Full opinion text
BEA ANN SMITH, Justice. Texas Utilities Electric Company, the Public Utility Commission, the Office of Public Utility Counsel, and the Cities of Arlington, et al. appeal from a district-court judgment rendered in a suit for judicial review of the Commission’s final order in an electric utility rate case conducted under the Public Utility Regulatory Act (PURA), Tex.Rev.Civ.Stat.' Ann. art. 1446c (West Supp.1994). The district-court judgment reverses and remands certain aspects of the Commission’s final order, and affirms the remainder. We will reverse the district-court judgment and remand the cause to the district court with instructions that the cause be remanded to the Commission for further proceedings consistent with our opinion. See Administrative Procedure Act (APA), Tex. Gov’t Code Ann. § 2001.174 (West 1994). THE CONTROVERSY Texas Utilities filed its application for a rate increase in January 1990 seeking to include in its rate base costs associated with Comanche Peak, a newly constructed nuclear power plant. The utility sought an agency adjudication regarding what portion of its costs it could include in its rate base as being a “prudent” investment, public interest findings on its reacquisition of a 12.2 percent ownership interest in the plant, final reconciliation of its fuel costs and revenues for the period April 1983 to June 1989, and a reduction of its fuel factor for the period May 1990 to April 1991. After the Commission issued its order, motions for rehearing were filed and the Commission issued a second order on rehearing. Subsequent motions for rehearing were overruled by operation of law, and five parties to the rate-making proceeding filed suit for judicial review in district court., See PURA § 69; APA § 2001.171. The district court affirmed the Commission order in part and reversed it in part, after which Texas Utilities, Public Utility Counsel, the Cities, and the Commission each appealed the district-court judgment. For clarity, we will provide additional facts germane to the various points of error throughout the opinion. CONFLICT OF INTEREST In their first point of error, the Cities and Public Utility Counsel argue that the chairman of the Commission, Paul Meek, was biased because he had a pecuniary interest in the outcome of the proceedings, and because he was prejudiced in favor of the gas industry. The allegations of impermissible bias center around Meek’s ties with American Petrofína (“Fina”). During the rate-making proceedings, Meek served as chairman of Fina’s board, received retirement benefits from Fina, and held shares of its publicly traded common stock. Fina’s direct sales of natural gas to Texas Utilities from 1989 to 1991 totalled $60,782; indirect revenue from sales to other Texas Utilities suppliers approximated $104 million. Because of his connections with Fina, the Cities and Public Utility Counsel claim that Meek’s participation in the hearings precluded the Commission from making impartial findings. The district court found the evidence insufficient to show that Meek’s service on the Commission led to unfair proceedings or prejudiced substantial rights of the parties. We agree. PURA provides that no commissioner may, during a period of service with the Commission, “have any pecuniary interest ... in any person or corporation or other business entity a significant portion of whose business consists of furnishing goods or services to public utilities or affiliated interests....” PURA § 6(b)(1). It is grounds for removal from the Commission if a member has interests in violation of section 6(b) at the time of his or her appointment. PURA § 6A. However, “the validity of an action of the commission is not affected by the fact that it was taken when a ground for removal of a member of the commission existed.” PURA § 6A(b). Meek resigned from the Commission effective April 20, 1992, after the Attorney General requested that he either sever all ties with Fina or resign from the Commission. Although Meek was not removed from the Commission because of a conflict of interest pursuant to PURA section 6A, he did resign in the face of a perceived conflict. Meek’s conflict, however, has no effect on the Commission’s order in Docket 9800. PURA § 6A(b). This Court is left, therefore, with the power to reverse and remand the Commission’s order only if Meek’s participation resulted in an order that prejudices substantial rights of the appellants. See APA § 2001.174. We understand appellants to contend that this Court should reverse the Commission’s order because Meek’s interests in Fina resulted in an order that is arbitrary and capricious and a violation of their constitutional right to a fair and impartial hearing. In order to prevail, appellants must overcome the presumption that agency members are persons of conscience and intellectual discipline, capable of judging a particular controversy fairly on the basis of its own circumstances. United States v. Morgan, 313 U.S. 409, 421, 61 S.Ct. 999, 1004, 85 L.Ed. 1429 (1941). Following the United States Supreme Court, we recognize a presumption of honesty and integrity in those serving as adjudicators. Withrow v. Larkin, 421 U.S. 35, 47, 95 S.Ct. 1456, 1464, 43 L.Ed.2d 712 (1975). One may overcome this presumption by demonstrating that the deci-sionmaker’s mind is “irrevocably closed” on the matters at issue. Federal Trade Comm’n v. Cement Inst. 333 U.S. 683, 701, 68 S.Ct. 793, 803-04, 92 L.Ed. 1010 (1948). During confirmation hearings conducted in May 1990, the Texas Senate fully explored the issue of Meek’s conflict. At that time, aware of Meek’s connections with Fina, the Senate satisfied itself that Meek could execute his duties as commissioner impartially and without prejudice in favor of the gas industry. Additionally, Meek promised to recuse himself from voting on any contested issue regarding contracts between public utilities and Fina, a promise he upheld by not reviewing contracts between Texas Utilities and Fina. It is well established that absent a showing of incapability to decide a particular controversy fairly, an administrative officer is not disqualified simply because he or she has previously taken a position, even in public, on a policy issue related to a particular dispute. Morgan, 313 U.S. at 421, 61 S.Ct. at 1004. In Morgan, the Supreme Court held that the Secretary of Agriculture’s strong views on a particular issue did not make him unfit to exercise his duties in administrative proceedings relating to those matters. Id. Similarly, in Cement Institute the Court held that members of the Federal Trade Commission were not disqualified from participating in adjudicatory proceedings simply because they had previously expressed their opinions that a pricing system at issue in the proceeding was illegal. Cement Institute, 333 U.S. at 700-01, 68 S.Ct. at 803-04. In this appeal, the Cities and Public Utility Counsel question Meek’s impartiality because of a newspaper interview in which he expressed his disappointment with the Commission’s decision to disallow $1.3 billion of Comanche Peak costs. The Supreme Court has decided, however, that public criticism “is a practice familiar in the long history of ... litigation,” and that while an administrator may have an underlying philosophy in approaching a specific ease, he or she may still be assumed to be a person of conscience and intelléctual discipline, capable of judging a particular controversy fairly. Morgan, 313 U.S. at 421, 61 S.Ct. at 1004. The Cities and Public Utility Counsel argue that this order should be invalidated, relying on American Cyanamid Co. v. Federal Trade Commission, 363 F.2d 757 (6th Cir.1966). In American Cyanamid, the court invalidated a commission order because one of the commissioners had previously served as counsel for a Senate subcommittee investigating many of the same facts and issues that later came before the commission. The court found that the commissioner’s dual investigative and adjudicative experiences with the issues involved in the hearing created a risk that commission decisions might be based on evidence outside the record. It was the presentation of nonrecord evidence, not the commissioner’s personal viewpoints, that led the court to invalidate the order. American Cyanamid, 363 F.2d at 767. In this case, however, appellants base their request for invalidation of the order on assertions that Meek’s personal views about the gas industry made it impossible for him to decide the issues fairly. Under the circumstances of this proceeding, we cannot agree. We do not express any opinion regarding whether Meek should have been removed from the Commission had he not resigned. This Court is limited to the judicial review enumerated in APA section 2001.174. We conclude that Meek’s involvement with Fina and his opinions about the gas industry have not been shown by the complaining parties to have resulted in a deprivation of the right to an impartial and fair hearing before the Commission, nor has it been shown that he exhibited bias such that his votes were necessarily arbitrary and capricious. The Cities and Public Utility Counsel’s first point of error is overruled. REACQUISITION OF MINORITY INTERESTS All appellants bring points of error related to the district court’s disposition of the Commission order disallowing more than $908 million spent to repurchase 12.2 percent of Comanche Peak from minority interest owners and to settle litigation arising from the joint ownership of the project. Section 63 of PURA permits the Commission to disallow certain expenses associated with transactions involving changes in public utility ownership. The Commission’s authority to make disallo-wances under section 63 is limited to three specific types of transactions: (1) the acquisition, sale or lease of any plant as an operating unit in the state of Texas for a total consideration in excess of $100,000; (2) a public utility’s merger or consolidation with another public utility operating in the state; and (3) the sale of fifty percent or more of a public utility’s stock. When any one of these transactions takes place, the utility must file a report with the Commission, which then investigates the transaction to determine whether it is in the public interest. In making this determination, the Commission is to consider the reasonable value of the property, facilities or securities involved. If the Commission finds that the transaction was not in the public interest, it must “disallow the effect of such transaction if it will unreasonably affect rates or service.” PURA § 63. In reviewing the costs associated with the construction of Comanche Peak, the Commission exercised its authority under section 63 to make a disallowance of $908,688,938. The Commission asserted that it had jurisdiction to make disallowances pursuant to section 63 because Texas Utilities’ repurchase of certain minority interests in the Comanche Peak project constituted the purchase of a plant or unit as an operating system for consideration in excess of $100,000. Texas Utilities’ second motion for rehearing filed with the Commission included an assignment of error stating: The Commission erred in concluding that PURA § 63 controls this Commission’s review of [Texas Utilities’] reaequisition of minority owner interests in Comanche Peak, for the reason that, as a matter of law, PURA § 63 does not apply to the transfer between joint owners of partial, undivided interests in a plant and does not apply to a plant under construction that is not operating. When this second motion for rehearing was overruled by operation of law, Texas Utilities sought review in the district court, and con-turned to maintain that the Commission had improperly applied section 63 to the repurchase of minority interests in the project. As part of its appeal to this Court, Texas Utilities contends in its second point of error that the Commission’s section 63 review was an error of law. The Cities, Public Utility Counsel, and the Commission each argue that Texas Utilities has waived its right to challenge the Commission’s decision to proceed under section 63 because it was the utility that initially identified section 63 as one of the provisions giving the Commission jurisdiction over the rate-making proceeding. Administrative agencies, however, have only those powers that are expressly conferred by statute, together with those necessarily implied from the authority conferred or the duties imposed. State v. Jackson, 376 S.W.2d 341, 344 (Tex.1964) (citing Stauffer v. City of San Antonio, 162 Tex. 13, 344 S.W.2d 158, 160 (1961)); Sexton v. Mount Olivet Cemetery Ass’n, 720 S.W.2d 129, 142 (Tex.App.—Austin 1986, writ ref d n.r.e.). Jurisdiction cannot be conferred upon the agency by the parties before it, but rather must emanate from the statute itself. See Nueces County Water Control & Improvement Disk v. Texas Water Rights Comm’n, 481 S.W.2d 924, 929 (Tex.Civ.App. — Austin 1972, writ ref d n.r.e.) (“If the statutes do not grant the board the power to do a thing, then it has no such power.”). If the utility’s reaequisition of minority interests in Comanche Peak is not one of the specific transactions identified in section 63 of PURA, the Commission has no jurisdiction to make disallowances based on the standards set forth in that section; such jurisdiction cannot be conferred on the Commission simply because the parties have requested or agreed to it. This Court has the power, as well as the duty, to review the agency’s interpretation and application of a statute. See Railroad Comm’n v. Lone Star Gas Co., 599 S.W.2d 659, 662 (Tex.Civ.App.—Austin 1980, writ refd n.r.e.) (stating that an agency’s duty is to carry forward the directives of statutes, and the courts review agency orders to ensure that statutes are enforced). In reviewing the Commission’s order, we are therefore obliged to determine whether the repurchase of minority ownership interests is a transaction contemplated by section 63 of PURA. If it is not, the Commission had no authority to conduct a section 63 review, and we may not uphold that portion of the order. Accordingly, we first examine the repurchase at issue in this case to determine if it falls within the scope of transactions the Commission is directed to review under PURA section 63. In August 1973, Texas Utilities’ corporate predecessors, Dallas Power & Light Company, Texas Power & Light Company, and Texas Electric Service Company, signed a memorandum of agreement to design, construct, and operate the Comanche Peak nuclear power plant. Texas Utilities originally intended to own the entire plant, but was required to sell ownership interests in the project in order to receive construction permits from the Nuclear Regulatory Commission (NRC). In 1974, Texas Utilities agreed to allow participation in the ownership of Comanche Peak, thereby eliminating antitrust concerns associated with the issuance of the construction permits. By 1979, Texas Municipal Power Agency and Brazos Electrical Power Cooperative had acquired ownership shares of 6.2 percent and 3.8 percent respectively. In 1982, Tex-La Electric Cooperative of Texas became another co-owner of the Comanche Peak project. Because Tex-La had raised antitrust issues with the Department of Justice and had filed a petition to intervene in the Comanche Peak antitrust review related to its application for an operating license, Texas Utilities agreed to sell Tex-La a 4.3 percent interest in the project. Before the closing, however, Tex-La reduced its purchase to 2.2 percent of the project. The joint operating agreement was amended to reflect this sale. The Commission granted certificates of public convenience and necessity for all three sales of ownership interests in the project. The joint ownership agreement began to deteriorate over time. In May 1985, Brazos Electrical Power Cooperative ceased making its contractual payments to Texas Utilities. In early 1985, Tex-La Electric Cooperative made several late payments, and thereafter stopped making payments altogether. Texas Municipal Power Agency continued to make payments, but it made them under protest. Thereafter, the minority interest owners claimed that Texas Utilities had failed to meet its responsibilities under the joint ownership agreement, resulting in rising costs, schedule delays, and licensing problems. The three minority interest owners contended that they were therefore relieved of any obligation to pay their percentage costs of the construction and operation of the project. Texas Utilities sued for breach of contract, seeking monetary damages and a declaratory judgment affirming the minority interest owners’ continuing obligation to pay their share of the plant’s remaining costs. The minority interest owners filed counterclaims alleging mismanagement of the project, breach of contract, and deceptive trade practices. Faced with mounting litigation costs, Texas Utilities settled with the minority interest owners by repurchasing their undivided interests in the project. The settlement agreements ended all litigation between Texas Utilities and the minority interest owners. The repurchases were approved by the Commission which, as previously noted, indicated its intention to review the repurchase of these minority interests under PURA section 63 in the future rate-making proceedings. As part of Docket 9300, the Commission did in fact conduct the section 63 review. The Commission determined that the repurchase was in the public interest “to the extent that [Texas Utilities] paid a reasonable value for the repurchased capacity.” The Commission found that the utility had reacquired the minority interests by paying $4,765 per kilowatt — the cost of building Comanche Peak. By contrast, the Commission decided that a “reasonable value” would be $1,865 per kilowatt, the cost of building a stand-alone “generic coal plant” with 12.2 percent of Comanche Peak’s capacity. As a result, the Commission determined that Texas Utilities had paid $2,900 more per kilowatt than was “reasonable.” Accordingly, the Commission disallowed the excess purchase price amounting to almost $812 million. The Commission also disallowed the utility’s reimbursement of the minority interest owners’ litigation costs, amounting to $72,684 million, and $24,662 million of the total consideration paid for the nuclear fuel. The district court concluded that although review under section 63 of PURA was appropriate, the Commission made disallowances that were arbitrary and capricious and not supported by substantial evidence. In two jointly raised points of error, the Cities and Public Utility Counsel assert that the district court erred in remanding some of the Commission’s findings of fact and that the Commission properly carried out its section 63 review. They do not challenge the propriety of the section 63 review. The Commission also brings two separate points of error relating to its section 63 review, contending that it properly applied section 63 and that its findings of fact were supported by substantial evidence. We do not address these points of error because our conclusion that the repurchase of the undivided minority interests in the plant are not transactions .reviewable under section 63 renders moot any farther controversy about what would constitute a proper disallowance under that provision. As previously noted, section 63 applies to three types of transactions: (1) the purchase, sale or lease of a plant or unit as an operating system for consideration in excess of $100,000; (2) sales of more than fifty percent of the stock of a public utility; and (3) a merger or consolidation of two public utilities. Texas Utilities’ repurchase of the undivided ownership interests sold to Texas Municipal Power Agency, Brazos Electrical Power Cooperative, and Tex-La Electric Cooperative falls into none of these categories. Rather than repurchasing a “plant or unit,” Texas Utilities acquired the undivided ownership interests of three tenants-in-eommon. Under the joint ownership agreement, the co-tenants had waived any right to partition the interests, thereby foreclosing the possibility of identifying any part of the plant as belonging specifically to any co-tenant. The fallacy in the Commission’s analysis is its assumption that the minority interests translate into a complete and independently operable portion of Comanche Peak, ownership of which changed hands when the repurchase took place. The Cities and Public Utility Counsel argue that excluding the repurchase of the undivided interests from the scope of a section 63 review renders the provision meaningless. They contend that it is illogical to conclude that “a statute concerned with transactions of at least $100,000 would not apply to a transaction 1,000 times greater than that amount.” This argument fails because the element that triggers section 63 review is not the amount of money involved in the transaction, although the legislature has set a $100,000 minimum presumably to exclude transactions so small that there is no real risk they will unreasonably affect rates or service. Rather, section 63 is concerned with certain types of transactions that result in changes of ownership of the utility or its operating units to ensure that the costs of transactions inconsistent with the public interest are not assessed against the ratepayers. We conclude that the Commission erred in reviewing the costs associated with the minority interests under PURA section 63. In its final judgment, the district court reversed and remanded for reconsideration on the existing record the following specific findings of fact related to the minority interest repurchases: 149. For the reasons discussed in Section VII.C.2. of this Report, [Texas Utilities] failed to prove that the consideration it paid for the repurchased 12.2 percent interest in the plant was reasonable. 150. For the reasons discussed in Section VII.C. of this Report, [Texas Utilities’] repurchases of the minority owners’ interests in Comanche Peak are consistent with the public interest to the extent that [Texas Utilities] paid a reasonable value for the repurchased capacity. 151. For the reasons discussed in Section VII.C. of this Report, all amounts in excess of the reasonable value of the repurchased interests should be disallowed from invested capital as unreasonably affecting rates. 152. For the reasons discussed in Section VU.B.2.d. and Section VII.D. of this Report, a reasonable value of the repurchased interests in Comanche Peak is $1,856 per kW. 153. For the reasons discussed in Section VII.D. of this Report, the reasonable value of $1,856 per kW should apply to valuating the repurchased interests in Unit 1 and Unit 2. 158A. Consistent with an estimated fuel cost for Comanche Peak of $11 billion, the test-year-end cost of $5,938 billion should be used to value the repurchased 12.2 percent interest in Unit 1 and an estimated cost of $5.0 billion should be used to value the repurchased 12.2 percent interest in Unit 2. 153B. The plant disallowances related to the repurchased 12.2 percent interest in Unit 1 is $462,764,691; the plant disal-lowance related to the repurchased 12.2 percent interest in Unit 2 is $348,578,-247. Taken together, the total plant disallowance related to the repurchased 12.2 percent interest in the entire plant is $811,342,938. 154.For the reasons discussed in Section VII.E. of this Report, the $72,684 million in minority owners’ litigation expenses reimbursed by [Texas Utilities] as part of the settlement agreements should be disallowed. ⅝ ⅜ ⅜ ⅝ ⅜ ⅜ 156. As modified by Findings of Fact 153A and 153B, Section VII.F. of this Report indicates the disallowances for Unit 1 and Unit 2, as calculated in Section VI. (Prudence) and Section VII. (Reacquisition of Minority Owner Interests). The total Unit 1 disallowance is $847,004,966; the total Unit 2 disallowance is $534,139,597. Taken together, the total disallowance is $1,381,144,563. The purpose of remanding these findings was to allow the Commission to reconsider the “reasonable value” it assigned the repurchased interests, presumably to make an upward adjustment in its $1,856 per kilowatt valuation to reflect the “intangible” benefits of repurchasing the minority interests. The district court instructed the Commission to consider not only the “economic value” of the property and facilities acquired, but also benefits gained from terminating expensive and time-consuming litigation that jeopardized the entire project. We affirm the district court’s rejection of these findings of fact based on our conclusion that the Commission erroneously reviewed the repurchases under PURA section 63 and failed to evaluate the repurchase price in light of the relevant statutory considerations. We reverse that portion of the district court’s judgment affirming the Commission’s disallowance of $24,662,000 of the cost to Texas Utilities of repurchasing nuclear fuel from the minority interest owners. This payment was part of the overall settlement cost and should be reviewed under the prudent investment standard along with all other costs related to the repurchase. The Commission has already approved the utility’s decision to settle the dispute with the minority interest owners; on remand, we direct the Commission to consider, under the prudent investment standard, the price paid for the repurchase, including the litigation costs and repurchase of nuclear fuel at its original cost. FEDERAL INCOME TAX EXPENSE In points of error seven through ten, the Cities and Public Utility Counsel complain that the district court erred in affirming the Commission’s calculation of the utility’s federal income tax expense. They contend that the Commission’s calculation (1) improperly employed the hypothetical rather than the actual-tax method, (2) failed to account for tax savings resulting from the utility’s consolidated tax return, (3) did not reflect deductions for actual interest expense, and (4) failed to reflect deductions taken for below-the-line expenses, including disallowed Comanche Peak plant costs. We sustain the seventh point of error complaining of the Commission’s use of the hypothetical tax method. The mandate from the supreme court is clear: “The utility’s rates must reflect the tax liability actually incurred.” Public Util. Comm’n v. Houston Lighting & Power Co., 748 S.W.2d 439, 442 (Tex.1987). This Court has repeatedly affirmed that statement by consistently requiring the Commission to employ the actual-taxes-paid doctrine. See City of Alvin v. Public Util. Comm’n, 876 S.W.2d 346, 359-60 (Tex.App.—Austin 1994, no writ h.); Cities of Abilene v. Public Util. Comm’n, 854 S.W.2d 932, 944 (Tex.App.—Austin 1993, writ requested); Public Util. Comm’n v. GTE-SW, 833 S.W.2d 153, 159 (Tex.App.—Austin 1992, writ granted). Furthermore, under the actual-taxes-paid test, “any utility tax savings must benefit ratepayers.” Cities of Abilene, 854 S.W.2d at 945 (emphasis added). In this case, as well, we reject the Commission’s refusal to adhere to binding precedent. The Cities and Public Utility Counsel’s eighth point of error asserts that the Commission erred when it failed to adjust its calculation of the utility’s tax expense to reflect savings that resulted from the utility’s filing a consolidated tax return. The Commission rejoins that its decision not to allocate any of the savings to the utility was consistent with PURA section 41(c)(2) and eases construing that statutory provision. Section 41(c)(2) states: If the public utility is a member of an affiliated group that is eligible to file a consolidated income tax return, and if it is advantageous to the public utility to do so, income taxes shall be computed as though a consolidated return had been so filed and the utility had realized its fair share of the savings resulting from the consolidated return, unless it is shown to the satisfaction of the regulatory authority that it was reasonable to choose not to consolidate returns. Texas Utilities argues that this statute only applies when the utility has not filed a consolidated return. We disagree. The statute provides that, regardless of whether the utility actually filed a consolidated return, the Commission must calculate the utility’s income tax expense as though it had received any tax benefits a consolidated return would provide. Once the Commission determines that a consolidated filing would have been, or was, advantageous to the utility, the Commission must adjust the utility’s tax expense to reflect those savings. If the Commission does not reduce the utility’s tax expense to reflect the utility’s tax savings, it violates the actual-tax doctrine’s underlying principle that rates must be set based on the utility’s actual tax liability. GTE-SW, 833 S.W.2d at 166. The Commission argues that it was not required to allocate any of the tax savings from the consolidated filing to the utility because it specifically found that the consolidated filing was not advantageous to the utility. See Finding of Fact 331A. In Cities of Abilene we held that no adjustment to income tax expense is necessary under PURA section 41(c)(2) if the Commission finds either (1) that it was not advantageous to the utility to consolidate returns, or (2) that the Commission has computed taxes as though a consolidated return were filed and the utility has received its fair share of the savings from the consolidated return. Cities of Abilene, 854 S.W.2d at 944. In this case, the Commission relied on its own conclusion that the utility’s fair share of the savings was zero to support its finding that the consolidated return was not advantageous to the utility. We will uphold the Commission’s decision only if it properly found that the utility’s fair share of the tax savings was zero. Finding of fact 331D states: The federal income tax savings resulting from the filing of a consolidated federal income tax return should accrue to the entity that provided the tax attributes that allowed for such savings, and [Texas Utilities] was not the entity that provided such tax attributes. This Court has previously decided that even when it is the utility’s affiliates that have suffered losses and provided “the tax attributes that allowed for savings,” those savings must be passed on to the ratepayers. GTE-SW, 833 S.W.2d at 167. In finding of fact 331F, the Commission asserts that it would be unfair to allocate to the utility tax savings resulting from the affiliates’ losses because the utility will never be responsible for paying the affiliates’ taxes when “timing differences reverse and those affiliates have taxable income.” Again, this Court has rejected that argument. GTE-SW, 833 S.W.2d at 167 n. 16 (inequity resulting from ratepayers’ benefitting from tax savings not offset by obligation to pay higher rates in the event of affiliates’ gains is a matter for the legislature to remedy by amending PURA section 41(c)(2)). Similarly, finding of fact 331H, that Texas Utilities should not benefit from tax savings attributed to affiliates because it bears none of the risks associated with those entities, conflicts with existing caselaw. The Commission’s finding that the consolidated tax return was not advantageous cannot rest upon its own improper refusal to allocate any savings to the utility. Having rejected several of the findings supporting the Commission’s conclusion that the utility’s fair share of the tax savings is zero, we are unable to uphold that conclusion. There is no indication that each finding is independently sufficient to support the conclusion. We therefore sustain the Cities and Public Utility Counsel’s eighth point of error. The ninth point of error objects to the Commission’s failure to adjust the tax expense calculation to reflect actual-interest-expense deductions. The Commission is required to allocate tax savings to ratepayers rather than to shareholders. The actual-tax doctrine requires that the ratepayers be held accountable only for “those tax expenses that are actually incurred by a utility.” Houston Lighting & Power, 748 S.W.2d at 442. If the utility enjoys a tax deduction based on interest expense, the benefits of that deduction must be passed on to the ratepayers. In City of Alvin, however, we rejected the argument that the Commission must pass on immediately the entire savings related to a utility’s tax deductions. City of Alvin, No. 3-92-459-CV, slip op. at 18 (“Section 27(e) of PURA directs the Commission to distribute [tax savings benefits] to all ratepayers, however, both present and future. We will not interpret Houston Lighting as mandating that present ratepayers receive all the benefits of accelerated depreciation.”). We sustain the point of error to the extent that we continue to require the Commission to pass through to ratepayers any tax benefits from interest expense deductions. However, the Commission must allocate those savings between present and future ratepayers, and the proper allocation is within the Commission’s discretion. The Cities and Public Utility Counsel’s tenth point of error contends that the Commission erroneously excluded tax benefits resulting from below-the-line expenses, including tax deductions related to expenses disallowed as imprudently incurred. This Court has already decided that PURA requires that the Commission reduce the utility’s income tax expense by the amount of tax deductions, even if they are associated with disallowed capital expenses. City of Alvin, No. 3-92-459-CV, slip op. at 17 (citing GTE-SW, 833 S.W.2d at 169). We remain unpersuaded by the Commission’s argument that the actual-tax doctrine conflicts with the normalization rules. See City of Alvin, No. 3-92-459-CV, slip op. at 18. We sustain the tenth point of error. BONDED RATES In their twenty-first point of error, the Cities challenge the Commission’s authority to allow Texas Utilities to implement bonded rates in both the municipal and non-municipal sections of its service area. Disposition of this point of error requires an interpretation of PURA section 43(e). This appeal presents the first opportunity for this Court to consider the bonded-rate provision of the statute since its amendment in 1983. When an electric utility wishes to change its rates it must follow the procedures outlined in PURA section 43. The utility initiates rate proceedings by filing a statement of intent to change rates with the regulatory authority having original jurisdiction. PURA § 43(a). In all proceedings involving major rate changes, the regulatory authority having original jurisdiction must hold a hearing on the proposed rate schedule. PURA § 43(c). Pending the hearing, the regulatory authority may suspend implementation of the new rate schedule. If the original proceeding involves a proposed increase in the rates charged in municipal areas, the municipality holds the hearing and has ninety days in which to come to a final decision. If the municipality has made no final disposition of the rate proceeding at the expiration of ninety days, the proposed rate schedule is deemed to have been approved and the municipality loses jurisdiction over the proceeding. PURA § 43(d). If an order is issued, any party to the proceeding may seek de novo appellate review in the Commission. PURA § 26(a), (g). Because most utilities provide services in both municipal and non-municipal areas, there is usually a parallel proceeding originating in the Commission to consider the same proposed rate increase as it affects non-municipal areas. The Commission, however, has a 150-day period of original jurisdiction over its portion of the rate proceeding. In addition, the Commission is allowed two days for each day of hearings in excess of fifteen days. The practical result of allowing the Commission a longer period of original jurisdiction is that it can wait for the municipality to issue a final appealable order and then consolidate de novo appellate review with its own consideration of the same proposed rate increase in non-municipal areas. Therefore, the Commission typically exercises its original and appellate jurisdiction concurrently. In these consolidated rate proceedings, the Commission has 150 days plus two days for each day of hearings in excess of fifteen days in which to make a final determination. When the Commission is faced with a particularly complex rate proceeding, protracted hearings can mean a utility’s proposed rate schedule may not take effect for a long period of time. The term “regulatory lag” is used to describe the economic consequences of this delay. In order to protect utilities from the financial harm engendered by prolonged regulatory lag, PURA section 43(e) provides that in cases in which the Commission has failed to render a final order within 150 days of the proposed effective date of the rate increase, the utility may put a changed rate, not to exceed the proposed rate, into effect upon the filing with the regulatory authority of a bond_ The utility concerned shall refund or credit against future bills all sums collected ... in excess of the rate finally ordered plus interest at the current rate as finally determined by the regulatory authority. PURA § 43(e). This practice is known as “bonding in” rates and is used to relieve the potential financial hardship imposed on a utility while it awaits a final Commission order on its requested rate increase. In Docket 9300, Texas Utilities requested the same rate increase throughout its entire service area, encompassing both municipal and non-municipal areas. As permitted by the 1983 amendments to PURA, the Commission reviewed the proposed rate increase in municipal areas under its appellate jurisdiction at the same time it considered the increase in non-municipal areas under its original jurisdiction. When 150 days had passed without the Commission’s having reached a final determination, the utility decided to implement bonded rates throughout its entire service area, and pursuant to PURA 43(e) requested that the Commission approve its bond. The Cities objected to Texas Utilities’ request for bonded rates in municipal areas, maintaining that PURA prohibits bonded rates in municipal areas once the municipality has lost its original jurisdiction over the rate proceeding. The Commission rejected this argument and determined that PURA’s bonding provision does not prohibit a utility from implementing-bonded rates in municipal areas when the underlying rate increase is subject to the Commission’s appellate jurisdiction. We conclude that the Commission’s interpretation of PURA section 43(e) is correct. Section 43(e) contains no language that limits the bonding provisions to rates being considered under the Commission’s original jurisdiction: If the 150-day period has been extended, ... and the commission fails to make its final determination of rates within 150 days from the date that the proposed change would have gone into effect, the utility concerned may put a changed rate, not to exceed the proposed rate, into effect upon the filing with the regulatory authority of a bond.... PURA § 43(e). In support of its contention that the utility may implement bonded rates only for those rates subject to the Commission’s original jurisdiction, the Cities rely on two pre-1983 eases holding that the former version of section 43(e) did not permit bonded rates in areas under the Commission’s appellate jurisdiction. See Lone Star Gas, 656 S.W.2d at 425; Arkansas Louisiana Gas Co. (Arkla) v. Railroad Comm’n, 586 S.W.2d 643 (Tex.Civ.App.—Austin 1979, writ refd n.r.e.). We conclude that the reasoning of those eases is so closely tied to the wording of PURA before the 1983 amendments that they do not support the Cities’ interpretation of amended section 43(e). In Lone Star Gas the supreme court recognized the hardship created by PURA’s failure to provide for bonded rates during an extended period of appellate review, but commented that “any changes in the protection afforded the utility should be made by the legislature.” 656 S.W.2d at 425. Perhaps responding to the court’s invitation to act, in 1983 the legislature significantly amended PURA and apparently cured this particular hardship. See GTE-SW, 833 S.W.2d at 173 (noting that an almost identical bonded rate provision in the new Gas Utility Regulatory Act cured the problems caused by the utility’s inability to implement bonded rates in municipal areas pending review de novo by the Commission). Without the ability to bond in rates, a utility’s only avenue for relief from regulatory lag in city rates, traditionally the lion’s share of its service area, would be to request interim rates. See PURA § 26(g) (allowing the Commission to authorize interim rates if “necessary to effect uniform system-wide rates”). This would necessitate a bifurcated process of considering the request for interim city rates while contemporaneously implementing bonded rates outside city limits. Such an inefficient and unwieldy process undermines the amended statutory scheme designed to consolidate consideration of system-wide rates in one proceeding. Furthermore, interim rates that require a hearing do not provide relief from regulatory lag equivalent to the bonding provision which permits implementation of new rates without Commission approval, subject only to a bond adequate to ensure possible refunds. We see no suggestion in the amended version of section 43(e) that utilities should be limited to seeking interim rates to cure regulatory lag in areas servicing cities. The Commission’s interpretation of section 43(e) is entitled to great weight, provided it is reasonable and does not contradict the plain language of the statute. Tarrant Appraisal Dist. v. Moore, 845 S.W.2d 820, 823 (Tex.1993). The Commission’s construction of the bonding provision is consistent with the statutory scheme embodied in the 1983 amendments designed to facilitate contemporaneous disposition of system-wide rates in a single proceeding. It also affords the utility protection from regulatory lag through bonded rates, whether inside or outside city limits. Nothing in the statute itself or the relevant case law supports the Cities’ restricted reading of section 43(e). We overrule the Cities’ twenty-first point of error. RATE BASE ALLOWANCES In points of error two through four, the Cities and Public Utility Counsel complain that the district court improperly upheld various aspects of the Commission’s order on rehearing relating to the prudence phase of the rate-making proceeding. Specifically, they contend that the Commission’s disallowance of Comanche Peak costs is contrary to substantial evidence and inconsistent with the Commission’s factual determinations regarding the insufficiency of Texas Utilities’ proof and with Texas law regarding the burden of proof. The Cities and Public Utility Counsel assert that reasonable minds could not reach the decision arrived at by the Commission regarding the reasonable cost of Comanche Peak, and that the Commission failed to disallow imprudent project costs as required by statute. See PURA §§ 39, 41. In August 1972, Texas Utilities announced its plan to build Comanche Peak, its first nuclear power plant. In 1977, the utility estimated that Comanche Peak Unit 1 would be commercially operable in 1981, and Unit 2 would achieve commercial operation in 1983. The total estimated cost of the project was $1.7 billion, including an allowance for funds used during construction (AFUDC). However, Unit 1 did not become commercially operable until August 1990. At the rate-making proceeding, the examiners attributed this substantial delay to Texas Utilities’ inability to obtain an operating license from the NRC. See Examiners’ Report at 5. Docket 9300 addressed the prudence of costs incurred by the utility in responding to the NRC’s concerns; the utility engaged in an unprecedented revalidation and reinspection program which caused Comanche Peak costs to nearly double. The Commission, which heard three explanations for these costs, was charged with determining which costs were prudent. Texas Utilities contended that the NRC unforeseeably and unreasonably applied stricter licensing standards to Comanche Peak, forcing the utility to implement an expensive and time-consuming revalidation and reinspeetion program in order to obtain an operating license. The utility took the position that all of these were regulatory costs that should be included in rate base. At the other end of the spectrum, the Cities and Public Utility Counsel argued that imprudent project management caused the NRC to lose confidence in Comanche Peak’s safety, and that all post-1984 costs incurred in responding to these concerns should be disallowed as imprudent. The Commission’s general counsel, supported by an evaluation conducted by the Nielsen-Wur-ster Group, an independent auditor, concluded that Texas Utilities’ inability to obtain an operating license resulted from the NRC’s significant, but unfounded, quality concerns. The general counsel maintained that certain costs should, however, be disallowed based on Nielsen-Wurster’s findings that the utility acted imprudently in discrete instances during the life of the project. The Commission reviewed the evidence presented by the parties and general counsel and determined that $537.90 million of Comanche Peak costs were imprudently incurred and should be disallowed. To support the assertion that the Commission erred in the prudence phase of Docket 9300, the Cities and Public Utility Counsel make three basic points: (1) Texas Utilities did not sustain its burden of proof on the prudence of its Comanche Peak expenditures, (2) Texas Utilities did not properly quantify its imprudent Comanche Peak costs, and (3) the Nielsen-Wurster report does not constitute substantial evidence to support the Commission’s determination of which Comanche Peak costs were imprudently incurred. Taken together, these points assert that the evidence presented during 203 days of hearings cannot support the Commission’s final order with respect to disallowances. See APA § 2001.174(2)(E); Texas Health Facilities Comm’n v. Charter Medical-Dallas, Inc., 665 S.W.2d 446, 452-53 (Tex.1984). In conducting a substantial evidence review, we must determine whether the evidence as a whole is such that reasonable minds could have reached the conclusion the agency must have reached in order to take the disputed action. Charter Medical, 665 S.W.2d at 453. We may not substitute our judgment for that of the agency and may consider only the record on which the agency based its decision. Texas State Bd. of Dental Examiners v. Sizemore, 759 S.W.2d 114, 116 (Tex.1988), cert. denied, 490 U.S. 1080, 109 S.Ct. 2100, 104 L.Ed.2d 662 (1989). The party bringing the appeal bears the burden of showing a lack of substantial evidence. Charter Medical, 665 S.W.2d at 453. If substantial evidence would support either affirmative or negative findings, we must uphold the agency’s order, resolving any conflict in favor of the agency’s decision. Auto Convoy Co. v. Railroad Comm’n, 507 S.W.2d 718, 722 (Tex.1974). The Cities and Public Utility Counsel essentially argue that because the Commission was not persuaded by the utility’s argument that all Comanche Peak costs were prudent, and because the utility then failed to quantify the impact of its imprudence by identifying costs related to imprudent management, the Commission was required to disallow all of these expenditures. We do not agree. The Commission determined the evidence presented by the parties did not provide an accurate foundation on which to base its disallowance decisions. It therefore turned to the report prepared by the Nielsen-Wurster Group. Nielsen-Wurster had previously performed twelve comprehensive prudence reviews of other nuclear plants, eight for commissions and four on behalf of utilities, before it was retained by the Commission to evaluate the planning and management of Comanche Peak. After an extensive investigation, Nielsen-Wurster offered its findings in ten days of testimony presented by five expert witnesses. The Cities and Public Utility Counsel argue that the evidence presented by Nielsen-Wurster cannot serve as a proper foundation for Commission decision-making because it does not provide a sufficiently detailed breakdown of all Texas Utilities’ expenditures identifying those related to utility imprudence and those that would have been necessary absent any imprudence. They assert that in order to provide evidence sufficient to support the Commission’s order, either the utility or Nielsen-Wurster was required to “produce a breakdown of the Company’s post-March 1985 expenditures, disaggregated between those that were ‘remedial’ and those that would have been incurred even absent the prolonged licensing delay.” The argument urged on appeal is that once the Commission has determined the utility’s evidence is insufficient to demonstrate that all expenditures were prudently incurred, the utility must then “isolate out the costs associated with its imprudent conduct” in order to avoid having the Commission disallow all the costs incurred. In support of their argument, the Cities and Public Utility Counsel direct this Court to Coalition of Cities v. Public Utility Commission, 798 S.W.2d 560, 568-64 (Tex.1990), cert. denied, 499 U.S. 983, 111 S.Ct. 1641, 113 L.Ed.2d 736 (1991), in which the supreme court stated that “[a] party who fails to meet its burden of proof loses.” In Coalition of Cities, the utility “lost” because neither the utility nor any other party satisfied the Commission that $1,453 billion in expenditures were prudently incurred. Nowhere does the supreme court state that a utility must segregate imprudent costs. When a utility fails to persuade the Commission of the wisdom of all its expenditures, that does not preclude the Commission from considering the other evidence presented in the rate-making proceeding. Indeed, it is the Commission that is charged with sifting through the evidence and deciding whether imprudent conduct caused certain expenditures. Having reviewed the utility’s evidence and the Nielsen-Wurster report, the Commission determined that $90.5 million of the Comanche Peak Response Team expenses and $79.9 million of the Corrective Action Program expenses were imprudent. The Commission made further disallowances for other imprudent conduct associated with the delay in licensing; it disallowed $54.1 million in time-driven indirect costs and $167.3 million in AFUDC. The Commission rejected Texas Utilities’ claim that the costs associated with the rein-speetion and revalidation program were entirely due to higher regulatory standards; it similarly rejected the Cities and Public Utility Counsel’s contentions that all such costs should be disallowed as imprudent. The Commission accepted the Nielsen-Wurster study as evidence that some, but not all, of the expenditures were imprudently incurred. The Commission found that the NRC’s Technical Review Team findings on the plant’s condition were partly unfounded, although they did identify weaknesses in the pre-1985 quality assurance program. The Commission also concluded that the growth of regulatory requirements increased the cost and extended the construction schedule beyond Texas Utilities’ control. These findings are supported by testimony adduced during the rate-making proceeding and provide substantial evidence upon which the Commission could base its decision to examine all the costs in detail and make discrete disallowanc-es associated with imprudent conduct. The Cities and Public Utility Counsel vigorously assert that the Commission erred in not making any disallowance for the costs of executing the Corrective Action Program. However, the Commission determined that although the imprudence of the utility was partially responsible for the need to carry out the Corrective Action Program, the changed regulatory climate would have made such a program necessary even in the absence of utility imprudence. The Commission’s findings are presumed to be supported by substantial evidence, and the Cities and Public Utility Counsel have failed to demonstrate that reasonable minds could not have come to that decision based on this record. Charter Medical, 665 S.W.2d at 453. The Cities and Public Utility Counsel also complain that the Commission improperly applied a “sliding” standard of prudence, assigning degrees of imprudence to utility decisions and making disallowances only when the imprudence reached a certain level or degree. After reviewing the record we believe this criticism is unfounded. The Commission determined that the costs associated with responding to NRC concerns were necessary in part because of utility imprudence and in part because of the NRC’s application of higher safety and inspection standards in the face of mounting concerns about the safety of nuclear power plants in general. The Commission’s finding of fact 138 expresses this conclusion. The Examiners’ Report notes that Texas Utilities’ conduct was not the sole reason for the expenditures necessary to regain the NRC’s confidence. The Commission then made partial disallowances for the costs of the remedial action program, not the wholesale disallowances recommended by the intervenors. After a careful and thorough review of all the evidence presented in 203 days of hearings, the Commission made findings of fact and conclusions of law based on that review. For each finding of imprudence in the construction and management of Comanche Peak, the Commission made a disallowance for the associated costs. The Commission also made significant disallowances for the cost of the delay in licensing, reflecting its opinion that the utility’s imprudence was partially responsible for that delay. The Cities and Public Utility Counsel next contend that the Commission’s order is improper because it is not supported by underlying findings of fact. We understand their complaint to be that the findings of fact do not meet the requirements of the APA. See APA § 2001.141(d) (“Findings of fact, if set forth in statutory language, must be accompanied by a concise and explicit statement of the underlying facts supporting the findings.”) The supreme court has concluded that an agency’s findings of fact need the additional support of findings of underlying facts only when the findings are stated in terms taken directly from the enabling legislation or when they “represent the criteria that the legislature has directed the agency to consider in performing its function.” Charter Medical, 665 S.W.2d at 451. The Cities and Public Utility Counsel contend that findings of fact 138 through 152 are “ultimate” findings by which the Commission fulfills its statutory obligation to exclude from rate base all imprudently incurred post-1985 remedial costs, and as such they require underlying findings of fact. We first consider whether the findings of fact at issue are indeed “ultimate findings.” In City of El Paso, this Court stated that although PURA does not expressly require the Commission to make a finding of prudence before including costs in rate base, once the Commission finds a major project to have been imprudently planned or managed, it should generally disallow project costs to the extent of the imprudence. City of El Paso, 839 S.W.2d at 908. A determination that an expenditure is imprudent carries the legal consequence of its exclusion from rate base. Such a finding must be supported by underlying findings because it embodies one of the criteria the Commission must consider in deciding whether to include the particular expenditure in rate base. The following findings of fact are here at issue: 138. As discussed in Section VI.Q.2. of this Report, the evidence does not support imprudence disallowances of the magnitude proposed by the intervenors. 139. As discussed in Section YI.Q.2. of this Report, the costs of executing the Comanche Peak Response Team and Corrective Action Program were prudent. 140. As calculated in Section VI.Q.2. of this Report, the total imprudent costs incurred by [Texas Utilities] through the end of the test year is $537.9 million, which allocates $382.05 million to Unit 1 and $155.85 million to Unit 2. To meet the criteria set forth in Charter Medical and City of El Paso, these findings must be accompanied by underlying findings connecting evidence to the conclusions expressed in the Commission’s ultimate findings. In support of finding of fact 138, the Examiners’ Report explains that the utility should not be prohibited from including any of the costs of the remedial action program in rate base because other factors contributed to the NRC’s application of stricter regulatory standards. See Examiners’ Report at 169. Those other factors are also identified in the Report: “On balance, although the inspection standards and procedures applied by the Technical Review Team were the same as those previously used by the project’s quality control inspectors, the Technical Review Team conducted its inspections and scrutinized its inspection results at Comanche Peak in a manner as never before.” See id. at 124. These findings support the Commission’s decision not to make the wholesale disallowances proposed by the intervenors. Nielsen-Wurster did not recommend disallowing any costs related to the post-effective date execution of the response team or the corrective action program. See Examiners’ Report at 139. Finding of fact 140 expresses the Commission’s final calculation of total imprudent costs incurred by the utility through the end of the test year. These calculations are supported by extensive explanations in the Examiners’ Report as well as specific findings of fact in the order on rehearing for each element of the total disal-lowance. We reject the Cities and Public Utility Counsel’s contention that findings of fact 138, 139, and 140 are not adequately supported by underlying findings of fact. Finally the Cities and Public Utility Counsel challenge the Commission’s failure to impose specific disallowances flowing from its finding that the utility imprudently failed to infuse its senior management with personnel having the appropriate nuclear experience. During the rate-making proceedings the examiners determined that it was impossible to state generally the effect of this lack of nuclear experience; rather, as in the entire prudence review, the examiners proposed an examination of the utility’s discrete actions and decisions throughout the project. The Commission adopted the examiners’ reasoning and made disallowances for costs associated with imprudent management. These disallowances represent the Commission’s exercise of its discretion in determining rate base; the findings are not arbitrary or capricious or unsupported by substantial evidence. After careful review and consideration of all the arguments raised by the Cities and the Office of Public Utility Counsel, we overrule points of error two through four. COMANCHE PEAK RESPONSE TEAM DELAY In its first point of error, Texas Utilities complains of the Commission’s disallowance of $194.4 million representing costs associated with an imprudent seven-month delay in Comanche Peak construction. Each of the utility’s arguments advanced under this point of error, however, was presented to the Commission during the rate-making proceeding and rejected with adequate accompanying findings supported by substantial evidence. We decline to substitute our judgment for that of the Commission, and will overrule the point of error. The utility first argues that there is not substantial evidence to support the Commission’s finding that Revision 2 to the Comanche Peak Response Team Program Plan was not a