7. Rate Base

7.1. 2002 Plant Additions

Test Year 2003 Distribution and General Plant-in-Service is based on recorded capital expenditures through 2001 and estimated 2002 capital additions. In its filed testimony, Southwest proposed 2002 plant additions totaling approximately $29,391,129 while ORA proposes 2002 additions of $17,649,029⁵³. As we note below, one reason for the difference between Southwest and ORA's proposal is attributable to pipeline replacement rates. In comments to the ALJ Proposed Decision ORA offered a compelling rationale for adopting actual recorded 2002 plant additions rather than Southwest's forecast expenditure level. We note that the recorded 2002 plant additions falls within Southwest's and ORA's proposal. As such, we adopt Distribution and General Plant-in-Service 2002 additions of $25,920,403.

7.2. Distribution Plant-Pipeline Replacement
Project

Almost all of the differences⁵⁴ between Southwest, ORA, and San Bernardino with respect to gas distribution plant-in-service, are attributable to different recommendations regarding the rate at which polyvinyl chloride (PVC) mains and services should be replaced in the Southern and Northern California Divisions. All PVC services have been replaced in Northern California. Thus, the replacement program includes mains in Northern California, and PVC services and mains in Southern California.

Southwest, ORA, and San Bernardino agree that eventually all existing PVC pipe mains and services must be replaced. However, parties disagree over the period necessary for pipe replacement, and amounts proposed for test year 2003, and the attrition years. ORA proposes to replace the pipe uniformly over a 20-year period; the County recommends a 25-year replacement period, while Southwest proposes to replace PVC pipe on an accelerated basis during the test year, and attrition years, and then scale back replacement over an additional 15-years.⁵⁵

7.3. Background

Southwest installed PVC pipe in the late-1950s, 1960s, and 1970s. Beginning in 1997, Southwest embarked on an accelerated PVC pipe replacement program for its PVC mains and services, using polyethylene (PE) pipe not subject to the problems encountered with PVC. Southwest conducted a Pipeline Integrity Assessment (PIA) to determine the factors that might cause mains and services to fail, and planned a process to address these factors. Southwest notes that, as a result of the PIA, Southwest spent $19.2 million on pipeline replacement between 1998 and 2001. Southwest estimates it will spend an additional $23.6 million in 2002 and 2003, and $36.6 million between 2004 and 2007.

ORA and San Bernardino contend there is little justification for the proposed increased rate of PVC pipe replacement. ORA asserts that a 1993-97 study of leak rates, measured in leaks-per-mile, indicate that there is no upward trend in leak rates, and that a review by a Commission certified pipeline safety inspector confirms leaks are negligible. San Bernardino, relying on analysis of leak rate data for the period 1987-1999, agrees. ORA adds that PG&E has a greater leak rate than Southwest, nevertheless, PG&E has operated under a 25-year pipeline replacement program, and that a PVC laboratory analysis confirms that piping materials have generally retained their design integrity.

San Bernardino argues that under Southwest's "scoring" system, that awards scores to particular pipelines using the PIA, a score of 77 constitutes a need for replacement in Northern California, while only a 65 is needed to justify replacement in Southern California. San Bernardino contends that applying a score of 77 to Southern California would result in reducing the amount of replacement piping to about 36% of the amount requested by Southwest for Southern California. San Bernardino and ORA question Southwest's justification for the separate criteria for Northern California and Southern California.

Finally, San Bernardino notes that Southwest proposes to replace more than 300,000 feet of additional replacements in Southern California in 2002 but provided no justification for these additional replacements. Further, although the company proposes to continue pipeline replacement spending at 2002 levels beyond 2002, San Bernardino notes that the company has not conducted a PIA assessment for 2003 and subsequent years.

Both ORA and San Bernardino contend that a longer replacement period will mitigate the rate impact of the PVC pipe replacement and allow for better coordination of joint trenching operations for main replacements. San Bernardino suggests a 25 year replacement program, consistent with the time frame we have adopted for PG&E, and ORA recommends a 20-year replacement program. ORA and San Bernardino each believes their respective recommendation will provide sufficient revenues for Southwest to focus on the most problematic portions of its system.

ORA also recommends that Southwest be required to file an annual progress report on its PVC replacement operations.

Southwest argues that the positions of ORA and San Bernardino are limited in scope, and therefore their conclusions are faulty. Southwest states that many factors, other than leak rates, determine the need for PVC pipe replacement.⁵⁶ Southwest points to later information on leak rates between 1997 and 2001, and the types of leaks,⁵⁷ that support an aggressive replacement program. Specifically, Southwest notes that Grade 3 leaks have increased dramatically since 1999. ORA responds that the more recent leak rate data reveals that the Grade 3 main leaks are concentrated in two areas within Southwest's service territory: Barstow and Big Bear. With regard to mains in Barstow, and services in both Barstow and Big Bear, ORA observes that between 80-90% of the pipe will have been replaced by the end of the test year, raising the point that the problems in those areas will be solved and is not a justification for the level of expenditures beyond the test year. With regard to main leaks in the Big Bear area, where leak rates rose to 0.48 leaks per mile, ORA wonders why Southwest nevertheless proposes to focus its main and service replacement program in Victorville for the test and attrition years, where main leak trends reflected 0.1 per mile, and were flat for service leaks.

Finally, Southwest argues that its PVC pipe replacement program is not comparable to PG&E's program. Southwest states PG&E's program replaces steel and cast iron pipe, and not PVC pipe, and that steel and cast iron pipe tend to suffer from pinhole leaks, not the type of major breaks at joints found in PVC pipes. ORA maintains that Southwest provided no evidence to show that catastrophic leaks are greater with PVC pipe than with other piping material, such as PE and steel, or that federal or California state rules for replacement of PVC pipe have changed. Southwest points out that there is no current industry standard, or regulatory federal or state standard for acceptable leak levels, and that ultimately there should be zero leaks.

7.4. Discussion

In other proceedings, we are often asked to encourage utilities to maintain, repair or replace existing plant. In the instant proceeding, it is not a matter of encouraging or directing Southwest to maintain its system, or whether the aging PVC pipe must be replaced. Parties agree that the PVC pipe, portions of which are over 40-years old, must eventually be replaced. The question before us is how quickly the PVC pipe should be replaced.

We agree with Southwest that reducing leak rates to zero is a laudable goal, or alternatively, leak rates should be minimized. Thus, we acknowledge the importance of reducing leaks in PVC mains and services in order to avoid accidents and improve safety. The record indicates, however, that is it not necessary to adopt the accelerated replacement program suggested by Southwest in order to provide sufficient revenues for Southwest to address those safety issues and focus on the most problematic portions of its system.

First, as we weigh the options for how best to achieve those goals, ORA's detailed regional analysis of mains and service leak data, and pipeline replacements since 1999, is particularly instructive. As ORA notes, most of the mains and service pipeline replacements in the two areas within Southwest's service territory (Barstow and Big Bear) that have experienced significantly more Grade 3 leaks since 1999, were completed by the end of the test year. Now that most of investment to address these trouble spots has been completed and Southwest focuses its efforts on areas where fewer increases in mains and services leaks have been observed, we believe it is reasonable to expect that the expense levels associated with this relatively accelerated work in 2001-2002 should flatten out. With that expectation in mind, we are not comfortable adopting rates going forward that reflect an accelerated replacement program for PVC mains and services as advocated by Southwest gas.

Second, like San Bernardino, we question the level of test year pipeline replacement expenditures based on the different criteria Southwest appears to apply to Northern and Southern California. Southwest has not adequately explained on this record why a PIA score of 77 (or greater than 2.0 leaks per mile) in its Northern California districts motivates a decision to replace pipe, while in Southern California its replacement strategy is governed by a PIA score of 65 (or an average of 1.64 leaks per mile) or greater. As San Bernardino notes, applying the more strict criteria of 65 to Southern California under the accelerated replacement program results in approximately 179% more miles of pipeline replacements in 2002 than would have been necessary under the Northern California criteria. Southwest has offered no basis for the difference. Furthermore, Southwest has not conducted a PIA for 2003 and beyond, although the rates reflecting the accelerated replacement program would continue relatively elevated funding. Absent an understanding of why replacement decisions are made based on different criteria between Northern and Southern California districts, and an analysis of the need to continue accelerated replacement beyond 2002, we cannot conclude that the accelerated replacement program will result in reasonable rates going forward.

We believe Southwest should spread its pipeline replacement program over a period of 20 years, as ORA suggests. This time frame results in replacing approximately 140,000 feet of main and 73,000 feet of services pipe per year, at a cost of $3,449,550. We believe this is adequate to allow the company to focus its expenditures on the most problematic portions of its system. Given the impact of the replacement program on Southwest's revenue requirement, this approach will mitigate the rate impact on the ratepayers in Southwest's Southern California divisions -- most of whom are still recovering from the significant rate shock of the last two winters' high natural gas bills.⁵⁸ A longer replacement period will also allow for better coordination of main trenching operations with other utilities in the area.

7.5. Pipeline Replacement Program (Northern California)

Although the majority of the PVC pipeline replacement is in Southern California, differences also exist between ORA and Southwest over the replacement schedule for PVC mains in Northern California. Most of the services installed in Northern California are non-PVC pipe, and therefore do not require replacement. Southwest proposes a 15-year replacement schedule for PVC mains that would replace about 30,000 feet per year. ORA proposes a 20-year replacement schedule that would replace about 21,000 feet per year of mains.

In adopting a replacement program for Northern California, unlike Southern California, there is little information on leak rates. Since Northern California services have already been upgraded, it appears that the overall problem is less severe. We will adopt a pipe replacement program that provides adequate main replacement in consideration of these factors. Our adopted Southern California PVC replacement program is over 20 years. For Northern California, we will adopt a similar program to replace PVC mains at a rate of 21,000 feet per year, at a cost of $533,000 per year. In the next GRC, we expect that Southwest will provide further information on the leak rates experienced in Northern California mains, and make recommendations on further PVC replacement, including changes in the proposed rate, as necessary.

We also adopt a reporting requirement similar to PG&E's reporting requirement for its gas pipe replacement program. We direct Southwest to provide an annual progress report to the Commission regarding the pipeline replacement program as shown in Exhibit 103. The report should include, among other items, the footage of mains and services replaced, costs, and information on leak rates.

7.6. Working Capital

7.6.1. Materials and Supplies

Southwest calculated materials and supplies (M&S) utilizing a five-year average of 13-month inventory balances divided by gas plant in-service. This ratio is then multiplied times plant-in-service balances to estimate M&S. ORA used different estimating methods for M&S including a four-year average and an adjustment for reduced PVC pipeline replacement. Southwest disputes the M&S reduction in Southern California and contends that a reduced pipe replacement program will actually increase M&S balances as a result of increased need for repairs.

We will adopt Southwest's estimate of M&S for Southern California decreased by 15%. Southwest did not dispute ORA's M&S adjustment for Northern California; thus we will adopt ORA's Northern California M&S estimate.

7.6.2. Working Cash

Southwest and ORA developed working cash estimates using the lead-lag methodology in Commission General Order U-16. However, ORA estimated revenue lag at 40-days, while Southwest used 44-days. ORA argues that its 40-day estimate is reasonable since it is based on historic lead-lag days. ORA also asserts, that due to expiration of high gas cost contracts, it is expected that customers will pay their bills more quickly, thus reducing revenue lag. Southwest, however, points out that during an ORA on-site audit in March 2002, ORA was informed revenue lag had actually increased to 46.7 days. We note that although gas costs may have decreased, and higher priced gas contracts expired, given the current economic climate it is unlikely that customers will pay their bills any more quickly. Thus, we will use Southwest's 44-day revenue lag in our calculation of working cash that is also more reflective of current information.

Southwest and ORA also differ in estimates for income tax payments' lag days used in the working cash calculation. Southwest based its estimates on statutorily mandated filing dates. ORA based its income tax lag days on a proxy reflecting income tax payments for Edison and PG&E. ORA alleges that in Southwest's last GRC, Southwest made an incorrect assumption on the payment of income taxes, and therefore a proxy is appropriate. Southwest states it provided the actual timing of income tax payments to ORA in response to a data request, and in response to questions from the ALJ.⁵⁹ Southwest's response states that actual federal income tax lag days were a negative 224 days for 2001, while state income tax lag days were a positive 39 days. We will not adopt these actual income tax lag days as they are an apparent anomaly resulting from unusual gas prices. We will not adopt ORA's proxy based on the tax payments for other utilities, as there is not evidence that this proxy should apply to Southwest's tax payments. Instead, we will adopt Southwest's initial estimate of tax lag days based on the statutorily mandated tax payment filing dates.

7.7. General Plant

Southwest and ORA initially disagreed over two general plant issues, Miscellaneous Intangible Plant (Intangible Plant), and 2001 Construction Work in Progress (CWIP) balances; however they have agreed to defer the 2001 CWIP balances until the next GRC.

Regarding Intangible Plant, ORA made a 40% reduction to reflect questionable costs for a Southwest affiliate, Utility Partners (UP). ORA's analysis and recommended disallowance are included in its Audit Review, Exhibit 122. As discussed in Exhibit 122, UP⁶⁰ developed software for use by Southwest and other utilities. Although ORA initially found that Southwest violated affiliate transaction rules, ORA now acknowledges that an exemption for affiliate transaction rules applies to Southwest's relationship to UP. ORA explains that its 40% reduction reflects Southwest's equity share in UP, and is based on the reasonableness of UP project costs charged to Southwest already included in plant in service.⁶¹

Southwest in its rebuttal provides extensive insight into the relationship between Southwest and UP, UP's software projects, and how these projects are useful for improving Southwest operations.⁶² Southwest states its investment in UP involved only shareholder earnings and no ratepayer funds. Southwest also explains that although it had a preferred customer agreement with UP that began in 1999, that agreement provided third-party oversight for costs charged from UP to Southwest. Finally, Southwest asserts that ORA's calculations are in error, as certain UP projects are amortized, or not included at their full value in rate base for the test year, and that deferred income taxes cause additional changes. Southwest calculates that including these adjustments results in a reduction in ORA's proposed disallowance from approximately $8.5 million to $3.3 million.

7.8. Discussion

We will not further address the $30 million in software projects currently in CWIP, except to remind Southwest of its commitment to demonstrate in its next GRC that these project costs are reasonable before any of these dollars may be included in plant-in-service. As stated previously, Southwest bears the burden of proof that such costs are reasonable, and not ORA. Southwest argues that it has not been required in prior proceedings before this Commission or in any other regulatory jurisdictions to provide independent proof.⁶³ Each proceeding before us has its own unique circumstances, with findings determined on the facts applicable to the specific proceeding; thus, we will not excuse Southwest on the basis that a specific showing has not been previously required.

We will not adopt ORA's recommended disallowance regarding the remaining UP project costs. Southwest provided substantial information regarding the usefulness of each of these projects to the company's operations and as improvements to efficiencies. ORA concluded that the project costs were unreasonable because these costs exceeded budgeted amounts. However, many projects, especially those involving complex projects, such as information technology, may exceed their budgets. Thus, a final cost that exceeds a budget is not of itself a measure of unreasonableness. Furthermore, this measure of cost unreasonableness would have to be significantly reduced by the various credits for both fully and partially amortized projects, and deferred income taxes in weighing the value of a project against its cost.

7.9. Contract Escalation

Southwest agrees with ORA's allocation of capital expenditures to labor, non-labor and contactor services, but argues that the contractor services component should also be escalated by 1.75%, or 3% for the test year. We have reviewed Southwest's calculations and agree with a 3% increase in the contractor component.

7.10. Truckee Operations Center

ORA excluded the building and furniture costs of the Truckee Operations Center from its estimates of plant arguing that Southwest had failed to provide a definite plan to construct the operations center, obtaining a building permit, or selection of an architect. Initially, Southwest included the operations center in its 2002 plant in service, although it now expects to finish construction by the end of 2003. Southwest contends it has committed to build the operations center, and through its witness provided an update of the status of design and permit process.

As a result of the Truckee expansion project, and the need to have an operations center in the Truckee area, it is not a matter of whether Southwest will have an operations center, but a question of when. We have already included, with ORA's agreement, expenses for a new Truckee district manger, who will eventually require an office. Our review of the record indicates that although design and construction of the operations center has been delayed in the past, it appears that construction is likely during 2003. Therefore, although we will not include the costs for the Truckee Operations Center in the test year estimates, we will include these costs in plant for attrition year 2004.