Word Document PDF Document

COM/SK1/ham

Decision 04-07-022

BEFORE THE PUBLIC UTILITIES COMMISSION OF THE STATE OF CALIFORNIA

Application of Southern California Edison Company (U 338-E) For Authority to, Among Other Things, Increase Its Authorized Revenues For Electric Service in 2003, And to Reflect That Increase in Rates.

Application 02-05-004

(Filed May 3, 2002)

Investigation on the Commission's Own Motion into the Rates, Operations, Practices, Service and Facilities of Southern California Edison Company.

Investigation 02-06-002

(Filed June 6, 2002)

(See Appendix A for a list of appearances.)

OPINION ON BASE RATE REVENUE REQUIREMENT

AND OTHER PHASE 1 ISSUES

TABLE OF CONTENTS

Title Page

OPINION ON BASE RATE REVENUE REQUIREMENT

AND OTHER PHASE 1 ISSUES

1. Introduction

1.1. Summary of Decision

Returning Southern California Edison Company (SCE) to conventional cost-of-service ratemaking after a six-year hiatus, we set the company's authorized base rate revenue requirement at $2.814 billion for the 2003 test year. On an annualized basis, this represents an increase of $73 million (2.66%) above SCE's present base rate revenue of $2.741 billion for 2003. SCE had requested an increase of $251 million (9.2%). The test year revenue requirement authorized herein will be implemented in accordance with Decision (D.) 03-05-076 and related determinations made in this decision.

SCE's base rate revenue requirement covers the costs of operating, maintaining and investing in the utility's generation, distribution, and central office functions. It excludes such costs as fuel, power procurement, and public purpose programs. In D.03-07-029 we provided for the reduction of SCE's retail rates by $1.249 billion annually upon the utility's recovery of the balance of its Procurement Related Obligations Account (PROACT). This reduction was calculated using an estimate of the total bundled service ratepayer revenue responsibility of $8.472 billion, which includes the share of the Department of Water Resources revenue requirement paid by SCE's customers. Thus, the base rate revenue requirement that we authorize today, while substantial, represents approximately one-third of the consolidated revenue requirement being paid by SCE's bundled service customers. The adopted test year base revenue requirement increases the bundled revenue requirement by 0.86%.

Pursuant to the Commission's order in D.03-07-029, SCE's electric rates will be increased on a system average percentage change (SAPC) basis to give effect to the base rate revenue requirement increase adopted today. In Phase 2 of this proceeding, the Commission is evaluating proposals regarding the allocation of revenue requirement responsibility to customer classes and the design of rate structures.

We approve SCE's request to establish a late payment charge for residential customers along with an exemption for customers enrolled in the California Alternate Rates for Energy (CARE) program. We also approve in part SCE's request to adjust its charges for returned checks, reconnects, service establishment, and field assignment. We do so to more closely align rates and charges with the principle of cost causation. We adopt standards for various areas of customer service and require that compensatory rebates be paid to affected customers when SCE fails to meet those standards.

We adopt, with revisions, SCE's proposed "post test-year ratemaking" (PTYR) mechanism to adjust the authorized revenue requirements for 2004 and 2005. The PTYR mechanism ties capital forecasts to actual projects in SCE's budget subject to a true-up procedure. In connection with the PTYR mechanism, we approve a refueling and maintenance outage expense recovery mechanism for San Onofre Nuclear Generating Station Units 2 & 3 (SONGS 2 & 3).

This decision reviews certain 1997-98 generation capital additions, consideration of which was transferred from Application (A.) 99-04-024 to this proceeding. SCE is authorized to recover costs associated with $30.937 million in capital additions found to be reasonable.

Proposals by SCE and other parties to establish a system of safety, reliability, and customer satisfaction performance incentives are granted.

Finally, in this decision, we examine certain of the roles fulfilled by SCE on behalf of its customers and other stakeholders. We review and comment on SCE's role with respect to integrated resource planning and whether it should be prepared to build or buy utility-owned generation capacity to serve its customers. We also review SCE's Women, Minority, and Disabled Veterans Business Enterprise (WMDVBE) program and the diversity of its workforce.

With this decision, Phase 1 of this general rate case (GRC) proceeding is concluded. Phase 2 of this proceeding addresses SCE's pricing proposals and will be resolved by future order of the Commission. This proceeding therefore shall remain open.

1.2. Background

In SCE's last GRC, D.96-01-011 established SCE's authorized revenue requirement for the 1995 test year. Pursuant to D.96-09-092, SCE has operated under a PBR mechanism since January 1, 1997. Pursuant to D.01-06-038 and D.02-04-055, the PBR mechanism remains in effect, with modifications, until it is superseded by the issuance of a decision in SCE's next GRC, i.e., the instant proceeding.

On May 3, 2002, SCE filed A.02-05-004 seeking, among other things, an increase in its authorized test year 2003 base rate revenue requirement. SCE originally sought authorization for revenues of approximately $3.065 billion for 2003, which represented an increase of $286 million (10.3%) above the currently authorized base rate revenue as then calculated. During the course of the proceeding, SCE revised both its request and its calculation of present revenue. Based upon its latest calculations, SCE now seeks authorization for base revenue of approximately $2.992 billion for 2003. This represents an increase of $251 million (9.2%) above the base rate revenue, now calculated at $2.741 billion. SCE also seeks authority to establish a post test-year ratemaking mechanism that would set the authorized base revenue requirements for 2004 and 2005. In addition, SCE seeks authority to establish a late payment charge for residential customers, and to increase various fees such as charges for returned checks, service establishment, and reconnection.

The Commission instituted Investigation (I.) 02-06-002 on June 6, 2002, to allow the Commission to hear proposals other than SCE's, and to enable the Commission to enter orders on matters for which the utility may not be the proponent. The Commission ordered that A.02-05-004 and I.02-06-002 be heard on a consolidated evidentiary record.

Prehearing conferences were convened on June 13 and November 1, 2002. Public participation hearings were held at 14 locations throughout SCE's service territory in October 2002. Direct and rebuttal evidentiary hearings were held before Administrative Law Judge (ALJ) Wetzell on 38 days from November 2002 to March 2003. Briefs were filed on April 18, 2003 and reply briefs were filed on May 28, 2003.1 SCE and San Diego Gas & Electric Company (SDG&E) served update testimony on May 9, 2003. Phase 1 was submitted for decision on October 23, 2003. Final oral argument before the Commission was held following the issuance of the ALJ's proposed decision.

In addition to SCE, the active parties in Phase 1 of this proceeding were the Office of Ratepayer Advocates (ORA), Aglet Consumer Alliance (Aglet), The Utility Reform Network (TURN), SDG&E, the Coalition of California Utility Employees (CUE), The Greenlining Institute (Greenlining),2 the California Disabled Veterans Business Enterprise Alliance (DVBEA), the Natural Resources Defense Council (NRDC), and the County of Los Angeles (LA County). The positions taken by the parties are described throughout this opinion.

ORA, Aglet, and TURN have taken positions affecting the forecast of SCE's base rate revenue requirement. As set forth in the March 2003 Joint Comparison Exhibit (Exhibit 403), ORA's base rate revenue requirement recommendation for 2003 is $2.625 billion, or $364 million less than SCE's request.3 Due to the complexities of calculating revenue requirements reflecting parties' positions on the various underlying components, Exhibit 403 does not include a calculation of the revenue requirement recommendations associated with the positions of Aglet or TURN.

1. Preliminary Matters

Our primary task in this decision is determining the just and reasonable base revenue requirement for SCE for the 2003 test year. We accomplish this task, as well as the resolution of the other Phase 1 matters at issue, by evaluating and resolving approximately 150 separate issues, most of which were contested. We will first address certain overarching matters that warrant discussion.

1.1. The Utility's Showing

In a 1992 SDG&E proceeding, the Commission stated its expectations for utility showings in GRCs:


The purpose of a general rate case is to develop and adopt sound, informed estimates of the reasonable costs to be incurred in the test year. We know that our adopted levels of revenues and expenses may be at variance with actual experience. However, we must be sufficiently informed to know that adopting a given estimate makes sense. Part of this process involves making sure that we do not repeatedly approve revenues to meet a one-time cost. When a utility's expense estimate includes the performance of a task it had planned to accomplish with previously authorized funds, we will want to know why the utility did not spend its funds as planned the first time around and will be hesitant to charge ratepayers twice for the same expense. In addition, we want to be confident that the activities being undertaken by the utility are lawful and otherwise consistent with public policy. (D.92-12-019, 46 CPUC 2d 538, 555.)


The company often does not even mention the name of major programs or activities and almost never adequately explains its basis for forecasting related costs. The application often makes only a general request for funds without providing a reasonable, well-explained justification.4 While approving [the settlement at issue in that decision], we wish to make it clear to SDG&E and other utilities that the initial showing in the current case does not meet our requirements.5 (Id.)

Discussion

SCE seems to have taken at least some of the 1992 Commission's concerns to heart in this proceeding. The volume of material that SCE submitted in its direct and rebuttal evidentiary showings was nothing short of massive, and SCE claims that it submitted "the most comprehensive showing SCE has ever made." (SCE Opening Brief, p. 1.) SCE states that it submitted more than 5,400 pages of prepared direct testimony (not including rebuttal testimony). For administrative and general (A&G) issues alone, SCE included 1,500 pages of prepared testimony supported by 10,340 pages of workpapers, sponsored by 35 witnesses. In comparison, in its last GRC SCE's showing included 400 pages of prepared A&G testimony by nine witnesses. Moreover, SCE continued to think big when it tendered an 856-page opening brief.6

Although we appreciate SCE's apparent attention to the Commission's stated concerns about guarded initial showings, we are compelled to observe that size alone does not constitute fulfillment of the utility's obligation to explain and justify its request. In fact, an overly massive utility showing can obscure the utility's substantial justification for its request (or lack thereof), thereby detracting from the parties' and the Commission's ability to conduct timely review and evaluation. We must question whether it is reasonable to attempt the complete processing of any case with a volume of documentation that even approaches "the most comprehensive showing ... ever" within the confines of evolving expectations for the timely conclusion of our proceedings.

Accordingly, we now request that in presenting their initial rate case showings, utilities work to provide the necessary justification with greater attention to the need for economy of words and data. We are not in any way retreating from our policy of requiring better initial utility showings than the one we encountered in the 1992 SDG&E proceeding. We are simply directing utilities to work at being more efficient in their presentations, which in turn should enable the Commission to administer its proceedings with greater efficiency. We invite utilities to consider, for example, whether the inclusion of a wiring diagram that depicts the type of excitation system used at coal-fired generation facilities adds needed evidentiary support for their funding requests. (See Exhibit 17, pp. 67, 69.) Elimination of duplicative material may be helpful. (See Exhibit 55, pp. 9, 57.) We also invite utilities to review the Commission's discussion in D.93-04-056, where it proposed the use of exhibit and chapter summaries to focus attention on what the request or issue really is, and whether there is an explanation for it being found reasonable. (49 CPUC 2d 72, 88.). We also ask that ORA and other intervenors make efforts to ensure that their participation contributes to the efficient processing of our rate proceedings.

ORA reminds us that SCE must meet the burden of proving by clear and convincing evidence that it is entitled to the relief it is seeking in this proceeding, and that the burden is not on ORA or other intervenors to demonstrate that SCE's request is unreasonable.7 We intend to hold SCE to this standard as we examine individually the myriad components of SCE's request.

As a general matter, with respect to individual uncontested issues in this proceeding, we find that SCE has made a prima facie just and reasonable showing unless otherwise stated in this opinion.

1.2. SCE's Financial Health

SCE's Chairman and Chief Executive Officer (CEO), Alan J. Fohrer, testified that in the wake of the California energy crisis that brought SCE to the brink of financial collapse, restoration of investor confidence is essential if the company is to provide the service its customers expect. He went on testify that:


Commission general rate case decisions have always been important to investors' assessments of regulatory support for a utility's continued financial well-being. In this instance, Commission approval of SCE's investment and operating plans is especially important to signal a return to the supportive regulation historically found in California. (Exhibit 1, p. 4.)

Discussion

We understand that the investment community is vitally interested in the decisions of this Commission. We also recognize that an investor-owned utility's credit rating and its access to capital are of critical importance to its ability to provide the infrastructure it needs to meet its customer service obligations. However, we find no evidence convincing us that granting SCE the full amount of its requested test year base revenue is a necessary precondition for the company to achieve the financial health it requires to provide adequate utility service. To the contrary, evidence introduced by Aglet shows that the company's financial condition has already improved greatly since the height of the state's energy crisis and SCE's financial crisis. Indicators supporting this assessment include an improved credit rating of BB by Standard and Poor's in March 2002, SCE's achievement of certain standards marking progress towards its qualifying for an investment grade rating, the payoff of defaulted obligations, a much-reduced threat of bankruptcy, and the recovery of the PROACT undercollection.

We believe that our actions in response to SCE's financial crisis have been and will continue to be supportive of the utility, and we will consider SCE's creditworthiness as we consider the components of its GRC request. Nevertheless, if we find that SCE has included unnecessary or unreasonable expenses or capital projects in its GRC request, we will not hesitate to exclude the item from ratepayer funding responsibility. Similarly, we will not approve higher depreciation rates than we otherwise would solely in an effort to help SCE achieve a better credit standing. In summary, our concept of "supportive regulation," unlike that of SCE's CEO, apparently, includes support for the interests of ratepayers as well as the interests of the utility and its investors.

1.3. Comparative Rate Levels

ORA points out that SCE's rates are among the highest in the nation even without the revenue increase sought by SCE in this GRC. The following facts make ORA's point and expand on it:


1. In 2001, SCE's average residential rate was 13.29 cents per kilowatt-hour (kWh), the California average was 11.40 cents per kWh, and the national average was 8.47 cents per kWh.


2. From 1996 through 2001, SCE's residential rates were between 9% and 18% higher than the California average and 37% to 62% higher than the national average.


3. In 2001, SCE's commercial rates reached 17.04 cents per kWh while the national average was 7.35 cents per kWh.


4. From about 1996 to 2000, SCE's industrial rates stayed at 8.05 cents per kWh, then, in 2001, increased to 12.13 cents per kWh, or about 240% of the national average.


5. In 2001, SCE's residential, commercial, and industrial rates were even higher than those of Pacific Gas and Electric Company (PG&E) and SDG&E, with the exception of SDG&E's residential rates.


6. In 2001, average California rates (including municipal utilities) were 11.4 cents per kWh for residential service, 11.2 cents for commercial service, and 7.6 cents for industrial service. While these averages were considerably higher than national average, they were still low compared to SCE's rates of 13.29 cents, 17.04 cents, and 12.13 cents.

Arguing that high electricity rates are damaging to California's economy and to SCE itself (because some customers are able to bypass SCE's system and will do so if high rates make it cost effective to do so), ORA asks the Commission to look at all rate case increase requests with skepticism, and to adopt only reasonable rates that are supported by record evidence.

Discussion

The evidence shows that the electric rates of investor-owned California utilities, particularly those of SCE, are comparatively high by several standards, and that ratepayers and the California economy are being harmed by these high rates. SCE's attempt to change the topic to customer bills rather than high rates does not change this fact, nor does SCE's attempt to assign part of the blame for high rates to the Division of Ratepayer Advocates (DRA), ORA's predecessor. ORA has also demonstrated that SCE's high rates add to the threat of utility bypass in various forms such as municipalization, firms moving operations out of the utility's service territory, and self-generation, and that residential ratepayers who have fewer alternatives to utility service are the most threatened by bypass.

Of course, most of the rate comparison data supplied by ORA pertains to the year 2001, a time of unprecedented electricity market disruptions and extraordinarily high commodity prices. It must be recognized that the high rate problem in 2001 was in large part a reflection of the commodity market conditions of that time. Indeed, while SCE's base rate revenue requirement is substantial, and determining its appropriate level is the major undertaking of this GRC, it is worth noting that the outcome of this proceeding can at best provide only a partial solution to the high rate problem. This can be seen by comparing the difference between SCE's and ORA's revenue requirement recommendations ($364 million) with the consolidated revenue requirements of $8.472 billion being paid by SCE's customers, as set forth in D.03-07-029. Even if ORA's revenue requirement recommendation were adopted in full, SCE's electric rates might remain high by national standards.

Regarding the high rate problem, ORA witness Phan testified that "[a] regulated utility will continue to expand its services, personnel, rates, and revenues until and unless it is halted from the outside." (Ex. 113, p. 2-13.) In the context of this GRC proceeding, we see our role as providing such an "outside" check by conducting a thorough review of SCE's request and, as discussed earlier, holding SCE to its burden of proving that its revenue requirement request is justified. In this manner, we ensure that SCE's revenue requirement is set at a level that does not contribute unnecessarily, if at all, to the high rate problem.

1.4. Forecasting Issues

A central feature of conventional cost-of-service method of ratemaking is forecasting future test year costs using historical cost information as well as current information regarding the utility's operational and investment plans. In this section, we address general principles applicable to forecasts of both expenses and capital expenditures.8

1.4.1. Averaging and Other Methodologies

Several different methods can be used to calculate test year estimates of expenses. These include linear trending, averaging, last recorded year, and budget based estimates. In PG&E's 1999 GRC, the Commission summarized certain methodological issues as follows:


The Commission has recognized that there are different valid and acceptable methods for account-by-account forecasting test year costs in a GRC, including using a single recorded year's expenses . . . and using multi-year average recorded costs . . . . The question at hand is which of these two methods yields the most accurate and reliable forecast of test year expenses. In PG&E's test year 1990 GRC the Commission described the following criteria for developing a base estimate of test year expenses:


If recorded expenses in an account have been relatively stable for three or more years, the 1987 recorded expense is an appropriate base estimate for 1990.


If recorded expenses in an account have shown a trend in a certain direction over three or more years, the 1987 level is the most recent point in the trend and is an appropriate base estimate for 1990.


For those accounts which have significant fluctuations in recorded expenses from year to year, or which are influenced by weather or other external forces beyond the control of the utility, an average of recorded expenses over a period of time (typical four years) is a reasonable base expense for the 1990 test year. (D.89-12-057, 34 CPUC 2d 199, 231.)

With respect to a particular account in that GRC (Account 588), the Commission went on to state:


Absent a specific explanation of why 1987 recorded data best reflects the estimated 1990 expenses of an account with fluctuating expense levels and no discernible trends, we find it most appropriate to use a four-year average as the base 1990 estimate. (Id., 238.)


(Re Pacific Gas and Electric Co., D.00-02-046, pp. 156-157 (quoting, D.89-12-057, 34 CPUC 2d 199).)

While the foregoing passage arose in the context of electric distribution expenses, the principles are consistent with findings the Commission has made in the context of other expense categories. For example, the Commission addressed forecasting methodologies in a 1989 decision on a water utility's request for a rate increase:


Where it is clear that there is a trend of increasing expenses which cannot be explained by inflation alone, and that such increases are necessarily incurred in providing utility service, less weight should be given to the constant dollar averaging method. On the other hand, where it appears that an expense category is subject to year-to-year variations, constant dollar averaging may be a more appropriate method to smooth out such variations. (Re California Water Service, D.89-04-060, 31 CPUC 2d 481, 506.)

Discussion

The forecasting principles discussed in D.00-02-046, D.89-12-057, and D.89-04-060, quoted above, are generally appropriate and applicable here. Many accounts reflect significant spending fluctuations from year to year, and in the absence of information to the contrary (for example, information that expenses have been stable over time and no discernible upwards or downwards trend exists), we would ordinarily expect a multi-year average of recorded data to yield a more reliable forecast than a forecast predicated upon a single year's data. Also, because utility spending plans may not always be implemented as intended, budget-based forecasts generally will be given less weight than forecasts based on recorded spending in the absence of a showing supporting the contrary approach. For ongoing functions, a multi-year spending pattern, as reflected in the recorded books of account, suggests a utility's willingness and ability to commit to a budgetary plan on a sustained basis.

1.4.2. Capital Expenditures & PBR

Aglet raises two forecasting issues associated with the utility's return to cost-of-service ratemaking. First, Aglet takes issue with SCE's position that capital spending associated with a program to replace aging distribution assets (Infrastructure Replacement Program or IRP) was not reflected in rates in recent years. Aglet acknowledges that IRP costs were not explicitly included in the base revenue requirement that formed the starting point for SCE's PBR mechanism. However, Aglet disputes any contention that this means that SCE has been unable to recover its IRP costs under PBR. This is because under PBR, where allowable capital costs are based on a formula and are not tied to specific projects, there is no way to determine with any precision what capital-related costs are included in rates. At the same time, Aglet contends that in light of the well-settled regulatory principle that utilities are not required to make specific expenditures in exactly the same manner and amounts that underlie authorized rates, it is reasonable to assume that all incurred costs are included in rates even when the utility spends more than the authorized revenue requirement.

Moreover, Aglet notes that SCE recorded positive earnings in most quarters that its PBR mechanism was in effect, suggesting that the company recovered its operating expenses, depreciation, and tax costs during that period. Accordingly, Aglet urges that in determining the need for increased capital costs for the IRP, we should recognize that present rates have recovered ongoing capital-related costs of plant installed to date.

Second, Aglet asks that we recognize that in recent years under PBR, SCE has had an incentive to recover capital-related costs in rates while not actually spending on the IRP. Aglet urges that given the existence of this incentive, we should be very cautious in approving additional capital costs for improvements that were deferred during the SCE financial crisis.

Discussion

We accept as valid the theory that capital expenditures such as those associated with the IRP have been included in rates during the term of SCE's PBR mechanism. To assume otherwise would require acceptance of the unreasonable and unproven hypothesis that SCE's PBR mechanism failed to provide the level of revenues needed by the utility to meet its service obligations. With the return to conventional cost-of-service ratemaking, this theory requires that we carefully review the incremental capital expenditures for which SCE seeks recovery. In particular, we will seek reasonable assurance that ratepayers do not contribute twice for the capital-related costs of installed plant. Moreover, as discussed in Section 3 of this decision, this theory applies as well to the return of SCE's nuclear generation from incentive to conventional cost-of-service ratemaking.

Aglet has also shown that SCE had an incentive to defer capital expenditures with the expectation of a return to cost-of-service ratemaking. We will heed Aglet's call for caution in reviewing deferred projects. However, if a deferred project is reasonably calculated to enable the utility to provide safe and reliable service on a cost-effective basis, ratepayers have not already paid for it, and the deferral did not cause cost increases that could have reasonably been avoided, we will not deny funding for that project solely because it has been deferred.

1.4.3. Witness Qualifications

In rebuttal testimony an SCE witness suggested "that judgment, experience, training and education are required to make an informed decision regarding forecasts." (Exhibit 299, p. 55.) SCE makes similar statements elsewhere in its testimony.

Discussion

We accept this proposition, but we reject any implication that such qualifications are the exclusive province of utility witnesses, who will invariably have the most intimate knowledge of their area of operations. In other words, ORA and intervenor analysts are entitled to offer and have us consider their expert opinions based on informed judgment, even if they have never been employed by a public utility.

Selecting the most appropriate method to forecast test year expenses is ultimately a matter of informed judgment, and whether a witness is employed by a utility or an intervenor, the exercise of judgment might be influenced by a concern for results. With this in mind, we will carefully review each witness's underlying assumptions, analysis and logic before accepting the forecast recommended by that witness.

2. Generation

SCE retains interests in nuclear, coal, and hydroelectric generation facilities. SCE seeks base rate recovery of the costs of these assets, excluding fuel costs recovered through the Energy Resources Recovery Account (ERRA) mechanism. SCE also requests rate recovery of certain 1997-98 capital additions, consideration of which was transferred from A.99-04-024 to this case.

2.1. San Onofre Nuclear Generating Station

SCE owns a 75.05% share of San Onofre Nuclear Generating Station (SONGS) Unit Nos. 2 and 3 and is also the operating agent for those units. SCE also owns 80% of SONGS Unit No. 1, which has been permanently shut down. In D.96-04-059, the Commission adopted the Incremental Cost Incentive Pricing (ICIP) ratemaking mechanism for SONGS 2 & 3, a form of PBR under which SCE recovers incremental operational costs through a cents-per-kWh payment. The ICIP mechanism will continue in effect until December 31, 2003, after which SONGS 2 & 3 will return to cost-based ratemaking consistent with D.01-01-061 and D.01-05-035. No party contests SCE's proposed framework for conventional cost-of-service ratemaking for SONGS 2 & 3.

2.1.1. SONGS 2 & 3 Capital

For SONGS 2 & 3 plant that will be placed in service in 2004 and 2005, SCE forecasts capital expenditures (excluding corporate overheads) of $81.7 million in 2004 and $63.2 million in 2005. SCE's share of those expenditures, which reflects its 75.05% plant ownership share, is $61.3 million in 2004 and $47.5 million in 2005. SCE reduced its estimate for the Used Fuel Pool Rack Modification by $520 thousand in 2004 and by $571 thousand in 2005 upon accepting ORA's position on that project. SCE's total capital expenditure forecast is reduced accordingly.

ORA proposes the inclusion of capital expenditures (SCE share) totaling $41.1 million in 2004 and $33.2 million in 2005. Compared to SCE's forecast, ORA's estimates represent reductions of $19.7 million in 2004 and $13.6 million in 2005. As shown in Exhibit 403, TURN proposes reductions to SCE's forecast of $17.9 million for Marine Mitigation, $25.7 million for Used Fuel Storage, and $1.9 million for Blanket Work Orders. The total difference between TURN and SCE is $45.5 million (SCE share) for the two years.

SCE contends that the correct way to forecast the SONGS 2 & 3 capital budget is to make a project-by-project investigation of the actual work to be performed. SCE's forecast reflects this methodology.

For its forecast, ORA used both averaging and project-specific information. ORA used five years of historical data (1996-2000) to estimate expenditures for those capital additions that will be basically the same from year to year. For two projects that ORA found to be fully vetted and approved through SCE's internal processes, the SONGS 2 & 3 Used Fuel Storage Project and Cycle 13 Used Fuel Rack Modifications, ORA used project-specific information to develop its forecast. ORA recommends disallowances for two other projects: Wetlands Reclamation ($9 million for 2004 and $9 million for 2005, SCE share), and Replacement of Offsite Sirens/Monitors ($2.7 million for 2004 and $300 thousand for 2005, SCE share).

TURN objects to SCE's forecast on the grounds that SCE substantially underspent on planned capital additions during the ICIP period, and that the company's project-specific forecast amounts to double charging ratepayers for certain activities that were planned in the past and included in ICIP rates. SCE spent $182 million during the 1996-2003 ICIP period, or approximately 36% of the $510 million forecast used in the establishment of the ICIP mechanism. The actual spending was an average of $22 million per year during this period, and the highest annual spending was approximately $36 million in 1998. In view of this underspending of nearly $330 million for the eight-year ICIP period, TURN submits that SCE's proposal to more than double the average spending of the ICIP period is unjust and inequitable.

SCE acknowledges that recent historic averages of capital expenditures for SONGS 2 & 3 can provide a helpful comparison for forecast capital expenditures levels, but it claims that ORA's forecast fails to provide sufficient funding for projects necessary for continued safe, compliant, and reliable operation. In addition, SCE finds fault with ORA's averaging methodology because it failed to escalate capital expenditures to constant 2000 dollars prior to applying the five-year average methodology. According to SCE, correcting this methodological error changes ORA's recommendation for SONGS 2 & 3 capital funding to $60.1 million for 2004 and $50.7 million for 2005 (nominal dollars at 100% level).

Discussion

A forecast that reflects the company's actual plans for future capital expenditures might seem to be more reliable than a forecast that is based solely on the level of past investments. After all, the utility experts having the most detailed knowledge of operational requirements should be in the best position to know what capital additions are needed for future operations and what their costs are.

However, utility spending plans are not necessarily carried out. In the first place, there is no specific obligation under conventional cost-of-service or incentive ratemaking to spend budgeted amounts during the relevant time period. Moreover, as SCE witness Perez stated:


SCE cannot rigidly "fix" the detailed specific scope of capital work to be implemented in future years. SCE requires flexibility to optimally respond to changing NRC requirements, plant reliability or operability changes, results of studies and conceptual or preliminary engineering, industry developments, replacement energy costs, and other evolving factors. (Exhibit 9, p. 4.)

Given these many reasons why spending plans may not be carried out, regardless of how much site-specific expertise went into the making of those plans, a utility's budget-based forecast may actually be less reliable than a forecast based on historical spending data. Historical spending patterns reflect not only past spending plans, but also the utility's willingness and ability to carry out those plans. It is instructive that SCE had forecast capital spending of $510 million during the 1996-2003 period for the purpose of establishing the ICIP mechanism, but it actually will have spent just slightly more than a third of the forecast amount. Also, as ORA notes, SCE's budget-based forecast of SONGS 2 & 3 capital expenditures represents more than a doubling of the 1996-2000 average level of expenditures. Specifically, SCE forecasts spending increases of 162% above the five-year average for 2004 and 103% above the average for 2005.

Rebutting TURN's position on SCE's underspending pattern, and TURN's related contention that SCE's forecast would lead to double charging ratepayers for the same investments, SCE goes to great lengths to explain the history of the ICIP mechanism. SCE focuses on the fact that the mechanism eliminated retrospective reasonableness reviews of SONGS capital expenditures, and claims that TURN is attempting to institute an impermissible review of the timing of certain SONGS 2 & 3 expenditures in this proceeding. SCE is missing the point of TURN's analysis, which is to determine the reasonableness of forecast expenditures by, among other things, determining whether it can reasonably be concluded that ratepayers have already paid, in whole or in part, for such expenditures. This is a legitimate inquiry that does not in any way represent an after-the-fact reasonableness review precluded by the ICIP mechanism. We find that TURN has raised a valid point, i.e., ratepayers have already made contributions to the Used Fuel Storage and Marine Mitigation projects through the ICIP rates.

All of these factors support the use of an average of actual expenditures as being more reliable than a straight budget-based forecast of SONGS 2 & 3 capital expenditures, and disallowances for specific projects to the extent necessary to prevent double charging ratepayers for those projects. While SCE has provided a substantial amount of material to justify various SONGS 2 & 3 projects included in its capital forecast, it has not convinced us that it is reasonable to (1) ignore its spending history, (2) assume that all of the justified projects will be carried out in 2004 and 2005 as planned, and (3) assume that ratepayers have contributed nothing to any of the deferred projects during the ICIP period.

We will therefore adopt ORA's blended forecasting approach as the more reliable method for determining SONGS 2 & 3 capital spending in this GRC. However, we do not adopt ORA's specific capital spending recommendations. First, ORA's averaging approach requires correction to account for the effects of inflation. Also, we do not adopt all of ORA's project-specific recommendations, and we adopt adjustments based on TURN's recommendations. We turn our attention to these project-specific recommendations, and then we present our adopted forecast for SONGS 2 & 3 capital expenditures.

SCE projects that the SONGS 2 & 3 fuel pools will run out of space in July 2007 and March 2008, respectively. Construction of a temporary Used Fuel Dry Storage Facility (dry cask storage) will provide necessary onsite spent fuel storage capacity. SCE forecasts that it will incur capital expenditures of about $34 million (nominal dollars, 100% level; SCE's share is $25.7 million) for the Used Fuel Storage project in 2004 and 2005.

Based on its position that ICIP rates have already fully covered the costs of the project, TURN recommends a disallowance of the entire $25.7 million sought by SCE.

Discussion

SCE identified a used fuel storage project that it planned to build during the ICIP period, but the project was deferred. While we reject any assumption that ratepayers will have contributed nothing to projects such as the Used Fuel Storage project during the ICIP years, we are not prepared to accept TURN's opposite assumption that ratepayers have fully paid for the project's capital costs. Nothing in the decisions leading to the establishment of the ICIP mechanism suggests such precise attribution of the ICIP rate to specific projects is warranted. TURN's proposal to explicitly disallow the entire amount of SCE's projected costs for the Used Fuel Storage project is denied. Because it is reasonable to determine that ratepayers have made contributions to the cost of this project, but it is impossible to calculate the precise amount of that contribution, the fairest outcome is to assign equal cost responsibility for remaining costs of the project. Accordingly, we allow 50% of SCE's forecast cost.

SCE is required to mitigate impacts on the marine environment in compliance with the Coastal Development Permit (CDP) issued by the California Coastal Commission (CCC) for SONGS 2 & 3. These mitigation requirements include development and implementation of a wetland restoration project to mitigate fish losses and construction of an artificial kelp reef to mitigate impacts on the San Onofre kelp reef. SCE forecasts that it will spend $13.2 million in 2004 and $20 million in 2005 on marine mitigation projects (nominal dollars, 100% level; SCE's share for both years is $24.9 million).

TURN proposes a disallowance of $17.9 million, or all but $7 million of the $24.9 SCE forecast for marine mitigation. TURN does so to prevent SCE from recovering the same mitigation costs twice. TURN contends that the ICIP rates provided for the recovery of $75 million in marine mitigation costs, and that the forecast underlying the ICIP mechanism estimated that $7 million would be spent after 2004.

ORA proposes disallowances of $9 million in 2004 and $9 million in 2005 associated with the restoration of 20 acres of wetlands at San Dieguito.

Discussion

As with the Used Fuel Storage project, TURN's recommended disallowance is based upon the assumption that ratepayers have already paid for the work (all but $7 million) during the ICIP period. Again, we are not prepared to make such an assumption, and we therefore deny the recommended disallowance. It is more reasonable to assume that a portion of the ICIP rate was associated with marine mitigation costs, and that the use of a forecast that assigns equal ratepayer and shareholder responsibility for remaining costs strikes the most reasonable balancing of interests.

SCE has shown that ORA's recommended disallowances associated with 20 additional acres of wetlands restoration at San Dieguito is inappropriate because the restoration has no effect on customer rates. This proposed disallowance will not be approved. The amounts in question ($18 million for 2004 and 2005, SCE share) are for the mandated restoration of 150 acres at the San Dieguito site.

SCE's capital expenditure forecast includes $3.935 million (nominal dollars, 100% level; SCE's share is $2.953 million) for replacement of off-site sirens and monitors in 2004 and 2005. SCE originally installed the off-site siren system in 1982, and determined it needed replacement when it showed visible signs of deterioration and spare components were no longer available from the manufacturer. In addition, the software for the monitoring system is no longer supported by its manufacturer. In 1999, SCE commenced a program to replace the off-site siren system, even though it was then, and remains today, fully functional and consistent with regulatory requirements.

ORA asserts that SCE unreasonably deferred the off-site siren and monitor replacement project to 2004 and 2005, and should have performed it earlier. ORA therefore proposes the removal of the forecast cost of the off-site sirens and monitors from SCE's capital expenditure forecast for SONGS 2 & 3.

Discussion

The evidence does not support a conclusion that SCE unreasonably, and without cause, deferred this necessary project to the test period. In fact, SCE began investing in the off-site sirens and monitoring replacement project in late 1999 under the SONGS 2 & 3 ICIP rate mechanism. SCE's requested funding for the off-site siren and monitoring replacement project is essential to providing a seamless transition to the new system without risking public health and safety, and should be allowed in rates.

TURN recommends disallowances of SCE's increased spending on several small work orders, including tools and equipment, computers, office equipment, facilities, and spare parts. As shown in Exhibit 403, TURN's recommendation would result in a disallowance of $1.9 million.

Discussion

TURN's analysis is not without merit. However, we are adopting ORA's forecast methodology, which relies on the averaging of historical expenditures, which in turn results in disallowances of certain of SCE's proposed expenditures. We will not order additional disallowances for capital expenditures that are covered by ORA's forecast. Doing so could result in an unfair duplication of disallowances.

Applying ORA's blended forecasting approach, the corrections set forth in Exhibit 280, Appendix B, and the project-specific determinations discussed above, we adopt the following capital expenditures forecast for SONGS 2 & 3 for 2004 and 2005:

SONGS 2 & 3 Adopted Capital Expenditures

($1,000)

2.1.2. SONGS 2 & 3 Base O&M Expenses

Although SONGS 2 & 3 will not be returned to conventional cost-of-service ratemaking until 2004, SCE presented its O&M forecast for 2003 with the intent that it will be escalated to 2004 and 2005 levels. SCE at first forecast O&M costs of $147.8 million for SONGS 2 & 3 in 2003 (2000 dollars, SCE share). In its May 2003 update testimony, SCE increased its forecast to $148.4 million to reflect the cost of increased annual fees assessed by the Nuclear Regulatory Commission (NRC). This forecast is for what SCE considers to be base O&M, i.e., for normal plant operations. It excludes refueling and maintenance outage-related O&M costs, the amount and the recovery of which are addressed in Section 3.1.3.

SCE developed its base O&M forecast by analyzing recorded costs for the years 1996-2000, making 34 "historical" adjustments to remove cyclical and unusual expenses incurred during the recorded period, and making 11 "future" adjustments to incorporate anticipated cyclical and unusual activities and expenses. It then selected what it considered to be the appropriate method for converting the recorded/adjusted data into test year forecasts for various functions and accounts, as described below.

SCE manages SONGS 2 & 3 in accordance with eleven functional groups: Operations, Maintenance, Engineering, Site Projects, RadChemical Control, Regulatory Affairs, Security, Training, Nuclear Support, Corporate Support, and Participants. To reconcile this breakdown of functions with the Federal Energy Regulatory Commission (FERC) Uniform System of Accounts, SCE organized O&M costs by these functional groups and then, within each such group, by FERC Account. Using the recorded/adjusted costs for 49 of these functional group/FERC account pairings, SCE developed labor, non-labor, and other cost estimates for each such pairing using a forecast method such as last recorded year; three-, four-, or five-year averaging; and three-, four-, or five-year linear trending.

ORA recommends several revisions to SCE's base O&M forecast. In combination, ORA's recommended adjustments result in a forecast of $130.4 million (2000 dollars, SCE share), or $17.4 million less than SCE's "pre-update" forecast of $147.8 million. TURN recommends a reduction of $1.7 million (2000 dollars, 100% share) for an adjustment related to workers' compensation claims.

In the remainder of Section 3.1.2 of this decision, we address issues that affect the parties' base O&M expense forecasts, including disputes over adjustments and forecast methods. Unless otherwise indicated, dollar figures in the following discussion are in 2000 dollars, 100% share. Also, some dollar figures stated below represent record evidence that predated submission of the Joint Comparison Exhibit (Exhibit 403) and the update testimony (Exhibit 411). Notation is made where these dollar figures are revised or updated in Exhibits 403 or 411.

In this historical adjustment (Adjustment #20), SCE removed credits to its training program received from the State of California as part of its Employment Training Program (ETP). The ETP provides for training for workers facing layoff because of technological advances and foreign and domestic competition. During the period May 1998 through May 2000, SONGS 2 & 3 was eligible to participate in ETP because it was facing out-of-state domestic competition. According to SCE, SONGS 2 & 3 is no longer eligible to participate in the ETP because, with the return to the cost-of-service ratemaking, it is no longer confronted with out-of-state domestic competition.

Noting that SCE has not opted out of the ETP altogether, ORA proposes to disallow this adjustment. This would reduce SCE's O&M forecast by $367,000.

Discussion

The evidence shows that SCE is ineligible to participate in the ETP with the return of SONGS 2 & 3 to cost-of-service ratemaking. Since SCE cannot participate in the program, it would be unreasonable to establish the authorized revenue requirement on the basis of an assumption that it will continue to receive credits from the program. ORA's proposed disallowance of this adjustment is rejected.

ORA takes issue with SCE's Adjustment #30, which SCE made to account for certain deferred O&M projects. This historical adjustment is not described in SCE's direct testimony on generation, though they did describe the adjustment in their rebuttal testimony as necessary to add expenditures back into the 1998-2000 period where they normally would have occurred. ORA describes the adjustment, and its position regarding the adjustment, as follows:

Adjustments #29 and #30 refers (sic) to expenditures incurred for the [Year 2000 (Y2K)] project. In Adjustment #29 SCE removes these one-time Y2K support costs from the 1998 through 2000 historical recorded period. However, in Adjustment #30 "Y2K Replenishment of Nuclear Support Costs & Deferrals" SCE restores these same amounts and adds additional amounts borrowed from other departments to deal with Y2K. It states that "to create funding for the Y2K project, SCE required business units to reduce their O&M expenditures. This one-time event resulted in deferral of work to be performed in the future." [Footnote omitted.] SCE provided no specific information on "deferred work" and furthermore stated that "all of the deferred activities are now complete." [Footnote omitted.] Accordingly, ORA disagrees with the inclusion of historical adjustment #30. However, ORA concurs with Adjustment #29. (Exhibit 113, p. 7-B-11.)

ORA's recommendation to exclude Adjustment #30 results in a reduction of forecast costs of $2.021 million.

Discussion

SCE does not take issue with ORA's assertion that SCE failed to provide specific information on the work deferred as a result of the Y2K project. Moreover, SCE does not deny that such project-specific information is important to the analysis of the adjustment. SCE did though, present a list of such projects in its rebuttal testimony.9

SCE's rebuttal testimony shows that SCE deferred 16 specific projects plus miscellaneous work totaling $8.8 million during the period 1998-2000. SCE indicates that if it were not for the diversion of personnel to the Y2K project, it would have performed work on the 16 projects during the period, and the recorded costs would have reflected the costs of doing so. It was because of the emergent need of the Y2K project that maintenance required for plant reliability was not carried out and was thus deferred to the year 2000. SCE states that all of the deferred work is now complete. SCE likens the need to resume work on these projects to the need to resume a regular oil change schedule on a vehicle on which a particular scheduled oil change was deferred. They believe that the deferral of this work would not have a long-term impact on the reliability of the plant, but resuming the work is necessary to avoid irreparable damage to the facility.

SCE's direct evidentiary showing, combined with its rebuttal testimony indicates that the deferred maintenance was indeed related to the reliability of the plant and that the maintenance did occur. SCE's witness testified that Adjustment #30 is needed because recorded O&M for the historical period was lower than is necessary to maintain reliability of SONGS 2&3 and this adjustment would bring the level of expenditures back to a consistent level. Because we are concerned that inadequate funding for SONGS O&M could have a reliability impact on the plant, we will grant SCE's request along with Adjustment #30 so that the Test Year estimate is sufficient to maintain the reliability of SONGS 2&3. In this instance, we are concerned about the reliability of the plant and believe this is an overarching policy goal, but we do direct SCE to make a complete and robust showing in its direct testimony in its next GRC.

ORA proposes an adjustment to remove the costs of an employee awards and recognition program from the historical recorded period (1996-2000). ORA objects to the inclusion of these costs because, according to ORA, they are for rewards set by SCE management and shareholders, and they are based on criteria devoid of ratepayer input. In response to a data request by SCE, ORA stated that it views the program as no more than social/cultural activities having no bearing on standard O&M functions. ORA construes the inclusion of these expenses as an involuntary levy on ratepayers of the sort prohibited by Commission policy set forth in D.67369.10

SCE opposes ORA's proposed adjustment, which would result in a reduction to SCE's forecast of $476,000.

Discussion

The SONGS 2 & 3 awards and recognition program provides employees with incentives to perform above and beyond already high performance standards. Such a program is consistent with current human performance theories and is utilized at many corporations. ORA has not shown why ratepayer input is a necessary condition for ratepayer funding for the program. Even though ratepayer dollars may be involved, SCE management is entitled to a reasonable degree of discretion in determining how to motivate employee performance. Moreover, the costs at issue are not so large as to warrant a cost-benefit analysis to determine the program's effectiveness.

It is well established that dues, donations, and contributions are not eligible for ratepayer funding. However, the record evidence establishes that the awards and recognition program includes no such costs. Moreover, there is no basis for concluding that the SONGS 2 & 3 employee awards and recognition costs are in any way equivalent to chamber of commerce dues or charitable contributions. To the extent that ORA's proposed disallowance relies upon the Commission's policy set forth in D.67369, it is without merit. SONGS 2 & 3 employees, and reasonable incentives to motivate their superior performance, are not just a "worthy cause" to which ratepayers might want to contribute separate and apart from utility rates. ORA's proposed adjustment will not be approved.

When SONGS 2 & 3 ICIP ratemaking was implemented in April 1996, SCE no longer needed the SONGS Nuclear Rate Regulation (NRR) group that prepared GRC information and attrition year filings, and provided other support for Commission proceedings. SCE therefore eliminated the NRR group, and, with the exception of 1996, the recorded data from 1996 to 2000 do not reflect the costs of NRR. With the return of SONGS 2 & 3 to cost-of-service ratemaking in January 2004, SCE will once again fund NRR activities.

SCE's Adjustment #38 incorporates into the 2003 O&M forecast SCE's $776,000 estimate of funding needed for preparing testimony, responding to data requests, meeting with ORA and intervenors, preparing for and attending hearings, and interfacing with other SCE departments to prepare justification for O&M expenses and capital additions.

ORA does not take issue with the need to recognize the resumption of work on rate regulation, but it disputes the amount of Adjustment #38. ORA used a five-year average of NRR costs from 1992-1996, the most recent pre-ICIP five-year period of SONGS 2 & 3 cost-of-service ratemaking, to determine an adjustment amount of $552,000. This is a difference of $224,000 from SCE's forecast.

Discussion

Although an averaging-based forecast might be more reliable than SCE's estimate, in this case it is based on incomplete cost data. During the 1992-1996 period, SCE utilized some SONGS 2 & 3 personnel to perform work for NRR who were temporarily "matrixed" to NRR. SCE did not charge the costs of the matrixed personnel to NRR accounts and consequently did not identify them as NRR costs. Because the 1992-1996 NRR expenditures do not include these matrixed personnel costs, the averaging methodology does not capture all of the NRR costs from the earlier period, rendering it less reliable. We will therefore approve SCE's proposed Adjustment #38.

SCE funds certain SONGS 2 & 3 activities through the Site Projects Functional Group. The Site Projects budget includes major O&M activities that are not required as a part of routine plant operations, occur on a cyclical basis, or require special focus. The majority of the O&M expenses for this group are for projects developed in response to employee Action Requests, and the projects typically result in design modifications to the plant. The scope and quantity of these projects can vary from year to year.

SCE chose the last recorded year method (2000) to forecast test year 2003 costs for the Site Projects Functional Group. SCE also determined that it will need additional funds to implement the Site Projects that are forecast for 2003. SCE's Adjustment #39 identifies the net increase between the last recorded year and the company's estimate for Site Project work activities in 2003. SCE estimates that $12.460 million will be needed in 2003 for Site Projects (consisting of $6.390 million in FERC Account 517 and $6.070 million in FERC Account 532). This represents an increase of 23% from 2000 expenditures.

ORA takes issue with SCE's selection of the last recorded year to forecast the test year expense for site projects as well as SCE's Adjustment #39. ORA instead used a five-year average (1996-2000) of historical expenditures based on its determination that averaging is more appropriate for capturing annual variations. This yields an estimate of $7.517 million ($4.843 million in FERC Account 517 and $2.674 million in FERC Account 532), which is $4.943 million less than SCE's request.11 ORA objects to Adjustment #39 because, among other things, Site Projects identified by SCE have not been approved by the Site Integrated Project Committee (SIPC) even though the SIPC is responsible for approving and prioritizing the projects. ORA also notes that there has been confusion over which projects should be covered by base funds and which ones by Site Projects funds, there has been possible underspending in the past three years, and underruns due to uncommitted funds are allocated to a contingency fund. ORA further notes that SCE did not provide detailed project outlines with cost breakdowns for ORA review. ORA concludes that SCE has not substantiated the need for Adjustment #39.

Discussion

At issue is whether SCE's budget-based forecast of Site Projects expenditures or ORA's averaging approach produces a more reliable estimate for the test year. We first note that, as SCE's own testimony and the following table show, Site Projects O&M expenses can vary from year to year:

Site Projects O&M Expenditures

(2000 Dollars X 1000, 100% Level)

1996

1997

1998

1999

2000

2001

2001

2002

2002

5,234

4,773

13,248

8,779

10,103

11,100

10,000

13,050

15,973

Recorded/Adjusted

Budget

Rec.

Budget

Rec.

Sources: Exhibit 7, Tables VIII-3 and VIII-4; and Exhibit 283, Tables II-2 and II-3.

Given the year-to-year variability displayed in this table, an averaging approach should produce a more reliable forecast in the absence of information explaining why any single recorded year or a departmental budget produces a better estimate of future spending. SCE's rebuttal testimony makes several points in an effort to provide such information, but, as discussed below, it fails to make a convincing case for rejecting averaging.

First, SCE contends that ORA's averaging-based forecast does not recognize recent recorded expenditures. Apart from the fact that SCE's direct testimony also did not recognize recent expenditures, since SCE based its O&M forecast on 1996-2000 data plus budget-based adjustments, this contention does not explain why recent expenditures are more predictive than those of prior years. In fact, reliance on recent years may be particularly problematic given ORA's finding of possible past underspending.

Second, SCE states that it needs flexibility in the project implementation schedule; that work scope is contingent upon plant conditions, needs, and evolving regulations; and that adherence to budgets never interferes with regulatory compliance and safety requirements. If anything, this testimony supports the proposition that departmental spending plans may not always be the best predictors of actual future spending.

Third, SCE states that the five-year average of recorded spending from 1998 through 2002 is $12 million, which it asserts is consistent with its forecast of $12.460 million.12 However, this does not explain why the 1998-2002 period is more predictive than the 1996-2000 period.

Fourth, to counter ORA's testimony regarding the lack of approval for all projects underlying Adjustment #39, SCE states that each identified project now has approval for engineering analysis and most projects have final approval. However, SCE is unwilling to claim that all of the projects have been approved, and it does not state the dollar value of those projects or the minority of projects that have not been approved.

Finally, noting that it spent approximately $3 million more than it planned to spend in 2002, largely due to a $2.4 million gantry crane project that "emerged" after the budget was approved in January 2002, SCE contends that similar emergent projects can be expected in future years. This last point is at best incomplete, as SCE fails to explain why projects with expenses of such magnitude did not emerge during the historical period.

We will reject SCE's budget-based forecast because the company has not shown that it is more predictive than an averaging forecast. However, we find that ORA's five-year average gives insufficient weight to what may be an increasing trend in this functional group. We will instead adopt a forecast based on a three-year average (1998-2000) as shown in the following table:

Adopted Site Projects O&M Forecast

(2000 Dollars X 1000, 100% Level)

 

FERC Account 517

FERC Account 532

Total Site Projects

Labor

280

12

292

Non-Labor

5,666

3,235

8,901

Totals

5,946

3,247

9,193

Sources: Exhibit 7, pp. 80, 82; Exhibit 136, pp. 38-40. The non-labor forecast for Account 517 reflects the $4.378 million Y2K adjustment for 1998 calculated by ORA.

SCE's Master Insurance Program (MIP) provided workers' compensation to contractors prior to 1999. Since October 1999, SCE has required all contractors to carry their own insurance coverage. SCE states that it must cover all future claims for contractors covered under the original MIP. To provide funding to cover claims for contractors' work prior to October 1999, SCE's forecast for the Nuclear Support/FERC Account 528 pairing includes SONGS-related MIP costs of $1.710 million for the test year. The three-year averaging method used by SCE for this function includes MIP claim funding of $920 thousand. SCE's Adjustment #40 adds $790 thousand to the test year forecast to yield the total MIP cost of $1.710 million.

TURN objects to the inclusion of the $1.710 million for MIP claim costs in the O&M forecast. TURN takes the position that the cost of pre-1999 claims should not be borne by current ratepayers beginning in 2004. TURN believes that claims made prior to the beginning of the ICIP period should be extinguished as generation-related regulatory assets that cannot be collected after the end of 2001, while the cost of claims begun during the ICIP period should be assumed to have been covered in full by ICIP prices. If the MIP costs are allowed, TURN believes that it is questionable whether a constant amount should be allowed each year, and that in any event the amount allowed should not be subject to escalation.

Discussion

SCE has shown that these workers' compensation costs are appropriately charged to ratepayers. The ICIP proposal adopted by D.96-04-059 included a provision (Section 4.8.5) by which SCE could request recovery of certain expenses separately from the ICIP price, including worker and third party claims. The evidence does not support a finding that ICIP prices included all claims begun during the ICIP period. With respect to earlier claims, we find no basis for characterizing them as stranded costs. TURN's proposal to disallow the costs therefore will not be approved. With respect to TURN's secondary proposals, the amount requested is based on historical costs and is therefore reasonable, and no cause has been shown to exclude the costs from escalation.

According to SCE, the nuclear industry is facing a labor shortage for certain categories of "nuclear trained" personnel.13 SCE has determined that while the demand for nuclear trained and qualified personnel is likely to remain fairly stable, the supply of such personnel has decreased and will continue to do so. SCE bases this determination on recent nuclear industry reports and its own analysis of SONGS 2 & 3 demographic and regional demographic data.

Industry-wide, the nuclear workforce is aging and there has been a decline in the number of college students interested in nuclear careers. Certain craft positions are also in short supply due in part to fewer departures from the U.S. Navy. According to SCE, a nuclear industry trend to shorter outages means that craft workers needed for outages cannot obtain enough work to make a living, and the resulting workforce attrition has led to a situation where remaining workers are able to successfully request higher wages.

SCE finds that the industry-wide challenge of ensuring an adequate number of workers to operate nuclear power plants is particularly acute at SONGS 2 & 3 for several reasons. These include the high cost of living in Southern California, limited career opportunities due to the fact that SONGS is the only nuclear station operated by SCE, SCE's financial challenges, California's chaotic energy market, and California's law against future nuclear development. The average age of the SONGS nuclear workforce is 48, and SCE expects this will result in substantial retirements of qualified workers in the coming years. By 2004, 14% of the SONGS workforce will be of retirement age.

To address the labor scarcity challenge, SCE has developed a strategy that includes a job skills program aimed at high school students interested in technician jobs, a career program for high school students interested in professional jobs, a summer hire and intern program, signing bonuses based on 10% of base pay, referral bonuses, a relocation allowance with reimbursements of $34,000 for homeowner, supplemental relocation allowances, and training. SCE expects to incur added costs of nearly $4 million in 2002 and again in 2003 to recruit and train key talent during this anticipated labor shortage.

According to SCE, these added costs were not included in the historical period (1996-2000) costs because the impending labor shortage had not yet occurred during those years. As shown in the Joint Comparison Exhibit, SCE's Adjustment #41 adds $3.937 million to the test year 2003 O&M forecast to account for these anticipated costs.

ORA finds that SCE did not adequately support Adjustment #41. ORA proposes rejection of the adjustment on several grounds, including the following:

· A Nuclear Energy Institute (NEI) report prepared by Navigant Consulting projects a shortage of nuclear engineers as narrowly defined, but ORA found that SCE's labor scarcity testimony combined all engineers who work in the nuclear industry as nuclear engineers.

· According to ORA, a 1999 study of SONGS staffing commissioned by SCE and conducted by Martin & Associates found that SONGS is staffed 53% to 117% above the industry median.

· An October 2000 NRC study of labor market trends for nuclear engineers noted that the scarcity of nuclear engineering graduates was based on the assumption of a strong U.S. economy.

· Documentation that SCE submitted support of the adjustment does not justify a scarcity of critical positions, according to ORA. For example, SCE did not show past experience indicating a need for hiring incentives in relevant positions, and SCE admitted to ORA that there had not been any failed recruitment efforts. Also, SCE did not provide specific data on new hires based on attrition and vacancies, and it did not provide job description and classification data. ORA found that SCE used broad definitions for critical positions for which the scarcity adjustment is sought.

· SCE did not substantiate to ORA the actual training requirements for anticipated new hires.

· SCE did not explain to ORA's satisfaction why historical expenditures for training and certification are inadequate to meet the training and certification requirements for anticipated new hires.

· ORA noted that SCE's plans for the 2002 summer hire program had not been developed and implemented, and historical data and analysis on the summer hire program was not available.

· ORA noted that SCE did not address the current state of the economy and how it might impact the adjustment.

Discussion

While SCE has demonstrated that the labor market for certain nuclear job categories is expected to remain tight during the test period, it has not translated the existence of this challenge into a demonstrated need for an adjustment of nearly $4 million. ORA's analysis, summarized above, casts considerable doubt on the reasonableness of SCE's justification for the adjustment. Particularly troubling is a lack of substantiation by SCE for the amounts requested for the summer and training programs that represent a substantial portion of proposed Adjustment #41.

SCE had difficulty filling technician positions in 2001 and 2002, and since mid-2002 it has only been able to hire nuclear trained engineers with signing bonuses and relocation allowances. This corroborates the need for the company to incur scarcity related expenses at some level. SCE stated in its rebuttal testimony that it recorded approximately $1 million in additional costs in 2002 to attract, hire, and train qualified replacements for nuclear trained personnel. This demonstrates that SCE's direct evidentiary showing, in which it projected expenses of $3.85 million for 2002, has not been adequately substantiated. We find that it justifies a labor scarcity adjustment of $1 million in lieu of the requested adjustment of $3.937 million that is set forth in Exhibit 403. We will apply the $1 million adjustment to the various FERC accounts in the same proportions as set forth in Exhibit 403, pp. 97-108.

SCE made a future year adjustment of $5.650 million (Adjustment #45) to reflect increased security costs at SONGS 2 & 3. SCE upgraded its security programs as a result of the terrorist attacks of September 11, 2001 and subsequent NRC directives. Among other things, the NRC required nuclear power plants to increase patrols, augment security forces, and heighten coordination with law enforcement and military authorities. In response, SCE established new security posts that must be covered 24 hours per day, seven days a week. Adjustment #45 captures the increased security costs by increasing the 2001 estimate to reflect an increase in overtime costs for the initial posts that were added, and increasing the 2002-2003 estimates to reflect the increased number of security guards that SCE plans to hire.14

ORA recommends rejection of Adjustment #45 on several grounds. ORA believes that it is premature to accept SCE's increased number of security guards because cost-sharing arrangements at the state and federal levels have not been determined. Also, ORA believes that national security matters fall largely under the jurisdiction of the federal government, and that California ratepayers should not be exclusively obligated to bear the burden of this type of expense. Further, since SCE has not previously raised the issue of any inability of SONGS security personnel to deter vehicular intrusion, ORA assumes that historical/recorded costs are adequate to allow SCE to meet this type of threat.

Discussion

The evidence shows that as a result of the September 11 terrorist attacks and subsequent NRC directives, added security costs have been created for SONGS 2 & 3 in 2003 that did not exist during the 1996-2000 historical period. It also shows that the SONGS 2 & 3 co-owners are at present solely responsible for these increased costs. SCE cannot rely on the presence of U.S. Marines near SONGS 2 & 3 to avoid increased security costs. Moreover, ORA has not shown that it is reasonable to wait until Congress and the NRC fully evaluate security requirements, and implement new statutes and regulations that might reduce SCE's security cost responsibility, before providing ratepayer funding for increased security at SONGS 2 & 3. Adjustment #45 is a reasonable means of recognizing added security costs that SCE must incur, and it will therefore be adopted.

SCE records expenses for training, certification, qualification, safety classes, and employee recognition awards for Maintenance Division personnel in the Maintenance/FERC Account 524 pairing. SCE used the last recorded year to forecast labor costs and a five-year average to forecast non-labor costs. SCE forecasts a total of $3.494 million in O&M expenses for this pairing, which represents an increase of 12% above year 2000 expenditures.

SCE used the last recorded year for labor costs because it includes the "final effects" of certain "programmatic changes" for maintenance personnel training and certification requirements. SCE implemented these programmatic changes during the 1998-2000 period in accordance with Institute of Nuclear Power Operations (INPO) Academy Documents that raised industry standards for craft and direct supervisory qualifications.

Since the programmatic changes were reflected in 1998-2000 period, ORA believes that it is appropriate to use a three-year average (1998-2000) for the labor forecast for the Maintenance/FERC Account 524 pairing. This results in a forecast of $2.604 million, or $890,000 less than SCE's forecast.15 ORA rejects SCE's justification for using the last recorded year as unsubstantiated.

Discussion

SCE takes the position that the year-2000 recorded expenses include the final effects of programmatic changes for maintenance personnel training and certification requirements. SCE argues that the expenses for 1998 and 1999 were low and that including them in a three-year average unreasonably distorts the forecast because the three-year average does not include the final effects of programmatic changes. We find that SCE's "final effects" argument adequately justifies the use of the last recorded year. ORA's proposal to use a three-year average for labor costs is denied.

For the Maintenance/FERC Accounts 530, 531, and 532 functional group pairings, SCE used five-year averages to develop its forecasts. SCE notes that the labor costs recorded in Accounts 530 and 531 showed gradual declines during the 1996-2000 period, while the amount recorded in Account 532 showed an increase. This is due to a change in how maintenance support activities are reported. Costs formerly recorded against the direct maintenance activity account are now recorded in Account 532. SCE notes that in the aggregate, the labor cost in the three accounts was relatively constant.

ORA used the last recorded year (2000) for Accounts 530 and 531, for which there was a declining trend in labor costs. ORA used a five-year average for Account 532, for which there was an increasing trend. This method, along with ORA's recommendations with respect to adjustments, results in lower ORA forecasts for Account 530 ($6.954 million, or $818,000 less than SCE's forecast of $7.772 million) and Account 531 ($8.179 million, or $1.248 million less than SCE's forecast of $9.427 million).16 ORA acknowledges SCE's explanation that the aggregate labor costs in all three accounts were relatively stable from 1996 to 2000. However, ORA does not accept this explanation as justification for SCE's decision not to use the last recorded year for Accounts 530 and 532. ORA believes that the year 2000 recorded costs incorporate the shift from direct maintenance accounts to Account 532.

Discussion

SCE has shown that Maintenance/FERC Accounts 530, 531, and 532 are linked because of how activities were reported, and that a consistent methodology should be used for the labor component of all three accounts. Thus, at a minimum, ORA should have used the last recorded year for Account 532 since it did so for Accounts 530 and 531. SCE has also shown that five-year averaging is appropriate because of the cyclical nature of the underlying activities and related expenditures. SCE's five-year averaging forecast method for these functional group/FERC account pairings will therefore be adopted.

SCE forecasts expenses of $19.219 million for this functional group/FERC account pairing. Between 1996 and 1997 there was a significant labor cost decrease that was almost fully offset by an increase in non-labor costs. SCE explains that this was mainly due to a 1996 reversal of estimated severance costs as a credit to non-labor and recording of the actual cost to labor. To exclude the effects of this transaction, SCE selected a four-year average for both labor and non-labor costs for the Nuclear Support/FERC Account 524 pairing. SCE notes that the work scope and cost for this pairing remained relatively constant over the period 1997-2000.

ORA follows the basic framework of SCE's analysis but believes that both 1996 and 1997 recorded data should be excluded for the same reason that SCE excluded 1996. ORA therefore used a three-year average (1998-2000) for the labor and non-labor estimate for Nuclear Support FERC Account 524. ORA's forecast is $18.35 million, or $869,000 less than SCE's forecast.17

Discussion

ORA states that the impact of severance costs for both 1996 and 1997 were omitted because "these two years are not reflective of costs associated with this account." (Exhibit 113, p. 7-B-56.) However, the evidence (Exhibits 8 and 283) shows that the severance transaction affected 1995 and 1996 only. No good cause has been shown for excluding 1997. SCE's four-year averaging method will therefore be adopted.

SCE forecasts expenses of $2.295 million for this functional group/FERC account pairing, in which SCE records costs for miscellaneous corporate adjustments that relate to SONGS functions. The corporate adjustments may include settlement costs due to bankruptcies, fines, and other litigation. SCE's forecast, based on a three-year average, excludes 1996 and 1997 because those years reflected one-time offsetting adjustments. SCE expects the costs for the most recent three years to continue in the future. SCE states the forecast includes the MIP workers' compensation adjustment, discussed above, as well as $555,000 for the return of rebuilt plant equipment to inventory and $277,000 for miscellaneous other costs. These include normalized paid time-off expenditures associated with SONGS employees.

ORA does not believe that SCE provided sufficient information to evaluate this functional group/FERC account pairing, and recommends removal from the O&M forecast all amounts except $1.71 million in MIP costs. This results in a proposed disallowance of $585,000.

Discussion

SCE's forecast for this pairing includes MIP workers' compensation costs, the return of inventory, and normalized paid time-off expenditures. While the amounts and purposes of expenditures in this pairing have varied considerably over the five-year recorded period, we find no basis for assuming that no expenses excepting the MIP costs will occur in the test year. SCE's use of a three-year average of recorded/adjusted data is reasonable and will be adopted.

For Nuclear Support/FERC Account 532, in which SCE records costs incurred by the Material Support section, SCE used the last recorded year for labor and a three-year average (1998-2000) for non-labor. SCE forecasts O&M costs of $4.559 million.

This pairing includes debits and credits for obsolete and salvaged inventory and credits from the sale of this material. It also includes costs for all services performed for SONGS by SCE's Procurement And Material Management division (PAMM). PAMM activities include procurement, expediting, contract administration, and support of the Material Management System. SCE states that PAMM costs charged to this account increased in 1998 due to corporate accounting changes, and it therefore used a three-year average to exclude pre-1998 costs.

Finding SCE's explanation of the 1998 accounting change insufficient, ORA proposes a five-year average to forecast the non-labor forecast for this pairing. ORA accepts SCE's forecast for labor. ORA's forecast is $4.052 million, or $507,000 less than SCE's forecast.18

Discussion

SCE has shown that the 1998 accounting change justifies its use of a three-year average. The three-year average provides a better forecast because in 1998, SCE implemented an internal market mechanism (IMM) process throughout the company.19 Beginning in 1998, procurement costs charged to SONGS 2&3 were increased to include support personnel located at the corporate office.20 This additional allocation increased Nuclear Support/FERC Account 532 costs beginning in 1998.21 The five-year average includes 1996 and 1997 which don't reflect these increased cost allocations and therefore do not provide full funding of known allocated costs. We therefore adopt SCE's proposal to use a three-year average.

In its rebuttal testimony SCE identified four functional group/FERC account pairings for which methodological differences between SCE and ORA resulted in forecast difference of $250,000 or less. These pairings, and the amounts by which SCE's forecast exceeds ORA's forecast, are Operations/FERC Account 524 - $52,000, Maintenance FERC Account 528 - negative $214,000 (i.e., ORA exceeds SCE), Radchemical Control/FERC Account 523 - $64,000, and Nuclear Support FERC Account 529 - $184,000. The net difference for these four pairings is $86,000.

Discussion

For both Operations/FERC Account 524 and Nuclear Support/FERC Account 529, SCE used five-year averages. ORA used last recorded year and four-year averages, respectively. We find that five-year averages are likely to be more reliable for these pairings. For Maintenance/FERC Account 528, SCE used the last recorded year. SCE has justified such use by showing that (1) for labor, it gained efficiencies through programmatic enhancements and (2) for non-labor, only the year 2000 fully reflects cost reductions associated with inventory reductions. For the Radchemical Control/FERC Account 523 pairing, SCE used the last recorded year for both labor and non-labor. SCE has shown that this is necessary to reflect cost shifts from Account 520. For the foregoing reasons, we will adopt SCE's forecasts for these four pairings.

Consistent with its proposal for recovery of expenses for refueling and maintenance outages (see Section 3.1.3 below), SCE intended to remove all outage-related costs from its base recorded costs for 1996-2000. This included costs related to Cycles 9, 10, and 11 outages. However, upon review of several FERC accounts, ORA found that not all such costs had been removed by SCE. ORA recommends removal of these additional outage-related amounts from the base O&M forecast. As set forth in Exhibit 283, Appendix B, as well as Exhibit 403, pp. 125-142, ORA's proposal to remove the outage-related amounts as recommended by ORA affects 17 functional group/FERC account pairings for a total proposed disallowance of $2.557 million.

In its rebuttal testimony, SCE states that the amounts identified by ORA may be removed from the base O&M forecast. However, SCE asserts that the identical amount of costs should be added to the refueling and maintenance outage cost recovery mechanism.

Discussion

ORA has identified refueling and maintenance outage-related expenses that SCE included in its base O&M forecast for SONGS 2 & 3, even though SCE proposes recovery of outage-related expenses separately and apart from the base revenue requirement. SCE's rebuttal testimony does not deny that the subject costs are outage related, and SCE's reply brief acknowledges that the company's "initial testimony incorrectly identified them in the wrong category." (SCE Reply Brief, p. 33.)

In order to remove all refueling and maintenance outage O&M costs from the base O&M forecast as intended by SCE, we will adopt ORA's proposal to remove a total of $2.557 million in outage costs from the affected functional group/FERC Account pairings underlying the base O&M forecast. SCE's proposal to add this amount to its outage-related cost recovery mechanism is addressed in Section 3.1.3.

The NRC assesses license holders two types of fees under the Omnibus Budget Reconciliation Act of 1990. License and inspection fees recover the NRC's costs of providing special benefits to identifiable applicants and licensees. Annual fees recover generic and other regulatory costs not otherwise recovered through license and inspection fees. SCE's Adjustment #36 reflected net fee increases in 2001 and 2002 and a decrease in 2003. This adjustment affects the Nuclear Support/FERC Account 517 pairing.

As set forth in the Federal Register of April 3, 2003, the NRC is amending its regulations to increase annual fees for fiscal year 2003. The updated annual fee for each power reactor is $3.278 million for a total of $6.556 million in 2003 dollars, or $6.138 million in 2000 dollars. This represents an increase of $852,600 (2000 dollars, 100% share) over the $5.286 million amount that SCE had previously forecast in this proceeding. In its May 2003 update testimony, SCE presented an updated base O&M forecast reflecting the revised NRC annual fees.

Discussion

No party took issue with SCE's Adjustment #36 or the May 2003 updated forecast of NRC fees. Since the expenses are attributable to readily identifiable government action and are readily quantifiable, they will be approved.

Incorporating the foregoing resolution of various O&M issues, the following table sets forth the adopted test year 2003 SONGS 2 & 3 base O&M forecast by FERC Account and functional group.

Test Year 2003 Base O&M Adjusted Costs

(2000 Dollars, SCE Share)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Operations

Maintenance

Engineering

Site Projects

RadChemical Controls

Regulatory Affairs

Security

Training

Nuclear Support 3/

Corporate Support

Participants 2/

IT-Post IS Transformation 1/

Total

517

879

0

19,398

5,946

347

2,341

1,126

0

16,937

0

(11,721)

3,723

38,976

519

655

0

0

0

0

0

0

0

0

0

(163)

0

492

520

4,207

0

128

0

10,476

0

0

0

1,843

0

(4,155)

0

12,499

523

5,882

0

0

0

193

0

0

0

0

0

(1,516)

0

4,559

524

5,538

3,482

5,472

0

2,174

4,672

12,666

6,474

19,001

(12,266)

(11,780)

0

35,433

525

0

0

0

0

0

0

0

0

1,110

0

(277)

0

833

528

0

8,699

3,723

0

0

0

0

0

2,295

0

(3,672)

0

11,045

529

0

5,870

0

0

0

0

0

0

3,843

0

(2,424)

0

7,289

530

0

7,479

6,402

0

0

0

0

0

0

0

(3,464)

0

10,417

531

0

9,412

3,390

0

0

0

0

0

0

0

(3,194)

0

9,608

532

0

11,311

1,621

3,247

0

0

0

0

4,451

(1,402)

(4,797)

0

14,431

Total

17,161

46,253

40,134

9,193

13,190

7,013

13,792

6,474

49,480

(13,668)

(47,164)

3,723

145,581

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1/ Exhibit 47, Section III, page 25, plus $204 for IT O&M adjustment so that the adopted 2003 level for Acct. 517 equals year 2000 recorded [$3,723].

See Section 6.5.2.

 

 

 

 

 

 

 

 

 

 

 

2/ Participant credit calculation at 24.95% for all above activities except IT-Post IS Transformation.

 

 

 

 

3/ Includes increase for Nuclear Regulatory Commission fees from May 9th update testimony of $0.640 million ($0.853 million [100%] less participants

share of $0.213 million).

 

 

 

 

 

 

 

 

 

 

 

2.1.3. SONGS 2 & 3 Outage O&M

While there was some confusion among SCE and ORA witnesses regarding SCE's proposal for recovery of SONGS refueling and maintenance outage O&M costs, the record regarding the nature and intent of SCE's proposed mechanism is clear. SCE requests that a "flexible outage schedule mechanism" similar to that adopted and affirmed in SCE's last three general rate cases be established. SCE's proposed mechanism works in conjunction with its Post-Test Year Ratemaking (PTYR) proposal. A standard per unit per outage cost would be established in this GRC, and SCE would include with its PTYR filings a refueling and maintenance outage O&M forecast based on the number of outages forecast to occur in the next year. SCE requests that the Commission establish in this proceeding a forecast for its refueling and maintenance outage costs of $52.462 million (2000 dollars, 100% level) per outage per unit, plus the $2.557 million that was removed from the base O&M forecast.

ORA does not take issue with the $52.462 million outage cost, but it objects to SCE's proposal to add the $2.557 million in outage costs removed from the base O&M forecast. Aglet disputes the level of SCE's estimated outage cost.

Discussion

Since it is not possible within the context of a GRC to predict with reasonable reliability the number of refueling and maintenance outages that will occur at SONGS in any one of the future years of the GRC cycle, inclusion of outage-related expenses in the base O&M forecast could lead to an unreliable forecast. SCE's proposed mechanism uses more reliable annual forecasts of the number of outages in the following year. It will therefore be adopted, provided that this approval does not constitute full approval of SCE's PTYR proposal, which is further addressed in Section 11 of this decision. Sections 3.1.3.2 and 3.1.3.3 below address disputes affecting the amount of the outage cost to be adopted. Because, as discussed therein, we do not adopt any revisions to SCE's original forecast, we adopt a per unit per outage cost estimate of $52.462 million (2000 dollars, 100% level). The SCE share is $39.373 million.

Since we have removed $2.557 million in outage related O&M expenses (2000 dollars, 100% level) from the SONGS 2 & 3 base O&M forecast for 2003, SCE contends that the identical amount should be added to the refueling and maintenance outage cost estimate. As a result, SCE contends that the cost estimate should be $55.1 million per unit per outage instead of $52.5 million.

Discussion

SCE's evidentiary support for its proposal to reinstate the excluded outage costs on a dollar-for-dollar basis consists of the following quote from its rebuttal testimony:


ORA's proposed disallowance of Refueling Outage (RFO) costs are (sic) based on outage related function numbers appearing in workpapers in historical years 1996-2000. These costs may be removed from the historical recorded period, and thus the Base O&M Test Year estimate, but must be added to the RFO cost estimate. Therefore, the RFO estimate must be increased by the $2,557K (2000$, 100% level) for each outage. (Exhibit 283, p. 21.)

Apart from SCE's insistence that the $2.557 million that it mistakenly included in its base O&M forecast simply must be added to its per unit per outage forecast, there is no evidence to support the proposal. Moreover, there appear to be conceptual problems with it.

First, we do not find that SCE has addressed the unit of measurement problem associated with removing annualized expenses from the base O&M forecast and adding them to the per unit per outage forecast at issue here. In addition, for purposes of estimating its per unit per outage cost, SCE relied on a budget-based approach (in the form of adjustments for numerous specific, major activities) combined with adjusted Cycle 11 costs. (Exhibit 8, p. 100.) SCE has not shown whether, or how, it reconciled the amounts removed from the base O&M forecast with the budget-based per outage cost estimate.

The record before us does not allow us to conclude that any amount should be added to SCE's per unit per outage forecast as a result of the correction to SCE's base O&M forecast error, let alone the precise amount that was removed from the base forecast. SCE's proposal to reinstate that amount is denied.

SCE developed its estimate of per unit per outage costs by including an adjustment of $3 million for SONGS Unit 3 Cycle 11 "permanent Unit 3 Mob/Experience Factor/Timing." Because the outage estimate uses an average of Unit 2 and Unit 3 costs, the effect of this adjustment is to add $1.5 million to the unit/outage cost forecast. Aglet recommends the removal of this adjustment.

Discussion

SCE's witness testified that this adjustment reflects the need for SCE to mobilize its outage forces. In past years, outages at the SONGS units occurred back to back, and outage workforce mobilization was only required once. Due to a fire and forced outage in Unit 3 in 2001, the outages are now nine months apart, and mobilization costs are incurred for each outage. SCE has justified this adjustment, and Aglet's proposed reversal of it is therefore denied.

SONGS Unit 1 was permanently shut down in November 1992 and placed in a SAFSTOR (safe long-term protective storage prior to dismantling) configuration in 1993. Decommissioning and dismantling of SONGS Unit 1 began in June 1999.

D.92-08-036 provided that rate recovery of SONGS 1 shutdown O&M expenses would be addressed in SCE's GRC's until SONGS 1 is decommissioned. Pursuant to D.97-08-056 and D.99-06-007, SONGS 1 O&M expenses are recovered through SCE's Nuclear Decommissioning Charge until used fuel is removed from the SONGS 1 used fuel pool and placed in dry storage. SCE anticipates this will occur prior to 2005. The scope of SONGS 1 shutdown O&M expenses is limited to assuring safe storage of used fuel until it is removed, and preservation of safe physical conditions in the areas of SONGS 1 not under the control of the decommissioning project. When all of the SONGS 1 used fuel is removed from the SONGS 1 used fuel pool, and the used fuel pool is ready for transition to decommissioning, SCE will make an advice letter filing to (1) remove the variable SONGS 1 shutdown O&M expenses from rates and (2) to reallocate the fixed SONGS common costs to SONGS 2 & 3 O&M and capital as well as the SONGS 1 decommissioning project.

SCE estimates direct shutdown O&M expenses of $3.864 million (2000 dollars, 100% share) for 2003. In addition, SCE estimates the allocation of SONGS common costs to SONGS Unit 1 shutdown O&M as $3.310 million (2000 dollars, 100% share). ORA reviewed SCE's account specific estimates for SONGS Unit 1 shutdown O&M expenses. ORA recommends no changes to, and therefore stipulates to, the estimates.

Discussion

SCE's estimate for direct shutdown O&M expenses is approximately one-half the level adopted in the 1995 GRC. This is consistent with SCE's testimony that certain functions required for the long-term storage of the plant are now complete, reducing the need for future use of contract labor. We find SCE's estimates of direct shutdown O&M and of allocated common costs to be reasonable, and therefore adopt them.

2.1.5. SDG&E's Share of SONGS Costs

SDG&E owns a 20% share of the SONGS units. To ensure consistent treatment of SONGS expenditures and to avoid duplicate litigation, the Commission has addressed SONGS-related expenses that SCE bills to SDG&E in SCE's GRCs. In this proceeding, SDG&E requests that the Commission:

· Approve SCE's forecasted SONGS costs as set forth in A.02-05-004.

· Approve $15.806 million as SDG&E's share of SONGS 2 & 3 capital additions for 2004 and authorize SDG&E to reflect this approved amount in calculating the depreciation expense and other costs associated with these capital additions in its Test Year 2004 cost of service proceeding (A.02-12-027/A.02-12-028).

· Approve $67.585 million as SDG&E's share of SONGS 2 & 3 O&M expenses for 2004 (other than refueling outage expenses) and authorize SDG&E to reflect this revenue requirement in rates effective January 1, 2004.

· Approve $12.468 million as SDG&E's share of each SONGS 2 & 3 refueling outage that occurs in 2004 and 2005 and authorize SDG&E to file annual advice letters on November 1, 2003 and November 1, 2004 to specify the number of SONGS refueling outages expected to occur during the following year and the escalated cost per outage.

· Approve $2.635 million as SDG&E's share of SONGS 1 shutdown O&M expenses for 2004 and authorize SDG&E to reflect this revenue requirement in rates effective January 1, 2004.

Discussion

Except for SDG&E's request that we approve SCE's SONGS-related proposals in full, no party takes issue with the framework of SDG&E's proposals, as set forth above and explained in its direct testimony (Exhibit 261) and update testimony (Exhibit 414). We find the framework to be reasonable and consistent with our prior decisions for addressing SDG&E's SONGS costs in SCE's GRCs. For the most part, SDG&E's proposals consist of determining its costs by applying the 20% ownership share to SCE's costs.

However, the specific dollar amounts for which SDG&E seeks approval are calculated with the assumption of our full approval of SCE's requests for SONGS. Since SDG&E's costs for SONGS are predicated upon its 20% ownership share, the amounts requested by SDG&E must be adjusted to reflect the corresponding 100% level of capital and O&M costs for SONGS 2 & 3 as well as the amortization period adopted in this decision. We will approve SDG&E's requests as set forth above, subject to the adjustments required to reflect SONGS-related determinations made in this decision.

2.2. Palo Verde Nuclear Generating Station

SCE owns a 15.8% share of Palo Verde Nuclear Generating Station Units 1, 2, and 3 (Palo Verde), which are located 50 miles west of Phoenix Arizona. Arizona Public Service (APS) is operating agent of Palo Verde, the nation's largest nuclear installation. SCE implements its ownership responsibilities through participation in Administrative, Engineering and Operations, and Audits Committees.

Since 1997, SCE has recovered its share of Palo Verde operating costs through a balancing account ratemaking mechanism adopted D.96-12-083. Pursuant to D.01-09-041, this mechanism will continue until the effective date of a decision on this GRC, at which time the adopted 2003 capital and O&M forecasts for Palo Verde will be implemented.

2.2.1. Palo Verde Capital Expenditures

SCE forecast the capital budget for Palo Verde work on a project-by-project basis. SCE forecasts capital expenditures of $19.4 million in 2003 (nominal dollars, SCE share). ORA used a five-year average of nominal historical expenditures of $8.2 million, plus the $9.4 million cost of Unit 2 installation of steam generators, to arrive at a forecast of $17.6 million.

SCE contends that ORA's use of averaging plus extraordinary items is wrong because that method assumes that capital additions will be basically the same from year to year and that fixed rules exist for determining what type or size of projects are extraordinary. SCE also contends that ORA's averaging methodology was applied incorrectly because ORA used nominal dollars to calculate the five-year average. SCE determined that using common year dollars instead of nominal dollars would change ORA's adjusted 2003 estimate from $17.6 million to $19.07 million.

Discussion

We concur with SCE that the most accurate method for estimating an average of numbers from different years is to escalate them all into common year dollars. We note that when this adjustment is made to ORA's forecast, the resulting estimate is similar to SCE's. While we do not necessarily accept SCE's criticisms of the use of averaging for forecasting capital expenditures generally,22 we find that SCE's project-specific forecast of capital expenditures for Palo Verde is reasonable and should be approved.

2.2.2. Palo Verde O&M Expenses

SCE initially forecast Palo Verde O&M expenses of $42.6 million (2000 dollars, SCE share) based upon three-year averages for all FERC accounts except Account 517, which contains substantial SCE labor costs. SCE rejected averages for longer periods because Palo Verde operations were stabilizing in 1996 and 1997 after a period of elevated costs. SCE used a budget-based approach for Account 517, in which SCE's Palo Verde oversight labor costs are recorded. SCE estimated $42.3 million in base O&M expenses and $300,000 in oversight costs.

Due to declining Palo Verde O&M costs and what it believed to be inadequate justification for an increase above year-2000 costs, ORA recommended use of the last recorded year method for all Palo Verde estimates. The resulting ORA forecast of $41.2 million was $ 1.4 million less than SCE's forecast of $42.6 million.

TURN recommended a reduction to SCE's forecast of $3.267 million based upon updated APS budget data. SCE agrees that the approved budget for APS is an appropriate forecast, and it accepts TURN's recommendation provided that NRC annual fees are properly accounted for by booking fees that are in FERC Account 928 in APS' budget in SCE's Account 517. TURN stipulated to this accounting and revised its recommended reduction to $1.744 million. SCE calculated and supported a revised forecast of $39.517 million based upon the updated APS budget data and the accounting change.

Discussion

As shown in Exhibit 3, p. 15, Palo Verde O&M costs declined consistently throughout the recorded period, from $45.6 million to $41.2 million in 2000. The budget-based forecast of $39.517 million in Exhibit 327, which reflects the most recent budget information available, shows continued declines in Palo Verde O&M expenses that are not reflected in historical costs. We will adopt this budget based forecast as the most reasonable estimate of Palo Verde O&M expenses. SCE has justified its request for $300,000 for Palo Verde oversight activities, and ORA's proposal to reduce the oversight cost to $66,000 is denied. Similarly, we find no basis to reduce the estimate to exclude certain new plant security costs, as recommended by ORA.

2.3. Mohave Generating Station

SCE is the operating agent and owns 56% of Mohave Generating Station Units 1 and 2 (Mohave), located at Laughlin, Nevada. The Mohave Operating Agreement provides for its operation only through 2005. Other contracts, including those for the supply and transportation of coal fuel from the mine to the plant, also only provide for operation through 2005. While the future of the Mohave facility is in question (see A.02-05-046), it will operate through 2005.

SCE had expected to divest its share of Mohave until enactment of Assembly Bill (AB) X1-6 on January 18, 2001. Earlier, on May 10, 2000, it entered into an asset sale with AES with anticipated ownership transfer during the last half of 2000 or the first quarter of 2001.

SCE requests recovery of Mohave capital expenditures of $7.475 million in 2002 and $10.412 million in 2003, or a total for the two years of $17.887 million (all figures in nominal dollars, SCE share). SCE states that this level of investment is necessary to maintain reliable operations through 2005, and that it includes work necessary to ensure employee safety and environmental protection. According to SCE, no investment to extend operations beyond 2005 is included in its estimate of capital spending.

SCE stated that its capital spending at Mohave would yield O&M savings, but ORA found that the capital expenditures for 2002 would result in zero O&M savings and that its 2003 request would result in just $40,000 in O&M savings. Moreover, ORA found even those savings to be unsubstantiated. ORA also questions the reasonableness of SCE's capital spending given the likelihood of a shutdown of Mohave.

In arriving at its recommended capital expenditures, ORA considered only those projects earmarked by SCE as regulatory, safety, and environmental. These totaled $514,000 for 2002 and $1.180 million for 2003. Adding estimated blanket work order expenditures of $383,590 for each of the years 2002 and 2003, ORA arrived at estimated expenditures of $897,590 for 2002 and $1.564 million for 2003. ORA's recommended expenditures represent disallowances of $6.577 million for 2002 and $8.848 million for 2003, or a total of $15.425 million.

TURN took issue with a single capital project underlying SCE's capital forecast for Mohave-replacement of Cooling Tower 1EE. TURN recommends a permanent disallowance of the $1.23 million capital expenditure associated with the project.

Discussion

To explain its position on Mohave capital spending, ORA uses the analogy of a 15-year old car that has had hard service. ORA posits that its owner will think differently about repairs and maintenance than will the owner of a newer car. ORA also posits that, for safety reasons, the older car owner must maintain the brakes and tires as long as she operates the car, even if she plans to sell it within one week. Except for safety considerations, ORA assumes that further repairs are problematic, and that a convincing reason is needed to make them.

Like the oil change analogy that SCE used attempting to explain deferred spending at SONGS (see Section 3.1.2.3), we find ORA's vehicular maintenance analogy less than fully supportive of the associated recommendation. For one thing, we do not understand that SCE can easily sell or junk Mohave if needed repairs are too expensive, as the owner of a clunker vehicle might do. Also, we assume that the car in this analogy is required for transportation, and that reliability is important to the owner despite the vehicle's age. In fact, reliability issues may weigh heavily on the owner's mind precisely because the car is so old. A prudent owner who is in a position of having to operate the car for at least two more years will not necessarily limit expenditures to tires, brakes, and other safety components, but will instead perform all repairs and maintenance necessary to assure reliable automotive performance for those two years. An experienced automobile mechanic might advise the owner to install new spark plugs even though that might not improve safety. A prudent owner would heed that advice even though the mechanic did not determine through rigorous statistical analysis the probability that the old spark plugs would fail within two years. It almost certainly would not make sense for the vehicle's owner to plan on operating the vehicle on seven cylinders during its final years.

ORA's capital spending recommendation for Mohave reflects the concern that SCE will be spending substantial sums at ratepayer expense on projects for a generation facility that could be shut down in the relatively near future. This concern is not unreasonable, but the evidence does not support ORA's conclusion that SCE's planned capital spending should be limited to the bare minimum needed for regulatory requirements, environmental protection and safety. SCE's testimony shows that most of its planned investments, such as those for steam turbine buckets (blades), boiler tubes, electrical cables, and other components whose failure could cause a shutdown, are important for reliable operations at Mohave through 2005.23

ORA proposes to cut SCE's capital spending plans at Mohave by 86%, from $17.887 million to $2.462 million. We are concerned that cutbacks as severe as these may unduly impact production reliability. Whether or not Mohave continues to operate after 2005, determination of which is beyond the scope of this GRC, we intend to authorize the capital funding that is necessary for

continued safe, reliable, and environmentally responsible operation of the plant through 2005.

TURN's proposed disallowance of the Cooling Tower 1EE investment will not be adopted. Even though TURN has shown that SCE's economic analysis showing a benefit to cost ratio of 2.55 may be based on faulty assumptions, and that the direct economic benefits of the investment are at best questionable, SCE has shown that the project was needed for safety as well as reliability purposes.

One point in ORA's car repair analogy is valid-SCE must convince us that the reliability projects and other proposed expenditures are necessary. SCE's direct testimony (Exhibit 17) adequately fulfils this requirement. For the foregoing reasons, and because the level of SCE's planned capital spending is generally consistent with Mohave capital expenditures of the 1996-2000 period, as shown in Exhibit 3, Figure II-5, p. 21, we will approve SCE's proposed spending plan for ratemaking purposes.

Recorded/adjusted Mohave O&M expenses were $37.057, $31.105, $32.502, $26.971, and $25.359 million (2000 dollars, SCE share) in the years 1996 through 2000, respectively. SCE states that while operations expenses were relatively flat during this period, maintenance spending was temporarily but significantly decreased in 1999 and 2000, due largely to the expected sale of Mohave. SCE states that with the cancellation of the sale, the reduced level of maintenance spending in 1999 and 2000 cannot be sustained. For example, while the Mohave sale was pending SCE did not fill various vacancies as they occurred. After reducing its Mohave workforce by 46 employees beginning in 2000, SCE is now in the process of filling 35 vacancies at Mohave. In addition, training programs that were suspended are being reestablished.

Another O&M cost reduction at Mohave was realized in 2000 due to an abbreviated planned outage program. This was made possible when SCE determined on the basis of equipment assessments that a turbine overhaul scheduled for that year could be deferred. In addition, SCE notes, California needed the energy in 2000, so Mohave was kept operating as much as feasible. Finally, SCE states that it had no incentive to spend large sums on the 2000 outage program to benefit the new owner. According to SCE, the conditions that led to reduced planned outage expenses in 2000 will not recur in the test period.

Having determined that maintenance expenses were abnormally low in 1999 and 2000, SCE based its maintenance O&M forecast on three-year averages (1996-1998) for most FERC accounts. For operations, SCE used three-year averages for the years 1998-2000 for most FERC accounts after determining that use of those years properly captures a change in how employees charge their time to various accounts. SCE points out that its resulting O&M expense forecast of $30.633 million is consistent with the adjusted average O&M expenses for years prior to 1999.

ORA proposes a Mohave O&M forecast of $24.661 million, or $5.972 million less than SCE's forecast.24 ORA based its forecast on expenses for the years 1999 and 2000 for all but one FERC account. ORA did so for two reasons. First, ORA believed that using only the earlier years, as SCE did, would result in omission of savings achieved through SCE's "Condition-Based Maintenance" policy. Second, ORA believes that there are serious issues and uncertainties about Mohave's status after 2005. ORA determined that costs for 1999 and 2000 best represent the required O&M expenditures since that period may be reflective of a possible divestiture scenario.

For FERC Account 501.13, ORA used the last recorded year to forecast labor expenses to reflect reductions in fly ash management costs, and it supported additional reductions in that account to reflect reduced centrifuge overhaul costs. ORA also recommended adjustments totaling $345,000 for one-time occurrences that assertedly are not reflective of 2003 expenditures.

Discussion

SCE has presented evidence justifying the exclusion of 1999 and 2000 recorded expenses from its Mohave maintenance forecast. Due largely to the then-pending sale of Mohave, SCE spent less on maintenance than it would have in those years if it had not been planning to divest the plant. Such reduced expenses are not likely to recur in the test period. We note that SCE's total O&M forecast of $30.633 million for Mohave is less than the recorded/adjusted expenses for any of the three years from 1996 to 1998.

At the same time, we find that ORA's reasons for using only 1999 and 2000 lack merit. The Condition-Based Maintenance policy has been in place since 1992, and the associated savings are reflected in each of the years from 1996 to 1998. Also, ORA has not established that the circumstances underlying the period during which the Mohave sale was pending are more representative of the GRC test period circumstances than those of the earlier three years. The planned sale involved the immediate transfer of ownership to a known third party, whereas the test period circumstances involve future uncertainty regarding the plant's operation after 2005 combined with an obligation to continue reliable operations until then. We find no basis for concluding that the low-cost performance of 1999 and 2000 will be replicated in the test period.

Finally, SCE has shown that ORA's proposed adjustments for fly ash management, centrifuge overhauls, litigation settlement, and other costs are not warranted. We find that SCE's O&M forecast for Mohave is more reliable than ORA's forecast. SCE's forecast is therefore adopted.

SCE owns 48% of Four Corners Generating Station Units 4 and 5 (Four Corners), located near Farmington, New Mexico. APS is the operating agent for Four Corners.

According to SCE, the capital spending pattern at Four Corners is heavily influenced by plant overhaul schedules. Historical expenditures vary considerably from year to year. Between 1996 and 2000, SCE's share of Four Corners capital expenditures ranged from a low of $446,000 in 1998 to a high of $6.215 million in 1996, and the five-year average expenditure was approximately $3.1 million. (Exhibit 3, p. 24.) SCE requests recovery of Four Corners capital expenditures of $9.173 million in 2002 and $4.462 million in 2003, or a total for the two years of $13.635 million (nominal dollars, SCE share).

Finding that the Four Corners capital expenditures yield negligible O&M savings, and confronting a lack of information regarding priorities and costs for each capital project, ORA believes that capital expenditures should be limited to a five-year average of historical costs plus projects that SCE identified as having safety, environmental, or regulatory objectives. Based upon this position, ORA recommends that SCE's capital expenditures be limited to $2.858 million in 2002 and $2.876 million in 2003, or a total of $5.734 million. The difference between SCE and ORA for the two years is $7.901 million.

Discussion

Over 82% of the capital expenditures planned for Four Corners are focused on production reliability. Planned expenditures include replacement of turbine blades, boiler tubes, steam condensers, tubular air preheaters, and other components that would cause a shutdown if they failed. In its direct testimony, SCE identified and described 48 separate capital projects with a total cost of $18.858 million (SCE share). Two projects alone, replacement (retubing) of the Unit 5 main steam condenser and replacement of the tubular air preheater, required more than $3.8 million to complete in 2002. In other words, the costs for just these two projects exceed ORA's entire capital forecast for 2002.

We conclude that a forecast based upon a five-year average is inappropriate for Four Corners. Because the greatest part of Four Corners' capital expenditures are for reliability, we are concerned that adoption of ORA's recommended level of expenditures could, if implemented by the operating agent, significantly impact production reliability at Four Corners. We will adopt SCE's project-specific test year forecast for Four Corners capital expenditures.

Recorded/adjusted Four Corners O&M expenses were $25.727, $20.921, $19.691, $22.363, and $27.208 million (2000 dollars, SCE share) in the years 1996 through 2000 respectively. SCE explains that while expenses were relatively flat from 1997 to 1999, they were higher in 1996 and 2000 due to large overhaul outages. Using five-year averages for all FERC accounts to smooth out the effect of the outages, SCE forecast test year 2003 O&M expense of $23.182 million for Four Corners. SCE notes that the 1996-2000 recorded expenses used in its forecast are among the lowest recorded over the last 21 years of data examined (using constant 2000 dollars).

ORA accepted SCE's five-year averaging methodology for FERC operations Accounts 500.15 through 507.15, because their recorded expenditures were constant. ORA proposes that three-year averages (1997-1999) be used for FERC maintenance Accounts 510.015 through 514.015 because the year 2000 indicates substantial increases in maintenance costs as a result of a major overhaul. Since no major overhauls are planned for the test year, and the next one had been planned for 2006,25 ORA concluded that the test year forecast should exclude such costs. ORA's method yields a forecast of $9.994 million for Four Corners O&M, which is $2.384 million less than SCE's forecast of $12.378 million for maintenance accounts.

Discussion

SCE contends that by excluding 1996 and 2000 from the average, ORA's proposal would not compensate SCE for major overhaul outages such as those experienced in 1996 and 2000. Although there is no major outage overhauls planned for Four Corners in 2003, SCE has shown that a major overhaul will occur in 2004. ORA's use of a three-year average would underfund SCE for an overhaul in 2004 by inappropriately excluding years in which a major overhaul occurred. We believe the appropriate forecast is SCE's use of a five-year average because although there is no major overhaul planned in the test year of 2003, there is indeed one planned for 2004. We are cognizant that the purpose of this GRC is to determine the just and reasonable rates for the 2003 test year, yet we do not want to create the incentive for utilities to schedule major projects for the test year because we believe this will unnecessarily overfund the utilities in the post-test years before its next GRC.

SCE's methodology for Four Corners maintenance accounts is reasonable and yields the better forecast. SCE's proposal to use a five-year average to forecast test year 2003 O&M expense of $23.182 million for Four Corners is therefore adopted.

SCE operates and maintains 36 hydroelectric (hydro) generating plants consisting of 79 generating units with an aggregate 1,156 MW of capacity. Some of the facilities have exceeded 100 years of age, and most were constructed at least 50 years ago. Initial plant service dates range from 1893 to 1999.

Using project-specific analysis, SCE forecasts capital expenditures of $17.691 million in 2002 and $15.121 million in 2003 for its hydro system, a total of $32.812 million. These forecasted amounts are somewhat lower than the 1996-1999 average annual level of expenditures of $19.670 million.

As with SCE's other generation forecasts, ORA found that in its opinion SCE failed to provide adequate project-specific information. ORA therefore recommends a four-year average of historical expenditures as the basis for the capital expenditures forecast. ORA's approach yields a forecast of $14.432 million for each of the years 2002 and 2003, or a two-year total of $28.864 million. This represents a forecast that is $3.948 million less than SCE's two-year forecast of $32.812 million.

Discussion

ORA does not take issue with any particular capital project that SCE plans to undertake. Instead, ORA faults the methodology used by SCE to develop its forecast. As we have indicated earlier, a project-specific forecast has the potential for being the most reliable basis upon which to forecast capital expenditures. On the other hand, because projects may not always be implemented as planned, it may be appropriate to use other forecast methodologies such as averaging. With respect to SCE's hydro facilities, we find that SCE's project-based forecast is consistent with a properly determined averaging approach, i.e., one that uses five instead of four years and that accounts for inflation.

SCE has shown that when these corrections are made to ORA's averaging-based forecast, the result is a reduction of $942,000 from SCE's project-based forecast for 2002 and an increase of $1.888 million from SCE's forecast for 2003. This confirms the reasonableness of SCE's original forecast for hydro capital, which we therefore adopt.

SCE forecasts hydro O&M expenses of $27.771 million for 2003. This represents an increase of 12% above the average level of expenditures for the 1996-2000 recorded period ($24.789 million). SCE asserts that the hydro facilities' advancing age is the major factor underlying this increase. In addition, SCE plans to undertake certain maintenance initiatives, notably painting. SCE has not painted most of the powerhouses, flow lines, penstocks, and generation equipment for over 20 years (Exhibit 18, p. 20). Other planned activities that are driving O&M increases include penstock and flow line condition assessments, wicket gate maintenance, and buttress repairs at Florence Dam.

ORA recommends an O&M forecast of $25.661 million for SCE's hydro system, which represents a reduction of $2.110 million from SCE's forecast.

Discussion

SCE has shown the unquestionable importance of maintaining its hydro facilities by, among other things, maintaining protective coatings (painting), ensuring the safe condition of penstocks and flow lines, ensuring efficient and reliable operation of wicket gates that control the amount and distribution of water to the turbine, and ensuring the safety of Florence dam by repairing its buttresses because they are subject to spalling due to extreme freeze/thaw conditions. We have no doubt that SCE's failure to carry out these activities could jeopardize the safe and reliable operation of its hydro system.

However, with the exception of the Florence Dam buttress repairs, which were suspended in the 1990's while seismic retrofit work was anticipated, SCE has not adequately explained why these important activities have taken on the degree of added importance during the GRC test period reflected in its forecast. More to the point, SCE has not satisfactorily explained what events have occurred requiring the expenditure of an additional 12% each year to operate and maintain the hydro system. For example, SCE has not adequately explained why it did not paint certain hydro facilities for more than 20 years but must do so during the ratemaking test period.26

We would expect a system most of whose components are 50 to 100 years old to have relatively high maintenance costs. However, during the recorded period from 1996 to 2000 the system was already old. SCE has not explained to our satisfaction why its test period maintenance costs should be so much higher than those of the same system were when those components were 45 to 95 years old. The fact that SCE did not apply paint for 20 years suggests inappropriate deferred maintenance as a possible answer.

The questions about SCE's hydro O&M forecast cast significant doubt as to the forecast's reasonableness. With one exception, ORA's forecast gives appropriate weight to past spending and needed projects, and is more reasonable than SCE's. We note that where SCE proposes a 12% increase, ORA's forecast represents an increase of 3.5% above the historical level of 1996 to 2000. ORA's forecast is therefore adopted, provided, however, that we will add $845,000 to ORA's forecast to include costs for the Florence Dam buttress repairs.

SCE's Pebbly Beach Generating Station on Santa Catalina Island consists of six diesel generators that represent the principal power source for 1800 residential and 500 commercial customers. SCE forecasts test year O&M expenses of $1.518 million. SCE forecasts capital expenditures of $2.5 million for a Selective Catalytic Reduction system to reduce emissions ($2.2 million in 2002 and $0.3 million in 2003), $0.8 million in 2002 to repair and upgrade the fuel pier, and $0.3 million in 2003 to cover other routine capital related work. ORA agrees with SCE's O&M estimate. In a data request response, SCE indicated that the cost of the fuel pier project is less than is stated in SCE's testimony. ORA therefore proposes to remove $445,000 from SCE's capital request.

Discussion

SCE agrees with ORA's proposed $445,000 adjustment for the fuel pier capital expenditures. Therefore, both SCE's O&M forecast and its capital forecast, as adjusted, are uncontested. We will adopt these forecasts for Other Generation.

In response to Public Utilities Code Section 367,27 an electric industry restructuring statute that addressed utility recovery of uneconomic generation costs, as well as D.97-09-048, in which the Commission provided for the implementation of Section 367, SCE filed A.99-04-024 to recover approximately $83 million in capital additions that it made to certain non-nuclear generation plant in 1997 and early 1998.28 Earlier, by A.97-10-024, SCE had applied for "Competition Transition Charge" (CTC) recovery of its generation-related capital additions for 1996.

Commission restructuring policy promulgated in D.95-12-063 and legislative direction embodied in AB 1890 contemplated the divestiture of certain utility generation assets as part of the transition to a more competitive electric services industry. However, the subsequent breakdown of the wholesale electricity market led to significant changes in the policy for divestiture. Among other things, the Legislature amended Section 377 in January 2001 to (1) require public utilities to retain electric generation facilities until January 1, 2006 and (2) require that the Commission ensure that these facilities remain dedicated to service for the benefit of California ratepayers.29

The amendment to Section 377 was adopted after the evidentiary record in A.99-04-024 was closed. By ruling dated September 7, 2001, the assigned ALJ in A.99-04-024 noted that parties in that proceeding had not had an opportunity to comment on any effects that this statutory amendment might have on the criteria for evaluating the application. The ALJ set aside submission of A.99-04-024 and provided parties an opportunity to file comments and reply comments on the effects of amended Section 377. SCE, ORA, and TURN filed comments, and TURN filed reply comments. Upon review, the ALJ determined that the record in A.99-04-024 remained incomplete with respect to capital additions made for the purpose of plant reliability and obsolescence.

A subsequent joint ruling issued by the ALJs assigned to A.99-04-024 and this GRC on September 9, 2002 stated the following:

The ALJs assigned to A.99-04-024 and this proceeding jointly determined that this is the more appropriate proceeding to address the 1997 and 1998 reliability and obsolescence capital additions to SCE's coal and hydro facilities. Their September 9, 2002 ruling directed SCE to submit testimony on the reasonableness of those capital additions and how they support continued reliable operation of utility-retained generation through January 1, 2006. The ruling specified, by individual project, $32.479 million in reliability and obsolescence capital additions to be addressed in this GRC.30 The ruling further provided that consideration of capital additions for environmental, regulatory and safety and for divested fossil plants would remain with A.99-04-024.31

SCE served its supplemental testimony regarding the 1997-98 capital additions on October 1, 2002; this was later identified as Exhibit 71. SCE's policy witness also addressed the capital additions in Exhibit 69, Section I. SCE maintains that all of the $30.937 million in projects identified in the September 9 ruling comply with the four review criteria established in D.97-09-048. SCE also maintains that Section 367 does not apply to the capital additions reviewed in this docket. Nevertheless, SCE contends that all of the projects at issue meet the Section 367 requirement that the capital additions must be necessary to maintain the facilities through 2001.

ORA and TURN submitted written testimony on the 1997-98 capital additions on December 6, 2002. ORA takes the position that reasonableness of a capital addition must be evaluated on the basis of facts and circumstances known when the utility made the decision to make the addition. ORA states that it remains committed to the project-specific disallowances that it recommended in A.99-04-024. With respect to the projects transferred to this proceeding, ORA proposes disallowances of $14.932 million and $12.463 million, respectively, to SCE's coal and hydro additions, or a total of $27.395 million of the $30.937 million sought by SCE. TURN submitted testimony proposing disallowances totaling $7.642 million for Mohave Unit 1 centrifuge replacements, and on brief it recommends additional disallowances for investments at SCE's coal plants.

SCE submitted rebuttal testimony addressing the testimony submitted by ORA and TURN.

SCE witness Fielder introduced SCE's policy perspective on the 1997-98 capital additions as follows:

ORA and TURN urge that the reasonableness review criteria established by D.97-09-048 be observed in this GRC proceeding. Both of these parties base their proposed disallowances on their contention that SCE has failed to demonstrate compliance with those criteria with respect to various projects.

Discussion

As a condition precedent to utility recovery of uneconomic costs associated with capital additions made to generating facilities that were in existence as of December 20, 1995, Section 367 required that such additions be necessary to maintain those facilities through December 31, 2001, and reasonable as determined by the Commission. For capital additions made in 1996 and 1997 (and in 1998 until the ISO and PX commenced operations), D.97-09-048 further provided that cost recovery would be based on the Commission's after-the-fact (ex post facto) reasonableness review of recorded expenditures. Among other things, the Commission established the following reasonableness review criteria:


In their applications [for transition cost recovery], PG&E, SDG&E, and SCE shall demonstrate how their requests for recovery of capital addition costs meet the following criteria, among others: (1) consistency with recent capital budgets and expenditures for respective power plants, (2) the need for compliance with other regulatory requirements, (3) cost-effectiveness and (4) the impact of the capital addition on the unit's heat rate. PG&E, SDG&E, and SCE, and other interested parties may propose additional evaluation criteria for Commissions consideration. (D.97-09-048, Ordering Paragraph 3, 75 CPUC 2d 449.)

At issue now is whether the reasonableness of the generation capital additions that SCE made in 1997 and early 1998 should be evaluated according to the criteria used for other capital additions reviewed in this GRC, as SCE urges, or according to the criteria established in Section 367 and D.97-09-048, as ORA and TURN urge. If the higher standards urged by TURN and ORA are applicable, then shorter payback periods may apply for cost-effectiveness analyses of capital projects, and SCE must show that improvement of a generating unit's heat rate was not the purpose of the expenditure.

This issue harkens back to the recent past, when the unambiguous policy of this state was promotion of a competitive generation market. As the Commission explained in D.97-09-048, in which it established the criteria for determining eligibility for transition cost recovery for generation capital additions generally:


[T]he Legislature has given us latitude to adopt one or a combination of approaches for determining the reasonableness of transition cost recovery for capital additions. In doing so, we keep foremost in our minds the objective of creating a level playing field for all market participants during the transition to a fully competitive electric services industry. How we handle the issue of capital additions is a critical aspect of creating this level playing field. We do not wish to establish standards of reasonableness that afford utilities an unfair advantage in the market, particularly at ratepayers' expense. At the same time, we wish to encourage utilities to make cost-effective investments that will maintain the reliability of the electric system. As intended by [Section 367], utilities should have the opportunity to recover the costs of those investments during the transition period. (D.97-09-048, 75 CPUC 2d, 442.)

To implement the objectives of a level playing field and encouraging cost-effective investments that would maintain reliability, the Commission determined in D.97-09-048 that a "market control" approach was appropriate once the ISO and PX began functioning. The Commission determined that ex post facto reasonableness review was appropriate only for those capital additions made in 1996 and 1997. (As noted earlier, D.98-03-054 extended this period through March 1998.)

With the enactment of ABX1-6, we have new responsibilities to fulfill.32 Where it was once the policy that utility generation would either be divested or subjected to market prices if retained, and that a level playing field be established by, among other things, preventing inappropriate ratepayer subsidies of generation offered into the market, Section 377 now requires that utilities retain their generation facilities at least through 2005, subject to Commission regulation. It further requires that we act to ensure that these generation facilities remain dedicated to service for the benefit of California ratepayers.

In D.02-11-026, we commented on how ABX1-6, including its amendment of Section 377, impacts Section 367. The following excerpt from that decision bears repetition here:


These provisions of ABX1-6 clearly and expressly confer on the Commission jurisdiction over regulation of the utilities' retained generation assets, including rates. Such jurisdiction includes, for example, authority to determine whether and to what extent the utilities may recover in rates their investments in these retained generation assets. Moreover, by conferring upon the Commission the authority to continue to regulate the utilities' retained generation under a cost-of-service approach, and deleting provisions requiring generation to be transitioned from regulated to unregulated status, these provisions removed any danger that the investment in such assets "may become uneconomic as a result of a competitive generation market."33 (§ 367.) In other words, the investment in these assets no longer is a stranded or transition cost within the meaning of AB 1890. Thus, recovery of these investments is no longer barred by AB 1890's prohibition on the recovery of stranded costs after the end of the rate freeze.34 (D.02-11-026, pp. 13-14.)

No party has explained how we can fulfill our new obligation to ensure that SCE's coal and hydro facilities remain dedicated to service for the benefit of ratepayers through at least 2005 while limiting ratepayer funding responsibility only to those capital additions needed through 2001. Additionally, as we determined in D.02-11-026, the requirements of Section 367 are no longer applicable to the extent that those requirements conflict with those of ABX1-6. We also note that with the enactment of ABX1-6, utilities do not have the opportunity they previously had to recoup through divestiture capital investments that are not funded by ratepayers.

It follows that we must now provide for the utility's opportunity to recover the reasonable costs of maintaining and operating those facilities through 2005. It is no longer appropriate to hold the utility to the Section 367 requirement that capital additions be necessary to maintain the facilities only through 2001, or to the D.97-09-048 requirement that capital additions to improve the unit's heat rate be excluded. Doing so could deny the utility the opportunity to recover its reasonable costs for capital additions that benefit ratepayers, and it would be no more appropriate than it would be to adhere to the requirement that post-1998 capital additions considered in this GRC be subjected to the now-irrelevant market control approach.

ORA and TURN state the correct rule for retrospective (ex post facto) reasonableness reviews generally, as applicable to capital additions --reasonableness should be evaluated on the basis of facts and circumstances known or knowable when the utility made the decision to make the addition. However, for the reasons explained above, ABX1-6 has changed how we must evaluate these capital additions. We find that this general rule for reasonableness review is inapplicable in the very limited circumstances here. Accordingly, we will not hold SCE to the requirement that the 1997-98 capital additions be necessary to maintain the facilities through 2001, or to the requirement that capital additions made to improve a unit's heat rate be excluded.

Appropriately, SCE no longer requests CTC recovery of the 1997-98 capital additions at issue here. Instead, SCE requests authorization to include these additions in rate base, just as the other generation capital additions considered in this GRC are. We will adopt this approach for the coal and hydro capital additions that we find to be reasonable, as discussed below. For the reasons discussed earlier these, are not "uneconomic" costs within the meaning of Section 367, and we reject TURN's assertion that Section 367(a) prohibits their recovery in this GRC.

In its reply brief TURN raises the concern that SCE may be seeking double recovery of capital additions to the extent that such costs are currently included in rate base pursuant to D.02-04-016. SCE's testimony (Exhibit 71, Section VII) appears to adequately explain the company's ratemaking proposals. Nevertheless, to avoid any possibility of double recovery, SCE should include in the advice letter filing it makes pursuant to this order a demonstration that the costs associated with these capital additions for 1997 and 1998 are included in rate base only once.

2.7.3. Cost-Effectiveness Analyses

SCE maintains that all of the 1997-98 capital additions projects transferred to this docket were cost-effective using reasonable assumptions. For its cost-effectiveness analyses, SCE assumed that capital additions costs would be paid back over the economic life of the addition or the economic life of the plant in which it was installed, whichever was shorter. SCE limited the economic life used in its cost-effectiveness analyses to no greater than 20 years for hydroelectric projects and coal projects evaluated prior to May 1996. For coal projects evaluated in May 1996 and thereafter, SCE limited the economic life used in its cost-effectiveness analyses to no greater than 10 years. Using these payback periods, SCE determined that all projects had benefit-to-cost ratios greater than 1.0, i.e., all projects had more benefits than costs.

ORA and TURN advocate payback periods of six years and 7-10 years respectively. These shorter payback periods affect the cost-benefit analyses for several projects, leading to substantial recommended disallowances.

Discussion

In D.99-03-055, the Commission found that a 20-year payback assumption is reasonable for 1996 capital additions. The Commission did, however, indicate that it would consider different payback periods for subsequent investments in generation plant.

The shorter payback period advocated by ORA for 1997 and 1998 capital additions is premised upon the ORA position that " [a] long-term useful life assumption is unreasonable and inconsistent with a decision to divest..." (Exhibit 229, Appendix B, p. 2-3.) TURN similarly contends that the shorter payback periods it recommends are appropriate in light of the advent of industry restructuring. Since we have determined that it is appropriate to evaluate the subject capital additions according to GRC criteria, and not the stricter standards of Section 367 and D.97-09-048 that were associated with the transition to a more competitive generation market, we find that the shorter payback periods advocated by ORA and TURN are not properly applicable here. Their recommended project disallowances that are based upon the shorter payback periods will not be adopted. The appropriate way to measure the cost-effectiveness of a capital addition is to review its benefits over the remaining life of the generating station or the life of the capital addition, whichever is shorter.

2.7.4. Performance Improvements

ORA and TURN recommend disallowances of SCE's capital additions for which significant efficiency improvements resulted. ORA did not include benefit/savings from enhanced unit performance when the percent of total benefit/savings was greater than 50%. TURN found that heat rate benefits accounted for nearly 70% of the total benefits from Mohave centrifuge projects, and recommends disallowance of the associated capital additions.

SCE responds that its capital additions meet the heat rate standard set forth in D.97-09-048 because none of the projects resulted in a significant improvement to the generating unit's heat rate even though, for some projects, a significant portion of the project's benefit was due to heat rate improvements.

Discussion

In this GRC, for the reasons discussed previously, we are not enforcing the heat rate standard adopted in D.97-09-048. Accordingly, we will not adopt recommended disallowances to the extent that those disallowances are based upon the asserted failure of a project to meet the standard.

2.7.5. Budget Variances

ORA identified 10 hydro projects for which actual project expenditures exceeded SCE's budget by more than 10%. In some cases the budget overruns were substantial - as much as 416%. For six of these projects, ORA recommends disallowances solely on the basis of the budget variances. For the other four projects, ORA recommends disallowances on the basis of unfavorable benefit/cost ratios as well as budget variances.

SCE introduced evidence showing that for the six projects for which ORA recommended disallowances solely on the basis of budget variance, the projects had benefit/cost ratios greater than one. SCE contends that merely because a project varies from budget does not mean it was not cost-effective. For the other four projects for which ORA bases its recommended disallowances on both inadequate benefit/cost ratios and budget variances, SCE has shown that the benefit/cost ratios exceed one.

Discussion

The record evidence shows that the subject hydro projects are cost-effective in that their benefit/cost ratio exceeds one. Since the benefits exceed the costs, thereby benefiting ratepayers, we will not disallow the projects' capital costs on the basis of budget overruns.

2.7.6. Project Timing

ORA takes the position that SCE is entitled to receive recovery for capital additions completed after December 31, 1996 and before April 1, 1998. ORA recommends disallowances for projects completed after March 31, 1998. According to ORA, these projects are not eligible for cost recovery by Commission authorization, but SCE may seek and obtain recovery through competitive market revenues. For example, a project to replace a road and parking lot destroyed by storm flooding at the Mammoth Pool Powerhouse had a work order completion date of August 1998, so ORA recommends disallowance of $392,000 in capital additions associated with the project.

SCE contends that it has requested recovery only for capital additions that were completed and in service during the period from January 1, 1997 through March 31, 1998. SCE further contends that the date that a work order is closed is immaterial. If there is a portion of work in a work order that is completed and in operation, then that portion is closed to plant in accordance with FERC accounting rules.

Discussion

SCE has shown that during the period of time between the opening of a work order and its closure, there could be many capital closings to plant in service. The fact that certain work orders associated with the 1997-98 projects at issue in this GRC were closed after March 31, 1998 does not demonstrate that SCE has impermissibly sought capital cost recovery for projects outside the scope of this proceeding. ORA's proposed disallowances that are based upon assertions of improper timing will not be adopted.

2.7.7. Casualty Loss Projects

For recovery of project costs related to plants damaged by flood, fire, etc., ORA recommends allowing only 60% of the projects' costs. ORA reasons that SCE can claim the loss on its income tax return, and that since the company's tax rate is approximately 40%, SCE should be allowed recovery of only 60% of the capital additions cost.35 SCE contends that these 40% disallowances are improper.

Discussion

Casualty losses are tax benefits associated with the early retirement of capital assets. The amount of a casualty loss is measured by the remaining undepreciated basis of the asset, not the cost to replace the asset. Casualty losses are deductible for income tax purposes, and benefit ratepayers. Because the benefit of a casualty loss goes to ratepayers, the benefit does not provide for recovery of the cost to replace the asset. ORA's proposed 40% disallowances are therefore improper, and will not be adopted.

2.7.8. Conclusion - 1997-98 Capital Additions

We have not found it necessary to examine in detail whether each of the generation capital additions that SCE made in 1997-98 fully complies with the Section 367 requirement that the additions be necessary to maintain the generation facility through December 2001. We also have not examined in detail whether certain of the disputed capital additions contributed substantially to the efficiency of the plants' operations. This is because ABX1-6 has changed how we must evaluate the reasonableness of the capital additions and their eligibility for cost recovery from utility ratepayers.

The provision of amended Section 377 requiring that we ensure that utility-retained generation facilities remain dedicated to service for the benefit of ratepayers means that we must give reduced or no weight to previous review requirements that were intended to promote a competitive generation market. Even if it would have been appropriate to deny recovery of otherwise reasonable capital additions because they did not meet the standards of Section 367 and D.97-09-048, it would now be inappropriate to deny recovery of cost-effective capital additions that are consistent with Section 377's dedicated service requirement.

With respect to the GRC standards applicable here, SCE has made a showing of reasonableness for the 1997-98 generation capital additions underlying its request for $30.937 million. SCE's request will be approved.

3. Transmission and Distribution (T&D)

3.1. Introduction

When California's electricity industry restructuring program was implemented in 1998, SCE transferred operational control of its transmission facilities to the ISO, subject to FERC regulation. SCE's distribution and subtransmission facilities remain under this Commission's jurisdiction. SCE's showing in this GRC includes a total O&M and capital forecast for both Commission and FERC jurisdictional T&D facilities. A transmission revenue credit offsets the costs of FERC jurisdictional transmission assets and the associated O&M expenditures.

SCE's Transmission and Distribution Business Unit (TDBU) manages the company's T&D facilities. The TDBU has approximately 4,400 full-time employees who plan, design, construct, operate and maintain approximately $11.3 billion in utility assets. The SCE system serves 4.4 million customers over a 50,000 square mile service territory, and includes approximately 1.5 million poles, 95,000 circuit miles of wire, 306,000 underground structures, 667,000 transformers, 43,000 switches, and 750,000 streetlights.

SCE forecasts $274.7 million (2000 dollars) for T&D O&M expenses for test year 2003.36 SCE also forecasts more than $2 billion for new capital expenditures during the years 2001 through 2003. ORA recommends a $7.6 million O&M expense reduction ($1.5 million transmission and $6.1 million distribution) and a $459 million reduction in plant additions.37 TURN recommends a $1.7 million O&M expense reduction, a $29.7 million reduction in 2002-2003 capital additions, and a penalty of at least $48.8 million and as much as $61 million for SCE's alleged failure to adequately manage its wood pole inspection and replacement program. In addition, TURN has offered recommendations regarding SCE's line and service extension rules. Section 4 of this decision addresses these issues as well as contested issues regarding SCE's funding request for electric transportation.

3.2. Level of Reliability Performance

As measured by Average Customer Minutes of Interruption (ACMI) and Frequency of Circuit Interruptions, SCE's year 2000 reliability performance was among the best of the past 15 years. SCE's ACMI over the years 1996-2000 was 56.32 minutes, an improvement of 2.86 minutes compared to the 15-year average of 59.18 minutes. The five-year average of frequency of circuit interruptions was 9,546, a reduction of 512 interruptions from the 15-year average of 10,058.

SCE states that it is not seeking to enhance reliability through its T&D funding request in this GRC. Instead, SCE maintains, its requested funding level is necessary to avoid a significant degradation of the current level of reliability and to avoid significantly higher replacement costs.

ORA agrees that SCE generally provides reliable service to customers over its T&D system. In fact, ORA believes there is reason to be concerned that SCE's reliability may actually be too high, and that rates may therefore be too high. ORA states the following:


Reliability costs ratepayers money, as Edison has reminded us during the rate case. That is why the reasonable rates requirement is at tension with reliable service. Edison has not assessed, at least for this rate case phase, how different customers and customer classes value reliability. How much decreased reliability would customers accept in exchange for more reasonable rates? The evidence provides no answer, although we know that Edison customers are less than happy about their rate levels. (ORA Opening Brief, p. 33.)

ORA proposes that we require SCE to provide a comprehensive report with its next GRC application that provides detailed evidence on the specific monetary trade-off associated with specific reliability levels, and also provides empirical findings on customer willingness to accept lower reliability in return for lower rates. ORA also proposes that SCE's next GRC be used to review and address whether all customers receive a similar level of service for the rates they pay as do other customers in the same customer class.

Discussion

SCE's T&D policy witness testified that the company's reliability is generally high, customer satisfaction is moderately rising, controllable costs are moderate, and that "[w]e have accomplished this balanced performance through our focus on long-term planning, management oversight, and a commitment to long-term investments." (Exhibit 21, p. 3.) Unfortunately, it is not clear whether the planning, management and oversight performed by SCE's TDBU executives is adequately informed by empirical analyses such as those proposed by ORA.

It is at least theoretically possible to create an overly reliable T&D system that adds unnecessarily to ratepayer costs. In other words, ratepayers might be willing to accept somewhat diminished reliability performance if sufficient rate reductions could be realized in return. At the same time, as noted earlier in this decision, we are mindful that SCE's retail prices are comparatively high according to several standards, and that high rates are harmful to ratepayers and the California economy. While the prospect of "too much reliability" might seem remote, it is reasonable for regulators to take steps to avoid the realization of that theoretical possibility. We therefore accept ORA's proposal to carefully consider the issue in SCE's next GRC. We emphasize that nothing in the evidentiary record before us shows that SCE's T&D system is in any sense too reliable, or that any significant cost savings could necessarily be expected from a decision to target reduced reliability levels.

SCE notes that it performed a value-of-service study in 2000, and claims that ORA's proposal for a showing in the next GRC is therefore moot. We will direct SCE to include cost/reliability tradeoff and value-of-service analyses with its next GRC filing that address the concerns raised by ORA. If SCE believes that its 2000 study adequately addresses these concerns, it is free to offer that study in fulfillment of this requirement. In the event that SCE does so, it will be incumbent upon it to show that the study remains valid despite the passage of time, and otherwise fulfills this requirement.

We find no record evidence prompting us to comment on ORA's request that SCE's next GRC be used to examine whether all customers receive a similar level of service. ORA may wish to raise this issue at the commencement of SCE's next GRC. We suggest that SCE and ORA confer on this issue before SCE makes its next GRC filing, so that SCE has an opportunity to understand and address ORA's concerns in its direct showing.

3.3. Wood Pole Inspections

3.3.1. Introduction

TURN, which is "extremely critical" of SCE's past T&D maintenance practices (TURN Opening Brief, p. 42), asks that we impose a penalty for deferred pole maintenance of at least $48.8 million and as much as $61 million. As grounds for this penalty, TURN contends that SCE failed to adequately manage its wood pole inspection and replacement program over the past 20 years. TURN also recommends that SCE's shareholders be required to contribute more than $1 million annually to fund approximately 30,000 annual intrusive pole inspections, while ratepayers would fund 95,000 inspections per year. TURN also requests that SCE be directed to report annually on intrusive pole inspections completed in the prior year.

3.3.2. Historical Pole Inspections

Underlying TURN's proposed penalty and its related recommendations is a two-decade pattern of ratemaking authorizations for pole inspections and SCE's actual spending on inspections during the period. Before addressing TURN's recommendations, we examine disputed factual contentions associated with TURN's analysis, which is summarized in the following table.

TURN's Analysis of Number of Intrusive Pole

Inspections: Ratepayer Funded and Completed

Year

Funded Inspections

Completed Inspections

Cumulative Funded

Cumulative completed

Annual excess

Cumulative excess

 

(a)

(b)

(c)

(d)

(e)=(a)-(b)

  

1984

77,800

50,864

77,800

50,864

26,936

  

1985

77,800

66,248

155,600

117,112

11,552

  

1986

77,800

48,155

233,400

165,267

29,645

  

1987

77,800

22,197

311,200

187,464

55,603

  

1988

66,248

37,239

377,448

224,703

29,009

  

1989

66,248

11,367

443,696

236,070

54,881

  

1990

66,248

47,315

509,944

283,385

18,933

  

1991

66,248

54,539

576,192

337,924

11,709

238,268

1992

30,000

8,821

606,192

346,745

21,179

  

1993

30,000

0

636,192

346,745

30,000

  

1994

30,000

0

666,192

346,745

30,000

81,179

1995

113,000

51,371

779,192

398,116

61,629

  

1996

113,000

67,508

892,192

465,624

45,492

  

1997

113,000

90,609

1,005,192

556,233

22,391

  

1998

113,000

47,486

1,118,192

603,719

65,514

  

1999

113,000

9,476

1,231,192

613,195

103,524

  

2000

113,000

102,991

1,344,192

716,186

10,009

  

2001

113,000

120,796

1,457,192

836,982

-7,796

  

2002

113,000

125,000

1,570,192

961,982

-12,000

288,763

TOTALS

1,570,192

961,982

      

608,210

SCE's 1983 "Deteriorated Pole Report" incorporated the idea that a strategic, cost-minimizing program would optimize maintenance versus capital investments by coordinating pole inspections, pole repairs and pole replacements. That study was the genesis of SCE's pole inspection, repair and replacement programs. It recommended intrusive inspection38 of all poles older than 20 years on a 15-year cycle, a goal adopted by the company and proposed and accepted in SCE's 1985 GRC.39 It resulted in the recommended inspection of 77,800 poles annually. At the heart of the current issue is TURN's contention that beginning in 1984, SCE repeatedly deferred spending on the pole inspections envisioned in this program so that by 2002, the company had inspected only 961,982 poles, or about 61% of the 1,570,192 inspections that ratepayers funded during that period.

With respect to its contention that ratepayers funded more than 1.5 million intrusive inspections from 1984 to 2002, TURN's position is summarized below:

· During this time period SCE's rates were approved in five GRCs, with test years in 1983, 1985, 1988, 1992, and 1995. The Commission explicitly addressed funding for intrusive pole inspections in the 1985 and 1995 GRCs, and even though the Commission did not explicitly do so in the 1983, 1988, and 1992 GRCs, it is possible to determine the amounts implicitly authorized for intrusive inspections.

· In the 1985 GRC, the Commission explicitly approved 77,800 intrusive inspections annually for 1985-1997.

· In the 1995 GRC, the Commission approved 113,000 intrusive inspections annually from 1995 through 2002.

· In D.84-12-068 (the 1985 GRC decision), the Commission noted that in the 1983 GRC, SCE received funding for a deteriorated pole replacement program. As stated in the 1983 study, 77,800 inspections could be undertaken "within the parameters of the committed dollars in the 1983 rate case filing," and SCE's witness in this GRC agreed that the 1983 GRC funds were apparently adequate. TURN contends that it is reasonable to conclude that 77,800 inspections in 1984 were funded by ratepayers in the 1983 GRC.

· In the 1988 GRC, SCE's funding level was based on the 1985 recorded spending level. In 1985 SCE completed 66,248 inspections. TURN thus uses 66,248 inspections as the ratepayer funded level for 1988-1991.

· In the 1992 GRC, SCE was funded for at least 30,000 intrusive pole inspections per year for 1992-1994, as shown in evidence in SCE's 1995 GRC.

SCE takes issue with several aspects of TURN's analysis and conclusion that SCE completed only 61% of the intrusive inspections that ratepayers paid for, as discussed below.

Discussion

While it acknowledges that pole inspection forecasts were explicitly addressed in the 1985 and 1995 GRCs, SCE implies that neither forecasts nor commitments regarding pole inspections were made in the other GRCs over the past 20 years. However, TURN has demonstrated that it is unreasonable to make this assumption. Even when not explicitly addressed, SCE's T&D O&M costs have included a pole treatment and inspection component in every GRC beginning with the test year 1983 GRC. Moreover, it is possible to make an estimate of the number of pole inspections connected to funding received in the 1983, 1988 and 1992 GRCs, as TURN has done. This is consistent with evidence showing that SCE's prevailing policy for pole inspections from 1983 to 1997 was the strategy outlined in the 1983 study. The fact that SCE temporarily halted pole inspections in 1993 and 1994 in order to address a backlog of pole replacements does not change this conclusion.

Moreover, even though SCE's 1995 GRC was succeeded by a PBR mechanism in more recent years, it is reasonable to conclude that the level of pole inspections included in the 1995 GRC continued throughout the PBR period, since the GRC authorization was used as the base year for the PBR. Significantly, SCE's testimony in the 1995 GRC stated "starting in 1995 and continuing through 2003, we will schedule 113,000 pole inspections yearly until the remaining inventory of over one million uninspected poles have been inspected and treated."

While we can determine a reasonable estimate of the number of inspections that have been funded by ratepayers over the past two decades, TURN's contention that all of these inspections were to be intrusive is not supported by the record. The 1983 study on wood poles recommended that three categories of poles undergo an intrusive type inspection: those in service for 20 years or longer, second hand set poles, and Cellon poles. Other poles would undergo visual inspection only. SCE has demonstrated that that the 77,800 annual inspections forecast in the 1985 GRC and the 113,000 annual inspections forecast in the 1995 GRC included visual-only inspections as well as intrusive inspections. This was confirmed by an SCE rebuttal witness who was familiar with the 1983 study and testified that it was SCE's intent to intrusively inspect only a subset of the poles it would visually inspect each year.

Unfortunately, the witness did not specify the size of the subset. However, given the age of SCE's pole inventory, it is clear that the majority of the T&D poles were and are greater than 20 years old and would therefore have qualified for intrusive inspections under SCE's criteria. TURN points to evidence that only 20% of SCE's poles were in service less than 20 years. We will use this data to correct TURN's estimate of the number of intrusive inspections reflected in the GRC forecasts for the last 20 years. Accordingly, we will reduce TURN's calculation of 1,570,192 total inspections by 20% to arrive at a corrected estimate of 1,256,154 intrusive inspections funded in GRCs over the past 20 years. We note that this estimate is conservative (i.e., it favors SCE) because it does not account for second hand set poles and Cellon poles.

SCE points out that the approval of General Order (GO) 165 with an effective date of March 1, 1997 impacted its obligation to perform intrusive inspections. SCE contends that in light of the new obligations created by GO 165, TURN errs when it asserts that SCE's forecast of inspecting 113,000 poles per year extended beyond the 1995 GRC cycle. This contention is without merit. Nothing in the GO 165 requirement to intrusively inspect wood poles older than 15 years by 2007 affected the forecast of 113,000 visual/intrusive inspections per year underlying SCE's ratemaking authorization. As the Commission stated in D.97-03-030, which adopted GO 165, the new standards were maximum acceptable lengths for inspection cycles. Nothing in GO 165 suggested that SCE should slow the pace of intrusive inspections that was set in the 1995 GRC and continued through the PBR mechanism.

3.3.3. Proposed Penalty

We have determined that over a period of nearly 20 years and five GRCs, SCE received funding authorization associated with forecasts for 1,570,192 pole inspections. We have also determined that an estimated 80% of these inspections, or 1,256,154 million, were to be intrusive inspections that included a treatment to extend the pole life. During that same period, exercising its managerial discretion to devote authorized funds to other uses, SCE actually completed 961,982 intrusive inspections, or 77% of the funded level. We next consider whether this performance by SCE constitutes wrongdoing for which a penalty is appropriate. We then consider (in Section 4.3.4) whether a ratemaking disallowance should be adopted in this GRC as a result of this performance.

Discussion

It is a fundamental tenet of forecast test year ratemaking that inclusion of a particular expense category in a GRC authorization does not create a specific obligation for the utility to spend the authorized amount during the test year. Thus, utility management is generally provided discretion regarding use of funds and is not bound by the adopted forecast.40 However, as we observed previously in this decision, the Commission held in an SDG&E GRC that there are limits to that management discretion:


When a utility's expense estimate includes the performance of a task it had planned to accomplish with previously authorized funds, we will want to know why the utility did not spend its funds as planned the first time around and will be hesitant to charge ratepayers twice for the same expense. (D.92-12-019, 46 CPUC 2d 538, 555.)

More recently, in PG&E's test year 1999 GRC, the Commission once again addressed the problem of deferred maintenance and a utility's failure to perform activities that have been funded by forecasts adopted by the Commission:


It would be unjust and unreasonable to make ratepayers responsible for expenses directly attributable to deficient or unreasonably deferred maintenance, or to make ratepayers pay a second time for activities explicitly authorized by the Commission in the past. (D.00-02-046, Conclusion of Law 15, p. 536.)

SCE has provided no reason why we should exempt it from these ratemaking principles, and we will not do so. With respect to intrusive pole inspections, SCE's exercise of its managerial discretion over a period of nearly 20 years has led to a situation in which the company has completed an estimated 77% of the intrusive inspections for which it received funding through the GRC/PBR ratemaking process. This raises the question whether SCE is seeking funding in this GRC for pole inspections that could have and should have been performed already with previously authorized funds. We will return to this ratemaking question, but we first consider whether any aspect of the 23% shortfall in completed intrusive inspections warrants the imposition of a penalty as proposed by TURN.

This is not the first occasion we have had to critically review SCE's management of its pole inspection program. In the 1995 GRC, taking into account overriding safety concerns, the Commission rejected a proposal to reduce expenses for SCE's forecast level of inspections. However, the Commission imposed annual reporting requirements regarding the number of inspections completed in 1995, 1996, and 1997, and stated that it would consider imposing penalties on SCE for compromising the safety of its distribution system if, at the end of each of those years, SCE had failed to meet its inspection target for the year. (D.96-01-011, 64 CPUC 2d, 282, 392.)

Utilities are required to comply with relevant statutes, Commission GOs, and decisions, and the Commission has the statutory obligation to require utilities to do so. (See, e.g., Pub. Util. Code §§ 702, 2101 and 2106.) A utility's failure to comply with these statutes, GOs and decisions means a utility has violated them. The Commission may impose a penalty upon a utility when a violation by that utility has occurred.41 With respect to the penalty proposed by

TURN, the relevant Commission orders are a series of ratemaking decisions issued in five GRCs over a period of 20 years.

Notwithstanding SCE's failure to achieve its planned and funded pole inspection targets over the past 20 years, and the fact that SCE was on notice that it might incur a penalty if it failed to perform the target number of inspections in 1995, 1996, and 1997, we do not find that imposition of a penalty for this failure is justified.

We cannot find that a utility has failed to comply with, and therefore violated, a Commission order to perform a certain number of inspections within a certain time period unless the order actually exists, or at a minimum, can be readily discerned from the decision adopting the order. With the exception of Ordering Paragraph 15 of D.96-01-011 in SCE's 1995 GRC, no reasonable argument can be made that over the past two decades SCE was under a specific mandate pursuant to a GRC decision to perform a given number of intrusive inspections in a particular time period.42 If we cannot find that SCE failed to comply with an order to perform inspections, no violation has occurred warranting the imposition of a penalty.

In addition, there is a significant timeliness issue with TURN's penalty recommendation. In effect, it requires that we hold that SCE has failed to comply with a Commission order or orders with respect to acts or omissions that occurred as long as 20 years ago. Even with respect to D.96-01-011, where we placed SCE on notice that it was subject to penalties, we are asked to determine whether SCE complied with inspection requirements by acts or omissions that occurred in 1995, 1996, and 1997. Put simply, too much time has elapsed for us to determine fairly whether a violation has occurred. Indeed, the need to contemporaneously evaluate compliance is consistent with the D.96-01-011 requirement for annual reporting on completed inspections by March 1 of the succeeding year. Bringing a penalty action in December of 2002 (when TURN submitted its testimony in this GRC) is inconsistent with due process requirements of an enforcement action.

3.3.4. Ratemaking Adjustments for Deferred
Inspections

Although we have determined that imposition of a penalty for SCE's pole inspection management practices is not justified, SCE's failure to perform an estimated 23% of the intrusive inspections it was funded to complete over two decades requires further consideration. This is necessary to ensure that ratepayers are not required to pay a second time for activities explicitly authorized by the Commission in the past, and to ensure that ratepayers are not held responsible for costs that are directly attributable to deficient or unreasonably deferred maintenance practices.

Discussion

We first consider whether SCE's funding request for intrusive inspections would represent "double charging" for previously funded activities. Second, we will determine whether SCE's proposed capital expenditures include pole replacements that would have been unnecessary in this GRC cycle if intrusive inspections had been timely completed in the past.

As to the first question, TURN recommends that SCE's request for funding for 140,000 intrusive inspections annually be reduced by 44,500 to 95,500 intrusive inspections. This results in a proposed disallowance of $1.7 million (2000 dollars), or approximately $38 per inspection. In addition to 95,500 ratepayer-funded inspections, TURN recommends that SCE be directed to perform 30,000 intrusive inspections annually at shareholder expense for a total of 125,500 inspections. TURN states that the 90,000 shareholder-funded intrusive inspections that would be performed over the three-year GRC cycle would partially offset the inspection "deficit" of 317,000 funded but non-completed inspections since the last GRC, so that ratepayers do not pay twice for inspections.

In its rebuttal testimony, SCE determined that it needs to perform an average of 125,000 intrusive inspections per year through 2006 to meet the requirements of GO 165. Thus, for what appear to be different reasons, SCE and TURN are in virtual agreement with respect to the number of annual intrusive inspections that need to be performed in this GRC cycle (125,000 v. 125,500). Their disagreement is over funding responsibility. We will adopt SCE's estimate of 125,000 annual intrusive inspections as necessary and adequate to achieve compliance with GO 165.

The main premise of TURN's recommended disallowance, the idea that shareholders should fund a portion of the required number of inspections, is the correct one because it gives effect to the principle that ratepayers should not be required to pay twice for the same authorized expense. Even if it was appropriate for SCE exercise its spending discretion in order to address emergent issues in any one year, it is also reasonable and appropriate to hold SCE accountable for completing the pole inspections that ratepayers paid for over a longer period. This is especially true beginning with the issuance of D.96-01-011 in January 1996, when SCE was on notice that, for safety reasons, we fully expected that the funded inspections were to be timely completed.

We will adopt TURN's approach for addressing the "double charging" problem, along with corrections and modifications, as follows. First, we are reluctant to reach back over several GRC cycles to address underperformance that could have and should have been addressed previously. We will instead focus on the period 1996-2002, i.e., the period following the issuance of the 1995 GRC decision in January 1996. We determine that SCE was funded to perform an estimated 632,800 intrusive inspections during this period (113,000 annual inspections, 80% of which are estimated to be intrusive, times seven years.) During the same period SCE performed 563,866 such inspections (67,508; 90,609; 47,486; 9,476; 102,991; 120,796; and 125,000, respectively, in 1996, 1997, 1998, 1999, 2000, 2001, and 2002). The difference, 68,934, represents the number of inspections that were funded for the seven-year period but not performed. We will disallow the costs for that number of inspections over the current three-year GRC cycle. Accordingly, we disallow $2.619 million for intrusive inspections that have already been funded by ratepayers (68,934 times $38 per inspection), or $873,164 annually. In addition, we will remove $570,000 annually (15,000 times $38 per inspection) to reflect SCE's estimate of 125,000 required inspections in lieu of its requested funding for 140,000 inspections. The total adjustment is $1.443 million per year for FERC Account 583.400, Inspections. While we adopt this ratemaking adjustment, SCE must fully comply with GO 165 inspection requirements.

In addition to showing that SCE has performed fewer inspections than it was funded for, TURN has demonstrated that, because intrusive pole inspections were delayed, SCE missed an opportunity to extend the life of those poles and reduce future replacement costs. SCE contends that pole treatments simply reduce the frequency at which poles will be replaced. We accept this assertion, but we are not so dismissive of the benefit of reducing the replacement frequency. SCE acknowledges that a pole's life can be extended 10 to 15 years with an intrusive inspection that includes treatment. The fact remains that SCE now has to replace poles that otherwise would have remained in service had they received timely fungicide treatment.

SCE also claims that TURN fails to recognize the prior useful life of poles whose inspections were delayed for a short period of time, and that TURN erroneously concludes that a missed pole treatment creates a cost to SCE ratepayers amounting to the entire cost of a pole replacement. This claim is without merit, as TURN puts forth no such conclusion. Similarly, even if SCE's contention that "short delays in the application of insecticide/fumigant treatments have minimal impact on pole life" were true, an inspection that has not been performed at all between 1996 and 2002 is not evidence of a short delay. Again, this contention is inapposite. SCE's delays in completing pole inspections do not merely represent "minor slippage" as the company claims.

TURN has calculated that each missed intrusive inspection translates into a current pole replacement cost of approximately $330 after 20 years, or a net present value of $59 at 9%, and on the basis of this calculation it proposed that a penalty of at least $50 be imposed for each missed inspection. Although we do not impose a penalty, TURN's analysis remains valid for purpose of calculating a ratemaking adjustment to offset the unreasonably high costs of pole replacements that result from SCE's failure to timely perform inspections from 1996 to 2002.

Based upon the foregoing, we will reduce SCE's capital forecast for pole replacements by $3.447 million (68,934 intrusive inspections that were funded by ratepayers but not performed by SCE times $50 per missed inspection). This disallowance is directly attributable to SCE's deficient or unreasonably deferred maintenance practices, i.e., its failure to perform intrusive inspections that it was funded and required to perform.

3.3.5. Reporting Requirement

TURN proposes that SCE be directed to report annually on the number of intrusive pole inspections completed in the prior year.

Discussion

This recommendation in effect asks that we reinstate the three-year reporting requirement that we imposed in D.96-01-011, Ordering Paragraph 15. However, in the intervening period the Commission has adopted GO 165, which governs inspection practices. We are hesitant to overlay GO 165 requirements with additional reporting requirements. In addition, SCE will need to address the status of its pole inspection program in its next general rate case. We are not persuaded that additional reporting is required in this GRC docket. This proposal will not be adopted.

3.4. T&D Capital

3.4.1. Introduction

SCE's T&D policy witness testified that the company's plan to invest more than $2 billion in T&D capital additions between 2001 and 2003 includes what he characterized as an aggressive capital replacement program to minimize the adverse consequences of SCE's aging system. SCE refers to this replacement program as the Infrastructure Replacement (IR) program. The IR program is budgeted for $621.938 million in 2001-2003, with wood distribution pole replacement accounting for $181.6 million of this total. The figure below depicts the wood pole "age bubble" that is addressed by the IR program.

3.4.2. ORA's Plant Recommendations

SCE forecast T&D plant additions of $614.399 million in 2001, $718.166 million in 2002, and $811.209 million in 2003, or a three-year total of $2.144 billion. ORA estimated additions of $456.728 million in 2001, $509.407 million in 2002, and $561.311 million in 2003, or a three-year total of $1.527 billion. SCE stipulated to a reduction of $157.761 million for 2001. Thus, ORA's three-year plant additions estimate is lower than that of SCE by $458.657 million ($208.759 million for 2002 and $249.898 million for 2003.)

SCE based its T&D plant additions forecast on a budget-based plan of expenditures. ORA compared SCE's forecast with recorded gross plant additions for 2001 and the first half of 2002, and determined that recorded additions were 25.7% less than forecast for 2001 and 35.3% less than forecast (on an annualized basis) for 2002. ORA concluded that SCE's capital budget was not an accurate basis for developing plant additions for 2001 and 2002, and should not be used to forecast plant additions for 2003. ORA believes that it is more appropriate to use actual, recorded plant additions to forecast additions for 2002 and 2003.

Discussion

ORA does not question the reasonableness of any particular T&D capital spending item associated with SCE's budget-based forecast. Instead, ORA questions whether the full amount of SCE's forecast capital spending for 2002 and 2003 will occur during that period. ORA relies on the fact that SCE's plant additions in 2001 and early 2002 were below the level of SCE's forecast to conclude that plant additions in late 2002 and 2003 will similarly be below budget. However, the evidence does not support this conclusion.

Even though T&D plant additions were lower than forecast for 2001 and 2002, that was an anomalous period. In 2001, SCE temporarily reduced capital spending because of the company's financial condition brought on by the energy crisis. This delayed many T&D projects that were nearing completion, which delayed the closing of those projects' work orders to plant in service. This in turn resulted in a $158 million shortfall in the 2001 T&D plant addition target. The problem persisted into early 2002 because of additional capital spending deferrals associated with the financial crisis. However, despite the additional deferrals, SCE was able to complete and record $723 million in actual T&D plant additions in 2002, exceeding the GRC forecast by $5 million.

SCE`s plant additions were lower than forecasted because the coordination and completion of paperwork involved in closing these capital expenditures to plant in service was delayed for several non-recurring reasons. Significantly, however, T&D capital expenditures during 2001-2002 were $2 million above the amount that SCE forecast in the GRC application. In other words, SCE has actually incurred the capital expenditures that it is requesting in this proceeding.

SCE has stipulated to the reduced plant additions recorded in 2001, and it has exceeded its 2002 target in plant additions. Apart from speculation that SCE accelerated closings to plant in 2002 in part because the GRC was pending, which speculation we discount, no reason has been shown why this trend will not continue into 2003. Since no issue as to the reasonableness of SCE's budget-based forecast of T&D plant additions has been raised (except for pole replacement costs), and SCE was on target to spend the budgeted amounts, we conclude that its T&D capital forecast is more reliable and should therefore be adopted along with the pole replacement cost adjustments discussed previously and in the following section.

3.4.3. Wood Pole Replacement Costs

SCE's T&D capital forecast includes estimated capital expenditures of $72.315 million in 2002 for 9,000 wood pole replacements and $76.509 million in 2003 for 9,300 replacements. Finding SCE's per pole replacement cost of $7,661 (year 2000 dollars, $8,228 in 2003 dollars) to be excessive, TURN recommends adoption of a substantially lower unit cost of $6,129 per replacement (2000 dollars), or 20% less than SCE's requested unit cost. TURN's proposal would result in disallowances of $14.454 million for 2002 and $15.264 million for 2003, or a total of $29.718 million. As an alternative to its proposed 20% reduction, TURN submits that SCE's forecast for wood pole replacements should be reduced by at least 16%.

TURN's analysis in support of its conclusion that SCE's pole replacement costs are unreasonably high is summarized below:

· SCE's forecast unit replacement cost for 2002 is $8,036, a figure derived by escalating the SCE forecast of $7,661 in 2000. Recorded data for 2002 show that the actual unit cost was $6,719 (Exhibit 384, weighted average), or 16.4% less than SCE's forecast. TURN submits that, at a minimum, SCE's own data substantiate the need for at least a 16% reduction in unit costs to $6,405 in year 2000 dollars.

· In the 1995 GRC, SCE used $2,400 (in 1992 dollars) as its unit replacement cost. Escalating this value by a non-labor escalation factor, the unit pole replacement cost in 2000 would be $2,916. Using the Handy-Whitman index for distribution poles in the Pacific region as an escalation factor results in a 2000 unit cost of $3,017. According to SCE, accounting changes adopted since the 1995 GRC require that indirect costs be added to the estimate used in that proceeding to yield a comparable value, and the comparable cost reflected in the 1995 GRC was $5,360 in 1992 dollars. TURN acknowledges the accounting change, but contends that this data still supports its primary recommendation because, according to TURN, a 1992 unit cost of $5,360 is equivalent to a cost of approximately $6,511 (2000 dollars).

· TURN contends that while both the 1995 GRC data and the 2002 recorded costs support a unit replacement cost between $6,400 and $6,500, a lower unit cost estimate of $6,129 is warranted due to anticipated cost savings. According to TURN, SCE's Work Management System (WMS) will reduce costs of field crew construction and maintenance. In addition, TURN expects a decrease in unit costs due to economies of scale, since SCE has doubled the number of annual pole replacements to over 10,000.

· In its 2004 GRC, PG&E stated that its unit replacement costs, before deducting the joint revenue credit, would be $6,724 in 2001 (nominal dollars), approximately 14% less than SCE's unit cost forecast of $7,850 (nominal) for 2001.

Discussion

TURN's conclusion that recorded 2002 unit costs were 16.4% below SCE's forecast for that year is not demonstrably based on a comparison of equivalent cost figures (i.e., an "apples to apples" comparison). In particular, the data upon which TURN relies to draw this conclusion do not appear to include adjustments for joint poles and other items that may be necessary to make a valid comparison of SCE's forecast and recorded costs for 2002.43 Thus, while the cost figures in Exhibit 384 may cast some doubt on the validity of SCE's forecast, in light of the comparability problem they do not provide a sufficient evidentiary basis for adopting a unit cost forecast of 16% below SCE's forecast.

TURN's conclusion that SCE's forecast pole replacement costs are excessive compared to the costs underlying the 1995 GRC relies upon on data from 1992, stated in 1992 dollars. SCE draws a different conclusion from the same data. Depending on the escalation method used, the year 2000 unit cost estimates (excluding additional work items) range from $6,511 (TURN) to $7,155 (SCE). Because TURN used a non-labor escalator and SCE used a labor escalator, but the underlying cost includes both labor and non-labor components, we believe the more reliable conclusion to be drawn from the 1992 data is that the reasonable unit replacement cost falls between these two estimates. In short, with respect to data from the 1995 GRC, SCE and TURN have cast doubt on each other's recommendations for pole costs in this GRC.

With respect to the comparison of SCE's and PG&E's unit costs, the evidence shows that what are loosely termed urban versus rural conditions can significantly impact pole replacement costs. A pole that is located in a congested urban setting (e.g., surrounded by buildings or structures that make access difficult) will be more costly to replace than a pole that is located in an open rural

environment where construction tends to be simpler, there is less equipment on the poles, and there are few problems with accessing structures. However, SCE's efforts to show that its rural-urban distribution of pole installations relative to that of PG&E explains most of the difference in the two companies' unit costs are neither convincing nor persuasive. SCE's evidentiary showing on this point can be accorded little weight. Accordingly, accepting SCE's estimated adjustment for "additional repair items" of $600 to arrive at a unit cost that is most comparable to PG&E's, the evidence shows that SCE's forecast cost of $7,061 ($7,661 less $600) is $526 greater than PG&E's forecast cost of $6,535.

We conclude that SCE has not sustained its burden of proving that a per unit pole replacement cost of $7,661 is reasonable. At the same time, there are enough problems with the evidence underlying TURN's proposed disallowance of 20%, and its alternative disallowance of 16%, to prompt us to reject those proposals as well.

Taking into account the potential for savings from economies of scale associated with a substantial increase in the size of the pole replacement program, it is reasonable to reduce SCE's authorized per unit cost to conform to PG&E's forecast cost. Accordingly, we reduce SCE's unit replacement cost of $7,661 by $526, and adopt a unit replacement cost of $7,135. This represents a reduction of 6.87% from SCE's requested capital expenditures of $72.315 million in 2002 and $76.509 million in 2003, or disallowances of $4.968 million for 2002 and $5.256 million for 2003 ($10.224 million for both years). This disallowance is in addition to the $3.447 million disallowance we have adopted in recognition of SCE's failure to perform timely intrusive inspections. (See Section 4.3.4.)

3.5. T&D O&M Expenses

3.5.1. ORA's Proposals

We have previously addressed TURN's recommendation to disallow approximately $1.7 million for intrusive pole inspection costs, adopting instead a reduction of $1.443 million. (See Section 4.3.4.) In this section, we address ORA's proposed reductions to SCE's T&D O&M forecast of approximately $7.6 million.

SCE largely relied on a budget-based approach to forecast O&M expenses for its T&D system. While accepting SCE's forecasts for several FERC accounts, ORA included the use of 2001 recorded FERC account data to arrive at its O&M forecasts for FERC Accounts 568, 571, 591, 594, and 598. For Account 582, ORA used a five-year average of data from 1996 to 2000.

Discussion

ORA's O&M recommendations that are based on the use of 2001 FERC account data are inconsistent with the intent of the Commission's Rate Case Plan, which provides for the use of a five-year recorded period (1996-2000 in this GRC). Moreover, the recorded FERC Form 1 data for 2001 are unadjusted, and therefore are not consistent with adjusted data for the 1996 to 2000 period that represents the base record period in this GRC. We note that SCE was under no obligation to provide ORA with adjusted FERC account data for 2001. In any event, even if it were otherwise appropriate to use FERC account data for 2001, the unusual conditions associated with the company's financial crisis render that year's recorded data less reliable for forecasting test year expenses, as the Commission acknowledged in D.02-04-055. Accordingly, ORA's proposals for FERC Accounts 568, 571, 591, 594, and 598 will not be accepted.

ORA has also raised concerns about the significant rise in costs in FERC Account 582 (Distribution Station Expense), in which SCE records costs for dispatching distribution operations and field operations crews, processing switching programs, directing switching, monitoring circuit loading, and responding to circuit interruptions, equipment trouble, and customer inquiries. SCE's forecast of $11.918 million for this account is based on recorded expenditures for the year 2000. SCE claims that averaging is inappropriate for this account because the cost increases incurred by SCE will continue into the test year.44 SCE contends that the main factors for the increasing costs were the addition of substation staff in 2000 needed to fill vacant apprentice operator positions and because of severe labor shortages caused by the Independent System Operator's demand for skilled distribution stations operators, retention bonuses were required to be paid. 45 This evidence was undisputed by ORA. We find that ORA's use of a five-year average for Account 582 is contrary to the facts in evidence and does not examine the reality behind the more recent recorded data. We accept SCE's forecast of $11.918 million for Account 582.

3.5.2. Jurisdictional Separation

Resolution E-3712 directed SCE to verify in its next GRC that certain transmission O&M costs included with its 1998 PBR operating expenses were distribution related and reasonable. Resolutions E-3771 and E-3772 made similar requests with regard to 1999 and 2000 transmission O&M expenses. SCE presented testimony in this GRC in response to these directives. No other party addressed this issue.

Discussion

SCE has fulfilled the relevant requirements of Resolutions E-3712, E-3771, and E-3772, and has demonstrated that certain subtransmission expenses discussed therein are subject to our jurisdiction and have been properly included in PBR revenue requirements.

3.6. Line and Service Extension Rules

Line and service extension costs associated with connecting new customers are categorized as distribution capital investments and recorded to plant in service. When an applicant selects SCE to perform line extension construction work, SCE estimates the total construction cost and calculates an allowance based on the revenues expected from the project load. SCE charges the applicant the difference between the allowance and the estimated costs.

TURN takes the position that ratepayer liability should never exceed the amount of the allowance, and that either the applicant or SCE should be liable for any cost overruns. TURN recommends that the Commission adopt the ratemaking principle that the revenue-based allowance is a cap to ratepayer liability for line extension project costs. TURN also requests that the Commission adopt rates subject to refund, order SCE to provide the necessary data to evaluate historical cost overruns, and subsequently modify rates based on a review of the data. TURN contends that, at a minimum, the Commission should order SCE to properly track its line extension costs going forward to allow an evaluation of cost overruns.

Discussion

The evidence does not support the claim that SCE has pursued practices in the administration of the line and service extension rules that are in contravention of its tariffs or Commission orders. No basis exists in this GRC for a ratemaking disallowance related to the reasonableness of SCE's T&D capital additions associated with line and distribution extensions.

In D.94-12-026 (58 CPUC 2d 1), the Commission modified the gas and electric utilities' main and distribution extension rules so that the allowances would be based on the expected revenues from the load served by the extension. In D.97-12-098 (77 CPUC 2d 785), the Commission again modified the line and service extension rules and practices for gas and electric utilities to reduce the amounts by which ratepayers subsidize the costs caused by new ratepayers requiring new line and service extensions. The Commission noted in D.97-12-098 that the adopted modifications would result in more uniform and consistent practices among the utilities.

TURN is now asking that a new rule be adopted, applicable only to SCE, to further advance the principle addressed by D.97-12-098, i.e., that subsidies by existing ratepayers to newly connected ratepayers should be reduced. We remain committed to the promotion of more uniform and consistent practices among the gas and electric utilities, and are not persuaded that it is necessary or appropriate to revise SCE's extension rules at this time by the adoption of a cap on ratepayer funding responsibility.

TURN's underlying concern is that without a cap, the utility lacks adequate incentives to control construction costs, and that ratepayers bear the risk of actual job costs being greater or less than estimated costs. The only SCE evidence on this point is the opinion of an SCE witness that making accurate construction cost estimates is a utility responsibility, and that cost overruns and cost underruns will balance out over the long run; and a compilation of a small sample of jobs showing an aggregate 5% overrun. However, SCE failed to provide documentation to support this opinion, despite having been asked to do so in a data request, and the sample compilation is not demonstrably reliable.

We place SCE on notice that, in the event that the Commission further considers the line and service extension rules, SCE cannot escape its responsibility to provide the Commission and parties documentation necessary to support such assertions. Based upon SCE's discovery responses in this proceeding, it appears that the company may need to modify its record keeping so that it is able to keep track of the data in a manner that would allow

calculation of total line extension job costs recorded to rate base, versus total estimated costs and total allowances for the same work orders.

3.7. Electric Transportation

3.7.1. Introduction

SCE proposes to continue its Electric Transportation (ET) Program activities in the 2003 GRC cycle. SCE states that its ET program allows it to implement its electro-drive programs and activities and continues its ability to comply with the federal Energy Policy Act of 1992 (Pub. L. 102-486, codified in 42 U.S.C.; hereinafter, EPAct) as well as state and local goals and programs. SCE forecasts $6 million in expenses (FERC Account 588 - Miscellaneous Distribution Expenses) for the test year, consisting of $1.875 million for labor and $4.125 million for non-labor. ORA and Aglet have proposed reductions to SCE's forecast, as discussed below.

Certain issues related to the Low Emission Vehicle (LEV) program were addressed in the consolidated LEV proceeding (A.02-03-048 et al.). By ruling entered in the LEV proceeding on June 26, 2002, later affirmed by the Commission in D.02-12-056, the Assigned Commissioner determined that certain "discretionary" utility activities would be addressed in the LEV proceeding and that "mandatory" activities listed below would be addressed in the utilities' respective GRC's:

· Acquisition of alternative fuel use fleet vehicles pursuant to EPAct.

· Operation and maintenance costs associated with use of alternative fuel use fleet vehicles and associated infrastructure.

· Infrastructure (fueling facilities and related equipment) needed to support alternative fuel use fleet vehicles.

· Employee training and instruction necessary for the use of alternative fuel use fleet vehicles.

· Accounting for the costs of these mandatory activities.

3.7.2. ORA's Recommendations

ORA recommends that discretionary LEV expenses of $677,000 not be considered in this GRC. ORA also recommends a disallowance of $463,000 of ET costs based on a belief that these costs were under review in the LEV proceeding.

Discussion

While discretionary LEV activities were reviewed in the LEV proceeding, the result of the review (as applicable to SCE) will be reflected in the revenue requirement adopted in this GRC. Thus, of necessity, discretionary LEV activities are "considered" herein. Moreover, a disallowance of $677,000 of discretionary LEV program expense is excessive given that the discretionary program costs under review in the LEV proceeding were $303,600. Our intent is to include in the adopted GRC revenue requirement the discretionary funding approved in the LEV proceeding.

ORA's proposed disallowance of $463,000 relates to electric trolley and high-speed electric rail, not LEVs. Because they relate to mass transportation, these programs were not part of the LEV proceeding. These are on-going ET program costs and were authorized in SCE's 1995 GRC.

3.7.3. Aglet's Recommendations

Aglet takes issue with the non-labor component of SCE's budget-based forecast of $4.125 million. Using recorded costs for 2000 as the basis for its forecast, Aglet recommends instead a forecast of $2.257 million. This represents a proposed disallowance of $1.868 million.

Nominally a dispute over forecasting methodology (budget-based v. last year recorded), this dispute largely pertains to the estimated costs of compliance with the EPAct requirement that utilities to purchase Alternative Fueled Vehicles (AFVs). SCE's fleet contains approximately 2,550 Light Duty Vehicles (LDVs) covered by the EPAct mandate, and in a normal year SCE may replace as many as 500 LDVs. SCE's GRC forecast assumes that it will purchase 500 LDVs in 2003, triggering an EPAct requirement to purchase 450 AFVs. However, SCE budgeted to acquire 200 AFVs, assuming that the rest would be made up with credits authorized under EPAct. In addition, SCE demonstrated that it must meet EPAct requirements in 2003 by having 90% of its fleet vehicles be AFVs. This requirement is 20% higher than for the base year 2000.46

Aglet faults SCE for assuming that compliance with EPAct requires the purchase of electric vehicles exclusively. Aglet asserts that while EPAct allows compliance through use of nine different types of alternative fuels, SCE has not considered pursuing cost estimates for any fuel type except electricity. SCE believes that its commitment to electric vehicles is reasonable because it provides electricity, not natural gas or propane, and it needs to "be smart" about electric vehicle technologies in order to provide customers with related information. SCE opposes compliance through use of natural gas or propane vehicles because it does not have the necessary infrastructure, and use of those technologies would ignore the investments made to support electric vehicles. SCE also opposes Aglet's proposal on the basis that it would be more costly to undertake because EPAct requires that SCE own the infrastructure to refuel its AFVs regardless of the fuel source.47

Aglet does not object to SCE's continued assessment of electric transportation technologies, but argues there is no convincing evidence that SCE must purchase 450 electric vehicles or compliance credits every year in order to "be smart." Aglet believes that a better strategy would be to purchase enough electric vehicles to perform reasonable technology studies and maintain an electric vehicle presence in SCE's fleet, and to pursue remaining EPAct compliance in ways that minimize ratepayer costs.

Discussion

SCE asserts that the purchase of propane or natural gas vehicles would cause significant costs of adding the necessary refueling infrastructure. SCE's witness testified that Aglet's proposal would in fact cost customers much more because it would mean having to build costly propane and natural gas refueling infrastructure in addition to the existing electric infrastructure.48

SCE's also relies on Commission findings made in 1995 to support its commitment to the exclusive use of electric vehicles. It is true that in 1995, the Commission found it "sensible and appropriate" for electric utilities to use electric vehicles to meet their purchase obligations under federal law while gaining an understanding of any safety and reliability implications from their use (D.95-11-035; 62 CPUC 2d 395, 456). We note however, that in this decision, the Commission found that utilities are under no obligation to buy either natural gas or electric vehicles. (Id.)

Although there is no legal obligation that SCE must buy electric vehicles, SCE has in this GRC justified its use of electric AFVs. SCE is an electric utility. This, combined with the EPAct requirement that SCE must own its refueling infrastructure regardless of fuel source provides us with a strong basis upon which to grant SCE's exclusive use of electric AFVs in this GRC. We do expect the utility to provide cost estimates in its next GRC, which show relative cost comparisons for the use of other AFVs besides electric. We require this in an effort to ensure ratepayers are receiving the lowest cost possible for SCE to comply with federal requirements. Aglet proposes to use recorded costs for 2000 as the basis for its forecast. SCE has put forth evidence to show that use of the last recorded year would understate its funding requirements for 2003. In 2000, automakers had limited production of electric vehicles, and SCE was only able to acquire 31 new electric vehicles. SCE therefore had to use a large part of its available EPAct credits to meet its 2000 EPAct requirements. SCE states that these credits will not be available in 2003. In addition, the EPAct AFV requirement was 70% of LDV purchases in 2000, whereas it is 90% in 2003. We will adopt SCE's budget-based forecast of $4.125 million, as it is the most reliable forecasting method supported by the totality of this record. Aglet's use of the recorded costs for 2000 would substantially underfund SCE in compliance with federal mandates.

4. Customer Service

4.1. Introduction

SCE's Customer Service Business Unit (CSBU) has 2,900 employees. Among its functions are meter reading; billing and payment services; call centers for inquiries regarding service establishment, billing and payment arrangements, tariff matters, conservation, load curtailment programs, and other issues; field services such as turn-ons, turn-offs, billing inquiries, investigations of meter accuracy and energy theft, and meter replacements; credit functions; Electric Service Provider (ESP) support; account management for non-residential customers; customer satisfaction research; responding to industry restructuring inquiries; and resolution of complaints.

Section 5 of this decision addresses SCE's requested O&M expenses for customer service functions, proposals for service fees, and CSBU related capital requirements. Also addressed are proposals regarding service guarantees and other customer service issues that do not have direct revenue requirement impacts in this GRC, including ORA's proposals regarding Authorized Payment Agencies (APAs) and LA County's proposals regarding metering, billing, and other customer matters.

4.2. O&M Expenses and Related Issues

4.2.1. Overview

SCE estimates that customer services will require O&M funding of $184.5 million (2000 dollars, excludes uncollectible expenses) for test year 2003. This request consists of $150.7 million for Customer Service Operations (FERC Accounts 901-905) and $33.8 million for the Customer Service and Information (CS&I) function, which addresses what SCE describes as the more complex needs of primarily non-residential customers (FERC Accounts 907, 908, and 916).

Noting that its requested funding level is $20.1 million greater than the 2000 recorded/adjusted level of $164.4 million, SCE's Senior Vice-President of Customer Service identified the following factors underlying its request that have caused the increase: (1) direct access costs would be recovered as an O&M expense for the first time in 2003; (2) Public Goods Charge (PGC) funding would no longer cover SCE's technology centers, the pump test program, and customer load control programs; (3) customer growth requires greater expenditures in almost all CSBU areas; (4) increased postage costs; and (5) regulatory mandates such as Real-Time Energy Metering (RTEM), new tiered-rate structures, and rotating outage notification procedures.

4.2.2. Customer Service Operations

SCE used multi-year averages to forecast expenses for four of 12 Customer Service Operations accounts or sub-accounts, and otherwise used the base-year method. As a matter of consistency, ORA advocates the use of the base-year amount where possible. Taking issue with SCE with respect to Account 901 (Business Unit Management Support) and Account 903.200 (Credit Services), ORA recommends rejection of the averaging methodology for those two accounts. SCE used three-year averages to estimate expenses for these accounts to avoid year-to-year variations and/or anomalies that may be present in the base year recorded costs. ORA recommends using the last year recorded method to forecast costs in these accounts because ORA finds that there has been very little variation from year to year in the 1998-2000 period. ORA's approach results in a proposed disallowance of $283,000.

Discussion

Use of the last year's recorded values would be appropriate if there were stronger evidence supporting the proposition that those values are more predictive of test year expenses than those for the other years. Here, if anything, the opposite is the case. For Account 901, recorded expenditures in 2000 were the lowest of the last three years due primarily to cash conservation measures implemented in the fourth quarter of 2000. For Account 903.200, SCE expects future collection costs to increase due to the less stable economy and higher electric rates. While this expectation of increasing collection costs is not fully supported by independent evidence, it is nevertheless based on professional judgment and is more reasonable than the unsupported assumption that year 2000 expenses are more predictive than the average of the years 1998-2000. SCE's averaging approach yields the more reliable forecast and is therefore approved.

ORA proposed a reduction to postage costs of $100,000 to reflect reduced postage rates proposed by the Postal Service. SCE generally agrees to the reduction, but contends that it should be corrected to $97,000. ORA agrees with the corrected reduction of $97,000, which we adopt.

SCE used the five-year average of uncollectible factors (1996-2000) of 0.319% for its test year uncollectible factor. SCE added to this factor 0.007% to reflect additional uncollectibles associated with its proposed residential late payment charge (LPC) and increased Field Assignment Charge (discussed below), for a total factor of 0.326%.

ORA proposes using the last recorded year's factor of 0.311% to estimate future uncollectible expenses. This methodological difference yields an O&M expense difference of $539,000. Aglet proposes using SCE's recorded 1995 uncollectible factor of 0.264%, yielding a difference of $1.648 million.

The parties agree that to the extent that SCE's proposed LPC and its proposed increase to the Field Assignment Charge are rejected, the 0.007% adder associated with those charges should be adjusted or eliminated accordingly.

Discussion

SCE's five-year average (0.319%) is based on the historical uncollectible data set forth in the following table.

Uncollectible Factors

ORA disagreed with SCE's use of a five-year average because, using a process known as POSID, SCE has improved its assessment of the creditworthiness of new service applicants. ORA believed that the use of older data in a five-year average would exclude likely reductions in uncollectibles due to the new procedures. However, the five-year average already includes the effects of the POSID process, which was implemented prior to 1997. ORA is also concerned that SCE's uncollectibles rate rose from 1996 to 1999 even as unemployment and poverty rates in SCE's service territory declined. ORA faults SCE for failing to explain this increase.49

We take a different view regarding this data, i.e., unemployment and poverty rates simply may not be as reliable as predictors of SCE's uncollectible factor as SCE and the other parties seem to assume. We are also not persuaded that corporate bankruptcies are as explanatory as SCE assumes. We are left with the data shown in the table above and non-compelling explanations for any trends that may be occurring. In the absence of more compelling evidence that the uncollectible rate from any single year (whether 1995 as Aglet recommends or 2000 as ORA recommends) is more predictive than an average of several years, SCE's proposal to use a five-year average is more reasonable and will be adopted. As we are adopting SCE's proposed LPC, but not its proposed Field Assignment Charge, we adopt an adder of 0.005% instead of 0.007%.

The majority of Edison's customers (72%) pay their bills by mail. Another 20.4% of customers pay their bills in person at SCE's Local Business Offices (LBOs) or, increasingly, at one of SCE's 369 APAs. These are independent businesses that provide bill payment services for SCE in exchange for a transaction fee.

ORA found that some customers who pay their bills in person have complained about poor service. ORA recommends that SCE improve APA services and expand the APA network where needed. Specifically, ORA recommends that APAs be:

· Located a reasonable distance from the customer's home or office, recognizing the differences among urban, suburban and rural locations.

· Located in retail centers that provide parking, access via public transportation, safety and convenience.

· Located in businesses that are clean and pleasant, and that process payments professionally, efficiently and courteously.

· Located in businesses that process payments from Edison as well as other companies to provide one-stop service for customers.

· Located in businesses that are open for business six days per week or more and open extended hours each day (e.g., 8:00 a.m. to 9:00 p.m.).

In addition, ORA recommends that SCE undertake an education program to migrate customers away from in-person payments to lower cost payment options. Finally, ORA recommends that SCE evaluate the feasibility of replacing or supplementing APAs and LBOs with unmanned payment processing stations.

SCE objects to ORA's characterization of LBO and APA customer issues and believes that its customers are reasonably satisfied with the bill payment options offered by the company.

Discussion

While it contends that SCE should improve the level of service provided to customers who use APAs and LBOs to pay their bills, ORA agrees that, overall, SCE is performing satisfactorily in this area. Moreover, despite their differences regarding the nature and extent of customer dissatisfaction and how to interpret customer satisfaction surveys and focus group results, there appears to be little actual disagreement between SCE and ORA regarding ORA's specific recommendations for the APA program. In effect, except for ORA's recommendation to study the feasibility of unmanned payment processing stations, to which it did not respond, SCE accepts those recommendations. SCE has already established criteria for locating and establishing APAs that essentially include ORA's recommended criteria regarding accessibility to transportation, convenience, etc. We endorse these criteria, and expect SCE to continue to ensure that they are implemented. Similarly, SCE in effect agrees with ORA's recommendation for educating customers on the availability of lower cost payment options by noting that it has undertaken such an information campaign.

We do not find it necessary to undertake additional regulatory efforts at this time to improve SCE's APA program. We do find ORA's recommendation for an evaluation of the feasibility of replacing or supplementing the APA/LBO network with automated payment processing intriguing. If cost-effective, this payment option may provide a higher level of customer service and flexibility. SCE should report on its efforts to evaluate the feasibility of such options in its next GRC filing.

The volume of SCE's Internet communications has been expanding in recent years. Within Account 901, SCE requests $681,000 in additional expenses for Internet content maintenance, including $310,000 for personnel and $321,000 for technical development.

Aglet is concerned that SCE made little or no effort to determine the effect of growing Internet use on costs for other forms of communication. SCE's website has been open for public use since January 1995, and it allows customers to establish new service, and transfer or close service through on-line

transactions. Aglet believes that those who use the Internet to gain information from or about SCE will, to some extent, forego the use of other forms of communications such as telephone calls to SCE's customer service center and written communications. Aglet notes that the volume of customer calls has been declining. Aglet witness Weil has used SCE's website to review or download tariff sheets and financial information, which would otherwise have required him to write or call SCE. According to ORA witness Kinser, SCE customers have indicated that Internet use is more convenient than telephone use or visiting utility offices.

Aglet contends that in requesting an additional $681,000 in ratepayer funding to maintain and operate its website, SCE has not met its burden of showing that improved Internet capability will not affect other customer information revenue requirements. Aglet also contends that SCE has not shown that the new Internet expenses will be cost-effective. Aglet therefore recommends a disallowance of 50% of the new program expenses, or $340,000, to compensate ratepayers for reductions in other customer communications costs caused by the Internet expansion, and for the uncertainty of program benefits.

SCE asserts that increased Internet expenditures will become cost effective in future years as more customers use the Internet.

Discussion

SCE's contention that cost savings benefits from its investment in its Internet capabilities will not be realized until a future rate case lacks merit. The evidence shows that customers now have the ability to use the internet to interact with SCE, and that to some extent they do so instead of using other forms of communication. Accordingly, and since SCE has not adequately studied the issue, Aglet's recommended disallowance of $340,000 is reasonable under the circumstances and will therefore be adopted.

Pursuant to D.99-09-064, SCE's customer service O&M forecast includes a forecast of direct access customer costs. SCE's forecast of $3.799 million is based on the assumption that the number of direct access customers will range from 30,000 to 70,000. This represents a reduction from the recorded direct access costs of $7.331 million in 2000 and $4.474 million in 2001, but an increase over the recorded cost of $3.690 million in 2002.

TURN contends that because direct access has been closed to new customers since September 2001, and SCE serves 42,521 direct access customers, or 40% less than the maximum number assumed by SCE, there is no rationale for an increase in direct access costs in 2003. TURN submits that, barring new legislation, the only foreseeable likelihood is that the number of direct access customers will decrease. TURN therefore recommends one-way balancing account treatment for SCE's entire $3.8 million estimate of direct access costs. In the alternative, TURN recommends a 20% reduction to SCE's estimate.

Discussion

SCE's request rests on the proposition that its direct access related customer expenses are essentially fixed at $3.8 million if the number of direct access customers remains within the 30,000 to 70,000 range. SCE states that these are ongoing costs that have stabilized over the last two years. While we accept SCE's showing regarding the relatively fixed nature of these costs, SCE has not adequately demonstrated why its test year expense would be higher than the 2002 level. We will adopt a forecast of $3.690 million based on the 2002 recorded level. TURN's request for one-way balancing account treatment is not justified, as it is just as conceivable that SCE's estimate is too low as it is too high.

4.2.3. Customer Service & Information

SCE has two technology centers that provide customers with information on electromagnetic fields, metering, energy management, energy efficiency, and environmental solutions. These are the Customer Technology Application Center located in Irwindale and the Agricultural Technology Application Center located in Tulare. SCE also has a pump test program that serves agricultural customers, water agencies, municipalities, and other pumping customers. SCE has funded these activities in whole (pump test) or in part (technology centers) with Public Goods Charge (PGC) money.

With an expectation that PCG funding would be eliminated, SCE proposed to add $2.531 million in O&M funding for these activities. In 2000, SCE recorded $1.513 million in O&M funding and $2.45 million in PGC funding for the two technology centers. For the 2003 test year, SCE seeks to add an additional $1.272 million for a total O&M forecast of $2.785 million for the technology centers. In 2000, the pump test program was funded by $1.27 million in PGC money. For 2003, SCE seeks to fund the pump test program with $1.259 million in O&M money.

ORA opposes SCE's request for an additional $2.531 million in O&M funding for these activities. ORA takes the position that with the enactment of Assembly Bill 995 in 2000 (Stats. 2000, Chapter 1051), PGC funding will be available through the year 2011.

Discussion

This dispute does not address the reasonableness or effectiveness of the technology centers or the pump test program. It pertains to the appropriate funding mechanism, i.e., whether these functions should now be fully funded through base rates by their inclusion in O&M expenses, or whether they should remain part of the PGC funding process. Apparently acknowledging that PGC funding will not be eliminated any time soon, SCE nevertheless believes that some of the functions provide basic customer care, and it prefers the more stable funding of the GRC mechanism to the less certain funding of the PCG process. However, SCE's preference for more stable funding for the technology centers and the pump test program does not justify a shift in funding at this time, and SCE's basic customer care argument is unconvincing. Moreover, as ORA points out, SCE has not adequately explained why these of all programs that are now covered by the PGC should be singled out for O&M funding. As to SCE's concern about the future administration of these programs, we note that if, in the future, the Commission considers the selection of other administrators for these functions, SCE will have an opportunity to participate in relevant proceedings. At best, SCE's proposal is premature. Accordingly, SCE's request to fund the technology centers and the pump test program through an O&M allowance in this GRC is denied.

Established in 1983, the Air Conditioner Cycling Program provides load relief during periods of high demand in return for a rate discount. D.01-04-006 reopened the program to new participants and ordered program enhancements. SCE estimates that participation will increase to 170,000 customers by the end of the test year. The majority of the additional customers will require the installation of cycling devices, and SCE's O&M request includes material costs of $3 million (Account 908) for their acquisition.

ORA wishes to encourage this program but it does not believe that SCE's projected enrollment growth is borne out by the data. Participation was 114,002 in October 2001 and it dropped to 110,624 by July 2002. ORA recommends that SCE's annual estimate be reduced from $3 million for 19,900 cycling devices per year to $1 million

SCE responds that a $2 million disallowance for devices will limit it to installing or replacing only about 6,700 devices per year. SCE contends that this is inadequate to meet expected demand and replace existing defective devices in the coming years. While SCE believes that ORA has underestimated the potential number of enrollees, it acknowledges ORA's concern, and therefore proposes to establish a one-way balancing account that would reconcile differences in the actual cost incurred for new or replace air conditioning cycling devices and the allowed revenue requirement in this case.

Discussion

Forecasting enrollment levels in this program is particularly difficult at this time. On the one hand, ORA's estimate of annual growth of 5,000 customers per year reflects the first year after the cycling programs were authorized, which may not be an accurate reflection of test year activity since the cycling programs had been scheduled to terminate on December 31, 2002 and potential customers were just coming off a year of rotating outages. Further, SCE has stepped up its marketing activities this winter season and it expects that the current rate of enrollment will increase in 2003 and beyond. On the other hand, SCE's expectations notwithstanding, the actual enrollment data do not support a need for 19,900 new cycling devices in 2003.

In order to promote this potentially valuable load control program, we will approve SCE's forecast subject to a one-way balancing account as proposed by SCE. In addition to maximizing the potential benefits of this program, this approach will ensure that ratepayers provide funding only for cycling devices acquired and installed by SCE.

TURN initially recommended that forecast costs for all SCE's load-control programs receive balancing account treatment, not just the air conditioning cycling program. SCE opposed this proposal, and TURN did not address it on brief.

Discussion

In contrast to the costs for air conditioning cycling devices, administrative costs for the cycling programs and for all the other load-control programs are relatively stable. A large change in customer participation numbers has only a modest impact on administrative costs. We find insufficient justification for balancing account treatment for load-control programs except for air conditioning cycling devices, and therefore deny TURN's recommendation.

Section 740.4 directs the Commission to authorize utilities to engage in programs to encourage economic development (subdivision (a)) and to allow recovery of reasonable expenses for such programs to the extent of ratepayer benefit (subdivision (b)). Subdivision (h) of Section 740.4 states the Legislature's intent that rate recovery shall be allowed "to the extent the utility incurring or proposing to incur those expenses and rate discounts demonstrates that the ratepayers of the public utility will derive a benefit from those programs."

SCE's Director of the Commercial and Commodities Segment in the Major Customer Division of the CSBU testified that current economic conditions in California "reinforce the fundamental need for the State of California to sustain an integrated economic development function and a vital role for SCE to maintain active and aggressive participation in that process." (Exhibit 39, p. 64.) As a consequence, SCE requests $2.514 million in O&M funding (Account 908, $1.467 million labor and $1.047 million non-labor) for Economic and Business Development (E&BD) activities. SCE states that these activities serve to retain, expand, and attract business facilities and operations that would otherwise locate outside of California. SCE's witness testified that the primary customer target class is manufacturers, and that while the program can help all customers, SCE has "more likelihood of helping a large manufacturing customer." (Tr. V. 17, p. 1228.) SCE takes the position that all customer rates are reduced by maintaining and/or increasing revenues to cover fixed costs.

The Commission has applied the Rate Impact Measure (RIM) test for assessing the cost-effectiveness of demand-side management (DSM) programs. It has described the RIM test as follows:

The RIM test measures what happens to customer bills or rates due to changes in utility revenues and operating costs caused by the program. The benefits calculated in the RIM test are the savings from avoided supply costs (to which non-price factors would apply). The costs for this test are the program costs incurred by the utility, the incentives paid to the participant, decreased revenues for any periods in which load has been decreased and increased supply costs for any period when load has been increased." (D.95-06-016, 60 CPUC 2d 265, 293, footnote 5.)

Adapting the RIM test to an analysis of E&BD costs and benefits, SCE analyzed the difference between revenues and marginal costs for the kilowatt-hours affected by the E&BD program. SCE determined that the E&BD program yields a benefit/cost ratio of 1.26 using rates in effect during 2000 and assuming continuation of the PBR mechanism. SCE also calculated a RIM test benefit/cost ratio of 2.11 under the assumption of a revenue adjustment balancing account ratemaking mechanism. SCE contends that this provides the necessary justification for its E&BD funding request.

Aglet contends that SCE incorrectly applied the RIM test and failed to demonstrate ratepayer benefits of E&BD activities. Aglet proposes to eliminate all ratepayer funding for this function, resulting in a disallowance of $2.514 million. Aglet's concerns with SCE's RIM test analysis include the following:

· Aglet contends that SCE failed to measure "what happens to customer bills or rates due to changes in utility revenues and operating costs" caused by the E&BD program. In its initial showing, SCE calculated only the effects of proposed E&BD activities on electric distribution, transmission, nuclear decommissioning and public purpose program costs and revenues, and excluded generation related revenues and costs from the RIM test. Aglet contends that this exclusion is internally inconsistent and wrong. Aglet claims that SCE ignored the effects of E&BD efforts on generation and procurement needs, and that SCE does not know how future generation costs would affect the cost-effectiveness E&BD activities.

· Rebutting Aglet's position regarding the exclusion of generation from the RIM test, SCE provided RIM test results from a DSM report that included "Commission-specified avoided generation costs." Based on this DSM report, SCE asserts that the E&BD programs reduce rates because benefit/cost ratios exceeded one. Aglet responds that those avoided generation costs are not included in the record, and they appear to have been based on forecasts made prior to the years 1995-2001. Aglet is skeptical of their

· applicability here, as SCE's actual avoided costs during the 2000-2001 financial crisis were much higher than average rates. Aglet notes that SCE has presented no information on test year 2003 incremental generation and procurement costs.

· Aglet contends that SCE did not complete all of the necessary elements of the RIM test. Pursuant to D.95-06-016, SCE as the program proponent must also identify effects on core customers of programs that increase load in noncore markets and identify expected program benefits in terms of rate effects, resource planning effects, and other effects. Aglet submits that SCE has not shown how load increases caused by the E&BD program will affect overall customer rate levels or resource planning.

· If incremental supply costs are higher than average rates, then incremental load increases will raise average rates. SCE agrees that this principle applies to business expansion programs, but disputes Aglet's position that it also applies to business retention programs. Aglet responds that retention programs are equivalent to attraction programs if one uses the correct definition of incremental costs. Aglet submits that since incremental supply costs are likely higher than average costs, at least at times, there will almost certainly be periods where the load building resulting from E&BD activities will cause higher rates for SCE customers.

· Aglet contends that there is no evidentiary support for SCE's assumptions that program participants are influenced by the specific percentages upon which SCE relies. For example, there is no rigorous showing that up to 90% of increased loads installed by participants in SCE's programs have been or will be attributable to the programs.

Discussion

Pursuant to Section 740.4, we should approve the requested E&BD funding to the extent that SCE has demonstrated that ratepayers will benefit from the activities funded through the request. On the other hand, to the extent that SCE has failed to make such a demonstration, we must deny the funding request. As explained below, we find that SCE has established that its RIM test analysis provides a demonstration of ratepayer benefit in one aspect, and, therefore, we are bound to approve its E&BD funding request.

As a preliminary matter, we note that Aglet accepts SCE's position that the RIM test for DSM programs is an acceptable means of testing for ratepayer benefits of the E&BD program. We will therefore focus on the disputes over SCE's adaptation and implementation of the RIM test. Also, SCE points out that Aglet did not perform a RIM test calculation or recalculate the impact of Aglet's recommendations on the RIM test. However, the burden of proof rests with SCE to demonstrate the validity of its RIM test analysis. Aglet is not the proponent of SCE's E&BD funding, and it is not obligated to perform a RIM test.

Although we approve SCE's request for funding its E&BD costs because of the demonstrated ratepayer benefit due to the use of the RIM test, we order the utility to use this test as one part of the ratepayer benefit demonstration for this program. The benefits calculated in the RIM tests are the savings from avoided supply costs. Avoided supply costs alone cannot capture the potential ratepayer benefit in a program designed to promote and retain business in California. Section 740.4 directs the Commission to authorize utilities to engage in programs to encourage economic development (subdivision (a)). In addition to the RIM test, SCE should supplement the benefits showing of this program in its next GRC with recorded examples of how this program has not only encouraged economic benefit in the state of California, but also examples of how this

program has resulted in the actual retention, expansion or attraction of businesses in California. Basically, SCE must be able to demonstrate that but for this program, business in its service territory would have left the state or not moved into the state. We also note that SCE has recently filed a rate discount proposal with this Commission designed to help retain or expand existing Southern California businesses and to attract new businesses. If this proposal is approved by this Commission, SCE should report back to the Energy Division on how it is coordinating those efforts with its E&BD program.

LA County has more than 2,500 separately metered accounts with SCE, making it the company's second-largest customer. It asks that the Commission require SCE to be more responsive to the needs of large, multi-account customers. Section 5.2.3.5 addresses three specific LA County requests.

LA County seeks to establish a mechanism to obtain metered consumption data for its time-of-use accounts and billing data for all of its accounts in a manner that would allow it to better manage its energy consumption and expenditures. Under the existing program offerings provided by SCE to large multi-account customers, it would have cost the County anywhere from $100,000 to $687,500 a month to obtain the information it needs in a useful format. This expense is not acceptable to the county.

SCE committed in this proceeding to provide the County with billing data through a modified version of SCE's Electronic Data Interchange (EDI) offering, which was previously only available to customers who were able to pay their bills electronically. Under this commitment, LA County will continue to be able to pay its bills by paper, and SCE will provide all of the bill information currently provided on the paper bill to a value added network (VAN) with whom the County will contract for the purpose of obtaining the information. SCE estimates that it will cost the County approximately 32 cents per month per bill to obtain the data. LA County will purchase the necessary hardware and software to convert the data into a format consistent with the energy management software the County has purchased. LA County welcomes and requests that the Commission endorse SCE's commitment to provide a modified EDI service through a VAN in addition to the paper billing SCE now provides.

Discussion

SCE's existing customer offerings to provide consumption data and billing data are not well suited to customers such as the County with many accounts. SCE and LA County have worked cooperatively to achieve a solution that meets the County's needs and is workable for the utility. We endorse the results of their efforts, with the understanding that SCE will make the offer available to all similarly situated customers.

SCE makes third-party financing available to customers for meters and other equipment, including energy efficiency equipment. LA County is concerned that SCE's financing program is not designed for local governments to use, and it notes that no municipality has taken advantage of it. Because financing is critical to the ability of local governments to invest in energy efficiency, the County requests that the Commission order SCE to modify its third party financing program for energy efficiency investments in a manner that will enable the participation of municipal governments.

Discussion

The record does not support an order directing SCE to make particular modifications to its third-party financing program at this time. However, SCE states that it is working with the County to better understand the County's borrowing limitations, and that it is fully prepared to work with the County and participating lenders to determine how the County's needs can be addressed through a third-party lending program. LA County has suggested that we order a workshop so that SCE can obtain the input it requires to tailor such a program to the needs of local governments. In light of SCE's stated willingness to work with the County, we direct SCE to convene a workshop as proposed by the County, not later than 90 days after the effective date of this decision. SCE should provide notice and an opportunity for municipalities in its service territory to participate. Our objective in making this directive is to explore whether and how SCE's third-party financing can be made effectively available to local governments so that they may more aggressively pursue cost effective DSM projects.

LA County states that it is overwhelmed by the number of proceedings before this Commission that could significantly impact its energy budget. The County needs to constantly prioritize and allocate its resources in order to ensure that it participates in proceedings that have the greatest impact on the County as a ratepayer. The County requests that we order SCE to provide a free-standing, easily identifiable "ratepayer impact analysis" on its website in conjunction with any proposal made by SCE in a Commission proceeding that could increase rates for any single rate class by 5% or more. The County had initially requested that SCE provide additional information such as summaries of the positions of other parties, but has withdrawn those proposals.

SCE opposes this request, contending that such detailed information would be useful to only a small number of customers; that it already provides ongoing information to customers about regulatory, legislative, and marketplace developments; and that providing the information required by the County would require sizable staffing additions.

Discussion

Even though some ratepayers would undoubtedly benefit from a requirement that SCE provide a ratepayer impact analysis as recommended by the County, we must weigh the costs as well as benefits of requiring such information to be produced. The record is clear that requiring SCE to provide the information that LA County initially sought would have been unduly burdensome, although it is somewhat less clear that a requirement limited to the provision of ratepayer impact analyses would be as burdensome to SCE. However, we remain concerned that the benefits of such a proposal may not justify the costs. SCE already provides a bill insert to customers describing the impacts of its rate proposals by customer class, and we can only presume that LA County is asking that SCE provide significantly greater detail than bill inserts provide.

We will not, by today's decision, order SCE to provide more detailed rate impact information than it is currently required to provide to customers. Nevertheless, we urge SCE to explore cost-effective means of providing customers with more meaningful information about the impacts of its rate proposals on the various rate classes. SCE should implement those means as appropriate. In addition, we will continue our existing practice whereby Commissioners and ALJs assigned to individual proceedings may, as they deem appropriate, determine that additional notice to customers about rate proposals and their impacts may be needed.

We will implement one aspect of LA County's proposal. Whenever SCE files a rate proposal before this Commission that requires bill insert or similar notice to customers, SCE should ensure that the content of the notice is also posted on its website in a manner accessible to customers. While we decline to specify the detailed manner in which this should be accomplished, it should be possible for those who access the website to easily locate such notices for all open proceedings for which the requirement is applicable.

4.3. Service Fees and Other Operating Revenues

4.3.1. Ratemaking Policy Considerations

Although Phase 2 of this GRC addresses most issues regarding the utility's rate and tariff structure, i.e., how the utility's authorized revenue requirement is to be collected from the various classes and categories of customers, SCE, ORA, and TURN have introduced and responded to proposals for various service fees in this phase of the proceeding. Other Operating Revenue (OOR), which includes the revenue for such fees, is credited to the cost of service. Therefore, OOR reduces on a dollar for dollar basis the revenues that must be collected through base rates. To the extent that our adopted service fees are lower than requested by SCE, and produce less revenue than SCE forecast, SCE's base revenue requirement would be increased accordingly.

SCE proposes increases in service fees for returned checks, service establishment, and disconnects and reconnects for nonpayment. In addition, SCE proposes to institute a residential late payment charge (LPC). SCE states that these proposed fees would result in an increase in OOR of approximately $21 million.

ORA and TURN oppose SCE's proposed LPC due to its impact on low-income families, a concern echoed by Greenlining. ORA proposes more moderate increases to the various fees than those proposed by SCE. TURN proposes elimination of the service establishment fee and initiation of a direct access customer charge. We will address these proposals in the remainder of Section 5.3. We first address overarching policy for the various proposals.

Discussion

SCE takes the position that, where possible, the cost of services should be assessed directly on those customers responsible for those costs. SCE maintains that doing so sends appropriate price signals and serves to reduce rates for all customers. SCE's proposals for service fees are offered to further this objective.

To the extent that an identifiable service is provided to a specific customer without charge or at a charge that is below the incremental cost of providing the service, the general body of ratepayers subsidizes the customer receiving the service. Moreover, if a cost-based service charge induces customer behavior (for example, timely bill payment) that results in lower overall utility costs, then both the utility and ratepayers alike may benefit from increased efficiency.

However, we must balance costs with a host of other ratemaking considerations. In D.96-01-011, 64 CPUC 2d 241, 287, when considering a proposed service connection fee, the Commission reiterated that:

[C]ost is not the sole factor relevant to our inquiry. As we stated in D.91-12-075, 42 CPUC 2d 566, 591-592:

"Our past decisions have never held that just and reasonable rates, the statutory standard (PU Code §§ 451 and 728), had only one component - costs. We have always held that factors such as conservation, affordability, market price and equity had to be factored into the rates. Cases which most strongly supported cost-based rates invariably tempered those statements with language which showed our concern for other ratemaking factors...

"A reading of the PU Code leaves no doubt that the Commission must look beyond costs when setting rates. ... There is nothing in the Code which equates cost-based rates as being a synonym for just and reasonable rates, or as the sole standard by which rates are considered just and reasonable."

In the same decision, in adopting a fee at a lower level than had been proposed by SCE, the Commission took note of the fact that an increased charge would disproportionately impact low-income customers. (Id., 288.)

In D.96-04-050 (65 CPUC 2d 362, 436-440), the Commission ordered the phased elimination of rate discounts for SCE's employees and retirees. The phase-out was suspended when Section 368(a) was adopted, and in Phase 2 of this proceeding SCE proposes to continue the 25% discount unless the Commission terminates employee/retiree discounts for other utilities. (SCE's Phase 2 testimony, Exhibit SCE-16.) We will address this proposal in Phase 2. At this time, we merely note that SCE apparently agrees that cost considerations are not exclusive when setting tariff structures.

4.3.2. Late Payment Charge

SCE proposes the institution of a Residential Late Payment Charge (LPC) of 0.9% per month for customers who fail to pay their bill within 19 days of its receipt. The Commission denied a similar proposal by SCE in the 1995 GRC, while stating that it would reconsider the matter when the Southern California economy significantly improved and SCE's then new computer system, which can accurately track the payment receipt date, became operational. (D.96-01-011, 64 CPUC 2d 241, 289.) SCE claims that both of these conditions now exist, as evidenced by the decline in unemployment from 1995 to 2001 and by the fact that the CSS is now fully operational and virtually 100% of payments are posted by the next day after their receipt.

ORA opposes the LPC based on its concern that the residential LPC will have a disproportionate impact on low-income customers. ORA believes that an LPC will compound the problems of those customers whose payments are late owing to inability to pay. ORA also believes that some customers who face a late payment charge will contact the company to dispute the charge, adding to customer service costs. Moreover, ORA claims, the LPC is inefficient due to the relatively high uncollectible rate estimated by SCE (5%) compared to the overall uncollectible rate (0.311%) recorded in 2000. Finally, ORA believes the LPC is particularly inappropriate at this time due to the recent rate increases to cover higher generation costs.

Joining ORA in opposition to the LPC, TURN submits that economic indicators are declining and that instituting such a charge will create an economic hardship. Referring to data provided by SCE, TURN notes that unemployment rates for the SCE service territory were 8.9%, 7.9%, 7.6%, 6.5%, 6.1%, 4.8%, 4.5%, 5.5%, 6.1%, and 6.0 % for the 10 years from 1994 to 2003, respectively (forecast data for 2002 and 2003). Moreover, TURN contends, U.S. unemployment data for November 2002 suggest that SCE's data may be out of date. If the Commission adopts a residential LPC, TURN proposes that it be limited to bills under $40, and that it be waived if the customer is put on a payment plan.

Discussion

SCE's proposed residential LPC upholds the basic principle that those customers who pay late should be responsible for the costs associated with late payments. Moreover, the Commission's stated conditions in D.96-01-011 for consideration of an LPC have been met. No party contends that SCE's CSS is incapable of accurately tracking and recognizing payment receipt dates. TURN disputes SCE's contention that economic conditions, as measured by unemployment in SCE's service territory, have improved sufficiently, but we note that at worst, unemployment was forecast to rise to 6.1% and 6.0% in 2002 and 2003, respectively. These rates are significantly below the 1994 and 1995 rates of 8.9% and 7.9%, respectively, that immediately preceded the issuance of D.96-01-011. The 6.0% forecast for 2003 is also significantly below the 10-year average of 7.1%. TURN's reference to late 2002 national employment data does nothing to change this conclusion.

The parties' concern for the impact of a residential LPC on low-income customers is not without merit. However, the evidence shows that the financial impact on individual low-income customers would not be significant. In any event, SCE has agreed to exempt CARE customers from the application of the LPC, a proposal that we approve. This is consistent with D.95-12-055, where a residential LPC was approved for PG&E with a CARE customer exemption.

ORA is concerned that a new residential LPC could result in increased customer calls, and increased costs. However, there is no study or analysis to support this opinion, and it is just as likely that the residential LPC could reduce the number of calls relating to credit extensions and disconnect notices, since more people will have an incentive to pay on time.

SCE has shown that the Commission's conditions for consideration of a residential LPC have been met, that the LPC will promote cost causation principles, and that low-income customers will not be unduly impacted. SCE's proposal is therefore adopted. TURN's alternative proposal to limit application of the residential LPC to bills under $40 and to waive the charge if the customer is put on a payment plan will not be approved due to the costs and complications involved in setting up and maintaining the processes and systems needed to implement these exceptions.

SCE proposes an interim balancing account for treatment of LPC revenue, as it expects the revenue to vary significantly over the next two years. We will approve this uncontested proposal.

4.3.3. Service Establishment Charge

TURN recommends that the service establishment charge for customers requesting a new service be eliminated due to its disproportionate impact on

low-income ratepayers and renters, who move more often than homeowners and the more affluent. In the Los Angeles-Long Beach Standard Metropolitan Statistical Area (SMSA), 28% of renters (including 39% of renters below the poverty line) moved in the past year, while only 7% of homeowners moved in the past year. For the Riverside-San Bernardino SMSA the data showed that 47% of renters (including 54% of renters below the poverty line) moved in the past year, compared to 9% of homeowners. Because renters have lower average incomes than homeowners, the percentage of households moving in the past year is strongly related to income.

TURN maintains that the existence of this charge does not reduce the utility's costs because a customer's decision to move is generally not affected by the charge. TURN contrasts this to a late payment and reconnection charges, which may influence customer behavior and thereby reduce costs.

SCE accepts most of TURN's evidence and analysis but it does not concede these are reasons to abolish the charge. SCE asserts that not all people who move often are poor or less fortunate, and that those who are have other remedies to make electricity available to them at a reduced rate. SCE maintains that costs related to new service establishment are directly attributable to customers moving into a new business or home, and that these costs should be borne by those who cause them to be incurred.

Discussion

Even though low-income ratepayers and renters move more often and therefore incur the service establishment charge more often than other ratepayers, that does not establish conclusively that low-income ratepayers and renters are unduly impacted by the charge. Clearly, low-income ratepayers and those who rent would benefit from waiver or elimination of the charge, but the evidence does not persuade us to weigh that benefit more heavily than our concern for establishing service charges on principles of cost causation. Moreover, as SCE points out, these customers have other remedies to make electricity available to them at a reduced rate, including the CARE discount. Further, there are customers who attempt to avoid paying their bills by reapplying for service under a new or assumed name, leaving an unpaid balance on the former account, which would eventually go to write-off. While this is not a major factor (for the reasons explained by TURN), neither should we ignore it. Finally, as explained below, we are adopting the more moderate service charge increases advocated by ORA instead of those proposed by SCE, which will mitigate somewhat the impact of this charge on low-income ratepayers and renters. For all of these reasons, TURN's proposal to abolish the service establishment charge is denied.

4.3.4. Direct Access Customer Charge

TURN recommends that a $5 per month customer fee be charged to all direct access customers over 20 kW on an interim basis, with the intent to further evaluate and design direct access service fees that are more cost based in the second phase of this proceeding. TURN states that this would generate $378,420 in OOR. Asserting that there are many unsettled aspects of direct access service that should be resolved first, SCE contends that an interim $5 fee should not be adopted at this time but perhaps reconsidered later.

Discussion

A $5/month fee is a minimal cost increase to large direct access customers, resulting in rate impacts of less than 1/100th of a mil per kilowatt-hour. It would not materially change the economics of their direct access transactions. It would, however, reduce the current subsidy to direct access customers resulting from the fact that bundled customers currently pay about 90% of the direct access costs recorded in Accounts 901, 902 and 903. The fact that there are many unsettled aspects of direct access service is not dispositive, since TURN is proposing the customer charge solely on an interim basis. We will adopt a $5 per month direct access customer service charge on an interim basis, along with $378,420 in OOR, subject to further consideration in Phase 2 of this GRC or other appropriate proceeding.

4.3.5. Level of Service Charges

SCE has proposed cost-based service charges, resulting in significant increases. In one case the increase is 90%. ORA proposed a schedule of more moderate fee increases under which no fee rises by more than 26%. The following table shows the current and proposed fees.

Current and Proposed Service Charges

Discussion

SCE has agreed to ORA's proposed fee for returned checks but otherwise continues to oppose ORA's fees proposal. We find that ORA has presented a schedule of fees designed to reflect cost responsibility, while also moderating percentage increases and taking account of affordability concerns. ORA's proposal reflects a more reasonable balancing of ratemaking considerations, including affordability, whereas SCE's proposal reflects too much emphasis on the role of costs as a ratemaking consideration. ORA's proposed schedule of fees is more reasonable and will therefore be adopted.

Since we are adopting ORA's proposals for service charges, we will likewise adopt OOR associated with those charges. We also adopt ORA's correction to SCE's proposed Direct Access Service Fee OOR amount of $256,000. The correct amount, which we adopt, is $368,000.

4.4. Capital

4.4.1. SCE's Showing

Through inadvertency, SCE did not provide justification for $3.408 million in CSBU-related capital additions (FERC Account 101) in its direct showing. The capital additions were, however, reflected in SCE's results of operations showing. ORA pointed out this omission in its testimony, and SCE subsequently provided justification for the expenditures with its rebuttal showing. The capital additions at issue are for furniture and equipment, building improvements, and tools and test equipment.

ORA does not dispute the reasonableness of SCE's CSBU capital additions but it nevertheless proposes their disallowance on procedural grounds. Also, ORA concedes that SCE's omission of the necessary justification for the capital additions was inadvertent. ORA contends that providing justification in rebuttal testimony is too late for parties to analyze the showing. ORA therefore argues that SCE has failed to meet its burden of proof.

Discussion

The Commission has held that it is not permissible for utilities to hold back on the presentation of salient information until the submission of rebuttal testimony.50 We would be well within our rights and responsibilities if we were to disallow these capital additions on procedural grounds as advocated by ORA. However, taking into account the special circumstances applicable here, we find that our procedural requirements may be waived in this instance.

SCE obviously made a simple mistake. Its failure to include the justification with the application was clearly not part of a litigation strategy whereby SCE would wait until rebuttal to spring this information on unsuspecting parties. The results of operations study submitted with SCE's initial showing reflected the capital additions. Presumably, if ORA had identified this deficiency during its review of the Notice of Intent, SCE would have been alerted to the omission and could have included the justification with its application.

ORA has made no claim or showing that it attempted to have SCE provide the justification through a data request. Given the relatively small amount of material provided by SCE to justify inclusion of the capital additions (approximately two pages), and the relatively small dollar amount of additions at issue, we are not prepared to conclude that ORA and other parties would have had insufficient time to review the showing had they made a timely data request upon learning of the mistaken omission. As a matter of equity in this proceeding, we will consider SCE's justification on the merits rather dismiss the entire CSBU capital additions request on procedural grounds. This is based on consideration of the circumstances described herein and is not a license for SCE or any other utility to depend on similar waivers of our procedural rules in the future.

ORA expresses concern that "failure to deny Edison's inadequate and tardy request will lead to the belief among utilities that they do not have to provide justification for their requests when they file their applications and during their direct cases-that the Commission will allow them to file their cases during the rebuttal phase of the case." (ORA Reply Brief, p. 3.) We understand ORA's concern, and we state emphatically that any utility harboring such beliefs should be disabused of them. Notwithstanding today's decision, we reserve the right to deny consideration of any "rebuttal" evidence that could have and should have been included with the utility's direct showing, even where, as here, a simple mistake of omission has been made by the utility.

Although ORA declined to do so, we have reviewed the content of SCE's justification for its CSBU capital additions. SCE contends that the capital additions at issue are typical, noncontroversial expenditures. However, of its total request for $3.408 million, $956,000 is for the energy centers and $150,000 is for the pump test program. Since we have denied SCE's request for O&M funding of these programs, and SCE has not demonstrated that GRC funding of the program's capital requirements is reasonable given the O&M denial, we will disallow the associated capital expenditures, for a total disallowance of $1.106 million.

4.4.2. Real Time Energy Metering (RTEM)

In 2001, SCE embarked on an RTEM project. In AB1X-29 and SBX1-5 the Legislature mandated that $35 million shall be used to provide time-of-use or real time meters for customers whose usage is greater than 200 kilowatts. This RTEM project involves the purchase and installation of advance metering and related communication, information and notification systems for all bundled end-user accounts with monthly maximum demands of 200 kW or more. The project provides these customers with the ability to access their interval usage information via the Internet. The scope of work focused on purchase and installation of 12,000 meters and communications equipment, development of communications infrastructure, operational modifications, and customer education. SCE estimates the total cost of its RTEM project to be $27.649 million for capital requirements and $3.462 million for ongoing operational expense.

A significant portion of these RTEM costs will be offset by funds from the California Energy Commission (CEC) that are derived from appropriations authorized by ABX1-29 and SBX1-5.51 In April 2001, SCE responded to a CEC RFP on the use of state funds for the RTEM project. In May 2001, CEC agreed to allocate funding to SCE on a flat cost per meter basis, and SCE determined that this funding was insufficient to recover its total implementation costs. As a result, SCE filed Advice Letter 1549-E requesting a memorandum account to track all the RTEM project costs for future recovery. In fact, SCE stated in the advice letter that,

"Without authorization from the Commission to establish the memorandum account, SCE asserts it will not be able to install the communication infrastructure enabling the RTEM system to function, which in turn, will cause the state to forgo the conservation benefits of the RTEM during the summer and fall of 2001."52

The advice letter was approved by Resolution E-3746 dated June 28, 2001.

TURN takes the position that SCE should not receive ratepayer funding for the portion of its RTEM capital additions request that is not covered by the $16.8 million CEC contract. According to TURN, SCE should not have signed a contract with CEC for RTEM funding with any expectation of such ratepayer funding. TURN proposes a capital additions disallowance of $10.8 million for what it calls cost overruns. SCE disputes several aspects of this proposal.

Discussion

TURN has not shown that RTEM project cost overruns have occurred. SCE knew at the outset that CEC funding would be inadequate for the project that SCE had in mind, and its Advice Letter 1549-E identified this fact. However, the fact that there were no cost overruns does not resolve the issue. SCE must still show that it was reasonable for it to embark upon a project requiring $10.8 million in ratepayer funded capital additions.

In support of its expenditures, SCE emphasizes the mandatory nature of the RTEM program. It is true, as we stated in D.01-08-021 "that ABX1 29 does require mandatory TOU or RTP metering for customers whose usage is greater than 200kW of peak demand" and "that meter installation is mandatory under ABX1 29." (D.01-08-021, p. 5.) SCE has shown that the Legislature felt it was necessary for the installation of interval meters by the summer of 2001. We agree that the installation of these meters was crucial and was consistent with the desire of the Legislature. We are cautious to not make a precedent here that just because we permitted SCE to establish a memorandum account they can spend any amount and be assured recovery. We have analyzed the facts and testimony in this situation and feel that SCE was acting in good-spirit to further the goals of the Legislature. In addition, SCE made it well known from the start that the CEC funds would be inadequate. Because of the urgency and importance of this situation, we grant SCE's request for recovery of $10.8 million. TURN's proposed capital additions disallowance of $10.8 million is denied.

4.5. Service Guarantees

SCE's Senior Vice President of Customer Service testified that the company strives to provide a high level of customer service and satisfaction, and that in 2000, under the customer satisfaction measure of the PBR mechanism, 76% of SCE's customers responded that they were "completely satisfied" or "delighted" with SCE's customer services. Stating that it seeks to ensure that SCE lives up to this service quality commitment, and to compensate customers when SCE fails to do so, ORA proposes that an eight-point service guarantee program be mandated by the Commission.

Under the program proposed by ORA, SCE would guarantee that it will (1) meet agreed appointment times; (2) investigate non-emergency situations and communicate results to customers within seven days of a customer request; (3) decide on a course of action to resolve a complaint and communicate it to the customer within three working days, and communicate the complaint's resolution to the customer within 10 working days, or 30 working days when an off-site meter test is required or an on-site home audit is requested; (4) meet the agreed date for installing a new meter and turning on service for a customer; (5) respond to service interruptions within four hours after receiving a customer report by either restoring service or informing the customer of when they can expect service to be restored; (6) restore service within 24 hours; (7) provide at least three days notice of a planned interruption in service; and (8) issue an accurate first bill to a new customer account within 60 days of service initiation. Each time SCE fails to meet one of these guaranteed standards, it would be required to pay the affected customer $50 from shareholder funds. SCE would be required to report program results (number of claims made, claims paid, and amounts of money paid) to the Commission on a quarterly basis.

ORA states that its proposal is based on a voluntary program that SCE operated from 1995 to 2001. ORA notes that this was a period when SCE was faced with potential competition for its customers. ORA is concerned that now, with no competition for customers, SCE could merely pay lip service to ensuring customer satisfaction without having to immediately answer to them. ORA states that its proposal is not meant to be punitive, but instead is meant to provide an incentive for Edison to maintain high-quality customer service.

SCE opposes this proposal, although, under specified conditions, it would agree to consider implementing some aspects of ORA's proposed program (agreed date for meter installation and service turn-on, response to service interruptions within four hours, service restoration within 24 hours). Among other things, SCE is concerned with the cost of the proposed program. SCE estimates that start up costs would be $708,000, annual operating costs would be $534,000, and "base level" credit costs would be $838,000 annually.

SCE maintains that the following design principles should be applied in any adopted customer guarantee program:

· A service guarantee program should enhance overall customer satisfaction at a reasonable cost.

· All of the costs to implement and administer a service guarantee program should be recoverable through rates.

· A service guarantee program should be based on standards and expectations that are achievable.

· Service guarantee credits should be fair, balanced, and consistent with the other California utilities.

· A service guarantee program must be simple, clearly defined, easy to understand, and include appropriate exclusions.

· A service guarantee program should complement existing customer satisfaction efforts and avoid duplication.

· A service guarantee program should focus on a few critical services that are most important to customers.

Discussion

This GRC is the appropriate venue to address and propose a customer service guarantee program; the General Rate Case is the traditional forum in which the revenue requirement associated with providing reasonable adequate service is traditionally made. Defining the level of customer service and appropriate enforcement mechanisms in this GRC will assist SCE and the public in finding common ground on the level of serviced expected of a franchised provider. ORA has presented a program that would address certain areas of customer satisfaction performance by providing compensation to certain customers who have been inconvenienced by SCE. We find this to be the proper place to adopt certain customer service standards. The standards we adopt today from ORA's proposal are detailed below. Although we believe that SCE is currently providing satisfactory customer service overall, we feel that customer service is a core element of utility service and thus wish to ensure there is no degradation to SCE's current level of customer service.

Additionally, a central feature of ORA's customer service guarantee program is the use of "compensatory rebates" penalties; penalties that accrue to the customer when SCE fails to meet a customer service standard. This is a self-enforcing mechanism that can create a significant incentive for SCE to meet the adopted standards. However, we are concerned that the levels proposed by ORA may be excessive. Although we do not adopt all of the standards proposed by ORA, we will make modifications to those proposed standard penalties that we endorse in this decision as follows:

Adopting a rebate enforcement element to the mandatory customer service guarantees we adopt today requires that we monitor the program to ensure efficacy and avoid potential abuse. We believe that ORA's recommendation to require SCE to report program results (number of claims made, claims paid, and amounts of money paid) to the Commission on a quarterly basis is too onerous and thus will require that reporting be done on a semi-annual basis.

5. Administrative and General

5.1. Introduction

Section 6 addresses contested Administrative and General (A&G) issues raised by ORA, which addressed A&G issues comprehensively, as well as issues raised by Aglet, TURN, and Greenlining. As shown in the Joint Comparison Exhibit (Exhibit 403, p. 315), SCE's total A&G expense forecast is $552.1 million (constant 2000 dollars), while ORA's forecast is $476.348 million. This represents a difference of $75.752 million.

5.2. Financial Organizations and Capitalized
Expenses

5.2.1. Account 930 - Participant Credits

SCE bills the minority participants of facilities such as SONGS for their share of the A&G costs to operate these facilities. Also, the majority owner of Palo Verde and Four Corners bills SCE for their A&G costs. These transactions are tracked in Account 930.

SCE forecast an overall non-labor credit of $8.087 million in 2003 for Account 930, an amount equal to recorded expenses for 2000. Using a five-year average of 1996-2000 data, ORA forecast an overall non-labor credit of $12.192 million in 2003 for Account 930, yielding a difference of $4.105 million.

Discussion

The recorded non-labor amounts in Account 930 are shown in the following table:

FERC Account 930 Non-Labor

(Constant Dollars X 1000)

The consistent decline from 1997 to 2000 to approximately one-half the 1997 level suggests that a declining trend may exist, throwing into doubt the reasonableness of using a five-year average. These numbers do not suggest mere variability as ORA contends, nor do we find convincing evidence within the underlying data in Exhibit 211 that variability warrants the use of a five-year average for this account.

SCE has presented evidence explaining why these credits have declined from 1997 to 2000. The methodology used to calculate participant credit rates is defined in the Participant Billing Operating Agreements between SCE and the participants. The rates have been declining from the mid-forty percent level in 1997 to the mid-thirty percent level in 2000. Also, the SONGS participation rate, which is the largest component of the Account 930 expenses, will remain constant through at least 2003 pursuant to a negotiated agreement.

We find that recorded data for 2000 is most representative of test year amounts for Account 930. We therefore adopt SCE's test year 2003 forecast of $8.102 million for Account 930 (Labor and Non-labor).

5.2.2. Capitalized Pension and Benefits (P&B)

SCE computes the total amount of P&B expenses, subtracts the capitalized portion, and assigns the remainder as O&M expenses. A capitalization rate is determined to estimate the capitalized portion. Capitalized P&B expense is recorded as a credit to Account 926.900 (Employee P&B Transferred) and a debit to Account 107 (Construction Work In Progress).

SCE initially used a capitalization rate of 24% but during the proceeding changed its recommendation to 25% to reflect a correction identified by TURN. TURN recommends a capitalization rate of 29.43% based on a regression analysis of data from 1990-2000 that relates capitalized labor to gross plant additions and assumes SCE's forecast of capital additions is adopted. (TURN recommends that this rate be recomputed to reflect the adopted capital additions.) Compared to SCE's forecast, TURN's recommended approach results in an estimated difference of $14.893 million in Account 926 and an additional $4.703 million in capital overheads.

Discussion

TURN's regression methodology is potentially more reliable as a forecast methodology than SCE's selection of a single year's data, but SCE has raised doubts about its validity in this GRC because of the fact that SCE plans to accomplish the increased capital additions associated with its Infrastructure Replacement Program with contract labor, which does not receive SCE P&B.

The capitalization rate percentages associated with recorded spending, as shown in Exhibit 299, p. 10, are set forth in the following table:

Capitalization Percentages Based on Recorded

Capitalized and Total Utility Labor

Based on the method used in prior rate cases and 1999 recorded amounts, SCE proposes a capitalized P&B rate of 25% for 2003 (25.4% rounded down to 25%). In light of the apparent increasing trend as shown in the table above, we will adopt TURN's secondary recommendation to use the 2000 value of 27.8%.

5.3. Legal And Regulatory, Workers'
Compensation, Insurance

5.3.1. Account 923 (ORA Audit
Recommendations)

ORA recommends that $407,071 be removed from Account 923 for outside counsel expenses incurred in 2000 to prepare for a potential involuntary bankruptcy filing. Because SCE used a three-year average to forecast 2003 expenses, this would result in a $136,000 reduction to SCE's 2003 forecast. ORA also proposes a $62,629 reduction to recorded 2000 expenses in Account 923 to remove costs of litigating SCE's Filed Rate Doctrine lawsuit against the Commission, which results in a test year reduction of $21,000.

Discussion

ORA's recommendations rest on the premise that these outside counsel expenses are non-recurring (bankruptcy) and on the basis that a settlement was reached (Filed Rate Doctrine lawsuit). We find ORA's showing with respect to the bankruptcy expenses persuasive. The argument that each litigation matter is unique, and the implication that there can be no non-recurring litigation expense, is without merit. We are not prepared to assume that it is reasonable to expect that the level of spending on bankruptcy litigation in 2000 will continue. ORA's proposed removal of $136,000 associated with bankruptcy expenses in the test year is reasonable and will be adopted.

Even if all appeals regarding the Filed Rate Doctrine lawsuit were exhausted, we are persuaded that the costs SCE incurred are nonetheless representative of future litigation costs. ORA's proposed test year reduction of $21,000 will not be adopted.

5.3.2. Accounts 923 & 928 (Forecast
Methodology)

Aglet contests SCE's use of a three-year average for outside counsel expenses (1998-2000), proposing instead that a two-year average be used (1999-2000). Aglet's recommendation would result in a reduction from SCE's forecast of $2.005 million in Account 923 and $0.729 million in Account 928, or a total of $2.734 million

Discussion

According to Aglet, SCE's three-year average methodology does not adequately reflect the reduced legal workload associated with the decline of restructuring implementation work, the signing of long-term procurement contracts by the DWR, reduction of the residual net short, generation divestiture, and the financial crisis. However, 2000 expenses were low due to the energy crisis, and legal and regulatory work associated with the energy crisis has been increasing. The assertion that the execution of the DWR contracts and the decrease in average residual net short will lead to a decreased legal workload is contrary to SCE's direct showing and is not adequately supported. Finally, the assumption that the declining legal work associated with power plant divestiture should result in a lower forecast for Accounts 923 and 928 outside counsel expenses is not adequately supported, as legal costs incurred for divestiture were charged to capital work orders. We conclude that use of a three-year average to forecast outside counsel expenses is reasonable and should be adopted.

5.3.3. Account 928 (GRC Expenses)

Within Account 928, SCE recorded expenses of $1.3 million in 2000 for GRC expenses. SCE bases its test year forecast of GRC expenses on the 2000 recorded value, i.e., it did not use three-year averaging. SCE believes that the $1.3 million spending level will recur with the return to conventional cost of service ratemaking.

Aglet proposes that SCE's 2003 Account 928 forecast be reduced by $130,000 (10% of $1.3 million) based on the contention that SCE will gain experience in this GRC that will improve its efficiency in preparing the next GRC. SCE counters that rate case expenses are increasing due to additional requirements imposed on the company, the increased scrutiny on SCE's cases, and the involvement of new parties. SCE questions the qualifications of Aglet witness Weil to make the professional judgment that a 10% efficiency gain is reasonable to achieve.

Discussion

We accept Weil's qualifications to make the professional judgment that a 10% efficiency gain is reasonable. One need not have direct experience with a utility function as a prerequisite to rendering expert opinions regarding that function. However, while SCE should be more efficient, the evidence also suggests that the scope of work to be performed in connection with GRCs has been growing. We are not prepared to conclude that a $130,000 reduction from the 2000 recorded cost is reasonable.

5.3.4. Account 930 (Board Meetings)

SCE's test year 2003 forecast for Corporate Governance activities in FERC Account 930 is $3.3 million. ORA recommends that 2000 recorded expenses in Account 930 be reduced by $133,250. This recommendation is based on limiting ratepayer funding to seven corporate board meetings per year (SCE held 11 board meetings in 2000), resulting in a proposed reduction of $64,500, and removal of $68,750 for retainer fees paid to Directors.

Discussion

ORA's recommendation is based on the assumption that the number of board meetings held annually prior to 2000 will resume. The evidence does not support this assumption. SCE has amended its bylaws so that certain officers could call a meeting for any purpose. SCE held more than seven board meetings in 2000, 2001, and 2002. SCE continues to face pressing matters such as its financial health and the shake-up of energy markets. The requirements of the Sarbanes-Oxley Act are likely to lead to greater board oversight of corporate activity, and more board meetings. We accept SCE's forecast without the reduction proposed by ORA.

5.3.5. Property and Liability Insurance
Expense

SCE purchases nuclear property insurance from Nuclear Electric Insurance Limited (NEIL), a mutual insurance company. From time-to-time, based on current expected losses, surplus funds, and reinsurance available, the board of directors of NEIL may approve a refund to its members.

Based on statements from NEIL's board of directors, SCE forecast a $5.451 million refund in 2003, $5.387 million of which is for SONGS, with the remaining $64,000 for Palo Verde. ORA recommends forecasting the NEIL refund for 2003 by using six-year average of recorded data, or $11.501 million. However, during the proceeding NEIL's board of directors determined the actual refund amount to be allocated to SCE as $7.499 million ($6.890 million for SONGS and $608,302 for Palo Verde). SCE now proposes using the actual refund instead of a forecast.

Discussion

The timing of the processing of this GRC made it possible to use actual instead of forecast data for this particular expense. We find that it is reasonable to do so, and therefore adopt SCE's revised request. ORA's concern that the NEIL refunds may be different in 2004 and 2005 is not compelling, since our purpose here is to forecast expenses for 2003.

As it did with nuclear property insurance, ORA recommends a six-year average of $560,000 to estimate nuclear liability insurance refunds for the test year. This recommendation would raise SCE's forecast of $239,000 by $321,000.

Discussion

ORA has demonstrated that a six-year average is reasonable for this forecast, even though it may actually underestimate the refund. In its opening brief, SCE points to SCE's testimony that "[n]uclear liability insurance is expected to increase over the forecast period" for the same reasons that nuclear property insurance is increasing. (Exhibit 43, p. 52.) This hardly represents convincing evidence that its estimate of $239,000 for refunds is more reliable than ORA's estimate based on a six-year average. ORA's adjustment is reasonable and will be adopted.

5.4. Shared Services

5.4.1. Shared Services Expenses

SCE's Business Resources Unit provides document and drawing management, event and travel services (E&TS), mailing services, and records management and storage services. ORA takes issue with certain non-labor

E&TS-related costs. ORA states that it is proposing a forecast of $1.829 million

for Business Resources, which would be a reduction of $863,000 from SCE's non-labor forecast of $2.692 million in Account 921. (Exhibit 45, p. 7; Exhibit 188, p. 14-D-6.) ORA states that its reduction is made to normalize non-recurring, unusual, and/or one-time expenditures. Noting that E&TS scheduled employee events that included basketball games at Staples Center and baseball games at Edison Field, music concerts, etc., ORA states that it removed all such expenses not appropriately charged to ratepayers.

Discussion

SCE and ORA have managed to inject considerable confusion into the framing of this issue. SCE's rebuttal testimony and the Joint Comparison Exhibit state that ORA is proposing a reduction of $1.416 million to the Business Resources forecast (Exhibit 299, pp. 32, et seq., passim; Exhibit 403, p. 348), but ORA stated earlier in its errata testimony that it is proposing a reduction of $863,000 (Exhibit 188, p. 14-D-6).53 Although ORA states in its testimony and in its opening brief that it is recommending a reduction of $863,000, ORA jointly agreed with SCE in the Joint Comparison Exhibit that the issue represents a difference between the parties of $1.045 million. In its rebuttal testimony, SCE acknowledged that it failed to provide needed clarification in a data request response, and that this caused confusion. (Exhibit 299, p. 34.) Also, SCE changed its position on at least four occasions. (Id., pp. 36-37.)

Whether the dispute is over $863,000, $1.045 million, or $1.416 million (and as discussed below we find it is none of these), SCE is only requesting $818,000 for E&TS, and only $359,000 for the non-labor portion (Exhibit 45, p. 12). Thus, ORA's request to remove expenses that exceed the company's entire forecast for E&TS is, on its face, at least questionable. SCE clarified that ORA's recommendation affects several areas in the company, not just shared services, but ORA does not acknowledge this clarification.

After reviewing the record, we find that the most reliable statement of the dispute is set forth in SCE's rebuttal testimony, Table IV-3 of Exhibit 299 and the accompanying text at pp. 37-38. SCE's last revised forecast of expenses for events handled by E&TS is $865,936, and ORA's recommendation is $374,061, a difference of $491,875.

SCE contends that the remaining events, i.e., those covered by its final request for $865,936, were conducted in furtherance of SCE's utility activities and are representative of the types of meetings that will be conducted in 2003 and beyond. SCE contends that they are not non-recurring, unusual, and/or one-time expenditures. The largest area of disagreement between SCE and ORA is employee awards and recognition ($345,741 of the total $491,875 difference). Among other things, SCE's Information Technology (IT) department held a celebratory "Y2K" event in 2000 at a cost of $265,898, and the Law Department held a retirement function at a cost of $4,348.

Whether or not these are "normal and recurring business expenses within the business world that any given company would incur as an operating expense" (Exhibit 299, p. 40), SCE has not demonstrated that utility ratepayer funding of such celebrations is reasonable or necessary. We conclude that SCE has failed to provide justification and substantiation for the costs of the disputed events, and therefore order a reduction of $491,875. We will reduce SCE's Business Resources forecast by that amount.

SCE forecast $4.671 million for its Corporate Security and Emergency Planning and Preparedness Unit (CS&EPP). The Investigations Division of CS&EPP consists of 14 SCE employees who are responsible for investigating threats and other hostile or criminal acts directed at SCE or its employees. Examples of its functions include a 2000 incident where a contractor falsified expense reports and timesheets for over $10,000, and a 1998 incident where two persons impersonated SCE employees and fraudulently offered customers a 50% bill discount in return for an advanced fee. In the 2000 incident SCE's investigation led to criminal charges, conviction, and restitution to SCE. In the 1998 incident SCE's investigation led to identification of the suspects and felony convictions.

ORA proposes a $70,000 reduction to SCE's forecast for the non-labor portion of the Investigations Division. These non-labor expenses consist primarily of costs related to contract services, as well as professional expenses associated with background checks. Due to the variability of the non-labor expenses, SCE used a five-year average of recorded data to forecast a test year expense of $204,000. ORA's proposed reduction is based on ORA's use of the last recorded year.

Discussion

The historical non-labor recorded/adjusted costs for Investigations are set forth below:

Investigations Expense - Non-Labor

(Constant Dollars X 1000)

SCE explains that the reduced expense in 2000 was due to temporary cost reduction measures implemented in 2000 in response to the California energy crisis. (Exhibit 45, p. 26.) SCE maintains that the lower spending level of 2000 is not sustainable. (Id., p. 27.) Absent persuasive evidence to the contrary, ORA's recommendation to use the last recorded year would not be reasonable. ORA's position that a hiring freeze that SCE implemented in 2000 should reduce the need for background investigations lacks persuasiveness. SCE's forecast of $204,000, based on a five-year average, will be approved.

The Shared Services Support Group consist of seven employees responsible for providing centralized support to the organization Vice President. Their functions include business planning and strategy, business improvement, and financial services. Using the last recorded year, SCE forecast expenses of $1.049 million ($767,000 labor and $282,000 non-labor).

ORA proposed reducing the non-labor portion of SCE's request from $282,000 to $177,364, a reduction of $104,636. ORA recommends this reduction to remove the costs of employee awards, mentor luncheons, employee contributions, flowers, and sports events.

SCE agreed to remove $21,129 from its request for items that are non-recurring and items for which SCE agrees there are no ratepayer benefits. These include sporting events and employee contributions, among others. SCE's revised request is $260,871, which is $83,507 greater than ORA's recommendation.54

Discussion

Although SCE removed some of the disputed expenses for Shared Services Support, SCE contends that expenses for food vendor services, mentor luncheons, and employee awards are appropriate because they support valid business purposes. The disputed expenses support working lunches for the Vice President and managers, which, SCE contends, results in greater organizational effectiveness. They also support lunches for mentor programs that, according to SCE, strengthen the organization, provide for career enhancement, professional growth, and job effectiveness. Finally, SCE maintains that employee awards and recognition programs foster continuous improvement and achievement of long-term objectives, and create an environment of valued contribution that promotes employee retention.

We find SCE's justification for the disputed expenses unconvincing. In particular, SCE has not adequately demonstrated that ratepayer funded lunches for executives and managers and for mentor program participants is necessary or appropriate. ORA's proposed reduction of $83,507 will be adopted. The adopted non-labor expense for the Shared Services' Support Group in Account 921 is $177,364.

ORA removed costs in Account 921 for market studies and increased title and mapping services that were performed by SCE in connection with studies for its Revenue Enhancement section. ORA removed a total of $323,000 on the basis that these were one-time, nonrecurring expenditures.

SCE disputed ORA's recommendation, contending that the services are necessary to prepare properties for marketing to developers, which ultimately benefits ratepayers by the generation of OOR.

In its opening brief, TURN pointed out that under the gross revenue sharing mechanism adopted in D.99-09-070, costs for land title services and mapping support are not to be included in rates.

Discussion

We appreciate TURN's pointing out that compliance with D.99-09-070 requires exclusion of this expense. SCE readily admits in its reply brief that TURN's recommendation is appropriate. SCE states that it overlooked the adjustment due to an internal miscommunication. We find this oversight troubling, and we therefore adopt TURN's proposal to require SCE to certify in an advice letter filing that none of its GRC requests include expenses that, pursuant to D.99-09-070, are to be borne by shareholders. SCE's request for Account 921 is reduced by $323,000.

ORA takes issue with SCE's inclusion of $96,000 in expenses for landscape maintenance activities consisting of such activities as turf replacement, removal and replacement of ground cover, irrigation work, replacement of dead trees, etc. ORA believes that these expenses should be removed on the basis that they are one-time, nonrecurring costs.

Discussion

ORA has not adequately demonstrated why the disputed landscaping expenses would be nonrecurring activities ineligible for consideration in the forecast. This proposed reduction will not be approved.

SCE has 70 non-electric facilities encompassing 170 separate buildings and tens of thousands of square feet of parking surface. SCE used the last recorded year to forecast test year expenses $6.740 million for Account 935 - Maintenance of General Plant. These expenses cover maintenance of structures and parking areas at SCE's non-electric facilities and include roof patching, minor asphalt repairs, interior drywall patching, and painting. They also include repair of office furniture and equipment, minor building alterations, and remodels. SCE states that these expenses are increasing due largely to the increasing average age of the facilities.

ORA used a five-year average to arrive at a forecast of $6.041 million, or $699,000 less than SCE's forecast. ORA contends that a five-year average is appropriate to account for normal year to year fluctuations.

Discussion

Because SCE's non-electric facilities are aging, and facility age significantly influences maintenance costs, use of the last recorded year instead of a five-year average is appropriate for this account. SCE's forecast is therefore approved.

5.4.2. Shared Services Capital

SCE's forecast of shared services capital expenditures is $25.000 million in 2002 and $34.617 million in 2003, or a total of $59.617 million. ORA recommends adoption of capital expenditures of $10.896 million in 2002 and $8.255 million in 2003, or a total of $19.151 million. ORA's recommendation represents a proposed disallowance of $40.466 million. Section 6.4.2 addresses ORA's claims regarding the adequacy of SCE's showing, projects over $1 million, and blanket work orders.

ORA contends that SCE did not provide sufficient documentation to fully substantiate its shared services capital expenditures forecast, that SCE did not provide cost benefit analyses to demonstrate the value of demolishing and major remodel projects, and that SCE did not provide sufficient documentation for ORA to independently reconstruct the forecast. ORA believes that the forecast costs for certain projects appear to be based primarily on "wish lists."

Discussion

SCE's direct testimony that accompanied the application included approximately 60 pages of text, tables, and figures explaining and supporting SCE's shared services capital expenditures request. (Exhibit 46, pp. 15-75.) SCE provided additional, comprehensive information to ORA in response to data requests (e.g., Exhibit 193), and ORA has not shown that SCE unreasonably failed to respond to any ORA data requests pertaining to this issue. While we will address the reasonableness of various components of the disputed capital additions in the remainder of Section 6.4.2, we find no basis for concluding that SCE has failed generally to justify its requests, or has merely presented wish lists.

We will not disallow disputed shared services capital additions on the basis of such general claims of an inadequate showing. Thus, ORA's recommended disallowances of $1.593 million for Various Rights-of-Way Acquisitions, $15.107 million for Field Facility Infrastructure Improvements, and $2.314 million for projects under $1 million, which are not based on any more specific ORA analysis, will not be approved.

SCE determined that certain of its non-electric facilities needed replacement or major remodeling, but for some time it did not implement remodeling projects due to changes in the business environment. By 1998, SCE determined that the combination of the facilities' age, upgrades required by law (such as the Americans With Disabilities Act or ADA), and the need for telecommunications and electrical upgrades were converging. To address this concern, SCE developed a Strategic Facilities Plan (SFP) that focused largely on the General Office (GO) complex in Rosemead and also addressed facilities in Westminster, Santa Ana, Alhambra, and Irwindale. SCE expended $32.342 million on the SFP prior to 2001, did not make SFP expenditures in 2001, and forecast SFP capital expenditures of $5.247 million in 2002 and $3.757 million in 2003 for a total of $9.004 million for the latter two years.

Included among the various projects under the SFP umbrella were air conditioning upgrades and replacement; ADA upgrades; fire, life, and safety upgrades; electrical and telecommunications infrastructure upgrades; and reconfiguration of the GO1 building from hard wall offices to an open landscape office environment to facilitate and reduce the cost of employee moves.

ORA believes that some of the SFP costs included in SCE's forecast of capital costs for 2002 and 2003 are excessive and redundant, and that SCE may be seeking duplicate recovery for certain security enhancements. ORA also questions the reasonableness of SCE's decision to reconfigure GO1 to an open landscape environment. ORA recommends disallowance of the entire $9.004 million amount of SFP capital additions forecast by SCE for 2002 and 2003. In addition, ORA recommends disallowance of the costs of a new corporate fitness center that was included with the SFP. We address this latter recommendation in the following section.

Discussion

As shown in SCE's direct testimony (Exhibit 46, p. 44, Table VIII-7), expenditures on the first, third, and fourth floors of the GO1 building account for all of the $9.004 million in SFP expenditures that SCE forecast for 2002 and 2003. Most of the $41.346 million in planned spending on SFP projects, including $7.526 million for air conditioning chillers, occurred prior to 2001. Since the GO air conditioning chiller project was funded with expenditures made prior to 2001, and, except for the fitness center, ORA's SFP recommendations pertain to forecast expenditures for 2002 and 2003, we are perplexed by SCE's persistent attempts to show that ORA misunderstands the importance of the chillers. In fact, ORA does not recommend any disallowance with respect to the chillers.

A major focus of ORA's concern regarding the SFP pertains to the reconfiguration of GO1 to an open environment. However, SCE's direct testimony adequately addressed the value of this SFP component by noting the reduced cost of employee moves. In 1997 an average of 1,184 employees moved per year in GO1 at a cost of approximately $2,500 per move, or $2.96 million annually. With the reconfiguration, SCE expects the future average cost per move to be $400. SCE also expects the reconfiguration to improve the utilization of available space at the GO complex.

ORA also expressed concern regarding expensive artwork in the GO complex, based on observations by ORA's witness during a field trip. While we would be reluctant to have ratepayers fund expensive artwork, ORA's concern is not backed up by information or analysis that would allow us to determine the extent, if any, to which SCE's GRC request includes unreasonable costs for such funding. There is no evidence with respect to this issue that justifies a disallowance.

SCE demonstrated that its 2002 decision to earmark $4.5 million from the Various Major Structures blanket work order for certain security related improvements is separate from and addresses different requirements than those that the SFP improvements were intended to address. Thus, ORA has not demonstrated that its redundancy concerns are applicable to the SFP. We will further address this concern in connection with the blanket work orders, below.

Notwithstanding SCE's diversion to the issue of the chillers, SCE has demonstrated that the company's forecast SFP capital expenditures for 2002 and 2003 are reasonable. ORA's recommended disallowance of the forecast SFP capital expenditures for 2002 and 2003 will not be adopted.

As part of the SFP, SCE constructed a corporate fitness center, opened in 1999, for use by all GO occupants as well as employees that work at other locations. The objectives of the fitness center are to increase the number of employees who engage in regular physical activity and to leverage the knowledge and expertise of the fitness center staff across the company by expanding health education to address emerging health and safety issues such as back injuries, stress reduction, ergonomics, and nutrition education. SCE believes that meeting these objectives will lead to reduced absenteeism, increased productivity, and reduced work-related injuries. SCE also believes that it will assist in the attraction and retention of productive employees. The cost of the fitness center was $1.530 million. (Exhibit 302, p. 193.) SCE's Human Resources Department is responsible for operating the fitness center, the operating costs of which are addressed within Section 6.7.

ORA believes that the costs of the fitness center should not be funded by ratepayers, and it recommends removal of those costs for ratemaking purposes. ORA notes that SCE did not request prior Commission approval to construct the center. ORA also notes that SCE employees can enroll in private fitness programs and receive reimbursement through SCE health benefits that are also funded by ratepayers. ORA believes that shareholders benefit from healthy and productive employees, as expenses for sick and injured employees are reduced.

Discussion

ORA has not demonstrated that there are regulatory requirements that SCE secure Commission approval prior to constructing a corporate fitness center. SCE's failure to do so is not a reason for disallowance of the center's costs. Nevertheless, we are not convinced that ratepayer funding of the center is necessary or appropriate at this time. We do not doubt that there are productivity benefits from having healthy employees. However, SCE has not adequately demonstrated that ratepayer funding of the fitness center represents a cost-effective means of improving employee health and producing productivity gains that benefits ratepayers.

SCE's health benefits program already provides full time employees who choose to enroll in a medical plan a Preventive Health Account, which provides reimbursement of up to $150 per year for completing approved fitness, diet, nutrition, smoking cessation, and other health promotion programs. Approximately 250 participants use the fitness center each day, which means that most of SCE's 12,000 employees either choose not to use it or are unable to do so. If the ratepayer benefits of the corporate fitness center were as demonstrable as SCE implies (but fails to demonstrate) they are, then we would need to consider significant disallowances in this GRC for SCE's failure to construct enough fitness centers, or provide comparable access to fitness facilities, so that all employees had access to their benefits.

Because SCE has not demonstrated that ratepayer funding of the corporate fitness center is reasonable, the associated construction costs should not be included in rate base.

SCE's GO complex is within one mile of the epicenter of the 1987 Whittier Narrows earthquake. SCE's shared services capital request includes a forecast of $1.5 million in expenses in 2003 for seismic upgrades to GO2. This building was constructed in 1974 to accommodate computing/information services requirements. Lessons learned from the 1994 Northridge earthquake indicated that welded moment-frame buildings similar to the GO2 office area had sustained severe cracking in the beam-column connections. To minimize structural damage during moderate earthquakes, the seismic upgrades will add diagonal bracing. The upgrades will also include ceiling system bracing, anchoring and cross-bracing computer access floors, bracing partition walls, and installation of seismic bracing on pipe and duct supports and equipment.

ORA proposes a disallowance of the entire $1.5 million in capital expenditures for the seismic upgrade project. ORA did not discuss the reasons for its proposed disallowance in its prepared testimony. On cross-examination, ORA's witness suggested that SCE's decision to construct a corporate fitness center indicated that the seismic upgrades project was not a priority for SCE.

Discussion

SCE's decision to seismically upgrade its GO2 building was the result of a study conducted by the SAC Joint Venture, a partnership of the Structural Engineers Association of California, Applied Technology Council, and California Universities for Research in Earthquake Engineering, under contract with the Federal Emergency Management Agency (FEMA). The findings and recommendations were published in June 2000, whereas the corporate fitness center was completed in December of 1998.

SCE's decision to construct a fitness center has no bearing on the reasonableness of the proposed expenditures for GO2 seismic upgrades, and ORA has presented no other reason why the upgrades are unreasonable or imprudent. We find no basis for concluding that the seismic upgrades were not a priority. ORA's proposed disallowance will not be approved.

Despite routine preventative maintenance and planned capital repairs and replacements, there are times when unplanned major structure work must be undertaken. In 1997 SCE established blanket work orders for "Various Major Structures" to provide funding for such occurrences. The budgeted funds have historically been required in each year since 1997, except for 1998. An example of their use is an upgrade to the overloaded electrical system of the Long Beach Telephone Center in 2000.

SCE's request in this GRC includes $10.498 million for Various Major Structures. ORA recommends a complete disallowance of SCE's $10.948 million forecast for Various Major Structures. ORA's testimony states that:


ORA believes that some of the costs SCE has forecasted relating to its implementation of its SFP are excessive and redundant. For example, SCE forecasted $10,948,000 for Various Major Structures, however SCE only briefly describes projects amounting to $4,500,000 of the $10,948,000 for security improvements and remains silent on its forecasts for the difference of $6,448,000. SCE forecasted an additional $1,471,000 for Security System Enhancements and another $1,288,000 for Access/Intrusion Security Control and some of the security line items listed in the forecasts for the three areas mentioned above appear to be some of the same or similar items and are thus redundant and not prudent investments, and should not be funded 100% by its ratepayers. (Exhibit 188, p. 16-D-3.)

Discussion

ORA's proposed disallowance for Various Major Structures appears to be based upon a misunderstanding regarding the SFP and its relationship to blanket work orders. In fact, the SFP is a separate, specific project of defined purpose

and duration, whereas the Various Major Structures blanket work orders provide funding for unplanned occurrences on an ongoing basis. SCE has demonstrated that there is no redundancy warranting their disallowance, and ORA has provided no other basis for its proposed disallowance. We therefore adopt SCE's request for Various Major Structures.

5.5. Information Technology

5.5.1. Introduction

SCE's Information Technology (IT) Department oversees SCE's infrastructure of large processors, storage media, communications links, operating system and application software, and personal computing and communications devices. As of January 31, 2001, the IT Department had 1,056 regular and part-time employees and 208 supplemental and contract employees.

5.5.2. IT Expenses

As set forth in its direct testimony (Exhibit 47, pp. 1-2), SCE forecasts test year IT O&M expenses of $117.891 million, consisting of $32.073 million in non-A&G Accounts 517, 561, 588, and 903, and $85.817 million in A&G Accounts 920/921, 923, 926, and 931. SCE states that while recorded expenses were an important starting point for the forecasting analysis, its test year O&M forecast is not based solely on past spending. SCE states that it also considered new and changing business requirements and the collective judgment of its IT experts. In other words, SCE used a budget-based approach for its IT O&M forecast.

ORA rejects this approach. Based on proposed adjustments to remove from recorded data the costs of a large (O&M expenses of $42.987 million in 1998, 1999, and 2000), nonrecurring IT project that addressed anticipated Y2K computer problems, and use of a five-year average of the resulting recorded/adjusted IT O&M expenses, ORA proposes an IT O&M forecast of $99.324 million. This represents a reduction of $18.566 million from SCE's aggregate forecast of $117.890 million for test year 2003 IT O&M expenses.

Discussion

SCE's forecast of $117.890 is $27.485 million (30.4%) higher than the 1996 recorded level of $90.405 million, and $6.055 million (5.4%) higher than the 2000 recorded level of $111.835 million. SCE states that two factors explain the increase from 1996 to 2003. First, the company decided "to retain and improve [its] in-house IT function," so it "developed and implemented several new practices (such as company wide planning, management of cross-business-unit-issues, adoption of common application development and maintenance processes and practices, and the establishment of common IT infrastructure hardware)." (Exhibit 47, pp. 25-26.) "Second, critical systems, such as CSS, were implemented during this time frame that required incremental application maintenance expense to support these systems on an ongoing basis. The increase in O&M expenses reflects our consolidation of IT application and development resources that were previously decentralized in our internal Business Units in the 1995 GRC." (Id., p. 26.)

In its testimony, SCE describes three main areas in which the ratepayers benefit from its IT service department. These areas are:

SCE filed its IT expenses in an account-by-account format. The table and graph below show IT's recorded O&M costs for 1996-2000 by FERC account, plus SCE's forecast costs for 2001-2003. A visual analysis of this figure shows that SCE's IT O&M expenses have shown a trend in an upward direction, save year 2000, over three or more years.

Another issue is whether $42.987 million in Y2K project costs should be removed from the recorded data before a five-year average is calculated. SCE's argument is that it temporarily reprioritized other work when the Y2K project was underway, and that such work must now be done. ORA counters that Y2K expenditures should be removed because of the abnormal, nonrecurring nature of the project. ORA does not accept the contention that the deferral was temporary. In effect, ORA is saying that just as the Y2K project was undertaken in 1998, O&M expenses for other non-Y2K work were reduced by the same amount on a permanent basis. While such a reduction is not implausible, SCE has demonstrated that the deferral of work on governance activities such as company-wide IT planning, and other activities, is not sustainable going forward. We will not adopt ORA's proposed Y2K adjustment.

As shown in Figure I-1 of Exhibit 47, recorded IT O&M expenditures during 1996-2000 totaled $539.606 million. Neither ORA nor any other party contested the reasonableness of SCE's individual account estimates for IT O&M expenses. Indeed, parties chose to aggregate the account numbers in lieu of scrutinizing certain accounts. SCE presented an account-by-account analysis of its 1996-2000 recorded and 2003 forecast IT expenses, with a separate analysis of the labor and non-labor components of each account. Because of the upward trend exhibited by SCE's IT O&M expenditures, we will approve the year 2000's actual recorded data of $111,835 as SCE's appropriate test year estimate. This approach is consistent the methods the Commission has endorsed in other rate cases.56

5.5.3. IT Capital Expenditures

As set forth in its direct testimony (Exhibit 47, p. 3), SCE estimated cumulative IT capital expenditures of $148.334 million over the period 2001-2003 ($33.870 million for 2001, $47.062 million for 2002, and $67.402 million for 2003). These expenditures consist of (1) upgrades and replacements of personal computers, servers, and mainframe ($66.352 million), (2) upgrades and replacement of wires, communications, and support equipment ($35.145 million); (3) other major projects ($25.630 million); (4) minor projects and blanket purchase orders ($14.614 million); and (5) a Disaster Recovery Program ($6.593 million).

ORA took issue with SCE's proposed capital expenditures in the Blanket Budget Items category only, and only for the year 2003. ORA used average annual recorded capital expenditures for 1996-2000, excluding $25.107 million for Y2K related expenditures, to arrive at a forecast of $49.642 million in 2003, or $17.759 million less than SCE's IT capital expenditures forecast of $67.401 million for 2003.

Discussion

As with the forecast of IT O&M expenditures, the issue of IT capital expenditures involves disputes over whether to exclude Y2K related capital expenditures from recorded data, and whether to use a budget based approach or an average of recorded/adjusted data to forecast 2003 capital expenditures. Since the issue pertains to Blanket Budget Items, we will focus on that category of capital additions. Blanket Budget Items are routine, ongoing capital projects, with no single completion date, that provide for the expansion, replacement, and upgrades of the IT infrastructure.

ORA's case for excluding Y2K related costs from recorded capital expenditure data is no more compelling than it was for excluding Y2K related expenses. Although SCE was able to temporarily delay a portion of these ongoing projects to offset Y2K costs, continuing to delay this work is not sustainable, and could impact the company's ability to serve ratepayers. We will not adopt ORA's proposed Y2K adjustment to recorded IT capital expenditures.

As discussed in Section 3.1.1.2 of this decision, an averaging approach to capital expenditures estimates requires correction to account for the effects of inflation. A five-year average of inflation adjusted IT capital expenditures would yield $46.669 million. Although this amount is less than SCE's requested amount of $52.646 million, we will nonetheless adopt the five-year average, as it is consistent with our averaging methodologies.

5.5.4. Y2K Retention Bonuses

SCE paid $547,965 in retention bonuses in 2000 to six employees involved in Y2K project. ORA contends these costs are non-recurring and should be removed from SCE's recorded base. Since 65% of this cost was expensed and 35% was capitalized, the recorded base for expenses (Account 920) would be reduced by $356,177 and the recorded base for capital (Account 101) would be reduced by $191,788.

Discussion

SCE contends that in recommending this adjustment, ORA implicitly assumes that SCE will not have any additional projects, like the Y2K project, requiring retention of critical IT talent with specialized technical skills needed for future IT projects. However, SCE has not demonstrated that it is necessary to expect to routinely pay certain IT employees an average bonus of nearly $100,000 for unspecified future projects. ORA's proposed expense and capital adjustments to remove the costs of these bonuses will be adopted.

5.5.5. IBM Charges

SCE paid $112,388 to IBM Corporation for professional services rendered to assist in the initial development of an e-business strategy that set the tactical and strategic direction and overall vision of the IT business unit. ORA recommends that this expense be removed from Account 923 because the service was performed once and is nonrecurring.

Discussion

The e-business strategy originally developed with this work by IBM is still evolving as with all internet related initiatives. SCE will incur as much as $112,388, if not more, in the 2003 Test Year for this initiative to refresh these strategies and determine what revisions are necessary due to industry changes that will occur in the e-business environment. ORA's proposed adjustment for IBM charges will not be adopted.

5.6. Capitalized Software

Under the terms of a settlement agreement adopted by D.92-11-051, SCE was authorized to capitalize application software development costs amounting to $100,000 or more for projects with a useful life of at least five years. SCE capitalizes only software projects over $1 million. SCE forecasts a total capitalized software rate base for the test year of $222.982 million.

SCE requests $31.862 million in direct additions for 2002 and $40.754 million in additions for 2003. ORA accepts SCE's request for 2003 but recommends removal of $860,000 from 2002 additions for contingency for SCE's Outage Management System Enhancement (OMSE).

Discussion

SCE included a $1.28 million contingency in the OMSE estimate. Of this amount, $420,000 was for unforeseen hardware (capacity and storage) needs, which ORA finds reasonable. ORA takes issue with the remaining $860,000, alleging SCE has not provided adequate support. Reviewing an article by William H. Roetzheim, author of the book Software Project Cost and Schedule Estimating - Best Practices, who SCE relied upon in part to justify its proposed contingency, ORA concluded that SCE's rationale for including the $860,000 was not persuasive. ORA believes that Roetzheim has indicated that contingencies can be positive or negative.

SCE has shown that the $860,000 contingency for software was based on industry standards and best practices for application development, which indicate including a contingency in budgeting for capitalized software projects. This contingency is actually 8.4% of the total software costs. When the OMSE contingency was developed, the project was in the design phase, for which the industry literature indicates a 10% contingency is appropriate. ORA's proposed reduction will not be approved.

5.7. Human Resources (HR)

5.7.1. HR Departmental Costs

SCE forecasts $4.823 million in A&G expenses in Accounts 920, 921, 923, and 926 for the Total Compensation Division of the HR Departmental Organization. This is composed of $3.206 million for labor and $1.617 million for non-labor. Within Account 920/921, SCE forecasts non-labor expenses of $63,000, based on a three-year average (1998-2000). ORA proposes a $33,000 reduction to the Total Compensation Division's non-labor expenses in FERC Account 921, based on the last recorded year's expenses.

Discussion

The recorded non-labor amounts in Account 921 for the Total Compensation Division are shown in the following table:

Total Compensation Division

FERC Account 921 Non-Labor

(Constant Dollars X 1000)

SCE selected a three-year average of $63,000 after determining that a four- or five-year average would overstate expenditures due to the inclusion of 1996 and 1997 costs which reflect higher than anticipated use of supplemental workers. ORA believes that the 2000 recorded spending level provides the best forecast because costs have been declining every year and because ORA did not believe that SCE provided adequate documentation for the forecast increase of $33,000 above the year 2000 level.

We find SCE's proposal to use a three-year average to be reasonable because it normalizes for yearly variations. The last recorded year is the lowest value of the five years from 1996 to 2000, but that alone does not indicate that it is a reasonable predictor of test year expenses. We therefore adopt SCE's forecast.

ORA made a normalized adjustment of $53,000 to Account 921 to remove the costs of dues and memberships that, ORA contends, do not relate to the utility business. The dues and memberships at issue are for such organizations

as the Corporate Executive Leadership Board, the Academy of Business Leadership, the Human Resource Planning Society, the American Statistical Association, and Inroads, a trade and technical association. Because of the use of a four-year average, this results in a proposed $13,000 reduction to FERC Account 921.

Discussion

This dispute turns on whether the dues and memberships at issue are related to the provision of utility service. In SCE's 1995 GRC decision, the Commission addressed dues and memberships in Account 921 as follows:


We have a long-standing policy not to allow recovery in rates of dues to chambers of commerce and service clubs. In Pacific Tel. & Tel. Co. v. Public Util. Comm. (1965) 62 Cal.2d 634, 669, the California Supreme Court upheld this policy. We apply this policy here and grant DRA's recommendation not to fund $82,000 for chamber of commerce dues. We also concur with DRA's second recommendation [not to fund $19,000 for certain membership dues] because Edison did not meet its burden of proof in demonstrating how these organizations relate to the utility's business and offer ratepayer benefits. (D.96-01-011, 64 CPUC2d 241, 316.)

Thus, it should have been clear to SCE that it needed to sustain its burden of proving that the dues and memberships at issue offer ratepayer benefits. SCE's brief does not point to any portion of SCE's direct showing where such proof is provided. SCE focuses on the fact that it is not seeking ratepayer reimbursement for chamber of commerce or service club dues, but the point of the foregoing passage is not that all memberships and dues are recoverable from ratepayers so long as they are not for chambers of commerce or service clubs. SCE must show that any memberships or dues for which rate recovery is sought offer ratepayer benefits. SCE's rebuttal testimony (Exhibit 302, p. 120, line 1 through p. 124, line 4) addresses this issue, but as noted elsewhere this is improper litigation, and we therefore strike that testimony.

Because SCE has failed to meet its burden of proving that the memberships and dues at issue offer ratepayer benefits, ORA's proposed reduction of $13,000 is adopted.

SCE records outside services costs associated with strategic and operational change initiatives to FERC Account 923 in the HR Service Center. SCE used a four-year average of 1997-2000 recorded amounts to estimate test year expenses of $427,000 for the HR Service Center in Account 923, whereas ORA proposes a five-year average of the years 1996-2000 to arrive at a forecast of $342,000. This results in a difference of $85,000.

Discussion

The recorded non-labor amounts in Account 923 for the HR Service Center are shown in the following table:

HR Service Center

FERC Account 923 Non-Labor

(Constant Dollars X 1000)

In 1996, SCE implemented a company-wide workforce reduction effort and HR curtailed activities relating to strategic and operational change initiatives. Moreover, 1996 was a start-up year for the HR Service Center. We conclude that the test year forecast should not be based on consideration of 1996 data. ORA's proposed reduction will not be approved.

SCE used a three-year average to estimate costs for outside services for executive in Account 923 to arrive at a forecast of $1.066 million. ORA used a five-year average to arrive at a forecast of $787,000, which is a proposed reduction of $279,000. ORA noted that SCE did not substantiate how ratepayers benefit from an increase in outside services for executive studies and compensation statements that "increase executive understanding and appreciation of the compensation benefits they receive."

Discussion

Outside services that help executives understand and appreciate their compensation benefits are, at best, of questionable value to ratepayers. We note that SCE does not contest ORA's proposed reduction, which is hereby adopted.

5.7.2. Employee Compensation Issues

In accordance with Commission direction in prior GRCs (D.87-12-066, D.91-12-076, and D.96-01-011), SCE and ORA jointly selected an independent expert, Hewitt Associates, to perform a total compensation study. SCE and ORA jointly managed the study. In the context of this study, "total compensation" represents cash (including base salaries and annual incentive compensation) and non-cash compensation (i.e., pensions and benefits) received by SCE's workforce.

Based on samples of job categories representing 41% of SCE's workforce of 12,818 employees, the study shows that SCE's aggregate compensation is 4.3% above market levels. However, the sampling error is plus or minus 5%. The Hewitt study results are summarized in the following table:

Summary Results of Total Compensation Study

(Percentage above/below market)

ORA and Greenlining addressed aspects of the total compensation study in their opening briefs, as described below.

Discussion

The submission of the total compensation study comports with prior Commission directives. We appreciate SCE's and ORA's cooperative efforts in this respect. Since SCE's total compensation is shown to be 4.3% above the comparable market total compensation, and the study margin of error is plus or minus 5%, we conclude that SCE's total compensation for all employees is equivalent to the market level.57 We will not make a ratemaking adjustment on the basis of above market total employee compensation.

ORA states that because the total compensation study shows that SCE's overall total compensation is 4.3% above market, and ORA does not object to such above market payments to labor, this represents a major concession to SCE worth millions of dollars. However, as noted previously, we cannot conclude with certainty that SCE's compensation, on an overall basis, is above market because of the margin of error in this study.

Greenlining, which indicates it is concerned with excessive executive compensation, believes that the total compensation study is of limited value with respect to executives because, according to Greenlining, it is predicated upon comparisons with inflated compensation packages. We believe Greenlining misunderstands the use and purpose of the study. Our interest is in assuring that SCE is positioned to provide safe, reliable public utility service to its customers. This requires that SCE be able to attract, retain, and motivate qualified employees, including employees in the management and executive ranks, which means it needs to be a competitive employer. We look to market comparators to determine the level of employee total compensation needed to provide reasonable assurance that SCE is a competitive employer that is not providing above-market salary/benefit packages that would add unnecessarily to ratepayer costs.58 The total compensation study allows us to do that. The executive compensation of comparator firms may be considered by some observers to be excessive according to their standards of societal worth, but it is not our objective to define such notions of excessive compensation, or to apply our own values of societal worth as a substitute for market comparisons.

SCE and ORA, along with Hewitt, have fulfilled our directives in earlier decisions, and have provided the value we expected with respect to the total compensation study. Any contention that the study is flawed because it uses comparative values, or because the comparative values used are excessive, is without merit.

Greenlining also takes issue with the total compensation study's value because "it ignores the information that we have recently discovered from the Jack Welch case at General Electric and the John W. Snow case at CSX about unreported compensation unknown to even the [Securities and Exchange Commission (SEC)], shareholders, or board members." (Greenlining opening brief, p. 5.) However, there is no evidence in this proceeding that SCE's executives are receiving unreported compensation, that any unreported compensation that does exist is included in the GRC request, or that the Hewitt study failed to consider all executive compensation relevant to ratemaking. The media reports of problems with unreported executive compensation do not undermine the evidentiary value of the Hewitt study for our purposes, even if, as Greenlining witness Phillips testified, "the entire financial institution of the United States has been shaken to its core to find out that annual reports and filings with the SEC can exclude very significant parts of the executive compensation..." (Tr. V. 38, p. 3448.)

Finally, Greenlining takes issue with the total compensation study process, calling it "highly secretive." We do not find any lack of transparency to be problematic. The study methodology is extensively outlined in Exhibit 77. The study was jointly managed by SCE and ORA, which contradicts contentions of undue secrecy.59 Even if Hewitt has proprietary intellectual property that was utilized in the study, we find no basis for concluding it was in any way flawed because of the protection of such property.

Greenlining asks that we adopt a requirement that SCE report annually on the total compensation packages for each of the top ten executives, including the value of stock options and retirement plans. We are not persuaded of the merits of such a requirement at this time, and therefore decline to adopt the proposal. Greenlining also asks that we adopt a requirement that SCE report annually on the total compensation packages for each of the top ten executives, including the value of stock options and retirement plans. This information would be instructive as part of the Commission's effort to better understand the nature of SCE's executive compensation practices. We will adopt this requirement.

SCE's request for executive salaries in Account 920 includes the costs of incentive compensation received by executive officers. Based upon its position that ratepayers and shareholders should contribute equally to the costs of executive bonuses, ORA recommends that SCE's forecast be reduced by one-half the costs of such bonuses, or $2.358 million. SCE opposes this recommendation.

Discussion

SCE states that the total compensation study shows that total

compensation for the executive category is 1.1% lower than competitive market norms. As a preliminary matter, we note that in making this statement, SCE commits the same error it accused ORA of committing when ORA said that SCE's overall total compensation is 4.3% above market. If the study's margin of error does not permit ORA to make such a statement, it also does not permit SCE to claim that its executives receive less than competitive norms. In fact, we cannot rule out the possibility that SCE's executives receive above market total compensation. Nevertheless, neither the total compensation study nor any other record evidence provides us with any basis for concluding that SCE's executive total compensation, including base salary, incentive pay, and non-cash compensation, is unreasonable. At issue is which of differing policies regarding ratepayer funding of incentive programs that have been applied in the past should be pursued here.

The approach favored by SCE is that adopted in GRCs of PG&E and SoCalGas in 1992 and 1993, respectively. In D.92-12-057, the Commission noted the following conclusions of a workshop conducted by the Commission staff:


"The consensus reached in the workshop was that the Commission should not attempt to micromanage utility incentive compensation programs. Instead of adopting a `cookie cutter' approach, workshop participants recommend that the Commission review incentive compensation programs utility by utility, as a component of the total cash compensation requested in each utility's general rate case. They proposed, moreover, that the allocation of total cash compensation between salaries and incentives should be left to each utility's discretion.&