STATE OF ALASKA BEFORE THE REGULATORY COMMISSION OF ALASKA. Paul F. Lisankie T.W. Patch Norman Rokeberg Janis W. Wilson

Transcription

1 STATE OF ALASKA BEFORE THE REGULATORY COMMISSION OF ALASKA Before Commissioners: In the Matter of the Consideration of the Revenue Requirement Designated as TA - Filed by ENSTAR NATURAL GAS COMPANY, A DIVISION OF SEMCO ENERGY, INC. ) ) ) ) ) Robert M. Pickett, Chairman Paul F. Lisankie T.W. Patch Norman Rokeberg Janis W. Wilson Docket No. U-1- PREFILED DIRECT TESTIMONY OF BRUCE H. FAIRCHILD 1 1 Q. Please state your name and business address. A. Bruce H. Fairchild, 0 Red River, Austin, Texas 1. Q. By whom are you employed and in what position? A. I am a principal in Financial Concepts and Applications, Inc. ( FINCAP ), a firm engaged in financial, economic, and policy consulting to business and government. Q. Describe your educational background, professional qualifications, and prior experience. A. I hold a BBA degree from Southern Methodist University and MBA and PhD degrees from the University of Texas at Austin. I am also a Certified Public Accountant. My previous employment includes working in the Controller's Department at Sears, Roebuck and Company and serving as Assistant Director of Economic Research at the Public Utility Commission of Texas ( PUCT ). I have also been on the business school Docket No. U-1- : September, 01 Page 1 of 1 { ;1}

2 faculties at the University of Colorado at Boulder and the University of Texas at Austin, where I taught undergraduate and graduate courses in finance and accounting. Q. Briefly describe your experience in utility-related matters. A. While at the PUCT, I assisted in managing a division comprised of approximately twenty-five professionals responsible for financial analysis, cost allocation and rate design, economic and financial research, and data processing systems. I testified on behalf of the PUCT staff in numerous cases involving most major investor-owned and cooperative electric, telephone, and water/sewer utilities in the state regarding a variety of financial, accounting, and economic issues. Since forming FINCAP in 1, I have participated in a wide range of analytical assignments involving utility-related matters on behalf of utilities, industrial consumers, municipalities, and regulatory commissions. I have also prepared and presented expert witness testimony before a number of regulatory authorities addressing revenue requirements, cost allocation, and rate design issues for gas, electric, telephone, and water/sewer service. I have been a frequent speaker at regulatory conferences and seminars, and have published research concerning various regulatory issues. A resume that contains the details of my experience and qualifications is attached as Appendix A, with Appendix B listing my prior testimony before regulatory agencies since leaving the PUCT. Q. What is the purpose of your testimony? A. My purpose in providing this testimony on behalf of ENSTAR Natural Gas Company and Alaska Pipeline Company (collectively, ENSTAR) is three-fold. First, it is to present certain of the schedules comprising ENSTAR's (a) filing 1 (Attachment B) 1 A (a) filing refers to a filing in compliance with the Commission s regulations at AAC.(a), which specifies the information necessary to support a new revenue requirement filing. Docket No. U-1- : September, 01 Page of 1 { ;1}

3 and, along with Daniel M. Dieckgraeff, Steven W. Warsinske, Ronald H. Brown, and Robert Hevert, sponsor various adjustments to test year data for the year ended December 1, 01 to make that data suitable for ratemaking purposes. Second, it is to present ENSTAR's (h) filing, which consists of a cost-of-service ( COS ) study that allocates ENSTAR s requested revenue requirement among its various customer classes. Third, I develop specific rates for each of ENSTAR s proposed classes that reflect the costs of providing service, as determined in the COS study. Q. Please summarize the basis of your knowledge and conclusions concerning the issues to which you are testifying in this case. A. In preparing my analysis and testimony in this case, I utilized a variety of sources of information that would normally be relied upon by a person in my role. I am generally knowledgeable about the natural gas industry from my prior work with many of the major intrastate gas distribution and transmission companies in the Southwest and elsewhere. I have worked with ENSTAR for almost thirty years on various rate and regulatory matters, a listing of which is included in Appendix B. I. AAC.(a) Filing Q. What was your role in the preparation of ENSTAR's (a) filings? A. Under the direction of Mr. Dieckgraeff, and with his assistance and that of his staff, I was assigned the primary responsibility of preparing the normalized test year and revenue requirement schedules contained in ENSTAR's (a) filing, which is based on a test year ended December 1, 01. AAC.(h) states: In addition to any other supporting studies required by this chapter, if a proposed tariff revision includes a rate redesign, other than an across-the-board increase, a cost-of-service study and a narrative explaining the methodology used in the study must be submitted Docket No. U-1- : September, 01 Page of 1 { ;1}

4 A. Summary Q. Please briefly describe the normalized test year and revenue requirement schedules, which are pages 1- of ENSTAR s (a) filing (Attachment B). A. Page 1 summarizes the filing and shows the actual results of ENSTAR s operations for the year ended December 1, 01 (column ), followed by the various adjustments (column ) made to convert the historical test year to a normalized test period representative of ENSTAR's ongoing operations (column ) for the rate effective period, which in this case is 01 and beyond. Columns (), (), and () briefly describe the adjustment, specify the related supporting schedule, and identify the primary sponsoring witness(es), respectively. Page of the filing develops ENSTAR s rate base, beginning with 1-month average test year balances (column ), followed by various adjustments (column ) made to reflect the ongoing level of ENSTAR s investment in assets used to provide service to customers (column ). Page of the filing displays ENSTAR s proposed capital structure, cost of debt, and requested return on equity. Page of the filing reconfigures the information on pages 1,, and into a revenue requirement format. Pages through 1 of the filing provide supporting schedules that detail the various adjustments made on pages 1 through. Pages through 1 provide the required historical comparative information. Q. Please turn to pages 1 and of ENSTAR s (a) filing and describe the information shown on these pages. A. The last two lines on page 1 show the rate of return on rate base and rate of return on Docket No. U-1- : September, 01 Page of 1 { ;1}

5 equity, respectively, produced by ENSTAR s existing rates on a normalized basis. Those lines show that, using the rate base from page and the capital structure and cost of debt from page, ENSTAR s current rates are producing a.% overall rate of return and a.% return on equity. Meanwhile, the last two lines of column on page show that, to achieve its requested overall rate of return of.%, ENSTAR s existing rates need to be increased by $1,0,, or.0%. B. Overview of Adjustments Q. What types of adjustments were made to ENSTAR s historical results of operations for the year ended December 1, 01? A. A historical test year serves only as the starting point for evaluating the adequacy of base rates. Various adjustments are required to convert actual test year revenues, expenses, and rate base to a representative test period data reflective of normalized, ongoing operations. Adjustments to the historical test year are generally of three types. The first are normalization adjustments designed to eliminate unusual or non-recurring items during the test year. The second are adjustments to reflect known and measurable, or pro forma, changes in test year revenues, expenses, and rate base (or investment). The third are regulatory adjustments intended to account for items in a manner consistent with currently accepted ratemaking principles and objectives. C. Revenue Adjustments Q. What was the first adjustment made on page 1 of ENSTAR s (a) filing package labeled 01 GCA Revenues? A. This adjustment was to remove from 01 revenues amounts recovered through Docket No. U-1- : September, 01 Page of 1 { ;1}

6 ENSTAR s Gas Cost Adjustment ( GCA ) mechanism. These amounts will be replaced subsequently with an amount reflecting normalized customer usage and the 01 GCA, as described more fully in Mr. Dieckgraeff s testimony. Q. Please describe the adjustments on page 1 labeled Normalized Use per Customer. A. As explained above, adjustments are made to actual test year results so that test period data is representative and reflective of normalized, ongoing operations when rates are in effect. Accordingly, 01 revenues for General Service customers were adjusted for the normalized use per customer developed by Mr. Brown and further discussed by Mr. Dieckgraeff. The details of these adjustments, which affect the gas volumes and revenues for the G-1, G-, G-, G-, and T- classes, are shown on Schedule B of the filing. Q. What is the purpose of the adjustment to 01 test year revenues on page 1 of ENSTAR s (a) filing labeled 01 Cost of Gas? A. As shown in Schedule C, the normalized use of each of the four general service sales classes from Schedule B was multiplied times ENSTAR s 01 cost of purchased gas of $.00 per Mcf to calculate a representative level of GCA revenues for each sales class. Q. Please explain the three adjustments labeled SPP Normalization. A. As discussed in Mr. Dieckgraeff s testimony, in January 01 ENSTAR began providing transportation service to the Southcentral Power Project ( SPP ), which is T- is not a separate rate class, but an internal ENSTAR designation for G- customers who purchase gas from a third party supplier and only receive transportation service from ENSTAR. The revenues from these customers are shown as Commercial Transport on page 1 of Attachment B. Docket No. U-1- : September, 01 Page of 1 { ;1}

7 jointly owned by Chugach Electric Association ( CEA ) and Anchorage Municipal Light and Power ( ML&P ). To remove the impact of any abnormalities or distortions associated with the plant s start-up, the volumes and revenues during the first six months of operations were replaced with SPP s volumes and revenues for January-June 01. As further explained by Mr. Dieckgraeff, because the operations of SPP also affect volumes transported to CEA s International plant and ML&P s generating facilities, similar adjustments were made to normalize the volumes and revenues for these facilities as well. The details of these volume and revenue adjustments are shown on Schedule D. Q. What is the adjustment to ML&P for Off-System Sales? A. As explained by Mr. Dieckgraeff, ML&P is not expected to continue making off-system sales to other utilities. The adjustment detailed in Schedule E removes the revenues and volumes associated with ML&P s off-system sales during the test year. Q. What are the next three adjustments on page 1 of ENSTAR s (a) filing? A. As explained by Mr. Dieckgraeff, ENSTAR s arrangement for transportation associated with CEA s Beluga power plant changed during 01. Because ENSTAR has not received any revenues under the new agreement with CEA, the Beluga volumes and revenues received during 01 were removed. Similarly, Mr. Dieckgraeff explains that the Bernice Lake power plant is no longer being operated by CEA, but is a back-up facility on the Homer Electric Association ( HEA ) system. Accordingly, the revenues and volumes during the test year when CEA operated the plant were removed and replaced with the revenues expected under the contract for service with HEA. This latter adjustment is shown on Schedule F. Finally, a non-recurring item in the amount Docket No. U-1- : September, 01 Page of 1 { ;1}

8 of $,00 was also removed from test year revenues. Q. What are the adjustments related to HEA s Soldotna, and Matanuska Electric Association s ( MEA ) Eklutna Generation Stations? A. Neither HEA s Soldotna nor MEA s Eklutna generation facilities were in service in the test year. However, as explained by Mr. Dieckgraeff, ENSTAR expects to be providing transportation service to both facilities during the period when the rates being set in this proceeding are in effect. Because the volumes associated with each facility are unknown, only revenues from customer and demand charges were included in the HEA- Soldotna and MEA-Eklutna adjustments (Schedules G and H, respectively), and they are based on the current customer and demand charges for service under ENSTAR s Very Large Firm Transport ( VLFT ) tariff. Q. What adjustments were made to Miscellaneous Revenues? A. Four adjustments were made to test year Miscellaneous Revenues and are detailed in Schedule I. The first adjustment reclassifies the fees that ENSTAR charges customers for insufficient fund checks from below-the-line to above-the-line revenues. The second adjustment removes the surcharge revenues from customers located in Homer because these revenues are accounted-for elsewhere in their rates. The third and fourth adjustments are related to late fees and are described by Mr. Dieckgraeff. Q. After making the above adjustments, what were ENSTAR s normalized operating revenues? A. As shown in the middle of page 1 of the (a) filing, the various adjustments described above produce ENSTAR s normalized revenues, including sales revenues, transport revenues, and miscellaneous revenues, totaling $,,. Docket No. U-1- : September, 01 Page of 1 { ;1}

9 D. Operating Expenses Q. Please refer to the Operating Expenses portion of page 1 of ENSTAR s (a) filing. What were the first two adjustments made to test year operating expenses? A. Consistent with removing 01 GCA revenues and replacing them with the 01 GCA applied to normalized gas sales, the first two adjustments to operating expenses were to remove the 01 cost of purchased gas (Schedule A) and replace it with a normalized 01 cost of purchased gas (Schedule C). Q. What were the next adjustments to operating expenses? A. The next adjustments to operating expenses were: (1) a payroll adjustment to account for changes in the number and composition of personnel at ENSTAR during the test year and reflect current salary, wage, and payroll tax levels (labeled Payroll ); and () an adjustment for higher electric utility costs (labeled Utilities ). Mr. Dieckgraeff sponsors these adjustments, with additional detail being shown on Schedule J. Q. What is the Software Maintenance adjustment? A. As noted by Mr. Dieckgraeff, ENSTAR and Cook Inlet Natural Gas Storage Alaska, LLC ( CINGSA ) use a volume accounting and nominations management software package in their respective operations. In November 01, the vendor will begin charging an annual fee to maintain this software, which is $1, for the first year. Accordingly, transmission operation and maintenance ( O&M ) expense were increased $,0 to reflect ENSTAR s 0% share of this ongoing software maintenance expense (Schedule K). This adjustment is consistent with the treatment of this expense item in CINGSA s recent true-up filing (TA1-). Separately, the maintenance fee for ENSTAR s customer service and field order software systems has Docket No. U-1- : September, 01 Page of 1 { ;1}

10 increased by $,1 per year, which is also shown on Schedule K and reflected as an increase to administrative and general ( A&G ) expenses on page 1 of the (a) filing. Q. What is the adjustment for Credit Card Program? A. As explained by Mr. Dieckgraeff, ENSTAR is proposing to allow its retail customers to use credit cards to pay their bills. While ENSTAR will have to pay processing fees for this service, its uncollectible accounts expense will be lower because bad debts are borne by the credit card issuers. The net cost to ENSTAR of the proposed credit card program is shown on Schedule L. Q. Please explain the adjustment for Uncollectible Accounts. A. The purpose of this adjustment is to match uncollectible accounts expense with normalized retail revenues (i.e., taking into account the various adjustments made to revenues discussed earlier). The calculation of this adjustment to test year uncollectible accounts expense is shown on the upper portion of Schedule M (the lower portion will be discussed later). Q. How did you calculate the adjustment labeled Insurance? A. The insurance expense recorded on ENSTAR s books during the test year reflects both the amortization of previous premiums paid and premium refunds attributable to prior years. To reflect current and ongoing insurance expense levels, the most recent premium paid by ENSTAR on its various insurance policies was annualized and compared with test year amounts (Schedule N). The sum of the differences between annualized current premiums and test year amounts, after adjusting for portions capitalized or reimbursed, served as the adjustment to test year insurance expense. Q. What is the Regulatory Asset Amortization adjustment? Docket No. U-1- : September, 01 Page of 1 { ;1}

11 A. Included in Account Regulatory Commission Expenses for the test year is the amortization of various regulatory assets; namely, rate case expenses from Dockets U- 0- and U-0-0, capitalized costs associated with the gas sales agreement with Marathon, and costs incurred in connection with a proposed gas transmission line ( Bullet Line ). The capitalized expenses from ENSTAR s last rate case were fully amortized in 01 and the Marathon Sale Agreement expenses will be fully amortized in 01; therefore, both were removed from test year expenses. Because the settlement agreement in Dockets U-0- and U-0-0 excluded Bullet Line amortization expense from future rates, an adjustment was also made to remove this amount from test year expenses. The details of these adjustments are shown in the upper portion of Schedule O. Q. Please describe the adjustment labeled Rate Case Expenses. A. As discussed by Mr. Dieckgraeff, it is estimated that if the present case is fully litigated, ENSTAR will incur $00,000 in rate case expenses. Amortizing this amount over five years, and removing approximately % of the costs that are capitalized, results in rate case expenses of $,1 per year. Q. In the Depreciation and Amortization category, what is the adjustment labeled 01 YE Software Amortization? A. At the end of 01, approximately $1 million in software costs were capitalized on ENSTAR s books. The purpose of the adjustment labeled 01 YE Software Amortization is to reflect the amortization expense associated with this new software. The calculation of this additional amortization expense is shown on Schedule P. Q. Please describe the Additional Plant adjustment. Docket No. U-1- : September, 01 Page of 1 { ;1}

12 A. This adjustment includes the depreciation expense associated with the construction completed but not classified at test year-end, complete and in-progress non-revenue producing plant being placed in service in 01, and the planned CINGSA/KPL bypass pipeline, as discussed in Mr. Dieckgraeff s testimony. The details of this additional depreciation expense are contained in Schedules T, U, and V, respectively. E. State & Federal Taxes Q. Please describe the adjustment labeled 01 Property Taxes to Ad Valorem & Miscellaneous in the State & Federal Taxes category on page 1 of ENSTAR s (a) package. A. Because property taxes for 01 are known amounts, this adjustment simply increases test year amounts to 01 levels. The details of the adjustment are contained in Schedule Q. Q. What is the Additional Plant adjustment to Ad Valorem & Miscellaneous taxes? A. Similar to the adjustment to depreciation expense, the purpose of this adjustment is to increase property taxes associated with the construction completed but not classified at test year-end, complete and in-progress non-revenue producing plant being placed in service in 01, and the planned CINGSA/KPL bypass pipeline. These adjustments are also developed on Schedules T, U, and V. Q. What is the purpose of the two adjustments to State Income Tax and Federal Income Tax? A. In the left-hand column of Schedule R (captioned Current Rates ), state and federal income taxes are calculated to correspond to the income generated by the normalized revenues, expenses, rate base, capital structure, and cost of debt developed elsewhere in Docket No. U-1- : September, 01 Page 1 of 1 { ;1}

13 this filing for the 01 test year. These normalized incomes taxes, calculated by applying state and federal statutory tax rates of. and percent, respectively, were then compared with 01 test year income taxes to arrive at the respective income tax adjustments shown on page 1 of ENSTAR s (a) filing. Q. What is ENSTAR s normalized Net Utility Operating Income? A. After making the adjustments to revenues and operating expenses described above, ENSTAR s normalized Net Utility Operating Income is $1,,, as shown near the bottom of the right-hand column on page 1. F. Rate Base Q. Please turn to page of ENSTAR s (a) filing, which develops ENSTAR s rate base, or net investment in assets used to serve customers. What served as the starting point for calculating ENSTAR s rate base? A. Schedule S displays monthly balances during the 01 test year for the components comprising ENSTAR s plant in service, working capital items, regulatory assets, and non-investor supplied capital. The 1-month average balances of these items served as the starting point for determining ENSTAR s rate base and are shown in the second column of page of ENSTAR s (a) package. Q. What is the adjustment labeled Remove Acquisition Adjustment? A. This adjustment explicitly removes from rate base the acquisition adjustment recorded on ENSTAR's books to demonstrate that none of the amounts paid by AltaGas Ltd., SEMCO, or Continental Energy Systems above the original cost of property, plant, and equipment to acquire ENSTAR are included in rates. Q. What is the adjustment to Remove CWIP? Docket No. U-1- : September, 01 Page 1 of 1 { ;1}

14 A. Construction work in progress ( CWIP ) is removed from rate base because the majority of the plant involved was placed in service during the test year will be revenue producing or is accounted for elsewhere. Q. Please explain the Additional Plant adjustments to rate base. A. As previously discussed, these adjustments include in rate base ENSTAR s investment in the construction completed but not classified at test year-end, complete and inprogress non-revenue producing plant placed in service and software costs being capitalized in 01, and the planned CINGSA/KPL bypass pipeline. Mr. Dieckgraeff further describes these investments and the associated adjustments. Additional details are shown on Schedules T, U, and V of the filing. Q. What is the next item of rate base labeled Gas Stored Underground? A. To meet the demands of firm customers during peak periods, ENSTAR must maintain an inventory of stored gas from which it can draw. Gas stored underground reflects ENSTAR s average investment in working gas stored in the CINGSA storage facility during the test year. Although the amount of gas stored underground varies throughout the year (i.e., inventory is usually built up in off-peak months and then drawn down during peak periods), it represents an ongoing and substantial investment in assets used and useful in providing service to customers. Accordingly, just as with plant and other working capital items (i.e., materials and supplies and prepayments), the normal ratemaking practice is to include a representative, average level of gas stored underground in rate base, which allows the utility to earn a return to cover the capital carrying costs associated with this significant asset investment. Q. What is the adjustment labeled Cash Working Capital Allowance? Docket No. U-1- : September, 01 Page 1 of 1 { ;1}

15 A. In Schedule W, a cash working capital allowance for inclusion in rate base is calculated using the conventional one-eighth (or days) formula applied to normalized O&M expenses and taxes other than income. This calculation follows the methodology used in ENSTAR s previous rate cases. Q. What are the adjustments to Rate Case Expenses - 00, Gas Costs - Marathon, and Bullet Pipeline under the Regulatory Assets category of rate base? A. These adjustments remove ENSTAR s investment in these regulatory assets from rate base and are the counterparts to the adjustments to Account -- Regulatory Commission Expense described earlier and detailed in Schedule O. Q. Please describe the adjustment for Rate Case Expenses. A. This adjustment includes the rate case expenses estimated for the present proceeding in ENSTAR s rate base, as discussed by Mr. Dieckgraeff. Q. What is the purpose of the adjustment labeled DIT Associated with Acquisition Adjustment? A. Consistent with excluding the acquisition adjustment from rate base, this adjustment excludes the deferred income taxes ( DIT ) recorded on ENSTAR's books that are associated with the acquisition adjustment (Schedule X). Mr. Warsinske describes and develops the DIT related to the acquisition adjustment. Q. Finally, please describe the adjustment for Deferred ITC. A. This adjustment relates to the treatment of investment tax credits ( ITCs ) allowed under previous income tax law. Specifically, the accumulated balance of ITCs is amortized over the life of property, plant, and equipment, with the average unamortized balance during the test year serving as a reduction to rate base. The details of this Docket No. U-1- : September, 01 Page 1 of 1 { ;1}

16 adjustment are shown on Schedule Y. Q. What is ENSTAR s normalized rate base? A. As shown at the bottom of page of ENSTAR s (a) filing, after making the various adjustments described above, ENSTAR s rate base totals $,1,. Q. What then is ENSTAR s normalized rate of return on rate base? A. As shown on the next-to-the-bottom line on page 1, dividing ENSTAR s normalized Net Utility Operating Income of $1,, by a rate base of $,1, indicates that its current rates are only producing a rate of return on rate base of.%. G. Capital Structure Q. Please describe page of ENSTAR s (a) filing. A. Page calculates ENSTAR s requested rate of return, which combines its capital structure ratios with its cost of debt and requested rate of return on common equity. Q. What is ENSTAR s capital structure? A. At test year-end, ENSTAR s capital structure was comprised of $. million of debt and $1. million of common equity, which equates to capital structure ratios of.% debt and.% equity. Q. What is ENSTAR s cost of debt? A. ENSTAR s cost of debt is developed in footnote (d) at the bottom of page of its (a) filing. As shown there, after removing a debt issue that matured in 01 and including the costs associated with issuing and refinancing its debt, ENSTAR s average embedded cost of debt is.0%. Q. What rate of return on equity is ENSTAR earning? A. Subtracting $,,0 in interest expense (shown on page ) from ENSTAR s Docket No. U-1- : September, 01 Page 1 of 1 { ;1}

17 normalized Net Utility Operating Income (shown on page 1) results in Net Income Available for Common of $,,0. Dividing this by common equity of $,, (also from page ) indicates that ENSTAR s current rates are only producing a rate of return on common equity ( ROE ) of.%, which is well below ENSTAR s currently authorized ROE of 1.%. H. Revenue Requirement Q. Please describe page of ENSTAR s (a) filing. A. As indicated earlier, page reconfigures pages 1- into a revenue requirement format. In other words, rather than looking at what rate of return current rates are producing, as on page 1, page calculates the total revenues ENSTAR needs for it to earn its requested rate of return of.%, including a rate of return on equity of 1.%, as recommended by Mr. Hevert. Q. Are there any differences in the numbers on page versus pages 1-? A. There are several. Rate base remains the same at $,1,, but there is one change in O&M expenses, state and federal income taxes change, and return on investment reflects ENSTAR s requested rate of return instead of its implied earned rate of return. Q. What O&M expense changes on page? A. Because ENSTAR s revenue requirement is greater than its normalized revenues, uncollectible accounts expense is correspondingly higher. The uncollectible accounts expense associated with ENSTAR s requested retail revenue requirement is developed in the lower portion of Schedule M. Q. What state and federal income taxes are associated with ENSTAR s revenue requirement? Docket No. U-1- : September, 01 Page 1 of 1 { ;1}

18 1 1 A. In the right-hand column of Schedule R (captioned Requested Rates ), state and federal income taxes corresponding to ENSTAR s requested return on investment are calculated. These normalized incomes taxes, also calculated by applying state and federal statutory tax rates of. and percent, respectively, were compared with 01 test year income taxes to arrive at the respective income tax adjustments shown on page. Q. Based on these changes in O&M expense and income taxes and ENSTAR s requested rate of return, what is ENSTAR s total revenue requirement? A. As shown at the bottom of the right-hand column of page, summing operating expenses, state and federal taxes, and return on investment produces total a revenue requirement of $1,,1. Comparing this with adjusted normalized revenues of $,, demonstrates that ENSTAR s current rates are deficient by $1,0,, and need to be increased by.0%. II. AAC.(h) Filing Q. Please describe the purpose of this section of your testimony. A. As indicated earlier, the purpose of this section is to present a COS study that allocates ENSTAR s requested revenue requirement among its customer classes. Using the results of the COS study, specific rates for ENSTAR s customer classes are developed. A. Overview Q. Would you briefly describe a COS study? A. A COS study is an engineering, accounting, and economic analysis designed to allocate a utility s total cost of providing service to specific customers or customer classes. Many of a utility s operating expenses and much of its capital investment are incurred Docket No. U-1- : September, 01 Page 1 of 1 { ;1}

19 to serve all customers, to a greater or lesser extent. Because these joint and common costs cannot be directly tied to specific customers, they must be apportioned among customers and/or customer classes. This apportionment is accomplished through a COS study, in which operating and capital costs are allocated using factors developed from various operating data reflecting cost causation. The sum of the costs allocated to each customer and/or customer class in the COS study represents class revenue requirements, or that portion of the utility s total costs for which a particular customer or customer class is responsible. Q. What guided ENSTAR's present (h) filing? A. Because ENSTAR s current rates were arrived at through a negotiated settlement, the present COS studies follows the methodology adopted in ENSTAR s last rate design case, Docket U-00-, which was generally patterned after that ordered in its previous rate design case, Docket U--. Q. Please describe how the COS study was conducted. A. The first step in conducting the COS study is to assign the components of ENSTAR s requested revenue requirement and rate base to Federal Energy Regulatory Commission ( FERC ) accounts. The next step is to examine each account to judge its nature and cost-causal characteristics. This initially involves classifying a particular account as primarily customer-related (i.e., varying with the number of customers), capacity-related (i.e., varying with the need to meet peak demands), or commodity-related (i.e., varying with the amount of gas delivered). The manner in which each operating and capital account is classified is identified in the COS studies under the heading Cost Classification. Docket No. U-1- : September, 01 Page 1 of 1 { ;1}

20 The next step is to develop various allocation factors to reflect the responsibility of each customer class for different costs based on cost-causal relationships. These allocation factors are derived from usage, operating, and other data for the normalized test year, and are intended to capture the relative contribution of each customer class to the system as a whole. The allocation factor used to apportion each operating and capital account between customer classes is identified in the COS studies under the heading Alloc. Factor. The details underlying the derivation of each allocation factor are contained in separate pages of each COS study. Finally, the amounts in each operating and capital account are allocated among customer classes using the indicated allocation factor. The amounts allocated to each customer class are then summed to calculate class revenue requirements. Q. Are there any features of the COS studies that are particularly noteworthy? A. Yes, there are three. First, most of ENSTAR s capacity-related costs, which are primarily related to its transmission activities, are apportioned between customer classes using an allocation factor calculated by weighting equally the relative contributions of each customer class to the test year coincident system peak demand and the average day demand, which is the equivalent of volumes. This is the same allocation used in Dockets U-00- and U--. The rationale for using an allocation factor reflecting both coincident peak demand and average day demand is that ENSTAR s transmission facilities were intended not just to meet customers demands on peak days, but also to improve access to gas supplies. Second, in Docket U-- approximately $. million of investment in Account (Distribution Mains) was ascribed to power customers (e.g., ML&P). This same Docket No. U-1- : September, 01 Page 0 of 1 { ;1}

21 amount was carried forward and assigned to power and large volume transport customers in the present COS studies, although it is now fully depreciated. Under the premise that depreciation expense is intended to recover the original investment in plant, large transmission customers no longer have responsibility for any depreciation expense on Accounts and because the amount assigned them has been fully recovered. Accordingly, an allocation factor was developed that does not allocate any Accounts and depreciation expense to large transmission customers. Third, all general service customers, regardless of whether they are a sales customer or served via transport, are included in their corresponding class for cost allocation purposes. The reason for this is to maintain a uniform margin (i.e., rate net of gas costs) and avoid creating artificial incentives for customers to switch from one type of service to another (i.e., sales to transport, or vice versa). Q. What revenue requirement and rate base are allocated among ENSTAR's customer classes in the COS study? A. The basis for the COS study is ENSTAR s requested revenue requirement and rate base is summarized on page of the (a) filing. However, these amounts are adjusted for two items. First, the 01 cost of gas that will be recovered through the GCA mechanism for sales customers is excluded, so the COS study only allocates amounts that are incorporated into base rates. Second, some of ENSTAR s current and expected customers are not well suited to inclusion in a cost analysis based on an historical test year. For example, ENSTAR transports gas to CEA s International power plant, a standby, or back-up, generator that Docket No. U-1- : September, 01 Page 1 of 1 { ;1}

22 1 1 seldom runs, but can be called on to meet unexpected outages. ENSTAR also transports gas to HEA at Bernice Lake plant under a special non-cost-based bypass rate designed to retain this load on the system (approved in Letter Order L00, TA - ) as well as on an interruptible basis to a storage facility. Finally, because HEA s Soldotna plant was only recently placed in service and MEA s Eklutna plant is not yet completed, there is limited to no historical data on which to allocate costs. Accordingly, these customers were not included in the COS study; instead, their normalized revenues based on ENSTAR s proposed rates were used to offset the cost of providing service to other customers. Thus, ENSTAR s revenue requirement of $1,,1, shown on page of the (a) filing, was adjusted as follows: Requested Revenue Requirements $ 1,,1 01 Cost of Gas (,1,) Miscellaneous Service Fees (1,1,0) CEA International (1,1) HEA Bernice Lake (,00) Interruptible Storage Service (1,00) HEA Soldotna (,1) MEA Eklutna (1,,0) Net Revenue Requirement $,, The $,, in net revenue requirement and the $,1, in rate base from page of the (a) filing is allocated among customer classes in the COS study. B. Customer Class COS Study Q. How are ENSTAR s customers grouped for purposes of allocating costs using its existing customer classes? A. For COS purposes, ENSTAR s customers are grouped into the following classes: 1 General Service 1 ( G-1 ) Docket No. U-1- : September, 01 Page of 1 { ;1}

23 General Service ( G- ) General Service ( G- ) General Service ( G- ) Very Large Firm Transport ( VLFT ) Fairbanks Natural Gas ( FNG ) Municipal Light & Power ( ML&P ) The General Service classes include both sales and transport customers. The VLFT class presently consists of SPP, although the VLFT cost-based rates are applied to HEA-Soldotna and MEA-Eklutna and also serve as the basis for the Interruptible Industrial Transport ( IIT ) rate. Because of their unique circumstances, FNG and ML&P were treated as separate customer classes. Q. What are the results of the COS study based on these customer classes? A. The complete COS study based on these customer classes is attached as Exhibit BHF-1. The following table summarizes the cost of providing service (excluding gas costs) to each customer class. In addition, the COS study breaks out the total cost of providing service for each class into customer-related costs (i.e., those associated with providing service irrespective of consumption), demand-related costs (i.e., those associated with having sufficient capacity available to meet customers demand for gas), and commodity-related costs (i.e., those associated with gas itself): Docket No. U-1- : September, 01 Page of 1 { ;1}

24 Class Customer Capacity Commodity Total G-1,,,0,,,,1 G- 1,,1,,,0,0,0 G-,0,,0,0 1,,, G-,0,1,,0 1,0 1,, VLFT,1,1, 1,0,1,0 FNG,0, 1,0, ML&P,,, 1,,0, Total,, 1 Q. How do the results of this COS study compare with the normalized revenues being produced under existing rates? A. The following table compares the results of the COS study with normalized test year revenues (excluding gas costs) by customer class: Class Cost-of-Service Current Base Revenues Base Revenue Shortfall Percent Increase G-1,,1 0,,0,, 1.% G-,0,0,,,00 1.1% G-,,,1,00 1,.0% G- 1,, 1,0,,.01% VLFT,1,0,,,, 1.% FNG,,,1.% ML&P,0,,00,1 1,0, 1.% Total,, 1,01, 1,, 1.0% As shown above, the rates currently being charged to each customer class are well below ENSTAR s cost of providing service, especially for the VLFT, FNG, and ML&P classes. Q. Please explain why the percentage increase across the General Service classes is not more uniform. A. The primary reason for this disparity relates to the relative cost of meters for the various General Service customer classes. Previous COS studies were based on G- and G- Docket No. U-1- : September, 01 Page of 1 { ;1}

25 meter installations being approximately and times more costly, respectively, than the average G-1 meter. Recent analyses of meter installation costs, which are performed annually as part of ENSTAR s construction fee and allowance filings (TA - being the most recent), show that the cost of installing meters for the G- and G- classes is now only approximately and times greater, respectively, than the installation cost of the average G-1 meter. Because the relative cost of meter installations is a major driver in the allocation of customer- distribution-related costs, smaller proportions of these total costs were allocated to the G- and G- classes, which largely accounts for the differing percentage increases between the General Service classes. Q. Why are the percentage increases to the VLFT, FNG, and ML&P classes so much greater than the system average? A. There are four primary reasons. First, FNG s current rates were the result of a negotiated settlement in ENSTAR s last rate case agreed to by all the parties and accepted by the Commission, but which did not reflect a full allocation of system costs. Second, as noted earlier, most of ENSTAR s capacity-related costs, which are primarily related to its transmission activities, were apportioned between customer classes in the present COS study using an allocation factor calculated by weighting equally the relative contributions of each customer class to the test year coincident system peak demand and the average day demand. This contrasts with the allocation method used in the settlement of ENSTAR s last rate case, which is the basis of existing rates, that was based only on coincident system peak demand. Because a peak/average demand allocator typically apportions more costs to relatively higher volume users than a peak- Docket No. U-1- : September, 01 Page of 1 { ;1}

26 only demand allocator, the VLFT, FNG, and ML&P classes were assigned a greater share of capacity-related costs than those reflected in ENSTAR s existing rates. Third, none of the costs associated with the South Peninsula Pipeline (also known as the Anchor Point Pipeline) were included in the existing rates of the VLFT, FNG, and ML&P classes, whereas these three classes share a portion of these transmission-related costs in the present COS study. Finally, since ENSTAR s last rate case, its net investment in transmission plant has increased from approximately $ million to $1 million, while its net investment in distribution plant has remained essentially unchanged. Because the rates of the VLFT, FNG, and ML&P classes are largely driven by ENSTAR s transmission-related costs, they have increased more dramatically than those of the General Service classes. Q. What would be the result if a peak-only instead of a peak/average demand allocator was used to allocate ENSTAR s revenue requirement? A. The following table compares current revenues with the results of a COS study where transmission-related costs are allocated using coincident peak demand (excluding gas costs): Docket No. U-1- : September, 01 Page of 1 { ;1}

27 Class Cost-of-Service Current Base Revenues Base Revenue Shortfall Percent Increase G-1 0,1,0 0,,0,1,00.00% G-,00,1,,, 1.% G-,,1,1,00 0,1.% G- 1,0, 1,0, 1,0,1.% VLFT,,,, 1,01,.0% FNG 1,, 1, 1.1% ML&P,,0,00,1,.% Total,1, 1,01, 1,0, 0.% 1 As expected, using a peak-only demand allocator versus a peak/average demand allocator results in greater increases for the General Service classes and lower increases for the VLFT, FNG, and ML&P classes. C. Rate Design Q. What is rate design? A. Having determined the cost of providing service to each of the various customer classes using a peak/average demand allocation of most capacity-related costs, the next step was to convert the class revenue requirements into specific rates to be charged customers. This process of developing rates corresponding to class revenue requirements is often referred to as rate design. Q. What rates does ENSTAR propose for its General Service classes? A. As described earlier, the COS study classifies ENSTAR s costs as primarily customer-related, capacity-related, or commodity-related. Virtually all of the customer- and capacity-related costs comprising ENSTAR s revenue requirement are fixed, and ENSTAR incurs these costs regardless of how much gas is sold. Historically, gas rates were designed to include only customer-related costs in the monthly customer charge, with capacity- and commodity-related costs being included in the volumetric charges. Docket No. U-1- : September, 01 Page of 1 { ;1}

28 Increasingly, gas rates are now being designed to recover fixed costs through a monthly service charge and variable costs through a volumetric charge. As a move towards a contemporary rate design, ENSTAR is proposing that, for its General Service classes, 0% of customer- and capacity-related costs be included in a monthly service charge, with the other 0% and commodity-related costs being reflected in uniform volumetric rates. This produces the following rates for each of the General Service customer classes: Class G-1 G- G- G- Monthly Service Charge Volumetric Rate (Mcf) $1.00 $.00 $.00 $0.00 $1.1 $1.1 $1.1 $ Q. What rates does ENSTAR propose for FNG and ML&P? A. The current tariff provisions applicable to FNG and ML&P consist of a fixed monthly charge, which was designed to recover one-half of each customer s customer- and capacity-related costs, and a volumetric charge intended to recover the remainder of the cost of service. ENSTAR proposes to continue this rate design. From the table on page, the customer- and capacity-related costs of providing service to FNG and ML&P total $, and $,00,1, respectively. Multiplying these amount times 0% and then dividing the product by twelve months produces a fixed monthly charge of $, for FNG (rounded to $,000), and $1,1 for ML&P (rounded to $1,000). Dividing FNG s remaining revenue requirement not recovered by the monthly charge of $0, by FNG s test year volumes of 1,00,00 Mcf produces a volumetric charge of $0.00 per Mcf. Likewise, dividing ML&P s remaining cost of service of Docket No. U-1- : September, 01 Page of 1 { ;1}

29 $1,, by adjusted test year volumes of,, Mcf produces a volumetric charge of $0. per Mcf. Q. How were the rates for ENSTAR s VLFT customer class developed? A. ENSTAR s proposed rates for the VLFT class are based on the costs allocated to it in the COS study and follow the manner in which they were designed in Docket U-00- and U-0-/U-0-0. Customer-related costs of $,1 were divided by one customer and twelve months to calculate a monthly customer cost of $,, which was rounded to $,00. Rates for the VLFT class include monthly demand charges for each Mcf of the customer's contracted peak demand. Because ENSTAR's capacity-related costs are based one-half on peak demand and one-half on average demand, 0% of the capacity-related costs were assigned to the contracted peak demand charge. Taking one-half of capacity-related costs of $,1,1 and dividing it by the adjusted test year contracted peak demand of,00 Mcf and then by 1 months produced a demand cost of $. (rounded to a demand charge of $. per Mcf per month). Finally, the remaining VLFT class revenue requirement of $,00,, after accounting for rounding of the customer and demand charges, were built into rates applicable to volumes. Consistent with the current rate structure, these volumetric rates were based on declining blocks. Unit rates for each block were calculated by decreasing the previous block by percent so that the rate schedule, when applied to the test year usage volumes, produces revenues equal to the remaining cost of service. The block intervals were then extended to accommodate a wide range of potential loads, resulting in the following schedule of volumetric rates: Docket No. U-1- : September, 01 Page of 1 { ;1}

30 First 0,000 Mcf Next 0,000 Mcf Next,000 Mcf Next 00,000 Mcf Next 0,000 Mcf Next 00,000 Mcf Next 0,000 Mcf Next 00,000 Mcf Remaining Volumes $0. per Mcf $0.1 per Mcf $0. per Mcf $0.0 per Mcf $0. per Mcf $0.0 per Mcf $0.1 per Mcf $0. per Mcf $0.1 per Mcf Q. What rates does ENSTAR propose for IIT and Interruptible Transport Storage ( ITS ) service? A. ENSTAR is proposing that rates for IIT service be equal to the VLFT schedule of volumetric charges displayed above (there are no demand charges because IIT service is interruptible). In an effort to prevent the IIT tariff from being misused by customers that are not true large volume users, ENSTAR proposes to continue a minimum monthly volume charge equal to full use of the first block, which equates to $,0. For ITS service, ENSTAR is proposing to use the $0. per Mcf rate from the IIT tariff as the rate for the first block, with the remaining declining block rates and the minimum annual bill being calculated by increasing the existing rates by the average percentage increase in the VLFT, FNG, and ML&P rates, which is 1.%. Q. What rates does ENSTAR propose for CEA s International plant? A. ENSTAR is proposing that the existing rates charged to CEA for service at its International plant of $,00 per month and $0. per Mcf also be increased by the same 1.% average increase in VLFT, FNG, and ML&P rates. This results in a monthly charge of $,00 (rounded) and a volumetric rate of $1.1 per Mcf. Q. What rates does ENSTAR propose for HEA-Soldotna and MEA-Eklutna? A. ENSTAR s proposes to charge rates for transportation to HEA-Soldotna using its VLFT Docket No. U-1- : September, 01 Page 0 of 1 { ;1}

31 tariff, as provided for in the current special contract. Meanwhile, MEA has indicated that it is agreeable to transportation service to its Eklutna plant being under the VLFT tariff. Q. Does this conclude your testimony? A. Yes, it does. Docket No. U-1- : September, 01 Page 1 of 1 { ;1}