead, it is displacing more efficient, less polluting generating units. Moreover, as an 18 intermittent resource, its availability is highly uncertain and fossil plants are often called 19 upon to operate on a less efficient, more carbon-emitting "ramping" basis than if they 20 were running as pure baseload. Thus, as a tool for carbon reduction, solar PV DG is very 21 expensive. 22 Those conclusions have also been borne out by developments in Germany. In Germany, 23 where there has been a very dramatic increase in reliance on intermittent energy, prices REBUTTAL-WEPCO/WG-BROWN-47 1 have risen since 2005 and were accompanied by a spike in carbon emissions. (See Eddy, 2 Melissa. German Energy Push Runs Into Problems. The New York Times, March 19, 3 2014. (Ex.-WEPCO/WG-Brown-9). While there are very significant differences 4 between Wisconsin, or MISO, and Germany, (perhaps most notably that Germany has 5 decided to close down its nuclear plants), the experience in that country is very telling. It 6 clearly demonstrates that an increased dependence on renewable energy resources, and 7 particularly intermittent resources, does not, as many solar proponents claim, ipso facto 8 mean fewer carbon emissions, and may, in fact, cause the opposite to occur. It also 9 demonstrates that prices will escalate dramatically if the feed in tariffs are as far in excess 10 of market as NEM prices are, as shown by the DTE graph I discussed earlier in my 11 testimony. The Germans, incidentally, have recognized their miscalculations and are 12 dramatically recalibrating their strategy. 13 Q. 14 15 What would happen to the cost effectiveness of solar PV DG for reducing carbon if WE’s proposal to use LMP for pricing solar were in place? A. The cost-effectiveness of solar PV DG to reduce carbon would improve. This would 16 occur not because solar PV would reduce more carbon, but because its value as a carbon 17 offset would be priced more appropriately. In other words, the actual price of solar PV 18 DG would more accurately reflect the value of the resource both economically and 19 environmentally. 20 Indeed, because of the nature of the dispatch model, LMP is almost elegant, in that it 21 effectively aligns both environmental and economic values. It also obviates any need to 22 debate the on- or off-peak nature of the energy being produced, because LMP 23 automatically captures that. REBUTTAL-WEPCO/WG-BROWN-48 1 Q. How does NEM distort price signals for energy efficiency? 2 A. Arbitrary price NEM does not reflect value, and it also provides absolutely no price 3 signal to customers regarding the value of demand side management and energy 4 efficiency. Compounding the problem is that the flat price paid for solar DG through net 5 metering and many forms of incentives actually deprives customers of the price signal 6 necessary for truly efficient use of energy. In fact, the subsidy inherent in NEM distorts 7 the price signal for efficiency. NEM completely distorts and dilutes the very precise 8 price signals produced by the LMP pricing in the ISO. 9 Under net metering, not only is the price likely to be higher (although in theory it could 10 also be lower) than it might be if assessed based on what is available in the market, but 11 also non-solar customers are compelled to pay both their distribution costs and part of the 12 distribution costs of solar customers. In fact, a recent study by E3 Consulting for the 13 California Public Utilities Commission projects the annual costs of net metering to be 14 $1.1 billion by 2020. This cost will have to be absorbed by the utilities, non-solar 15 customers, or both, unless there is a change in rate design. 16 Arguably, the payment of a single price for solar DG regardless of its capacity and energy 17 value may actually contribute to a need for more capacity than would be needed if the 18 price signal were more appropriate. Some may also argue that adverse energy efficiency 19 effects are a result of solar DG users (1) failing to pay for fixed costs and (2) failing to 20 recognize the costs of the solar-produced energy because the energy is subsidized by the 21 high prices the utility pays and passes on to non-solar customers through net metering. 22 The intermittent nature of solar DG (also true of wind and large scale solar) has still 23 another effect that may serve to dampen environmental expectations. Because the REBUTTAL-WEPCO/WG-BROWN-49 1 capacity required to supplement the renewable is ramping rather than baseload, the 2 signals to investors to build new, more efficient generators is diluted, and is therefore less 3 attractive from a financial point of view. (This is a particularly interesting issue in the 4 context of the California duck chart.) If the new investments are not made, then older 5 and likely “dirtier” plants will have to have their lives extended and/or be operated on a 6 ramping basis for which they were not designed. The result will be less efficiency and, 7 therefore, likely increased emissions. As discussed earlier, developments in Germany 8 have clearly demonstrated this outcome. Thus, to be truly meaningful and intellectually 9 honest, any analysis of environmental impact must take into account the change in 10 dispatch and operations 11 One of the more interesting aspects of net metering is that it subsidizes, and therefore 12 incentivizes, a highly inefficient use of solar DG. Without getting into a debate over 13 whether subsidies in general are a good or bad idea, it seems obvious that if a subsidy is 14 to be deployed to support some technology, that it ought to be designed to enable 15 resources to be more efficient and more commercially viable so that the subsidy can 16 eventually be eliminated. In the case of solar DG, one of its critical shortcomings is its 17 intermittent availability. Indeed, that intermittency not only reduces its economic value, 18 but as discussed it also reduces the likelihood of attaining the desired environmental 19 results. 20 Q. What is the social impact of substantial cross-subsidies for distributed solar PV? 21 A. There are social effects beyond the environment that have to be taken into account if 22 externalities are to be factored into ratemaking. Any failure to examine environmental 23 externalities without recognizing that there are other social externalities to be considered REBUTTAL-WEPCO/WG-BROWN-50 1 as well will yield highly skewed results. Perhaps the most important of those is the social 2 impact. 3 The social impacts of solar DG are caused by three main factors. First, as noted above, 4 solar DG users have their electricity costs cross-subsidized by their neighbors who 5 completely rely on the grid. Second, some data suggests that solar DG users are unusual 6 electricity users. Third, not everyone can afford to be a solar DG user. To address the 7 second point: unlike typical residential customers, in some regions solar DG users use 8 little or no grid power when solar production peaks, but quickly ramp up demand on 9 peak, when PV production often wanes (as is demonstrated by the charts from the New 10 England and California ISOs). Utilities must be able not only to serve full load on days 11 when solar PV is not performing, but also to ramp up resources quickly to address the 12 peak created by solar DG users. In order to ramp up as needed, utilities will purchase 13 energy at the marginal price and then distribute those costs across all users, not just solar 14 DG users. Thus, users without solar DG may be penalized for the use patterns of their 15 solar DG neighbors. A comparison of residential electricity consumers in the western 16 United States may be found in the Pacificorp chart below: REBUTTAL-WEPCO/WG-BROWN-51 1 Further, the impact of net metering is not simply the creation of a cross-subsidy from 2 non-solar PV customers to solar PV customers, but, as has been pointed out in a recent 3 study by E3, a prominent economic consulting firm, it is a cross-subsidy from less 4 affluent households to more affluent ones. Indeed, the average median household income 5 of net energy metering customers in California is 68% higher than that of the average 6 household in the state, according to the study. Energy+Environmental Economics, Inc., 7 California Net Energy Metering Ratepayer Impacts Evaluation, October 28, 2013, p.11. 8 In a recent proceeding, the staff of the Arizona Commerce Commission noted the same 9 consequence. Net metering is “Robin Hood” in reverse. In order to install rooftop solar 10 panels, often individuals must be homeowners with high credit ratings or sufficient 11 capital. Leasing arrangements are also widespread, but are generally available only to 12 customers who own their own premises, and they require the assignment of most of the 13 rooftop solar benefits to the lessor. Leasing arrangements also customarily transfer most 14 of the subsidy benefits from the solar host to the lessor. Thus, more affluent solar DG 15 customers derive disproportionately higher benefits from DG than their less affluent 16 peers. Many electricity customers, particularly less affluent ones, do not own homes or 17 lost their homes in the most recent recession. The electricity customers who are unable to 18 afford rooftop solar are forced to subsidize those who are already in a more favorable 19 financial position. Thus, it is entirely fair to characterize NEM as a wealth transfer from 20 less affluent ratepayers to more affluent ones. 21 Q. What is the impact of Solar PV on jobs? 22 A. The impact of solar PV on jobs is often cited as an externality benefit. Indeed, RENEW 23 witness Rabago does exactly that in his testimony, albeit as a mere naked assumption REBUTTAL-WEPCO/WG-BROWN-52 1 with no supporting data. (Direct-RENEW-Rabago-13, 16, 46). Any analysis of the job 2 impact must be comprehensive and not an effort to cherry-pick data. Many aspirations 3 for more jobs manufacturing PV units in the United States have not materialized due to 4 the fact that China has captured more than 50% of the market. Other impacts to be 5 considered are the effect of solar PV on electric rates and the impact of that on the job 6 market, not only in terms of what happens with rates, but also in terms of the rate 7 structure that is implemented as a result of more market penetration by solar DG. If, for 8 example, the increased presence of solar PV DG paid for through NEM artificially drives 9 up prices, there will be an inevitable adverse impact on jobs offsetting the gains Mr. 10 Rabago so cavalierly assumes. In addition, if, as illustrated by the duck chart, California 11 has to embark on significant investments or contracts for redundant generating capacity 12 because of solar PV DG’s poor capacity factor, there is likely to be a secondary negative 13 impact on jobs. My point is not that solar has an inherently adverse impact on jobs, but 14 rather that any job impact is very difficult to predict or even to gauge, and for that reason 15 ought not be determinative of anything in DG decisions such as this one. 16 Q. Does this conclude your testimony? 17 A. Yes it does. REBUTTAL-WEPCO/WG-BROWN-53