IV. Cost Allocation Issues

A. Allocation of Ongoing Transition Costs

Section 367 generally defines transition costs and establishes the time frame for recovery of uneconomic costs. Generation-related transition costs must be recovered by December 31, 2001, with certain important exceptions. These exceptions include the following:

· employee-related transition costs (which must be recovered no later than December 31, 2006);

· power purchase contract obligations (which continue for the duration of the contract);

· costs associated with any buy-out, buy-down, or renegotiation of such contracts (which also continue for the duration of the agreement);

· costs associated with contracts approved by the Commission to settle issues associated with the Biennial Resource Plan Update (BRPU) (which may be collected through March 31, 2002, provided that only 80% of the balance remaining after December 31, 2001 are eligible for recovery);

· costs associated with entities exempted from transition cost recovery as delineated in § 374 (which must be recovered by March 31, 2002, provided that only $50 million of any balance remaining after December 31, 2001 is eligible for recovery);

· and costs associated with repaying the rate reduction bonds may be recovered until the fixed transition amounts are recovered in full.

Much of the controversy in this proceeding has centered on how these ongoing transition costs are allocated and the statutory interpretation of § 367(e) et seq.

The utilities propose to allocate post rate freeze transition costs using a System Average Percentage Change (SAPC) methodology. FEA, Farm Bureau, State Consumers, and Large Users support the utilities' proposal. They argue that § 367(e)(1) mandates that transition costs, both during and after the transition period, be allocated as similar costs were allocated on June 10, 1996. Rates frozen at June 1996 levels were allocated using a full Equal Percent Marginal Cost (EPMC) methodology. The SAPC, a proxy for the EPMC, adjusts these rate components (derived using EPMC) for usage.

Large Users state that it is unlawful to change allocation methodologies because this approach would involve cost shifting in conflict with § 367(e)(1) and approve a different allocation than that in effect in June 1996. Parties in favor of continuing the SAPC methodology argue that § 367(e) is specific to "transition costs" and does not distinguish between uneconomic costs during or after the rate freeze. Ongoing transition costs, which are primarily QF and power purchase agreement costs, are designated transition costs and should be treated as such under § 367(e)(1). In addition they argue that pursuant to § 371(a), transition costs should be adjusted for usage, which is accomplished by the SAPC methodology.

State Consumers support the SAPC methodology for now, but acknowledge equity concerns with locking in allocation factors over the long term. These parties recognize that, over time, rates will continue to diverge from those in effect on June 10, 1996. This is especially so considering the many ratemaking proceedings before the Commission.

On the other hand, TURN and ORA argue that ongoing transition cost responsibility should be allocated using cents-per-kilowatt-hour, which is inherently a usage-based allocation methodology. They argue that transition cost obligations should be allocated as generation costs since they are generation-related uneconomic costs. TURN and ORA recognize that D.99-06-058 (the decision in the 1998 RAP) mandated a SAPC transition cost allocation methodology during the rate freeze to avoid unlawful cost-shifting. However, these parties argue that there is no prohibition against cost-shifting after the rate freeze ends.

TURN also proposed an alternative allocation methodology, based on customer class demand in the top 100 hours of the year, which has previously been used to allocate generation costs for PG&E. TURN states that such a demand-based allocator is necessary to reflect that shortage costs that lead to construction of generation are incurred in more than a single peak hour.

Farm Bureau states that if the Commission determines that the statute permits deviation from the SAPC methodology, it supports TURN's and ORA's cents-per-kilowatt-hour approach. However, it recommends that if that method is chosen, allocation should be applied on a system wide basis as opposed to each tariff within the rate groups on a pro-rata basis. ORA agrees that this is reasonable.

In D.99-06-058, the Commission determined that transition costs must be allocated using a SAPC methodology during the rate freeze since that is the method used to derive frozen rates. The primary question to be resolved in this proceeding is whether the allocation methodology can and should be changed after the rate freeze ends.

Section 367(e)(1) states that transition costs must:

Be allocated among the various classes of customers...in substantially the same proportion as similar costs are recovered as of June 10, 1996.

Section 367(e)(3) establishes that the Commission shall retain existing cost allocation authority provided the firewall and rate freeze principals are not violated.

The proponents of maintaining the SAPC cost allocation methodology post-rate-freeze contend that there is no conflict between the §§ 367(e)(1) and (3). According to these parties, we cannot change the cost allocation methodology for transition cost recovery until all transition costs are collected, including those that the utilities are allowed to recover after the rate freeze.

We must interpret §§ 367(e)(1) (2), and (3) in a manner that harmonizes the statute, and makes sense in light of the language and intent of the statute as a whole.

We agree with the various parties who assert that the statute as a whole requires transition costs to be allocated in substantially the same way as they are allocated now both until the rate freeze ends and after, until transition costs no longer exist. However, this does not mean that the Commission cannot change the allocation at all. The Section 367 (e)(1) use of the modifiers "substantially the same proportion as" must be given effect as well as all other words. Further, if the Commission's allocation methodology can be "substantially the same proportion as" the current (and 1996) allocation, the discretion inherent in that phrase must mean that some classes may pay more and some classes may pay less in either dollar or percentage terms, than they pay now.

The statute (§367(e)(2)) provides that no customer pays more due to transition cost allocation changes (except for direct access customers in certain circumstances). During the rate freeze, this clause leads to residual calculation of transition costs. However, its implication must be considered after the rate freeze. After the rate freeze, an allocation of transition costs which is "substantially the same proportion as" - but not exactly the same as - today's EPMC allocation will raise some customer's costs and lower some other customer's costs. In order to take each part of the statute together, we must add the fact that overall rates will decrease after the end of the rate freeze. Therefore, it is possible to increase some customers's responsibility for transition costs (on a percentage basis) and still have the result of lower rates.

The statute (§367(e)(3)) allows the commission to retain its pre-1996 cost allocation authority, subject to certain conditions. Therefore, the commission may consider other cost allocation methodologies besides EPMC or SAPC, again as long as the constraints of §367(e)(1) and (2) are abided by. While not required, it is appropriate for us to explore our options.

We will not adopt the EPMC or SAPC method for post-rate freeze transition costs solely on the basis that these methods are consistent with the statute. Other methods are also consistent with the statute, and we wish to consider various allocation methodologies that take into account equity and economics as well as adherence to the law.

Continuing to allocate costs in a manner based on bundled rates is outdated and inconsistent with current ratemaking policies. The EPMC and SAPC methodologies derive from ratemaking approaches implemented when rates were fully bundled. While rates are frozen at June 1996 levels, it makes sense to allocate transition costs using an EPMC or SAPC because rates were fully bundled in 1996. Now that rates are unbundled and costs are assigned to general functions, it would be inappropriate and contrary to cost causation to continue to allocate transition costs after the rate freeze as though rates were still bundled. For this reason we believe the EPMC and SAPC allocation methodologies are inappropriate for allocating ongoing transition costs.

Transition costs are an unusual set of costs because they are the uneconomic costs of generation resulting from the onset of competition. Large Users argue that ongoing transition costs should not be allocated using a cents-per-kilowatt-hour allocation because the costs are not volumetric in nature as they do not vary with energy use. TURN and ORA argue that transition costs are appropriately assigned to generation since they are the uneconomic costs associated with that function. We agree with TURN and ORA. Transition costs should be allocated based on energy consumption or demand because the costs are most appropriately assigned to the generation function. Again, this methodology is consistent with our policy of unbundling rates and functionalizing costs. Further, allocation of costs based on energy consumption is consistent with our long- standing principle of allocation by cost causation.

Costs vary based on time of use; put simplistically, the market cost (the PX price) for energy is generally lower in the middle of the night than in the middle of the day. Costs also vary based on season and other factors. To a certain extent, rates reflect these cost differences. Customers with time-of-use meters and real-time meters receive price signals that correspond (more or less) to the changes in costs and prices over time. Even residential customers without sophisticated meters currently see prices that vary between season, based on cost and market differences between season.

An allocation methodology for generation-related transition costs should take into consideration the variances in generation costs over time. The current EPMC method of allocation for transition costs does not do this, as it allocates costs based on systemwide costs instead of generation costs. Further, it allocates transition costs based on a method in existence in 1996 (and from costs from earlier years) that likely has little or no relation to costs today. At the same time, the equal cents per kwh allocation method fails to recognize that different customer classes purchase power at different times on average. Certainly, it cannot be true that each class has the same load curve, thus each class cannot be seen as causing the same generation costs. A theme throughout this decision is to expose customers to market forces e.g., by rejecting price cap proposals and requiring hourly data be used to bill customers on hourly meters. Cost allocation should likewise align with actual market prices billed.

TURN proposed a transition cost allocation method that addresses cost causation in such a way as to directly link actual usage patterns and provide an appropriate proxy for actual generation costs. We believe such a methodology must be considered and analyzed for these purposes because it is the only proposal in the record which addresses cost causation in a way related to demands placed on the system.

Exhibit 91 is a comparison exhibit in which the three allocation methodologies ²⁸ are analyzed to determine what percentage of transition costs should be allocated to each class. This Exhibit shows that the allocation would change by about 10% or less for each class (compared with EPMC) using the top 100 hours method proposed by TURN. In comparison, the SAPC method would change allocations less than 10% in most cases, and the equal cents methodology would change allocations more than 10% in many cases.

Based on this data, we can conclude that the TURN methodology for allocating costs based on the top 100 demand hours would meet our obligations under §367(e)(1)(2) and (3). This methodology is different from the method in place today, but to adopt it would be an exercise of our continuing cost allocation authority. The top 100 hours methodology would not result in price increases for customers as compared to 1996 rates, because overall rates will have decreased substantially at the end of each utility's rate freeze.²⁹ Finally, the changes of 10% or less in allocation for each class maintain allocations which are substantially in the same proportion as the allocations on June 10, 1996.

However, while we can adopt the policy to use the top 100 hours methodology, we cannot adopt actual allocations consistent with it at this time. As is noted in Exhibit 91, TURN's methodology is converted to allocation percentages based on historical data from 1991-1993 which is neither confirmed nor supported by other parties. Therefore, we need to update the data to obtain actual allocations.

Therefore, we will order SDG&E to file an Application within 60 days of this decision which updates the top 100 hours methodology. PG&E should use A.99-03-014, Phase II of its General Rate Case, for this matter, and SCE should include the issue in the proceeding considering A.00-01-009. Utilities should use the class average hourly load profile data used to calculate the PX credit (with a reasonable allocation to streetlighting and any other class without a calculable load profile). The data should be used to identify the top 100 demand hours for 1998 and 1999 and identify the percentage allocation attributable to each class. This percentage should be averaged over the two years, and ongoing transition costs should be allocated based on these averaged percentages by class. We note that the 1998 and 1999 data may or may not be representative of normal years. However, we believe updated data using load profiles will produce superior results to the data underlying the 1991-1993 calculation in Exhibit 91. We reserve the right to further update the data or otherwise change the ongoing transition cost allocations consistent with §367 in the future.

For SDG&E, which ended its rate freeze in July 1999, it is appropriate to implement the new transition cost allocation as soon as possible. While Exhibit 91 for SDG&E is not updated sufficiently for long-term purposes, it provides a short-term record-based proxy. We will order SDG&E to file an Advice Letter within 30 days to re-allocate transition costs per Exhibit 91, and this allocation should remain in place until SDG&E's Application on this issue is resolved. SDG&E may propose in its Advice Letter a balancing account to allow for any changes made when the Application is resolved.

Assuming the Application provides data consistent with this decision, we will adopt the results as a proxy for TURN's top 100 hours proposal, with the caveat that we will need to review the results once again to ensure compatibility with §367(e) if there are significant allocation differences from those indicated in Exhibit 91 and to ensure that no individual customer would experience a rate increase from changes in transition cost allocation (as compared to rates during the rate freeze). For example, we may need to consider caps or floors on the allocation to specific classes or rate schedules.

B. Allocation of Restructuring Implementation Costs

Restructuring implementation costs are costs resulting from the implementation of direct access, the PX, and the ISO. Treatment of these costs is addressed in Pub. Util. Code § 376³⁰. We discussed the eligibility of such costs for § 376 treatment in D.99-05-031, D.99-09-064, and D.99-12-032. In D.99-06-058, we determined that these costs should be allocated using a SAPC methodology during the rate freeze. Again, the question before us now is whether the allocation methodology can and should be changed after the rate freeze.

As with ongoing transition costs, SCE, SDG&E, FEA, and Large Users propose to continue to use a SAPC method for restructuring implementation cost allocation. These parties believe that maintaining the SAPC methodology is most consistent with the cost-shifting principles of § 367(e)(1).

They point out that implementation costs are not recoverable after the rate freeze, only the displaced transition costs. Therefore, the displaced transition costs should have the same allocation as all transition costs. These parties maintain their position that like transition costs, costs displaced by recovery of restructuring implementation costs should be allocated using a SAPC methodology.

TURN, UCAN, and ORA argue that these costs should be allocated using equal cents-per-kilowatt hour methodology. These groups argue that allocating restructuring implementation costs using SAPC would spread the costs disproportionately to those classes that have not benefited equally from electric restructuring. These parties state that in D.99-06-058, the Commission mandated a continuation of the SAPC through the transition period due to cost-shifting considerations, but implied consideration of an alternative treatment of 376 costs at the end of rate freeze. PG&E proposes to attribute these costs to a function such as distribution and to allocate the costs in a similar manner to others in the particular function.

The restructuring implementation costs themselves are not recoverable after the rate freeze. Pursuant to § 376, the only costs recoverable are the transition costs that were displaced because of recovery of restructuring implementation costs. These costs should not be singled out from other transition costs for separate treatment, but instead should be allocated according to the same methodology as other ongoing transition costs. Therefore, transition costs displaced because of recovery of restructuring implementation costs should be allocated using a top 100 hours methodology applied on a system-wide basis.

C. Nuclear Decommissioning and Public Purpose Costs

UCAN and TURN argue that nuclear decommissioning costs should be allocated based on equal cents per kilowatt-hour. These costs are currently allocated using an EPMC methodology. They argue that nuclear costs were incurred to meet base as opposed to peak demand and that costs should be borne by those consuming larger volumes of power.

Large Users and FEA argue that decommissioning costs are the costs of doing business for a utility with nuclear plants and are not a function of energy demand or how many kWh are produced. They argue that these costs have traditionally been recovered residually on a SAPC basis (D.97-08-056, p. 36). FEA points out that decommissioning is a result of the very existence of the plants and not the amount of power produced.

Edison maintains that TURN and UCAN provide no justification to change the allocation of nuclear decommissioning costs. PG&E also opposes this proposal. PG&E contends that neither nuclear decommissioning costs nor public purpose costs are rate components that are associated with ending the rate freeze; therefore, PG&E contends that proper discussion of these components should take place in A.99-03-014, its Phase 2 general rate case (GRC) proceeding.

TURN and UCAN propose to continue the current cents-per-kilowatt-hour cost allocation for costs related to the CARE program. They recommend that, until decisions are made on how energy efficiency and other non-CARE public purpose programs will be funded and administered after 2001, the current SAPC cost allocation methodology for non-CARE costs continue and be tracked in a one-way balancing account. When a final determination regarding funding for public purpose programs is made, allocation of the funds should be decided at that time.

ORA recommends that we direct the utilities to propose a consistent format and process for separately tracking, reporting, and reconciling all public purpose revenues in the 2000 Annual Earnings Assessment Proceeding (AEAP). Edison believes the ATCP is appropriate place to assess these programs. PG&E agrees that a balancing account must be created to track the difference between the revenues collected under the public purpose rate component and the authorized public purpose revenue requirement and proposes the Public Purpose Program Revenue Adjustment Mechanism (PPPRAM), a two-way balancing account. PG&E contends that a discussion of public purpose cost allocation should take place in its Phase 2 GRC proceeding.

We previously considered nuclear decommissioning cost allocation in D.97-08-056, in which the Commission was constrained by the cost shifting provisions of AB 1890. That decision states:

We direct the utilities to allocate these program costs using PG&E's system average percent method, which is closest to the current cost allocation methods and therefore accommodates AB 1890's rate freeze and prohibition against cost-shifting (Id., mimeo. at p. 36).³¹

In this decision, we have stated our intent to further our policy of unbundling rates and functionalizing costs. Consistent with our approach to transition cost allocation, nuclear decommissioning costs should be assigned to function. We agree with TURN's and UCAN's argument that nuclear decommissioning costs are most appropriately assigned to the generation function. Therefore, we will adopt a cost allocation methodology based on energy consumption. Once the rate freeze ends, nuclear decommissioning costs shall be allocated using a cents-per-kilowatt-hour methodology. This approach is not only consistent with our treatment of transition costs, but is also the most equitable allocation methodology given the cost saving disparity between large and small customers.

CARE costs should continue to be allocated on a cents-per-kilowatt-hour basis. For non-CARE public purpose programs, we agree that it is reasonable to continue SAPC cost allocation after the rate freeze. We acknowledge that a new mechanism must be established to ensure that the UDCs collect the authorized revenue requirement for public purpose programs. We approved the Public Purpose Programs Adjustment Mechanism (PPPRAM) and Edison's PPPRAM in D.99-10-057 on an interim basis. We affirm that approval here and establish that these accounts should be two-way balancing accounts. Further issues regarding funding for the period after 2001 should be determined in the public purpose rulemaking, R.98-07-037. We will not make any recommendations here as to what issues should be considered in the 2000 AEAP.

D. Reliability Must Run Cost Allocation

Reliability Must Run (RMR) contracts ensure the ISO's ability to summon generators to provide reliability and system stability when the market fails to provide the necessary support. The RMR contracts are subject to FERC jurisdiction. During the rate freeze period RMR costs are being recovered through the Commission-established Transition Revenue Account (TRA). Recovery and allocation of RMR costs through the TRA during the rate freeze is being reviewed in the Revenue Adjustment Proceeding (A.99-08-022 et al.). Once the rate freeze terminates the utilities will no longer have the TRA cost recovery mechanism and must seek authorization at the FERC to recover RMR costs³².

Several parties, including TURN, FEA, SCE, and PG&E, address RMR costs in this proceeding. The Phase 1 decision explicitly stated that FERC has jurisdiction of RMR costs and defers to the FERC on all related matters (D.99-10-057, mimeo. at pp. 29-30, Finding of Fact 9, Conclusion of Law 11). Since this matter was fully litigated in Phase I of this proceeding we will not readdress the issue of post-transition treatment of RMR costs here.

E. Inclusion of Non-CTC Costs on the Ongoing CTC Rate Component

The utilities may continue to offer interruptible or curtailable service until March 31, 2002 pursuant to § 743.1. TURN and ORA object to PG&E's and Edison's proposal to include any discount costs and rate limiter adjustments in the ongoing CTC component on the customer bill. Edison and PG&E argue that the cost separation proceeding assigned these costs to the generation function. While the utilities concede that these costs are not transition costs as defined in § 367, the ongoing CTC component is the only remaining generation related component. ORA and TURN argue that inclusion of non-CTC costs should be rejected because it will be a misrepresentation of the costs in that component and is not easily explained as related to a particular utility function or activity. ORA and TURN further argue that the CTC component was created for the specific purpose of collecting transition costs as defined in § 367 and the costs stemming from interruptible programs and discounted rates are not transition costs. ORA suggests that these costs are more appropriately assigned to the distribution function. TURN does not specify which function the costs should be assigned to, but states that consistent with Commission policy of disclosing the source of utility costs, the utilities could add a line item on the customer bill for these costs.

We agree with TURN and ORA that including such non-transition costs in the CTC component is improper and unlawful and misrepresents the costs in that component since it was created specifically for the collection of transition costs. Although D.97-08-056 designated these costs as generation-related, we agree with ORA that they are more appropriately assigned to distribution function. Including the costs of interruptible discounts and rate limiter adjustments as a separate line item on the customer bill is administratively burdensome, although technically more correct. Since these costs are relatively short-lived, we will require this item to be assigned to the distribution function. We will not allow additional costs that are not transition costs to be collected in the CTC rate component.

ORA & PG&E agree that the power factor adjustment should be recovered through the distribution rate component.

F. Rate Group Transition Cost Memorandum Account

The Rate Group Transition Cost Memorandum Account (RGTCOMA) tracks transition costs obligations and payments by rate group. The first issue to be resolved concerning this account is whether during the rate freeze the transition cost allocators should be adjusted for energy use profiles among classes. FEA, SCE, and Large Users propose to adjust the allocators for changes in class energy use profiles. These parties state that class allocators must be adjusted for changes in energy usage pursuant to § 371(a). ORA opposes that approach stating that § 367(e) mandates that the allocators remain fixed as they were on June 10, 1996, notwithstanding changes in energy usage patterns.

The second issue involves post rate freeze reconciliation of RGTCOMA balances. ORA recommends reconciling the difference between rate group CTC obligations and CTC revenues received. That is, CTC will continue to be collected after the rate freeze for rate groups that have a remaining obligation determined by the RGTCOMA balance.

SCE, Edison,PG&E, and the Large Users argue that such a reconciliation of RGTCOMA balances after the rate freeze is unlawful as determined in D.99-10-057. In that decision, we determined that any carry over of costs into the post rate freeze period is unlawful pursuant to §§ 368(a) and 367(a) (Id., mimeo. at p. 36, Conclusions of Law 3 and 4.)

Section 371(a) states:

Except as provided in Sections 372 and 374, the uneconomic costs provided in Sections 367, 368, and 376 shall be applied to each customers based on the amount of electricity purchased by the customer from an electrical corporation or alternative supplier of electricity, subject to changes in usage occurring in the normal course of business.

Section 371(b) states:

Changes in the usage occurring in the normal course of business are those resulting from changes in business cycles, termination of operations, departure from the utility service territory.....

We have explained that, during the rate freeze, § 367(e)(1) requires that transition cost be allocated in substantially the same proportion as similar costs were allocated in June of 1996. However, we do not believe that the language "in substantially the same proportion" is necessarily in conflict with the § 371(a) provisions mandating that transition cost allocation reflect changes in usage profiles. We agree with FEA, Large Users, and Edison that the statute mandates that the allocation of transition costs be adjusted for changes in usage patterns. Therefore, during the rate freeze, while transition costs continue to be allocated using an EPMC or SAPC methodology, allocators shall be updated to reflect changes in class usage profiles occurring in the normal course of business.

Regarding reconciliation of RGTCOMA balances, D.99-10-057 is explicit that costs cannot be carried over after the rate freeze period. In that decision, we also established refund accounts and mandated that over-collections must be returned to ratepayers using the allocation method used in the collection of those costs. (Id., mimeo. at p. 16). In addition, it established that the rate freeze will end at the same time for all customers.

We reiterate that the rate freeze should end for all customers at the same time notwithstanding the class transition cost obligations in the RGTCOMA. Rate groups that have not met their transition cost obligation cannot continue to pay CTC post rate freeze as it would constitute a carry over of costs to the post rate freeze period. Such a carryover is unlawful pursuant to §§ 367(a) and 368(a). The RGTCOMA should be eliminated for each utility.

D.99-10-057 provides for the difference between the amount of CTC authorized and the actual amount collected to be returned to ratepayers at the utilities authorized rate of return. The utilities shall propose a method to return the funds in the first ATCP following the end of the rate freeze.