15. Depreciation

15.1. SCE's Request

SCE's depreciation rates for transmission and distribution (T&D) accounts have not been updated since its 1995 GRC. SCE asserts that these rates are out of date and the accumulated depreciation deficit is growing each year. SCE estimates that the current T&D depreciation rates have resulted in accumulated depreciation that is approximately $1.4 billion behind where depreciation rates would be based on current authorized levels of net salvage.

Accordingly, SCE requests that the Commission adopt its proposed depreciation expense of $934.8 million for 2006 which represents a $238.3 million or 34% increase over the authorized level in year 2003. The largest contributor to the increase is the recovery of SCE's past deficit in accumulated depreciation.

According to the company, its request is fully supported by its depreciation study, which was conducted in accordance with the Commission's Standard Practice U-4, Determination of Straight Line Remaining Life Accruals - a methodology used by this Commission for over fifty years. In conducting its study, SCE indicates that it performed a thorough analysis of its accounting records, drew upon the observations and expertise of field personnel with many years of operations experience, and applied the collective judgment of depreciation experts with many years of experience.

15.2. ORA's Recommendation

ORA also conducted its analysis of depreciation rates in accordance with the procedures set forth in Standard Practice U-4. ORA agrees with SCE's proposed average service lives. However, ORA disagrees with SCE's net salvage analysis. ORA's estimates for net salvage are about $101,000,000 less than SCE's for three reasons: (1) ORA is proposing a number of adjustments to SCE's requested capital additions for the test year, (2) SCE used 10 years of historical data as the basis for calculating its proposed net salvage ratios, while ORA primarily used 15 years of historical data, and (3) ORA recommends that the increase in negative salvage rates be capped at 25% above current levels for FERC Accounts 364 and 369.

ORA provides four reasons why using 15 years of historical data is more appropriate than using 10 years. First, a 15-year band provides a more accurate and balanced picture of transactions occurring over a greater time period. Second, a 15-year band is consistent with the 15-year historical band associated with the depreciation study SCE performed in its last GRC. Third, a 15-year band in this case is consistent with the 15-year historical bands used by both SDG&E and PG&E to perform their depreciation studies in their last GRCs. Fourth, a 15-year band, in contrast to a shorter time frame, mitigates the adverse impacts on ratepayers.

For Account 364, poles, towers and fixtures, ORA's 15-year historical average results in a net salvage rate of -190%. This contrasts with SCE's requested net salvage of -250% and the currently authorized rate of -100%. Account 364 consists of approximately $857 million of investment and represents the most significant portion of the increase associated with net salvage in this case. Under SCE's proposal the company would recover approximately $2.1 billion in future negative net salvage costs above the $857 million over the remaining lives of the assets. Because of the potential size of revenue requirement increase, ORA urges the Commission to cap the increase in order to mitigate the impact on ratepayers. Under ORA's recommendation the negative net salvage rate of 125% would provide approximately $1.1 billion over the remaining lives of the Account 364 assets.

Similarly for Account 369, services, which consists of approximately $752 million of investment, the currently authorized net salvage rate is, -60%, SCE requests a rate of -100% and ORA recommends the rate be capped at -75%. Over the remaining lives of the Account 369 assets, SCE's proposal would result in negative net salvage costs of $752 million, while ORA's recommendation would result in costs of $564 million.

15.3. TURN's Recommendation

As background to its recommendations, TURN provided the following:

· The Financial Accounting Standards Board (FASB) has issued Statement of Financial Accounting Standard (SFAS) 143 and, in doing so, changed the financial reporting requirements for retirement obligations. Where the entity has a legal obligation to remove an asset, it has an "asset retirement obligation" or ARO for which it must capitalize the discounted "fair value" and depreciate that amount as a component of the original asset cost. SFAS 143 reminded regulated entities such as Edison that the treatment of retirement obligations that did not meet the definition of an ARO might still meet the requirements of SFAS 71 for reporting as a regulatory liability.

· Concurrent with its implementation of SFAS 143, Edison reported a regulatory liability for its costs of removal for such non-ARO assets. In doing so, Edison acknowledged that under present ratemaking practices the Commission expects these costs to be incurred in the future and, to the extent they are not, understands that "future rates will be reduced by corresponding amounts."⁶⁸ But at this point the Commission's expectation and understanding appear to be implicit; certainly Edison did not point to any specific Commission decision that explicitly set forth such an expectation and understanding.

· Under Edison's current depreciation rates, the non-ARO regulatory liability grew by $90 million in 2004, to reach a total of $2.112 billion at the end of that year. If the Commission approves Edison's requested depreciation rates, it can expect this non-ARO amount to grow even more rapidly in 2006 and beyond.

15.3.1. Explicit Recognition of
SCE's Non-ARO Liability

According to TURN, in past years SCE has collected in rates several billion dollars based on the expectation that it will spend those amounts at some point in the future for the costs of removing assets that will be been retired. TURN expects this pattern to continue for the foreseeable future; that is SCE will collect in current rates an amount for costs of removal that far exceeds the current removal costs, with the excess intended to recover removal costs that are expected to be incurred in the future.

Prior to the enactment of SFAS 143, the full amount of the amounts collected-for-but-not-yet-spent-on removal costs appeared as an undifferentiated amount of "accumulated depreciation" on SCE's financial statements. SFAS 143 distinguished between removal activities that companies were legally obligated to undertake, and those that were not compelled by any such legal obligation. The former were deemed AROs and, while amounting to a substantial past and ongoing expense for SCE's customers, AROs are not the subject of TURN's dispute.

According to TURN, When FASB issued SFAS 143, it concluded that asset retirement costs that are not associated with an ARO might still warrant treatment as a regulatory liability "if the requirements of Statement [of Financial Accounting Standards No.] 71 are met."⁶⁹ In recent years SCE has reported a regulatory liability for its accumulated depreciation amounts associated with plant removal costs that do not meet the definition of an ARO (non-ARO). TURN argues that with this action SCE has demonstrated its determination that the requirements of SFAS 71 are indeed met for those costs.

TURN recommends that the Commission explicitly recognize, for ratemaking purposes, the regulatory liability associated with the non-ARO accrual. TURN asserts that, given the amount at stake and that such recognition is already implicit (as evidenced by the regulatory liability created for financial reporting purposes), the Commission should make such explicit recognition and eliminate any future doubt or dispute about the ratemaking treatment of the non-ARO balance.

TURN acknowledges that there probably is very little risk that anything will occur over the next years and decades that would jeopardize ratepayers' interest in the funds collected to date, as well as those collected going forward, for the costs of removing Edison's utility assets. But given how high the stakes are, and how relatively easy it is to mitigate, if not eliminate, the risk, TURN asserts that the Commission should make explicit the obligation to either spend the funds on costs of removal or return the balance to ratepayers.

TURN notes that asset removal costs are just one of several examples of costs funded in current rates even though the utility is unlikely to incur those costs until many years in the future. TURN specifically argues that just as the Commission directed the establishment of a Post-Retirement Benefits Other than Pensions (PBOP) regulatory asset for regulatory accounting purposes after SFAS 106 was implemented, it should recognize the non-ARO regulatory liability for regulatory accounting and ratemaking purposes.

15.3.2. Reporting Requirements

TURN also recommends that SCE separately identify and report non-ARO costs of removal in all future reports, rate cases, and depreciation studies. According to TURN, this is consistent with the separate subsidiary records the utility is required to maintain for the purposes of identifying the amount of specific allowances collected in rates for non-legal retirement obligations included in the depreciation accruals. TURN notes that (1) SCE's witnesses stated that they would not oppose such a separation of the utility's depreciation showing between plant recovery and cost of removal collection; (2) such a showing should require minimal additional effort, since SCE already maintains a subsidiary ledger in this manner; and (3) the greater specificity will be more consistent with the requirements of SFAS 143 and FERC Order 631 and should provide an opportunity for improved regulatory analysis of these matters.

15.3.3. TURN's Analysis of SCE's Cost
of Removal Proposal

Regarding SCE's proposal for determining costs of removal, TURN makes the following observations and criticisms:

· Nothing in SCE's depreciation study attributes any specific recorded increase in removal costs, much less any proposed increase in future removal costs, to any particular factor other than inflation affecting the associated labor and other costs related to that removal.

· Given that SCE's proposed depreciation rates are so driven by the forecast of future costs of removal, which are in turn extremely dependent on assumptions about future inflation, the Commission must require a demonstration of the reasonableness of the future inflation assumptions. In preparing a forecast of future costs of removal, the utility should attempt to use information that it believes is going to be the most accurate in terms of what the actual cost will be at that point in the future when the cost is incurred.

· SCE failed to make any attempt to develop an accurate inflation rate for use in its forecast of future costs of removal. Instead, it calculated ratios for plant installed decades in the past, a process that means the ratios reflect the level of inflation or cost escalation SCE experienced over those past decades. In other words, if the plant was originally installed in 1950 and removed in 2003, the inflation from 1950 through the present is reflected in the resulting ratio.

· The Handy-Whitman Index, a standard measure of cost escalation, indicates that the cost escalation experienced in the last 45 years averaged 5% per year. However, for the equivalent costs over the past ten years, cost escalation is 2.82% per year on average. Also, in its 2003 GRC, SCE relied upon a forecast of future inflation of 2.65% over the life of its transmission and distribution equipment.

· SCE is seeking a rate increase of approximately $130 million attributable entirely to changed depreciation rates for plant in service as of 2003. Before the Commission finds such an increase reasonable, it must assure itself that the underlying calculations are valid. The fact that the calculations do indeed reflect the extrapolation of future net salvage costs based on retirements that, with rare exception, amount to less than 10% of the plant in service, is cause for concern.

· SCE largely if not entirely ignores potential reductions to future removal costs from a number of improvement initiatives it is undertaking. Specifically, SCE's "infrastructure replacement program" has as one of its underlying goals the replacement of more utility equipment before failure, rather than at or after failure. The successful strategic targeting of replacements should reduce the associated costs of removal, due to lower labor costs and reduced inflation impact. The removal can be performed on a scheduled basis, thus minimizing the risk that the associated work will entail overtime or contract labor. Furthermore, the earlier removal of a piece of equipment will reduce the impact that inflation has on the removal costs for that equipment. Also, SCE is undertaking "process improvement initiatives" in its "Business Process Integration" program. SCE states that the improvements are "expected to yield cost benefits associated with replacement and removal costs." Yet nowhere are such benefits reflected in the costs of removal that underlie SCE's proposed depreciation rates. This failure to include such future costs reductions in the forecast depreciation rates is another reason for the Commission to reject the utility's proposed rates.

15.3.4. TURN's Cost of Removal Proposal

In its prepared testimony, TURN proposed a number of alternatives for determining future costs of removal for purposes of establishing depreciation rates for this GRC cycle. After considering the points raised in the utility's rebuttal and the record evidence developed during hearings, TURN recommends that the Commission adopt a net present-value based approach to calculating the future costs of removal and, by extension, the net salvage ratios used to derive depreciation rates. Specifically, TURN recommends that the Commission determine the net present value in 2006 dollars of Edison's forecast removal costs, and then add to that amount a component intended to reflect inflation likely to be experienced during the rate case cycle. In the utility's next GRC, the Commission can compare the forecast inflation with the inflation the utility experienced and make any necessary adjustments on a going-forward basis. TURN submits that any error between a forecast of inflation and actual inflation over the next three or four years is likely to be far smaller than the error between forecast and actual over the next three or four decades.

TURN recommends that the Commission could leave the non-ARO regulatory liability as an offset to rate base for the time being, noting that should the SCE's depreciation accrual continue to grow at a rate that causes the Commission concern, it could consider in the future whether to amortize some or the entire amount of that liability.

15.4. SCE's Response

In response, SCE notes that neither ORA nor TURN object to SCE's depreciation life estimates. Their differences stem entirely from differences in net salvage estimates and methodologies.

15.4.1. SCE Response to ORA

SCE notes that ORA and the company agree on a number of significant issues:

· The methodology used for depreciation rates - i.e., the Commission approved straight-line remaining life method;

· The depreciation life estimates;

· The depreciation levels for 89 of the 101 plant categories, including various coal;

· Hydro, nuclear, T&D, and general plant accounts;

· Increased removal costs; and

· A need to increase the accrual for net salvage costs.

ORA took issue with SCE's net salvage proposals for 12 out of 18 T&D FERC plant accounts. According to SCE, ORA (1) skews its net salvage estimates by choosing a simple 15-year average in those transmission and distribution accounts where SCE's net salvage costs have been increasing, thus dampening the effect of recent increasing; (2) ignores net salvage cost trends unless they are decreasing; (3) fails to use plant-weighted averages, which gives undue influence to smaller retirements; (4) rounds to less negative (or more positive) net salvage estimates instead of using the 15-year average; and (5) limits the net salvage estimates to an arbitrary 25% increase in two distribution line accounts in order to mitigate the needed increase in the depreciation rate. SCE argues that ORA's approach is results oriented and has a negative effect on depreciation rates, especially on distribution line accounts, which comprises of about 85% of SCE's depreciation expense request. SCE's authorized composite rate for distribution lines is 4.35% compared with ORA's proposed 4.33% for these assets.

While believing that its proposal provides the best estimates at this point in time and are even a bit conservative, SCE understands that the Commission might decide to take a measured approach to addressing the required change in depreciation rates in order to reduce rate impacts. With that in mind, SCE would support the use of levels of net salvage costs that are less than those it proposed, but equal to or greater than the net salvage cost estimates of ORA, as a good first step to establishing appropriate depreciation rates, understanding that these estimates will be re-evaluated in SCE's next GRC. Under this approach, SCE would support the following net salvage estimates as representing a reasonable middle ground between SCE estimates and ORA's proposed mitigation:

		SCE's	Compromise	ORA
		Application	Proposal	Proposal
Account
	Transmission
353	Station Equipment	-5%	0%	5%
354	Towers and Fixtures	-85%	-70%	-70%
355	Poles and Fixtures	-85%	-70%	-70%
356	OH Conductors & Devices	-95%	-80%	-80%
357	UG Conduit	-10%	0%	0%
	Distribution
362	Station Equipment	-15%	-10%	-10%
364	Poles, Towers and Fixtures	-250%	-190%	-125%
365	OH Conductors & Devices	-120%	-110%	-100%
367	UG Conductors & Devices	-70%	-60%	-60%
368	Line Transformers	-10%	-10%	0%
369	Services	-100%	-90%	-75%
373	Streetlighting	-30%	-25%	-15%

15.4.2. SCE Response to TURN

15.4.2.1. Recognition of the
Non-ARO Liability

SCE disputes TURN's recommendation that the Commission should explicitly recognize the non-ARO liability. According to SCE, SFAS 143 does not dictate how either legal AROs or non-AROs should be treated in ratemaking. SFAS 143 is a financial accounting requirement that deals with the identification, measurement, and recording of legal liabilities associated with retirements of tangible, long-lived assets like SCE's nuclear generating stations and is designed to standardize the way that companies report removal costs when there is a legal obligation to remove or dispose of an asset.

SCE states that FERC Order No. 631 did not change the accounting for non-ARO removal expenses. It recognizes SFAS 143 by amending FERC's Uniform System of Accounts to account for AROs. Like SFAS 143, it adheres to existing accounting by allowing recognition of timing differences that may arise for rate-regulated entities. According to SCE, FERC expressly concludes that there is no fundamental reason to change accounting concepts for costs that do not qualify as legal requirement obligations (i.e., non-AROs). SCE also provided examples of state commissions that, after evaluating the impact of using the mechanics of SFAS 143 and FERC Order No. 631 in rate-regulation to recover net salvage costs, agree with FERC on this issue.⁷⁰

SCE points out that SFAS 143 is not the first instance for which SCE has recorded regulatory assets and liabilities to account for differences between ratemaking and financial accounting. SCE reports several regulatory assets and liabilities in its financial statements. SCE's largest regulatory assets include flow-through taxes, transition cost deferral of rate reduction notes, and its unamortized nuclear and coal investments. SCE states that its year-end 2004 financial statements, shows that the total amount of SCE's regulatory assets is about the same as the total amount of its regulatory liabilities.

It is SCE's position that financial reporting changes required by SFAS 143 do not affect the underlying regulatory economics of the retirement obligations, because the goals of ratemaking and those of SFAS 143 are not the same and require very different approaches. SCE states that while SFAS 143 prescribes the measurement of legal retirement obligations on the balance sheet to provide investors a better idea of a company's future legal asset retirement obligations, in ratemaking, proper depreciation principles are concerned with measuring the service value of an asset (including the future removal cost expenditure) used during an accounting period for purposes of determining a fair revenue requirement to charge ratepayers.

SCE also states that TURN overlooks the fact that the Commission already recognizes the entire accumulated depreciation as a liability (not just that portion related to non-legal AROs) and therefore offsets the rate base by that amount. SCE also argues that the Commission has exercised prudent regulatory oversight regarding differences between the amounts collected by a utility and the amount spent after utility plant is retired. To demonstrate this, SCE points to the case of the plant divestitures that took place as a result of industry restructuring. In 1998, when SCE divested its twelve oil-/gas-fired generating stations, the purchasers assumed the responsibility for the decommissioning. Consequently, the Commission ordered SCE to refund to ratepayers the full amount of accumulated depreciation through the gain/loss calculation, including those amounts collected for plant decommissioning. SCE states that there was never a risk of SCE "disappearing" with ratepayer monies. SCE also cites the divesture of its Fuel Oil Pipeline Facilities in 2003 (again the buyer assumed the future decommissioning obligation), the Commission directed SCE to return to electric utility ratepayers the accumulated decommissioning expenses that would not have to be spent. This amounted to a $39.7 million refund to ratepayers.

Finally, SCE claims that TURN's proposal may unnecessarily limit the Commission's options. SCE witness Umbaugh explained:

"I don't think it's an uncertainty that has been an issue in the past and shouldn't be a concern, because the Commission can always make the decision at some point in time as to how to treat it. To require it to be a refund today kind of locks them in and eliminates one of the options that they have essentially to continue to adjust future rates going forward. I mean they could also have decided that if the cost of removal turns out to be more than they've allowed, they could have a one-time surcharge rather than to spread that out in the future. I mean the Commission has alternatives available to it."⁷¹

15.4.2.2. Separate Identification of
Removal Cost Depreciation

Regarding TURN's assertion that it is critical that the Commission require that SCE separately identify the accumulated depreciation and depreciation rates associated with non-ARO removal costs, SCE states that TURN is being unnecessarily alarmist in its appeal. SCE it already separately accounts for non-ARO removal costs within FERC Account 108, Accumulated Provision for Depreciation, in accordance with regulatory accounting requirements, and has disclosed these costs in the audited financial statements filed with the Securities and Exchange Commission in accordance with financial reporting requirements. SCE unbundles its depreciation rates to separately record its removal cost accrual component in order to support this accounting.

15.4.2.3. TURN's Analysis of SCE's
Cost of Removal Proposal

SCE asserts that it properly, even conservatively, reflected inflation in its cost of removal proposal. TURN's critique ignores the fact that the age of future retirements will be substantially older than past retirements (for example, future distribution overhead conductor will be about four times as old as the retirements for 1994-2003). Because of this, SCE's net salvage estimates reflect a substantial reduction in future inflation.

TURN also asked SCE's depreciation witnesses a series of questions regarding a hypothetical distribution pole example to suggest that SCE's net salvage ratios reflect past levels of inflation and do not appropriately reflect future expectations. According to SCE, what TURN failed to address in this line of questioning is SCE's actual net salvage estimates, which were based on judgments that considered the representative nature of the recorded retirements. Contrary to TURN's hypothetical, SCE's actual proposed net salvage ratio for distribution poles understates the impact of future inflation. Recent recorded net salvage costs (2000-2003) have amounted to $1,490 per pole (nominal dollars). Over the 36-year remaining life of the existing distribution pole investment, these costs can be expected to increase with inflation. However, SCE's proposed 250% net salvage ratio provides future cost recovery of only $1,340 per pole for the existing distribution pole investment. According to SCE, its proposal actually reflects a cost deflation.

15.4.2.4. TURN's Cost of Removal Proposal

SCE criticizes TURN's NPV proposal as a last minute tack-on to the number of alternatives contained in its original testimony where, according to SCE, there is less than eight lines of explanation on the NPV proposal and no discussion of its impact.

According to SCE there is a logic gap in TURN's NPV proposal. That is if the service value of the asset is to be adjusted to current price levels, then the future net salvage and the historical original cost should both be adjusted. Such a modification to TURN's NPV approach would require an adjustment to the historical cost of the asset. SCE also refers the computational problems associated with TURN's NPV approach, especially for mass property. A properly calculated present value approach would require, by vintage, the determination of the timing of widely dispersed future retirements, consistent with an account's survivor curve. SCE asserts that TURN's proposal fails to do this and more importantly, it also fails to include an annual interest accretion in its determination. According to SCE, the complexity of the calculation necessary to do the NPV method is little different from that of the SFAS 143 approach, which TURN's witness ultimately rejects as too complicated.

SCE notes that TURN attempted to remedy some of its flaws by revising the NPV approach in its Opening Brief by providing periodic updates, changing the discount year from 2003 to 2006, adding an inflation adjustment between GRC cycles, and so forth. However, SCE asserts that the revisions do nothing to solve the inherent problems underlying the NPV method. Also, NARUC points out other reasons why interest-rate methods like TURN's NPV approach should be rejected, including "problems of annuity mathematics" and "heavy accruals due to greater interest toward the end of a property's life [which] can produce wide differences between accumulated accruals and the cost being recovered if retirements occur only a year or two from the estimated time."

15.5. SDG&E's Response to TURN

For many of the same reasons given by SCE, SDG&E opposes TURN's recommendation that the Commission make explicit the understanding that amounts received by SCE in rates for future cost of removal must either be spent on such removal or returned to ratepayers. SDG&E states that it is neither necessary nor wise for the Commission to make such an unequivocal declaration.

Regarding TURN's proposal ratemaking treatment for cost of removal, SDG&E states that TURN's opening brief makes a series of factual and conceptual errors in its arguments in favor of deviating from the Commission's long-standing ratemaking treatment of the cost of removal, and then it virtually abandons all of if its own witness' alternative approaches and advocates an approach that was not sponsored by any witness. For many of the same reasons given by SCE, SDG&E asserts that TURN's proposal unreasonably shifts recovery of removal costs from current ratepayers to future ratepayers, and is inconsistent with other aspects of ratemaking.

In support of SCE's methodology for calculating cost of removal, SDG&E argues that SCE has used a reasonable level of future inflation in estimating future nominal removal costs. SDG&E claims that on an original-cost basis, more recently installed plant has a greater weight, so the result is more reflective of inflation in a more recent time period than the full average useful life. SDG&E states that contrary to TURN's claim that SCE used the average inflation over the past 45 years, the methodology used by SCE actually reflects average inflation over a much shorter period.

15.6. PG&E's Response to TURN

For many of the same reasons given by SCE and SDG&E, PG&E opposes TURN's recommendation that the Commission make explicit the understanding that amounts received by SCE in rates for future cost of removal must either be spent on such removal or returned to ratepayers. PG&E states the Commission should recognize that removal costs are ratepayer funded and that any excess accruals should be considered as such, if and when excess accruals become apparent. In PG&E's opinion, while returning such funds to customers should not be precluded by retroactive ratemaking or other concerns, neither should it be mandated, without taking into account all possible shortfalls in collections or other pertinent factors. PG&E concludes that all of these issues are more appropriately addressed at the time the issues arise, not in the abstract in SCE's GRC.

For many of the same reasons given by SCE and SDG&E, PG&E opposes TURN's cost of removal recommendation. While recommending that TURN's proposal be rejected, based on the record in this case, TURN believes that in the future, technical and generic issues such as the alternatives proposed by TURN would be more effectively and efficiently addressed in a generic statewide proceeding, if at all. PG&E recommends that the Commission should establish as a future policy that it is generally not appropriate for this Commission to consider proposals by interveners for technical adjustments to generic ratemaking policy in individual utility general rate cases.

15.7. Discussion

15.7.1. Recognition of a Regulatory Liability

TURN's request that the balance of funds collected for cost of removal related to non-ARO assets be recognized as a regulatory liability for ratemaking purposes is reasonable and will be adopted. The balance of this asset is substantial, amounting to $2.1 billion as of the end of 2004. This balance is already recognized as a regulatory liability for financial reporting purposes. SCE has not demonstrated any potential harm to the company. In fact, SCE indicates that in some ways the Commission already recognizes and treats such assets in the manner requested by TURN. SCE points to the Commission actions that refunded to ratepayers the decommissioning funds no longer needed for 12 oil/gas generating stations and the fuel oil pipeline, which were all divested. TURN acknowledges that these actions were consistent with the explicit recognition that it now requests. Formal recognition of our ratemaking responsibilities is a reasonable course of action and will establish regulatory certainty regarding ratemaking treatment and principles that all parties generally agree is appropriate.

SCE also argues that adoption of TURN's request might limit the Commission's options in dealing with unspent funds. We understand our options to be a refund through future rate reductions or payment of future costs with no corresponding effect on future rates.⁷² There is some flexibility in these options. For example, the period over which the refund in rates should occur is left open. Even so, such limitations are not unreasonable when considering the magnitude of the asset balance that has accumulated, and which will be increased in the future, with ratepayer funding.

15.7.2. Reporting Requirements

Regarding TURN's request that the CPUC require SCE to separately identify the accumulated depreciation and depreciation rates associated with non-ARO removal costs, there is no issue. SCE already separately accounts for non-ARO removal costs within FERC Account 108, Accumulated Provision for Depreciation, in accordance with regulatory accounting requirements, and has disclosed these costs in the audited financial statements filed with the Securities and Exchange Commission in accordance with financial reporting requirements.

15.7.3. Cost Removal

There are two considerations in determining what the appropriate annual accrual for net salvage (cost of removal and salvage) should be. One consideration is the details of the determination of the accrual. In this case SCE and ORA provide showings that analyze recorded net salvage as a percentage of original cost and then determine and apply a factor to all such properties placed into service. Costs would be recovered over the remaining lives of the properties.

The second consideration is the state of the accumulated accrual as it relates to existing plant. That is whether, based on the most recent determinations and assumptions regarding annual net salvage accruals, sufficient funds will be recovered over the remaining lives of the existing assets to remove them when they are retired.

Regarding the details of the determination of the accrual, in the past, SCE and the Commission have relied on the historical relationship of recorded net salvage costs and recorded retirements to develop rates to apply to future plant additions. This is consistent with practices of many other state commissions. However, when projected net salvage become substantial, in some cases substantially exceeding the original cost of the associated plant, we also have a responsibility to determine whether past practices are consistent with producing the most reliable net salvage projections.

Both ORA and TURN criticize SCE for not demonstrating the reasonableness of the escalation implicit in its cost of removal estimates. There is reason for such concerns. Inflation is the primary reason for the significant increases in historic and projected costs of removal. Variations in assumed inflation over a plant asset's life can substantially affect the cost of removal accrual over that time period. Consider the following net salvage analysis for a distribution pole replacement.

Assumptions:

$5,499 - Cost of Installation of new pole in 2005

$2,331 - Estimated Cost of Removal in 2005 dollars

$ 233 - Estimated Gross Salvage in 2005 dollars

$2,098 - Estimated Net Salvage in 2005 dollars

45 years - Expected and Actual Life of New Pole

There is an implicit annual escalation related to the accumulated net salvage in each of the conditions indicated in the table. For the currently authorized condition (100% net salvage rate), the table shows $5,499 would be accrued for net salvage over the 45 life. This implicitly reflects an annual escalation increase of 2.17% when compared to the 2005 dollar estimate of $2,098. If escalation related to net salvage increases by an average of 2.17%/year for the next 45 years, the net salvage rate for account 364 could be left at 100%. However if escalation reflects the historical compounded rate of 4.75%, the net salvage ratio would have to be increased to 308%. As the net salvage rate increases, the implicit annual escalation likewise increases. In examining the results, none of the conditions result in an implied escalation that is absurd. While 2% may be low and 5% may be high, either number or anything in between is not out of a zone of reasonableness. TURN has pointed out that for the 1993- 2003 time period, the Handy Whitman Index estimates annual cost escalation of 2.82% for distribution pole related costs. However, whether that would be an appropriate average rate for the next 45 years is questionable.

We note that the record in this proceeding does not include a forecast of inflation over the next 45 years. We do not even know if such forecasts are even made. However, in its next GRC, SCE should, as part of its account-by-account analysis, analyze the effects of past inflation on its proposed cost of removal rates and justify the implicit inflation rates reflected in its proposed rates.

Regarding the state of the accumulated accrual as it relates to existing plant, SCE has provided evidence indicating that with its proposed net salvage rate for distribution poles included in Account 364, it would not accumulate sufficient funds to retire the existing poles, even if the removal costs remained at recent recorded levels, unadjusted for inflation over the remaining lives of the existing poles. This supports the need for a significant increase in the net salvage rate, at least as it relates to distribution poles. However, there is not much of an explanation of why this situation is likely to occur. For instance, was it solely due to recent Commission decisions which held the net salvage rate constant or was it due to past understating of the net salvage rate due to methodological flaws? Depending on the cause, there may be more appropriate ways to account for the increased removal costs not covered by net salvage rates. In its next GRC, SCE should, as part of its account by account analysis, provide analyses similar to the one for distribution poles, which quantifies potential accrual deficiencies for the future removal costs of existing assets. SCE should provide an analysis of what is causing any likely deficiency. With that information, we can determine the proper course of action to address the deficiency.

We do note that despite the distribution pole situation described above, by the nature of the established methodology where SCE is paying off current removal costs, while rates are being collected to fund future costs that are much higher than current costs, the non-ARO balance, which is already over $2 billion, will continue to grow. At no time in the foreseeable future will SCE be short funds to cover its removal or net salvage costs.

In that regard it is not urgent that this issue be definitively decided at this time. Due to the large dollars at stake, and the wide range of possibilities, we prefer to be conservative in adjusting net salvage ratios, rates or accruals. In general, ORA's use of the 15-year historical average accomplishes that. Also, SCE did not dispute that it has used 15 years of historical data in the past, nor did it dispute that both SDG&E and PG&E used 15 years of historical data in establishing their current rates. Therefore, except for Accounts 364 and 369, we will use ORA's recommended net salvage rates based on the 15 year average.

Because of the additional information provided by SCE to support its request for Account 364, we will adopt its proposed compromise net salvage rate of -190%. SCE did not provide such information for Account 369, and again due to our preference to proceed in a conservative manner, we will adopt ORA's proposal to cap the increase at -75%.

TURN now recommends that cost of removal be determined using a net present value methodology that provides for updating the effects of inflation from one GRC to the next. The focus of the cost of removal issue has been in accounts such as wood poles where cost of removal and depreciation expenses determinations are subject to mass accounting where properties are continually being placed into service while others are being retired. PG&E's witness indicated that the accounting necessary under the NPV methodology for calculating net salvage costs, while more complicated than that under current procedures, could be done. However, it is not clear that, in the long-term, the results using TURN's proposal would be significantly different from that derived using the traditional net salvage procedures. Assuming costs are fully recoverable under TURN's proposal, in the long-term, the newer poles will be accruing lower removal costs than the older poles. However under mass accounting, it is not clear that the accumulated removal costs and subsequent rate effect would be significantly different than if the same annual removal cost were applied to all poles. We would prefer not to change the methodology for calculating costs of removal until we are convinced there is a need to do so, there is means to do so, and the means provide results that are meaningfully different and appropriate. At this time, we are not convinced that the net present value methodology as proposed by TURN should be adopted.

The conservative measures for determining net salvage in this decision are not permanent. In the future, we expect a more thorough record in order to make more definitive decisions. In its next GRC, by whatever method SCE proposes to estimate net salvage, it must provide a detailed analysis justifying the reasonableness of applying that method on a forward going basis. For example, inflation rates that are implicit in the proposed cost of removal rates justified. Also, if TURN wishes to reintroduce its net present value recommendation, it should make a full and more detailed showing on how it would be implemented and calculated for all the different classes of plant and what the long term difference is when compared to the methods used by ORA and SCE. Detailed cost of removal showings in the next GRC, which address our concerns expressed in today's decision, will provide the principal guidance as to whether future net salvage should be increased, be decreased, or remain the same.

Regarding PG&E's proposal to limit technical adjustments, we do not feel it is appropriate or necessary to institute a generic proceeding every time a party, other than a utility, proposes technical adjustments to existing methodologies. Generic statewide proceedings should be reserved for broader topics that would present all policy and technical proposals for consideration. In situations where technical adjustment proposals are the same as those proposed in prior proceedings, Commission precedent can be used as reason to accept or reject such proposals, unless new and relevant information indicates otherwise.

⁶⁸ SFAS 71, ¶11(b).

⁶⁹ Ex. 348 (Majoros Testimony for TURN), p. 10, citing SFAS 143 ¶B73.

⁷⁰ SCE cites Washington Gas Light Co., Case No. 7689, Maryland PSC, 1984 Md. PSC LEXIS 49 (1983); PacifiCorp, Idaho Public Utilities Commission Order No. 29385, 2003 Ida. PUC LEXIS 173, *5 (December 2, 2003);.and Central Vermont Public Service Corp., Docket Nos. 6946 and 6988, Vermont PSB, 2005 Vt. PUC LEXIS 65, * 226-27.

⁷¹ SCE, Umbaugh, Tr. 25/2508.

⁷² Statement of Financial Accounting Standards No. 71, Appendix C: Basis for Conclusions, Paragraph 79 (b) states: "A regulator can provide rates intended to recover costs that are expected to be incurred in the future. Paragraphs 38 and 39 illustrate that possibility. The resulting increased charges to customers are liabilities and not revenues for the enterprise-the enterprise undertakes to provide the services for which the increased charges were collected, and it is obligated to return those increased charges if the future cost does not occur. The obligation will be fulfilled either by refunding the increased charges through future rate reductions or by paying the future costs with no corresponding effect on future rates. The resulting increases in charges to customers are unearned revenues until they are earned by their use for the intended purpose."