7. Resource Issues
7.1. Monthly Peak Method vs. Load Shape
Method
While the nature of the RA obligation was addressed in D.04-10-035, the Phase 2 workshop discussions revealed deep divisions about an important aspect of the obligation. Specifically, the issue of whether LSEs should be required to acquire capacity to meet each month's peak demand for all hours of that month remains unresolved. By ruling dated April 7, 2005, the ALJ scheduled additional workshops to consider this issue, deferred issuance of the Phase 2 workshop report, and rescinded the previously established procedural schedule to accommodate the additional workshops.
Two general approaches to defining the RA obligation emerged from this process. The first, supported principally by the CAISO and by generator parties, would establish each LSE's procurement obligation for each month as the LSE's peak day load for that month, measured in megawatts (MW), plus 15%. This monthly peak approach was commonly referred to as the "top-down" (TD) method.
According to the workshop report, the TD method recognizes that many of the qualifying resources will not be available in all hours of the month. The report describes two alternative proposals for implementing the TD method. The IEP alternative would create special rules with limits on the CAISO's ability to call on energy-limited, environmentally limited fossil, pumped storage hydro, non-pumped storage hydro, Qualifying Facilities (QFs), and intermittent resources. Another alternative for the TD method, advanced by Mirant/WCP, would create interim exceptions for limited availability contracts with specific resources or portfolios of resources and standard delivery attributes (e.g., 6x16 firm energy contracts), limited availability call contracts with specific resources or portfolios of resources and non-standard dispatch attributes (e.g., 6x16 options with two-day-ahead call), and firm liquidated damages contracts for delivery within California. Mirant/WCP advocate a maximum cumulative contribution of specified resource categories with physical and contractual availability limitations to alleviate over-reliance on resources that could not be counted on to serve a large portion of a month outside of the peak period. There would be no need in the LSE compliance showing to acquire additional resources to "make up" for those resources that are not available at all hours due to allowable constraints.
The second general approach to defining the RA obligation, supported by LSEs and their customers as represented by Joint Parties as well as SDG&E, is based on calculating each LSE's load duration curve to determine the hourly resource need (in MW) for each hour of each month, then adding 15% across all hours. This LSE-specific resource duration curve approach is commonly referred to as the "bottom-up" (BU) method. A resource eligibility factor (REF), a proposed measure of the percent of time resources can be counted against an LSE's RAR, would be used to compare the LSE's resource portfolio to the resource duration curve. Non-energy-limited resources without planned outages would have a 100% REF, i.e., they would count towards an LSE's monthly RAR 100% of the time. Certain energy-limited resources could meet the 100% REF standard in specified conditions; otherwise, however, energy-limited resources would not be able to qualify for a 100% REF.
Apart from the TD and BU methods, no other alternative means of defining the RA obligation were proposed. We must adopt one of the two methods so that the RAR program may go forward. Accordingly, we will determine the method that appears to more effectively promote our policy objectives for RAR.
Availability of Resources to the CAISO - Proponents of TD argue that their approach meets CAISO's needs since the availability requirement is similar to the current MOO mechanism. The CAISO also contends that the TD method is more consistent with its operations because resources are offered for all hours they are physically capable of running (subject to environmental and other regulatory limitations). TD advocates note that under the BU approach, resource availability is limited not only by their physical capabilities but also by contract, creating a possible impact on the CAISO's ability to optimize resources.15 The CAISO also argues that the BU approach may limit the CAISO's ability to respond effectively to energy deficiencies because such deficiencies can occur in any hour, and the BU method does not require resources to be available outside of their contracts.
BU advocates on the other hand maintain that their approach not only ensures an adequate supply of resources in all hours to meet load plus reserves, it ensures an appropriate resource mix and reveals planned outages in all hours. They argue that under the TD method, it is unclear how RA is impacted by use limitations in off-peak hours.
The major availability advantage that we see for the BU approach-that it would reveal planned outages in all hours-is outweighed in our view by the major availability advantage that we see for the TD approach-that resources are available to the CAISO by rule and, increasingly as non-unit specific contracts are phased out, not restricted by contract terms. Another reliability advantage claimed for the BU approach is that it would better preclude over-reliance on energy-limited resources. However, the CAISO notes that high load periods can occur during off-peak times, especially on Sundays. CAISO is concerned that under the BU approach, the load duration curve could indicate that load would be satisfied during these times by a peaking resource that may not be available due to contract terms.
We find that on balance, the TD approach is likely to be more effective than the BU method in terms of ensuring that resources are available to the CAISO.
Joint Parties point to another form of availability problem. They contend that the BU method would lessen the chance that the CAISO would face circumstances where it would have to manage an excessive amount of generation during off-peak hours. Joint Parties believe that excessive generation could lead to minimum load conditions in which the sum total of operating generation exceeds load, which in turn could lead to congestion that would undermine reliability. PG&E makes a similar point. Neither Joint Parties nor PG&E provide us with adequate information to assess the frequency with which minimum load conditions might occur, how severe their reliability impacts might be, or how much more effective the BU method would be in preventing it than TD would be. Accordingly, while we do not discount the possibility of minimum load conditions occurring, we are not presented with an adequate basis for preferring the BU approach because of possible minimum load conditions. We find it significant however, that the CAISO itself has not raised this as a significant concern.
Infrastructure Investment - The other principal means by which we seek to promote reliability through RAR is establishing an environment conducive to investment in the resources needed to serve retail load in the IOUs' service territories. As discussed earlier, we have determined that a capacity-based RAR program that encourages forward commitments is a key to accomplishing this.
As the workshop report notes, the BU method and its use of an LSE's load duration curve would serve to promote the creation of differentiated capacity products. On its face, this appears to be inconsistent with development of a standard capacity product. TD advocates, on the other hand, argue that their approach ensures that appropriate forward capacity demand signals are registered in the marketplace as soon as possible, resulting in price signals that will establish the true value of capacity.
We conclude that the TD method would do more to move LSEs towards adopting capacity products and away from mixed products for which the capacity value is at best implicit within the energy value but not recognized or known. Moving toward a rational pricing approach for capacity, where the true market value of capacity is revealed, should provide the appropriate incentives for needed investment to occur.
While we have not determined that a centralized capacity market should be developed, we have determined that the question of whether to do so should be studied in additional proceedings. We have also determined that actions we take here should not preclude development of a capacity market. To the extent that use of a resource duration curve to define RA obligations promotes the development of differentiated capacity products, the BU approach may hinder development of a capacity market.
Cost Differences - There are two dimensions to considering the cost differences between the TD and BU methods. Not only should we consider whether the costs that accrue to individual LSEs and their customers are lower under one of the methods, we should also consider whether one method or the other yields lower overall costs for all participants. The Commission noted in D.04-07-028 that LSEs were able to schedule infeasible and undeliverable energy contracts, and that the CAISO therefore had to incur congestion, must-offer, and re-dispatch costs to serve load. We seek to avoid such outcomes except in circumstances where the CAISO is able to undertake procurement at demonstrably lower total cost and assign those costs to those who cause them to be incurred. We understand such circumstances to be extremely limited at best.
One of the advantages claimed for the BU approach is that it is more compatible with the existing procurement strategies of LSEs. The converse of this argument is that the TD method may require LSEs to revise those strategies at potentially higher cost. However, to the extent that current LSE procurement strategies yield cost savings because they fail to provide adequate opportunities for the recovery of investment costs, such savings may be illusory, and in any event they would be inconsistent with the achievement of our RAR objectives.
A major concern raised regarding the TD proposal is that it would lead to significant and costly over-procurement for off-peak periods. We think that this concern is overstated, and it may miss the point that the RAR program is intended to promote needed infrastructure investment. As the CAISO pointed out, the value of capacity in off-peak times is minimal. Moving to a 24x7 obligation should not significantly increase costs because sellers are unlikely to recover significant capacity value in the off-peak periods. Therefore, any cost benefit in favor of the BU approach is unlikely to be meaningful. In any event, as the workshop report correctly observed, if there is a substantial difference between the TD and BU approaches, it is because of the former's provision for fixed costs being paid to suppliers providing needed capacity. To the extent that resource owners are currently absorbing the cost of making capacity available without adequate compensation, and the TD approach is more effective than the BU approach in providing for adequate compensation, then any higher costs of the TD approach that are due to providing such adequate compensation are consistent with our RAR objectives.
Joint Parties argue that if the ability of LSEs and existing California generating resources to engage in exchanges or off-system sales is undermined through the imposition of a must-offer requirement that extends to all hours, then the imposition of the TD method would result in substantial revenue losses and possibly a substantial increase in costs to meet needs that otherwise would have been met through exchanges. We find this concern to be overstated as well. As the CAISO and others have noted, sellers are unlikely to recover significant value in off-peak periods. Moreover, the exchanges at issue are short term in nature and unlikely to be major components of RAR portfolios.
Implementation - This issue boils down to which method was more fully developed in workshops and is therefore more ready to be implemented. We do not find that either method has a clear advantage with respect to either ease of or readiness for RAR program implementation. On the one hand, the TD approach is aligned with practices in eastern markets, and the lessons learned from those markets may be helpful to some participants in dealing with California's new program, including the CAISO itself. On the other hand, California LSEs are accustomed to operating in an environment that resembles the BU method, and implementation may be easier for them under BU. As we noted earlier, however, preservation of the status quo is not a reason to refrain from pursuing our adopted RAR policies.
Compliance - Compliance topics in general received relatively little attention in the Phase 2 workshops, and it is probably for this reason that no party presented persuasive arguments that one method or the other should be preferred because of compliance considerations. We note that the TD method is inherently simpler in that the key indicator of compliance is whether the LSE's resources are adequate for the monthly peak, whereas for the BU method compliance will have to be assessed across all hours.
Procurement Policies - ORA contends that the BU method is consistent with integrated resource planning and the loading order of the Energy Action Plan (EAP). ORA notes that the resource duration curve approach reflects the quantity and mix of resources that the Commission has ordered the IOUs to procure and can account for changes in the quantity, priority, and mix of resources in the IOUs' portfolios. By contrast, ORA contends, the TD method does not address off-peak resources, and exception rules would likely be static.
ORA's arguments strike us as little more than a variation of the status quo preservation approach discussed earlier. We see no compelling reason to conclude that movement to a capacity-based resource adequacy system significantly or unduly impedes progress in the implementation of the EAP loading order.
Conclusion - While the TD and BU methods appear in stark contrast to each other conceptually, their differences diminish when practical implementation concerns are introduced. Whether exceptions are adopted in connection with the TD method or an REF mechanism is adopted in connection with the BU approach, both methods have to address the fact that not all resources are available 100% of the time.
Nevertheless, differences remain. The BU approach is more in line with current LSE procurement strategies, but we are concerned that those very strategies may be at odds with RAR objectives. We find that the TD approach is more closely aligned with our policy objectives for RAR and should therefore be adopted. We take this action not because it is essential to the development of a centralized capacity market, as some suggest, but rather because it is essential to carrying out the policies adopted in D.04-01-050 and D.04-10-035.
We adopt the alternative version of the TD method suggested by Mirant/WCP. This alternative provides for exceptions that reflect the transitional nature of the program we are adopting. Specifically, as discussed later herein, we provide for a transition away from reliance on non-unit specific contracts in order to mitigate impacts on current practices and arrangements.
7.2. Dispatchable Demand Response (DR)
Programs
D.04-10-035 found that in most circumstances dispatchable DR programs should be classified as resources that are eligible to count toward RAR. The Commission stated that it is "strongly supportive of demand response" and "willing to create special rules that permit it to qualify provided that [it does] not endanger system reliability in doing so." (D.04-10-035, pp. 26-27.) The Commission determined that DR resources should be available at least 48 hours each summer season to count as qualifying capacity, and that DR resources that operate two hours per day should be eligible but subject to a limit of 0.89% of monthly peaks.
The Phase 2 workshops addressed issues regarding how the DR programs will be dispatched. Some programs are only used under declared CAISO emergencies, which, arguably, is inconsistent with counting DR resources for RA so that they can be used to avoid emergencies and load interruptions. Also, because certain DR resources are only required to be available under emergency provisions, they do not currently have to bid or be scheduled in the day-ahead timeframe, or available in real time to the extent physically possible. This is arguably inconsistent with the concept of must-offer requirements for all RA resources.
In accordance with our prior determination that special RAR rules may be appropriate for DR programs as long as system reliability is not endangered, we will not adopt the CAISO's recommendation that emergency-only DR resources should not count for RAR. Nor will we require DR resources to bid or be scheduled a day ahead. We are concerned that these actions could effectively negate the value that these programs provide to the ratepayers who fund them. If the programs do not qualify as RA capacity, ratepayers would have to provide additional funding for the equivalent capacity value of the programs. The recommendations also appear to be inconsistent with the EAP, which gives policy preference to DR. We note that even though the CAISO states that emergency-only DR resources conflict with the objectives of RAR, it does not claim that allowing these resources to count would endanger reliability. For the same reasons, we find that it is appropriate to plan to use dispatchable DR programs up to the limits now established for each such program.
We recognize that some DR programs have been developed without the needs of the RAR program in mind, and, in particular, that the CAISO's ability to call on them may be sub-optimal. We anticipate that as the RAR program goes forward and that as DR programs continue to evolve, the programs can be better coordinated over time. However, at this time, we believe it is appropriate to recognize that ratepayers have been funding these programs and will continue to do so, and that the programs do provide capacity value even if they also create operational issues for the CAISO.
The workshop report asked parties to comment on how DR programs that are included in an LSE's RAR portfolio will be triggered. PG&E does not believe that the CAISO needs explicit dispatch authority for the DR programs as long as they can be executed in a timely fashion by the LSE. SCE believes that whether the CAISO or the LSE dispatches a program, the LSE should retain physical control over the program's use. Both PG&E and SCE refer to the need for protocols governing program dispatch established by the CAISO and LSEs.
We are generally supportive of DR program triggering protocols that would create a hierarchy in which price responsive DR programs are dispatched before emergency programs. We urge the CAISO and the LSEs to pursue this approach in conjunction with agency staff. As suggested in the workshop report, their discussions should include an evaluation of whether and when to apply DR triggers throughout the CAISO control area rather than just IOU service territories. We emphasize, however, that nothing in today's decision is intended to amend or revise current tariffs or other rules for existing or planned DR programs, whether with respect to triggering protocols or otherwise.
As noted earlier, D.04-10-035 established two restrictions on the use of DR resources in RAR portfolios-the 48-hour minimum availability requirement and the 0.89% cap on two-hour resources. The workshop report asked for comment on possible additional limitations on the countability of DR resources. Specifically, the report asked whether a program with a maximum call capability of four days per summer month should count.
As PG&E notes in its comments, the issue of limits on DR resources was discussed in the Phase 1 workshops and resolved in D.04-10-035. SCE believes a four-day call capability is generally insufficient, but it also believes that a single threshold cannot be applied to all types of DR. CAISO is concerned that a program providing only four days per month is potentially insufficient for RA, but it notes that the concern is mitigated by limiting the magnitude of DR capacity. CAISO also believes the topic warrants further discussion. We concur, and further determine that such discussion should take place in future RAR proceedings before additional restrictions on DR are adopted. We do not find it necessary or appropriate to modify the decision at this time.
7.3. Deliverability Issues
D.04-10-035 adopted the principle that to qualify for fulfillment of RA obligations, resources should be subject to both within-control area and out-of-control area deliverability screens. It also adopted the CAISO's proposal for a baseline analysis to determine deliverability of qualifying resources. The CAISO published a preliminary deliverability baseline analysis report and conducted a stakeholder meeting in May 2005, after the Phase 2 workshops were concluded. According to the Phase 2 workshop report, CAISO's May 2005 analysis found that (1) historical imports were deliverable and (2) while certain generation within generation pockets is not deliverable, that deficiency can likely be mitigated with transmission upgrades. The workshop report noted, however, that the deliverability test is only applicable to physical resources in the LSE portfolios, i.e., contracts that specify where the contract is being sourced.
Methodology - While the CAISO deliverability analysis methodology has general support, some parties questioned how often import data would be updated and whether import levels would be adjusted to reflect abnormal operating circumstances that could have affected the analysis. CAISO responded to these questions by proposing annual assessments timed to coincide with annual transmission grid planning. In these assessments, CAISO would adjust historic data to reflect both unusual operating circumstances and incremental capability from transmission upgrades. In response to specific direction in the workshop report, CAISO also included with its opening comments a description of how it would account for anomalous conditions that might be reflected in the deliverability assessment.
The comments on this topic describe several unresolved methodological issues that parties seek to have addressed. Fortunately, CAISO's determinations that historical imports are deliverable and that non-deliverability issues for generation within pockets can be mitigated by transmission upgrades allows us to proceed with the first cycle of RAR showings, before the methodological issues are resolved. We urge the CAISO to consider, through its stakeholder process, the concerns raised in comments on the methodological issues, including those raised by Calpine, Constellation, FPLE, PG&E, and Sempra Global.
Allocation of Import Capacity - The ability of LSEs to count import resources towards their RA obligations depends on the extent to which they can count upon access to inter-tie capacity. The workshop report describes three alternative proposals for allocating to LSEs the CAISO-determined level of import capacity:
1. Allocate inter-tie capacity in proportion to each LSE's contribution to the CAISO's transmission access charge (TAC). Parties favoring this approach support bilateral trading of unused inter-tie capacity among LSEs.
2. Allocate inter-tie capacity based on the TAC as in Option 1, but grandfather existing resource commitments. Bilateral trading or selling of an LSE's capacity share would not be required.
3. Allocate import capacity according to each LSE's share of CAISO system peak load. LSEs would assign their total intended RAR use to specific import paths and provide that information to the CAISO. The CAISO would then determine if the LSE's shares are feasible. If the CAISO determines that the allocation on a particular path is not feasible to meet a local requirement, then it would allocate first based on `evergreen' priority, and then based on the load share percentage. LSEs could trade and sell their load share provision on a path in advance of the determination for feasibility, but reselling or re-trading would not be allowed.
Selection of the most appropriate allocation option turns in large part on questions of equitable treatment of LSEs (and their customers) that have extensive import commitments on the one hand and those that do not on the other hand. The comments underscore the important question of whether LSEs with extensive import commitments should benefit from increased allocations that would result from grandfathering their commitments. As the workshop report explains, it is arguably inequitable to give preference to existing commitments because the costs of the transmission grid are socialized through TACs. The argument holds that LSEs that use the transmission system less because they rely on resources closer to load centers not only pay a socialized rate, they would also be disadvantaged by preferential treatment for existing commitments.16 The counter-argument is that long term commitments made on behalf of IOU customers prior to industry restructuring should be recognized.
It is our judgment that, on balance, the interests of LSEs and their customers who have made and invested in long-term commitments for imports outweigh the interests of those who pay socialized TACs. Moreover, we are concerned that failure to recognize long-term commitments here could discourage long-term contracting in the future. Accordingly, we will approve the grandfathering (i.e., evergreening) approach.
It is also our judgment that the third option is the most appropriate approach for allocating import capability among LSEs. Its use of load share rather than TAC charges as an allocator appears more in line with the capacity-based nature of the RAR program, and the TAC may be less valid as an allocator in that it covers all costs, not just those transmission lines used to import into the CAISO control area. We note that even though Option 3 did not receive specific attention as a package during the workshop discussions, it appears to have benefited from those discussions by addressing the underlying issues. In this respect, Option 3 actually appears to be a more complete package, and one that is more ready for implementation by the CAISO than are the other options. We note that it avoids the problem of LSEs with unneeded allocations withholding unused capacity as well as market power issues that could be associated with a secondary market for import capacity rights.
DWR Contracts - D.04-10-035 determined that DWR contracts will be subject to the adopted deliverability screens. As noted in the workshop report, grandfathering existing commitments (i.e., adoption of an "evergreen" provision for existing resource commitments) raises the issue of how to account for the deliverable portion DWR contracts. SCE proposed that DWR contracts be considered firm resource commitments eligible for evergreen treatment. SCE maintains that this will ensure that ratepayers will not have to pay for capacity to replace portions of the DWR contracts, which cannot be counted due to insufficient import capability. For contracts with sellers' choice provisions, SCE proposes that the contract's historical delivery be used to assess the path on which the contract will most likely be delivered.
We find that basing the allocation of the import capability of the DWR contracts on historic usage of the paths to deliver such supplies is consistent with grandfathering non-DWR contracts as well as our prior determination that DWR contracts should be subject to deliverability screens. SCE's evergreening proposal is adopted.
Deliverability in Generation Pockets - The CAISO's May 2005 deliverability analysis found approximately 2,300 MW to be undeliverable to the aggregate of load in the control area. Of this amount, 933 MW is located in PG&E's service territory, 1,270 in SCE's and 160 in SDG&E's. Staff reports that overall, relatively minor transmission remedies or operating solutions would resolve the deliverability limitations found in the study.
In its stakeholder meeting on the deliverability study results, the CAISO recommended that the existing units and imports be considered deliverable so long as the Participating Transmission Owners (PTOs) agree to complete the transmission upgrades by a date certain. Given that the reliability criteria violations that resulted in the undeliverable resources are "low level" and require relatively low costs fixes, the CAISO thinks the resources should be counted for RA purposes.
The Phase 2 workshop report identified and invited comments on two options for proceeding in light of the CAISO's deliverability findings:
1. Count all the generation as deliverable assuming that the transmission upgrades will be completed by the Participating Transmission Owners (PTOs). This option requires a commitment by PTOs to complete the transmission upgrades within a reasonable amount of time.
2. Disallow undeliverable capacity for counting toward the RAR until the transmission upgrades are completed. This option will require allocating the de-rates among the generators within the load pocket. A pro rata allocation of the de-rates to suppliers within a generation pocket seems the most equitable and simplistic approach.
In view of the CAISO's confidence that generation pocket non-deliverability can be mitigated by the PTOs by June 1, 2006, we will adopt the first option for the first compliance cycle of RAR. In the event that anticipated transmission upgrades are not completed and it is necessary to allocate de-rates among generators, we support a "first-come, first-served" approach as recommended by SCE and TURN rather than a pro rata allocation. Under this approach, if a constraint exists, capacity would first be allocated to generators that paid for firm transmission upgrades to make them deliverable or who did not need to add transmission capacity to be deliverable.
7.4. Liquidated Damages Contracts
In the context of this proceeding, liquidated damages (LD) contracts are bilateral agreements that provide energy, capacity, or ancillary service products without reference to a specific unit or resource backing the obligation. The enforcement mechanism for breach of these contracts is their liquidated damages provisions.17
D.04-10-035 noted that LD contracts are widely used in California and it acknowledged that they provide economic value. They are considered firm, and as AReM has pointed out, they have a track record of dependability as evidenced by the fact that four LSEs have never experienced circumstances where a liquidated damages clause in a CAISO-market LD contract was triggered, and a fifth LSE only experienced such a trigger for one hour. SCE also notes that these contracts performed according to their terms during the 2000-2001 energy crisis.
However, despite their proven performance, their failure to identify a specific resource that backs a capacity obligation could still undermine the integrity of the RAR program. Two concerns are especially problematic: LD contracts are not subject to deliverability screens, and they allow the possibility of double-counting resources that are nominated by LSEs in fulfillment of their RA obligations. Either of these shortcomings could affect the CAISO's ability to operate the grid. Additionally, while a supplier's failure to perform may result in financial compensation to the buyer through the LD clause, this incentive for the supplier to perform does not necessarily translate into availability of capacity to the CAISO when and where it is needed. Finally, the proven performance record of LD contracts does not mean they are an effective means of inducing forward capacity commitments.
The Commission's expressions of such concerns led to further consideration of LD contracts in Phase 2. The workshop discussions and subsequent comments underscore the concerns previously expressed by the Commission and they add to our concerns about the suitability of LD contracts for the RAR program. In particular, it is now apparent that LD contracts cannot meet the needs of local RAR due to their inherent deliverability and dispatchability constraints.
Some parties suggested mechanisms intended to make LD contracts more workable for RAR, such as year-ahead assessments by CAISO, day-ahead scheduling, and annual audits. We are not persuaded that these techniques would be effective, and they could raise new issues. For example the proposed year-ahead assessment would require that the qualified capacity of LD contracts be reduced pro rata in the event that the CAISO's assessment found a load/resource imbalance. We are concerned that such pro rata reductions could lead to new "free-rider" issues unless CAISO were able to identify the individual LSE and/or the specific LD contract responsible for the imbalance. Also, as Constellation points out, an assessment of what is available, while useful, does not ensure the commitment of resources.
We find that LD contracts are fundamentally incompatible with achieving the objectives of a physical capacity-based RAR program and that, ultimately, their eligibility for fulfillment of LSEs' capacity obligations should be disallowed. At the same time, however, we recognize that California's IOUs and ESPs have relied and continue to this day to rely extensively on the use of these contracts to serve their retail customers. Despite the shortcomings of LD contracts with respect to RAR, they have been valuable in other respects and no doubt will remain so. Terminating their eligibility to count for RAR showings too rapidly would be unnecessarily disruptive and costly to LSEs. This could be particularly problematic for ESPs that enter into contracts for energy services with their end-use customers. For them, any added costs that result from a capacity requirement could not be easily passed on to existing customers, at least in the short term. Accordingly it is our policy that the eligibility of in-area LD contracts to qualify for the LSEs' RAR showings should be phased out in a manner that fairly and effectively balances the needs of the RAR program and the interests of LSEs that rely on LD contracts.
We turn to alternatives for how to accomplish this phase-out. The options identified in the workshop report include (1) a grace period during which LSEs can continue to enter into new LD contracts and have them count in future RAR showings; (2) a sunset of and limitations on the countability of LD contracts (both existing and newly signed, if any); (3) term limits for LD contracts signed after the Phase 2 decision but before LD contracts are no longer permitted to count for resource adequacy; (4) limits on the extent to which LD contracts may count for resource adequacy, and (5) waivers.
Grandfathering Existing LD Contracts - The workshop report describes Calpine's position that only those LD contracts that were signed on or before October 28, 2004, the effective date of D.04-10-035, should count for resource adequacy. It also notes PG&E's proposal that current-form LD contracts signed after the effective date of this Phase 2 decision should not count for RAR. In effect, PG&E proposes that LD contracts signed before the date of today's decision should be grandfathered.
Grandfathering existing LD contracts is consistent with our decision to phase out rather than totally disallow the use of LD contracts for RAR at the commencement of the program. Even though D.04-10-035 referred to problems with LD contracts in the RAR program, and it provided for further evaluation of them in Phase 2 workshops, it did not definitively state an intention of this Commission to terminate their usage for RAR. We conclude that D.04-10-35 did not constitute fair notice to LSEs that, as of October 29, 2004, they should only enter into new LD contracts with the understanding that they were at risk that those contracts would not qualify for RAR. Nor did any other event prior to today constitute such notice.
We conclude that LD contracts executed on or before October 27, 2005, the date of this decision should be grandfathered, i.e., they should count for RAR showings subject to the sunset provisions and portfolio limitations described below. We again emphasize that this action affects only the extent to which LD contracts qualify for the RA obligation. We are not precluding their use for other purposes.
Grace Period for New LD Contracts - Proponents of a grace period hold that LSEs should be permitted to enter into new LD contracts after the effective date of this decision and have those contracts count for future RAR showings. During the workshops, AReM, TURN, and ORA proposed grace periods ranging between 90 and 180 days after the Phase 2 decision is issued. This would arguably permit a transition in the market to establish capacity products and allow LSEs to continue with business as usual. To avoid a rush on newly signed LD contracts, TURN further proposed that any firm LD contract signed within the grace period should only be grandfathered for one year.
We determine that the need for making progress towards full implementation of the physical capacity-based RAR program outweighs the interests of LSEs in an extended period during which they can continue historic procurement practices that do not provide identified physical capacity. Accordingly, we do not adopt proposals for a grace period beyond today.
Sunset Date - The workshop report describes general agreement that a sunset date should be established after which LD contracts would no longer count towards RAR, although some parties oppose any such sunset date. Most of the parties advocating a sunset date suggest that LD contracts should continue to count for capacity only through 2008. They do so based upon indications that the State's need for physical capacity will become more urgent in the 2008-2009 time frame. While we do not necessarily concur in the view that capacity additions are not critically needed until 2008 or 2009, we accept the common judgment that a transition of approximately three years is warranted. Therefore, we determine that LD contracts will not count for purposes of RAR showings after December 31, 2008. D.04-10-035 provided that DWR contracts should fully count for purposes of RA showings, subject only to deliverability screens. (D.04-10-035, Conclusion of Law 21.) Accordingly, to give effect to that decision, the sunset date adopted herein and the portfolio limitation discussed below shall not apply to DWR contracts.
Portfolio Share Limitation - SCE proposed that the use of LD contracts be capped at 25% of an LSE's overall RAR portfolio. PG&E proposed that in addition to being subject to a sunset date, LD contracts should also be subject to a scheduled phasing out in the interim. Thus, for 2006, an LSE could rely on LD contracts for 25% of its portfolio, but this limit would be reduced to 15% for 2007, 5% for 2008, and 0% for 2009 and beyond. ORA notes that a 25% of portfolio cap on LD contracts may be inappropriate for small LSEs.
We will adopt a variation of the declining share of RAR portfolio approach to balance the LSEs' need for time to adjust their portfolios away from LD contracts, and the need to make progress toward our goal of a physical capacity-based RAR program. Each LSE will be allowed to include LD contracts in partial fulfillment of its RA obligation, subject to the declining limits set forth in the following table:
Compliance Year |
Maximum LD Contract Limit (Expressed as a percentage of the LSE's RA Portfolio |
2006 |
75% |
2007 |
50% |
2008 |
25% |
This will apply to all LD contracts, regardless of the date signed, to eliminate incentives for parties to rush to sign large quantities of additional LD contracts.
As noted later in this decision, we are adopting a local capacity requirement and ordering its implementation after additional details of that RAR program element are considered. In light of the unsuitability of LD contracts for meeting local RA obligations, it is possible that the implementation of such local obligations will result in some LSEs having to acquire additional physical capacity in lieu of LD contracts to meet those obligations. We place all LSEs on notice that while we grandfather existing LD contracts and allow their continued use subject to the sunset date and phase-out schedule adopted herein, they may be subject to further limitations on the use of LD contracts to fulfill local RA obligations.
Finally, we note that by phasing out the ability of LD contracts to count in LSEs' RAR showings, we are not abrogating those contracts as has been claimed. The contracts will remain in effect until they expire on their own terms.
Waivers - AReM proposed that a waiver process be approved to protect LSEs against the possibility they will be unable to meet their RA obligations. AReM raises two primary concerns: (1) the market may fail to develop RA products and (2) a generator or generators may have opportunities to exert market power, particularly if a local RA obligation is adopted.
AReM's concern that the market may fail to develop needed RAR products is highly speculative in our view. With respect to market power concerns, as AReM and the Phase 2 workshop report point out, the Commission stated in D.04-10-035 that it would not require LSEs to sign contracts that meet RAR requirements at any cost. (D.04-10-035, p. 15.)
We stand by our earlier commitment to ensure that LSEs are not placed in a position whereby they would have to pay any price to acquire the capacity needed for their RA obligations. However, we are not persuaded that a specific waiver mechanism needs to be adopted to give effect to our commitment. Moreover, as TURN points out, the ability to submit comments on the workshop report does not provide adequate opportunity to develop a robust and well-conceived waiver process. We therefore decline to adopt such a mechanism at this time. As we give further consideration to the implementation details for the local capacity element of the RAR program, we will revisit the need for a waiver protocol.
7.5. Imports
D.04-10-035 approved several uncontested counting conventions that were addressed in the Phase 1 workshops and described in the June 2004 Phase 1 workshop report. These include counting conventions for import resources. Pursuant to D.04-10-035, the qualifying capacity for import contracts is the contract amount if the contract (1) is an Import Energy Product with operating reserves, (2) cannot be curtailed for economic reasons, and either (a) is delivered on transmission that cannot be curtailed in operating hours for economic reasons or bumped by higher priority transmission or (b) specifies firm delivery point (i.e., is not seller's choice).
In Phase 1, the CAISO had raised concerns about the possibility of sellers curtailing deliveries to meet native load requirements and about the definition of economic curtailments. However, after researching the applicable terms of the Western Systems Power Pool (WSPP) Agreement, the CAISO determined (and reported in its Phase 1 workshop comments) that the WSPP terms represented "acceptable and appropriate risk." D.04-10-035 provided that concerns about the use of firm transmission rights would be taken up in Phase 2.
In Phase 2, Powerex, the marketing subsidiary of British Columbia Hydro and Power Authority, offered a white paper consisting of a discussion of and proposals for the treatment of imports in the RAR program. (See Phase 2 Workshop Report, Appendix E.) The Phase 2 workshop report invited comment on the Powerex white paper. It also invited parties to address the exemption of imports from the determinations made with regard to resource availability, and related matters.
Powerex believes that even if, or when, intra-CAISO control area LD contracts are not allowed to qualify for RAR, firm LD import contracts should still be allowed because they are, Powerex contends, as reliable as unit-specific contracts if not more so. The only limit on import LD contracts would be the CAISO's deliverability test for intertie capability. Powerex further believes that CAISO Firm Transmission Rights (FTRs) should not be required of import resources for RAR purposes. Powerex notes that import contracts are not limited to energy products, and that an "import-backed Day Ahead capacity call option" product is useful and available to LSEs. In connection with this capacity product, Powerex proposes that imports not be subject the to the must-offer requirements established by D.04-10-035 because this would prevent it from re-marketing power that the LSE chooses not to use.
Firm import LD contracts do not raise issues of double counting and deliverability that led us to conclude that other LD contracts should be phased out for purposes of RAR. We note that firm import contracts are backed by spinning reserves. Accordingly, we approve the exemption of firm import LD contracts from the sunset/phase-out provisions applicable to other LD contracts as adopted in Section 7.4. We also approve the request of Powerex that import contracts not be required to have FTRs, as import transmission capability will be allocated to LSEs. We will not at this time approve the proposed exemption of call option contracts from the must-offer protocols adopted in D.04-10-035 and further elaborated here. Absent more definitive information that would enable us to weigh the trade-off between the business opportunities for the suppliers and the reliability benefit of the must-offer protocols that we have adopted, we are compelled to decide in favor of reliability. Powerex may present such information and renew its request in future RAR proceedings.
FPLE has proposed elimination of the requirement adopted in D.04-10-035 that imports have operating reserves. FPLE notes that the Western Electricity Coordinating Council (WECC) has been studying the question of reserves, and in recent months WECC committees have released draft documents that seek to clarify the origins and obligation of the reserves requirement. We do not find that these recent developments constitute a persuasive case for modifying our earlier decision, and we therefore deny this request.
7.6. Allocation of Capacity to Non-IOU LSEs
The Phase 1 Workshop Report described workshop discussions leading to the issue of whether any portion of the capacity value of the DWR contracts, QF contracts, and other utility retained generation should be allocated to non-utility LSEs. This issue was not resolved in the Phase 1 decision, but was instead addressed in the Phase 2 workshops.
Most direct access (DA) customers (i.e., those who are not "continuous" DA customers) pay a Cost Responsibility Surcharge (CRS). AReM maintains that by paying the CRS, these DA customers pay for DWR contract capacity that is assigned to the IOUs. AReM also contends that all DA customers pay a share of the costs of capacity associated with utility-retained generation (URG), including QF contracts, through the Competition Transition Charge (CTC). On the basis of cost causation principles and basic fairness, AReM takes the position that customers who pay for capacity should receive a capacity credit toward meeting their RA obligation. AReM proposes that the capacity credits would only be allocated to non-utility LSEs for RAR purposes. IOUs would retain full use of the contracts to meet their loads.
As described in the workshop report, opponents of allocating any portion of DWR contract capacity to non-utility LSEs maintain that DA customers only pay a portion of the above-market component of DWR and QF contracts. In contrast, the opponents argue that bundled service customers pay the full amount of the market value of such resources and, due to the 2.7 cent cap on the DA CRS, they currently pay a greater than proportionate share of the above-market component of such costs. Moreover, they contend, due to the deferred recovery of the balance of above-market costs from DA customers, it is impossible to determine how much, if any, of the above-market DWR costs will ultimately be paid by DA customers.
AReM relies on principles of equity and cost causation to support its case for capacity credits. However, as noted above, the cost responsibility of DA customers is capped. Moreover, that cap does not appear to be governed by the cost causation principles that AReM espouses. We find that it is not reasonable to craft remedies for possible cost shifting in this proceeding, where only a portion of the cost shifting issue is reviewed. Such remedies should be evaluated in proceedings where the totality of DA customer cost responsibility can be considered, including any cost shifting that may benefit DA customers. AReM's proposal to allocate RA capacity credits is denied.
7.7. Wind and Solar Resources
D.04-10-035 addressed issues of determining the qualifying capacity of wind and solar resources without backup by selecting an historic performance approach rather than using Effective Load Carrying Capacity adjustments. Under the adopted approach, the Commission requires that monthly performance differences be revealed, and that historic performance be computed during the peak period as defined in QF Standard Offer 1 (SO1) contracts. The Commission directed that methods for carrying out these determinations be taken up in Phase 2. Additionally, D.04-10-035 determined that proposals to segregate historic performance by different wind resource area would have to be supported by persuasive data, and that such proposals would be taken up as a second generation issue rather than in Phase 2.
The Phase 2 workshop report invited comments on how long the historic period for assessing generator capacity should be, what specific hours should be used for evaluation of the peak period, whether different types of generators should be measured separately, and the process for updating renewable-specific capacity assessments.
Averaging Period - Unlike hydroelectric generation, where rainfall and generation statistics are available over many years, there is not a large body of historical evidence regarding the performance of solar and wind generation. In addition, performance of solar and wind resources has been improving, and relying on old information about their performance could understate future estimated capacity factors. On the other hand, a longer history would smooth out the variability among individual months due to weather and other variables.
Workshop participants reached consensus that the best compromise would be a three-year rolling average of performance history. For example, for June 2006 the generation results for June 2003, 2004, and 2005 would be averaged. If 2005 data is not available, the most recent available data from the previous three years could be used. Workshop participants considered this a sufficiently short time to avoid downward bias because of technology changes, yet enough time to smooth out the variable results of any one particular year.
Comments on the workshop report echoed the consensus of the workshop discussions. Parties either support or accept the use of a three-year rolling average. We adopt the use of month-specific three-year rolling averages for determining the qualifying capacity of wind and solar generation because this approach strikes a reasonable balance between the need to recognize technological improvements and the need to smooth out recorded performance variations due to weather and other variables.
Peak Period Definition - As noted above, D.04-10-035 adopted the use of the standardized peak hour definition in SO1 contracts for purposes of calculating the qualifying capacity of wind and solar resources. The SO1 contract summer peak hours are noon to 6:00 p.m., and while this is reasonably consistent with the CAISO summer load profile, the Phase 2 workshops addressed problems that this raises for non-summer months. The SO1 contracts define mid-peak or partial-peak hours for the non-summer months but not peak hours. The CAISO non-summer peak generally falls in the range of 5:00 p.m. to 8:00 p.m. in the non-summer hours.
In its workshop report comments, CAISO proposes using the SO1 summer peak hours of noon to 6:00 p.m. on a year-round basis. We find this is a reasonable compromise of this surprisingly complex issue. CAISO notes that wind and solar production can vary dramatically across the afternoon hours, and that the wider six-hour window of SO1 hours could give a "somewhat added boost to these resources." As the workshop report notes, reasonableness and ease of administration argue in favor of a simpler method for defining peak periods. We therefore adopt the simplified approach of using SO1 peak hours year-round as recommended by the CAISO.
Differentiating Generator Types - The Phase 2 workshops addressed the idea of breakouts by technology and/or by vintage. This discussion reportedly yielded only partial consensus. The workshop report observes that benefits of differentiating resources by technology or by vintage would be small from a resource adequacy perspective. As also noted in the workshop report, our adoption of the three-year rolling average has the additional benefit of updating the sample by one-third each year. We think that this is an appropriate means of recognizing the addition of newer technologies for RAR, and that further consideration is not warranted at this time. As PG&E notes, qualifying capacity need not distinguish between technology types or vintage.
Renewables - The workshop report observed that the adopted methodology for assessing wind and solar generation capacity and expected output should not unduly disadvantage renewable generation, and it invited comment on this topic. None of the comments expressed any disagreement with the principle that the use of renewables should not be disadvantaged in or by the RAR program. SCE recommends using slightly higher expectations such as a 3% adder for newer wind technologies to compensate for data lags associated with the introduction of those new technologies. CAISO on the other hand recommends against methods that over-estimate peak-hour production from renewables.
An adder such as that suggested by SCE appears to have merit as an appropriate means to prevent possible disadvantaging of renewable resources that are offered to meet the RA obligation. We will adopt an adder of 3% for newer wind technologies for 2006 only, and provide for further consideration of the need for such an adder in future RAR proceedings. Additionally, as the operation of the RAR program unfolds, if we become aware of unintended consequences that unduly impact renewable resources we will be prepared to consider and make any necessary RAR program adjustments.
7.8. Energy-Limited Resources
D.04-10-035 determined that to qualify for RAR, a resource must (1) be able to operate for a minimum of four hours per day for three consecutive days and (2) be able to run a minimum aggregate number of hours per month based on the number of hours that loads in the CAISO control area exceed 90% of peak demand in that month. The second prong of this test (i.e., the 90% rule) is applicable to the summer months only. D.04-10-035 referred to Phase 2 the development of an appropriate rule for energy-limited resources for non-summer months.
Using data from 1998 through 2003, the CAISO calculated the number of hours in each summer month that load was greater than 90% of the monthly peak. The range is from 30 hours (for May) to 60 hours (for August). However, load shapes are less peaked in the non-summer months, with the result that the number of hours with loads in excess of 90% of the peak could be much higher, as much as 300 hours in some months.
The workshop discussions confirmed the view that the 90% rule is unworkable for the non-summer months. As the workshop report points out, if the qualifying capacity of energy-limited resources has to meet expected run times of up to 300 hours, the capacity will be severely degraded. PG&E notes this could affect the availability of hydroelectric units to the CAISO. Moreover, operation in the non-summer months for long periods would not be the best use of such resources. Rather than development of a substitute rule that would accomplish for the non-summer months the equivalent reliability value that the 90% rule accomplishes for the summer months, the workshop discussions yielded general agreement that there should be no second prong of the test for energy-limited for the non-summer months.
We are unwilling to adopt a rule that could cause LSEs to contract for large amounts of capacity that will not be called upon, because there is little assurance that such a rule would create reliability benefits that outweigh the cost of that capacity.
Based on the foregoing, we affirm the applicability of the two-prong test for energy-limited resources for the summer months as adopted in D.04-10-035. For the non-summer months, the first prong of the test, i.e., the four hour-by-three consecutive day rule, shall apply. The second prong, i.e., the 90 % rule, is waived for the non-summer months. We concur with the CAISO that aspects of this rule, including a limit on the total MW and a priority order, may have to be reviewed in future RAR proceedings.
7.9. Commercial On-Line Dates
The Phase 1 workshops and decision addressed the need for conventions for when and how to treat resources that are under construction as qualifying capacity. D.04-10-035 noted that project databases maintained by the CEC and the CAISO are the appropriate foundation for determining the commercial operation dates (CODs) for resources nominated for RAR, and referred the topic to Phase 2. The Phase 2 workshop discussions addressed criteria that the CAISO and CEC cooperatively used to develop a working proposal for counting rsources under construction and estimating CODs. The CAISO-CEC working proposal is attached to the Phase 2 workshop report as Appendix F.
The essence of the CAISO-CEC proposal is that the CAISO and the CEC would jointly create and post monthly on a public website a report for the use of LSEs. The report would list the expected date of commercial operation as reported by the developer of each resource under construction or with an expected date of commercial operation of one year or less, that has a nameplate capacity rating of one MW or greater. For the annual year-ahead showing of resource adequacy, an LSE would be able to include, for any given month, a resource that is still under construction provided that the latest revised date of commercial operation posted on the public web site is no later than the first calendar day of the applicable month, and the operational status is expected to be achieved no less than 60 days prior to that date. A resource that meets those criteria would be considered to have achieved qualified status for the year-ahead showing. For example, a resource that the developer reports is expected to achieve commercial operation no later than July 1, 2008 could be used by an LSE in its September 2007 report to demonstrate compliance in its year-ahead showing for the month of July 2008. For a month-ahead showing of resource adequacy, qualification of a resource is dependent upon the unit having achieved operational status 60 days prior to the month in which it would be counted for resource adequacy purposes. For example, if a month-ahead showing for the month of August 2011 required a month-ahead filing by June 30, 2011, that filing could only include a resource that had achieved operational status no later than June 1, 2011.
All parties that commented on this proposal either support or accept it. SCE notes that the 60-day provision may have to be revisited when more experience is gained. It may be possible to reduce that delay to 30 days or even less. We appreciate the joint efforts of the CAISO and the CEC in developing this proposal as well as their commitment to maintain the listing. We hereby adopt it as reasonable.
7.10. Local RAR
Through its Local Area Reliability Service (LARS) process, the CAISO identifies generators that must be available in or for a particular area due to transmission constraints. To assure operational reliability, the CAISO enters into reliability must run (RMR) contracts with those generators. RMR costs are paid by all load through CAISO uplift charges.
Addressing the local reliability challenges posed by constrained transmission limits, D.04-07-028 stated that "a utility scheduling practice or procurement plan that focuses solely on least cost energy, without regard to deliverability of the procured energy to load or to local reliability, is not in compliance with our prior decisions, approved short-term procurement plans, and Assembly Bill 57." (D.04-07-028, pp. 9-10.) The Commission also stated that "it is our intention to minimize the use of RMR contracts, and that the utilities should include local reliability in their long-term procurement plans for the purpose of reducing the need for RMR contracts." (Id., p. 13.)
Concerns about local reliability and CAISO's reliance on RMR contracts led to consideration of localized RAR for all LSEs in Phase 1. D.04-10-035 determined that adding a local component to the RAR program would be consistent with the Commission's prior decisions in which it has been held that LSEs are responsible for procuring the resources needed to meet their customers' needs. Discussing the costs of local RAR (higher procurement costs, higher forecasting and planning costs for LSEs, program complexity, and possible market power) as well as the benefits (contracts with longer terms than RMR contracts would assure revenue streams to generators, LSEs would be better able to identify cheaper and environmentally friendly alternatives to RMR contracts, and possible incentives for transmission upgrades) the commission determined that the likely benefits of local RAR outweigh the likely costs.
D.04-10-035 directed parties to address the implementation details of local RAR in future proceedings. It also laid out the sequence of events for how this should be done. First, when completed, the deliverability baseline analysis that was being conducted by the CAISO would be an important data source for identifying conditions that define load pockets, the geographic scope of load pockets, and methods for updating them.18 Next, once the first step is completed, the extent to which customers reside in load pockets, methods for tracking customers, and other LSE-specific load forecasts would be addressed. Finally, once LSE-specific load forecasts in load pockets are known, the timing of LSE local procurement would be coordinated with the expiration of existing RMR contracts.
Development of localized RAR was taken up in the Phase 2 workshops. The CAISO presented a working proposal for establishing local capacity requirements, and in response to initial workshop discussions it revised the proposal. The CAISO's January 25, 2005 working proposal is attached to the Phase 2 workshop report as Appendix G, and an alternative proposal by Mirant is attached as Appendix H. CAISO issued its "RAR Local Capacity Straw Proposal" and "Local Capacity Technical Analysis-Overview of Study Report and Preliminary Results" on June 23, 2005, and it convened a stakeholder meeting on June 29, 2005. The Phase 2 workshop report states that given that issues such as cost allocation and pricing for CAISO supplemental procurement, and local market power mitigation are FERC-jurisdictional, the locational capacity procurement framework is "somewhat incomplete."
We reaffirm our intention to establish a local capacity component of our RAR program as we determined in D.04-10-035. As DENA correctly observes, "the local area requirement runs to the heart of the `where and when needed' aspect of the RAR policy." (DENA comments, p. 12.) We note that parties appear to be unanimous in their concurrence that a local dimension to the RAR program is required. The principal issue that divides the parties is whether the local component should be implemented for 2006, when the basic RAR program is to be implemented, or for 2007.
We concur with DENA that local reliability should be reflected in RAR and implemented as soon as possible. We cannot concur, however, with those who advocate that the local component of RAR should go into effect with the initial wave of RAR implementation. We will provide for implementation of local procurement requirements in 2007 for several reasons. Most significantly, several important aspects and possible consequences of the proposed local program have not been fully or fairly considered in the Phase 2 workshops, as underscored by the fact that the CAISO's preliminary baseline deliverability analysis was published after the Phase 2 workshops were completed and its "straw proposal" for local capacity was distributed several weeks after that.
Among other concerns, the record before us does not allow us to find that the reliability benefits of the CAISO's straw proposal justify the costs and operational burdens that will be imposed on LSEs and their customers. We are also concerned that for some areas, local generation capacity may exist but not be available to smaller LSEs. Further, there is currently no mechanism available, such as a CAISO backstop procurement contract or tariff, to mitigate the market power of local generation capacity. Absent a showing that the local procurement obligations that would be imposed on LSEs will be more cost-effective than current local procurement through the RMR mechanism, we are not prepared to order LSEs to pursue such obligations for 2006.
Additionally, we have not been presented with an adequately developed method by which local capacity requirements can be allocated to individual LSEs, and it remains unclear how LSEs with small portions of the overall capacity requirements in any one load pocket could acquire necessary capacity from eligible generation. Parties discussed allocation of local capacity requirements using locational attributes included within LSE customer billing systems without apparently undertaking the effort to conduct the assessments needed to implement the concepts discussed. Given the potential volatility in customer relationships among ESPs, CCAs, and the default IOU, it is apparent that these allocations would need to be updated frequently, perhaps annually.
While the CAISO has pursued development of its local capacity procurement proposal through its stakeholder process and is reportedly continuing to do so, implementation of the proposal without an opportunity for it to be vetted before this Commission is not consistent with the three step sequence of events that we outlined in D.04-10-035. More generally, such implementation would be inconsistent with the processes and the authority of this Commission. We have committed to working cooperatively with the CAISO towards the development of complementary RAR and MRTU programs that serve California's needs for reliable electricity supply at reasonable costs. In carrying out this commitment, we are mindful of the respective roles of each entity. In view of the fact that important details of the CAISO's proposal for local capacity requirements are being developed by the CAISO through its stakeholder process but have not been considered on the record of this or any other Commission proceeding, approving the CAISO's preliminary proposal at this time would, in effect, constitute an inappropriate delegation of our own authority to make determinations regarding the balancing of reliability and the costs of achieving that reliability.
As set forth in Section 9 of this decision, we are providing for further proceedings to complete the implementation of our RAR policy framework. Those proceedings will be the forum to complete the development and the evaluation of the various details of the local RAR component so that it can be implemented in 2007. They will provide parties with an opportunity to present us with information regarding the appropriate overarching policies for local RAR;19 costs and benefits of alternative approaches to reliability critereia used to define the local obligation; means of preventing or mitigating market power; mechanisms that will allow LSEs, especially smaller ones, to acquire capacity to meet their localized obligations; whether there is a need for waivers and if so what form they should take; cost allocation issues; whether the MOO mechanism should be retained until the CAISO has authority to enter into backstop local capacity contracts; and assurance that the need for transmission upgrades to address load pockets is considered and weighed against the need for local capacity. To ensure that we are presented with a comprehensive proposal for implementation of a local RAR that can be timely implemented for 2007, we hereby direct the IOUs and authorize other parties to file such proposals in this or the successor RAR proceeding within 75 days of the date of this order.
For 2006, the local procurement policies we adopted in D.04-07-028 remain in effect. We also note that, despite its deficiencies in terms of promoting infrastructure investment and allocating costs on the basis of cost causation, as PG&E notes the existing CAISO RMR mechanism remains available and effective for achieving local operational reliability.
15 However, as staff notes and CAISO reiterates in its comments, the TD approach may also be subject to this inefficiency to the extent that existing contractual arrangements are deemed eligible to satisfy the RA obligation.
16 The comments refer to SCE's reliance on imports from coal and nuclear resources that it owns in the Southwest, but we understand the issue is generic and could apply in other circumstances as well.
17 As DENA points out in its comments, the term "LD contract" may not be most descriptive term. The real issue at hand is not whether contracts between LSEs and resource suppliers have liquidated damages clauses, but whether they identify specific, committed assets or units (i.e., physical resources) that back up the contractual obligations. The term "unit-specific contract" may be more descriptive. We also recognize that liquidated damages clauses are not uncommon in commercial contracts. We nevertheless elect to use the term "LD contract" in this decision because the parties have used it widely in both Phase 1 and Phase 2.
18 CAISO issued its "Preliminary Deliverability Baseline Analysis Study Report" on May 3, 2005, after the Phase 2 workshops were completed.
19 We note that the policy principles suggested by AReM in its opening comments may represent an appropriate starting point for discussions of local RAR policy; however, we do not necessarily endorse the AReM positions stated therein.
