20. Post Test Year Ratemaking

SCE proposes to extend its current Post Test Year Ratemaking (PTYR) mechanism that was adopted in its 2003 GRC. Its proposal has the following features:

· An annual advice letter providing notice of the revenue requirement change for the following year;

· O&M escalation using the GRC escalation rate methodology, updated at the time of the advice letter filing;

· Capital-related cost increases based on SCE's forecast of post test year capital expenditures, updated for changes in SCE's authorized cost of capital;

· An annual revenue adjustment to reflect the number of nuclear refueling outages at San Onofre Nuclear Generating Station (SONGS) and cost per refueling outage as adopted in this proceeding and updated for escalation;

· A revenue adjustment if necessary to reflect major maintenance outages at Four Corners Generating Station; and

· A mechanism to address major exogenous changes in SCE's costs.

In authorizing post test year ratemaking for 2004 and 2005, the Commission imposed a requirement that if SCE's revenue requirement increase were to exceed $150 million in either year, SCE would be required to submit an application for that year, rather than an advice letter. SCE asserts that unlike its previous GRC application, this application contains testimony supporting its proposed capital expenditures through 2008, not just through the test year. Because of this, SCE states that there is no substantial component of SCE's post test year ratemaking mechanism that is not addressed by testimony in this application, and the Commission should not require SCE to submit a second application in 2006 or 2007 to reapprove its proposed mechanism.

Based on its proposed mechanism and test year 2006 estimates, SCE estimates a 2007 revenue requirement increase of $108,485,000 over its proposed 2006 level and a 2008 revenue requirement increase of $113,015,000 over its estimated 2007 level.

SCE argues that adoption of its proposal is necessary for it to be able to cover its costs of doing business in 2007 and 2008. The proposed increases will cover cost increases caused by increased capital spending, including the need to replace aging infrastructure facilities, and the impact of price inflation on operating expenses.

ORA does not oppose a mechanism that provides SCE the opportunity to earn its authorized rate of return for its GRC related operations during the years 2007-2009. However, ORA does not agree with SCE's PTYR proposal and recommends the following:

· A Consumer Price Index (CPI) indexing method to determine SCE's PTYR rates between general rate cases. This method will allow SCE's rates to increase with inflation, and encourage SCE to be efficient, productive, and innovative.

· As an alternative to its proposed CPI indexing method ORA recommends that SCE's PTYR capital additions be based upon historical attrition rates rather than estimates that are based on SCE's preliminary planning.

· ORA recommends one additional attrition year in 2009 be incorporated into the post test year ratemaking cycle.

ORA states that its proposed CPI indexing method is predicated on the following:

· The Commission's position is that attrition rate changes are not an entitlement and some utilities have been denied attrition rate increases recently.

· Using a properly developed CPI indexing method between rate cases, give utilities an incentive to minimize their costs.

· The Commission has adopted indexing methods for utilities in the past including:

    · Telephone utilities using the New Regulatory Framework authorized in D.89-10-031;

    · Class C and D water utilities in D.92-03-093;

    · Pacific Gas and Electric Company in D.04-05-055; and,

    · Southern California Gas Company and San Diego Gas and Electric Company in D.05-03-023.

· A CPI indexing method encourages efficiency, cost savings, and innovation.

· SCE's method discourages efficiency, cost savings, and innovation.

· ORA's CPI indexing method allows for a balanced and reasonable method to calculate base margin revenue requirement during the PTYR period.

· ORA's CPI indexing method is easy to use.

ORA's recommendation that one additional year in 2009 be incorporated into the post test year ratemaking cycle is based on a number of factors. First, inflation has been modest over the past fifteen years. Second, a longer attrition period gives more time between the issuance of a decision and the filing of the next case. Third, extending the attrition period will provide the Commission with more data to use in determine SCE's future rates. Fourth, another year of attrition will give SCE another year free of the Commission's oversight. Fifth, utilities, because of the regulatory protections, no longer need general rate cases every three-years. Finally, a four-year attrition period is consistent with the current longer regulatory cycle used by PG&E, SoCalGas, SDG&E, Southwest Gas, and most small utilities operating in California.

Based on its proposed CPI mechanism and test year 2006 estimates, ORA estimates a 2007 revenue requirement increase of $2,030,000 over its recommended 2006 revenue requirement level and a 2008 revenue requirement decrease of $9,825,000 to its estimated 2007 revenue requirement level.

As an alternative to ORA's primary recommendation to use a CPI indexing method to set SCE's Attrition rates, ORA recommends annual increases that do not exceed the level based on the historical attrition rate setting method. This involves: (1) increasing operational expenses for inflation and (2) increasing capital-related costs based on recent historical plant additions.

Regarding capital related costs, ORA states:

The capital related portion of SCE's revenue requirement estimate for the post test year is based in part on the accumulated plant balances estimated for 2003 and annual plant additions for each of the years 2004 and 2005. Capital related costs such as net return on rate base, income taxes, property taxes and depreciation expense are directly related to the accumulated plant balance for that year. ORA does not dispute the use of the estimated accumulated plant balance at the end of 2006 as the starting point to estimate plant balances for both 2007 and 2008. ORA used its estimate of the end of year 2006 plant balances for this purpose. However, SCE's estimates of plant additions for 2007 and 2008 are budget-based and are not based on SCE's historic capital additions consistent with traditional attrition relief. ORA is not able to conduct a detailed analyses of SCE's post test-year plant additions. For these reasons, ORA based its estimates of plant additions primarily on recent historic recorded plant additions, which is consistent with the historical attrition method used by this Commission.106

Based on its alternative methodology and test year 2006 estimates, ORA estimates a 2007 revenue requirement increase of $97,840,000 over its recommended 2006 revenue requirement level and a 2008 revenue requirement increase of $12,080,000 over its estimated 2007 revenue requirement level.

Aglet also does not agree with SCE's PTYR proposal. Consistent with ORA's primary recommendation, Aglet recommends a CPI indexing method to adjust SCE's base rate revenue requirements between GRCs. Aglet makes the following points in supporting the use of a CPI index:

· Simplicity - Calculation of CPI-based revenue requirements is much easier than SCE's calculations.

· Comprehension by Consumers - The CPI is a measure of inflation that is widely recognized by consumers. Small customers identify the connection between retail price changes and the CPI, and large customers have seen the CPI used as a price index in long-term contracts. Reliance on the CPI will help customers understand the reasons behind rate changes between general rate cases, leading to increased consumer confidence in Commission regulation.

· Verification - The CPI is objective and easily verified. The utilities and Commission staff subscribe to Global Insight publications, which include CPI forecasts. These attributes lead to public trust in the relevant calculations, and use without controversy.

· Revision - The CPI is not generally subject to revision. Reliance on an October forecast of the next year's CPI introduces a modest amount of uncertainty, but that risk is no different from the uncertainty the Commission accepts in forecasting test year revenue requirements for utilities generally.

· Stability - The CPI is generally less volatile than utility price indices. SCE has reported that volatility is a serious problem with industry-specific indices because it destabilizes utility rates and earnings, which could lead to an increase in cost of capital.

· Bias - Aglet is aware of no record evidence that shows long-term CPI bias compared to utility price indices.

Aglet asserts that, taken as a whole, the evidence shows that for ratemaking purposes the CPI is superior to utility price indices as a measure of price escalation.

With regard to SCE's post test year capital budgets, Aglet notes that only $40 million of the capital costs (compared to more than $1.8 billion in capital additions for each of the years 2007 and 2008) are supported by cost effectiveness analysis. In agreeing with ORA's argument that budget-based ratemaking in 2007 and 2008 may encourage excessive capitalization of plant additions without sufficient safeguards to protect ratepayer interest as to the need or reasonableness of the plant additions, Aglet states:

The current and proposed post-test year ratemaking schemes give SCE an incentive to overestimate its capital forecasts, and later reduce revenue requirements so that virtually all recorded expenditures are recovered in rates. This outcome is very close to recorded cost ratemaking for SCE. Due in part to the scope of general rate cases, no other party has the resources to test the reasonableness of actual capital costs. There is no showing in this proceeding of the reasonableness of SCE's 2004 and 2005 capital spending, and I expect there will be no showing in the next general rate case of the reasonableness of SCE's 2007 and 2008 spending. These are fundamental flaws to budget-based ratemaking.107

TURN recommends that the Commission should adopt a CPI-based escalation of the entire base revenue requirement as recommended by Aglet and the ORA. Regarding SCE's request to continue a "budget based" attrition mechanism, as it relates to capital related revenue requirement for 2007 and 2008, TURN observes that:

The primary responsibility of the ORA and intervenors has traditionally been to review the prudency and reasonableness of historical capital additions and the reasonableness of the test year forecast. Adding to this the need to review the reasonableness of capital budget forecasts two or three attrition years results in an unmanageable task. There is no disincentive for the utility to provide realistic capital forecasts, as it has every incentive to over inflate expected costs and no incentive for cost control.108

In response to the ORA, Aglet and TURN PTYR proposals, SCE makes the following arguments:

· The CPI is not an accurate escalation rate to use for SCE's revenue requirement with respect to expense or capital because it ignores the realities of SCE's need for adequate revenues to support its capital investments. The CPI is a measure of price inflation in the goods and services purchased by consumers, not the goods and services purchased by electric utilities.

· If ORA had used the model for the CPI estimate as it did for the hybrid-attrition forecast, it would have accurately forecast its three-year GRC revenue requirements for the entire GRC cycle. By taking this shortcut, ORA did not recognize that SCE would be required to cut an additional $706 million in capital expenditures in the test year-beyond the adjustments already recommended by ORA in its testimony. This additional reduction in capital expenditures occurs because ORA did not recognize a forecast year-end 2006 Construction Work in Progress (CWIP) balance of $742 million.

· The ORA proposed alternative for cost recovery in 2007 and 2008 is not equivalent to the Attrition formulas adopted by this Commission in SCE's general rate cases in the 1980s and early 1990s. When the Commission adopted Attrition ratemaking for SCE in previous general rate cases, the formula was a combination of the average of SCE recorded capital additions, plus the forecast of large capital projects (usually generation), which comprised the capital portion of the attrition year revenue requirement (adjusted for customer growth). In the attrition methodology, large capital projects were forecast on a work order or project-specific basis and the balance of the capital forecast was the product of a historic average of capital additions. In those earlier years, the large projects were usually in the areas of generation. However, given the capital needs of SCE's transmission and distribution system, initiatives such as the Infrastructure Replacement program are similar to those large generation projects that were separately forecast in the attrition formula when the Commission approved SCE's general rate cases in the 1980s and early 1990s.

· Regarding the ORA and Aglet claim that the CPI-indexing methodology is a simpler approach than SCE's PTYR proposal, the Commission has previously rejected this argument in SCE's 2003 GRC decision, D.04-07-022, "This [CPI] approach may be simple, but it has no other known benefit. Simplicity alone does not prompt us to prefer it over SCE's approach, which provides for separate and therefore, we believe, more accurate treatment of O&M expenses and capital related costs."

· The CPI is projected to increase less rapidly than SCE's O&M escalation indexes over the post-test year period. Thus, the CPI is an inadequate escalator for SCE's O&M expenses.

· Many of SCE's capital assets have long lives and an escalation rate for capital related costs must account for more than contemporaneous changes in prices. The CPI which is based on current information about prices cannot do this. Therefore, reliance on the CPI to project increases in SCE's capital costs should also be rejected.

· ORA and Aglet's reliance on recent settlements with PG&E and SDG&E cases to support their CPI methodology should be rejected. It is well established that settlements in regulatory proceedings are compromises between the parties involved, the terms of which are not precedential.

· The Capital Additions Adjustment Mechanism (CAAM) protects SCE customers by ensuring that authorized revenue requirements are adjusted to recover only the costs of capital investments that are actually made and placed into service. If SCE does not make investments equal to or greater than the authorized level, SCE will return the excess revenue requirement to ratepayers. The reasonableness of SCE's capital investments, scope changes or capital addition substitutions can be reviewed by the Commission in any subsequent general rate cases, just as they have in the past.

· In opposing SCE's PTYR proposal, the ORA, Aglet and TURN implicitly argue that the Commission should reverse its policy decision that authorized funding for the Infrastructure Replacement program. In the 2003 GRC decision, D.04-07-022, the Commission recognized the importance of this program and made a policy decision to authorize increased capital expenditures to support SCE's investments to meet customer growth, load growth and to increase our level of investment in the infrastructure replacement program.

· SCE's proposed capital expenditures, when compared with the level proposed by ORA, will result in greater overall economic activity in Southern California, even after absorbing the cost of the associated revenue requirement. These economic benefits are in addition to the direct benefits from maintaining SCE's system that will be realized by SCE's customers if SCE's proposed capital expenditures are adopted.

· ORA's proposal to extend the attrition mechanism through 2009 lacks merit and should be rejected. ORA's proposal would require the Commission to set rates on recorded data that is over five years old, an approach the Commission has rejected. Also, there is no evidence in this GRC of SCE's capital forecast in 2009. In addition, ORA's reliance on prior settlements as authority for its proposal to extend the attrition mechanism should be rejected because settlements are not precedential.

· The ORA, Aglet and TURN proposals would reverse a policy decision the Commission made only one year ago to support SCE's capital investment program. The adoption of any of these proposals will also lead to the inevitable degradation of the electric system reliability. Therefore, SCE respectfully requests that the Commission resist the temptation of issuing a compromise decision between the ORA, Aglet or TURN proposals and SCE's PTYR proposal.

Rates for post test year 2007 and 2008 could be determined in the same manner as for test year 2006. Estimates of sales, revenues, operating expenses, capital additions, and capital related revenue requirement can be determined for each of the post test years 2007 and 2008 similar to what is done for test year 2006. However, this would be time consuming and complicated in that it would expand the scope and analysis of many aspects of the GRC by a factor of three (three test years versus one test year). Rather than subjecting post test years to the same scrutiny as test years, the Commission has adopted "attrition," and subsequently "post-test year," (PTY) methodologies as substitute measures for determining rates for the time period between test years.

The attrition methodologies adopted in D.85-12-076 and used during much of the 1980's and 1990's, among other things, determined attrition year revenue requirements by escalating operation and maintenance expenses by forecasted inflation factors109 and determining capital related costs based on forecasted attrition year plant additions. SCE essentially followed this procedure in determining its PTY revenue requirements, although it used a budget based approach for forecasting PTY plant additions. ORA also used this methodology in its alternate PTY recommendation. ORA used a four-year average of escalated historic and forecasted plant additions (2004 - 2006) to estimate plant additions for 2007 and 2008. There is merit to this type of analysis because the increase in revenue requirements can directly be tied to the principal factors that cause the operational attrition - inflation as it relates to expenses and annual plant additions. For capital-related costs, the reasonableness of the increases relates directly to the reasonableness of the adopted plant additions.

A consumer price index (CPI) methodology, ORA's principal recommendation, has been recently adopted by the Commission in determining attrition for PG&E and SDG&E.110 Various forms of indexing have also been used in other proceedings such as performance based ratemaking (PBR) for energy utilities and the new regulatory framework (NRF) for telecommunication utilities. Indices can be applied to rates or to revenue requirement.

In one respect, ORA's CPI proposal is similar to previous attrition methodologies. That is it could be assumed its proposal escalates operation and maintenance expenses by a forecasted inflation factor, that being the CPI rather than specific O&M escalation factors. However, it would then follow that ORA's proposal also escalates capital related costs by the CPI. This is in contrast to previous attrition methodologies where capital related costs are calculated based on the effect of forecasted plant additions being added to the test year base. By the CPI method, the escalated capital related costs do not specifically take into consideration the magnitude of post test year plant additions, the increases to depreciation expense and the effects of accumulated depreciation and deferred tax reserves. A determination of the reasonableness of the results is therefore more complicated. A test could be whether the results provided reasonable capital related costs based on a reasonable level of plant in service or plant additions for each of the post test years. For instance, in D.97-07-054, the Commission determined that an indexing methodology applied to capital related costs was not appropriate for Southern California Gas Company because, in fact, annual depreciation accruals exceeded annual plant additions and rate base was therefore declining. Certain capital related costs, such as return on rate base, should therefore have been forecasted to decline rather increase as was reflected in the proposed indexing methodology.111 While a modified indexing methodology was ultimately adopted, the Commission did state:

"We would prefer to adopt a method to take rate base changes into account outside of the indexing formula. A methodology such as a direct revenue offset or adjustment of the benchmark rate of return could accomplish this. However, no party has proposed such a method, and we must rely upon the indexing methodology, in which rate base factors are effectively translated into productivity. SoCal estimates in its comments on the Proposed Decision (p. 4) that the impact of the TURN/DGS formula may result in an effective productivity factor as high as 2.9 percent, which is 1.4 percent above the 1.5 percent final stretch "X" factor. This suggests that it may be possible to translate directly the TURN/DGS formula into a straight productivity figure and thus roughly reconcile the TURN/DGS concept with the indexing methodologies adopted in other PBRs."112

In that proceeding, the Commission considered the effect of an indexing methodology in light of evidence that showed that rate base would likely decline. Because the indexing methodology, as proposed, would likely have resulted in an excessive rate increase, it was rejected and modified.

Under different circumstances, indexing may well be appropriate. In the last PG&E and SDG&E general rate proceedings, there were no claims by any of the parties that a proposed CPI indexing methodology would result in excessive or inadequate rates to cover capital related or other costs. Applicants, as well other interested parties, were apparently satisfied that, in those cases, the CPI methodology would provide reasonable results. The Commission did not find otherwise and adopted a CPI methodology for PG&E and for SDG&E. We are therefore open to the use of CPI indexing to set rates for the post test year period. However, as in A.95-06-002, we must consider the views of dissenting parties and ensure the reasonableness what is adopted. In this case, SCE objects to the use of the CPI methodology and has provided information that casts doubt on the reasonableness of its use for determining post test year revenue requirements.

The CPI methodology proposed by ORA and endorsed by Aglet and TURN increases authorized revenue requirement (essentially expenses and the capital related revenue requirement) by the percentage increase in the CPI and reduces that amount by increased revenues from increased sales. The net amount represents the change in revenues that SCE would see and would have to cover increases caused by expense inflation and incremental costs to serve new customers and increased load. The revenue increases calculated by increasing the authorized revenue requirement by the CPI increases for 2007 and 2008 are offset by increases in the revenues generated by increased sales to the extent that there is essentially no effect on the rates charged to ratepayers under ORA's primary PYTR proposal.

As discussed above, CPI increases, or inflation increases in general, are not linked to the capital expenditure cost increases that the utility incurs but instead relate to capital related costs such as return on rate base, income taxes and depreciation, which are the items that are directly reflected in the revenue requirement. For that reason, a CPI increase may not fairly represent reasonable overall cost increases to the utility. There has to be some kind of check for reasonableness. The check is usually some type of calculation of the revenue increases generated using accepted ratemaking principles and specific assumptions related to those principles. In this case, SCE calculated the capital expenditures associated with ORA's proposal and prioritized projects that could be done within that funding level. The major point of SCE's rebuttal testimony is that ORA's proposed revenues do not support a capital expenditure level that will enable it to provide safe and reliable service.

The question of whether simplified methods to basic ratemaking principles are worth pursuing can only be answered on a case-by-case basis. In this case, because of the effect on SCE's future capital expenditure levels, we will not use the CPI methodology proposed by ORA. The plant additions implicit in ORA's PTY CPI proposal are significantly less than what we, by this decision, are adopting for SCE for test year 2006. Such reductions are counter to our commitment, as discussed previously, to facilitate the replacement of SCE's transmission and distribution infrastructure in a timely and efficient manner.

We note that a CPI methodology could be constructed in other ways depending on what the CPI increase is intended to represent. It could be argued that the CPI should represent the effect on ratepayers. That is, if the CPI is projected to increase by 2% in a year, the utility rates would increase 2% for that year. That would be simple to implement. Also, the link between increases in the CPI and increases in rates should be fairly understandable to utility customers. In fact, in the customer/CPUC correspondence generated by this GRC, a number of customers complain that the proposed increases exceed the CPI changes or cost of living increases in general. However the difference between escalating rates by the CPI increase and escalating the revenue requirement by the CPI, as recommended by ORA, is large. For example, the effect of escalating rates by the CPI can be approximated by first taking the increase in revenue requirement calculated by applying the CPI increase to the last authorized revenue requirement. That would represent the increase to the existing customers at existing sales levels. On top of this, additional sales related to increased usage and new customers would be coming in under this CPI escalated rate and the utility would capture the increased revenues. Using ORA's recommended 2006 revenue requirement of $3,592,407,000113 and CPI increases of 1.90% for 2007 and 2.10% for 2008 results in increases of $68,256,000 in 2007 and $76,947,000 in 2008. In its PTYR calculations, ORA used revenue growth due to sales increases of $65,382,000 for 2007 and $85,752,000 for 2008. Adding these elements results in revenue increases of $133,638,000 in 2007 and $162,699,000 in 2008, for a total of $296,337,000. The total is larger than that requested by SCE in its application showing ($159,447,000 in 2007 and $121,522,000 in 2008 for a total of $281,969,000).114 In this case, a CPI methodology would result in providing SCE with more money than what it though it needed. This is not a recommendation by any party in this proceeding. It merely demonstrates that while attractive, simplified methodologies may not be appropriate and may not produce reasonable results.

In rejecting the CPI methodology, we must now determine appropriate levels of plant additions for 2007 and 2008. For calculating the capital related costs, SCE has included its proposed capital budget for 2007 and 2008. In its alternate recommendation, ORA bases 2007 and 2008 plant additions on a five year average of historic and estimated plant additions for the 2002 through 2006 time period, in 2003 constant dollars.

We note that SCE's budget for 2006 has been scrutinized in this rate case and certain reductions to that budget are reflected in our decision today. That, along with the fact that no other party performed a detailed review of SCE's post test year budget for the years 2007 and 2008, leads us to conclude that it would not be appropriate to base plant additions on SCE's budgets for those years. We also note that ORA's alternate recommendation for 2007 and 2008, in constant dollars, is less than its estimate for 2006 plant additions. That, along with the fact that our decision today adopts 2006 plant additions greater than that recommended by ORA, leads us to conclude that use of ORA's alternate recommendation for post test year plant additions is not consistent with our commitment to facilitate SCE's infrastructure replacement and meet its other capital needs. Rather than choosing between SCE's budget and ORA's estimate, we will use the plant addition level adopted for 2006, escalated for inflation to 2007 and 2008 levels.115 Plant additions for 2006 have been fully scrutinized in this rate case. For the post test years, it is reasonable to assume a level of plant investment similar to that for the test year. While not as ambitious as proposed by SCE, adopted levels for 2007 and 2008 will allow SCE to continue its infrastructure replacement at the adopted 2006 level, which is substantial.

In its alternate recommendation, ORA recommends that O&M inflation be based on its recommendations for labor and non labor inflation for the 2004 - 2006 period. While benefit escalation is not discussed in escalation testimony, ORA concurred with SCE's labor and non labor escalation estimates. We will adopt SCE's methodology for O&M expense escalation, including that for benefits. We will also adopt SCE's proposed annual advice letter procedure for implementing post test year rate changes and the proposed mechanism to address major exogenous changes in SCE's costs.116 Appendix D illustrates the calculation of the post test year requirements using the methodology adopted by this decision.

ORA requests the Commission extend the rate case cycle associated with SCE's test year request to four years by including the additional post test year 2009. SCE opposes the request. We will not extend the GRC cycle as requested by ORA. It is contrary to the current rate case plan that allows major energy utilities the opportunity to file GRC applications every three-years. We would certainly consider the extension if SCE were in agreement. However, since SCE opposes the extension, changes to the length of the GRC cycle should be addressed through modification of the rate case plan, where the issues regarding the proper length of the cycle can be addressed by all affected parties and where the resultant decision could be applied fairly and consistently among the parties.

106 ORA/ Bumgardner, Ex. 202, pages 16-12.

107 Aglet, Exhibit 407, page 31.

108 TURN, Opening Brief, page 174.

109 In D.85-12-076, expenses related to customer growth were assumed to be offset by productivity.

110 See D.04-05-055 and D.05-03-023.

111 In its performance based ratemaking application, A.95-06-002, SCG proposed to increase rates by price inflation less a productivity factor.

112 D.97-07-054, mimeo, page 37.

113 Exhibit 899, page 4.

114 This analysis excludes the effects of the one-time PBOP refund in 2006.

115 Adopted gross additions for 2006 amount to $1,622,147,000. The escalated amounts are $1,643,050,000 for post test year 2007 and $1,668,348,000 for post test year 2008.

116 Earlier in this decision we adopted SCE's proposed annual revenue adjustment to reflect the number of nuclear refueling outages at SONGS and rejected SCE's proposal for a similar adjustment mechanism at Four Corners.

Previous PageTop Of PageNext PageGo To First Page