Table of Contents I. SUMMARY . . . . . . . . . . . . . . . . . . . . . . . . . .2 II. THE QF PROGRAM IN CALIFORNIA. . . . . . . . . . . . . . . .3 III. THE CRITERIA FOR A MARKET THAT IS "FUNCTIONING PROPERLY" FOR THE PURPOSE OF SETTING SRAC PRICES. . . . . . . . . . . . .6 A. SDG&E. . . . . . . . . . . . . . . . . . . . . . . . .7 B. Power Exchange . . . . . . . . . . . . . . . . . . . .8 IV. PX-BASED SRAC PRICING . . . . . . . . . . . . . . . . . . 10 A. The Use of the Zonal, Day-ahead PX Price . . . . . . 10 1. Edison / ORA's Use of Merchant Plants as the Basis for SRAC10 2. ORA's Heat Rate Cap . . . . . . . . . . . . . . 14 B. Capacity Value in the PX Price . . . . . . . . . . . 15 1. Edison / ORA. . . . . . . . . . . . . . . . . . 15 2. Power Exchange. . . . . . . . . . . . . . . . . 20 C. Other Issues . . . . . . . . . . . . . . . . . . . . 21 1. QF Administrative Costs . . . . . . . . . . . . 21 2. "Forecast Risk Adjustment". . . . . . . . . . . 22 BEFORE THE PUBLIC UTILITIES COMMISSION OF THE STATE OF CALIFORNIA Order Instituting Rulemaking into ) Implementation of Pub. Util. Code 390 ) R. 99-11-022 ) Prepared Rebuttal Testimony of R. Thomas Beach on behalf of the California Cogeneration Council and Watson Cogeneration Company Q: Please state for the record your name, position, and business address. A: My name is R. Thomas Beach. I am principal consultant with the firm Crossborder Energy. My business address is 2560 Ninth Street, Suite 316, Berkeley, California 94710. Q: Have you previously present direct testimony in this proceeding? A: Yes, I have. I sponsored direct testimony in this case on behalf of the California Cogeneration Council (CCC) and the Watson Cogeneration Company (Watson). I mailed that testimony to the other parties on February 14, 2000. Q: Are your experience and qualifications set forth in that direct testimony? A: Yes. My experience, qualifications, and prior testimony before this Commission are described in Attachment RTB-1 to my direct testimony. Q: Are you sponsoring this rebuttal testimony on behalf of the CCC and Watson? A: Yes. I. SUMMARY Q: Please summarize your rebuttal testimony. A: My rebuttal testimony addresses the following issues: 1. Criteria for a Properly Functioning Market. The direct testimonies of San Diego Gas and Electric (SDG&E) and the California Power Exchange (PX) conclude that the PX market is already "functioning properly" for the purpose of determining SRAC payments, as required under Pub. Util. Code Section 390(c). This conclusion is premature, is based on little more than the observation that the market is working in accordance with its FERC-approved tariff, and ignores important market issues that remain unresolved. 2. SRAC Prices Based on Merchant Plant Energy Costs. Southern California Edison (Edison) and the Office of Ratepayer Advocates (ORA) have sponsored testimony that would base SRAC energy payments on the energy production cost of a new gas-fired merchant power plant. This approach fails to reflect the IOUs' avoided costs, and thus violates PURPA. It also violates Section 390(c), by basing SRAC energy payments on merchant plant costs, instead of on market-clearing prices in the PX. Finally, the Edison / ORA approach would return the Commission to the days of contentious, administratively-determined SRAC prices. 3. SRAC Prices Based on Heat Rate Caps. ORA advances an alternative proposal to cap PX-based SRAC prices using the current interim SRAC formula times a multiplier based on the ratio of the heat rate of the most inefficient plant on the system divided by the heat rate implicit in the interim SRAC formula. ORA's proposal would determine the capacity value in the PX price in a manner contrary to Section 390(d). ORA itself notes numerous implementation problems with this proposal. Finally, ORA's heat rate cap will not reflect the highest marginal cost- based bids in the PX. 4. Capacity Value in the PX Price. Edison / ORA. The Commission should reject Edison's and ORA's attempts to evade the implementation of Section 390(d), which specifies how to determine the capacity value in the PX price. Edison and ORA have advanced no sound reasons why this statute cannot be implemented. My rebuttal testimony discusses why the Section 390(d) formula remains the conceptually correct way to calculate the capacity value in the PX price. California PX. The testimony of the California PX raises an issue concerning the feasibility of implementing Section 390(d). This testimony presents several workable solutions to the PX's concern. 5. Other Issues. QF Administrative Costs. The Commission should reject ORA's proposal to deduct QF administrative costs from SRAC prices. These costs do not vary depending on the IOUs' purchases of short-run energy from QFs. Therefore, they are not appropriate for consideration in the SRAC methodology. Forecast Risk Adjustment. ORA has proposed a "forecast risk adjustment" to SRAC prices, based on the worry that the IOUs are incurring additional costs in the ISO's imbalance market due to differences between scheduled and actual amounts of QF power. ORA has not examined any actual data on QF schedules versus deliveries to see if its concern has any basis in reality. In addition, ORA's adjustment could encourage the IOUs to game their forecasts of QF generation, in an effort to depress SRAC prices. This poorly-developed proposal should be rejected as well. II. THE QF PROGRAM IN CALIFORNIA Q: Southern California Edison's (Edison) witness Mr. Bergmann comments that, measured by the standard of cost-effectiveness, California's implementation of PURPA has been "a failure." Edison Testimony, at 7. Do you agree with this assessment? A: No. I am reluctant to expand the issues in this case to include a historical debate over the merits of the QF program in California. However, a response to Edison's claim is warranted because the utility uses this assertion to set the tone for its testimony. Mr. Bergmann neglects to mention what would have happened in California but for the QF program. He does quote a Commission order that describes the circumstances that gave birth to PURPA and QFs: "the severe dislocations of the U.S. electric industry stemming from the fuel supply disruptions of the 1970s." Edison, at 8. The 1970s will be remembered as the decade of the "energy crisis." Oil prices skyrocketed during the Arab oil embargo of 1973, and OPEC exercised its collective market power to keep prices high. In the electric industry, high oil prices hit ratepayers hard, particularly in states such as California where the utilities burned oil in a substantial portion of their generating capacity. The traditional alternatives to oil seemed increasingly suspect. Natural gas resources were in short supply, and gas was widely perceived to be a declining resource. The electric utilities responded to the crisis by seeking to build ever-larger central station coal and nuclear power plants. These plans foundered on safety and environmental opposition, as exemplified by the 1979 Three Mile Island accident and the California utilities' aborted efforts to build the huge Kaipairowits coal-fired plant in southern Utah. The nation needed a new approach to meeting its energy needs. That new paradigm emerged in the Public Utilities Regulatory Policies Act (PURPA), enacted in 1978 as part of the Carter Administration's energy plan. PURPA sought to reduce the country's dependence on oil through the development of new resources for electric generation, including renewable resources (solar, wind, biomass, and small hydro), waste fuels, and the more efficient use of oil and gas in cogeneration projects. In addition to fostering these new generation resources, PURPA embodied a completely new approach to the building of electric generation a change that was just as important as the technologies that PURPA encouraged. In essence, PURPA broke the monopoly that the electric utilities enjoyed in the building of new electric generation, and created a competitive industry in the development, ownership, and operation of power plants that were independent of the utilities. The utilities were required to purchase the power output of qualifying cogeneration and other small power production facilities (so-called "QFs") at "avoided cost" prices that reflected the costs that the utilities avoided by not generating or purchasing that power themselves. This mandatory purchase requirement ended the utilities' monopoly control of all generation in their service territories, and created the independent electric generation industry that is one of the foundations of today's restructured electric market. QF projects were not subject to state or federal regulation as utilities. QFs simply were provided with the right to sell power to the utilities at the utilities' avoided cost. The QF itself bore all of the risks of designing, siting, building, and operating its plant. In contrast to traditional, rate-based utility generation, if the QF did not produce power, the ratepayer bore no costs. This reform was of particular significance to state regulators such as this Commission. Regulators were painfully aware of spiraling utility construction costs for large central station power plants, and of the difficulty that regulators faced in controlling and reviewing those costs. The QF program offered a new approach: the regulator simply put out an avoided cost price signal and standardized power purchase contracts, and allowed private capital to bear the risks associated with the siting, construction, and operation of power plants. Q: What would California's energy industry look like today, but for the QF program? A: If 10,000 MWs of QF projects had not been built in the state, the state is likely to have built another five 2,000 MW central station nuclear plants on its coastline or inland waterways, similar to Edison's San Onofre nuclear plant. In October 1996, Edison filed with the Commission in A. 96-08-071its net present value (NPV) projections of the transition costs associated with its nuclear and QF resources over the period 1998 to 2030. At that time, Edison projected $4.1 billion (1998 NPV) of transition costs for its 2,200 MWs of nuclear capacity, or about $1,864 in CTC costs per kW of nuclear capacity. By comparison, Edison's expected QF CTC costs were $7.5 billion for 5,000 MWs of QF power, or $1,500 per kW of QF capacity. Compared to the other options available to California in the 1970s and 1980s, the QF program was the cost effective choice, even without including the many difficult-to-quantify social and environmental benefits of QF power. Those social and environmental benefits have been substantial. California today enjoys one of the most diverse electric resource bases in the world. Distributed QF generation helps to make the California electric grid more resilient and reliable, and has deferred the need for additional electric transmission lines. As well, due to QF power, California's environment and coastline are cleaner. Cogeneration projects use fossil fuels more efficiently than conventional utility plants, and have significantly reduced energy costs for California industry, helping those industries to remain competitive in a global marketplace. California QFs pioneered the development of new, renewable electric generation technologies using wind, solar, biomass, and geothermal resources. Companies nurtured in California's QF program including unregulated affiliates of the California IOUs have become world leaders in the development, construction, and operation of state-of-the-art electric power plants. Finally, as the Commission's "Blue Book" correctly recognized in 1994, without the independent generation industry spawned by PURPA, today's restructured electric market simply would not be possible. Q: With the benefit of hindsight, are there changes that we would make in the QF program, if we were implementing it today? A: Yes. Clearly, we would not offer QFs the ten years' of too-high fixed energy prices included in Interim Standard Offer No. 4. Q: Is Edison correct is saying that ratepayers are paying above-market costs for QF power? A: Although total QF contract costs today are above current market costs, mostly due to ongoing QF capacity payments, this is not the correct way to assess the long-term benefits of QF power, and simply creates a climate of hysteria around the issue. The relevant long- term comparison is between the costs of QF power and the costs of the nuclear plants that would have been built by the utilities, in the absence of QFs. History will judge the success or failure of the QF program in California. The Commission does not need to do so in this case. III. THE CRITERIA FOR A MARKET THAT IS "FUNCTIONING PROPERLY" FOR THE PURPOSE OF SETTING SRAC PRICES Q: Do any of the parties recommend that the Commission find that the California Power Exchange (PX) market is functioning properly for the purpose of determining SRAC payments? A: The direct testimony of San Diego Gas and Electric (SDG&E) and the California Power Exchange (PX) state that the market is functioning properly for the purpose of determining SRAC payments. With respect to this testimony, the Commission should ensure that all parties stay within the ruling which limited the scope of this phase of the proceeding to developing the criteria for evaluating whether the market is functioning properly. Phase II is designed to actually examine whether the adopted criteria have been met, and if they have, to make a "functioning properly" finding. Thus, any such finding in this phase would be premature, and would prejudice those parties who stayed within the designated scope for this phase. I turn next to the specifics of the SDG&E and PX evaluations of the "functioning properly" issue. A. SDG&E Q: SDG&E's witness L.D. Schelhorse asserts that the market is functioning properly for the purpose of determining SRAC payments. Do you agree? A: No. Dr. Schelhorse is alone among the utility witnesses in this case in asserting that the present market meets the "functioning properly" criteria of Public Utility Code Section 390(c). He bases this conclusion on ten lines of text in which he makes "observations" about the operations of the PX and ISO markets. SDG&E Testimony, at 3. Dr. Schelhorse's cursory observations support no more than a conclusion that the new markets are functioning. No one can dispute that fact. However, the Commission's task in this proceeding is to determine if the new markets are functioning properly for the purpose of determining SRAC payments. If Dr. Schelhorse were to dig a little deeper, he might discover that the simple fact that the markets are functioning is not the entire story. In particular, his "once over lightly" testimony ignores what is really happening:  The ISO and PX market monitors that Dr. Schelhorse believes are "adequate to the task" have themselves concluded that there are serious problems in the ISO and PX markets, such that the market monitors do not believe that the current markets are even "workably competitive."  Dr. Schelhorse ignores the fact that the buyers' side of the PX market remains highly concentrated, as I demonstrated in my calculation of the HHI index for the buyers' side of the PX market. He also does not comment on the fact that there is very little price responsiveness on the demand side of the market.  Dr. Schelhorse ignores the fact that the ISO has had to keep price caps in place that effectively limit the maximum prices in all of the PX and ISO markets.  Section 390(c) requires that the market must be functioning properly for the purpose of determining SRAC payments. SDG&E's testimony thus ignores the central issue of whether current market prices adequately strike the balance that PURPA requires between the interests of ratepayers and QFs. Q: Dr. Schelhorse suggests examining the correlation between the PX prices and prices in other nearby wholesale markets, such as at Palo Verde and the California - Oregon border, as the basis for assessing in Phase II of this case whether the market is functioning properly. What are your views on the merits of this criteria? A: Establishing such a correlation could be one element of the Commission's assessment of whether the PX market is functioning properly. My direct testimony proposed a similar test, based upon a broader "market basket" of prices, in order to ensure that the PX day- ahead price was broadly representative of the market for electric energy in California. I would note that Dr. Schelhorse appears to have already concluded that the market is "functioning properly," even though he does not yet appear to have performed the study that he recommends as the criteria for making that finding. Perhaps SDG&E should reserve its judgement until after it has done the study that its witness asserts is necessary to reach that judgement. B. Power Exchange Q: Mr. Kritikson of the California PX also asserts that the PX is functioning properly for the purpose of determining SRAC payments. Please respond to his testimony. A: Like Dr. Schelhorse, Mr. Kritikson spends half a page asserting that the PX market is working, trades most of the power consumed in California, and operates in accordance with its FERC-approved tariff. He does admit that "some minor concerns about market power have been expressed," but notes that the FERC has not revoked the market-based rate authority of any PX market participant. He then states that the market is not yet "fully mature," due to the "still-evolving" nature of ancillary service markets, demand-side options, the PX forward markets, reliability must-run (RMR) units, and transmission rights. He concludes that these factors do not have enough of an impact on PX prices to warrant not using the PX price for SRAC payments at this time. Again, as with Dr. Schelhorse, the Commission cannot base its "functioning properly" determination on such cursory evidence. Mr. Kritikson dismisses as "minor" the market power concerns of the PX's own market monitoring committee (PX MMC), as described in the committee's two reports to the FERC and its first annual report. He also fails to discuss the market "distortions" documented by the PX MMC, including the impact of the large percentage of zero-bid power and the IOUs' underscheduling practices. He appears to believe that the lack of demand-side price responsiveness is somehow "external" to price-setting in the PX market, even though if customers were exposed directly to PX prices and could bid their demand directly into the PX, it is obvious that the shape of the aggregate PX demand curve would change markedly. IV. PX-BASED SRAC PRICING A. The Use of the Zonal, Day-ahead PX Price Q; Do the CCC and Watson continue to support the use of the zonal, day-ahead PX market-clearing price as the basis for SRAC energy payments? A: Yes. In fact, the CCC and Watson are pleased that a wide range of parties agree that the day-ahead PX price should be the basis for SRAC energy payments. The other parties supporting this position include Pacific Gas & Electric, SDG&E, FPL Energy, the Independent Energy Producers (IEP), and the California Association of Cogenerators et al. (CAC). Only Edison and ORA appear to propose a different basis for SRAC energy payments. The California PX raises potential implementation issues associated with the use of the PX's market-clearing prices. I address below the testimony of Edison, ORA, and the PX. 1. Edison / ORA's Use of Merchant Plants as the Basis for SRAC Q: Edison and ORA both propose to use the energy costs of a new merchant power plant to set SRAC energy prices. Edison Testimony, at 36 - 49; ORA Testimony, at 40 - 48. Does this proposal comply with PURPA? A: No. The Commission is trying to set short-run avoided cost (SRAC) energy prices in this case. In the short-run, one does not consider the addition of new resources, such as new merchant plants. In the short-run, the energy costs that a QF avoids are the short-run marginal costs of the system to produce the last MWh, as reflected in the market-clearing supply bid of the last unit dispatched in the PX. A new merchant plant is very unlikely to be the marginal unit dispatched in the PX. New combined-cycle merchant plants may achieve heat rates as low as 7,000 Btu per kWh. However, the ISO system still has more than 15,000 MW of conventional gas-fired thermal generation with heat rates in the range of 10,000 Btu per kWh or higher. For the foreseeable future, these plants will remain the marginal units in the PX in most hours. The energy costs that Edison avoids as a result of QF purchases will continue to reflect the marginal costs of this large block of conventional generation. If the new capacity additions exceed load growth, then the new merchant plants may displace some older, less efficient gas-fired capacity, and thus may reduce overall price levels in the PX. However, they are very unlikely to displace all of the older units, and thus are unlikely to be the marginal, price-setting units. When new merchant plants do come on line, QFs will not avoid energy production from the new plants. New merchant plants with heat rates as low as 7,000 Btu per kWh are being built because their energy production costs will be infra-marginal, that is, much less than the energy costs of the marginal, market-clearing units. The developers of new merchant plants hope to recover their capital costs and make a profit based on the margin that they will earn between their energy production costs and the energy costs of the conventional gas-fired thermal units that will continue to set market-clearing prices in the PX. This is illustrated in Figure 1, where Generator A is an new, infra-marginal merchant plant. The new plants will operate at high load factors, as base load units, due to their low heat rates. They will rarely be on the margin in the PX and will not change their generation when a QF comes on-line or goes off-line. Thus, the energy production costs of a new merchant plant will not reflect the IOUs' short-run avoided energy costs in the PX. The Commission plainly would violate PURPA if it used such costs as the basis for SRAC energy payments. Q: Does the Edison / ORA proposal comply with Section 390(c)? A: No. Section 390(c) requires that SRAC prices "shall be based on the clearing price in the independent power exchange. . . . " (emphasis added). The Edison / ORA proposal bases SRAC prices on the production costs of a new merchant plant, not on the PX market- clearing price as the law requires. Edison and ORA attempt to cover their failure to comply with Section 390(c) by calling their approach an "energy %" (Edison) or "an energy value multiplier" (ORA) to the PX price. Edison and ORA would calculate such percentages or multipliers as the ratio of a new merchant plant's energy cost to the average PX price over some prior period. They would then multiply this "energy %" or "energy value multiplier" by the PX price in an hour to determine the SRAC price in that hour. In effect, all that Edison and ORA are using the PX price to do is to escalate the merchant plant's energy costs over time. The merchant plant energy costs are simply escalated by the ratio of the current PX price to the base period PX price used to calculate the "energy %" or "energy value multiplier." Thus, Edison's and ORA's SRAC energy price is based on a merchant plant's energy cost, not on the PX market clearing price as Section 390(c) requires. At times, Edison does not even use the fig leaf of an "energy %" to cover its naked attempt to circumvent Section 390(c). The heading of this section of Edison's testimony states candidly that "The Avoided Cost of Energy Paid to QFs Should Be Based on the Variable Operating Cost of the New Market Entrant" (emphasis added). Edison, at 43.The variable operating cost of a new merchant plant is not the basis for the market- clearing price in the PX, and thus the Edison / ORA proposals do not comply with Section 390(c). The energy costs of a new merchant plant also are not the costs that the IOUs avoid through their purchases of QF power; thus, the Edison / ORA proposals do not accurately determine SRAC prices. Q: Would the Edison / ORA proposals to use a merchant plant's energy production costs simplify the determination of SRAC prices? A: No. Their approaches would continue the administrative determination of SRAC prices with which the Commission has struggled for the past two decades. Contrary to the assertions of Edison's witness Mr. Davis and ORA's witness Ms. Sabino, setting SRAC prices on the basis of a merchant plant's energy production cost will not be straightforward. Edison, at 45-46. All merchant plants are not the same, as ORA admits: "different power plants have different cost streams which may entail varying estimates of the new entrant's marginal costs." ORA, at 47, lines 8-9. Among the issues that would have to be determined include: What plant or plants among the 15,000 MWs of merchant plants proposed in California should be used to set SRAC prices? ORA admits that "technology and resource will thus affect the new generator's marginal cost ratio to the PX price." ORA, at 47, lines 9-10. What heat rate should be used? Table 1 on page 46 in Edison's testimony shows a range of 6% between the low and high heat rates for the same plant (La Paloma). What are the variable O&M costs? Edison's Table 1 shows a variation from $1.1 per MWh to $2 per MWh. What gas cost is appropriate for the merchant plant's fuel cost? As shown in Table 1 of Edison's testimony, gas costs for merchant plants can vary significantly by location and by the gas utility or pipeline that serves each plant. Edison's testimony admits that the choice of which gas cost to use can cause the "energy %" to vary by 10 - 15% due to this factor alone. Edison, at 47, lines 13-16. Gas costs also vary over time: the Commission will recall the complexities and controversies surrounding the "index methodology" used in the early- and mid-1990s to determine avoided gas costs. For determining the "energy %" or "energy value multiplier," what base period should be used to set the percentage or multiplier? The Commission has sought to simplify the SRAC price-setting process, and to move from an administrative to a market-based determination of SRAC energy prices. The current interim SRAC formula has represented major progress in this regard, and the use of PX market prices to set SRAC prices offers the potential to keep the SRAC determination in the market and out of the Commission's courtrooms. The Edison / ORA proposals are a step in the wrong direction: they represent a return to the "SRAC Wars" of the past, and would invite regular, contentious administrative proceedings to set and revise merchant-plant-based SRAC energy prices. The Commission should consider carefully whether it wants to choose that litigious path. 2. ORA's Heat Rate Cap Q: ORA also advances an alternative proposal to cap SRAC energy prices using a so- called "heat rate cap." ORA, at 41 - 43. Please present your evaluation of this proposal. A: ORA's alternative proposal would cap PX-based SRAC energy prices based on a proxy for what ORA believes to be "the highest production cost of energy." ORA, at 41.Although ORA's proposal is not entirely clear, ORA appears to propose to cap PX-based SRAC energy prices at the product of the interim SRAC formula price times a multiplier. The multiplier would be the ratio of the heat rate of the least-efficient generator on the system divided by the heat rate implicit in the SRAC formula. In any hour in which the PX day-ahead price is above the cap, the amount in excess of the cap would be deemed to be the capacity value in the PX price, and would be excluded from the SRAC price. There are numerous problems with this proposal. Section 390(d) certainly does not provide that the capacity value in the PX price should be established using the interim formula specified in Section 390(b). The clear intent of Sections 390(c) and (d) is to supersede completely the formula specified in Section 390(b). Furthermore, ORA's testimony on pages 42 - 43 candidly acknowledges the many implementation problems with this proposal: Access to heat rate data is now difficult, as the IOUs have divested all of their conventional gas-fired plants. Picking the least-efficient unit will be controversial. This method assumes that the interim SRAC formula continues to reflect the heat rate profile of the system. This approach would continue to require administrative proceedings to determine and then revise the heat rate cap. Finally, I question the accuracy of the proposed cap as a measure of the "highest production cost of energy." Bids based on marginal energy costs easily can exceed such a cap, as bidders seek to recover their variable O&M costs during summer peak periods. I discuss this point further in the next section. B. Capacity Value in the PX Price 1. Edison / ORA Q: Edison's and ORA's testimonies criticize the use of the method specified in Section 390(d) for determining the capacity value in the PX price. Edison calls the Section 390(d) method "the capacity subtracter" approach, and argues that the approach will not yield "reliable" results. Edison, at 24 - 25. ORA says that the Section 390(d) formula is "an unreasonable method of determining capacity value." ORA, at 34.What are your reactions to these critiques of Section 390(d)? A: I would observe that Section 390(d) is a statute that this Commission has no discretion to change. If Edison does not want to use "the capacity subtracter" method, then its recourse should be to the Legislature, not to this Commission. Q: Edison asserts that the Section 390(d) formula has always yielded a capacity value of zero. Edison cites this as evidence that Section 390(d) is "patently unreasonable" and as "the clearest refutation of the validity of the formula." Do you agree with these assertions? A: No. Edison's own exercise of market power in the PX is the reason that no capacity value has been observed in the PX to date. My direct testimony summarizes the compelling evidence from PX bid data and from the ISO and PX market-monitoring reports that the IOUs the dominant buyers in the PX market have artificially underscheduled demand in the day-ahead PX market, in order to limit day-ahead PX prices. This practice ensures that the PX supply and demand curves cross, and that there is no capacity value in the PX price. Thus, the IOUs only have their own market manipulations to blame if, to date, the Section 390(d) formula would have found no capacity value in the PX price. Table 10 of the March 9, 1999 PX Market Monitoring Report (included as Attachment RTB-3 to my direct testimony) cites 116 hours in 1998 in which the aggregate PX supply and demand curves would not have crossed, but for the IOUs' underscheduling practices. In these hours, but for IOU underscheduling, the Section 390(d) approach might well have determined a non-zero capacity value in the PX price. This indicates to me that in a properly functioning market, without IOU underscheduling, the Section 390(d) formula may well produce a significant number of hours in which there is a non-zero capacity value. Q: Edison complains that the assumptions underlying Section 390(d) were not implemented in practice. Edison offers the lack of demand-side price responsiveness as its first, "and perhaps most important," example of such changed assumptions. Do you agree? A: To a point, I do. The CCC and Watson agree with Edison that the market presently lacks adequate demand-side price responsiveness. However, this is a "functioning properly" issue, not a Section 390(d) concern. Once customers have developed greater price responsiveness to demand, and once the market is functioning properly in this regard, then Edison's concern on this score should be alleviated. Edison also points to the fact that the market design for bidding in the PX changed after the passage of AB 1890, which included Section 390. Yet Edison fails to explain how this change has made the implementation of Section 390(d) infeasible. The change that Edison cites simplified the PX bidding process from the use of multi-part bids to single-part bids. Even after this change, however, the PX market design continued to set a market clearing price based on supply and demand bids. Thus, there continues today to be a basis to calculate the capacity value in the PX price using the Section 390(d) formula. Furthermore, as I discuss below, the Section 390(d) formula remains the economically correct way to determine the capacity value in the PX price. Q: Did the settlement that the CCC and IEP signed with Edison in 1996 include the formula for the capacity value in the PX price that AB 1890 codified in Section 390(d)? A: Yes, it did, and the CCC is disappointed that Edison apparently has decided not to respect this settlement agreement (see Attachment RTB-6 of my direct testimony). Edison's testimony acknowledges that Section 390(d) incorporated verbatim the language of this settlement concerning the calculation of capacity value in the PX price. The settlement contained certain market attributes as conditions on the parties' support of the deal: The ISO separate from the PX; A uniform market clearing price; Competitive procurement of ancillary services; Maintenance of the requirement that the IOUs buy and sell all of their power out of the PX for the period required in D. 95-12-063, with significant purchases out of the PX thereafter; and Full utility recovery of QF-related transition costs through the Competition Transition Charge (CTC). All of these conditions have been met, including Edison's ability to recover 100% of QF- related CTC. The multi-part PX bidding protocol, whose demise Edison cites as the reason for not respecting either the settlement or Section 390(d), was not one of the conditions precedent for the effectiveness of the settlement. Edison has obtained a significant benefit from the 1996 settlement and AB 1890 the assurance of full recovery of QF CTC. It is unfortunate that Edison is unwilling to uphold its end of the bargain with respect to the Section 390(d) formula. Q: Edison claims that there is "strong empirical and anecdotal evidence of a capacity value in the market-clearing price." Please comment on this evidence. A: Edison believes that the variable cost of the most expensive unit selling power into the PX is no higher than $90 per MWh, without citing any actual data from a real generator to support this belief. There is no evidence before the Commission that Edison's $90 per MWh limit accurately captures how marginal units will bid their costs into the market. In my opinion, a marginal cost-based bid could easily exceed $90 per MWh. Gas-fired units can have significant O&M costs that vary with the number of hours they operate, including overhaul costs, catalyst replacement, and hot path consumables. Marginal generators must recover these variable O&M costs in the hours of peak summer demand. Thus, a generator may well bid very high prices but still his marginal costs during peak summer hours. For example, Edison cites 186 hours in 1998 and 1999 with day- ahead PX prices in excess of $90 per MWh. Assume that a generator's fuel costs are $45 per MWH, and his avoidable O&M costs are $1 per MWH when spread over all hours of the year. If this generator tried to collect these costs over just 186 hours, he would bid $82 per MWH above his fuel costs, or $127 per MWH, in these hours. Such a bid is well above Edison's arbitrary $90 per MWH, yet would fully reflect the generator's marginal costs. Edison also cites statements by the new owners of Edison's divested plants to the effect that "today's generators must recover their fixed operating costs and a return on their investment solely through the market based energy prices." It is important for the Commission to appreciate why this statement does not support Edison's suggestion that PX market-clearing prices necessarily contain capacity value. If a generator bids his marginal cost of production and those costs are less than the market-clearing bid, then the generator will be dispatched, and will earn a contribution to his fixed costs equal to the difference between his bid and the market-clearing price. That is how generators are able to recover fixed costs in the PX by producing power at a cost less than the marginal, market-clearing price. Figure 1 illustrates this basic fact; in the figure, Generator A earns a contribution to his fixed costs (labeled "margin" in the figure) by producing power at less than the market-clearing price. If the PX market is workably competitive (i.e. "functioning properly"), a generator will have no incentive to bid higher than his marginal costs, because by so doing he will risk not being dispatched and losing any ability to make a contribution to his fixed costs. In a properly functioning market, generators will maximize their revenues by bidding at their marginal costs. Edison also cites statements by merchant generators that they bid into the PX not their fuel costs, but their opportunity costs in other markets. If the PX is functioning properly, efficient arbitrage should leave little difference between the day-ahead PX price and prices in the other markets in which generators may sell their output. Thus, there should be no substantial differences over time between the day-ahead PX price and generators' oppor- tunity costs in other markets. In any market, generators have no incentive to bid above their marginal costs if they want to be dispatched in a way that maximizes their revenues. Q: Setting aside the facts that it was the product of a settlement and that it is now the law, do you think that the Section 390(d) formula is the appropriate way to calculate the capacity value in the PX price? A: Yes. As I have discussed, in a properly functioning market, suppliers will bid their marginal costs of production, in order to maximize their revenues. If the aggregate supply curve falls below the demand curve at all points, then the market will clear on a demand- side bid at the level of generation of the highest supply bid. The difference between the market-clearing demand bid and the highest supply bid dispatched represents the additional value to the buyer of ensuring that he is supplied with power. In other words, it represents the value to the buyer of the system's capacity to provide him with power. It is the capacity value in the PX price. If the supply and demand curves intersect, then the value of power to the buyer is the same as the marginal cost of producing another MWh of energy, and the capacity value in the market-clearing price is zero. More generally, both Edison and ORA commit a fundamental error in trying to equate the capacity value in the PX price with the fixed costs that a generator is recovering through its PX sales. This reflects an outdated view of the nature of capacity. In today's market, capacity is not a category of costs, it is a service the ability to call on a generator to deliver energy when it is needed. As noted above, when the demand side is willing to pay more for electricity than the highest supply bid, then the difference between the demand- and supply-side bids is capacity value, because if reflects the premium that the buyer is willing to pay to ensure that he receives electricity. Supply bids in the PX reflect only energy, because they capture only the cost of supplying another megawatt-hour of energy to the market. The value of the capacity to supply energy is best captured outside the PX, in the ISO's capacity markets for generating reserves. 2. Power Exchange Q: Mr. Kritikson of the PX raises a practical difficulty with calculating the capacity value in the PX price as specified in Section 390(d). This problem arises from the constraints imposed by the current software that the PX uses. Do you have a solution to this problem? A: Yes. Mr. Kritikson notes that the PX requires market participants to bid piece-wise linear supply curves, with the final point on the curve being the amount that the generators would supply at the maximum PX price of $2,500 per MWh. He observes that this constraint on PX bids will guarantee that the supply and demand curves cross (see his Figure PX-2). He also notes that generators do not necessarily have to bid a vertical supply curve for the final segment that ends at a price of $2,500 per MWh. He opines that this circumstance could make it difficult for the PX to calculate the capacity value in the PX price, and could allow bidders to seek to "game" the Section 390(d) determination. I would note that, unless QFs with a direct interest in the Section 390(d) value are bidding into the PX, suppliers do not have an interest in bidding a non-vertical final segment of their supply curve. Assuming that the last price point before $2,500 per MWh on their supply curve represents their marginal cost at their highest possible production, supply bidders will sacrifice some potential production if the final segment is not vertical. There appear to be several solutions to this problem. The first would be for the PX to define a slope to the aggregate supply curve that it would consider to be vertical (for example, $100 per MWh per MW). If there continue to be gaming concerns with this definition, then another approach that definitely solves the problem is simply to break the supplier bid segment that clears the market into a "stair-step" of horizontal and vertical components, for the purpose of the Section 390 determination. Then the capacity value in the PX price simply will be the difference between the market-clearing demand price and the price at the base of the vertical portion of the market-clearing supply segment. See, for example, Figure PX-3 of Mr. Kritikson's testimony. As Mr. Kritikson notes, the ISO's congestion management process uses such "stair-step" bids. Thus, this adjustment will only be required to calculate the Section 390(d) value in unconstrained hours. The CCC and Watson would support allowing the PX to recover from the IOUs the PX's cost for the software modifications required to comply with Section 390(d). C. Other Issues 1. QF Administrative Costs Q: ORA proposes to deduct the IOU's costs of administering QF contracts from SRAC prices. Is this appropriate? A: No. The IOUs have administered QF contracts for almost two decades. This is the first time, to my knowledge, that such a suggestion has been made. Unfortunately, the QF contracts are in place, and the IOUs' costs to administer them will not change depending on QF energy production in the short run. QF contract administration costs simply are not short-run avoided energy costs, so long as the contracts remain in place. ORA's testimony on this issue clearly confuses short- and long-term avoided costs. ORA, at 18, lines 13-14. ORA is also unclear as to the magnitude of the IOU's QF administration costs, and exactly how those costs would be subtracted from energy prices. 2. "Forecast Risk Adjustment" Q: ORA recommends adoption of what it calls a "forecast risk adjustment." ORA, at 15-17. ORA recommends that the Commission develop the specifics of such an adjustment in Phase II. Should the Commission pursue this issue in Phase II? A: No. This adjustment is designed to compensate the utility for the price risk that the IOU bears due to the uncertainty in whether the QF will actually deliver the power that the utility has scheduled through the PX and ISO. ORA admits that it does not even know if there is a problem with the IOUs bearing significant price risk due to differences between forecasted and actual QF deliveries; ORA simply has not done the work to examine the necessary data. ORA, at 17, lines 1-2. I doubt that the difference between scheduled and actual QF deliveries amounts to a signi- ficant price risk for the IOUs. When there is a difference, it is reconciled at the ISO real- time price. If the PX day-ahead and ISO real-time prices converge over time, as they should in a properly functioning market, then there should be no systematic price risk to ratepayers from such differences. In addition, I disagree strongly with ORA's assertion that QF deliveries are less firm than PX power. ORA, at 15, line 27. If a PX generator fails to produce the power it scheduled, the generator simply must pay to replace the scheduled generation with imbalance energy at the real-time price. A QF has a stronger incentive to produce power, particularly during on- and mid-peak periods: it faces the loss of capacity payments and, in the summer, risks being put on probation unless it produces. Finally, I am concerned that ORA's proposed adjustment would open the door to IOU manipulation of its QF schedules, in order to create a downward "forecast risk adjust- ment" to SRAC prices. For example, an IOU could underschedule QF power whenever it expected high real-time prices, in order to establish a "forecast risk adjustment" less than 1.0. Q: Does this conclude your rebuttal testimony? A: Yes, it does.