Opinion for the Court filed by Circuit Judge WALD.
WALD, Circuit Judge:The Cities of Batavia, Naperville, Rock Falls, Geneva, Rochelle, St. Charles and Winnetka, Illinois (“Cities”) bring this consolidated appeal challenging decisions of the Federal Energy Regulatory Commission (“Commission” or “FERC”) approving a wholesale rate increase filed by Commonwealth Edison Company (“Com Ed”). Cities, wholesale Rate 78 customers and retail Rate 6 competitors of Com EM, allege that the wholesale rate increase violates the Federal Power Act section 205 requirement that rates be “just and reasonable.” 16 U.S.C. § 824d(a). They allege certain defects in the calculation of rate base, cost of service, rate design and rate of return, which they claim produce a wholesale rate so excessive that Cities are “price squeezed” from the large industrial retail Rate 6 market. Our review reveals that the rate increase was reasonable and so we affirm all but one of the Commission’s many rulings challenged in this appeal. We remand to the Commission one cost of service issue, involving a fuel adjustment clause no longer in effect, because the Commission ap*215pears to have misunderstood the extent of its powers under the statute.
I. BACKGROUND
This case has taken over ten years to travel through the administrative process. It began in 1970, when Com Ed departed from a 1968 proposed settlement agreement with Cities, whereby wholesale Rate 78 and retail Rate 6 were brought into parity.1 The departure took the form of a wholesale Rate 78 increase unaccompanied by a comparable retail Rate 6 increase. An ALJ sustained a challenge by Cities to the increase, Commonwealth Edison Co., 51 F.P.C. 97 (1972), finding it discriminatory, and ordered the wholesale rate to be brought back into parity with the retail rate. Id. at 116. However, the Commission reversed, ruling (1) that it lacked jurisdiction to compare jurisdictional rates (wholesale Rate 78) with non-jurisdictional rates (retail Rate 6),2 and (2) that service to the two customer groups was not comparable because the Cities’ (Rate 78) peak demand tended to coincide with the Com Ed system’s peak, while the retail industrials’ (Rate 6) peak did not. Commonwealth Edison Co., 51 F.P.C. 86 (1974).
Shortly thereafter, Com Ed submitted a revised Rate 78, which went into effect on October 31, 1974 subject to refund. Cities intervened, Record (R.) 3795, challenging the revised rate as discriminatory and alleging defects in rate design, cost of service and rate of return. R. 3799-3815. Before completion of the Commission’s investigation of the revised rate, Com Ed made a further filing, revising its old fuel adjustment clause,3 which had been lifted from a formally approved rate schedule.4 The Commission thereupon terminated a section 206 investigation5 of the old fuel adjustment clause and accepted the new filing, suspended the revised clause for one day and then allowed it to become effective subject to refund. See pp. 75-77 infra.
Meanwhile, this court issued its opinion in Conway Corp. v. Federal Power Commission, 510 F.2d 1264 (D.C.Cir.1975), aff’d, 426 U.S. 271, 96 S.Ct. 1999, 48 L.Ed.2d 626 (1976) (hereinafter Conway). That case held that although the Commission lacked jurisdiction to fix a utility’s retail rates, its jurisdiction to set a utility’s wholesale rates allowed it to take the utility’s retail rates into consideration and to press wholesale rates to the lower end of the range of reasonableness when wholesale customers who compete with the utility for retail customers were price squeezed from the retail market. On the basis of Conway, we remanded, Cities of Batavia v. Federal Power Commission, 548 F.2d 1056 (D.C.Cir.1977) (hereinafter Batavia), and the Commission, finding already submitted evidence sufficient to require further study of the price squeeze issue, reopened proceedings in the case so that additional evidence could be taken on that issue. R. 4003-04.
The ALJ issued his initial decision on May 22, 1978. He found, inter alia, that (1) Com Ed’s rate base was not inflated by allegedly excessive generating capacity reserves, R. 4646; '(2) Cities’ claim regarding the old fuel adjustment clause was no longer viable, R. 4653; (3) Cities’ attack upon Com Ed’s attempt to normalize its taxes was an inappropriate collateral attack upon the Commission’s policy permitting tax normalization, R. 4648-51; (4) Com Ed’s meth*216od of allocating demand costs6 'among classes of customers was inappropriate, R. 4656-61; (5) Com Ed’s proposed 100% and 75%-23 month ratchets7 were unjustified, R. 4661-64; (6) a 13% rate of return on common equity was reasonable, R. 4666-69; and (7) the prima facie case of price squeeze established after the Batavia decision, R. 4003-04, was not sustained because there was no showing that Com Ed specifically intended to restrain competition for industrial retail customers, R. 4677. The Commission subsequently upheld the ALJ’s findings on rate base, R. 4969-76, fuel adjustment, R. 4983-84, tax normalization, R. 4981-82, and rate of return, R. 4992-96. However, it departed from the ALJ’s decision and approved Com Ed’s method of allocating demand, R. 4985-89, and its use of a 100% ratchet, R. 4989-91. It also held that specific anticompetitive intent was not material to a finding of illegal price discrimination, found that Cities had established a prima facie price squeeze case and shifted the burden to Com Ed to refute the existence of an illegal price squeeze, requiring it to submit additional evidence. R. 4997-5022. On January 16, 1980, Cities appealed the Commission’s price squeeze as well as non-price squeeze decisions to this court. Meanwhile, it continued to fight the price squeeze issue during further agency proceedings.
To refute Cities’ price squeeze allegation, Com Ed first submitted a cost of service study comparing the rate of return earned by Com Ed from a select group of 52 large industrial retail customers with that earned from the seven cities. See R. at 2462-81. That study indicated that the difference between wholesale and retail prices for customers with similar cost and demand characteristics was cost justified. Cities responded, in turn, by disputing the Company’s sample as unrepresentative because it allegedly included customers whose demand exceeded the maximum load of the largest municipality and excluded customers whose demand was comparable to that of the Cities. See R. at 2833-37. They submitted an alternate study which eliminated five members of Com Ed’s select group of 52, said to have demands not comparable to the municipalities, and included 146 additional customers whose demands allegedly fell within the same demand range as the Cities. Id.; see also R. at 5006. Cities’ study indicated unjust price discrimination. Using Com Ed’s study, but substituting a different method of demand allocation,8 the Commission’s Staff concluded that the Cities were being price squeezed from the large industrial retail market, R. 4350. The Staff nevertheless refused to recommend relief because the problem was caused by the actions of the state regulatory commission which apparently resulted in a retail rate considered by the Staff to be unreasonably low. R. 4351-52. Although the ALJ’s decision against Cities was ultimately based upon Cities’ failure to prove specific anti-competitive intent, the ALJ observed:
Com Ed serves about 225,000 customers under Rate 6. Depending on which industrial customers are selected for cost of service analysis purposes, and depending also upon the impact of the demand charge, it is possible to calculate that Com Ed derives a lower return from a selected group of retail customers than it earns from Cities. It is also possible to select other groups of industrial customers from whom materially higher returns are earned. The fact is that the group of industrial customers considered by Com Ed and Staff is reasonably representative of industrial customers with electric demand and use patterns that are comparable to Cities, and the results of this study are acceptable for the purpose of this proceeding.
*217R. 4678. The Commission rejected the specific intent requirement, R. 5016-18, and requested the submission of additional evidence. That request was based upon its view that the parties had misapplied the rate of return methodology for determining whether price discrimination existed:
The parties have carved out a group of retail customers similar to the wholesale customer in terms of billing demand and then attempted to use that group in determining relative rates of return. The rate of return test for price discrimination does not depend on retail and wholesale customers having comparable cost-determining characteristics.
This is in contrast to a test for price discrimination wherein the retail rates are compared directly with the wholesale rates. A precondition to this comparison is a determination that the costs to serve the retail class are the same as the costs to serve at wholesale. If cost causative factors, such as diversity, coincidence or load factors are shown to be comparable between retail and wholesale service, then a detailed retail to wholesale cost analysis and comparison can be avoided. This test was not used here, even though a group of retail customers of the same size as the wholesale customers (in terms of billing demands) was selected for rate of return comparisons.
Ordinarily, where the evidence is not adequate to sustain a finding on a significant issue, it would be necessary to remand the proceeding for further development. This is not the case here, however, since the rate of return comparison is sufficient to provide a framework for analysis in determining the presence of price discrimination.
Under a rate of return analysis, the relevant question is which retail customers comprise the group for which there is competition between Com Ed and the Cities, not which group of retail customers is of like size to the wholesale customers. If there is price discrimination in the rate under which this group is served with respect to the wholesale rate, then that is the price discrimination that should be the focus of the price squeeze proceedings. We understand that, in this case, all types of customers served under retail Rate 6 are potential targets of competition between Com Ed and the Cities.36 Certainly there is no evidence that the Cities compete only to serve those types of retail customers with billing demands similar to their own. Under these circumstances, the rate of return comparisons should be made between all customers served under Rate 6 (not just the selected group of industrials) and the customers served under the wholesale rate. The wholesale rate to be used for this comparison is the rate found to be just and reasonable, prior to a consideration of whether there exists price discrimination.
R. 5006-09.
Applications for rehearing of this Commission decision were denied. R. 5278. In defending the theory upon which its evidentiary request was based against Cities’ charge that Conway and Batavia were being evaded, the Commission stated:
In Opinion No. 63, we took a close look at whether an intra-class subsidy possibly imposed by Rate 6 in favor of the customers most similar to the wholesale customers would give Com Ed a competitive advantage over Cities. Our analysis was based on two premises. The first premise, which is not disputed by Cities, is that there is competition between Cities and Com Ed for the entire Rate 6 class. The second premise, which cannot yet be proven but is said by Cities to be *218irrelevant, is that the realized rate of return for serving the Rate 6 class is the same as the realized rate of return for serving the wholesale class. Under these sets of conditions, if a subgroup of the Rate 6 class is subsidized by other members of that class, then to the extent there is a competitive disadvantage to the Cities in their efforts to serve the subsidized group within the Rate 6 class, there is a compensating competitive advantage in serving those members of the Rate 6 class who are forced to subsidize other Rate 6 customers. We do not see this situation as ousting Cities from the market, because the market includes the subsidizing customers as well as the subsidized customers.
We should emphasize that the record in this proceeding gives us no reason to believe that Cities are more interested in serving the industrial customers that look like themselves than they are in serving the other members of the Rate 6 class.
R. 5281.
Com Ed made its compliance filing on November 13,1979. R. 5027-43. The filing indicated that the price discrimination between wholesale and retail customers was cost justified:
The rate of return under the “Demand Responsibility” related to Annual System Peak method is 7.95% for the Cities and 8.41% for the retail customers based on the rate level in effect on October 31, 1974. The rates of return for the retail customers based on four other rate levels of Rate 6 effective in the locked-in period of the wholesale rate are 10.14%, 11.83%, 13.83% and 14.78% as shown on Pages 1 and 2 of Schedule 1.
The 7.95% rate of return for the Cities is less than the 8.87% return authorized by the Commission in Opinion No. 63 and less than any of the returns computed for the retail Rate 6 customers.
R. 5036. Cities objected, arguing that Com Ed’s evidence was flawed9 and therefore did not overcome the prima facie price squeeze case that they had previously established and reinforced. Com Ed acknowledged imperfections with the compliance filing, but contended that it was based upon the best evidence available. R. 5403-15. The Commission agreed10 and accepted Com Ed’s compliance filing. R. 5416 — 18. Cities’ application for rehearing, again disputing the validity of the data and requesting full hearings, R. 5419, was denied. R. 5489. The Commission reaffirmed its view that the evidence submitted was adequate to overcome Cities’ prima facie price squeeze case and the evidence submitted to bolster that case, concluded that Cities used an improper statistical method in their challenge to the compliance filing, and rejected the claimed need for full hearings.11 A petition for review was filed with this court on March 6, 1981. On April 8, we consolidated that appeal with the one filed by Cities on January 16, 1980, raising price squeeze as well as related non-price squeeze issues.
*219II. JURISDICTION
We are met at the outset with a challenge to our jurisdiction to decide the non-price squeeze issues raised by Cities.12 The Commission contends that Cities’ January 16,1980 petition for review of the Commission’s Opinion No. 63-A, which concluded agency proceedings on the non-price squeeze issues, was filed one day late.13 The Commission calculates that the final day for filing was January 15, 1980 — sixty days after Opinion No. 63-A “issued”14 (November 16, 1979).
Section 313(b) of the Federal Power Act, 16 U.S.C. § 8251(b), provides that a petition for review of an agency order15 must be filed within sixty days of that order.16 The Commission reminds us that *220that statutory provision is “jurisdictional and unalterable.”17 Nevertheless, under the agency’s own regulations, 18 C.F.R. § 1.13(a),
in computing any period of time '. . . the last day of the period .. . shall be included, unless it is a Saturday, Sunday, or a legal holiday in the District of Columbia, in which event the period shall run until the end of the next day which is not a Saturday, Sunday, or a holiday.18
The Commission (and Cities) lamentably failed to notice that January 15th is the birthday of Dr. Martin Luther King and, under section 28-2701 of the D.C.Code, a legal holiday in the District of Columbia. See also D.C.Code § 1-503; Washington Post, Jan. 14, 1982, at B-7, 60,000 Expected For King Birthday Parade (17 states and the District of Columbia observe Dr. King’s birthday). Thus, the statutory period for filing a petition for review of Opinion No. 63-A ran until January 16, 1980. Cities’ petition was timely filed and so we reject the Commission’s jurisdictional challenge, which has already been rejected by a previous panel of the court. Commonwealth Edison Company v. Federal Energy Regulatory Commission, No. 80-1063 (D.C.Cir. July 7, 1980) (motion to dismiss denied by per curiam order); Cities of Batavia v. Federal Energy Regulatory Commission, No. 80-1072, (D.C.Cir. Aug. 7, 1980) (petition for rehearing of July 7, 1980 per curiam order denying motion to dismiss denied; leave to renew jurisdiction issue before merits panel granted).
III. EXCESS GENERATING CAPACITY
Cities contend that Com Ed’s excess generating capacity during the test year was 12.9% higher than normal for the longer period during which the proposed rate would be effective, and that the Commission therefore erred in refusing to “normalize” or “adjust” the reserves that it allowed in the test year rate base so that an averaging of excess capacity would be effected.19 Although inartfully advocated in the briefs and not pressed during oral argument, Cities’ position appears to be that rates set on the basis of a test year, which is characterized by maximum unused capacity, will allow over-recovery in subsequent years unless an adjustment is made. As rehearsed by the Commission, Cities’ argument is:
If no adjustment is made to the test year, the rates are determined by dividing the capital carrying charges on generating plant in the test year by the KW of billed demand during the same period. Rates calculated in this manner will yield the *221revenue requirement associated with generating plant in the test year. In subsequent years, however, if billable demand increases but generating plant does not (until the next cycle begins), the utility will receive revenues in excess of cost (including permitted return on equity).20
The phenomenon described by Cities cannot be gainsaid. Calculating rates by dividing capital carrying charges on generating plant by demand during a test year, when reserves are highest and demand lowest of any period during which the rates are in effect, will mean increased recovery on generating capacity in subsequent years: demand will grow, reserves will shrink, but the utility will still be allowed to charge the test year unit price, but for many more units.
The Commission nevertheless rejects the argument, acknowledging that the “test year approach” to rate-making presupposes certain inexactnesses. Depletion of reserve capacity subsequent to the test year will, it says, inevitably be accompanied by some increased costs to serve increased demand as well as increased revenue. Thus, the increased recovery on generating capacity will be somewhat offset. However, the Commission, recognizing that increased costs and revenues will not usually be synchronized, declares itself receptive to “a showing that there will be no offsetting cost increases” — though it warns that such a showing could probably not be made. R. 4976.
It is clear that Cities does not intend to challenge the entire test year approach. We have already declared that approach to be a reasonable and perhaps even a necessary one, though inevitably flawed. See Villages of Chatham v. Federal Energy Regulatory Commission, 662 F.2d 23, 35 (D.C.Cir.1981) (hereinafter Chatham); see also Illinois Cities of Bethany v. Federal Energy Regulatory Commission, 670 F.2d 187, 200-201 (D.C.Cir.1981), as modified on rehearing (hereinafter Bethany). Cities concede that a certain reserve capacity is always necessary. The basic question they raise is how much inexactness can be tolerated in the test year approach before rates calculated thereby cease to be “just and reasonable.”
The record indicated that Com Ed’s excess generating capacity was 30.8% during the test year. R. 450 ct seq. The Commission’s recommended margin is 15-20%. See Chatham at 35. In Bethany we tolerated a 30.2% excess for the test year, id. at n.59, but we indicated then that such a figure was near the point where a different rate *222treatment of excess capacity might be necessary to attain a “just and reasonable” price. Id. at n.65. Because, as the Commission recognized, it will be extremely difficult, if not impossible, for Cities to make out a case that increased revenues in post-test years will not be offset by increased costs incurred in serving increased demand, there will come a point in future cases when it may be basically unfair to impose such a burden on similarly situated parties. At that point, the presumption that costs will rise to offset increased revenues may be “arbitrary and capricious” and so the burden of proof might have to shift to the utility to show that increased revenues will likely be offset. Certainly, at that point, the Commission would be required to give a fuller statement of reasons than provided here for sanctioning such inexactness. In this case, we have come close to that point, but have not crossed over.
IV. COST OF SERVICE ISSUES
A. Fuel Adjustment Clause
Cities challenge the Commission’s October 29, 1974 order which refused to reject or suspend Com Ed’s then effective fuel adjustment clause21 pursuant to its section 205 rate review authority, 16 U.S.C. § 824d(e). The Commission instead instituted a section 206, 16 U.S.C. § 824e, investigation. Commonwealth Edison Co., 52 F.P.C. 1072, 1073-74 (1974). It found that Cities’ challenge bore on the “justness and reasonableness” of the fuel adjustment clause, but because the challenged fuel adjustment clause had been approved by the Commission in a previous proceeding,22 the Commission believed that it lacked authority to proceed according to section 205, which concerns newly filed rate schedules.
The distinction between sections 205 and 206 was discussed in Otter Tail Power v. Federal Energy Regulatory Commission, 583 F.2d 399 (8th Cir. 1978) (hereinafter Otter Tail). Under section 205, the Commission has discretion to suspend new schedules for a period and to require the utility to refund any portion of the increased rate not found justified.23 Under section 206, the Commission may, upon its own motion or upon complaint, investigate existing rates and, if those rates are found unjust, unreasonable, unduly discriminatory or preferential, the Commission must deter*223mine the rate to be observed thereafter.24 The significance of .the distinction for present purposes is obvious. If an investigation indicated that the fuel adjustment clause in effect on October 24, 1974 resulted in an overcharge, the Commission would be able to refund the overcharge under section 205 but only to correct the overcharge in futuro under section 206.
While the section 206 investigation was pending, Com Ed filed a revised fuel adjustment clause to comply with a regulation change. 18 C.F.R. § 35.14, as amended by Order No. 517, 52 F.P.C. 1304. Because section 206 relief would have been prospective only, the investigation was terminated. R. 4984. Cities claim that the investigation would have revealed that they had paid $600,000 in overcharges while the old clause was part of the new schedule. Initial Brief of Cities at 50.
Cities argue that the investigation should have been conducted pursuant to section 205 because a fuel adjustment clause is an integral part of a filed rate schedule and is therefore subject to the Commission’s suspension and refund authority. Id. at 52. In support, Cities cite the definition of “rate schedule” in 18 C.F.R. § 35.225 and the Eighth Circuit’s opinion in Otter Tail. The definition of rate schedule, encompassing “all classifications, practices, rules, regulations or contracts which in any manner affect or relate to” the rate, sheds some light on the issue confronting us, but not enough to answer whether the Commission may institute a section 205 investigation of a provision formerly approved as part of an old schedule and included within a new schedule. Likewise, Otter Tail, cited to support the proposition that any change in any component of the rate schedule constitutes a change in the rate schedule itself and thereby subjects all aspects of the newly filed rate to the Commission’s suspension and refund authority fails to aid Cities’ position. Otter Tail, while acknowledging the agency’s discretionary power to suspend a newly, filed rate change, held only that whether a schedule was an initial rate — not subject to suspension — rather than a rate change — subject to suspension — was a threshold judicial question. It was not concerned with whether a previously approved fuel adjustment clause was within or beyond section 205 jurisdiction. Nevertheless, there is authority to suggest that the Commission may have section 205 authority over some previously approved fuel adjustment clauses when carried over into new rates. For instance, Electric and Water Plant Board of Frankfort, Ky. v. Kentucky Utility Co., 15 P.U.R. 4th 271, 280 (1976), states explicitly that the fuel adjustment clause involved there was “ ‘an integral part of the wholesale transaction, and cannot be isolated and considered apart from the remainder of the agreement between the parties.’ ”26 And in Public Service Co. of New Hampshire v. Federal Energy Regulatory Commission, 600 F.2d 944, 947 (D.C.Cir.1979), we announced that “[a] utility’s fuel adjustment formula must be approved by the Commission, for it is considered part of the utility filed rate.” It is not difficult to envision many situations where the interaction between a fuel adjustment clause and new components of a revised rate will make the operation of the clause itself unjust or *224unreasonable.27 If that happens, we cannot conceive that the Commission would be unable to subject the fuel adjustment clause to its suspension and refund authority. We certainly can find nothing — and the Commission has cited nothing — to command such an unlikely result.
We are aware that this court recently rejected the notion that simply “because a company files for higher rates, it bears the burden of proof on those portions of its filing that represent no departure from the status quo.” Transcontinental Gas Pipo Line Corp. v. Federal Energy Regulatory Commission, 642 F.2d 1335, 1345 (D.C.Cir. 1980) (hereinafter Transco). But we do not read Transco as precluding the Commission from reviewing a revised rate completely to assure that all its parts — old and new — operate in tandem to insure a “just and reasonable” result and from ordering refunds if the previously approved fuel clause operates with new provisions to produce an over-recovery. On this record, we are unable to determine the effect of the old clause in the revised rate schedule. And, in any event, that determination must be made initially by the Commission.
To hold that the Commission had section 205 jurisdiction to look at how the fuel clause operated is not, of course, the same as ordering it to exercise that jurisdiction and granted a refund. As the Commission points out in its brief, a decision by FERC not to suspend (or refund) is an exercise of discretion. See Papago Tribal Utility Authority v. Federal Energy Regulatory Commission, 628 F.2d 235 (D.C.Cir. 1980). However, the Commission’s decision not to review the clause under section 205 appears to be based upon its mistaken belief that whenever a previously approved provision is included within a revised schedule the Commission is precluded from proceeding under section 205. Therefore, we remand to allow the Commission to determine whether the old fuel adjustment clause interacted so differently in the context of the revised schedule that it effected an unjust and unreasonable rate and, if so, to determine whether this is an appropriate case in which to exercise its section 205 suspension and refund jurisdiction.
B. Tax Normalization
Whether certain tax benefits ought to flow through to customers or be “normalized” by the utility has been a controversial academic28 as well as practical problem for some time. We set out the terms of the debate recently in Public Systems v. Federal Energy Regulatory Commission, 606 F.2d 973 (D.C.Cir.1979) (hereinafter Public Systems):
Under normalization, a utility receives the benefit of the tax deferral but charges rates as though the postponed taxes had been paid. The difference between the taxes actually paid and the larger amount charged to rate-payers is ordinarily retained in a “deferred tax” account. Until the deferred taxes fall due, the funds “are available to the company as working capital free of any charges for interest or dividends.” Without normalization, intervenors claim, utilities may have to raise future rates to pay the deferred liability, thus shifting expenses incurred by current consumers to future ratepayers.
Advocates of flow-through, however, contend that under normalization the consumer pays “phantom” taxes. Unless tax benefits are passed on to the taxpay*225er, the argument goes, the utility will reap a permanent tax saving, since so long as the expenses subject to normalization remain stable or increase, the benefits of postponing present tax liabilities will more than offset previously deferred taxes that currently fall due. This proposition is strengthened in periods of inflation, when current liabilities will likely exceed previously deferred expenses.
Id. at 976 (footnotes omitted). Ultimately, determination of which policy to pursue is a matter reserved for the Commission, though matters incidental to that determination may require judicial guidance. The Commission presently has the matter under reconsideration on remand from this court.29 Our guidance is sought here to judge a challenge to the application of an interim policy allowing normalization subject to refund.
The rates filed by Com Ed on August 30, 1974, R. 3245, included adjustments that negated the flow-through of current tax savings to Cities. On June 18, 1975, the Commission announced a general policy permitting tax normalization when appropriate factual showings were made. Order No. 530, 5 F.P.S. 5-1017, 1024 (1975). In an order denying rehearing, Order No. 530-A, 8 F.P.S. 5-224 (1976), the Commission explained that an appropriate factual showing was one which demonstrated that tax normalization would result in tax deferral rather than permanent tax savings. Id. at 227. Opinion 530-A was subsequently modified and a general policy was adopted permitting normalization, as long as it involved “timing differences rather than . . . permanent differences between book and tax treatment.” Opinion No. 530-B, 10 F.P.S. 5-187, 188-89 (1976). In denying a further request for rehearing, the Commission stated that it intended
to permit normalization of items, as discussed in Order No. 530-B, only in those rate cases where the utility has requested normalization in its direct evidentiary presentation. The Commission will not entertain requests to amend pending rate cases to reflect normalization of items for which such treatment has not already been requested. This is sound administrative policy and will help prevent undue delays in the resolution of pending rate proceedings.
Order Denying Rehearing Of Order No. 530-B, Docket Nos. R-424, R-446 (September 3, 1976), slip op. at 78 (emphasis in original). That case then came before this court and because we found that the Commission failed to “(1) assess the consequences of its action for the industry, and (2) indicate ‘fully and carefully’ the purpose behind the order,” the case was remanded. Public Systems at 980. In Missouri Power & Light Co., 17 F.P.S. 5-823 (1979), the Commission decided to continue to allow normalization subject to refund pending reexamination of the matter pursuant to Public Systems. Shortly thereafter, a general policy allowing interim normalization was adopted:
Unless and until the Commission changes its position on remand, we believe that it would be administratively wasteful to require all utilities seeking normalization of tax items, in accordance with Order No. 530-B, to file evidence to demonstrate that a tax deferral rather than a tax savings would occur under tax normalization. However, a utility will be permitted to introduce such evidence. Only in the event that the utility submits evidence on the issue will the Intervenors or Staff be allowed to submit evidence in support of or in opposition to such a showing, or will cross-examination on the issue be permitted.
Order Establishing Interim Procedures (June 8, 1979), FERC Brief, September 11, 1979, Appendix (emphasis in original). The interim policy thus established a two-track procedure: one track’allowed normalization without evidentiary support; the other allowed normalization supported by evidence *226subject to challenge. In case tax normalization is ultimately rejected by the Commission, the option of postponing a showing of tax deferral rather than saving allows administrative resources to be spared. For those utilities more optimistic about' the fate of normalization, evidentiary support can be submitted before final agency action on the general policy so that the case need not be reopened after normalization is, if ever, finally adopted. That second track, followed here, R. 4969, 4982, in effect, continued the pre-530-B procedure, which allowed normalization on a case-by-case showing that normalization was appropriate.
Cities argue that the interim policy continuing Order No. 530-B should not have been applied to this case because: (1) it was initiated before the issuance of Order No. 530, the initial general policy permitting normalization, and therefore application of the interim policy would constitute a retroactive application without notice contradicting the Commission’s own rules and regulations; (2) the record here does not indicate that the tax benefits involved will not be permanent; and (3) the interim policy unnecessarily prolongs implementation of this court’s order in Public Systems. The Commission answers that Cities’ challenge is “an impermissible collateral attack on the Order Establishing Interim Procedures . . . . ” FERC Brief September 11, 1980, at 25. We find that answer unresponsive because Cities’ challenge here goes principally to the propriety of applying the interim order to this case, not to the general interim policy of tax normalization (although their challcnge below did seem basically to be an attack upon normalization per sc). Nevertheless, we do not find Cities’ challenge persuasive for the following reasons.
First, given the posture of this case, we do not find that there is any unlawful retroactive application of Order No. 530-B as carried forward by the interim policy. As already noted, the pre-530-B case-by-case procedure was, in effect, followed here. See Texas Gas Transmission Corp., 43 F.P.C. 824 (1970), aff’d sub nom. Federal Power Commission v. Memphis Light, Gas and Water Division, 411 U.S. 458, 93 S.Ct. 1723, 36 L.Ed.2d 426 (1973). Cities therefore had notice and an opportunity to respond to specific evidence offered by Com Ed supporting normalization. Thus, even if, as Cities argue, Com Ed were not eligible to benefit from Order No. 530-B or the interim policy carrying it forward because Com Ed did not make its “direct evidentiary presentation” in support of its use of normalization in its original (August 30, 1974) filing,30 it was still eligible — as it would have been had this case come up before Order No. 530-B — for a specific determination of the appropriateness of normalization.31 Hence, nothing in the Order Denying Rehearing Of Order No. 530-B prevents the Commission from allowing Com Ed to normalize its tax benefits.
Cities’ second argument is that there is no support in the record for the ALJ’s conclusion that “the taxes in issue do not represent a tax savings.” R. 4649. As we have noted above, there was testimony sufficient to support that conclusion. R. 2370-76. The Staff had acknowledged that Com *227Ed met the requirements of Order No. 530-B,32 and the Commission accepted the ALJ’s conclusion. R. 4982.
Finally, we recognize that the ultimate question whether to allow normalization as a general policy is not an easy matter to resolve. Actually, the Commission appears to be near closure on the question.33 There is nothing to suggest that the Commission is dragging its feet and so we cannot accept Cities’ argument that, by its interim policy, the Commission has unduly delayed implementation of our order in Public Systems.
V. RATE DESIGN
A. The Single Peak (1-CP) Allocation Method
Cities challenge FERC’s decision to accept a Single Peak (1-CP) method of allocating demand costs among classes of customers. Cities argue that use of a 12 monthly coincident peak average (12-CP) would be more accurate and more equitable. In order to fully understand the issue raised by Cities, it is necessary first to outline the allocation policy applied in this case.
Public utility rates are designed to do more than recover the cost of service; they are intended to allocate cost of service among customers in a reasonable manner. P. Garfield and W. Lovejoy, Public Utility Economics 135 (1964) (hereinafter Garfield and Lovejoy). The allocation theory involved in this case is the Hopkinson-type schedule, widely used for medium and large commercial and industrial consumers. See generally id. at 156. The Hopkinson-type rate schedule provides for a two part rate: (1) demand charge (or “load factor” rate), designed to account for fixed or capacity-related cost plus a return on the rate base;34 and (2) energy charge, designed to account for the variable costs (e.g., fuel) of supplying service.35 An individual customer’s energy use is easily calculated and so the energy charge is based strictly upon the individual customer’s use. However, an individual customer’s use of system capacity is not so easily calculated; there is an apparent theoretical and practical problem involved in allocating this joint cost36 among different classes of customers. See J. Bonbright, Principles of Public Utility Rates 350-68 (1961) (hereinafter Bonbright). Because each utility is uniquely structured, FERC’s approach has been flexible; no single method of cost allocation is considered appropriate for all systems. See Florida Power & Light Co., Opinion No. 784-A, 20 P.U.R. 4th 429, 499 (1977); Wisconsin Michigan Co., 31 F.P.C. 1445 (1964); see generally Bonbright at 351-66; Garfield and Love-joy at 143-44, 159-64; 1 A. Kahn, The Economics of Regulation 142-50 (1970) (hereinafter Kahn).
*228The competing methods of allocation involved here are the 1-CP (or peak allocation) and 12-CP methods. Under the 1-CP method, demand costs associated with the system’s peak demand are allocated among classes of customers in proportion to their contribution to that demand. The hypothesis underlying this method is that plant size is determined by the annual peak. Under the 12-CP method, on the other hand, demand costs are based upon the average of each month’s peak demand during the year; the allocation among customer classes is based upon their average contribution to the average monthly peak. The hypothesis underlying the 12-CP method is that plant size is designed to meet “the demands season to season, month to month, day to day and not just the maximum load on the system at any given time or any segment of the year.” Wisconsin Michigan Power Co., 31 F.P.C. 1445, 1455 (1964). In less abstract terms, application of a 1-CP method instead of a 12-CP method will cost Cities $189,000 more during the test year. R. 4656.
The Commission’s decision to accept the 1-CP method, proffered by Com Ed, reversed the decision of the ALJ, R. 4656-58, and rejected the recommendation of the Staff. R. 4883-88. In addition, the Commission itself had refused to accept a 1-CP method in prior ratemaking proceedings, a stance which undoubtedly affected the ALJ and Staff’s position in this case. The primary reason given by the ALJ and the Staff for rejecting the 1-CP method was that Com Ed’s system was designed to meet year-round demands rather than the system peak demand. Sec R. at 4658, 4886. The variance in the system’s monthly peaks was a salient consideration for the AU and the Staff.37 The ALJ relied upon a finding that for the test year (1974), the annual peak was 12,270 mw; the lowest monthly peak was 8,277 mw; the average monthly peak was 9,661 mw. The ratio of average monthly peaks to the system peak was IS.T’/r; the ratio of the lowest monthly peak to the system peak was 67.6%.38 After comparing Com Ed’s monthly peak variation to other companies,39 the ALJ concluded that Com Fid’s system was not single month or summer peaked.40 The Staff agreed with *229the ALJ, R. 4883-88, and bolstered his finding on monthly peak variation, but relied primarily upon reliability data,41 which indicated that Com Ed was able to achieve reliability and minimize, peak demand capacity “through the interconnections, capacity exchanges, energy purchases, off-peak scheduling of maintenance, and timing of other outages.” R. 4886. Such efforts, the Staff found, allowed the system to operate with “roughly the same excess capacity throughout the year.” R. 4887. The Staff refused to rely upon forecast data. R. 4888.
In rejecting the ALJ’s decision and the Staff’s recommendation, the Commission recognized that the record was “less than fully satisfactory.” R. 4989. The record revealed no specific information about the 1974 test year monthly reserves or monthly reserve variances. R. 4987. However, the forecast data for 1976-85 showed significantly lower reserves in the summer peaking months than during the rest of the year. R. 4986. In addition, the record showed “substantial summer peaking” during the test year, R. 4987-88, and during the ten-year forecast period (1976-85), R. 4986. Indeed, the forecast data indicated intensification of peaking in the future. See R. at 2571-81. That data supported a Com Ed spokesman’s assertion that the Company does, in fact, base its capacity planning upon projected peak demand rather than upon year-round demand. R. 446. Moreover, the method of demand allocation approved by the Commission closely resembled that used by the state commission in designating retail rates. R. 4988.
Our task here is to determine whether the Commission’s decision to uphold a 1-CP method of demand allocation was a reasoned one based upon substantial evidence. On the basis of the record here, we are not prepared to reverse the Commission. The Commission’s comments about the less than fully satisfactory state of the record were designed to emphasize the sui generis nature of its decision — a 1-CP was a just and reasonable method of allocation in this case. The Commission’s position here is in accord with its generally flexible policy toward demand allocation, which requires a consideration of the equities of each case. See Illinois Power Co., 12 F.P.S. 5-893, 899 (1977). There is adequate evidence in this record to support the Commission’s balancing of equities. It goes without saying that courts must accord great weight to an agency judgment. See City of Cleveland v. Federal Power Commission, 525 F.2d 845, 849 (D.C.Cir.1976). And where agencies have been assigned principal policy-development responsibility, a healthy dose of deference is in order. See Federal Election Commission v. Democratic Senatorial Committee, - U.S. -, --, 102 S.Ct. 38, 44, 70 L.Ed.2d 23 (1981). The Commission’s finding below that Com Ed’s system was single or seasonal peaked was not merely a technical determination about the system’s operation, but also reflected a policy decision about how system-wide costs ought to be distributed among customer groups. See Garfield and Lovejoy at 134 ct seq. No particular method of demand allocation has been prescribed by Congress; on the basis of the record here the one accepted by the Commission is not *230arbitrary and capricious; therefore we must affirm.
B. Billing Design
1. The “100% Demand Ratchet”
Cities argue that the Commission erred in approving a “100% demand ratchet.” A ratchet is a billing device that sets the minimum demand cost for a utility customer at some fixed percentage of the customer’s maximum demand during a particular period. See Garfield and Lovejoy at 157. Thus, the ratchet involved here is set at 100%42 of Cities’ peak demand during Com Ed’s peak season, the four summer months.43
While methods of demand allocation distribute the demand charge among classes of customers, a ratchet, as a method of demand billing, distributes a class’ allocated demand among members of that class. See Florida Power & Light Co., 56 F.P.C. 3581, 3609 (1976). There is no necessary relationship between a particular method of demand allocation and a particular method of demand billing. As the ALJ explained in approving a 65% summer ratchet instead of the 100% ratchet requested by Cities and ultimately approved by the Commission, the factors considered in determining demand billing rates are not always the same as those considered in determining demand cost allocation:
In determining an equitable allocation method, the variation in monthly peak loads of the company’s total system, the scheduling of maintenance, and other factors relating to the distribution of total system demand-related costs throughout the year, should be considered.
An appropriate demand rate should provide a reasonable basis for the company to recover demand-related costs from the wholesale customers. Justification for a demand ratchet is based on the relative importance of the individual customer’s demand for electric energy at the company’s peak in contributing to the company’s demand-related costs.
R. 4662-63 (emphasis in original). Although a ratchet is, initially, customer-focused, like demand allocation, it is also designed to bring about a more efficient system by providing an economic incentive for individual customers to level their yearly demand. By encouraging customers to spread their load more evenly, the utility seeks to improve its load distribution, i.e., to eliminate periodic excess capacity.44 A ratchet may therefore be an appropriate device even in a system without periodic (¡caking intense enough to warrant any method of allocation other than a 12-CP.45 Thus, even though the ALJ here failed to *231find monthly variance in the test year extreme enough to justify a 1-CP demand allocation, he recognized that without any ratchet, the disproportionately heavy seasonal demands of certain customers would not be taken into account.
Cities argue here that (1) the ratchet is inconsistent with a 12-CP method of cost allocation, and (2) it is unjustifiable because (a) there is nothing that sets Cities apart from other customers not subject to a ratchet, (b) a ratchet based upon four months is inconsistent with a single month demand allocation, and (c) ratchets do not serve the purpose for which they are designed. In light of our decision to uphold a 1-CP allocation method, Cities’ first argument is no longer viable. Cities’ second argument is likewise without merit. We find that the ratchet approved by the Commission works in harmony with the 1-CP demand allocation method to distribute demand cost in an efficient and fair manner. First of all, the record indicates that the need for new capacity to serve the municipalities has been growing at twice the rate of the Company’s system as a whole.46 Thus, there is support in the record for the distinctive treatment of Cities. Further, Cities were wrong to suggest at oral argument that a ratchet based upon the four summer months is somehow incongruent with a single month peak allocation method. As noted above, see p. 83 supra, there is no necessary relationship between a particular method of demand allocation and a particular billing method. Indeed, demand billing based upon the single highest peak month would only tend to level the maximum demand placed upon the system from the point reached in July to that reached in August, the next highest peak. See n.38 supra. Although the demand charge may be properly allocated on the basis of a single peak, to level demand throughout the year, all peaks, not only the single highest peak, must be targeted by the billing device. Ideally, the pressure applied to July, the peak month, should be such as to en-
courage the elimination of July’s variance from the average demand throughout the year, and the pressure applied to August and other peak months should be proportionate to the pressure applied to July so as to encourage the elimination of their respective variances from average monthly demand. Here, admittedly, the billing design was not optimal: the pressure applied throughout the four months of the peak season was equal rather than weighted to respond to each month’s different variance from the monthly average. Further, we find nothing in the record to insure that the impact of the ratchet will not be so excessive as to transform the slope into a slum]) — an improbable result, yet mentioned to indicate the paucity of evidence in this record about the likely effect of any particular ratchet. Nevertheless, we recognize that the billing design need only be reasonable, not theoretically perfect. Given the evidence in the record about this system, the ratchet does not appear unreasonable. Hence, the Commission’s decision constitutes a legitimate exercise of agency discretion. See Federal Power Commission v. Hope Natural Gas Co., 320 U.S. 591, 603, 64 S.Ct. 281, 288, 88 L.Ed. 333 (1944); Continental Air Lines, Inc. v. Civil Aeronautics Board, 551 F.2d 1293, 1301 (D.C.Cir.1974) (“A theory of ratemaking must be reasonable, explained, and supported, but is not subject to the same substantiation principle as the substantial evidence test applied to fact-finding.”). If the design of the ratchet proves so flawed as to produce any of the envisioned problems, the Commission has the authority to reconsider its approval. 16 U.S.C. § 824e.
2. The 75%-23 Month Ratchet
Com Ed had requested, but was denied, a 75%-23 month ratchet. R. 4661, 4991. Cities claim that they have paid substantial sums to Com Ed under the rejected ratchet provision and that they are entitled *232to a refund.47 The Commission refused to award a refund because:
Com Ed cannot retroactively collect more from any customer [than] that [which] has already been collected subject to refund, even though a redesigned rate presumptively would show some customers should be charged more and others less than under the rates in effect subject to refunds. Also, any change of this nature in the form of charging, to the extent that it has the potential for affecting the pattern and extent of energy usage by customers, can so operate only in the future when customers are aware of the revision and can modify their consumption patterns accordingly.
R. 4990.
Cities rely exclusively upon Federal Power Commission v. Tennessee Gas Transmission Co., 371 U.S. 145, 83 S.Ct. 211, 9 L.Ed.2d 199 (1962), to support their claim. That case upheld an interim refund order by the Commission. In reversing this court’s decision to set aside the interim order, the Supreme Court observed that it was “the duty of the Commission to look at ‘the backdrop of the practical consequences [resulting] . . . and the purpose of the Act’ ... in exercising its discretion under § 16 to issue interim orders and, where refunds are found due, to direct their payment at the earliest possible moment consistent with due process.” Id. at 155, 83 S.Ct. at 216 (emphasis added). It is clear from the passage from the Commission’s decision reprinted above that in denying a refund in this case the Commission also considered the practical consequences and the purpose of the Act; hence we are required to uphold its exercise of discretionary power.48
VI. RATE OF RETURN
The Commission upheld the decision of the ALJ and rejected the recommendation of the Staff49 in approving a 13% return on common equity. R. 4666-69, 4992-96. Cities challenge that decision as arbitrary and unreasonable because it is in excess of rates permitted other comparable utilities. We find the challenge meritless.
The scope of our review of the reasonableness of a particular rate was set out in Transco. Our role is “to insure that the Commission’s judgment is supported by substantial evidence and that the methodology used in arriving at that judgment is either consistent with past practice or adequately justified. ...” Id. at 1351. In the present case, the ALJ’s decision was based upon “factors such as ratio of equity to total capital, coverage of interest obligations, market acceptability of the Company’s securities, debt ratio, imbedded debt costs and related factors.” R. 4660. The ALJ also relied upon the rate authorized for comparable electric utilities. R. 4667. Cities do not, however, dispute the financial evidence underlying the decision. Cf. Anaheim at 803-804. Rather, they argue that the rate allowed by the Commission is “unusually high for a utility with an equity ratio and superior financial ratings” like Com Ed. Initial Brief of Cities at 68. Cities fail to cite any examples to support that contention. But see Public Service Company of Indiana v. Federal Energy Regulatory Commission, 575 F.2d 1204, 1214-15 (7th Cir. 1978) (13% rate of return found not unusual). We therefore affirm the Commission’s decision that 13% is a just and reasonable return.
*233VII. PRICE SQUEEZE
The gravamen of Cities’ appeal is that the wholesale rate increase approved by the Commission “price squeezes” municipal wholesale (Rate 78) customers from the large industrial retail (Rate 6) market.50 The price squeeze complaint developed in antitrust law, United States v. Aluminum Co. of America, 148 F.2d 416, 436-38 (2d Cir. 1945), and established itself only recently in the regulated industries field. See Conway; Batavia; Bethany. In its new environs, it was met with a scheme of dual regulation: wholesale rates regulated by a federal commission; retail rates regulated by state commissions. To adapt to this brave new world, flexibility was essential. Thus, in Conway, the Supreme Court upheld this court’s decision and acknowledged the Federal Energy Regulatory Commission’s discretion in price squeeze situations to press wholesale rates to the lower end of the zone of reasonableness as a means of fulfilling its obligation of insuring just and reasonable rates. Conway, 510 F.2d at 1274, aff’d, 426 U.S. at 278-79, 96 S.Ct. at 2004.
A. Establishment of a Prima Facie Case of Price Squeeze
In this case, Cities allege first that the Commission failed to live up to its obligation under the statute as defined by Conway and the agency’s own regulations implementing Conway. Following the Supreme Court’s decision in Conway, interpreting the Federal Power Act as requiring the Commission to consider the anticompetitive consequences of a wholesale rate in the retail market, the Commission began a rule-making proceeding culminating in the adoption of Order No. 563, 18 C.F.R. § 2.17, which set out the following guidelines for the establishment of a prima facie price squeeze case:
(1) Specification of the filing utility’s retail rate schedules with which the intervening wholesale customer is unable to compete due to purchased power costs;
(2) A showing that a competitive situation exists in that the wholesale customer competes in the same market as the filing utility;
(3) A showing that the retail rates are lower than the proposed wholesale rates for comparable service;
(4) The wholesale customer’s prospective rate for comparable retail service, i.e. the rate necessary to recover bulk power costs (at the proposed wholesale rate) and distribution costs;
(5) An indication of the reduction in the wholesale rate necessary to eliminate the price squeeze alleged.
18 C.F.R. § 2.17(a). Cities maintain that the Commission erred in Opinion No. 63 by allowing Com Ed to submit additional evidence to rebut their price squeeze claim. They maintain that having established a prima facie case, which at that point Com Ed had failed to rebut,51 they were entitled to judgment on the price squeeze issue:
*234[S]ince . . . Edison had not presented valid, substantial cost of service evidence to sustain its burden of proof to rebut the prima facie case of price squeeze established by Cities under the Commission’s Order No. 563, the Commission erred in not ruling in the Cities’ favor.
Cities Brief, April 27, 1981, at 39.
At various times throughout the administrative proceedings, it has been conceded that Cities had established a prima facie price squeeze case. Prior to Conway and promulgation of the Commission guidelines, Administrative Law Judge Ellis held that Com Ed’s rate increases were “unjust, unreasonable, and unduly discriminatory to the extent they are more burdensome to the wholesale customers than the rates set forth in Commonwealth’s Rate 6 ...” Commonwealth Edison Co., 51 F.P.C. 97, 115-16 (1972). That decision was reversed by the Commission on jurisdictional grounds, 51 F.P.C. 86, but the Commission’s decision was in turn reversed, Batavia, and evidence was subsequently taken on the price squeeze allegation since, on the basis of the previously submitted evidence, Administrative Law Judge McGowan found that Cities had established a prima facie case of price squeeze. R. 4003-04. Later, however, a different ALJ held that Cities’ prima facie case was not sustained because Com Ed did not specifically intend to restrain competition. R. 4677. The Commission then ruled that specific intent was not a sine qua non of a price squeeze case in regulated industries law and that
The Cities’ showing of comparable billing determinants, though not dispositive of the price discrimination issue, is sufficient for the parties to go forward in the litigation of the allegations and to shift the burden to Com Ed to demonstrate the cost difference it claims exist.
R. 5004 (Opinion No. 63).
Although Cities had presented a prima facie price squeeze case, we do not agree that the Commission should have held for Cities instead of requiring the submission of additional evidence. The Commission answered Cities’ argument on rehearing correctly, we believe, by noting:
The Commission’s understanding and treatment of price squeeze have evolved considerably. As mentioned in Opinion No. 63, this case was one of the very first price squeeze proceedings. Essential elements in our present approach to price squeeze were not established until well after hearings ended in this case. While we are aware of the problem posed for both seller and buyer by prolonged proceedings, we do not believe it to be fair or consistent with fundamental due process to find in favor of one party or another because an adversary has failed to make a showing it was not aware would be required of it. We are concerned that final disposition in this case based on a party’s failure to make such a showing, rather than on an affirmative finding of the presence or absence of price discrimination as between the wholesale customers and the retail class, would constitute fundamental unfairness, and we shall not take such a course.
R. 5285 (Opinion No. 63-A). In support, we repeat what we said recently in Bethany about the Commission’s price squeeze regulations:
To call 18 C.F.R. § 2.17 a “regulation” ... is to overstate its importance. It was promulgated as a statement of general policy to guide the Staff in dealing with price squeeze cases in the immediate aftermath of Conway; the Commission has felt free to disregard its strictures when conducting price squeeze proceedings.
Id., as modified on rehearing at n.39. The Commission is free, in any particular case, to set a very low threshold for the establishment of a prima facie case — though, of course, were the Commission ultimately to *235decide to exercise its Conway discretion and press wholesale rates to the lower end of the “range of reasonableness” to further competition, its action would have to be based upon substantial evidence and reasoned decision-making. But, a prima facie case need only be enough to inspire agency inquiry, not necessarily enough to support a remedy for a price squeeze. As we explained in Bethany, “unlike the ordinary situation where a party who establishes a prima facie case is, as a matter of law, entitled to relief unless the defending party rebuts the case, the establishment of a prima facie price squeeze case means only that enough has been shown to warrant inquiry into the price squeeze allegation to ascertain whether price discrimination exists.” Id. at 192. Once the Commission is persuaded to ascertain whether unjust or illegal price discrimination exists, the utility’s burden is, quite obviously, to dissuade the Commission of the existence of anticompetitive wholesale pricing. Its burden, therefore, will be defined by the quality of the prima facie case. As long as the Commission’s procedures for presenting and rebutting a prima facie case do not violate due process, those procedures will be judicially sanctioned. Vermont Yankee Nuclear Power Corp. v. Natural Resources Defense Council, 435 U.S. 519, 98 S.Ct. 1197, 55 L.Ed.2d 460 (1978). By allowing Com Ed additional opportunity to present its response to Cities’ allegation, the Commission was well within the boundaries of due process, especially in light of its reformulation of price squeeze analysis.
B. Theoretical Structure of the Commission’s Price Squeeze Analysis
Cities argue secondly that the approach followed by the Commission here to test for illegal price discrimination is contrary to Conway and Batavia in that (1) it ignores the impact of the actual filed rate and erroneously focuses upon the accepted (“but for”) rate; (2) it arbitrarily places wholesale customers in a different class from large industrial customers “merely” because of the former’s resale operation; and (3) it wrongly compares wholesale and retail rates of return rather than simply actual rates.
(1) To determine whether there was a price squeeze, the Commission compared the retail rate of return to the wholesale rate of return found just and reasonable rather than to the filed wholesale rate of return. We believe that this focus was appropriate, yet not entirely without problems. Cities’ objection, however, overstates the magnitude of the problem. They suggest that
Under the Commission’s new [test] regulated utility companies such as Edison are completely free to file a high wholesale rate creating a severe price squeeze, safe in the knowledge that no relief will ever be ordered against them because the Commission will only evaluate the existence or non-existence of a price squeeze based on the level of the reduced cost of service rate determined years later. The actual price squeeze thus completely escapes scrutiny.
The anticompetitive effect begins at the moment the new wholesale rates are allowed to go into effect even if subject to refund because the municipal system risks losing its existing and potential customers, since it is forced to charge higher prices. Alternatively, if it attempts to maintain competitive prices, the monopolist is able to cripple the municipality with the unjustified discriminatory wholesale rate.
[E]ven though at some later date the municipality may sue the monopolist under the antitrust laws, the remedies are statutorily different. Also, if the municipal has not been able to bear the weight of the price squeeze and has crumbled under its pressure, the remedies serve no purpose, since refunds cannot revive a corpus.
Initial Brief of Cities at 31, 32.
It is unfortunate, indeed, that cases like this one take so many years to resolve. The Commission has the ability to suspend the filed rate for only five months; after that time, the rate goes into effect subject to *236refund. 16 U.S.C. § 824d(e). Between the time the filed rate goes into effect and the time the Commission determines a just and reasonable rate but for the allegation of price squeeze, the refund offers some protection but only if any interim price squeeze is not too severe and/or the period between the time a rate takes effect and administrative proceedings are completed is not too long. It is possible, as Cities maintains, that refunds will come too late to revive a beleaguered wholesale customer/retail competitor.
In this case, of course, that possibility did not occur. Indeed, no municipal utility even claimed to be operating at a loss. See R. at 4679 (ALJ’s Initial Decision). Further, the 17.1% difference between wholesale and retail rates lasted only ten months before the state commission increased retail rates. R. 4674. Moreover although possible, it seems unlikely that the problem identified by Cities will have the threatened effect in other cases. We must assume that if a case arises where the evidence of a severe price squeeze is strong the Commission will react appropriately to reduce substantially the normal length of time between the effective date of the rate and final disposition of the price squeeze issue. To do otherwise might be to ignore the spirit, if not the letter, of Conway. Finally, as Cities acknowledge, traditional antitrust remedies are also available to price squeeze victims. See City of Groton v. Connecticut Light & Power Co., 497 F.Supp. 1040 (D.Conn.1980), aff’d in part, remanded for further proceedings, 662 F.2d 921 (2d Cir. 1981). See generally, Note, The Application of the Antitrust Laws To Price Squeezes In The Electric Utility Industry, 54 St. John’s L.Rev. 103 (1979); Note, Antitrust-Price Squeeze, 12 Ind.L.Rev. 637 (1979). While refunds “cannot revive a corpus,” antitrust remedies are designed to do just that.
The approach followed by the Commission here was dictated by its limited role of setting just and reasonable rates. The Commission can take certain anticompetitive consequences into account in setting just and reasonable rates; however, except for its ability to award refunds, it is unable to provide retroactive relief. Whatever is lost because of the pace with which the administrative process proceeds, can be sought along the public and private paths of the antitrust laws.
(2) In attacking the Commission’s approach for arbitrarily placing wholesale customers in a different class from large industrial customers “merely” because of the former’s retail operations, Cities rely upon Batavia. In Batavia, we warned that
If a substantial number of industrial consumers have their peak load at a time roughly the same as those of Cities, there is a substantial ousting from the market with no difference in conditions of service. And when there is a competitive advantage to the utility in blanketing such industrials along with other industrials (whose peak characteristics may produce cost savings), there is room for concern on the part of the regulating commission.
Batavia at 1058. Cities erroneously contend that Batavia requires the Commission now “to determine the key question as to whether the Cities should be considered in a separate class from large industrial customers ...” Cities Brief, April 27, 1981, at 44.
Cities and large industrial retail customers are placed in different classes because their rates are set by two different jurisdictions. The Batavia passage set out above elaborated upon the concern the Supreme Court and this court expressed in Conway. Conway instructed the Commission to consider the anticompetitive effects of a proposed wholesale rate by examining non-jurisdictional (retail) rates when fixing the wholesale rate. Batavia warned the Commission to be particularly attentive when a substantial number ' of non-jurisdictional (retail) customers have cost characteristics similar to jurisdictional (wholesale) customers but are being charged rates so low that jurisdictional customers are price squeezed from the retail market because those non-jurisdictional customers might be grouped with and subsidized by other non-jurisdic*237tional customers for whom jurisdictional customers do not and cannot compete. If there are a substantial number of retail customers for whom Cities could compete but for the subsidy those retail customers receive from other retail customers with whom they are grouped and for whom Cities could not compete, we suggested in Batavia that the Commission ought to consider whether its Conway discretion should appropriately be exercised. In any case, neither this court nor the Commission has the power to abrogate the classification scheme of which Cities complain.52
Significantly, in this case, however, the target of competition was the entire retail Rate 6 market, not just those large industrial retail customers with cost and demand characteristics similar to Cities’.53 Hence, Batavia has no application to the facts involved here. As the Commission explained: “if a subgroup of the Rate 6 class is subsidized by other members of that class, then to the extent there is a competitive disadvantage to Cities in their efforts to serve the subsidized group within the Rate 6 class, there is a compensating advantage in serving those members of the Rate 6 class who are forced to subsidize other Rate 6 customers.” R. 5281 (Opinion No. 63-A). The Commission conceded during oral argument — and we agree — that if Cities had claimed that the target of competition was a subsidized subgroup with cost characteristics similar to Cities and that subgroup was charged rates lower than those charged Cities, this would have been a very different case.
(3) Once the Batavia problem is removed from the case, the ultimate question becomes whether there was a cost justified reason for the 17% difference between wholesale Rate 78 and retail Rate 6 rates. Cities challenge the Commission’s method of addressing that question by comparing wholesale and retail rates of return. Cities would simply have compared wholesale and retail rates to determine whether wholesalers were being price squeezed. Such an approach would uncover price differences but would not disclose whether such differences were cost justified. Conway and Batavia were concerned only with non-cost justified discriminations. And, as we noted in Bethany, a cost of service study is an appropriate means of determining whether an illegal price squeeze is occurring: “If, after wholesale and retail costs are fully allocated,[54] the company’s wholesale profit margin is significantly greater than its retail profit margin, indications are that a[n illegal] price discrimination is occurring.” Bethany, as modified on rehearing at 199 n.53 (emphasis in original).
In this case, the cost of service study indicated no illegal price discrimination. After costs were allocated, the rate of return was 7.95% for Cities and 8.41% for retail Rate 6 customers. R. 5040 (Com Ed Compliance Filing). The 17% differential between wholesale and retail rates was apparently due to the higher costs of serving the wholesale group. R. 4674. Thus, if Cities were unable to compete with Com Ed for certain retail customers, that was the result of market conditions, not anticompetitive pricing. There was, therefore, no need *238for the Commission to exercise its Conway discretion. As we emphasized in Bethany, as modified on rehearing at 200, the Conway doctrine was not designed to subsidize particular retailers but rather to assure that genuinely competitive wholesale customers would not be price squeezed out of a retail market because of the imperfections of regulation, an assumed propensity of utilities for unjustified monopoly power or the coordination problems due to the division of regulatory power between federal and state government.
C. Evidentiary Base
Cities contend that the Commission’s decision is defective because an opportunity for a “full hearing” was not provided Cities so that they might challenge Com Ed’s compliance filing and because the Commission’s price squeeze determination was not based upon substantial evidence. In support of their alleged right to a full hearing, which was not claimed until rehearing of the Commission’s acceptance of Com Ed’s compliance filing was sought, Cities rely upon section 205 of the Federal Power Act, 16 U.S.C. § 824d(e). We read the language of that provision as allowing, rather than requiring, the Commission to conduct full-scale evidentiary hearings. It provides, in relevant part, that “the Commission shall have authority ... to enter upon a hearing” regarding the legality of a rate increase. (Emphasis added.) Cities also suggest that once having entered into hearings, the Commission must follow formal adjudicatory procedures throughout. Cities cite no authority in support of that proposition. In opposition, we note that this court recently acknowledged the Commission’s authority to dispose of issues summarily if there is no need for a hearing. Public Service Company of New Mexico v. Federal Energy Regulatory Commission, 653 F.2d 681, 692 (D.C.Cir.1981); see also Appalachian Power Co. v. Federal Power Commission, 529 F.2d 342, 351 (D.C.Cir. 1976); Citizens For Allegan County, Inc. v. Federal Power Commission, 414 F.2d 1125, 1128 (D.C.Cir.1969). Here we find there was no need for more than a paper hearing to resolve Cities’ challenge to Com Ed’s compliance filing. Cities’ written objections were considered by the Commission and no need for a supplementary oral hearing was shown.
Cities objected to Com Ed’s compliance filing alleging that it did not rebut Cities’ prima facie price squeeze case because much of the retail data submitted was based upon estimates, not test year data. R. 5361A-68. Furthermore, Cities maintain that the estimates are unreliable because they were not based upon a valid statistical sample. Id. at 5366. The Commission found the data reliable because “[t]he load estimates . .. were based on sampling and class load estimating methodology routinely used by Com Ed for its own purposes and in class cost of service studies for retail rate proceedings before the Illinois Commission.” R. 5418 (Order Accepting Compliance Filing). Cities’ objections were, then, fully considered and rejected. Even with the acknowledged imperfections in the data submitted, that data was enough to overcome the prima facie case that inspired these long proceedings and other evidence submitted to reinforce Cities’ position and discredit Com Ed’s response. That is all that is required of a defending party.
CONCLUSION
For the foregoing reasons, we affirm all decisions of the Commission in this case except for its decision to investigate the old fuel adjustment clause under section 206. With respect to that decision we remand so that the Commission can determine whether this is an appropriate case for the exercise of its section 205 suspension and refund jurisdiction.
. The two rates were “substantially” identical from 1955 to 1968. The 1968 settlement was entered into after an informal complaint filed with the Commission by the cities of Batavia and Geneva in 1966. The 1966 complaint contended that there was no valid basis for the small differential between Rates 6 and 78. Docket IN-989, IN-991.
. The Federal Power Act empowers the Commission to regulate “the sale of electric energy at wholesale in interstate commerce,” 16 U.S.C. § 824. Retail sales are regulated by state commissions.
. A fuel adjustment clause is a device that allows utilities to pass on the increasing cost of fuel to customers without filing a new rate schedule. See n.21 infra.
. A definition of rate schedule is set out at n.25 infra.
. See pp. 75 76 infra.
. See pp. 80-83 (demand costs are allocated among classes of customers); compare n.7; pp. 83 -85 infra (demand billing).
. A ratchet is a device for billing individual customers. See pp. 83-85 infra.
.The Staff used a 12-CP method of demand allocation, rather than the 1-CP method used in the Company’s study.
The parties, both in hearings and in their briefs, have addressed certain of the concerns voiced by the Court in Cities of Batavia regarding the reasonableness of a retail class which blankets all industrials regardless of their individual peak characteristics. The court’s concern appears to be that group subsidies within the industrial class may work to the disadvantage of the wholesale class. This is a proper concern. It would be a matter to be remedied if that part of the industrial class for which the wholesale customer wished to compete were being subsidized by higher rates to that part of the industrial class for which the wholesale customer did not wish to compete. There is no evidence that this situation prevails in this case.
. R. 5361-5400. Specifically Cities contend that Com Ed’s cost of service study was invalid because (1) it used retail revenues for subsequent periods beyond the test period to obtain a return on equity comparable to that earned by resale revenues in the actual 1974 test year, and (2) it was based upon estimated coincident demand data for Com Ed’s 225,000 retail customers. See p. 91 infra.
. The Commission rejected Cities’ objection to the use of post-test retail rates and to the use of estimated wholesale and retail coincident peak demands:
The load estimates which underlie the compliance filing were based on sampling and class load estimating methodologies routinely used by the Com Ed for its own purposes and in class cost of service studies for retail rate proceedings before the Illinois Commission. Com Ed’s method of developing the estimates of retail coincident demand appears to be reasonable considering the data available.
R. 5418.
.The Commission stated:
We had before us the compliance filing, Cities’ objections to the compliance filing, Com Ed’s response to the objections along with testimony and affidavits of its witnesses. • Since the issue raised by the Cities would rise or fall on the validity of their statistical method, a hearing was not necessary to resolve such issues upon a finding that their statistical method was invalid.
R. 5490.
. The Commission’s jurisdictional challenge does not extend to our authority to determine whether there is an unjustified or illegal price squeeze. Final agency action on the price squeeze issue was not taken until February 2, 1981, R. 5489, and a petition for review of that action was timely filed with this court on March 6, 1981. The Commission did not conclude the price squeeze issue with Opinions Nos. 63 and 63-A because it determined that additional information on that issue was required. R. 5019-20. Only after the Commission accepted Com Ed’s compliance filing, R. 5416, and determined that there was no anti-competitive price discrimination, could Cities appeal the Commission’s final disposition of that issue. When Cities appealed the final agency action on the price squeeze issue, we consolidated that appeal with their January 16th appeal from Opinion No. 63-A, which raised price squeeze as well as non-price squeeze issues.
. In response to that petition, the Commission also argued that the price squeeze and tax normalization issues were not yet ripe for review. FERC Brief, September 11, 1980, at 20-25.
. Cities have cast substantial doubt upon the issue date of Opinion No. 63-A. According to the records of Cities’ counsel, Opinion No. 63-A was not delivered across town to Cities’ counsel until November 30, 1979, two weeks after the Commission contends the Opinion issued. See Affidavit of Carol Vodra, Cities Supp’l Brief, October 15, 1980, Appendix B. See also id. Appendix C. While Cities do not dispute that the Opinion was placed in inter-office mail on November 16, they question whether it was given to the United States Postal Service on that day. See Affidavits of Louis D. Cashell and Ray A. Huber, FERC Supp’l Brief, Appendices A and B; Cities Supp’l Brief, October 15, 1980 at 2-3. The Cashell Affidavit states that “[ejach time an opinion is mailed out, the words ‘service-issued’ are stamped on the card and the date of mailing and opinion number are written in.” The affidavit goes on to state that “[t]he ‘service cards’ .. . indicate that one copy of Opinion No. 63-A was mailed ... on November 16, 1979.” The affidavit does not explain whether “mailed” means placed in the United States mail or simply placed in inter-office mail. It does not dispose of the real possibility that the Opinion may not have been deposited in the United States Mail until, at best, the day after it was placed in inter-office mail. Cf. 18 C.F.R. § 1.17(d) (“The date of service shall be the day when the matter served is deposited in the United States mail . .. ”). Thus, on the basis of the record before us, we would be unable to determine whether Opinion No. 63-A did, in fact, issue on November 16. For the reasons below, we need not resolve that problem. We should note, however, that if we were confronted with that problem — i.e., if we had to decide whether the Opinion issued on November 16 for jurisdictional reasons — we might not dismiss issues so important to the price squeeze issue (which the Commission concedes is now legitimately raised) as well as to whether the wholesale rate was “just and reasonable” but for the price squeeze, unless the Commission, as movant, was able to resolve our doubts about the issue date of Opinion No. 63-A.
. While a petition from an agency order cannot be filed after the statutory period for filing has run, it may be that some of the issues that might have been raised in that appeal are so inextricably linked to a subsequent agency opinion on another aspect of the same case, that those issues may be raised in a timely appeal from the second opinion. For example, the demand allocation issue in this case is an important element not only of Opinion No. 63-A, but also of the agency’s price squeeze decision. Therefore, even if Cities’ appeal from Opinion No. 63-A were not timely filed, Cities might still have been able to raise the demand allocation issue in their timely appeal from the agency’s final action on the price squeeze issue.
. Section 313(b) reads in relevant part:
Any party to a proceeding under this Act aggrieved by an order issued by the Commission in such proceeding may obtain a review of such order in the United States Court of Appeals for any circuit wherein the licensee or public utility to which the order relates is located or has its principal place of business, or in the United States Court of Appeals for the District of Columbia, by filing in such court, within sixty days after the order of the *220Commission upon the application for rehearing, a written petition praying that the order of-the Commission be modified or set aside in whole or in part.
. Cf. Selco Supply Co. v. Environmental Protection Agency, 632 F.2d 863 (10th Cir. 1980) (Federal Insecticide, Fungicide and Rodenticide Act, 7 U.S.C. § 136n(b)); Microwave Communications, Inc. v. Federal Communications Commission, 515 F.2d 385, 389 (D.C.Cir. 1974) (the sixty day requirement of the Communications Act, 28 U.S.C. § 2344, is “jurisdictional and unalterable”).
. Emphasis added. See also 18 C.F.R. 1.13(b):
(b) Issuance of orders. In computing any period of time involving the date of the issuance of an order by the Commission, the day of issuance of an order shall be the day the Office of the Secretary mails or delivers copies of the order (full text) to the parties or their attorneys of record, or makes such copies public, whichever be the earlier. Orders will not be made public prior to the mailing or delivery to the parties or their attorneys of record, except where, in the judgment of the Commission, the public interest so requires. The day of issuance of an order may not be the day of its adoption by the Commission. In any event, the Office of the Secretary shall clearly indicate on each order the date of its issuance. At the time any intermediate initial or tentative decision becomes effective as a decision of the Commission in the absence of Commission review as provided for by §§ 1.30 and 1.31, the Secretary will issue and serve upon the parties of record an appropriate notice of the date such decision became effective as a Commission decision or order.
.The ALJ’s decision against Cities was upheld but on different grounds. The ALJ understood Cities’ argument to be based upon alleged imprudent judgment by Com Ed’s management in developing excessive reserve capacity. R. 4646. The Commission recognized that Cities’ position went further and would deny Com Ed recovery of even certain prudently incurred capital carrying charges. R. 4971.
. R. 4975. Cities’ use of the term “normalization” instead of “adjustment” caused the Commission to respond to an argument, not presented by Cities although perhaps supportable, that the rate base ought to have been normalized'to reallocate the cost of excess capacity to those who will actually benefit from it. Normalization is an accounting device by which expenses are distributed over a period of time in a different manner from that by which they are incurred. Cf. pp. 77-80 infra (tax normalization). As the Commission recognized, it is arguable that customers taking power at the beginning of a cycle are paying for capacity costs incurred for the benefit of customers at the end of the cycle and therefore an interperiod reallocation is appropriate. This is because:
In order to achieve economies of scale, utilities characteristically add capacity in large lumps. As a result, when a new facility first goes on line, the utility may well have a significant amount of spare capacity. This spare capacity is gradually consumed by load growth, so that at the end of the cycle the utility is maintaining the bare minimum reserves.
R. 4973. The Commission did not pursue the theory because it was neither suggested by Cities nor sustained by the record. To sustain such a theory, the Commission noted that:
The analysis and calculation would not be simple, however. Among the factors that might have to be taken into account are the following: the possibility of irregular growth in demand; a discrepancy between the length of the described expansion cycle and the period between rate applications; fuel expense savings accruing to the benefit of customers in the early years of the expansion cycle due to the “premature” addition of presumably efficient new capacity; the fact that capital carrying charges per unit of capacity do not stay constant due to depreciation; and some recognition of the fact that customers in the early years receive some of the benefits of the economies of scale that were a basis for the decision to add capacity in a large lump.
R. 4974 n.7.
. As we explained recently in Anaheim, et al. v. Federal Energy Regulatory Commission, 669 F.2d 799, 806 (D.C.Cir.1981) (hereinafter Anaheim), “[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 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.” See generally Public Service Co. of New Hampshire v. Federal Energy Regulatory Commission, 600 F.2d 944 (D.C.Cir.1979).
. Commonwealth Edison Co., 51 F.P.C. 86, 111, reh. denied, 51 F.P.C. 978 (1974). The fuel adjustment clause was uncontested in that proceeding.
. Section 205, 16 U.S.C. § 824d(e) provides:
Whenever any such new schedule is filed the Commission shall have authority ... to enter upon a hearing concerning the lawfulness of such rate, charge, classification, or service; and, pending such hearing and the decision thereon, the Commission . .. may suspend the operation of such schedule and defer the use of such rate, charge, classification, or service, but not for a longer period than five months beyond the time when it would otherwise go into effect; and after full hearings, either completed before or after the rate, charge, classification, or service goes into effect, the Commission may make such orders with reference thereto as would be proper in a proceeding initiated after it had become effective. If the proceeding has not been concluded and an order made at the expiration of such five months, the proposed change of rate, charge, classification, or service shall go into effect at the end of such period, but in case of a proposed increase rate or charge, the Commission may by order require the interested public utility or public utilities to keep accurate account in detail of all amounts received by reason of such increase, specifying by whom and in whose behalf such amounts are paid, and upon completion of the hearing and decision may by further order require such public utility or public utilities to refund, with interest, to the persons in whose behalf such amounts were paid, such portion of such increased rates or charges as by its decision shall be found not justified....
.Section 206, 16 U.S.C. § 824e(a) provides: Whenever the Commission, after a hearing had upon its own motion or upon complaint, shall find that any rate, charge, or classification, demanded, observed, charged, or collected by any public utility for any transmission or sale subject to the jurisdiction of the Commission, or that any rule, regulation, practice, or contact affecting such rate, charge, or classification is unjust, unreasonable, unduly discriminatory or preferential, the Commission shall determine the just and reasonable rate, charge, classification, rule, regulation, practice, or contract to be thereafter observed and in force, and shall fix the same by order.
. 18 C.F.R. § 35.2 defines rate schedule as “a statement of . .. electric service . . . rates and charges for or in connection with that service, and ... all classifications, practices, rules, regulations or contracts which in any manner affect or relate to the aforementioned service, rate, and charges.”
. Quoting St. Michaels Utilities Commission v. Eastern Shore Public Service Co., 31 F.P.C. 1161, 1164 (1964).
. Cf. Commonwealth Edison Co., 52 F.P.C. 1072, 1073-74 (1974) (“Because Com Ed’s fuel adjustment clause provides that the cost of fuel shall be determined in accordance with the Uniform System of Accounts prescribed by the Illinois Commerce Commission, the question arises as to the propriety of the treatment of costs in the provision presently in effect. We note further that Com Ed’s presently effective provision reflects total system losses rather than wholesale losses. These two issues which bear on the justness and reasonableness of Com Ed’s fuel adjustment provision require this Commission to institute an investigation under Section 206 of the Federal Power.”). We do not mean, however, to prejudge this case.
. See, e.g., J. Bonbright, Principles of Public Utility Rates 218-23 (1961), 1 A. Kahn, The Economics of Regulation 32-34 (1970).
. On May 6, 1981, the Commission issued Order No. 144 holding tax normalization of miscellaneous timing differences in electric rate-making appropriate. That order has been stayed pending reconsideration.
. Evidence was not submitted until December 23, 1975, R. 2370-76, which was after Order No. 530 but before Orders Nos. 503-A, 530-B and the Order Denying Rehearing of Order No. 530-B (which announced that the general policy allowing normalization would be prospective).
. The Commission specifically held:
Since these deductions represented timing differences between book and tax accounting, the refunds represented a deferral of income taxes and were credited to accumulated deferred income taxes. In its evidentiary presentation, Com Ed proposed adjustments to the test year to reflect changes in deductions for 1974 and to reflect changes in deferred taxes for 1974 and prior years. Since these changes adjusted the actual test year data to reflect the company intended tax deductions, it cannot be argued that the Law Judge acted inconsistently in disallowing post test year wage adjustments while accepting adjustments to correct current year tax accounts.
R. 4982. It is clear also that the ALJ treated this case as one where evidence was submitted to prove that normalization was appropriate. See R. at 4648-51.
. Although the Staff conceded that Com Ed satisfied Order No. 530-B, it recommended against normalization because the adjustment due under the normalization policy was offset elsewhere. See R. at 4649 (“the higher allowance for Account No. 282 deferred taxes as calculated by Staff more than offsets the adjustments in issue”):
. See n.29 supra', Notice of Proposed Rule-making, Docket No. RM-80-42 (March 21, 1980).
. See P. Garfield and W. Lovejoy, Public Utility Economics 56-83 (1964) (hereinafter Garfield- and Lovejoy). So as not to confuse, we should note that the rate base is, very simply, “the net (or depreciated) valuation of the public utility’s tangible property, comprising the plant and equipment used and useful in serving the public. (Such tangible property is called ‘plant in service’ in the uniform systems of accounts prescribed for public utilities.) In addition, the rate base includes an allowance for working capital and, depending on the circumstances, may also include amounts for the overhead costs of organizing the business, intangibles, and going-concern value.” Id. at 56.
. As will be seen more fully infra, the energy charge has two parts: (1) the basic energy rate; and (2) the fuel adjustment clause. The basic energy rate is a base cost or an estimate of what fuel will cost during a specific period. The fuel adjustment charge represents the difference between base and actual costs. See generally Anaheim at 14-15.
. Cost of service is not to be confused with the billing of service, which is considered infra. See generally Re Virginia Electric & Power Co., 9 P.U.R.3d 225, 239 (1953): “Cost is only'one of the many elements to be considered in determining the appropriateness of particular rates. The value of service is a matter of great importance in the establishment of individual rates.”
. Other factors to consider are “maintenance and overhaul schedules, scheduled and unscheduled outages, diversity interchanges of capacity and reliance upon interconnections for reserve capacity.” Opinion No. 768, Nevada Power Co., Docket No. E-8721 (July 7, 1976), slip op. at 8.
. R. 2456 (Edison Exhibit C2D); R. 4657. The monthly peaks during the four summer months were: June, 10,773 mw; July, 12,270 mw; August, 11,518 mw; September, 10,714 mw. The average of summer month peaks constituted 92.2% of the annual system peak. Use during the eight off-peak monthly peaks of the test year amounted to 78% of the power used during the four summer month peak period. R. 4658.
. R. 4657-58, citing Opinion No. 821, Southern California Edison Company, Docket No. E-8176 (September 22, 1977), slip op. at 18; Opinion No. 813, Louisiana Power & Light Co., Docket No. E-8615 (July 21, 1977), slip op. at 18, 20. In Southern California, the Commission used a 12-CP method where average monthly peaks to the annual peak was 87.8% and the ratio of the lowest monthly peak to the annual peak was 79.1%. In Louisiana Power, the Commission used a 4-CP method where average monthly peaks to the annual peak was 62% and the ratio of lowest monthly peak to the annual peak was 56%.
.R. 4658. In addition the ALJ found:
Com Ed has 25 high voltage interconnections which materially reduce the amount of peak period reserve capacity by approximately two-thirds of what would be necessary without such connections (Tr. 450). These interconnection arrangements with other utilities require Com Ed to pay carrying charges throughout the year (Tr. 449) ano these charges are assessed against all customers on an annual basis.
Com Ed purchases capacity from Commonwealth Edison Company of Indiana, a subsidiary, and from the Ludington Pumped Storage facility (Ex. C1B, p. 1). Although Com Ed claims that these purchases are made solely to add capacity for the summer peak, it conceded that since the carrying costs were fixed “the Company uses energy from [this] capacity .. . whenever it is good system operation to do so.” (Ex. C1B, p. 2). Com Ed also has an agreement with Indiana and Michigan Power for purchase of 500 mw during the summer months, and a 200 mw diversity interchange in which Com Ed has arranged to relieve itself of excess capacity during the winter months, in exchange for power during the summer months.
*229Com Ed projects its peak loads on a month-by-month basis in order to assure that there will be sufficient capacity for load and maintenance on a year-round basis (Ex. C4, Tr. 456). It schedules its maintenance during non-peak months in order to have generating units available during the peak months. Com Ed’s power planning engineer testified on cross-examination that Com Ed’s 3% “net reserve” (reserve less interchanges, maintenance, and forced outages) on the peak day in 1974 remains in generally the same range throughout the 12 monthly peaks (Tr. 465-466).
Com Ed submitted a projection of its electric power reserves for the period of 1976-1985. This projection shows that the reserve power margin and surplus is expected to be substantially lower in July and August than in the other months (Ex. C1.3-C1.12). This forecast of future years conflicts with Com Ed’s actual experience in 1974 in which the surplus of available power in relation to power used remained in generally the same range on a year-round basis.
R. 4658-60. See n.37 supra.
. See n.40 supra.
. There is a dispute between the parties as to whether this is a “true ratchet.” Ordinarily, a ratchet simply sets a floor under maximum demand for billing purposes. Garfield and Lovejoy at 157. When set at 100%, it sets both the floor and the ceiling, i.e., it sets the demand cost. The debate has no substantive consequence and, although we adopt the term “ratchet” for present purposes, we do not presume to settle this semantic dispute for all time.
. The billing demand provision reads:
The billing demand (except for partial requirements customers) shall be the highest maximum demand established in any one of the four preceding summer months, including the billing month if it is a summer month.
R. 3278.
. Cities maintain that the purpose behind the ratchet will be unfulfilled. The ratchet, Cities argue, will not invite conservation but waste: “The ratchet imposes on the wholesale customer a fixed demand charge no matter what the customer’s actual usage is.” Initial Brief of Cities at 48. But as Com Ed notes, the ratchet is not designed solely to promote individual “conservation.” Rather, the purpose of the ratchet is also to promote system efficiency, i.e., to eliminate off-peak excess capacity or at least to charge those responsible for it. Brief of Com Ed in No. 80-1072 at 42. Cities’ response that they will be unable to affect consumption patterns by their customers is beside the point. See Reply Brief of Cities at 19.
.The disproportional seasonal demand of individual customers would not cause problems if that disproportional use were spread evenly throughout the year. In such a case, the system would, in fact, experience no periodic excess capacity. However, when disproportional impact is seasonal, as here, efficiency can be encouraged by billing those customers responsible for the seasonal variation.
. R. 2255 (“The load of the municipalities coincident with the Edison system peak was 47% higher in 1973 than in 1970, while the Edison peak itself grew only 24.3% over that period.”).
. At oral argument, Com Ed suggested for the first time that the 75%-23 month ratchet did not affect a single bill because the 100% ratchet was controlling. If that is the case, it is unclear to us why both ratchets were sought in the first place and how if both had been allowed they would have operated together.
. The plain language of the statute, 16 U.S.C. § 824d(e) (The Commission “may . .. require such public utility or public utilities to refund such portions of such increased rates or charges as by its decision shall be found not justified”) (emphasis added), and a recent decision of this court, Belco v. Federal Energy Regulatory Commission, 589 F.2d 680 (D.C.Cir. 1978), (Commission discretion to order and compute refund), also dictate affirmance of the Commission’s decision.
.The Staff recommended a 12% rate of return.
. Hjelmfelt, A Price Squeeze Theory For Implementation of Federal Power Commission v. Conway, 50 U.Colo.L.Rev. 459 (1979), notes that “a ‘price squeeze’ occurs when a wholesale supplier, who also sells at retail, charges such high rates to its wholesale customers that they cannot compete with the suppliers’ retail rates.” See also M. Peck, Competition In The Aluminum Industry 1945-1958, at 73 (1961); G. Hale & R. Hale, Market Power Size and Shape Under the Sherman Act 32 (1958); Meeks, Concentration In The Electric Power Industry: The Impact of Antitrust Policy, 72 Colum.L.Rev. 64, 119 (1972); Note, The Application of the Antitrust Laws To Price Squeezes in The Electric Utility Industry, 54 St. John’s L.Rev. 103 (1979).
. 18 C.F.R. § 2.17(e) provides:
The burden of proof (i.e. the risk of nonpersuasion) to rebut the allegations of price squeeze and to justify the proposed rates are on the utility proposing the rates under section 205(e) of the Federal Power Act.
See also Opinion No. 62, Southern California Edison Co., FERC Docket No. ER 76-205 (August 22, 1979), slip op. at 23-24, aff’d sub nom. Anaheim (price squeeze issue not resolved on appeal):
[I]t is clear that a judgment that a price squeeze exists follows a three-part analysis. First, we must determine whether there is price discrimination. Second, we must determine whether the price discrimination has had any anticompetitive effects. If the answer to both of these inquiries is yes, a price squeeze is said to exist and a presumption arises that the rates are unduly discriminato*234ry. We would then determine whether there were any countervailing public policy or factual circumstances which merit a finding that the discrimination is not undue. Absent such a finding, the determination of a proper remedy would be required.
(Footnotes omitted.)
. Further, neither Conway nor Batavia require the Commission to set a wholesale rate so that wholesale customers are guaranteed the ability to compete in the retail market. It may be, for instance, that the state commission is not allowing an adequate rate of return. If so, the Commission is not obliged to follow, and indeed may not follow suit, for the rates it sets must fall within the range of reasonableness. See Bethany, as modified on rehearing at 200.
. See R. 4339 (Order Reopening Record); R. 5008 (Opinion No. 63). All parties to this proceeding, including Cities, agree with this market definition. See Cities Brief, April 27, 1981, at 21-22, 24. See also R. 5278 (Opinion No. 63-A) setting out Cities’ argument, step one of which was “there is across-the-board competition between Com Ed and Cities, comprehending franchise competition and competition for every retail class and subclass .. . ”).
.The Commission should also be careful to insure that wholesale costs are not artificially high because of utility inefficiency (intended or unintended) in the wholesale market. Inefficiency in the wholesale market accompanied by efficiency in the retail market could naturally work to the disadvantage of wholesale customer/retail competitors.