17. Rate Base - Plant in Service

17.1. Recorded 2004 Plant Service

In its application showing, SCE uses the 2003 recorded plant balances as the starting point for determining the test year plant balances. SCE forecasts 2004, 2005 and 2006 plant additions in determining the test year 2006 beginning of year, end of year and weighted average plant balances. ORA recommends that the 2004 forecast of plant in service and accumulated depreciation and accumulated deferred taxes be updated for 2004 recorded data, thus providing a more recent starting point. SCE contends that plant should not be updated for 2004 recorded information, because the post test year ratemaking (PTYR) mechanism adopted by the Commission in D.04-07-022 gives SCE the opportunity to implement its authorized capital spending budget over a two year period (2004-2005). ORA's proposal to update for recorded 2004 capital additions along with its recommended 2005 capital additions would result in capital additions that are less than SCE's currently authorized capital additions for the post test years 2004 and 2005. SCE states its expectation that by year end 2005, it will have fully implemented the two year (2004-2005) capital budget approved by the Commission in D.04-07-022.

17.2 Discussion

In past GRCs, updating for more recent recorded information, especially for plant related items, was common. It was not unusual that the utility's forecast for the first estimated year (in this case 2004) would be different than that forecast in the application showing. There might be substantial differences between recorded and forecasted amounts despite the fact that the application was generally filed at the end of the year in question (in this case, SCE filed in December, 2004). The Commission has used updated recorded information in prior GRCs. For instance in A.90-12-018, the Division of Ratepayer Advocates (DRA, the predecessor of ORA) recommended a $162,649,000 plant reduction based on the use of recorded 1990 plant additions. SCE opposed the adjustment arguing, in part, that decreases in recorded plant may be offset by increases in forecast plant, as plant additions are deferred from the end of the recorded period (fourth quarter of 1990) to the forecast period (1991 and 1992). In D.91-12-076, we stated:

"We agree with DRA on this point. Although recorded and forecast plant additions do interact, as Edison claims, Edison's analysis ignores the likelihood that deferral of plant at the beginning of a forecast period will be offset by the deferral of plant additions at the end of 1992. Deferral of plant additions is not symmetric. It is more likely that forecast plant additions will be completed late than early. This is typical of construction projects, and may even be influenced by the perverse utility incentive to delay actual construction of new plant once it is put into rate base. We will adopt DRA's $162.649 million reduction."73

In this proceeding the difference between SCE's forecast for 2004 and the recorded amount is $118,045,000. However, compared to the conditions in A.90-12-018, the issue now is complicated by the PTYR mechanism adopted by D.04-07-022. By that mechanism, SCE was authorized plant additions for 2004 and 2005 based on its proposed budgets for those years, as presented in its 2003 GRC. Ratepayers are protected if SCE spends more than authorized for the 2004 - 2005 period in that rate recovery is limited to the authorized level. Ratepayers are also protected if SCE spends less than authorized, since the revenue requirement associated with any 2004 - 2005 forecasted additions that are not booked during that time period is subject to refund.74 Because of this, it is not appropriate or fair to incorporate 2004 end-of year recorded plant balance without somehow adjusting 2005 additions to consider that, under the PTYR mechanism, 2004 and 2005 plant additions are viewed as whole rather than separately. It is only because of the previously adopted PTYR mechanism that we will not adopt ORA's recorded 2004 plant balance adjustment in this proceeding.

However, for this GRC, it would be reasonable to consider the results of SCE's 2006 CAAM filing as it relates to both 2004 and 2005 recorded plant additions. Besides providing consistent treatment of recorded information used for the 2003 GRC, this would provide an opportunity to project the test year 2006 plant balances using the most recent recorded information, even more so than in previous GRCs.

For the 2004/2005 timeframe SCE was authorized gross plant additions amounting to $1,307,000,000 in 2004 and $1,143,000,000 in 2005.75 The PTYR total authorized plant additions for the two year period is therefore $2,450,000,000.

In this GRC, SCE forecasted gross plant additions of $1,262,000,000 for 2004 and $1,308,000,000 for 2005. The forecasted plant addition total for the two-year period is therefore $2,570,000,000, which is slightly higher than that previously authorized.

A final determination of the need to true up 2004/2005 plant additions through the CAAM will occur subsequent to this decision. In the meantime, for forecasting test year 2006 plant balances for this GRC, we will use SCE's forecasted plant balances through the end of year 2005. We will also establish a memorandum account to track the revenue requirement associated with recorded and SCE's forecasted 2004/2005 plant additions. When plant additions are evaluated for the CAAM, there are two potential outcomes.

The first potential outcome is that SCE records plant additions equal or exceed $2,570,000,000 for the period 2004 - 2005. In that case, no further action is necessary.

The second potential outcome is that SCE records plant additions that are lower than $2,570,000,000 for the period 2004 - 2005. In that case, SCE should credit ratepayers with the excess revenue requirement collected through this decision, that is the difference between the revenue requirement associated with the 2004/2005 plant additions forecasted in this GRC and the revenue requirement associated with the recorded 2004/2005 plant additions. The refund would be calculated from the effective date of this decision.

Although complicated, this process is necessary to avoid results in this GRC that would otherwise be inconsistent with the results of the 2003 GRC. It is difficult to rationalize potentially truing up 2004 and 2005 plant additions for recorded information in the CAAM and then ignoring that recorded information going forward through this GRC cycle. This process will not be necessary in considering plant additions for SCE's next GRC, since the PTYR ratemaking adopted for in this proceeding does not require the use of a CAAM.

Because of our use of the CAAM to determine the 2005 end-of-year plant balance for this GRC, issues related to 2004 and 2005 plant additions are moot. For this reason, as well as our resolution of the post test year ratemaking for 2007 and 2008 as discussed later in decision, we will only address plant addition issues that relate to the 2006 test year.

17.3 Plant Weighting Percentage

ORA proposes the adoption of the weighting percentage of 41.16% resulting from SCE's plant in service forecast for this GRC. ORA explains that this percentage is consistent with historical weighting percentages. SCE states that the weighting percentage is an informative ratio, indicating the amount of time the total annual additions are included in rate base. SCE argues that its forecast as to when projects are booked to plant should be found reasonable, not the resultant weighting percentage.

This issue was addressed in SCE's last GRC. In D.04-07-022, we stated:

"Notwithstanding SCE's claims that its method is more rigorous and sophisticated, and is based on the intimate knowledge of business unit managers, SCE has not demonstrated that rigor, sophistication, and intimacy yield more accurate and reliable forecasts than the historical record. SCE improperly attempts to shift the burden of proof to ORA in this GRC by pointing out that ORA provided no conclusive explanation of why an average of historical weighting percentages better represents the plant weighting than a detailed budget. The more pertinent question, not adequately addressed by SCE, is why its budget-based approach, which suffers from the problem that budgets are not always carried out as planned, is necessarily more accurate and reliable than data based on actual performance over an extended period."76

As discussed above, it is not uncommon for a utility to incorrectly estimate plant additions for the first of the forecast years, even though the estimates are made during the year in question. There is no evidence that any utility's ability to accurately forecast the timing of projects gets any better as the length of time related to the estimate increases (test year 2006 plant addition and timing estimates were prepared in 2004). Therefore, in general, we agree with ORA's proposition that a weighting percentage based on historical information is more reliable than that embedded in the utility's budget. However, in this case, the timing of projects as reflected in SCE's 2006 budget and a historical analysis of the weighting factors are apparently very close and ORA is recommending the use of the 41.16% weighting factor embedded in SCE's budget. This is also very close to the 42.554% weighting factor adopted for SCE in its last GRC.77 For this GRC, we will therefore use the embedded timing of projects as reflected in SCE's budget for the adopted projects to be included in rates in 2006.78

For purposes of forecasting capital additions in 2005 through 2008, SCE assumed there would be no short term debt available for construction activities when making its 2005 through 2008 forecast of AFUDC rates. SCE indicates that not all of its short term debt is available to finance construction activities. The majority is used to finance balancing account under-collections and fuel inventory. SCE states it would not oppose using a three-year historic average of short-term debt available for construction activities for computing the AFUDC rate instead of the value of zero. That amount would be approximately $17 million per year.

ORA recommends that the average short-term debt as a percentage of total capitalization, or 2.61%, be used to determine the short term debt to be included in forecasting the AFUDC rate for this GRC cycle. SCE estimates that ORA's recommendation would include up to $300 million of short term debt in the AFUDC calculation, depending on the year.

The full amount of short term debt cannot be used to finance construction activities, if there were other obligations for those funds. SCE's explanation that it only has a minimal amount of short term debt available for construction activities is convincing considering the large amounts necessary to cover balancing account under-collections and fuel inventory. Since, as discussed elsewhere in this decision, we decline to change the financing of fuel inventory from short term debt to the rate of return on rate base, we will assume an amount of short term debt for construction is available during this GRC cycle based on historic information. SCE indicates a three-year average of 2002 ($4.6 million), 2003 ($1.6 million), and 2004 ($43.4 million) would be acceptable. However, the more recent 2004 data better reflects SCE's return to financial health following the 2000/2001 energy crisis. We will therefore include $43 million of short term debt in the calculation of the AFUDC rate for this proceeding.

TURN recommends that SCE include, as a reduction to the plant in service forecast, allowances for costs transferred from CAC to CIAC. While SCE reflects the transfer, on a recorded basis, through 2003, it does not reflect the transfer on a forecast basis. TURN's adjustment would reduce the 2006 weighted average plant in service by $2,619,000.

SCE states that it did not explicitly include the estimates for costs transferred from CAC to CIAC in its forecast of plant in service, but argues that it is an insignificant factor that adds no value to the Results of Operations forecast and appropriately was not included. SCE notes that just as there are insignificant factors that would result in a decrease to the plant in service forecast, there are factors that would result in an increase to the forecast. SCE states that there are numerous parameters that affect actual recorded capital additions, and given the complexity of forecasting the results of operation it is unreasonable to factor every minor parameter into a forecast.

SCE prepared its plant related forecast based on factors it felt were important and determined which plant related items were significant and which were not. Those determinations were reflected in the development of the Results of Operations model. It is reasonable for other parties to question such assumptions and determinations when they are used as bases for ratemaking purposes. We will include the adjustment as proposed by TURN. The adjustment is small but not insignificant when compared to some of the other issues discussed in this decision. Also, SCE indicates that there are a number of such minor adjustments that are not factored into its forecast or the Results of Operations model. If it is not already a part of its filings, SCE's future GRC filings should include a listing and description of all such adjustments to support the reasonableness of its actions.

In SCE's last GRC, the Commission adopted a 50-50 sharing between ratepayers and shareholders for costs associated with Spent Fuel Storage and Coastal Mitigation. The Commission stated that because it was reasonable to determine that ratepayers have made contributions to the cost of the SONGS Used Fuel Storage project as well as marine mitigation costs, but impossible to calculate the precise amount of that contribution, the fairest outcome was to assign equal cost responsibility for the remaining costs of the projects.79 It was impossible to calculate the ratepayer contribution, because between April 1996 and December 31, 2003, SCE recovered SONGS 2&3 operating costs through a fixed "cents per kilowatt hour" price mechanism identified as Increment Cost Incentive pricing (ICIP). ICIP prices were not tied directly to SCE's cost forecasts during that timeframe.

In this case, SCE has reflected the previous adjustment to only assign 50% of the cost to ratepayers through 2005 but did not reflect that sharing for 2006 or beyond. TURN recommends that the 50% sharing of costs between ratepayers and shareholders continue. However, because Edison's actual spending was somewhat less than its 2003 GRC forecast in 2004 and 2005, TURN proposes a reduction of $9,200,000 in test year 2006 to return to the 50% level previously adopted by the Commission and further reductions of $16,800,000 in 2007, and $6,900,000 in 2008. TURN recommends that these should be permanent disallowances, although in its next GRC, SCE should be able to true up the actual disallowance to actual spending in the historical years.

SCE states that the SONGS 2&3 ICIP mechanism did not include any specific list of capital projects to be completed during the ICIP period, and TURN's proposed additional disallowances of Marine Mitigation and Used Fuel Storage project costs are not warranted.

In D.04-07-022, the Commission found that ratepayers had already paid at least some of the costs of these projects and, because the ratepayer contribution could not be determined, there should be equal cost responsibility for remainder of the project costs. We are not persuaded to reject our previous finding that ratepayers have already made contributions to the SONGS Used Fuel Storage and Marine Mitigation projects through the ICIP rates. The only reason to deviate from the sharing previously established would be if the ratepayer contribution could be determined and directly reflected. In that vein, specifically in the event that the Commission chose to adopt a continuing disallowance, SCE developed a proxy for determining the maximum that ratepayers could have contributed during the ICIP period and the maximum adjustment that should be imposed. SCE attempted to tie the assumptions in the test year 1995 GRC to what was in rates during the ICIP period and compare that to what was recorded. SCE asserts that the difference would be the maximum adjustment that should be made.

There is merit to SCE's proxy approach. While not definitive,80 it provides a more objective basis for assigning costs that were paid by ratepayers during the ICIP period. We will adopt it for this GRC cycle and reduce the 2006 beginning-of-year SONGS plant balance by $22,600,00081 (100% share). SCE's share of the adjustment is $16,951.000.

As discussed previously in the section dealing with Mohave O&M costs, we stated our preference to assume a temporary shutdown scenario as recommended by ORA and to reflect SCE's forecasted O&M and capital additions associated with that scenario in the test year. SCE's adopted share of Mohave capital additions is therefore $2,517,000 for 2005 and $2,821,000 for 2006. As discussed previously, the related capital costs will be used to establish the temporary rate recovery of Mohave costs. Recorded costs associated with the temporary shutdown scenario will be entered into a two-way balancing account; and permanent recovery will be determined in a future reasonableness review.

SCE's forecasted test year rate base includes $1,545,000 for buttress repairs at Florence Dam. In SCE's last GRC, these repairs were included in D.04-07-022 as O&M costs, amounting to $800,000, that were expected to occur during 2003. According to SCE, when it implemented the Florence Dam Buttress repairs, the scope of work changed from that forecasted and the costs almost doubled to $1,545,000. This change, caused SCE to conduct another review of this project, to assure the proper accounting was being used. Commensurate with its capitalization policies and accounting guidelines, SCE determined that the Florence Dam Buttress repair project costs should be capitalized. The project was completed in 2003 and is included in the recorded plant balances used in this GRC as the base for projecting test year plant balances.

TURN recommends that the $1,545,000 of Florence Dam Buttress Repair costs not be recovered in rate base but that these costs instead should be deemed to be an O&M expense. TURN maintains that the costs of this project were already recovered from ratepayers in the test years and attrition years through the adopted 2003 O&M expense, and it would be unreasonable to recover those costs a second time through the capitalization of the same costs, as proposed by SCE in this rate case. TURN states that its recommendation is not retroactively adjusting 2003 results but is the result of a reasonableness review in this case of a capital expenditure that was specifically not requested or authorized in the last rate case.

SCE argues that the Commission should not retroactively modify the capitalization of these costs. SCE indicates that Hydro overspent its 2003 authorized O&M by $0.8 million and overspent 2003 authorized capital (on a direct expenditure basis) by $3.1 million. Since it spent more for hydro maintenance expenses than what was authorized in 2003 rates, SCE asserts that there is no possibility of double recovery of the Florence Dam Buttress repair project.

SCE also cites D.04-07-022, Finding of Fact No. 8, which states:

Capital spending budgets are not necessarily carried out as planned, as there is no specific obligation under conventional cost of- service or incentive ratemaking to spend budgeted amounts during the relevant time period . . . SCE requires flexibility to optimally respond to changing circumstances.

Normally, we take a fairly broad view when looking at what was included in rates and what was actually spent. The general concept of test year ratemaking is to authorize a rate level based on a reasonable forecast of various revenues and costs. Once rates are set, the utility has the discretion and responsibility to spend its funds in the most cost effective manner to proved safe and reliable service. However, in D.92-12-019, the Commission stated:

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. (D.92-12-019, 46 CPUC 2d 538, 555.)

SCE has provided information to show that, for hydro O&M expenses and capital expenditures, it spent more than what was adopted for test year 2003. However, inclusion of the Florence Dam project was the basis for setting rates for 2003, 2004, and 2005. If that project, which was specifically identified and justified by SCE and included in the adopted test year 2003 O&M forecast, were excluded from that adopted forecast, SCE would have received approximately $2,000,000 less than they actually received over the test year 2003 GRC cycle. SCE has not provided any information on its recorded hydro spending in 2004 or 2005.

We must also consider the ratemaking implications of changing the projected O&M expense to a recorded plant addition. There is an advantage to SCE in making this change, since in 2003 it overspent its hydro O&M budget. The costs of the Florence Dam project would not have been fully covered by rates in that year. By switching to capital, most of the costs will eventually be recovered, since the non-depreciated balance would be covered in rates going forward. SCE explained the decision to capitalize, in a data request response to TURN. SCE stated:

...As stated in the email, Remark #3 of the subject CPR catalog account had been clarified to read surfacing instead of facing in August 2003, as a result of inquiries made by Northern Hydro employees regarding this project earlier in 2003. Prior to that clarification, the remark had been interpreted to apply only to the upstream face of the dam. This special remark was added back in 1992 instead of creating a new retirement unit. Although the special remark originally referred to the "facing" of the dam, the reference meant the entire surface. Based on these criteria, added in 1992, the Florence Dam buttress surfacing qualified as capital.82

It appears that this project should never have been included in the expense forecast for the test year 2003 GRC. The data request response does not indicate that the scope of the project changed.83 In fact, the response indicates the original reference meant the entire surface. It was a mistakenly included as an anticipated maintenance expense because it was described or interpreted as "facing" rather than "surfacing". If the project had been classified correctly all along, there would be no dispute now. It would have been included correctly as a plant addition in 2003.

SCE should not benefit, just because it made a mistake in originally classifying this project as expense. For this GRC cycle, we will exclude the Florence Dam buttress repair as recommended by TURN. The beginning of year 2006 plant balance should be reduced by $1,545,000. Before the costs are included in any future rate case, SCE must provide convincing evidence that it did not benefit unduly by switching the project from expense to capital and sufficiently address the Commission's concerns expressed in D.92-12-019, as indicated above.

SCE's expenditure forecast for meters is based on the number of new customer meter sets time the cost per meter (CPM) set with cost escalation. ORA proposes the recorded cost experience by SCE in 2004 be held constant for 2005 and 2006 at $2,922 per meter. Multiplying this cost per meter by SCE's estimates for additional meters results in an ORA adjustment of $7,170,000 for test year 2006.

SCE states that there is no evidence that productivity or cost reductions will offset cost escalation associated with these activities. SCE calculates that adding the T&D capital escalation to the recorded 2004 CPM of $2,922 would yield a $3,010 CPM for 2005 and $3,100 for 2006, both of which are higher than its forecasts and concludes that ORA's analysis, when adjusted for inflation, actually corroborates the reasonableness of SCE's forecast CPM.

We note that TURN proposed an adjustment to this plant category, but after SCE fixed a discrepancy in the 2004 and 2005 CPM that affected the test year CPM, TURN no longer opposes SCE's forecasts. We will therefore only address the SCE and ORA difference. Regarding its position to hold the 2004 recorded unit costs constant, ORA asserts it is reasonable to expect that overtime labor, contract labor, and contract overtime labor costs will stabilize at current levels. While there is no specific evidence which quantifies productivity or other cost reductions that would offset cost escalation, ORA points out that due to the increased number of linemen from 2003 (647 linemen) to 2006 (828 linemen), overtime embedded in the 2004 recorded CPM would be reduced and would offset cost escalation associated with the other CPM activities. That 77,437 actual 2004 meter sets exceeded the forecast of 73,749 meter sets implies additional overtime would have been necessary to some degree. To the extent that overtime may be reduced due to the 28% increase in linemen from 2003 to 2006, it is reasonable to assume some cost savings to at least partially offset cost escalation. Since this is a labor intensive activity cost reductions resulting from reduced overtime may be substantial. We are persuaded by ORA's argument to hold the CPM at $2,922 for 2006 and will incorporate it in determining the test year estimate of $210,124,000 for this capital activity.

Regarding line extension allowances for existing customers, TURN recommends that line extension allowances for new panel upgrades should not be granted because they are not justified. Also, no line extension allowances should be granted for home remodels that do not entail an electric panel upgrade or for conversions to underground service. The necessary data was not available for TURN to adjust SCE's capital budget to exclude ratepayer funding for providing new services to existing customers. Therefore, TURN recommended the following:

First, the Commission should change the language contained in Section F.1.a.of Rule 16 concerning service reinforcements to the following:

SCE Owned. When SCE determines that its existing Service Facilities require replacement, the existing Service Facilities shall be replaced and the Applicant shall pay Edison its total estimated cost of replacement.

Second, if the Commission believes this unduly harms applicants that must have their services replaced for a panel upgrade or service reinforcement, it could treat these panel upgrades as a nonresidential service extension and require Edison to calculate the actual incremental revenues associated with a panel upgrade on a customer specific basis. Instead of receiving the full residential line extension allowance that is based on total average annual residential distribution revenues, this alternative would only credit applicants for their incremental distribution revenue.

In its direct testimony, TURN also raised several objections to the treatment of line extension allowances including the calculation of line extension allowances in general, the exclusion of sub-transmission costs in the calculation of line extension allowances, and the utilities' data collection practices regarding line extension costs and projects. However, in its opening brief, TURN suggests that, in light of Resolution E-3921,84 these issues should be removed from this rate case and the Commission should order SCE to revise its calculation of line extension allowances according to the modifications adopted in Resolution E-3921.

SCE notes that TURN's opening brief states: "While Edison's interpretation of its service reinforcements under Rule 16 may be technically valid, it eviscerates the spirit of the Commission's long standing policies to revenue justify new customer connections." SCE argues that TURN thus acknowledges that SCE is complying with the language of that tariff rule, and what TURN seeks is a change to that tariff. SCE states this is not appropriate in this proceeding for three reasons:

First, such a change would affect not only SCE, but the other California utilities that have similar tariff rules.

Second, the Commission's line and service extension proceeding already provides a forum to review residential line extension allowances.

Third, TURN itself has stated that issues surrounding line and service extension allowances "should be removed from this rate case." Since line and service extension allowances are intertwined with the operating language of SCE's Rule 15 and Rule 16 tariffs, TURN should raise this issue in that other proceeding, where it properly belongs.

Regarding line extension allowances for existing customers, SCE is in compliance with its current tariff language. We agree with SCE's position that the changes to Rule 16 may well affect other utilities and a generic proceeding would be the appropriate forum to make such changes. TURN's concerns regarding line extension allowances for existing customers should be brought up in SCE's A.05-10-019, which addresses residential line and service extension allowances. It is likely this application will be addressed concurrently with similar applications by SDG&E and PG&E.

SCE proposes a number of load growth projects (primarily in the form of new or expanded substations) to meet projected growth for customer load throughout its service territory. Also included in this category are capital expenditures necessary to interconnect new generating plants to the system. For this activity, SCE forecasts plant additions of $73,240,000 for 2005 and $84,532,000 for 2006. ORA forecasts plant additions of $56,571 for 2005 and 76,327,000 for 2006. ORA recommends postponing two projects indefinitely and deferring seven projects for one year. Because of the proposed deferrals, ORA also reduces the distribution substation program because the purchase of certain distribution circuits can likewise be deferred for one year.

Four of the projects85 in question are budgeted to go into service in 2005. As discussed earlier in this decision, we will be truing up estimates for 2005 to conform to the amounts authorized in D.04-07-022 through SCE's compliance with the CAAM. In the meantime, we are including SCE's estimates for 2005 in our decision today. For this reason, we will adopt SCE's request for these four load growth projects subject to adjustment for 2005 through the CAAM. We note SCE's admission that two of the projects - San Bernardino and Arrowhead may be of lesser priority than other projects required to avoid significant overloads. However, if these projects are ultimately completed in 2005, they should be recognized in rates as SCE has provided sufficient information concerning the prudence of the projects.

For the remaining projects ORA is recommending a deferral of one year, from 2006 to 2007. ORA concludes that there is a low probability of exceeding the utilization factors identified by SCE, provides an alternative approach to calculating projected loads, and assumes an overload capability above 100 % utilization. While there may be merit to ORA's analysis, it is clear that the projects need to be done in the near future. Whether or not certain events will coincide such that 100% utilization will occur or whether and by how much name plate ratings can be exceeded are secondary to the fact that the projects are needed and are needed soon.

In summary, we adopt SCE's request regarding the load growth projects, with the understanding that certain costs may need to adjusted as a result of SCE's upcoming CAAM analysis/filing.

SCE forecasts a total of $907,700,000 (2004 - 2008) for its Distribution Capital Replacement Program in order to address an increasing volume of infrastructure components wearing out and needing to be replaced. SCE indicates:

· The increased volume of pole replacements and repairs reflect increased levels of inspection performed in order to meet the requirements of General Order 165 and that performing fewer than forecast pole replacements and repairs will put it in non-compliance with regulatory requirements.

· An increase in the volume of preemptive replacements of underground switches and cable is necessary to deal with the increasing number of circuit interruptions due to failures of underground equipment, as well as to enhance public and employee safety.

· Old and obsolete automatic reclosures need to be replaced at a rate slightly less than achieved in 2000 and 2002 in order to manage system reliability and to enhance public safety.

· Capacitor banks need to be replaced at a rate slightly less than achieved in 2003 in order to provide adequate voltage to customers and ensure grid reliability.

· A modest number of underground vaults and manholes are forecast for replacement, because these are showing signs of weakening and potential collapse.

· Refurbishment of the worst performing circuits is necessary to move all customers toward the same level of service.

In general SCE has provided information that supports a need to replace certain portions of its distribution infrastructure at rates in excess of recorded rates. We will evaluate SCE's requests for the various aspects of its proposed infrastructure replacement program with that in mind. However, SCE still has the burden to justify the need and costs of each of its various proposed elements of the program.

Discussions relate to test year 2006 costs only. As discussed earlier in this decision, we will be incorporating 2005 recorded information into this proceeding via SCE's CAAM filing in 2006. In the meantime, we will include SCE's forecast of capital expenditures for 2005.

As discussed below, we have evaluated SCE's test year 2006 proposals and considered ORA's recommendations in developing the test year forecasts. For the amounts at issue, SCE requests $253,900,000 for 2006, while ORA recommends an amount of $80,500,000. We adopt a test year 2006 forecast of $188,814,000.

SCE states that the increased volume of pole replacements reflect increased levels of inspection performed in order to meet the requirements of General Order 165. Most of the pole replacements for 2005 and 2006 have already been identified and scheduled. Performing fewer than the forecast number of pole replacements will put SCE in noncompliance with regulatory requirements. SCE forecasts 14,900 pole replacements in 2005 with an expenditure of $116,000,000 and 14,800 pole replacements in 2006 with an expenditure of $119,300,000.

ORA recommends 9,512 pole replacements in 2005 with an expenditure of $74,500,000 and 6,499 pole replacements in 2006 with an expenditure of $52,4000,000. ORA states that its recommendation is consistent with recent historical pole replacement levels and costs, and includes poles with priority codes 1 through 4. It also takes into consideration a normalized level of intrusive inspections for years 2006 through 2008. It is ORA's position that SCE has not provided any reasons or data to support an increase over the historical replacement level in its Application.

According to ORA, SCE could not identify the number of poles replaced historically as a result of intrusive inspections although its 2005 and 2006 pole replacement forecast is based on the number of 2005 and 2006 intrusive inspections.

ORA argues that SCE was over-ambitious in its forecast for deferred pole replacement based on the number of priority code 3 poles scheduled for replacement in 2005 and 2006. According to ORA, since SCE only replaced 65 deferred poles in 2002 and 124 deferred poles in 2003, SCE's forecast of 3,769 deferred poles for 2005 and 2,332 deferred poles for 2006, appears to be excessive.

Regarding the number of poles due for replacement for years 2004 through 2008 that SCE claims necessary as a result of pole inspections required by Commission General Order 165, ORA states that SCE should have been cognizant of the requirements of G.O. 165 since 1997, and the company should have been replacing affected poles all along, not deferring the replacement work until 2005 and 2006 when the company filed its GRC Application. ORA points out, between 1999 and 2003, the company replaced an average of 7,500 poles each year and that for 2005 and 2006, SCE is forecasting a replacement level that is almost twice this number: 14,900 for 2005 and 14, 800 for 2006.

In response to ORA criticism of SCE's forecast of 3,769 deferred poles for 2005 and 2,332 deferred poles for 2006 as being excessive, SCE states, these pole replacements will not have been deferred but will be performed on time. SCE's rebuttal shows that most pole replacements are Priority 4 and will not occur until three-years after their inspection. According to SCE, these pole replacements are not discretionary, as ORA suggests, but necessary to comply with procedures written, in turn, to comply with GO 95.

Regarding compliance with General Order 165, SCE argues that GO 165 simply establishes a deadline by which all utilities must have completed their inspections. It says nothing about the rate at which these inspections must or should be performed. SCE further states that it has not been "deferring" needed expenditures. It has not been earning its authorized rate of return due in part to expending more capital than authorized. Regarding ORA's claim that SCE's forecast of pole replacements as a result of intrusive inspections excessive and unsupported, SCE states that the data ORA wanted had been archived as they were not relevant to SCE's day to day operations, (e.g., the priority code assigned to a pole which was replaced years ago.)

According to SCE, SCE based its pole replacement forecast on a detailed analysis of historic inspections and their results in terms of rejection rates by priority code. SCE (1) broke down the historical rejection rates by geographical location; and (2) determined how many would be inspected in that specific location from 2005 through 2006. SCE argues that this level of detail represents the most reasonable forecast possible.

As discussed earlier in this decision, we will be incorporating 2005 recorded information into this proceeding via SCE's CAAM filing in 2006. In the meantime, we will include SCE's forecast of pole replacements and costs for 2005. Regarding estimates of pole replacements for 2006, ORA has proposed an alternate forecasting methodology that is consistent with historical levels of pole replacements and costs. However, ORA has not explained, under its proposal, if or how SCE can meet pole replacement requirements identified as a result of General Order 165 inspections. SCE has provided information on the number of poles identified for replacement by priority, both as a result of past inspections and forecasted replacements based on future inspections. Considering the General Order 165 requirement that all wood poles over 15 years, which have not been subject to intrusive inspection must be intrusively inspected within ten years,86 SCE's estimates appear generally reasonable. For ratemaking purposes, rather than assuming replacement of 14,800 poles for 2006, 11,134 poles for 2007 and 11,160 poles for 2008, we will use the average of 12,365 poles for each of the years. Use of SCE's proposed unit cost of $8,060 for 2006 results in our adopted test year 2006 estimate of $99,659,000 for the wood pole replacement program, as opposed to SCE's estimate of $119,300,000.

SCE is requesting $12 million in 2005 and $28 million in 2006 for the preemptive replacement of underground distribution switches and fuse cabinets. Switches are used for opening or closing electrical circuit connections. SCE states that its "planned annual replacements rely heavily on judgment."87

ORA states that since the 1990s, SCE has been aware of problems with Buried Underground Residential Distribution (BURD) switches and mainline switches and has been replacing them preemptively over the past few years. Between 2000 and 2004, with the exception of 2001 when zero switches were replaced, SCE has been replacing switches at an average rate of 69 switches per year under this program. ORA concludes that SCE has not provided any justification to deviate from past replacement levels.

Based on a lack of data available to support an increase in the replacement rate over historical levels, and the fact that switch failures have been an on-going issue, ORA recommends continuing the level of replacement that SCE has been performing most recently. SCE's 2004 recorded data shows a total of 90 switches with an expenditure of $4,000,000. There has been no replacement of fuse cabinets from 1999-2004. Based on this recent data, ORA recommends a total of 90 switches for 2005 and 2006 with an annual expenditure of $4,100,000 and $4,200,000, respectively.

SCE estimates a total of $27, 580,000 for the replacement of underground distribution switches. SCE plans to replace 143 mainline manual oil-filled switches in 2004 and 200 in 2005. Of the remaining 1857, SCE plans to replace 300 in 2006. At a replacement rate of 300 per year, SCE would replace the remaining manual switches over six years, which for the purposes of this GRC appears reasonable. Therefore, we will adopt SCE's estimate of 300 manual oil-filled switch replacements for 2006.

SCE has identified 131 mainline spring operated oil-filled switches with known problems and intends to replace 15 per year from 2005 to 2008. This appears reasonable and will be reflected in rates.

There are 6,343 spring operated oil-filled switches with no defects. SCE plans to replace 10 in 2005 and 85 in 2006. SCE has not provided a compelling reason to increase the number of replacements from 10 to 85, and we will use 10 replacements for 2006.

SCE states there are 1100 oil filled BURD switches older than 30 years old that pose the same reliability and safety issues as posed by the mainline oil switches. SCE plans to replace 125 in 2005. The immediate need to replace 900 switches over the 2006 - 2008 time period is not evident. For this GRC, we will provide a moderate increase over the amount planned for 2005, by assuming the remaining 975 switches are replaced over a six-year period at 162 per year.

Out of 957 submersible fuse cabinets SCE plans to replace 20 in 2005 and 250 in 2006. SCE indicates that while external inspections are performed every three-years, no internal inspections are made because opening the cabinets risks destruction of the water-tight seal and the cabinets are extremely old. SCE states that while not a safety issue, the proximity of many submersible fuse cabinets to their expected end of life will impact reliability. Almost 400 submersible fuse switches are older than 40 years old. It appears the replacement of all of the cabinets is a reasonable course of action. However SCE does not support the immediate need to replace 750 of the remaining 937 over the 2006 - 2008 timeframe. For this GRC, we will instead assume 125 cabinet replacements per year. While less than requested by SCE, it is still a substantial increase from the total of 20 planned for the 2004 - 2005 time period and provides funds to replace the switches that are older than 40 years.

Based on SCE's estimates of the unit costs of the switches, the adjustments described above, we calculate expenditures for 2006 to be $19,537,000. This reduces SCE's test year request for underground distribution switches by $8,043,000.

SCE states a need to increase its preemptive replacement of underground cable in order to avoid a significant decline in system reliability. SCE states that the number of circuit interruptions due to cable failure is increasing, and modest volumes of replacements are proposed (0.1% of the cable system in 2005 and 0.5% in 2006). Specific sections to be replaced will be determined by a combination of age, circuit performance, and judgment.

ORA believes SCE's forecast is excessively high and unreasonable. According to ORA, SCE only provided information which rates replacement factors for paper insulated lead covered (PILC) and cross-linked polyethylene (XLPR) cables, and has not supported its request for replacing high molecular weight polyethylene (HMW-PE) cable.

SCE has forecasted underground cable replacement costs to be $10,200,000 in 2005 and $35,000,000 in 2006. ORA recommends costs of $459,000 in 2005 and $0 in 2006.

SCE proposes a five year plan to replace 860 miles of PILC and HMW-PE cable (14% of current inventory). From 1999 to 2003, SCE has replaced 70 conductor-miles of underground cable. They planned 0 miles in 2004, due to budget constraints, 60 miles in 2005, 200 miles in 2006, 300 miles in 2007 and 300 miles in 2008. SCE also shows that the sustained interruptions due to underground failures ranged from about 200 to 250 sustained interruptions per year from 1994 to 1999. From 2000 to 2003 the range has increased to about 300 to 350 sustained interruptions per year, despite replacement of 70 miles of cable. From this information, it is reasonable to assume that a replacement at a rate greater than in the past is necessary to maintain or reduce the sustained interruption rate. What is not clear is what the replacement rate should be. SCE states precise engineering data is not available and that its proposed replacement volumes are admittedly heavily based on judgment. SCE argues that to delay replacement of cable pending the availability of precise engineering data will institute a defacto policy of running cable to failure and that the inescapable eventual result of such a policy would be significantly poorer system reliability than what customers experience today. In general, we agree with SCE. However, the proposed 200 to 300 miles of cable replacement per year is a significant increase over the recorded level of 14 miles per year over the 1999 - 2003 timeframe or the planned 60 miles of cable replacement in 2005. Without more engineering data, we would prefer to moderate the increased rate of cable replacement and will instead assume 100 miles per year of cable replacement for this GRC cycle, which is a substantial increase to the 60 miles of cable replacement planned by SCE for 2005. While less than that requested by SCE, it should be sufficient to provide information on the effect of an increased rate of cable replacement on the number of sustained interruptions. Hopefully precise engineering data will also become available for analysis in the next GRC. This reduces SCE's request for replacement of underground primary cable from $35,000,000 to $17,500,000 for test year 2006.

SCE states that old and obsolete automatic reclosers (ARs) need to be replaced at a rate slightly less than that achieved in 2000 and 2002 in order to manage system reliability and also to enhance public safety. In order to replace 20 ARs per year, SCE requests funding of $1,130,000 in 2005 and $1,1700,000 in 2006.

ORA states that SCE did not support its forecast and recommends the use of replacement history to forecast the number of replacements. ORA recommends costs of $566,500 for 2005 and $583,500 for 2006, based on 10 replacements per year.

ORA's recommendation is based on an average of 2002 to 2004 data. According to SCE 22 ARs were replaced in 2000 and 21 in 2003. No replacements were possible in 2001 due to the financial crisis, only nine were replaced in 2003 and none in 2004 due to lineman resource limitations and corporate financial restraints due largely to the priority of meeting higher than expected customer demand. The financial crisis was an extraordinary circumstance and its effects should be ignored for forecasting purposes. However, diversion of costs due manpower constraints or higher priorities is not extraordinary and may be reflective of what happens in the test year. While, as discussed previously, we will use SCE's estimate for 2005 subject to the CAAM review in 2006, we will base the 2006 forecasted number of AR replacements on the average of 2000 - 2004 closures, excluding 2001. This results in 13 replacements in 2006 at a cost of $759,000.

SCE maintains that capacitor banks need to be replaced at a rate slightly less than that achieved in 2003 in order to provide adequate voltage to customers and ensure grid stability. SCE requests funding amounting to $6,900,000 in 2005 and $7,100,000 in 2006.

ORA states that it requested detailed historical data regarding failed and obsolete capacitor units by type for the years 1999-2004, which SCE could not provide. Based on the limited data provided, ORA escalated recorded 2004 data to develop its forecasts for 2005 and 2006 in the amounts of $5,900,000 and $6,100,000, respectively.

As discussed previously, we will use SCE's estimate for 2005 subject to the CAAM review in 2006. For the test year, ORA recommends funding at the 2004 level. SCE criticizes ORA for providing less funding than what SCE expects to fund going forward. SCE, on the other hand, does not explain why its 2004 recorded amount is low compared to what it suggests it needs for the future years. It is not clear why the increased level of replacements is necessary. In this situation, the use of the most recent information is reasonable and we will adopt ORA's recommended funding level of $6,100,000 for test year 2006.

SCE forecasts $1,100,000 in 2005 and $8,300,000 in 2006 to replace underground vaults and manholes which are showing signs of weakening and potential collapse. According to SCE, this, as part of the infrastructure replacement program, is a new program that addresses an emerging problem. Collapse of these concrete structures poses a risk to public safety and system reliability.

Based on its perceived lack of information supporting SCE's request and understanding that damaged equipment should be tracked in a different account, ORA recommends no funding for this program in 2005 and 2006.

We recognize SCE's argument that this is a new program and that SCE has not replaced deteriorated vaults and manholes of the type proposed here prior to 2004, the year SCE began to replace pre-cast underground concrete structures as part of its infrastructure replacement program. For 2005 and 2006, it is reasonable to recognize the new program as part of infrastructure replacement.

Based on known problems, SCE has justified replacements planned for 52 vaults/manholes and 74 BURD structures from 2004 through 2008. For 2006 the allocated costs would be $3,520,000. However, SCE indicates its belief that there may be more underground vaults and manholes that are candidate for replacement beyond these amounts. SCE therefore allocated additional funding for analyzing and replacing 22 additional structures in 2006 at a cost of $4,780,000. We find insufficient justification for more than doubling the request for projects that may or may not be undertaken. We therefore include only $3,520,000, for known and needed projects, in the test year estimate.

In 1997, SCE instituted the Annual Circuit Review Program. SCE states that the objective of the program is to maintain the overall reliability of the distribution system despite the tendency toward less reliability due to infrastructure aging. According to the company, the basic premise of the program is that the most cost-effective way of impacting overall system unreliability is to direct resources toward the larges individual contributors to that unreliability. Consequently, SCE's practice has been to focus on the worst-performing circuits ranked using objective measures of reliability.

SCE proposes to remediate 5 circuits in 2005, 15 circuits in 2006, 20 circuits in 2007 and 20 circuits in 2008, at a cost of $1,000,000 per circuit. ORA accepts the 2005 proposal to remediate 5 circuits and extends that number to 2006. ORA also recommends a cost per circuit of $500,000 based on the use of more current information.

Regarding the forecast of the number of circuits to be remediated, SCE states that available funds and workforce resource limitations preclude it from doing this work in 2004. Also, they have only planned to remediate five circuits in 2005. While, as suggested by SCE, this program may be a very cost-effective way of staying the effects of infrastructure replacement, it does not appear to be a high priority for funding. For the years 1999 to 2003, excluding 2001 due to the energy crisis, SCE remediated an average of eight complete circuits per year. We will use that annual amount for the forecasted years for this GRC cycle.

SCE states that the cost per remediation can vary widely depending on circuit length, number of customers, age of circuit, and whether the circuit is located in urban or rural areas. In 1999 eight circuits were remediated at an average cost of $1,500,000. In 2002 and 2003, ten circuits were remediated at an average cost of about $500,000. Due to the wide variance in costs and SCE's explanation of the possible reason, it is not clear that use of the more recent 2002-2003 information, as recommended by ORA, would produce better estimate of future costs than would the 1999 information. SCE's rough estimate of $1,000,000 per remediation appears reasonable.

Use of our adopted forecast of eight remeditated circuits per year at an average cost of $1,000,000 results in a test year 2006 estimate of $8,000,000 for the Annual Circuit Review Program, as opposed to SCE's request of $15,000,000.

SCE's forecast of wood pole repairs was based on the number of poles already identified for repair which must be completed to avoid non-compliance with regulatory requirements. SCE forecasts costs of $13,900,000 for 2005 and $19,900,000 for 2006.

ORA is reluctant to rely on historical cost data and number of poles forecast to be intrusively inspected as the bases for its forecast, since it perceives recorded pole and cost data to be unreliable. ORA based its repair expenditure estimates of $1,500,000 for both 2005 and 2006 on recorded 2004 cost data for steel stubbing and fiberglass wrapping.

Attached to its rebuttal, SCE provided the structure numbers for all deteriorated wood poles identified to be repaired by fiberglass wrap or steel stub in 2005, 2006 and 2007. SCE states that in order to comply with GO 95:

    · By the end of 2005, 743 poles must be fiberglass wrapped and 733 poles must be steel stubbed.

    · By the end of 2006, 3,561 additional poles must be fiberglass wrapped and 5,446 additional poles must be steel stubbed.

    · By the end of 2007, an additional 4,809 poles must be fiberglass wrapped and an additional 7,926 poles must be steel stubbed.

SCE indicates that it will not be able to perform the nearly 13,000 pole repairs due in 2007. A significant number of repairs must be completed ahead of their compliance due dates. SCE's current plan is to perform 1,745 fiberglass wraps and 2,691 steel stubs in 2005, and 4,000 fiberglass wraps and 6,000 steel stubs in 2006. This will leave roughly the same number (about 3,500 fiberglass wraps and about 5,500 steel stubs) to be performed in 2007. SCE states that its GRC forecast (developed in early 2004) of 27,000 pole repairs in 2005-2007 was only slightly conservative and that its current forecast for repairing the deteriorated wood poles that it knows with certainty must be done, provides it with the best chance of compliance with CPUC regulations.

In general, we agree with SCE and will provide the opportunity to repair all identified poles needing repair through 2007. Based on SCE's plan as indicated above, it will need to wrap a total of 7,368 poles in 2006 and 2007 and stub an additional 11,516 during that same time period. In its original showing, SCE also estimated 2000 wraps and 3000 steel stubs in 2008. For ratemaking purposes, we will normalize the repairs over the three-year GRC cycle by providing for 3,123 wraps and 4,839 stubs for each of the three-years. Use of SCE's 2006 unit costs results in 2006 expenditures of $9,600,000 for wraps and $6,200,000 for stubs, for a test year total of $15,800,000, which we will adopt.

SCE has modified its request for bark beetle pole replacement, since the program will now end in 2005. For 2005 SCE now estimates expenditures of $3,500,000. ORA's estimate of $4,500,000 was developed using more recent data than was available when SCE wrote its original testimony and estimated 2005 costs of $7,964,000 and 2006 costs of $3,318,000.

Costs now will only be incurred in 2005 and, as discussed earlier in this decision, we will be incorporating 2005 recorded information into this proceeding via SCE's CAAM filing in 2006. In the meantime, we will include SCE's forecast of capital expenditures for 2005. We will adopt SCE's latest estimate of no expenditures in 2006.

SCE based its forecast of pole repair/replacements on work already identified and scheduled, compliance driven inspection frequencies and historic rejection rates. SCE expects to replace 986 poles, fiberglass wrap 105 poles and steel stub 201 poles in its subtransmission system in both 2005 and 2006. SCE forecasts expenditures of $19,500,000 for 2005 and $20,100,000 for 2006. SCE's estimated cost per pole replacement is $18,400, in 2004 dollars.

ORA escalated 2004 recorded expenditures to develop its forecast of $12,500,000 for 2005 and $12,900,000 for 2006. Embedded in ORA's forecast is a recorded cost per pole replacement of $14,197 per pole.

In general, SCE's forecast of work activity, which is based on repair/replacements on work already identified and scheduled, compliance driven inspection frequencies, and historic rejection rates, is reasonable. However, SCE's recorded pole replacement costs for 2004 were substantially less than projected, principally due to the cost per pole being significantly less than forecasted. SCE explains that a very high percentage of poles replaced in 2004 were located in the San Joaquin region, which has been shown to have a relatively lower replacement cost due to its rural nature and ease of work. While the explanation seems reasonable, SCE does not explain if or how its forecasted price per pole takes such variances into consideration. SCE does not relate its projected cost per pole to any particular percentage of rural work. In light of the low 2004 recorded cost per pole replacement, SCE showing does not support its forecasted replacement cost of $18,400 per pole. We will instead average the two costs, the $18,400 projected by SCE in 2004 dollars and the 2004 recorded cost of $14,197 to approximate the cost per pole for 2006, in 2004 dollars. This results in a cost per pole of $16,300, and reflects a lower percentage of rural pole replacements in 2006 than in 2004. Adjusting SCE's forecast, by the reduced cost per pole, results in our adopted test year 2006 forecast of $17,939,000 as opposed to SCE's request of $20,100,000.

Distribution Automation is an ongoing program to provide remote control and monitoring of various distribution devices, such as mainline distribution switches, automatic reclosers, fault indicators and capacitor banks. This is accomplished by installing controllers incorporating intelligent electronics and a wide area packet radio communication system to operate distribution equipment and provide real-time information to system operators and engineering personnel.

There are three ongoing distribution automation capital projects:

· Capacitor Automation or Programmable Capacitor Controls;

· Circuit Automation, and

· Distribution System Efficiency Enhancement Project.

At issue in this proceeding is the forecast of costs related to circuit automation. SCE forecasts $5,900,000 for 2005 and $6,100,000 for 2006. For all three distribution automation projects, SCE requests $12,900,000 in 2005 and $13,400,00 in 2006.

ORA believes SCE's forecast for circuit automation is unreasonable and recommend a reduction based on its installation forecast for the total number of underground and overhead remote control switches, as well as the unit cost calculation for these items and remote transmission switches and remote fault indicators. ORA recommends expenditure levels of $3,700,000 for 2005 and $3,800,000 for 2006. For all three distribution automation projects, ORA recommends $10,700,000 in 2005 and $11,100,000 in 2006.

In rebuttal, SCE claims that contrary to ORA's assertions, the material costs for automation are in line with unit prices used in SCE's cost calculation for each type of automation equipment; ORA has incorrectly compared average customer minutes of interruption (ACMI) reductions for SCE's circuit breaker replacement program to the distribution automation program, ORA uses inconsistent recorded data to arrive at its forecast for remote control software and circuit automation; and ORA's distribution reliability proposal takes credit for SCE's proposed distribution automation program, which ORA's capital expenditure recommendation would largely disallow.

SCE has provided information in rebuttal that shows its vendor contract prices are not two-to-three times lower that that used in its cost calculations as claimed by ORA. The vendor contract prices appear to be in line with those assumed in SCE's automation cost calculations. SCE also makes relevant observations regarding ACMI comparisons, ORA's use of inconsistent recorded data and assumptions related to ORA's proposed reliability mechanism. However, ORA's use of recorded data to forecast future expenditures is not misplaced, especially in light of the significant proposed increases in this program from the 2003 recorded amount of $3,141,000 to SCE's $6,100,000 forecast for 2006. Also, SCE's recorded amount for 2004 of $3,400,000 is less than the $5,804,100 forecast as part of this GRC. While SCE argues its forecasts are better because it accounts for such things as differences between the current and future mix of overhead and underground equipment, its forecasts can be affected significantly by other factors such as limitation of workforce and prioritization of projects which may overwhelm such planning precision. A five year average of historical costs for the period 1999 - 2004, excluding 2001, would provide a reasonable forecast based on fairly recent, applicable information. In rebuttal, SCE indicates if that average were escalated the result would be $4,800,000.88 We will include this amount for the test year forecast of circuit automation expenditures. The adopted test year forecast for the three ongoing distribution automation projects is then $12,100,000.

SCE requests capital expenditures in two major categories - Substation Capital Replacements and Other Capital Requirements. Substation Capital Replacements expenditures are further divided into two sub-categories - Substation Infrastructure Replacement Program (SIRP) and Routine Capital Replacements. The SIRP focuses on a proactive, planned replacement of aging infrastructures for the purpose of minimizing safety risk to employees and the general public, maintaining system reliability, and reducing O&M costs. Routine Capital Replacements, on the other hand, are expenditures for the purpose of improving substation infrastructures, including routine and reactive replacements of equipment due to failures and normal maintenance. Other substation apparatus not covered under the SIRP are also included in this category.

Other Capital Requirements covers forecast expenditures for tools, spare parts and equipment, facilities, furniture and office equipment, and other miscellaneous items such as easements.

ORA states that SCE has failed to demonstrate that its request is reasonable and necessary. ORA's recommendations for 2005 and 2006 are consistent with historical spending.

Discussions related to each of the substation projects in dispute follow. As discussed earlier in this decision, we will address issues as they relate to test year 2006 costs only. We will be incorporating 2005 recorded information into this proceeding via SCE's filing in 2006. In the meantime, we will include SCE's forecast of capital expenditures for 2005.

Also as discussed earlier in this decision, in general SCE has provided information that supports a need to replace certain portions of its distribution infrastructure at rates in excess of recorded rates. This also applies to substation element of its infrastructure replacement program. Again, we will evaluate SCE's requests for the various aspects of its proposed infrastructure replacement program with that in mind. However, SCE still has the burden to justify the need and costs of each of its various proposed elements of the program.

As discussed below, we have evaluated SCE's test year 2006 proposals for replacement of substation capital equipment and considered ORA's recommendations in developing the test year forecasts. For the amounts at issue, SCE requests $127,800,000 for 2006, while ORA recommends an amount of $49,500,000. We adopt a test year 2006 forecast of $84,400,000.

SCE forecasts expenditures of $16,100,000 in 2005 and $22,600,000 in 2006 to replace 130 distribution circuit breakers in 2005 and 187 in 2006. From 2004 to 2008, SCE plans to replace about 211 distribution circuit breakers per year, equivalent to a 50-year replacement cycle.

ORA is recommending the continuation of historical work and expenditures because of its belief that SCE has not justified its forecast. ORA states that while SCE claims that age, those circuit breakers 40 years and older, is a determinant factor in the replacement of the circuit breakers, between 1997 and 2004, out of a total of 1344 distribution circuit breaker replacements, 628 or 43% were under 40 years of age. ORA's recommendation of $12,200,000 for 2005 and $22,600,000 for 2006 is based on averages of 2002 - 2004 expenditures.89

In rebuttal testimony, SCE provided age demographic information on circuit breakers. SCE claims that information shows that while a significant number of circuit breakers were replaced below the age of 40, most were in the older part of that range. The data shows that of the 1624 circuit breakers removed from service, 44% were aged 41 years or older, 12% were aged between 36 and 40 years, and 11% were aged between 31 and 35 years. In general, we would agree that there is a greater likelihood that older rather than newer circuit breakers will be replaced. However, it is not clear why SCE's estimate of 187 circuit breakers to be replaced at a cost of $22,590,000 is reasonable. It is not clear, that at this time, a 50 year replacement cycle is necessary.

SCE has provided information on its aging circuit breaker population and technical reasons why certain circuit breakers may be prone to failure. However, while the circuit breaker replacement program appears to be successful in decreasing the average customer minutes of interruption (ACMI) since 1997,90 it was done with average expenditures of approximately $9,000,000 per year. SCE spent $12,605,000 in 2002, $14,650,000 (including $4,352,000 for the Santa Monica Substation) in 2003, and $11,800,000 in 2004 on the distribution circuit breaker replacement program. SCE attributes lower spending in 2002 and 2003 to residual effects of the energy crisis and prioritization of projects. This may be true, but the necessity of spending at the significantly higher level of over $22,000,000 per year has never been demonstrated, even during years in which SCE's infrastructure replacement program was in effect. However, due to the potential benefits of this program in reducing interruptions to customers as well as O&M expenses, we will include expenditures for 2006 based on SCE's forecast for 2005. 2003 recorded information shows SCE's commitment to spend at least $14,650,000, even if a large portion was just for the Santa Monica substation. The 2005 forecast of $16,100,000 is in the range of the 2003 recorded amount which is about $15,100,000 in 2005 dollars. We see, however, no justification to increase the 2006 adopted amount to the $22,600,000 requested by SCE.

SCE replaces transformers both proactively, that is, before in-service failure and reactively, that is, after failure in-service. SCE reiterates that replacement prior to imminent failure is one the main goals of its infrastructure replacement program, since it mitigates outages and the resulting costs to customers. SCE's A-Bank Replacement Program starts with expenditures of $25,500,000 in 2006, which is planned to allow replacement of 12 A-Bank transformers in that year. SCE indicates that it plans to replace a total of 46 A-Bank transformers by 2008. The 46 transformers were identified by a group of company experts assembled to rate the transformers for replacement. The group known as the Transformer Resource Management Committee (TRMC) looked at the following factors that contributed to in-service failure: (1) age, (2) manufacturer, (3) design, (4) dissolved gas analysis, (5) loading/fault history, and (6) maintenance history.

According to ORA, between 1989 and 2003, SCE experienced a failure rate of 0.9 per year or an average of 13 months between failures. Based on its perception of a lack of support for SCE's forecast, ORA used A-Bank replacement history to determine the forecast for 2006. Based on recorded number of projects and the total cost for A-Bank transformer replacements from in 2000-2002, ORA forecasts 2 replacements for 2006 at a total cost of $2,000,000.

SCE has provided information that shows that the average age of the transformers it plans to replace is 52 years, which is significantly higher than both the nominal design life of 20.55 years identified by the IEEE and the historical average age at replacement of 42 years. This justifies the need to consider increased proactive replacements of A-Bank transformers in the future. However, in recent years SCE has replaced only two A-Bank transformers per year proactively and has not replaced any due to failure. While the recommendation to replace 16 transformers in 2006 is based on the recommendations of a group of company experts who rated the transformers for replacement, we are not convinced that such a drastic increase from the recent experience of two per year is necessary.

SCE indicates that ORA's recommended two replacements per year would amount to a replacement cycle of more than 100 years, more than five times the transformers nominal design life. We will authorize 10 replacements per year which would then result in a replacement cycle close to the nominal design life. Additionally, we will reduce the cost per transformer from $1,700,000 to $1,500,000 in consideration of recent recorded unit costs that averaged about $1,000,000,000 and which were used by ORA in its estimate. SCE did provide cost estimate detail, which showed the cost of the transformer itself was about $1,000,000, but it did not explain why the recorded costs were so low. Our adjustment to the average unit cost, merely reflects the possibility that circumstances which occurred during the last three-years may occur in the future and result in costs less than estimated by SCE. Based on this discussion, we adopt a forecast of 10 A-Bank transformer replacements at a cost of $15,000,000 in test year 2006.

Failures of B-Bank transformers averaged 6.5 per year from 1993 to 2003. In 1998, SCE conducted an analysis of B-Bank transformers using the TRMC methodology. Of the initial 50 units assessed, six were identified for replacement. SCE later assessed an additional 150 units and based on the same TRMC methodology, planned to replace seven in 2004, 14 in 2005, 13 in 2006, 14 in 2007 and 13 in 2008. The forecasted expenditures for 2006 amount to $6,600,000.

ORA states that SCE only replace 4 transformers in 2004 at a cost of $2,900,000. None were replaced in 2001 and 2002. Also, ORA indicates that the 1998 TRMC identified six replacements but SCE identified 15 units for replacement as part of this GRC. ORA states that the available TRMC data shows that only nine units need to replaced and ORA has reason to believe all nine units have already been replaced. Based on its analysis, ORA recommends $0 for B-Bank transformer replacements for 2005 and 2006. ORA asserts that SCE should have embedded expenditures from previous GRCs for transformer replacements on as-needed basis.

In rebuttal testimony, SCE clarified that in 2004, it replaced seven transformer banks at a cost of $2,900,000. Also, ORA has apparently confused identification of units for the replacement plan analysis and the units replaced. SCE states that while the initial analysis in 1998 identified 15 units for the replacement plan, by 2001 it had performed analysis on 212 units. SCE also provided information on ORA concerns regarding the planned year of replacement and identification of units replaced under the substation infrastructure replacement program.

SCE has provided information that shows the ages of the units scheduled for replacement in 2005, including seven that are 81 year old, three that are 78 years old and six that are 77 years old. This justifies the need to consider increased proactive replacements of B-Bank transformers in the future. When considering an average of 6.5 failures per year and the replacement of seven banks in 2004, SCE's forecast of 13 B-Bank transformer replacements in 2006 appears reasonable. We adopt SCE's test year forecast of $6,600,000 for this program.

SCE states that the protection and control systems at many of its 900 substations are aging. The age of the equipment ranges from 30 - 100 years. The aging protection and control systems are made up of electro-mechanical devices such as relays and switches, which require routine testing, maintenance, and repair. This equipment has no self-monitoring capability and no remote monitoring or control functions. The modern protection and control equipment SCE is using provides self-monitoring as well as remote monitoring and control of all functions and will identify potential problems before they cause harm. Through its automation program, which ended in 2003, SCE has replaced the protection and control equipment in 187 substations. There are still over 700 stations with the old electro-mechanical equipment, which this program is designed to replace. SCE began this program in 2001 with engineering, design, and procurement. Construction on the first project began in 2003. The program is intended to be ongoing, with approximately 25 substations being retrofitted each year. SCE forecasts expenditures of $14,750,000 associated with 25 projects in 2005 and $14,880,000 associated with 25 projects in 2006.

ORA disagrees with SCE's projected 25 projects per year in 2005 and 2006 as well as unit cost of the replacements. ORA recommends using the escalated three-year average of the 2002-2004 expenditures and the actual number of substations replaced, as the basis for the 2005 and 2006 forecast. ORA's calculations yield an annual forecast of 17 substations with costs of $6,200,000 in 2005 and $6,400,000 in 2006.

The distribution protection and control replacement program appears to be replacing the Substation Automation Program which ended in 2003. In that respect, we can consider these activities as continuing in nature. What is not clear is whether the programs are comparable as to the number of protection and control replacements per year or the magnitude of the expenditures related to the replacements. While SCE points out it has over 700 substations with old electro-mechanical equipment, it has not provided much information on failures associated with such equipment or quantification of other factors which would justify a need to replace protection and controls at 25 substations per year in 2005 and 2006. A program of replacement as suggested by SCE is reasonable but there needs to be some specific information on the impacts of carrying out the program at increased or reduced levels in order to make a decision on what the adopted level should be. SCE has not provided information on why its request of 25 substations per year is better than the historical average of 17 recommended by ORA. For this reason, we will adopt 20 as the number of substations for the distribution protection and control replacement program for 2006. Regarding costs, SCE states that costs of each substation are driven by the protection requirements of each piece of substation equipment. On the other hand, SCE has not addressed the specific reasons why the recorded units costs of $376,000 (in 2006 dollars) recommended by ORA are so much lower than its engineering estimates that average $595,000 per unit for 2006. Similar to our adjustment related to the costs of the A-Bank transformer replacement program, we will reduce the average cost for 2006 to $500,000. This merely reflects the possibility that circumstances which occurred during the last three-years may occur in the future and result in costs less than estimated by SCE. Based on this discussion, we adopt a forecast for distribution protection and control replacement at 20 substations with expenditures amounting to $10,000,000 for test year 2006.

SCE states that the A/AA Control Room Upgrade and Replacement Program will provide control of seven of its large attended stations by replacing the existing manual controls and indicating devices with a networked system. The new system will make it possible to locate the system operator's workstation at any location, not just in close proximity to the manual controls. It will also make it possible to monitor and control these critical facilities from a remote location in the event that the local control room becomes un-inhabitable. SCE proposes expenditures of $7,500,000 in 2005, 5,800,000 in 2006, $360,000 in 2007, and $2,000,000 in 2008.

ORA states that SCE could not provide any support for the forecast and claimed that "the numbers in this table were arrived based on conceptual estimates." ORA indicates that SCE could only provide support for one project, that of the Villa Park substation and recommends that only this project be included in the 2005 and 2006 forecast. ORA recommends funding of $1,000,000 for each of the years 2005 and 2006.

While ORA does not appear to contest the need for this program, it challenges SCE's estimated costs alleging that SCE was unable to provide support for those estimates. SCE argues that it provided the cost breakdown in the same level of detail provided for the Villa Park and Mesa projects and that ORA simply asserts that SCE's costs estimates, which were based on industry accepted standard engineering methods are not acceptable while offering no objective alternatives. We agree with SCE on this point. If ORA disagrees with the estimating methodology, it should at least explain the problem so that we can determine the soundness of SCE's showing. If possible the suggestion of an alternative methodology would be helpful. Lacking this type of information, we will adopt SCE's estimate of the costs for this program. We note that until detailed engineering and design is performed, use of industry accepted standard engineering methods may be appropriate for estimating future costs. For ratemaking purposes, we will normalize the costs for this program over the GRC cycle 2006 - 2008. This results in an average expenditure of $3,800,000 which we will adopt for test year 2006.

This program addresses estimated expenditures to replace substation equipment and major equipment that fail while in service or as a result of inspections showing the risk of imminent failure. SCE uses a four-year average of recorded expenditures for the period 1999 - 2003, excluding 2001, to forecast a base. On top of that base, SCE adds adjustments for Butyl Current Transformer Replacement, Cable Trench Cover Replacement, Disconnect Switch Replacement and Environmental Remedial Action, forecasted expenditures not previously identified in this blanket. For this reactive replacement blanket, SCE forecasts expenditures of $26,800,000 for 2005 and $32,900,000 for 2006.

ORA disagrees with SCE's calculation of the four year average. ORA also disagrees with SCE's adjustments related to Butyl Current Transformer Replacement, Cable Trench Cover Replacement, and Disconnect Switch Replacement. ORA's estimate of $17,700,000 for 2005 and $18,200,000 for 2006 is based on the escalated four-year average for 1999 - 2003, ($15,600,000 in constant 2003 dollars as opposed to SCE's calculation of $24,300,000 in 2003 dollars). ORA's estimate also includes environmental remedial action costs as requested by SCE.

Regarding the four-year average, ORA opposes SCE's "adjustment" of historical costs through the inclusion of estimated expenditures for transformer bank replacements from 1999-2001, and expenditures for "Other Specific Engineered Projects" from 2001-2003. ORA argues that SCE artificially inflated the average. In rebuttal, SCE states:

    As an initial matter, SCE did not inflate the historical averages, but adjusted the recorded cost to properly account for all reactive replacement expenditures. Prior to 2002, a portion of reactive replacement activities, such as those involving greater degree of complexity in project scope (for example, the replacement of a failed A-Bank transformer) were funded by offsetting this reactive blanket against another budget item (such as SIRP), resulting in the reduction in the recorded expenditure in this reactive blanket.

    This same budget offset was also used on other reactive replacement projects of significant scope (i.e. more than a simple like-for-like swap-out) that required engineering and design work. In these cases, SCE's Project Management Organization manages those projects and blanket budgets are offset by the expended capital amount. Again, this effectively reduces historical expenditures in these blanket budgets for reactive replacement of greater complexity.91

SCE explains why the recorded amounts in this blanket are reduced, but omits any discussion of whether or not the types of activities that caused these reductions will continue in the future. Without such analysis, it is reasonable to assume that offsets to this blanket will continue to occur. We see no reason to add offset costs back into the blanket in determining an average for forecasting test year 2006 and will adopt the four year average as calculated by ORA. This results in a 2006 base of $15,600,000 in 2003 dollars, or $17,600,000 in 2006 dollars.

SCE adds four adjustments to the base for forecasted activities that are not reflected in the historical data and therefore not in the four year average. ORA objects to three of the adjustments, as discussed below.

ORA objects to the inclusion of $600,000 for butyl current transformer replacements, because SCE could not provide support related to a 1987 survey and information related to a more recent survey was incomplete. In response, SCE states that it provided condition codes based on actual inspection results of the butyl current transformers in the substations, which should be sufficient to identify the problem.

SCE shows this cost to be in 2005. As discussed previously, we will be incorporating 2005 recorded information into this proceeding via SCE's CAAM filing in 2006. In the meantime, we will include SCE's forecast of capital expenditures for 2005. To the extent that the work is actually done, it will likely be included in rates. In any event, it appears that problems with butyl current transformers have been identified, perhaps as long as 17 years ago, and it is reasonable to replace them as proposed by SCE.

ORA objects to the inclusion of $1,600,000 per year to replace cable trench covers, principally because of its understanding that these costs are embedded in historic data and thus a certain level are already included in the base. In response, SCE states:

    In response to DR-ORA-45, we provided detailed information on the expenditure in FERC Subaccount 570.400 to replace deteriorated redwood cable trench covers with new redwood covers. The reactive replacement information for the distribution substations is not included in the discussion of subaccount 570.400 which only covers transmission substations. Thus ORA's claim that the proactive replacement program proposed for 2006 is similar to the reactive program in 2001 and 2002 is based on incomplete information. ORA's testimony fails to recognize the scope of SCE's proactive approach regarding the safety of its employees. ORA assumes that the historical rate of replacement was sufficient to address the replacement need going forward.

    ORA failed to recognize that during recent years SCE has been in the process of developing a new trench cover with enhanced durability. While this development was taking place, SCE used redwood trench covers on an interim basis to replace trench covers that had failed and this historical replacement rate was insufficient to address the safety concerns of the increasing number of deteriorated trench covers.92

SCE's explanations are sufficient to justify it request for cable trench cover replacement and we will include $1,600,000 for this activity in test year 2006.

Lastly, ORA objects to the inclusion of $3,000,000 per year to replace disconnect switches, because SCE could not provide support for the number of switches it proposes to replace. In response, SCE states that it has provided sufficient information through its exhibits, workpapers and data request responses to support its request.

SCE's testimony identifies a need to establish a systematic method for identifying high risk disconnects for replacement, indicating that the present method for identifying disconnects for replacement is a reactive process. The testimony itself does not provide any quantification or indication of the magnitude of the perceived problem. A data request response (Exhibit 237) referenced in SCE's rebuttal provides unit cost information. The response also indicates that historic data related to the number of disconnect switches repaired and replaced was not available. Also it does not appear that any workpapers supporting SCE's request were offered in evidence. In concept, SCE's request is reasonable, but without any justification for the level of activity proposed, we must reject funding for this program.

Based on the discussion above we adopt a forecast of $19,800,000. This includes $600,000 for environmental remedial action to which ORA did not object.

SCE explains that older circuit breakers may not meet newer operational requirements. A number of the older 66kV and 115kV class circuit breakers are incapable of de-energizing underground cable beyond a certain length. When the Rule 20B projects cause the cable to exceed this length, the circuit breaker must be replaced. Based on an average of 1999 - 2002 costs, SCE includes $2,200,000 in 2005 and $2,300,000 in 2006 for this activity.

ORA states that SCE did not justify its request and uses a three-year average of 2002 - 2004 costs to calculate its estimate $300,000 per year.

In response to ORA's recommendation, SCE states that the use of a four-year average based on the years 1999 - 2002 and adjusted for inflation is consistent with its proposal for subtransmission capital expenditures for Rule 20B projects. SCE argues that ORA's forecast is inappropriate because (1) the accounting for Rule 20B circuit breaker replacement changed recently and (2) 2003 was abnormally low and not representative of future spending needs.

SCE has sufficiently explained the basis for it proposed averaging of 1999 - 2002 data to forecast Rule 20B circuit breaker replacement costs. We will adopt the company's estimate of $2,300,000 for test year 2006.

The only area at issue here is SCE's request for an additional increase of $1.9 million above historical spending levels, which breaks down to $100,000 in 2004 and $1.8 million in 2005, to construct three access bridges for the flood control channels that bisect its Mesa-Antelope 220kv line right-of-way along Interstate 605 Freeway. SCE claims that this requirement is the result of an agreement with Caltrans and accounts for the high forecast in 2005. ORA opposes this portion of SCE's request because it cannot verify the need.

The costs for this project appear in 2004 and 2005. As discussed previously, we will be incorporating 2005 recorded information into this proceeding via SCE's CAAM filing in 2006. In the meantime, we will include SCE's forecast of capital expenditures for 2005. To the extent that the work is actually done, it will likely be included in rates.

This work category tracks expenditures for three budget items: (1) the Grid Dispatch Annual Department Program, (2) Tools and Work Equipment, and (3) Substation Spare Parts and Equipment. SCE's forecasts are a combination of averaging and budgeting and amount to $8,000,000 for 2005 and $9,100,000 for 2006.

ORA states that the 2005 request is almost twice, and the 2006 request is more than twice, the actual expenditure of $4.3 million recorded for 2004. Also, SCE spent $2.4 million less than its original forecast of $6.7 million for 2004. SCE explains that the 2003 and 2004 expenditures were abnormally low because the company had a low number of failures in substation B-Banks, but that the company is anticipating a greater likelihood of failure in the future. ORA investigated the need to purchase additional power transformers to maintain an adequate inventory of spares, but was unable to determine whether or not additional transformers are needed or determine the cost to acquire these transformers. ORA based its estimate of expenditures on an escalated three-year average resulting in $4,900,000 for 2005 and $5,000,000 for 2006.

Costs for this category have varied significantly over time, generally showing a downward trend since 1999. An average of recent years escalated to test year dollars would provide a reasonable estimate. Because of the energy crisis, 2001 has generally been excluded from averages in this proceeding. For tools and grid dispatch, we will use an average of the post energy crisis years of 2002 and 2003.93 This results in tools cost of $4,000,000 and grid dispatch costs of $300,000 for the test year. For spare parts, SCE links the test year amount to B-Bank transformer replacements. ORA was unable to substantiate that need. However, since we have, in this decision, adopted SCE's capital request regarding B-Bank transformer replacements will therefore include costs of the related spare parts, which amount to $3,600,000. Our adopted total test year forecast for this category is therefore $7,900,000.

For the Non-Operational Facility Blanket, SCE plans to expend $5,000,000 in each of the years 2005 and 2006 for facility expansion and improvements, including new office spaces, permits and building additions, office reconfigurations, etc. SCE states the funds are necessary to meet the incremental facility requirements of the T&D business unit. These funds are for potential locations and scope additions not included in the Corporate Real Estate (CRE) business unit capital budget.

Based on SCE's statement that the $5 million was a blanket allowance of funds without itemized estimates and ORA's understanding that SCE has never recorded any spending under this category, ORA concluded that SCE has provided no justification for the expenditures and therefore recommends $0.

In rebuttal, SCE argues that ORA ignored a data request response that showed details on increases in employees and impacts on availability of office space and facilities. SCE states that from 1999 to 2004 total head count with direct impact to facility need grew from 4,328 to 4,853. SCE also describes the San Jacinto Service Project that was included in its CRE testimony. Given increased numbers of employees, SCE has still not explained why the CRE budget cannot accommodate its facility growth needs, as specific facility needs are identified. For example, for shared services capital projects over $1 million, SCE lists a number of projects totaling over $100 million in direct costs with operational dates from 2004 to 2008. Included in the list is the San Jacinto Building Addition and other facilities used by the T&D business unit. SCE has not justified including funds for other, potential projects or addressed ORA's concern that no money has ever been spent in this category. We will therefore not include any funding for the non-operational facility blanket.

The Fee Simple and Rights-of-Way category is to acquire real properties and rights-of-way that are necessary for our transmission and substation system expansion due to load growth. SCE's estimate is based on its expected property and right-of-way needs for the 2004 - 2008 timeframe and result in estimates of $3,900,000 in 2005 and $500,000 in 2006.

Based on its analysis of SCE's expected needs, ORA states:

    ORA has reviewed SCE's supporting documents for this work category. According to responses to ORA data requests, SCE has not yet begun work on the Oak Valley acquisition project. The Akers and Canine projects currently have no supporting data to show that these projects will be completed in the year forecasted. It appears that SCE has not yet located potential substation sites for the Akers project and that the target date for this project, January 28, 2005, has not been met. As for the Canine project, SCE provided no supporting data at all for this acquisition. Finally, ORA learned that the Las Lomas acquisition is currently on hold pending a municipalization decision by the City of Irvine. As such, SCE will not need any of the requested capital expenditures it has previously requested.94

Based on ORA concludes that none of the requested expenditures are necessary and recommends $0.

For 2006, SCE has included $500,000 associated with the Oak Valley project which SCE addresses in its load growth testimony. Since ORA has not opposed the project, we will include the fee simple/rights-of-ways costs in 2006 as requested by SCE.

For the remaining projects at issue, the costs fall in the year 2005. As mentioned previously, we will be incorporating 2005 recorded information into this proceeding via SCE's CAAM filing in 2006. In the meantime, we will include SCE's forecast of capital expenditures for 2005. To the extent that SCE meets its expected needs for fee simple and rights-of-ways costs in 2005, those expenditures will likely be reflected in rates.

73 42 CPUC 2d at 693-694.

74 Pursuant to D.04-07-022, SCE filed advice letter 1808-E that established the Capital Additions Adjustment Mechanism (CAAM) for 2004-2005 to track the difference between actual (recorded) and authorized total company 2004-2005 gross capital additions plus cost of removal amounts. The advice letter notes that if, by the end of 2005, SCE fully implemented its 2004-2005 capital spending budget that was adopted in D.04-07-022 no customer refunds will be required. However, if SCE's authorized capital additions are greater than its recorded capital additions over the entire two year period, an overcollection in revenue requirement will be recorded in the CAAM and this amount will be returned to customers. The Commission approved SCE's advice letter 1808-E in Resolution E-3895.

75 See D.04-07-022, mimeo, page 275.

76 D.04-07-022, mimeo, page 236.

77 See D.04-07-022, mimeo, page 236.

78 This also simplifies calculations related to the results of operations program.

79 See D04-07-022, Finding of Fact 14.

80 For the used fuel project, it is not certain that the amount of money identified for the 1995 -1997 timeframe was the total cost of the project or just the amount that would be spent through 1997. This may be relevant because the ICIP period lasted through 2003.

81 See Exhibit 89, pp. 35 -38 and Exhibit 91, Appendix G for the development of the adjustment.

82 Exhibit 357, Appendix 10.

83 The difference in the estimated expense and the recorded plant addition amount is not explained or detailed. For instance, there may be overheads in the capital addition that would be reflected otherwise as an expense (e.g., pensions and benefits).

84 In response to ORA and TURN protests to SCE Advice Letter 1847-E seeking approval to increase its current line extension allowance, Resolution E-3921, issued on June 16, 2005 reduced SCE's proposed distribution rate by its baseline credit (0.625 cents/kWh) and imposed a COS factor of 17.52% per year versus Edison's proposed COS factor of 16.20% per year. The resolution also directed the utilities to file applications within 90 days that address possible changes in policy and the methodology for determining line extension allowances. Among other issues, the applications will address alternative methods of calculating the net revenue on which future line extension allowances are based, revenues sources to be used when calculating the allowance (including that from substations, primary circuits, and sub-transmission), sources of data for calculating the allowances, and the criteria for requiring that a revenue impact estimates be included in an allowance change advice letter filing.

85 San Bernardino ($1,720,000), Arrowhead ($1,510,000), Rush ($740,000), and Kernville ($820,000) Substations.

86 Ten years from the issuance of General Order 165 will be March 2007.

87 SCE, Exhibit 42, page 37.

88 SCE/Exhibit 96, page 81.

89 See Comparison Exhibit, Exhibit 899, page 337. In its opening brief ORA, page 107, ORA states that it recommends $9.9 million per year based on an average of 1999 - 2004 expenditures. For discussion, we will use ORA's Comparison Exhibit recommendation.

90 Since this program has been established, SCE's ACMI measurement has decreased from 1.452 minutes per year to 0.5 minutes per year.

91 SCE/Exhibit 96, pages 126-127.

92 SCE/Exhibit 96, page 129.

93 ORA references 2004 recorded information, but does not provide the necessary detail for use in this analysis. Therefore, only 2002 and 2003 recorded data is used.

94 ORA/Exhibit 202, page 13-D-68.

Previous PageTop Of PageNext PageGo To First Page