Results of Operations (RO) Volume 4 - Depreciation Study

Transcription

1 Application No.: Exhibit No.: Witnesses: A SCE- Vol.0 T. Condit A. Varvis R. White A (U -E) ERRATA Results of Operations (RO) Volume - Depreciation Study Before the Public Utilities Commission of the State of California Rosemead, California June 1, 01

2 Table II- 0 Comparison of Traditional vs. Per-Unit Net Salvage Analysis Results Traditional Per Unit with SCE Traditional Results Relative Account Analysis.% Inflation Proposed to Per Unit Results 1% 1% % Higher 1% % 0% Lower % % 0% Higher % % % Higher % % 1% Lower % 01% % Lower 1% 1% % Lower % % % Higher 0% % 1% Higher These variances between the results produced by a traditional analysis versus a per-unit analysis do not demonstrate flaws in the per-unit approach; rather, they reflect the difference between past retirement experience and what one can reasonably expect about future retirements and costs. -1% For example, Account Transmission Towers shows the largest difference (1%) between the forecast net salvage of a traditional analysis (-1%) and the per-unit analysis (-1%). Although this account was not among the largest five accounts requested by the Commission to be analyzed with a per-unit analysis, SCE nonetheless performed a per-unit analysis to evaluate the -1% reasonableness of the -1% net salvage rate resulting from SCE s traditional analysis. The resulting analysis showed that the traditional analysis produced a net salvage forecast unduly biased by a relatively low volume of past retirements, which included removal of old transmission towers in hard-toreach locations. The high labor costs associated with the removal of these older towers resulted in more negative experienced net salvage ratios than what SCE expects to realize going forward. The per-unit analysis demonstrated a lower cost-per-tower, and anticipated retirements relatively younger than the sample from the traditional analysis. These differences are not a failure of the model, but instead recognize that the past retirements will not always reflect future expectations. TURN assumes the move to the per-unit analysis will result in excessive costs for customers. As demonstrated in Table II-, no such conclusion should be drawn. As shown in Figure II- below, the traditional analysis, without application of expert judgment, produces depreciation expense approximately as large as the results supported by SCE s per-unit analysis. 0 See WPSCE- Volume 0, pp. 0-0 (Net Salvage Analysis). 1

3 1 leading to reduced recovery of justifiable depreciation expense. Rather than comparing what SCE proposes to what various other commissions on average have authorized, it is more appropriate to compare SCE s experienced net salvage rate with that experienced by other utilities. Figure II- below compares SCE s experienced net salvage rates to those experienced by other electric utilities. The values are taken from retirement data, cost of removal, and salvage data for all electric utilities that filed FERC Form 1 between 1 and 01, the years in which FERC Form 1 data is available. Figure II- Three-Year Rolling Average Experienced Net Salvage Ratios FERC Form 1 Observations from 1 through 01 SCE and Industry 1.0% 0.0% Observed Net Salvage Rate.0% 0.0%.0% 0.0% Industry Median Southern California Edison.0% First through Third Quartile of Industry Observations (0% of Observations within this Band) 11 1 As shown in the figure above, in recent history SCE s net salvage is more negative than all participants in the industry filing FERC Form 1. However, the list of respondents filing FERC Form 1 s includes companies with operations that are not similar to SCE s. For example, Companies represented in this industry include generation companies, energy trading companies, transmission companies and utilities operating in largely rural areas. See Appendix C, pp. 1-, for the study methodology used to compile this comparison from recorded FERC Form 1 data. See WPSCE- Volume 0, pp. 0- (Industry Net Salvage Analysis). 1

4 1 leading to reduced recovery of justifiable depreciation expense. Rather than comparing what SCE proposes to what various other commissions on average have authorized, it is more appropriate to compare SCE s experienced net salvage rate with that experienced by other utilities. Figure II- below compares SCE s experienced net salvage rates to those experienced by other electric utilities. The values are taken from retirement data, cost of removal, and salvage data for all electric utilities that filed FERC Form 1 between 1 and 01, the years in which FERC Form 1 data is available. Figure II- Three-Year Rolling Average Experienced Net Salvage Ratios FERC Form 1 Observations from 1 through 01 SCE and Industry 1.0% 0.0% Southern California Edison Observed Net Salvage Rate.0% 0.0%.0% 0.0% Industry Median First through Third Quartile of Industry Observations (0% of Observations within this Band).0% As shown in the figure above, in recent history SCE s net salvage is more negative than all participants in the industry filing FERC Form 1. However, the list of respondents filing FERC Form 1 s includes companies with operations that are not similar to SCE s. For example, Companies represented in this industry include generation companies, energy trading companies, transmission companies and utilities operating in largely rural areas. See Appendix C, pp. 1-, for the study methodology used to compile this comparison from recorded FERC Form 1 data. See WPSCE- Volume 0, pp. 0- (Industry Net Salvage Analysis). 1

5 1 Because utilities can differ greatly (e.g., by population density, system density, and types of operations), a more relevant comparison would be between SCE and peer utilities. Of the companies filing FERC Form 1 s in this period, SCE identified 11 as a peer group of companies. These companies were identified as being similar to SCE because they are large, established companies with large transmission and distribution investment operating in large, densely-populated cities. Figure II-, below, shows the resulting comparison to this group of peers. Figure II- Distribution of Experienced (Recorded) Net Salvage Rates FERC Form 1 Observations from 1 through 01 SCE and Peer Group 1.0% 0.0% Observed Net Salvage Rate.0% 0.0%.0% Industry Median Southern California Edison First through Third Quartile of Industry Observations (0% of Observations are within this Band) 0.0%.0% As shown in the figure above, both SCE and its peer utilities have experienced a trend of increasingly negative net salvage rates. This comparison of SCE to its peers does not substantiate TURN s claim that SCE s net salvage rates are way out of line with that of the industry. For example, utilities serving rural areas are likely to have lower costs of removal than utilities serving urban areas. For further discussion, see SCE-0, Volume 0, pp. -. See Appendix C pp. 1- for the complete description of this selection process. See WPSCE- Volume 0, pp. 0- (Industry Net Salvage Analysis). 1

6 1 Because utilities can differ greatly (e.g., by population density, system density, and types of operations), a more relevant comparison would be between SCE and peer utilities. Of the companies filing FERC Form 1 s in this period, SCE identified as a peer group of companies. These companies were identified as being similar to SCE because they are large, established companies with large transmission and distribution investment operating in large, densely-populated cities. Figure II-, below, shows the resulting comparison to this group of peers. Figure II- Distribution of Experienced (Recorded) Net Salvage Rates FERC Form 1 Observations from 1 through 01 SCE and Peer Group 1.0% 0.0% Observed Net Salvage Rate.0% 0.0%.0% Industry Median Southern California Edison First through Third Quartile of Industry Observations (0% of Observations within this Band) 0.0%.0% As shown in the figure above, both SCE and its peer utilities have experienced a trend of increasingly negative net salvage rates. This comparison of SCE to its peers does not substantiate TURN s claim that SCE s net salvage rates are way out of line with that of the industry. For example, utilities serving rural areas are likely to have lower costs of removal than utilities serving urban areas. For further discussion, see SCE-0, Volume 0, pp. -. See Appendix C pp. 1- for the complete description of this selection process. See WPSCE- Volume 0, pp. 0- (Industry Net Salvage Analysis). 1

7 Table II- 1 Account Mix of Units Installed and Removed and Associated Labor Cost Allocation Units (M) Ratio Weighted Units Dollar Allocation Year Installed Removed Install Removal Total Install Removal A B C=A*% D=B*% E=C+D F=C/E G=D/E Increasing allocation to COR comes from higher proportional share of replacement projects, not increasing the allocation to removal Contrary to TURN s claims, SCE has not changed its per-project cost assignment practices, nor has SCE increased the percentages used to allocate costs to the retirement unit s cost of removal. As explained in Chapter II.C. of SCE s direct testimony (SCE-0, Volume ), SCE has consistently used a standard rates table to allocate capital between installation and removal for most of its Distribution plant. The table has been in continuous use since approximately the s and it sets forth allocation factors that have been studied but not materially modified over the years. To comply with the fourth directive from the 01 GRC decision, SCE set forth an account-specific discussion of the process for allocating costs to COR. TURN neither cites to nor discusses that testimony. TURN s comparison appears to confuse cost of removal as a percentage of total expenditures for all projects, on the one hand, with the allocation to removal cost for a single unit replacement project, on the other hand. The data TURN uses to draw its conclusions includes both install-only work (for which little to no removal costs are incurred) and work that involves the replacement of assets (for which install and removal costs are both incurred and require proper allocation/assignment). The mix of install-only versus replacement work changed over the time period due to load/customer 1 See WPSCE- Volume 0, pp (Cost Allocation). SCE-0, Volume, p., lines 1-1. D p.. See WPSCE-, Volume 0 p. 11; SCE-0, Volume, pp. 0-0.

8 Table II- 1 Account Mix of Units Installed and Removed and Associated Labor Cost Allocation Units (M) Ratio Weighted Units Dollar Allocation Year Installed Removed Install Removal Total Install Removal A B C=A*% D=B*% E=C+D F=C/E G=D/E % % % % % % % % % 1% % 1% % 0% % 1% % 1% % 0% % 1% % % Increasing allocation to COR comes from higher proportional share of replacement projects, not increasing the allocation to removal Contrary to TURN s claims, SCE has not changed its per-project cost assignment practices, nor has SCE increased the percentages used to allocate costs to the retirement unit s cost of removal. As explained in Chapter II.C. of SCE s direct testimony (SCE-0, Volume ), SCE has consistently used a standard rates table to allocate capital between installation and removal for most of its Distribution plant. The table has been in continuous use since approximately the s and it sets forth allocation factors that have been studied but not materially modified over the years. To comply with the fourth directive from the 01 GRC decision, SCE set forth an account-specific discussion of the process for allocating costs to COR. TURN neither cites to nor discusses that testimony. TURN s comparison appears to confuse cost of removal as a percentage of total expenditures for all projects, on the one hand, with the allocation to removal cost for a single unit replacement project, on the other hand. The data TURN uses to draw its conclusions includes both install-only work (for which little to no removal costs are incurred) and work that involves the replacement of assets (for which install and removal costs are both incurred and require proper allocation/assignment). The mix of install-only versus replacement work changed over the time period due to load/customer 1 See WPSCE- Volume 0, pp (Cost Allocation). SCE-0, Volume, p., lines 1-1. D p.. See WPSCE-, Volume 0 p. 11; SCE-0, Volume, pp. 0-0.

9 Figure II- 00 to 01 Weighted Average Net Salvage Per Pole Compared to TURN Calculated Growth Rate $,000 $,00 Weighted 00 to 01 Average Per Unit Cost per Pole TURN Assumed Net Salvage Per Pole $, Average Net Salvage Per Pole $,000 $,00 $,000 $1,00 $1,000 $00 SCE's Net Salvage Per Pole Basis Used in its Per Unit Analysis is $,00 $ Because SCE recognizes that there is some variability in its historical cost to remove assets, SCE took a conservative approach of averaging the net salvage cost per pole over the period 00 to 01 (in nominal dollars without adjusting for inflation during the period), resulting in a $,00 net salvage cost per-pole as the basis for estimating future net salvage. SCE s use of this $,00 is both conservative and supported by the data. % To demonstrate just how conservative SCE s per-unit net salvage of $,00 is, consider a comparison to TURN s proposed reductions to the per-unit pole replacement costs in TURN-1. SCE allocates approximately % of a pole replacement project s costs to the removal of the pole. TURN s per-unit pole replacement cost of $1, (in TURN-1) 0 would cause a net salvage per pole of approximately $,0 per pole. While SCE does not agree that TURN s estimate is reasonable (a point addressed in rebuttal testimony elsewhere), TURN s criticisms that SCE s per pole removal costs used in the per-unit analysis is not correct. See WPSCE- Volume 0, pp. 1-1 (Account Unit Costs). 0 TURN-1, p.. are unreasonable

10 The Data in Question Did Not Impact SCE s Uncontested Life Proposals TURN claims that SCE admitted that its data for was erroneous, suggesting that SCE s life analysis data was erroneous and unreliable. 11 This is an overstatement. More accurately, SCE acknowledged to TURN during discovery that it had discovered an error in how vintages were being assigned to retirements caused by a system issue related to the YK conversion at the start of the century. As SCE explained, the impact of this issue was confined to a small group of retirements with vintages around the year 000. Once this issue was identified during the discovery phase of the proceeding, SCE reviewed the data associated with the error and requested that SCE s depreciation expert, Dr. White, run a supplemental life analysis to ascertain any effects on SCE s proposals. 1 As a result of his analysis, Dr. White concluded that the YK data issue provided an insufficient basis for adjusting any of his recommendations because the impacts were minimal. In addition, SCE also found a small population of retirements with negative ages (i.e., a 0 vintage asset being retired in 00). Again, this problem was very isolated and constituted less than 0.% of retirements and did not affect the depreciation estimation. 1 The collective percentage of retirements from these issues represent less than % of all retirements in SCE s analysis. TURN s decision to seize on these minor and explainable errors to cast doubt on the entirety of SCE s request is not warranted. approximately. Retirements at Early Ages Are Not Implausible TURN also asserts that any plant amounts retired at an early age are not plausible and must result from retirement data integrity issues. 1 However, SCE has a large 11 TURN-0, p SCE provided the results to this analysis in response to a data request. See Appendix B pp. -11 for SCE s response to TURN-SCE-0, Question (a) and (a) Supplemental. 1 See Appendix B pp. 1-1 for SCE s response to TURN-SCE-0, Question (c). 1 TURN-0, p. 1 TURN states that In discovery we sent SCE a list of retirement ages in the data for the years that were not plausible. In fact, TURN asked SCE to research 11 capital projects (See Appendix B pp. 1-1 for SCE s response to TURN-SCE-0, Question (a) and provide details on the underlying reasons for the assets retirement. This request was impractical because the information needed to answer TURN s specific questions were not readily at hand but kept only in paper work order packages either maintained in distributed field locations or in storage facilities. Consequently, without any evidence TURN simply concludes that the age of these retirements must be erroneous.

11 depreciation accounting is the objective incorporated by Foster Associates in estimating service life and net salvage parameters presented in this proceeding. Q. WHAT IS YOUR UNDERSTANDING OF MR. DUNKEL S OBJECTION TO A PER-UNIT NET SALVAGE ANALYSIS? A. Citing SCE responses to discovery questions, Mr. Dunkel claims there is no known precedent anywhere in the country for the per-unit net salvage analysis. 1 He further claims to have proven the method is unapproved, untested, and not part of accepted depreciation practices. 1 Mr. Dunkel describes his attack on the per-unit net salvage model as the most significant depreciation issue in this proceeding. 1 Q. HAVE PER-UNIT NET SALVAGE ANALYSES BEEN FILED IN OTHER JURISDICTIONS? A. Yes. Discovery questions asked by Mr. Dunkel regarding the use of per-unit net salvage analyses were directed specifically to electric utilities and testimony filed during calendar years 0 through 01. Per-unit studies, however, have been filed before the New Jersey Board of Public Utilities in New Jersey Natural Gas Docket Nos. GR00 and GR111. Per-unit studies were also conducted for Union Gas Limited in 1 and 1. The 1 study was filed before the Ontario Energy Board in E.B.R.O 1 0. If Mr. Dunkel had inquired about precedents for per-unit net salvage analyses prior to 0 for any utility, he might not have claimed there is no known precedent anywhere in the country for the per-unit net salvage analysis. GR and 1 Q. ARE PER-UNIT NET SALVAGE ANALYSES ROUTINELY PRESENTED IN DEPRECIATION STUDIES? A. No. As noted in my direct testimony (and stated in other studies conducted by Foster Associates), a per-unit net salvage analysis of installation and retirement costs is the most desirable treatment of net salvage. Unlike other methods of analysis, the per-unit model is derived from the principle that the cost per unit to retire an asset is independent of the age of the asset when it is retired from service. 1 TURN-0 (Dunkel Direct Testimony), p., lines. 1 TURN-0 (Dunkel Direct Testimony), p., lines TURN-0 (Dunkel Direct Testimony), p., line. 0

12 Appendix C SCE Industry Study & Methodology

13 ) American Water Works Company ) Centerpoint Energy ) Consolidated Edison ) DTE Energy Company Holdings ) Dominion Energy, Inc. ) Duke Energy Holding Company ) Edison International ) El Paso Electric Company 11) Entergy 1) Eversource Energy 1) Exelon Corporation 1) FirstEnergy 1) NextEra Energy, Inco. 1) Pacific Gas and Electric 1) Public Service Enterprise Group 1) Southern Company 1) The AES Corporation 0) Xcel Energy SCE added San Diego Gas and Electric to this list because of their proximity to SCE s service territory, size, California regulatory environment, and similarity in operations. Peer Group: Step For the purpose of this analysis, SCE used Page CONTROL OVER RESPONDENT of FF1 to determine which FF1 respondents are subsidiaries of the listed utilities in Step 1 to include in constructing a FF1 peer group against which to compare SCE s experienced NSR. Doing so yielded the following list of respondents: 1) American Transmission Systems, Incorporated ) Baltimore Gas and Electric Company ) Boston Edison Company ) Central Maine Power Company ) Cleveland Electric Illuminating Company, The ) Commonwealth Edison Company ) Connecticut Light and Power Company, The ) Consolidated Edison Company of New York, Inc. ) Jersey Central Power & Light Company ) Metropolitan Edison Company 11) Monongahela Power Company 1) New York State Electric & Gas Corporation 1) North Atlantic Energy Corporation 1) NSTAR Electric Company 1) Ohio Edison Company 1) Pacific Gas and Electric Company a list of respondents Whether a FF1 respondent is a subsidiary of the utilities list in Step 1 was determined based on information available through FF1 Page s reported control relationship at the end of 01.

14 1) Path Allegheny Transmission Company, LLC 1) PECO Energy Company 1) Pennsylvania Electric Company 0) Pennsylvania Power Company 1) Potomac Electric Power Company ) Public Service Company of New Hampshire ) Public Service Electric and Gas Company ) Rochester Gas and Electric Corporation ) San Diego Gas & Electric Company ) The Allegheny Generating Company ) The Potomac Edison Company ) The United Illuminating Company ) Toledo Edison Company, The 0) Trans-Allegheny Interstate Line Company 1) West Penn Power Company ) Western Massachusetts Electric Company Peer Group: Step The list of FF1 respondents in Step was further refined by considering the respondents investment (between Transmission, Distribution, General, and Generation plant). SCE examined this investment mix for each respondent in Step for similarity to SCE. A respondent was deemed similar to SCE, and included in this study as a peer, if they have a significant Transmission and Distribution investment (at least 0% of the total investment in the company). For reference, approximately 0% of SCE s investment was in Transmission and Distribution infrastructure. The remaining 0% was split between generation plant (1%) and general plant (%). Balances for each functional asset type used in calculating the asset mix percentages were taken directly from the following FF1 Pages 0-0 summary line items : Consideration of the percent mix of functional asset types yielded the following respondents: Reporting in the FERC Form 1 changed over the year time period examined. As a result, similar sounding line items appear. Where line items are similar in nature they were grouped together. Because these were reported in different periods of time amounts were not double counted.

15 Peer Group: Step Respondents listed in Step were then evaluated based on the average population density of the counties served, the population density of each respondent s landmark county served, and the number of customers served. Population data and the land square mileage for counties were obtained from the 0 US Census. The counties and number of customers served by each respondent were obtained from the Energy Information Administration s Form 1. Using these criteria, a list of peer group companies comprised of the following FF1 respondents was assembled: 1) Baltimore Gas and Electric Company ) Boston Edison Company ) Commonwealth Edison Company ) Consolidated Edison Company of New York, Inc. ) Jersey Central Power & Light Company ) NSTAR Electric Company ) Pacific Gas and Electric Company ) PECO Energy Company Population, Housing Units, Area, and Density: 0 - United States -- County by State; and for Puerto Rico, Census 0 Summary File 1 (SF 1) 0-Percent Data SCE used the following files from the 01 EIA Form 1 data set: Service_Territory_01.xlsx and Sales_Ult_Cust_01.xlsx This respondent changed its name to NSTAR Electric Company in 00 (see Page.1 of its 00 FF1).

16 Percent of Asset Mix Respondent Distribution General PlantGeneration Transmission Total AEP Texas Central Company 0 0 AEP Texas North Company 1 0 Ameren Illinois Company Baltimore Gas and Electric Company Boston Edison Company CenterPoint Energy Houston Electric, LLC 0 0 Commonwealth Edison Company Connecticut Light and Power Company, The 0 0 Consolidated Edison Company of New York, Inc Delmarva Power & Light Company Duke Energy Ohio, Inc. 0 0 Entergy Mississippi, Inc. 0 Entergy New Orleans, Inc Entergy New Orleans, Inc. (Debtor-in-possession) 11 0 Entergy Texas, Inc. 0 Jersey Central Power & Light Company Kingsport Power Company Metropolitan Edison Company Northern States Power Company (Wisconsin) 0 NSTAR Electric Company 0 0 Ohio Edison Company 0 0 Ohio Power Company 0 0 PACIFIC GAS AND ELECTRIC COMPANY PECO Energy Company Pennsylvania Electric Company Pennsylvania Power Company 0 0 Potomac Electric Power Company Public Service Company of New Hampshire 0 0 Public Service Company of Oklahoma 1 0 Public Service Electric and Gas Company 0 San Diego Gas & Electric Company 0 0 Southern California Edison Company 1 0 THE POTOMAC EDISON COMPANY 0 0 Toledo Edison Company, The 0 0 WEST PENN POWER COMPANY Western Massachusetts Electric Company 0 Wheeling Power Company 0 0 Peer Group: Step Respondents listed in Step were then evaluated based on the average population density of the counties served, the population density of each respondent s landmark county served, and the number of customers served. Population data and the land square mileage for counties were obtained from the 0 US Census. The counties and number of customers served by each respondent were obtained from the Energy Information Administration s Form 1. Using these criteria, a list of peer group companies comprised of the following FF1 respondents was assembled: 1) Baltimore Gas and Electric Company ) Boston Edison Company ) Commonwealth Edison Company ) Consolidated Edison Company of New York, Inc. ) Jersey Central Power & Light Company ) NSTAR Electric Company ) Pacific Gas and Electric Company ) PECO Energy Company Population, Housing Units, Area, and Density: 0 - United States -- County by State; and for Puerto Rico, Census 0 Summary File 1 (SF 1) 0-Percent Data SCE used the following files from the 01 EIA Form 1 data set: Service_Territory_01.xlsx and Sales_Ult_Cust_01.xlsx This respondent changed its name to NSTAR Electric Company in 00 (see Page.1 of its 00 FF1). -A

17 ) Pennsylvania Power Company ) Potomac Electric Power Company 11) Public Service Electric and Gas Company 1) San Diego Gas & Electric Company Peer group statistics such as the average or median experienced NSR are calculated using this list of FF1 respondent observations for the 1-01 period.