Opinion for the Court filed by Circuit Judge WALD.
WALD, Circuit Judge:
The Federal Power Act regulates “the sale of electric energy at wholesale in interstate commerce,” 16 U.S.C. § 824, and requires rates to be “just and reasonable,” 16 U.S.C. § 824d, i.e., ample to allow recovery of a utility’s operating costs and a fair rate of return on capital investment. See, e.g., Public Systems v. FERC, 606 F.2d 973, 978 n.24 (D.C.Cir.1979). The only costs recoverable under federal rates are those allocable to the wholesale transactions, subject to federal jurisdiction, and not those allocable to retail sales, regulated by the state. See 16 U.S.C. § 824(b). In these consolidated appeals, we review two orders1 of the Federal Energy Regulatory Commission (“FERC” or “Commission”) which approved increases in wholesale electric power rates of Southern California Edison Company (“Edison” or “Company”). In No. 80-1527, Edison urges that FERC (1) set an unreasonably low rate of return, (2) understated the Company’s rate base by excluding non-cash expenses from the working cash allowance, and (3) improperly disallowed recovery of certain operating costs. In No. 80-1334, five California cities (“Cities”)2 assert that the approved rate erroneously allowed Edison (1) to charge wholesale customers for ojjerating costs associated exclusively with retail transactions subject to state regulation, and (2) to recover from wholesale customers a portion of losses stemming from a retail fixed-rate contract. For the reasons set forth below, we reject the contentions of petitioners in each case and affirm the Commission’s actions.
I. BACKGROUND
The Cities, wholesale-for-resale customers of Edison, operate publicly-owned utilities which distribute electricity to retail customers within and near their respective jurisdictions. On October 31, 1975, Edison filed [4]*4with the Federal Power Commission3 an increase in its wholesale rates and, as required by Commission regulations, 18 C.F.R. § 35.13, supported its filing with both historical cost of service data and estimates for the calendar year 1976, designated as the “test year.” Within a month of Edison’s filing, the Cities petitioned to intervene, challenging the Company’s cost justification for the proposed increase.4
After “[mjassive discovery,”5 Edison, the Cities and Commission staff participated in hearings held throughout the month of August, 1977, “producing 2,477 pages of transcript, 219 Exhibits and 17 Items by Reference.” 6 In June, 1978, the Initial Decision of the Administrative Law Judge (“ALJ”) decided a “host of more or less traditional electric rate issues” substantially in favor of the Cities7 and directed Edison to revise its filed cost of service figures and rate schedules accordingly. All parties took exception to the Initial Decision on various issues including those raised in these petitions for review. More than a year later, on August 22,1979, the Commission issued Opinion No. 62, affirming the ALJ on the cost issues. Both Edison and the Cities applied for rehearing as provided by section 313(a) of the Federal Power Act, 16 U.S.C. § 8257(a). Opinion No. 62-A followed on March 20, 1980, again affirming, with clarification, the Initial Decision.
Petitions for review of Opinion Nos. 62 and 62-A were filed with this court pursuant to section 313(b) of the Federal Power Act, 16 U.S.C. § 8257(b). Both Cities and Edison challenge those FERC rulings adverse to their positions and each intervenes on the side of FERC in the appeal brought by the other. The appropriate scope of review is defined by the Administrative Procedure Act, 5 U.S.C. § 706,8 and the Federal Power Act; the court is required to accept as conclusive the “findings of the Commission as to the facts, if supported by substantial evidence.” 16 U.S.C. § 8257(b). We must also determine if the Commission’s findings are “reasoned inferences from substantial evidence.” Memphis Light, Gas and Water Division v. FPC, 504 F.2d 225, 236 (D.C.Cir.1974); City of Chicago v. FPC, 385 F.2d 629, 637 (D.C.Cir.1967), cert. denied, 390 U.S. 945, 88 S.Ct. 1028, 19 L.Ed.2d 1133 (1968).
[5]*5II. ISSUES
Under regulations adopted in 1973 and approved by this court in American Public Power Assn. v. FPC, 522 F.2d 142 (D.C.Cir.1975), Edison was required to support its proposed rate increase with cost of service data for two time periods: (1) historical data from the most recent twelve months available (Period I) and (2) estimates of costs in a “test year” (Period II), twelve months beginning any time between the end of Period I and the effective date of the rate filing. 18 C.F.R. § 35.13. Consistent with the statutory requirement that a utility seeking to increase its rates bears the burden of proof to show that the new charge is “just and reasonable,” 16 U.S.C. § 824d(e), the Commission “ ‘will not approve rates based on unsubstantiated cost estimations. The burden [is] on such companies to establish the validity and accuracy for each of their cost estimates.’ ” Village of Chatham v. FERC, 662 F.2d 23, at 28 (D.C.Cir.1981), quoting Filing of Electric Service Tariff Charges, 50 F.P.C. 125, 127 (1973), aff’d sub nom. American Public Power Assn., supra.
Edison failed to meet this burden with respect to a number of cost items. In their stead, the ALJ adopted figures which he derived from the record evidence. We do not find that his determinations, affirmed by the Commission Opinions, are unsupported by substantial evidence or adequate rationale. We decide similarly with respect to the disposition of the allocation issues raised by the Cities.
A. Rate of Return
“Just and reasonable” rates should “enable the company to operate successfully, to maintain its financial integrity, to attract capital, and to compensate its investors for the risks assumed.” FPC v. Hope Natural Gas Co., 320 U.S. 591, 605, 64 S.Ct. 281, 298, 88 L.Ed. 333 (1944). Investors should receive returns commensurate with those earned on comparable risk stock. See Bluefield Water Works & Improv. Co. v. Public Service Comm., 262 U.S. 679, 692, 43 S.Ct. 675, 679, 67 L.Ed. 1176 (1923). Edison challenges as inadequate for these purposes the 12.75% rate of return on common equity set by the ALJ and affirmed by the Commission. Edison presented evidence to support three main contentions: first, that the low market value of its stock compared with book value evidenced a need for a much higher rate of return;9 second, that the lower rate would hurt Edison’s ability to attract capital on reasonable terms because its bonds would lose the favorable ratings of Moody’s10 and Standard and Poors11 utility investment services; and finally, that a reduction in the filed rates was unwarranted because actual recorded costs in the “test year” revealed that the rate of return in fact generated by the filed rates was unreasonably low. We conclude that the Commission has adequately responded to all these contentions, and its decision to reject them is supported by substantial evidence in the record.
The ALJ and the Commission found that a rate of return of 12.75% would allow Edison to adequately compete for the investor’s dollar. The ALJ adopted in large part a comparable earnings analysis in the traditional mode, see FPC v. Hope Natural Gas Co., supra, 320 U.S. at 604-05, 64 S.Ct. at 288-89, put forward by the Commission staff and supported by data on seven other large utilities and enterprises in other industries. A 12.75% rate of return was shown to be consistent with that yielded on equity investments in enterprises determined to be of comparable risk on the basis of thirteen financial factors. See Testimony of L. Cumberland, Exhibit 119 at 3-6, R. 3623-26, J.A. 212-15. As such, the Commission’s ruling meets the requirements of Hope and Bluefield.
In setting the 12.75% rate of return, the Commission agreed with Edison that the [6]*6depressed market price of Edison stock indicated a need for a rate of return higher than the 11.64% that Edison’s own figures showed it had averaged in recent periods. See Opinion No. 62 at 5, R. 5893, J.A. 662. But neither staff data nor Edison data on the nation’s twenty largest utilities supports the need for the 15% rate of return proposed by Edison. In fact, Edison’s evidence suggested that 12.75% would be adequate to bring market price in line with book value. In the five-year period ending a year before the beginning of the test year, these twenty companies averaged a 12.33% return on common equity while maintaining an average market price more than 30% above book value. As the Commission observed, “while the market climate in 1976 (and 1977) may not be strictly comparable to that of the pre-1974 period, the time periods are not so dissimilar as to provide no guidance as to the reasonableness of the rate that Edison has requested.” Opinion No. 62 at 5, R. 5893, J.A. 662. Although the ALJ set the rate near the low end of what he acknowledged as the “zone of reasonableness,” such a decision, when appropriately supported by the evidence, is consistent with the dictates of Hope that ratemaking requires a balancing of consumer interests with that of investors. 320 U.S. at 603, 64 S.Ct. at 288.
The Commission also disagreed that Edison bonds were in jeopardy of being derated if the filed rates were disapproved. Edison had attempted to demonstrate “the magnitude of the derating problem” with reference to a list of utilities that had lost favorable bond ratings between 1969 and 1975. Testimony of H. Fred Christie, Exhibit 3 at 12, R. 2775, J.A. 118. But the ALJ found that Edison bonds had maintained their high rating in earlier years even when return on equity ranged between 9.56% and 10.14%. The Commission also noted Edison’s continued high bond rating through the 1974-75 market decline. And both the Commission and the ALJ cited several indicators of Edison’s improved financial situation since 1975 which could be expected to affect its bond revenues favorably. In sum, FERC found that Edison did not appear to evidence the characteristics of those utilities that lost their high bond ratings. In addition, the Commission observed that state regulatory policies would be a more probable source of difficulty for Edison’s bond ratings since FERC governs only 10% of Edison’s revenues, while the State of California, which allows the lower return of 12.63%, regulates the remaining 90%.
Finally, the Commission correctly rejected Edison’s argument that the filed rates should stand because the rate of return actually produced by the filed rates during the test year fell below 10%, a level acknowledged by all to be unreasonably low. The data underlying the updated cost figures were found to be unreliable, incomplete, and lacking support from testimony and work papers. In comparison, the cost figures accompanying the filing for the test ■ year were heavily documented and had been refined through the extensive hearing process. To validate Edison’s actual recorded cost figures for the test year would involve a repeat of similar lengthy proceedings. Although recorded cost figures may properly be introduced as a check on the reasonableness of prior estimates, Village of Chatham v. FERC, 662 F.2d 23, at 32 (1981), test year results should not be “rejected merely because they varied from the actual results, since that would supplant period II ratemaking with historic-cost ratemaking contrary to Commission regulations,” id. at 29 (citing Indiana Municipal Electric Ass’n v. FERC, 629 F.2d 480, 482-83 (7th Cir. 1980)).
B. Working Cash Allowance
Edison also challenges the Commission’s calculation of the rate base upon which the rate of return is to be applied. Specifically, while acknowledging contrary precedent, Edison contends that the working cash allowance in rate base should include non-cash expenses such as depreciation, materials issued from stores, provision for uncollectibles, insurance provisions, injuries and [7]*7damages, and pension provisions.12 The ALJ disallowed all of these as being “in the nature of ... bookkeeping expense[s] which [are] recouped in cost of service from ratepayers.” Initial Decision at 24, R. 5236, J.A. 327.
Cash working capital is “the amount of cash needed by an electric utility to meet its operating expenses for the period during which the utility has provided services to its customers and has not yet been paid for these services.” FERC Notice of Proposed Rulemaking, Calculation of Cash Working Capital Allowance for Electric Utilities, 44 Fed. Reg. 33,410 (1979) (to be codified in 18 C.F.R. § 35.24). FERC permits an allowance in rate base equal to the amount of cash needed for this purpose. The ALJ explained that the “purpose of the cash working allowance is to compensate the investors for the use value of their money where the company is required to pay expenses prior to receiving from the ratepayers the revenues associated with those expenses.” Initial Decision at 24, R. 5236, J.A. 327. The ALJ in this case held that non-cash “expenditures” should not be included in the working cash allowance relying in part on testimony of a witness for the Cities that there is in fact no actual lag period between the time such expenses are recorded on the books and receipt of the revenue associated with those expenses. Testimony of W. K. Wilkins, Exhibit 89 at 20, R. 3347, J.A. 182 (“Non-cash expenditures do not require cash working capital as these expenses are considered paid when the revenue is received.”). The ALJ also relied on Commission precedent that only expenses that represent an “outlay of cash” are to be allowed in working capital. Union Electric Co., 47 F.P.C. 144, 175 (1972).13
Traditionally, the amount of the cash working allowance has been calculated under the “45 day formula,” which is explained in an early Federal Power Act Decision.
Electric energy furnished by the company during the current month is billed to the customer as of the first of the succeeding month with a fifteen-day discount period. The full period between the dates of, rendition of service and the payment was adopted as the period of lag and the working capital required for this period (exclusive of fuel and other supplies) was determined to be 45/365 of operating costs, other than taxes which in this case do not require initial outlays of capital and depreciation which is a non-cash expense.
Interstate Power Co., 2 F.P.C. 71, 85 (1939). Recently, however, FERC has relied on “lead-lag” evidence to calculate working cash allowance. See, e.g., Opinion No. 55, Southern California Edison Co. Opinion and Order On Rate Increases, FERC Docket No. E-8570, Aug. 1, 1979; Opinion No. 49, New England Power Co. Opinion and Order Establishing Just and Reasonable Rates, FERC Docket Nos. ER76-304,-317,-498, July 19,1979. A lead-lag study divides the total adjusted operating expenses for the test year by the number of days in the year to arrive at an average daily expense requirement. That figure is then multiplied by the average resale class revenue lag time (in days) to yield an actual working cash requirement. See Opinion No. 55, supra, at 9 n.24. The Commission staff submitted a [8]*8recommendation for cash working allowance based on the 45-day formula and Edison presented a lead-lag study. The ALJ adopted Edison’s operating cost and lag-time figures and then subtracted the non-cash expenses from total operating costs as had always been done under the 45-day method.
Edison objects to the Commission’s departure from the 45-day formula as “unfair when that departure involves using Edison’s lower working cash estimate developed by lead-lag analysis as a point of departure for applying adjustments [disallowing non-cash expenses] that have no basis in fact nor in sound ratemaking theory.’’ Brief of Petitioners, No. 80-1527, at 26. Of particular concern are the major items — depreciation, materials and supplies, and prepaid expenses — which are deducted from rate base (and so cease generating a rate of return) at the time they are recorded as operating costs. Edison claims that deduction from rate base occurs 47.56 days (the average revenue lag time for cash expenses) before receipt of the revenue that represents recovery of those costs.
We find that the ALJ was justified in relying on Edison’s cost and revenue lag figures which were supported by record evidence. Disagreement over which items to include in the expense component of the lead-lag formula does not affect the validity of other elements of the study. Further, the ALJ properly excluded the non-cash expenses because Edison failed to show the existence of a period during which investors were denied a rate of return on capital expended to provide service prior to receipt of revenue associated with those expenses. The primary evidence that Edison submitted to justify recognition of non-cash expenses in working cash was testimony that such is the practice of the State of California. Testimony of C. O. Chubb, Exhibit 13 at 2, R. 2923, J.A. 136. For accounting purposes, there are different ways to conceptualize the bookkeeping transfers of non-cash items from rate base to operating expenses. We cannot say that the ALJ was incorrect in excluding non-cash items from working cash allowance computations. Certainly they do not logically fall within the accepted definition of that term, i.e., a short-term loan from investors to provide for interim cash expenses. And although “[i]t is well settled that to deprive public utility investors of a return on capital currently dedicated to public use constitutes an unconstitutional confiscation of property,” Transcontinental Gas Pipe Line Corp. v. FPC, 518 F.2d 459, 464 (D.C.Cir.1975), vacated and remanded on other grounds, 423 U.S. 326, 96 S.Ct. 579, 46 L.Ed.2d 533 (1976), this court has repeatedly declined to “draw accounting rules for the Commission.” 518 F.2d at 465. “We think the petitioner bears a heavy burden when it challenges FPC accounting regulations .... The effect of the accounting rules should not be viewed and analyzed piecemeal .... Rather, the Commission should be free to fashion individual accounting rules unless a rule is arbitrary or capricious, or would clearly and necessarily have the effect of rendering any rate, which is based upon the rule’s use, confiscatory in the constitutional sense.” 518 F.2d at 465. Edison has failed to make such a showing.
C. Operating Expenses
Edison’s final arguments dispute FERC’s decisions to exclude from operating costs the fuel service fee charged by an Edison subsidiary, losses from a nuclear power project, and the shortfall of fuel revenues over fuel costs produced by the operation of Edison’s fuel adjustment clause.
1. Fuel Adjustment Clause
A fuel adjustment clause is a device that allows a utility to pass on to its customers the increasing cost of fuel without filing a new rate schedule each time the price of fuel rises. To take account of actual — as opposed to projected — fuel costs, the Company may add on charges to the base rate reflected in the filed rate schedule. There are two types of fuel adjustment clauses, “cost of service” and “fixed rate.” In a prior case, we explained the differences between the two adjustment clauses:
[9]*9“Cost of service” tariffs are designed to reimburse the utility for its actual fuel expenditures. These tariffs have the advantage of being accurate, but the inclusion of current costs in the monthly billing must be deferred while the utility collects and assimilates current data. “Fixed rate” tariffs incorporate a charge of a “predetermined price per unit based on costs incurred during a past test period, subject to some adjustments.” Fixed rate tariffs are less accurate than cost of service tariffs, but monthly billings need not be postponed until information on fuel costs is tabulated.
Public Service Co. of New Hampshire v. FERC, 600 F.2d 944 (D.C.Cir.1979) (footnotes omitted).
Edison uses a fixed rate fuel adjustment clause.14 Billings for any given service period are calculated by adjusting the base energy charge to reflect the price of fuel in the previous period as a proxy for the cost of energy during the service period rather than the actual cost. Under this method,15 Edison is neither entitled to recoup nor required to return differences between actual fuel costs and increases or decreases allowed under the fuel adjustment charge. In times of rising prices, which Edison predicted would continue through the test year, the customer will be charged a lower rate unit than the company pays during the billing period. See Maine Public Service Co. v. FPC, 579 F.2d 659, 661 (1st Cir. 1978). Edison claims that the monthly shortfall of revenues over costs should be added to its annual operating costs. The ALJ and the Commission, however, disagreed. We affirm their decision. The Commission concedes that, in times of rising fuel costs, a fixed rate fuel adjustment clause will underestimate fuel costs within a given time period. See Brief for Respondent at 25. But this consequence is a result of Edison’s deliberate choice to follow the fixed rate method. In addition, the Commission notes that when prices stabilize or decline, Edison may well recover or even over-recover its costs.
In determining whether Edison should be allowed to account for its fuel adjustment “shortfall” during the test year in base rates, we start from the premise that the “purpose of using a test year for the determination of rates is to match rate base, operating expenses and sales in an actual, representative relationship.” The Ohio Fuel Gas Co., 13 F.P.C. 280, 287 (1954). The test year fulfills this purpose in the context of estimating costs so as to set appropriate base rates. A fuel adjustment clause carries forward the function of matching cost and revenues as actual cost data become available. The problem in this case is that, in view of rising costs, Edison’s fuel adjustment clause will systematically underestimate fuel costs during the test year. Although “the goal of a fuel adjustment clause is to accurately compensate utilities for their fuel costs,” Public Service Co. of New Hampshire v. FERC, 600 F.2d at 952 (emphasis supplied), that goal cannot be fulfilled by a fixed rate fuel adjustment clause in times of rising costs. Edison, therefore, seeks to add the shortfall to its revenue requirement with the result that the base energy charge will be increased. In short, Edison seeks to recover in its base energy charge the projected difference between what its fuel adjustment formula yields and actual energy costs.
We cannot agree that Edison is entitled to accomplish through its base energy charge the task of the fuel adjustment clause. Base rates are to be built on the best estimate of costs in the test year. Moreover, that rate will normally be in effect well past the year on which the rate [10]*10is based. Thus, even if rising costs can be confidently predicted for the test year, that circumstance could change thereafter. Inclusion of the shortfall in the rate base would insure the company against rising costs but overcompensate it if and when costs fell. The purpose of the fuel adjustment clause is to allow modification based on experience rather than estimates. Including estimates of shortfalls in base rates would essentially allow Edison to bolster its rates against a contingency that it may never face, i.e., that the fuel adjuster will underrecover, without providing for a corresponding decrease in base rates if prices decline and Edison’s fuel adjustment clause actually generates overrecovery.
We have previously recognized that, while the cost of service method of fuel adjustment provides the most accurate matching of costs and revenues, utilities may adopt the fixed rate method despite inaccuracies associated with it. Public Service Co. of New Hampshire v. FERC, 600 F.2d at 952. The risk of such inaccuracy, however, is on the utility. See, e.g., Boston Edison Co. v. FERC, 611 F.2d 8, 11 (1st Cir. 1979). Indeed, utilities are given considerable latitude in designing fuel adjustment clauses. When faced with a situation in which, through hindsight, these chosen methods fail to account fully for costs, we do not think it is unfair to place the risk of such an error on the utility. Public Service Co. of New Hampshire v. FERC, 600 F.2d at 952. If Edison is dissatisfied with the recovery allowed under the fixed rate clause, the cure is through adjustment of the formula.
2. Mono Power Company Fuel Charge
Mono Power Company is Edison’s wholly owned subsidiary for fuel exploration and development. An agreement between the two firms provides that Edison advance money to Mono at a rate of return determined by the last rate approved for Edison by the California Public Utilities Commission (CPUC) and that Edison in turn recover all its costs, including the rate of return paid to Edison. This charge, in essence, allows Edison to purchase fuel, when it is developed, at cost. However, at the time Edison filed its rates, delivery of fuel by Mono was not anticipated before late 1977.
Edison argues that “Since . . . present customers use up portions of present finite known fuel sources, their sharing in the cost of finding and developing future supplies seems perfectly sound and reasonable.” Brief of Petitioner, No. 80-1527, at 43. The ALJ rejected Edison’s policy arguments as allowing a “current return on activities which provide no current benefit to current ratepayers,” Initial Decision at 35-36, R. 5247-48, J.A. 338-39, and the Commission affirmed, Opinion No. 62 at 9-11, R. 5897-99, J.A. 666-68. To allow Edison to recover the fuel charge as an operating cost would allow Edison to recover “all costs and the CPUC permitted rate of return on its fuel exploration and development projects conducted, via Mono, for unsuccessful as well as successful projects,” Initial Decision at 35, R. 5247, J.A. 338. As acknowledged by an Edison witness, the effect would be virtually the same as allowing Mono’s investment to be included in Edison’s rate base. Testimony of F. Clisby, Excerpt from Hearing Transcript, 8/12/77, R. 1252, J.A. 62. It was therefore appropriate for the ALJ to apply in this context the precept “that an item may be included in a rate base only when it is ‘used and useful’ in providing service. In other words, current ratepayers should bear only legitimate costs of providing service to them.” Tennessee Gas Pipeline Co. v. FERC, 606 F.2d 1094, 1109 (D.C.Cir.1979).16
3. Abandonment of Nuclear Project
Edison claims in operating costs the losses associated with the abandoned Vidal nucle[11]*11ar plant project. The ALJ correctly disallowed the write-off because Edison failed to meet the burden of showing the expenses were prudent.17
The Federal Power Act imposes on the Company the “burden of proof to show that the increased rate of charge is just and reasonable.” 16 U.S.C. § 824d(e). Edison relies on Supreme Court precedent for the proposition that a utility’s costs are presumed to be prudently incurred. See Missouri ex rel. Southwestern Bell Telephone Co. v. Missouri Pub. Serv. Comm., 262 U.S. 276, 289 n.l (1923). However, the presumption does not survive “a showing of inefficiency or improvidence.” West Ohio Gas Co. v. Public Utilities Comm., 294 U.S. 63, 55 S.Ct. 316, 79 L.Ed. 761 (1935); see 1 A.L.G. Priest, Principles of Public Utility Regulation 50-51 (1969). As the Commission has explained, “utilities seeking a rate increase are not required to demonstrate in their cases-in-chief that all expenditures were prudent.... However, where some other participant in the proceeding creates a serious doubt as to the prudence of an expenditure, then the applicant has the burden of dispelling these doubts and proving the questioned expenditure to have been prudent.” Opinion No. 86, Minnesota Power & Light Co. Opinion and Order on Rate Increase Filing, Docket No. ER76-827, at 14, 20 Fed. Power Service 5-874, 5-887 (June 24, 1980) (footnotes omitted). Further, the Commission has held as sufficient to raise such a doubt evidence that a state public service commission has disallowed an expense as improvident. Id. The ALJ relied on such evidence, see Initial Decision at 32, R. 5244, J.A. 335, submitted by one of the Cities’ witnesses. See Testimony of K. Wilkins, Exhibit 89 at 6, R. 3333, J.A. at 180. Edison failed to “come forward with specific evidence justifying the write-off.” Opinion No. 62 at 8, R. 5896, J.A. 665. As the Commission noted in affirming the Initial Decision, “the evidence presented by Edison merely consisted of vague generalizations about the problems inherent in all building projects.” Id.18
D. Allocation Issues
As Edison’s operations are subject to regulation by both state and federal governments, Edison’s filings must allocate costs between the rates set by the two jurisdictions. See 18 C.F.R. 35.13(b)(4)(iii), Statement M. Costs are directly assigned to customers where feasible, but some expenses are generally beneficial, and it becomes necessary to “allocate a fair proportion of costs to [each] appropriate category” of customers. Union Electric Co., 47 F.P.C. 144, 172 (1972).
The Cities object to inclusion in wholesale service operating costs of certain expenses that they assert are properly attributable to state-regulated transactions. The Cities argue first that, under the guise of allocating to Cities its proportion of costs of facilities benefiting all Edison’s customers, the Com[12]*12mission charged to the Cities; wholesale customers, a portion of Edison’s costs of retail distribution. Secondly, the Cities dispute the allocation to them of losses from a disadvantageous contract with the California Department of Water Resources (DWR), which is a non-jurisdictional customer. The ALJ, however, decided against the Cities on these issues, and his rulings are supported by “such relevant evidence as a reasonable mind might accept as adequate to support a conclusion.” Consolidated Edison v. NLRB, 305 U.S. 197, 229, 59 S.Ct. 206, 216, 83 L.Ed. 126 (1938) (defining “substantial evidence”).
1. Transmission Costs
Retail and wholesale customer classes are allocated proportionate shares of the total costs of the Company’s “electric transmission system which operates as an integrated, cohesive network in moving electric energy in bulk and which is designed and constructed to achieve maximum efficiency and reliability at minimum cost on a system-wide basis.” Public Service Co. of Indiana, 56 F.P.C. 3003, 3035 (1976). Whether hardware is part of the common system or instead serves a distribution function for specific customers is a question of fact.
The Cities claim that the Initial Decision, summarily affirmed on this issue by Opinions 62 and 62-A, adopted a novel and inappropriate method for definition of those transmission facilities that benefit ratepayers systemwide. The Cities confuse the issue by alleging that the ALJ deviated from previous practice and classified costs as allocable to the Cities based on the voltage of a facility rather than on whether it performs a systemwide function. The issue, as we understand it, however, was not whether function or voltage should be the basis of classification but whether voltage or the particular method proposed by the Cities should be used to determine whether a facility serves a systemwide function. The Initial Decision shows that the ALJ so defined the issue: he referred to “Edison’s method of functionalizing and allocating transmission investment and expense” and relied on Commission precedent which defined facility function by voltage, Opinion No. 821, Southern California Edison Co. Opinion and Order Affirming Initial Decision, FERC Docket No. E-8176, Sept. 22, 1977. Further, because the use of voltage to define function was consistent with precedent, there was no need, as urged by the Cities, for the Commission to justify a change of policy.
Edison met its burden of proof to show that costs of all facilities at or over the 66 kv level were of systemwide benefit and the ALJ, therefore, correctly approved allocation to the Cities of a proportionate share of their cost. Edison witness Lawrence Hedrick, who prepared Edison’s cost allocation statement, offered a plausible rationale for why 66 kv facilities should be regarded as part of systemwide transmission facilities of common benefit.
[I]n fact, the 66 kV system does provide support to the Bulk Power System and is an interconnected transmission network .... For example, peaker generation is transmitted over the 66 kV system and 66 kV sychronous condensors provide stability to the Bulk Power System. The fact that the 66 kV system can not fully backup the Bulk Power System does not dilute the fact that the 66 kV system is an interconnected transmission network and does provide support to the Bulk Power System ....
Testimony of L. Hedrick, Exhibit No. 17, at 11, R. 2958, J.A. 144. The Cities’ claim that 66 kv hardware serves only distribution functions, attributable to specific customers, is supported only by an unsubstantiated 1974 study, that was “prepared by unidentified Edison employees, is not complete, [and] has not been approved by the Company,” Initial Decision at 56, R. 5268, J.A. 359. As the Cities were unable to elicit or submit any information about the study that might sustain its evidentiary value as rebuttal of Edison’s testimony, the ALJ was entitled to rely on Edison’s explanation that 66 kv facilities serve a system-wide function.
[13]*132. Contract Losses
Under a contract (referred to as the “Supplier’s Contract”) entered into in 1966 with the DWR, Edison, along with two other private utilities and the City of Los Angeles, is obligated until 1983 to sell energy at a fixed rate which is now substantially below Edison’s costs. The Cities object that Edison’s filing allocates to the Cities a share of the losses from these below-cost sales. The ALJ allowed inclusion of these costs in the wholesale rates on the grounds that the Cities benefit from two related contracts, entered into in 1967 between the DWR and the private utilities party to the Supplier’s Contract, under which Edison purchases low-cost power from the DWR. Under one of the purchase contracts, Edison purchases at a favorable fixed rate the output of two hydroelectric generating plants. Under the other purchase contract, Edison provides transmission capacity and purchases power contracted for by DWR from Canada and the Northwest which is in excess of DWR’s needs. The losses from sales to the DWR are, therefore, in effect an element of the cost of purchases from the DWR which is properly allocable to those customers who receive that power.
The Cities argue that the three contracts are not in fact related because there is not a total identity of parties in all of them, there are different termination dates, and performance under one is not explicitly dependent upon performance under the others. The Cities also rely on an earlier contrary Commission determination of the lack of interdependence among the contracts, see Opinion No. 55, supra (finding that Edison had failed to establish a relationship between two of the contracts at issue here). That opinion records, however, that “no evidence” of a relationship was introduced by Edison, id. at 14; and the Cities conceded in oral argument that a letter agreement ber tween the contracting parties offered as evidence in this ease had not been submitted in the earlier proceeding. We find that this letter agreement, discussed below, is sufficient evidence, in conjunction with the contracts themselves, to support the ALJ’s factual finding that “the parties entered into an overall plan, embraced in three contracts, for the sale and purchase of power to and from the State Water Plan.” Initial Decision at 55, R. 5267, J.A. 358.19
The “letter agreement” consists of a letter to the DWR from Edison and the other parties to the Supplier’s Contract, executed the same day, which addresses itself to the three elements of the plan for the power exchange which were ultimately covered in three separate contracts. Supplier’s Letter Agreement, Exhibit 144, R. 3840, J.A. 282. The letter describes the terms and conditions to apply between the time the Supplier’s Contract became effective and the effective date of the contemplated purchase contracts. Further, recitations in the Supplier’s Contract itself suggest its interrelationship with the other two contracts.
7. WHEREAS, for the operation of the State Water Project, State will require the transmission of power between the points of supply and the points of use; and
8. WHEREAS, Suppliers generate, transmit, transform and distribute electric capacity and energy by means of their interconnected electric systems; and
9. WHEREAS, Suppliers are willing under the terms and conditions hereinafter set forth to sell, exchange and transmit, and State is willing to purchase and exchange, electric capacity and ener[14]*14gy to supply a portion of Project Power Uses; and
10. WHEREAS, City and Edison are willing, under the terms and conditions hereinafter set forth, to transmit State’s Northwest Dump Energy during Off-Peak Periods over transmission capacity in City EHV Line and over transmission lines which connect thereto;
11. NOW THEREFORE, in consideration of the mutual covenants herein set forth, the parties hereto agree as follows[.]
Contract Between California Suppliers and the State of California for the Sale, Exchange and Transmission of Electric Capacity and Energy for the Operation of State Water Project Pumping Plans, Answering Brief of Intervenor Southern California Edison Company In Docket No. 80-1334, App. C. Together, the letter agreement and the contracts support the conclusion that the contracts, and their offsetting losses and benefits, were intended by the parties to be interrelated components of one overall transaction between Edison and the State.
CONCLUSION
For the foregoing reasons, Opinion Nos. 62 and 62-A are affirmed.