Reliant Energy, Inc. v. Public Utility Commission

153 S.W.3d 174, 2004 Tex. App. LEXIS 11280, 2004 WL 2900360

• Court: Court of Appeals of Texas
• Decided: December 16, 2004
• Docket: 03-02-00246-CV
• Status: Published

Opinion

OPINION

DAVID PURYEAR, Justice.

Our opinion and judgment issued on August 26, 2004 are withdrawn and the following opinion is substituted.

Several parties appeal from the district court’s judgment affirming a final order of the Public Utility Commission (“the Commission”) setting cost-of-service rates for the transmission and distribution utility (“TDU”) of Reliant Energy, Inc. (“Reliant”). 1 After a contested case proceeding, the Commission entered an order setting rates below the level Reliant sought for its TDU. Reliant and several parties who intervened at the agency level sought review by the district court. The district court concluded that one of the Commission’s findings of fact was a prohibited advisory opinion, but otherwise affirmed the Commission’s order. Reliant, Gulf Coast Coalition of Cities (“Gulf Coast”), Office of Public Utility Counsel (“OPC”), and Consumer Owned Power Systems 2 (“COPS”) all challenge the district court’s affirmance of the Commission’s order. We will reverse the portion of the district court’s judgment affirming the Commission’s inclusion of $107.3 million for the interconnection of Merchant Plant 4. We affirm the district court’s judgment in all other respects. We remand the case to the Commission for further proceedings.

BACKGROUND

The general outline of Texas’s scheme for the transition from a regulated electric utility industry to a competitive marketplace has been addressed in detail. See, e.g., Reliant Energy, Inc. v. Public Util. Comm’n, 101 S.W.3d 129, 133-36 (Tex.App.-Austin 2003), rev’d in part sub nom. Centerpoint Energy, Inc. v. Public Util. Comm’n, 47 Tex. S.Ct. J. 672, 2004 WL 1386192 (Tex.2004); Reliant Energy, Inc. v. Public Util. Comm’n, 62 S.W.3d 833, 835-36 (Tex.App.-Austin 2001, no pet.). Under the regulated system, a single utility generated electricity, built and maintained the electricity distribution grid, and sold the electricity to consumers. In 1999, the legislature added chapter 39 to the Public Utility Regulatory Act (“PURA”), 3 finding that the “public interest in competitive electric markets requires that, except for transmission and distribution services and for the recovery of stranded costs, electric services and their prices should be determined by customer choices and the normal forces of competition.” PURA § 39.001(a).

PURA chapter 39 requires existing integrated utilities to separate, or unbundle, into three units by January 1, 2002: a power generation company, a TDU, and a retail electric provider. Id. § 39.051(b). Despite the emphasis on competition in the retail market, the Commission will continue to regulate the TDUs’ rates and services. As part of the transition, the statute required Reliant and other electric utilities to file an unbundled cost-of-service rate case to establish transmission and distribution rates for their TDUs, and the Commission to set transmission and distribution rates as of January 1, 2002. Id. § 39.201.

Cost-of-service rates are set to allow a utility to recover a return on its invested capital, also called rate base, plus its reasonable and necessary expenses. 4 Because many of the unbundled TDUs did not exist for a full year before January 1, 2002, the legislature required the rates for the new transmission and distribution service to be based on a forecasted 2002 test year. See id. § 39.201(b)(1); see also id. § 11.003(20).

PURA provides general guidelines for setting rates. The Commission is required to “establish the utility’s overall revenues at an amount that will permit the utility a reasonable opportunity to earn a reasonable return on the utility’s invested capital used and useful in providing service to the public in excess of the utility’s reasonable and necessary operating expenses.” Id. § 36.051. A utility is entitled to rates sufficient to repay its expenses, without a return or profit on those expenses, and to provide a return on the invested capital included in its rate base, without repaying that investment. Cities for Fair Util. Rates v. Public Util. Comm’n, 924 S.W.2d 933, 935 (Tex.1996). In determining the amount of invested capital used to serve customers, the Commission uses the “original cost, less depreciation, of property used by and useful to the utility in providing service.” PURA § 36.053(a). To establish the utility’s reasonable and necessary operating expenses, the Commission starts with the expenses incurred during the test year, and then adjusts those expenses for known and measurable changes. 16 Tex. Admin. Code § 25.231(b) (2004). Operating expenses include such things as depreciation, federal income taxes, and employee wages. See id.

Intending to streamline the rate proceedings for the individual TDUs, the Commission initiated a generic proceeding (Docket No. 22344) to determine issues common to all the affected TDUs, then apply the generic determination in the utility-specific cases. The Commission’s final order in the Reliant-specific case included both recovery of estimated stranded costs 5 for Reliant’s generation company and cost-of-service rates for the Reliant TDU. See generally Tex. Pub. Util. Comm’n, Application of Reliant Energy for Approval of Unbundled Cost of Service Rate Pursuant to PURA 39.201 and Public Utility Commission Substantive Rule 25.3Ü, Docket No. 22355, 2001 WL 1869949 (Oct. 3, 2001) (hereinafter, “Reliant Order ”).

The parties challenged the Commission’s order on several grounds in the district court. The court decided that the issue of interest on excess mitigation credits was not ripe for decision and thus that the finding of fact (and the related discussion) on that issue was a prohibited advisory opinion, but otherwise affirmed the Commission’s order.

DISCUSSION

In this appeal, the parties challenge the Commission’s decisions regarding ele ments of the rate base, rate of return, expenses, and rate design. Our review of this appeal is under the substantial-evidence standard. PURA § 15.001. That standard is largely deferential to the Commission’s decision, as set out in the following statute:

If the law authorizes review of a decision in a contested case under the substantial evidence rule or if the law does not define the scope of judicial review, .a court may not substitute its judgment for the judgment of the state agency on the weight of the evidence on questions committed to agency discretion but:

(1) may affirm the agency decision in whole or in part; and

(2) shall reverse or remand the case for further proceedings if substantial rights of the appellant have been prejudiced because the administrative findings, inferences, conclusions, or decisions are:

(A) in violation of a constitutional or statutory provision;

(B) in excess of the agency’s statutory authority;

(C) made through unlawful procedure;

(D) affected by other error of law;

(E) not reasonably supported by substantial evidence considering the reliable and probative evidence in the record as a whole; or

(F) arbitrary or capricious or characterized by abuse of discretion or clearly unwarranted exercise of discretion.

Tex. Gov’t Code Ann. § 2001.174 (West 2000).

We presume that the Commission’s findings are supported by substantial evidence, and the contestant bears the burden of proving otherwise. See Southwestern Pub. Serv. v. Pub. Util. Comm’n, 962 S.W.2d 207, 215 (Tex.App.-Austin 1998, pet. denied). A court reviewing an agency action shall reverse and remand the cause to the agency when substantial rights of the appellant have been prejudiced by an agency’s findings that are not reasonably supported by substantial evidence considering the reliable evidence in the record as a whole. Tex. Gov’t Code Ann. § 2001.174(2)(E). We may not substitute our judgment for that of the agency on the weight of the evidence. Southwestern, 962 S.W.2d at 215. “Substantial evidence” does not mean a large or considerable amount Of evidence, but rather such relevant evidence as a reasonable mind might accept as adequate to support a conclusion of fact. Pierce v. Underwood, 487 U.S. 552, 564-65, 108 S.Ct. 2541, 101 L.Ed.2d 490 (1988); Lauderdale v. Department of Agric., 923 S.W.2d 834, 836 (Tex.App.-Austin 1996, no writ). We must uphold an agency’s finding even if the evidence actually preponderates against the agency’s finding so long as enough evidence suggests the agency’s determination was within the bounds of reasonableness. Southwestern, 962 S.W.2d at 215.

The parties raise several complaints. Gulf Coast, COPS, and OPC argue that the amounts used to calculate invested capital should have included an average of the investment for transmission lines in 2002 rather than the total amount that the Reliant TDU will have invested in transmission facilities by the end of 2002. Gulf Coast also disputes amounts included for transmission lines to serve one particular generator. Gulf Coast further contends that the Commission should have returned the value of nuclear insurance premiums to customers. Gulf Coast and OPC complain that the return on equity used to determine the rate of return was determined in the generic proceeding. OPC complains about the transmission cost recovery fac tor and the amount the Commission included in rates for a minimum delivery cost necessary for each customer, regardless of amount of use. OPC also challenges the cost escalation and asset degradation factors the Commission assessed on various items on Reliant’s balance sheet. Reliant asserts that the Commission’s consolidated tax-savings adjustment amounts to improper retroactive ratemaking and includes companies that should not be part of the calculation. Reliant also complains that the Commission refused to include overfunding to its retirement plan as part of invested capital and that it effectively excluded increased reliability expenses.

RATE BASE

A utility’s rate base or invested capital consists of the cost, less depreciation, of property “used by and useful to the utility in providing service.” See PURA §§ 36.051, .053(a); 6 see also 16 Tex. Admin. Code § 25.231(c)(2) (2003). “ ‘Used and useful’ refers to ‘such property as has been acquired ... in good faith and held for use in the reasonably near future in order to enable [a utility] to supply and furnish adequate and uninterrupted ... service.’ ” Cities, 924 S.W.2d at 935 (quoting Lone Star Gas Co. v. State, 137 Tex. 279, 153 S.W.2d 681, 698 (Tex.1941)). The Commission has significant discretion in determining what property is used by and useful to the utility solely to provide service to its ratepayers; that discretion is even greater for property used for other purposes with less direct benefit to ratepayers. See El Paso Elec. Co. v. Public Util. Comm’n, 917 S.W.2d 846, 856 (Tex.App.-Austin 1995, writ dism’d). As the United States Supreme Court has written, “The economic judgments required in rate-making proceedings are often hopelessly complex and do not admit of a single correct result.” Duquesne Light Co. v. Barasch, 488 U.S. 299, 314, 109 S.Ct. 609, 102 L.Ed.2d 646 (1989) (quoted in El Paso, 917 S.W.2d at 857).

Several parties challenge the Commission’s determination of the rate base for the Reliant TDU in the forecasted 2002 test year. Reliant questions the propriety of the failure to include an overfunded retirement plan in the rate base. Other parties challenge the inclusion of facilities in use at year’s end rather than figuring a year-long average by pro-rating the cost of facilities brought on line during the year. Gulf Coast complains about the accuracy of the amount of transmission system costs allotted to a particular plant.

Overfunded Retirement Plan

Reliant argues that the Commission erred by excluding from Reliant’s rate base the amount in Rehant’s retirement plan exceeding the funds necessary to satisfy its retirement obligations. Reliant claims that (1) no substantial evidence sup ports the Commission’s decision and (2) the decision is arbitrary and capricious because it is inconsistent with previous treatment of underfunding of Reliant’s retirement fund.

Reliant’s witness, James Brian, supported including in Reliant’s rate base the excess amounts that had accumulated in the retirement plan due to better than expected returns on investments; Brian projected that the plan would be $40 million overfunded in the forecast test year. 7 Brian testified that the overfunding would occur even with no additional contributions from the company. Brian testified that ratepayers benefit from the excess in that the excess funds reduce Reliant’s cost of service because Reliant will not have to collect anything in rates to cover contributions to the retirement plan. He asserted that, because the Commission previously reduced Reliant’s rate base for underfunding in the retirement plan, the failure to increase Reliant’s rate base for overages in the retirement plan would be arbitrary and capricious.

Commission staff witness Dinah Arce disagreed with placing the retirement plan excess in the rate base. She testified that Reliant’s ratepayers had paid into the plan through the payment of rates and that, according to Brian, the overfunded amounts resulted from investment returns on the plan balance. Arce then opined that it is not reasonable for Reliant shareholders to earn an additional return from ratepayers based on the returns on the plan investments.

We conclude that the record contains substantial evidence to support the Commission’s finding that the overfunded amounts should not be included in Reliant’s rate base. If a fully funded retirement plan is used and useful in providing electricity, it can be in the rate base. But previous decisions by the Commission to deduct underfunded amounts from Reliant’s rate base 8 do not make the Commission’s decision to exclude excess funds arbitrary or capricious. Even assuming, as Reliant argues, that a fully-funded pension plan is an investment that is “used and useful” in attracting and keeping workers who help supply electric- service, the Commission could reasonably conclude that excess funds are not used or useful in that same way. Excess and deficit in the same plan do not necessarily represent different sides of the used-and-useful coin. If the plan is not fully funded, then any deficit can be deducted from the rate base to reflect the company’s failure to make the full investment necessary to provide the full benefit useful in providing electricity (ie.g., to recruit and retain workers); however, it is not unreasonable to conclude that excess plan funds are not additionally beneficial or useful in providing electricity and therefore should not be included in the rate base. Similarly, while lower rates may make purchasing electricity more pleasant for customers, the Commission could conclude that lower rates are not necessarily used or useful to the utility in providing service. We overrule Reliant’s fourth issue.

Year-end. Rate Base

Gulf Coast, COPS, and OPC each claim that the Commission erred by allowing Reliant to calculate its rate base using the amount it anticipated investing in transmission facilities by the end of 2002 (“year-end rate base”); they argue that the Commission should have used the amount of investment made by midway through the year on June 30, 2002 (“average-year rate base”). They contend that the use of a year-end rate base violates the PURA requirement that investments be included in rate base only if they are “used and useful” in providing service. PURA §§ 36.051, .053(a). They complain that rates could be paid all year long for plant used only on the last day of the year.

The construction of a statute by the administrative agency charged with its enforcement is entitled to great weight, as long as the construction is reasonable and does not contradict the plain language of the statute. State v. Public Util. Comm’n, 883 S.W.2d 190, 196 (Tex.1994); Dodd v. Meno, 870 S.W.2d 4, 7 (Tex.1994). The process of setting regulated rates for utilities is “far from a precise process; instead, ratemaking relies substantially on informed judgment and expertise and utilizes projections and estimates in virtually all areas.” Public Util. Comm’n v. GTE-Southwest, Inc., 901 S.W.2d 401, 411 (Tex.1995).

The mandated use of a forecasted test year makes calculating the rate base even more speculative. In a traditional rate-setting process, rates are based on a historical test year for which an objective analysis can be made of what property was acquired and is necessary to provide utility service. See generally Central Power & Light Co. v. Public Util. Comm’n, 36 S.W.3d 547, 552-53 (Tex.App.-Austin 2000, pet. denied) (“CPL ”). But estimates are necessary even when the ratemaker uses a historical test year. See GTE-Southwest, 901 S.W.2d at 411. Here, the legislature required the Commission to use the fore-casted test year of 2002. See PURA § 39.201(b)(1). Using a forecasted year means that, unlike when calculating rate base using a historical year, the Commission had to set rates before the capital additions to be made in the relevant year were acquired or “used and useful” in supplying electricity.

The City of Houston, Texas Industrial Energy Consumers (“TIEC”), and OPC sponsored witnesses who supported using the average-year method. Houston’s James Daniel testified that other jurisdictions use average-year to determine rate base for forecasted years. He testified that using the year-end method would allow Reliant to charge for plant and claim depreciation on it for up to a year before it was in service or “used and useful.” He said that, although year-end measures were appropriate for historical test years, they were inappropriate for a forecasted year. OPC witness Ben Johnson explained that, when using historical test years, regulatory bodies use year-end balances to better match the level of investment in use during the rate year because the level of investment on December 31, 2001 is near the level of investment of January 1, 2002. In this case, however, Reliant requested that investment be calculated based on plant projected to be in use on December 31, 2002 to set rates to be effective on January 1, 2002 — almost a year before the plant was in use. Johnson asserted that this would distort Reliant’s balance sheet because the rates in place all year would be set for end-of-2002 investment levels, while the revenue and income stream would represent average conditions throughout 2002. He said that, when historical test years were used, this revenue/investment mismatch was balanced by a deterioration in earnings due to inflation, but that the forecasting process should account for inflation in a forecasted test year. TIEC witness Michael Gorman agreed, testifying that using the average-year instead of a year-end method would reduce Reliant’s rate base by $192.5 million and lower its revenue requirement by $24.4 million; Houston’s Daniel estimated a $249 million difference, resulting in a correspondingly lower revenue requirement.

Reliant witness Brian disputed these witnesses’ testimony. He asserted that an average-year rate base would .create mismatches in the rate-filing package between such items as accumulated depreciation and depreciated expense. He argued that the Commission should not rely on the possibility of rate correction to address undervaluing of additions to the rate base because the lag between the application and award of rate changes would further damage the TDUs. Brian also testified that the year-end method was dictated by the Commission’s instruction in the rate filing package that stated, “All rate base items for the Forecast Year shall be reflected at the Forecast Year-end amounts.” Brian argued that the opponents of a year-end rate base should have voiced their concerns when the instructions were being formulated.

Expressly accepting the testimony of Daniel, Johnson, and Gorman, the ALJ proposed that the Commission use the average-year rate base, concluding that the average-year rate base would better match plant investment levels with corresponding expenses than would a year-end calculation. The ALJ opined that using a year-end measure would allow Reliant to charge customers for plant that had not yet become used and useful, potentially violating PURA section 36.053. The ALJ acknowledged that, under the average-year method, Reliant would not be able to recover for its plant-in-service for years following 2002 until a new ratemaking proceeding, but noted that the Commission’s rules allowed expedited updating of its rates. See 16 Tex. Admin.Code § 25.193(b) (2004). The ALJ also wrote, “It has been [Reliant’s] experience in recent years that new revenue has more than made up for these new expenses under existing rates.”

The Commission, however, rejected the ALJ’s recommendation. The Commission stated that revenue/investment mismatches were unavoidable when a forecasted year was used. The Commission opined that, while using a 2002 average balance might somewhat even out the imbalance over that year, capital additions made during 2002 would be undervalued in the rate base for 2003 and subsequent years until a new ratemaking proceeding is initiated. Therefore, the Commission allowed Reliant to use a 2002 year-end balance to calculate its projected rate base.

On appeal, Gulf Coast, OPC, and COPS reiterate their argument that the “used and useful” requirement prohibits use of a year-end rate base because the rates can include capital spent on facilities that are not in service until late in the year. They contend that this also violates a principle of matching revenues and expenses. They further complain that the commissioners considered irrelevant factors in deciding to use the year-end method. They cite comments by commissioners in the open meeting that the year-end method would provide an incentive for utilities to build new transmission projects by showing that the Commission was trying to ensure that utilities could timely recover their investments in such projects. These appellants complain that such incentives are not proper considerations for inclusion in rate base, and instead violate the requirement that plant be used and useful to be included in rate base; indeed, Gulf Coast argues that the rate of return exists to provide the necessary incentive and must not be supplemented by overstated investment in the rate base. These appellants complain that the Commission’s assertion that the year-end method better reflects the plant in use for 2003 is based on speculation that there will not be another ratemaking proceeding for 2003. They contend that the transmission cost recovery factor (see 16 Tex. Admin. Code § 25.193) exists to incorporate additions to the rate base without a full-blown ratemaking proceeding.

The Commission can change, modify or vacate an ALJ’s order where the Commission determines that the ALJ misapplied or misinterpreted Commission rules or policies, applicable law or prior administrative decisions, or issued a finding of fact not supported by a preponderance of the evidence. See Tex. Gov’t Code Ann. §§ 2001.058(e)(1), 2003.049(g) (West 2000). The Commission must “state in writing the specific reason and legal basis for its determination” to depart from the ALJ’s order. See id. § 2003.049(h) (West 2000).

We conclude that appellants have not shown that the Commission erred by adopting the year-end rate base. The legislature granted the Commission broad powers and discretion in regulating public utilities. GTE-Southwest, 901 S.W.2d at 409. “An administrative agency is created to centralize expertise in a certain regulatory area and, thus, is to be given a large degree of latitude by the courts in the methods by which it accomplishes its regulatory function.” Id. (quoting City of Corpus Christi v. Public Util. Comm’n, 572 S.W.2d 290, 297 (Tex.1978)). PURA does not specify when during the forecasted test year the TDU’s investments must occur in order to be included in that utility’s rate base. See PURA § 39.201(b)(1). The legislature thus left that decision to the Commission’s discretion. Further, the supreme court does not always require that plant be in use in order to be “used and useful” enough to be included in rate base. See Cities, 924 S.W.2d at 941-42. We find no statutory barrier to the Commission considering whether its rate-base decisions will encourage investment or how the test-year rate base will affect rates for future years — particularly in this rare use of a forecasted year for the unique purpose of regulating the unbundling of electric utilities to facilitate the transition from a regulated market to a competitive market. In this context, the Commission’s decision that the 2002 year-end rate better reflects the plant used and useful for 2002 and for 2003, during which time the rates will also be in effect — essentially creating an average-two-year method- — is not erroneous.

We overrule Gulf Coast’s third issue, OPC’s second issue, and all of COPS’s issues.

Merchant Plant 4

Gulf Coast argues that the Commission overstated Reliant’s transmission system capital costs by relying on a cost estimate of a particular project by a witness who in rebuttal reduced his estimate of the cost of that same project. Gulf Coast does not on appeal challenge the Commission’s inclusion of costs for interconnecting Merchant Plant 4, but disputes the amount included. Gulf Coast would rely on the revised estimate rather than the original estimate.

In March 2000, Reliant witness John Houston testified about expenditures necessary for the continued operation- of Reliant’s transmission delivery system and the effect of the interconnection of new merchant generators to the Reliant transmission system. He estimated that $300 million of the approximately $482 million to be spent on the transmission system during the years 2000-02 would be devoted to the interconnection of merchant generators to the Reliant transmission system. The City of Garland submitted a workpaper bearing Houston’s initials containing an estimate that the interconnection of Merchant Plant 4 would cost $107.3 million.

In testimony dated December 12, 2000, COPS witness Brian Gedrich asserted that he did not believe that Reliant would incur all the costs claimed. He opined that the market would not demand all the capacity that Reliant projected by the end of 2002. In a chart, he used Houston’s estimate that Merchant Plant 4 would cost $107.3 million, but then listed that amount under the column heading “$ Excluded.”

In his rebuttal testimony, dated December 29, 2000, Houston testified that the transmission cost estimates had changed. Some had increased because of demand but others decreased because new technology allowed Reliant to increase the capacity of existing transmission corridors, reducing the need for new transmission construction. Houston said that the new technology eliminated the need to construct a relief circuit to interconnect Merchant Plant 4. Houston said:

Originally, we estimated that the entire interconnection would cost approximately $107.3 million, including approximately $42.5 million for the Greens Bayou-White Oak Project. The project is now estimated to cost $50.2 million. If we had not been able to use the new ACSS conductor technology on the King-North Belt circuits, and a 345 KV Greens Bayou-White Oak circuit was still needed, then the estimated cost would be $92.7 million, roughly 14% below the original estimate.

This testimony was uncontroverted. 9 Houston further testified that, despite the decrease in the amount attributed to the Merchant Plant 4 interconnection, increased demand caused him to revise his estimate of Reliant’s overall projected interconnection costs upward by $96.4 million over his original estimates. Reliant did not request additional funds based on this new estimate, but used the revised estimates to bolster its position that its original estimate was reasonable.

In the proposal for decision (“PFD”) dated March 27, 2001, the ALJ recommended that the Commission not include in Reliant’s transmission system capital costs any of the costs associated with the Merchant Plant 4. The ALJ wrote, “Nobody pretends to know which plants will and will not actually be built by the end of 2002, but the ALJ finds that COPS witness Mr. Gedrich drew the most reasonable line.” The ALJ then used Gedrich’s table listing the to-be-excluded cost of Merchant Plant 4 as $107.3 million, based on Houston’s original testimony; the ALJ did not explain the use of the original Merchant Plant 4 estimate rather than the revised estimate.

The Commission, however, rejected the ALJ’s recommendation to exclude the costs of interconnecting Merchant Plant 4 during the relevant time period. The Commission added $107.3 million, the exact amount Houston originally estimated, ignoring his revised estimate, for the cost of the Merchant Plant 4 interconnection.

The issues raised for a reviewing court when an agency rejects uncon-tradicted evidence are complex. See Cities of Port Arthur v. Railroad Comm’n, 886 S.W.2d 266, 270 (Tex.App.-Austin 1994, no writ). An administrative agency’s decision is to be based on evidential facts and made by experienced officials with an adequate appreciation of the complexities of the subject which is entrusted to their administration. Id. But an agency is not obliged to accept opinion testimony from an expert— even if it is the sole evidence on the issue and is uncontradicted and unimpeached. See Fuel Distrib., Inc. v. Railroad Comm’n, 727 S.W.2d 56, 61 (Tex.App.-Austin 1987, writ ref'd n.r.e.). We must uphold an agency’s finding even if the evidence actually preponderates against the agency’s finding so long as enough evidence suggests the agency’s determination was within the bounds of reasonableness. Southwestern, 962 S.W.2d at 215.

Gulf Coast claims that the Commission’s inclusion of $107.3 million for the Merchant Plant 4 interconnection costs is not supported by substantial evidence. Gulf Coast argues that Houston’s original testimony is the only source of the $107.3 million estimate, and that his revised estimate — the validity of which is unchallenged — renders the original estimate essentially a nullity. Gulf Coast contends that the ALJ used the $107.3 million estimate only because it adopted Gedrieh’s summative chart, and probably did not amend the chart with Houston’s revised estimate because the revision did not affect the ALJ’s recommendation that no costs associated with the Merchant 4 plant be awarded.

Reliant and the Commission argue that we may examine only the overall allocation of $261.2 million in finding 33A for service to new generators as well as the total of all projected capital expenditures. Reliant argues that it is inequitable to consider only the forecasted reductions for Merchant Plant 4, but not increases in other transmission costs that outstrip the Merchant Plant 4 savings in deciding whether the evidence supports the Commission’s order.

The Commission may depart from ALJ recommendations so long as it explains its departure. Tex. Gov’t Code Ann. § 2003.049(g),(h) (West 2000). Houston’s increased estimates of overall costs could support an increase attributable to other expenses, but the Commission’s explanation for its departure from the ALJ’s recommendation was not that overall capital expenditures had increased, but that Merchant Plant 4 costs must be included. It is significant that the Commission included $107.3 million, the figure Houston originally gave for Merchant Plant 4, not the $96.4 million he attributed to other transmission cost increases, or some other figure. The Commission’s order is clear that the upward revision of the interconnection costs was due, not to a general increase in such costs, 10 but solely to the inclusion of Merchant Plant 4 interconnection costs:

In evaluating the ten planned merchant generator interconnection projects that Reliant sought to include in its rate base, the ALJs considered whether the generator requesting the service had made a substantial irrevocable commitment to the transmission project. The Commission agrees with the standard used by the ALJs, but finds that Merchant Plant I sufficiently meets the standard of substantial irrevocable financial commitment. Findings of Fact Nos. 33 and 34 are deleted and new Findings of Fact Nos. 33A and 34A are added and reflect the inclusion of $107.3 million for this project

Reliant Order at 57 (emphasis added). But Houston clearly and unequivocally stated in rebuttal to his own testimony that the project would cost only $50.2 million. He revised his calculation, not because of discovered mistakes or a changed opinion regarding the necessity of interconnecting Merchant Plant 4, but because technological advancement meant that a $42.5 million component of the original interconnection plan would not be built. This declaration is highly credible because Houston was the Vice President of Transmission and Substation Operations for Reliant Energy HL & P, and nothing in the record rebuts his assertion that Reliant would not build the new transmission line. The Commission’s findings, inferences, conclusions, or decisions must be supported by substantial evidence from the record as a whole. Tex. Gov’t Code Ann. § 2001.174(2)(e) (West 2000). Although deference is due the agency’s findings of fact, the nature of Houston’s rebuttal, the lack of any contradiction or impeachment of his evidence, and the way in which the Commission derived the figure persuade us that no evidence supports the finding that the interconnection of Merchant Plant 4 would cost $107.3 million. The allocation of money based on non-existent needs for Merchant Plant 4 cannot be supported by evidence that would support allocation of a greater amount of money for other expenditures. The Commission could have allocated money for those expenditures, but it did not; its order plainly states that it allocated money for costs associated with Merchant Plant 4.

In light of Houston’s uncontroverted downward revision of his estimate of Merchant Plant 4’s interconnection costs, no reasonable basis existed to use the original estimate in determining Reliant’s transmission system capital costs. We hold that the Commission’s inclusion of $107.3 million associated with interconnection costs for Merchant Plant 4 in its final order is not supported by substantial evidence. We sustain Gulf Coast’s first issue, reverse the inclusion of $107.3 million for the interconnection of Merchant Plant 4, and remand this issue to the Commission for further proceedings. See Tex. Gov’t Code Ann. § 2001.174(2)(E). On remand, the Commission must base its decision on the existing administrative record. See Texas Health Facilities Commission v. Nueces County Hospital Dist., 581 S.W.2d 768, 770 (Tex.Civ.App.-Austin 1979, no writ); First Sav. & Loan Assoc. v. Lewis, 512 S.W.2d 62, 64 (Tex.Civ.App.-Austin 1974, writ ref'd n.r.e.).

RETURN ON EQUITY

Gulf Coast and OPC raise several complaints on appeal regarding the determination of the return on equity. They argue that the consolidation of the determinations of the rates of return for several utilities into a single proceeding producing a generic return on equity for those utilities is not supported by law or substantial evidence. They argue that the generic proceeding was unlawful because the Commission did not make utility-specific determinations as required by PURA section 36.052. They also contend that the Commission’s failure to adjust the return on equity despite the substantial decline in the cost of capital results after the interim generic order but before the Reliant-specific order resulted in an inflated, unreasonable return on equity in violation of PURA section 36.051.

PURA requires that the Commission set a “reasonable” return on the utility’s invested capital. PURA § 36.051. In setting this return, the Commission “shall consider applicable factors, including: (1) the efforts and achievements of the utility in conserving resources; (2) the quality of the utility’s services; (3) the efficiency of the utility’s operations; and (4) the quality of the utility’s management.” PURA § 36.052. The Commission is not limited to considering these four factors; the statute merely includes these four factors among the applicable factors. See id. Although the Commission was required to consider the listed factors, it was not required to make “ultimate” findings on each issue. See Meier Infiniti Co. v. Motor Vehicle Bd., 918 S.W.2d 95, 100 (Tex.App.-Austin 1996, writ denied). “[T]he logical force of the findings of underlying fact must be such that the reviewing court can fairly and reasonably say that the underlying findings support the statutorily required criteria.” Id.

Return on equity is among the issues in the utility restructuring process that the Commission concluded could be most efficiently and appropriately resolved in a generic proceeding involving all the newly unbundled TDUs. 11 The Commission found underlying similarities among the unbundled TDUs — including the level of regulatory oversight and comparable levels of risk — and concluded that a generic proceeding generating a generic return on equity for the TDUs was appropriate. In the generic proceeding, the Commission left open the possibility of individualized returns on equity, but in this proceeding applied the generic return for Reliant. The Commission considered „ the recommendation by some parties in a non-unanimous stipulation and agreement of a 10.75% return as a reasonable starting point. Texas Pub. Util. Comm’n, Generic Issues Associated with Applications for Approval of Unbundled Cost of Service Rate Pursuant to PURA 39.201 and Public Utility Commission Substantive Rule 25.3Jf.lp, Docket No. 22344, Order No. 42, at 11, 2000 WL 33599088 (Dec. 22, 2000). This number lies in the middle of the ranges of reasonable returns on equity admitted into evidence. Id. The Commission decided to add 0.5% to the stipulated return on equity to account for factors such as the undetermined effect of higher debt (based on the adoption of 60% debt and 40% equity for capital structure in this proceeding) on the TDUs’ investment-risk ratings and the risk premium recalculation recommended by Commission staff witness Martha Hinkle. Id. Accordingly, the Commission approved a return on equity of 11.25%. Id.

Propriety of a generic proceeding

We first consider the broad issue of whether the Commission is authorized to conduct a generic proceeding to determine a single issue common to several utilities. The Commission has “the general power to regulate and supervise the business of each public utility within its jurisdiction and to do anything specifically designated or implied by this title that is necessary and convenient to the exercise of that power and jurisdiction.” PURA § 14.001. The Commission has the power to call and hold hearings and to adopt and enforce rules reasonably required to exercise its power and jurisdiction. See id. §§ 14.002, .051, .052. OPC argues that, because the Commission did not enact a procedural rule for considering an issue common to several utilities in a generic proceeding, the Commission could not do so. But the Commission’s rules permit severance of proceedings or issues if the severance would serve the interest of efficiency or prevent unwarranted expense and delay. 16 Tex. Admin. Code § 22.34(b) (2004). Proceedings may be consolidated if the proceedings involve' common questions of law or fact and consolidation would be more time and cost efficient. Id. § 22.34(a). OPC seizes on the omission of the word “issues” from the consolidation provision to argue that issues cannot be consolidated. We are not persuaded by this distinction. If Issue A from one docket can be heard with Issue B from another docket when the two dockets are consolidated, we see no reason that the Commission could not, after severing issues A and B from their respective dockets, consolidate its consideration of issues A and B. For purposes of rule 22.34, the severed issues in essence become “proceedings” of their own that can be consolidated.

Even.if rule 22.34 were insufficient on its own terms, this Court has recognized that the Commission has the power to control its own docket. City of El Paso v. Public Util. Comm’n, 839 S.W.2d 895, 926 (Tex.App.-Austin 1992), rov’d in part on other grounds, 883 S.W.2d 179 (Tex.1994). In that case, despite the absence of a rule expressly allowing severance, we found authority to sever in the Commission’s broad regulatory power and attendant powers necessary for efficient conduct of adjudicative hearings. Id.; see also Public Util. Comm’n v. Southwestern Bell Tel. Co., 960 S.W.2d 116, 119 (Tex.App.-Austin 1997, no pet.). In El Paso, we wrote, “Any other result would defeat the legislative intent in delegating duties to the Commission for more efficient administration.” 839 S.W.2d at 926. This Court has recognized that the legislature delegated to the Commission broad powers:

A delegation of power to an administrative agency, in such broad and general terms, implies a legislative judgment that the agency should have the widest discretion in conducting its adjudicative proceedings, including a discretion to make ad hoc rulings in specific instances, within the bounds of relevant statutes and the fundamentals of fair play.

Southwestern Bell, 960 S.W.2d at 119. The agency must have the flexibility to adjust to the variety of incidents encountered in particular contested cases. Id. We conclude that the Commission has the authority to consolidate issues severed from different dockets into a single generic proceeding.

OPC and Gulf Coast argue that the record lacks substantial evidence to support the conclusion that there are sufficient similarities among the TDUs to support consolidation of their return-on-equity proceedings into a single proceeding producing a single generic return on equity. It is undisputed that the TDUs are all new, spawned by the unbundling required by the statutes governing the move from a regulated environment to a competitive one. See PURA § 39.051(b). Thus, as independent companies, they share the same history — none. Commission staff witness Martha Hinkle testified 12 that TDUs share a reduction of risks by being separated from other aspects of the power industry. They would not be subject to risks from asset concentration due to owning generation facilities or risk that generating plants would run into regulatory difficulties. Further, their cash flow would be more stable than an integrated utility’s because they would not be subject to fuel-price fluctuations and because transmission and distribution are essential services. She testified that all TDUs would have risk ratings in the lower half of the business position risk assessment range (1 to 4 on a 10-point scale), in contrast to the existing integrated companies which were in the mid-range of risk (4 to 7; Reliant Energy had a rating of 6). She testified that TDUs would have similar risk profiles because of similarities in their business and regulatory risks, although differences would persist in the types of territory served, growth rates, and corporate structures. She opined that the differences in risk profiles “will not lead to the disparate bond ratings that have resulted from the risk associated with providing generating facilities.” The Commission even left open the possibility of setting individualized rates of return, and even set one TDU’s return determination into a separate docket; however, the parties challenging the use of the generic rate in this appeal did not cite any evidence in their briefs supporting using an individualized rate for Reliant’s TDU beyond promoting a different interpretation of Hinkle’s testimony— particularly her admission that differences exist among the TDUs. The Commission’s decision to use a generic proceeding to set a return on equity for all the TDUs was supported by substantial evidence.

Appropriateness of the generic return for Reliant’s TDU

Gulf Coast and OPC further contend that the Commission acted arbitrarily and capriciously by using the generic return on equity for Reliant without considering the utility-specific factors listed in PURA section 36.052: the TDU’s efforts and achievements in conserving resources, the quality of its services, the efficiency of its operations, or the quality of its management. Although Reliant argues that evidence on these issues was before the Commission, in its brief the Commission concedes that it did not consider the listed factors; instead, the Commission argues that the factors relate to a utility’s historical practices, which the newly unbundled TDU does not have. Gulf Coast and OPC contend that the statute requires the Commission to consider these factors. See PURA § 36.052. They also note that the TDU has a history as part of Reliant that could be examined to assess the unbundled TDU’s prospects regarding the listed factors.

In determining whether an agency act or omission is arbitrary and capricious, a reviewing court must ascertain whether the agency abused its discretion by basing its decision on legally irrelevant factors, or by omitting to consider factors that the legislature statutorily directs the agency to consider. Consumers Water, Inc. v. Public Util. Comm’n, 774 S.W.2d 719, 721 (Tex.App.-Austin 1989, no pet.) (citing Gerst v. Nixon, 411 S.W.2d 350, 360, n. 8 (Tex.1966)). An agency is not required to make ultimate findings as to each factor listed in the statute. See Meier, 918 S.W.2d at 100. The factors listed in section 36.052 are not exclusive. See Cities of Abilene v. Public Util. Comm’n, 854 S.W.2d 932, 942 (Tex.App.Austin 1993), rev’d, in part on other grounds, 909 S.W.2d 493 (Tex.1995). In its order adopting the generic return on equity, the Commission stated that it considered the following factors: (1) the levels of business and financial risk; (2) the Commission’s decisions in the rate design phase of this case; (3) the need to maintain reasonable rates; (4) the need for new transmission capacity; (5) the maintenance of adequate reliability standards; and (6) the ■companies’ ability to attract new capital. Reliant Order at 24. The Commission also noted that this proceeding was to set rates for the period of transition to a more competitive energy market.

We conclude that the Commission did not act arbitrarily or capriciously by not considering the factors listed in 36.052. The evidence and findings that the TDUs have a degree of uniformity supports the decision not to explore the issues in a utility-specific matter in every case (although the Commission’s generic order did leave open that possibility). But the stronger support comes from the fact that these TDUs are creatures born of the transition to competitive markets. The ew idence and finding that the new TDUs have reduced risk profiles supports a conclusion that they are different from when they were part of a larger utility. More critically, the statutory mandate creating the TDUs made it impossible to assess the TDUs’ conservation efforts and achievements and the quality and efficiency of their services, operations and management because the TDUs have no record as standalone entities. Although this is not a direct statutory conflict — i.e., PURA section 36.052 does not prevent unbundling— the unbundling mandated by PURA section 39.051 has rendered impossible a duty created by PURA section 36.052; in such instances, Chapter 39 provisions control. See PURA § 39.002. We conclude that, under the circumstance of the inception of the TDUs as part of the transition to competitive markets, the Commission did not err by failing to consider nonexistent factors in calculating a reasonable return on equity.

Failure to reopen record to adjust return for falling interest rates

Gulf Coast further complains that the return on equity of 11.25% is not reasonable because it fails to account for the Federal Reserve Board’s repeated reductions of short-term interest rates between December 18, 2000 (the date of the Conn mission’s decision in the generic proceeding setting the return on equity at 11.25%) and October 3, 2001 (the date of the final order in the Reliant-specific case). Gulf Coast reiterates on appeal its claim made in its motion for rehearing below that the Commission should have held another hearing to determine the effect of the interest rate changes on the reasonableness of the return on equity.

Whether to reopen an administrative record to allow additional evidence is generally a matter left to the discretion of the Commission. Lake Medina Conservation Soc’y v. Texas Natural Res. Conservation Comm’n, 980 S.W.2d 511, 518-19 (Tex.App.-Austin 1998, pet denied); City of El Paso v. Public Util. Comm’n, 609 S.W.2d 574, 578 (Tex.Civ.App.-Austin 1980, writ ref'd n.r.e.). In a similar case under federal administrative law involving a five-year gap between receipt of evidence and the decision, the Supreme Court held that the delay did not make the record too stale to support an administrative decision, stating:

Administrative consideration of evidence ... always creates a gap between the time the record is closed and the time the administrative decision is promulgated. If upon the coming down of the order litigants might demand rehearings as a matter of law because some new circumstance has arisen, some new trend has been observed, or some new fact discovered, there would be little hope that the administrative process could ever be consummated in an order that would not be subject to reopening. It has been almost a rule of necessity that rehearings were not matters of right, but were pleas to discretion. And likewise it has been considered that the discretion to be invoked was that of the body making the order, and not that of a reviewing body.

Bowman Transp., Inc. v. Arkansas-Best Freight Sys., Inc., 419 U.S. 281, 294-95, 95 S.Ct. 438, 42 L.Ed.2d 447 (1974) (quoting Interstate Commerce Comm’n v. Jersey City, 322 U.S. 503, 514-515, 64 S.Ct. 1129, 88 L.Ed. 1420 (1944)). In Bowman, the Supreme Court noted it had found an abuse of discretion for the failure to reopen the record of a proceeding to account for the “economic metamorphosis” wrought by the Great Depression. Bowman, 419 U.S. at 295, 95 S.Ct. 438 (citing Atchison, T. & S. F. Ry. Co. v. United States, 284 U.S. 248, 52 S.Ct. 146, 76 L.Ed. 273 (1932)). But the Supreme Court also noted several cases of relatively similar vintage in which it did not find an abuse of discretion. See Bowman, 419 U.S. at 295, 95 S.Ct. 438.

Reliant argues that other factors affecting rates also changed in the same period (they assert that natural gas prices affecting excess mitigation credits declined by 50%); Reliant also notes that Gulf Coast did not seek to supplement the record with the interest rate changes until after the October 2001 order. Gulf Coast responds that the interest rate drop was unprecedented, that no other party requested to revisit any other component of the rates, and that their October 29, 2001 motion for rehearing gave the Commission plenty of time to hold a hearing, based on the previous one-day return-on-equity hearing on November 6, 2000.

We find the reasoning in Bowman persuasive. The parties’ arguments regarding various changes in market forces show why the decision on whether to reopen the record is committed to the Commission’s sound discretion. Interest rates are subject to change and the failure of rates set through a regulatory process to keep pace with interest rate changes is inherent in the regulatory process. In this case, utility rates had to be set by January 1, 2002. Given the complexity of the rate-setting hearing and the imminence of the period for which rates were to be set, we cannot find that the Commission abused its discretion in not reopening the record to take into account changes in interest rates that occurred after the administrative record had closed.

We overrule Gulf Coast’s second issue and OPC’s third issue.

REASONABLE AND NECESSARY EXPENSES

In setting the Reliant TDU’s rates, the Commission considered the company’s reasonable and necessary expenses. Reliant complains that the Commission’s calculation of its consolidated tax savings and application of a generic escalator to its vegetation control costs made its rates too low. Gulf Coast assails the Commission’s failure to use surplus insurance funds to reduce the TDU’s rates.

Consolidated Tax Savings

Reliant argues that the method the Commission used to determine the TDU’s fair share of consolidated tax savings constitutes impermissible retroactive ratemak-ing, erroneously includes the losses of companies that will not be eligible to file a consolidated tax return with the TDU, and arbitrarily and capriciously fails to follow methods used in similar cases

Federal income tax is one of the “reasonable and necessary expenses” that a utility incurs in providing service to ratepayers. See 16 Tex. Admin. Code § 25.231(b)(1)(D); GTE-Southwest, 901 S.W.2d at 409. When considering income taxes as part of a utility’s cost of service, PURA requires the Commission to deter mine a utility’s “fair share” of the tax benefits that result when its parent company files a consolidated tax return:

(a) Unless it is shown to the satisfaction of the regulatory authority that it was reasonable to choose not to consolidate returns, an electric utility’s income taxes shall be computed as though a consolidated tax return had been filed and the utility had realized its fair share of the savings resulting from that return, if:

(1) the utility is a member of an affiliated group eligible to file a consolidated income tax return; and

(2) it is advantageous to the utility to do so.

PURA § 36.060(a). The legislature has granted the Commission broad discretion in determining a utility’s fair share of the savings arising from the filing of a consolidated tax return by the parent company. CPL, 36 S.W.3d at 554. A consolidated return can result in tax savings because one company’s income may be offset by the losses of affiliated companies, reducing the tax liability for the collective affiliated group. See id. at 555.

Reliant argues that the Commission’s application of the consolidated tax savings adjustment in this case constitutes impermissible retroactive ratemaking because it adjusts for tax savings over a previous 15-year period — years in which the Commission had already apportioned fair shares of the tax savings. Utility rates are set prospectively, and the Commission “may not set rates that allow a utility to recoup past losses or refund excess utility profits to consumers.” Id. at 554; see also Tex. Const, art. I, § 16. Reliant argues that the orders in the previous rate-setting cases are final and cannot be re-litigated in this case.

We discussed and rejected Reliant’s arguments when we reviewed the Commission’s similar analysis in CPL. See 36 S.W.3d at 554-57. We concluded then that, “[a]s long as the Commission is not trying to recoup past savings, but only trying to recover today’s benefit from those past savings, its calculation of consolidated tax savings does not constitute retroactive ratemaking.” 13 Id. at 557. Reliant’s attempt in its appellant’s brief to distinguish the discussion in CPL about retroactive ratemaking as dicta fails; the issue was central to the resolution of the appeal in CPL. See id. at 556-57. We decline Reliant’s request that we overturn our decision in CPL. The reasoning in that case applies to the arguments and evidence presented in this case. We overrule Reliant’s first issue.

Reliant also argues that its TDU should not be subject to the adjustment because some of the companies whose losses were used to calculate the Commission’s consolidated tax savings adjustment will not be affiliated with the TDU during the forecast year of 2002, and therefore would not be eligible to file a consolidated tax return with the TDU. Reliant argues that the losses of the non-affiliated companies should be deducted from the amount used to calculate the consolidated tax savings adjustment. However, the TDU’s affiliations during the forecast test year are not relevant to the consolidated tax savings calculation because, as discussed in CPL, the Commission adjusted forecast rates based on the value of a benefit that the Commission has determined the TDU possesses. See CPL, 36 S.W.3d at 555-57. It is a calculation based, not on affiliations in 2002, but on an economic advantage that the Commission has determined that the TDU possesses and from which ratepayers are entitled to benefit in the test year. See id. Reliant’s complaint fails.

Reliant further argues that the adjustment approved by the Commission is arbitrary and capricious because it differs from the treatment of other utilities ordered by the Commission in other proceedings. We must reverse the Commission’s order if the Commission’s action was arbitrary or capricious. See Tex. Gov’t Code Ann. § 2001.174(2)(F); Public Util. Comm’n v. Gulf States Utils. Co., 809 S.W.2d 201, 210-11 (Tex.1991). An agency decision is arbitrary when its final order denies parties due process of law, see Lewis v. Metropolitan Sav. & Loan Ass’n, 550 S.W.2d 11, 16 (Tex.1977), or when it fails to follow the clear, unambiguous language of its own regulations. Gulf States, 809 S.W.2d at 207; Power Res. Group, Inc. v. Public Util. Comm’n, 73 S.W.3d 354, 358 (Tex.App.-Austin 2002, pet. denied). A court reviewing a decision for arbitrariness should consider all relevant factors and may not substitute its judgment for that of the agency. See Gulf States, 809 S.W.2d at 211. Our review is limited to determining whether the administrative interpretation “is plainly erroneous or inconsistent with the regulation.” Id. at 207.

Reliant complains that the Commission’s decisions in this case regarding the consolidated tax savings adjustment differ from its treatment of other utilities in similar circumstances in three areas. Reliant complains that the Commission applied a gross-up factor 14 to its tax savings amount, despite not applying the same factor in CPL; thus, instead of adjusting for $19.1 million of tax savings, the Commission “grossed-up” that figure to $29 million. Reliant complains that the Commission failed to distinguish between companies and divisions as it did in CPL; Reliant contends that, as a result, the Commission overstated the former affiliates’ losses because some of the losses occurred in divisions and should have been used to offset profits from other divisions in the same company, thus drastically reducing the amount of tax savings the Commission should have found that the TDU accrued. Finally, Reliant complains that the Commission used the tax savings to reduce the TDU’s operating expenses instead of reducing the rate base, as the Commission did in a similar proceeding involving TXU; Reliant argues that the different treatments reduced its income by nine times more than if it had been treated like TXU.

Two of Reliant’s complaints rely on a comparison of the Commission’s treatment of Reliant to its treatment of other utilities in different proceedings. But our review is restricted to the record before us, and an accurate comparison of this case to the earlier cases is not possible without at least relevant portions of those records or agreed statements of the circumstances of and arguments made in those cases. Tex. Gov’t Code Ann. § 2001.175(e) (West 2000); Public Util. Comm’n v. GTE-SW, 833 S.W.2d 153, 159 (Tex.App.-Austin 1992), rev’d on other grounds, 901 S.W.2d 401 (Tex.1995). Reliant concedes that those records are not before us but argues that such records would be unwieldy and possibly inadmissible in this proceeding. Rendering those concerns moot, however, is the inaccuracy of the premise underlying Reliant’s argument that the Commission’s different treatment of other utilities in other cases necessarily means that its treatment of Reliant in this case is arbitrary and capricious. Even if we had the records of the other proceedings before us, the question would remain whether the arbitrariness and capriciousness manifested itself in this case, the other case, neither, or both. In short, regardless of the Commission’s actions in other cases and the presence of the records of those cases, we must assess the reasonableness of the treatment of Reliant by the record in this case.

The only evidence in this record supports application of the gross-up factor. OPC witness Candice Romines testified that the gross-up factor was necessary to adjust for the effect of taxes. Although Reliant disagrees vehemently with this calculation as part of what it deems retroactive ratemaking, it points to no evidence disputing her testimony or showing that application of a gross-up factor is inherently arbitrary or capricious; as to the gross-up factor, Reliant argues only that its application here is an impermissible divergence from the procedure used for CPL in its proceeding. 15 The record in this case, however, does not show that the divergence is unjustified. Romines’s testimony in this case supports the use of the gross-up factor and makes the Commission’s use of the factor in this case neither arbitrary nor capricious.

Reliant next complains that the Commission unfairly calculated Reliant’s consolidated tax savings adjustment by considering losses at the division-wide level rather than at the company-wide level, as it had done with CPL. Reliant argues that certain divisions reported losses that were offset by gains in other divisions within the same company, but that the Commission’s failure to combine these divisions’ gains and losses caused the Commission to overstate the size of the overall business entity’s total losses and, hence, the size of the necessary adjustment. OPC argues that differences in corporate structure between Reliant and CPL make a strict application of the CPL method not feasible; OPC’s assertion is undisputed, although no party cites evidence of the asserted differences in corporate structure. The record does not demonstrate that calculation of gains and losses at the division level, rather than the company-wide level, is arbitrary or capricious in this case.

Similarly, the record does not weigh against the use of the consolidated tax savings adjustment amount to reduce Reliant’s operating expenses. Without pointing to evidence in this record of TXU’s treatment, Reliant complains without contradiction that the Commission reduced TXU’s rate base instead of its operating expenses, resulting in disparate impact on the tax adjustments of the two companies. Reliant calculates that, assuming a $30 million dollar adjustment, applying the adjustment to its operating expenses reduced its income by about $30 million, while using it to reduce rate base as was done with TXU would have reduced Reliant’s income by only $3 million. Reliant does not argue and points to no evidence in this record showing that applying the adjustment to operating expenses is, in itself, arbitrary or capricious. Assuming arguendo that TXU and Reliant were treated differently does not establish that the treatment of Reliant was arbitrary or capricious. Perhaps the treatment of TXU is improper, or perhaps neither is improper. We conclude that Reliant has not shown error on this record.

We overrule Reliant’s second issue.

Vegetation Control Costs

In its third issue, Reliant argues that the Commission’s decision not to exempt its vegetation control expenses from the generic cost escalation factor is not supported by substantial evidence and is arbitrary and capricious. To set TDU rates based on estimated expenses, the Commission took expenses from the test year of 1999, then applied four generic escalation rates to different classes of expenses; the escalators permitted an overall increase in expenses of about 1.8% per year. The Commission exempted some expenses from the limitation of this generic rate, however, because they were “new” expenses created by unbundling; these new expenses include demand-side management expenses, transmission access charges, and corporate restructuring costs. See Tex. Pub. Util. Comm’n, Generic Issues Associated with Applications for Approval of Unbundled Cost of Service Rate Pursuant to PURA § 39.201 and Public Utility Commission Substantive Rule 25.311, Docket No. 22344, Order No. 25, 2000 WL 33599088 (Aug. 24, 2000) (hereinafter, Generic Order No. 25).

Reliant argues that the Commission also should have excluded'vegetation control costs from the generic escalation rate because those costs will rise sharply due to higher reliability standards set by the legislature and implemented by the Commission. See generally PURA § 38.005; 16 Tex. Admin. Code § 25.52(f) (2004). Reliant argues that the new standard quintuples the previous standard 16 and effectively doubles even Reliant’s heightened internal standard. Reliant argues that the Commission acted arbitrarily and capriciously and without support in the record by treating the additional vegetation control costs differently from other new charges, such as demand-side management costs. Reliant argues that the Commission’s decision violated the requirement that utilities have a reasonable opportunity to recover a reasonable return on their investments that exceeds their expenses. See PURA § 36.051.

Before reviewing the Commission’s order, we must determine the scope of our review of the record. The parties, including Reliant, cite evidence and testimony from both the generic proceeding and the Reliant-specific proceeding in support of their arguments regarding the Commission’s decision to apply generic escalators. Reliant, however, argues in its appellant’s reply brief that evidence adduced in the Reliant-specific proceeding after the Corn- mission’s August 2000 order in the generic proceeding was not before the Commission when it made the August 2000 order. This argument is especially compelling because of the following language in the August 2000 order:

[T]he resolution of an issue in this generic proceeding is to be applied in each utility’s UCOS proceeding. Issues that are resolved in this proceeding will not be litigated or re-litigated in the utility-specific UCOS cases other than as specified by an order issued in this proceeding.

Generic Order No. 25 at 1. The scope of our review of the August 2000 order on use of the generic escalators will not include evidence or testimony from the Reliant-specific proceeding that was admitted after the August 2000 order. 17

Reliant points to the testimony of James Brian to support its claim that the generic escalator would be ineffective to compensate for increased vegetation control. Brian argued against the use of generic escalators. He argued that, even if generic escalators are used, utilities should be allowed to show that the generic escalators would not account for some major expenses because either the expenses did not exist in the test year or would increase from the test year at a rate that exceeded the escalator rate. Brian testified that Reliant’s tree-trimming expenses would increase $11.6 million to comply with heightened reliability standards, but he did not provide a context for that increase — ie., show what percentage increase that would be. 18 He calculated that Reliant’s actual expenses for 2002 would outstrip the expenses estimated by using a single generic escalator (an approach not adopted by the Commission) by more than $67.7 million overall: $393.9 million actual expenses compared to $326.2 million escalated expenses.

Defenders of the generic escalators cite evidence showing that generic escalators can account for variances in the actual rate of increase of expenses that existed during the test year. Steven Andersen testified that productivity increases could offset some increased expenses, and cautioned against making exceptions to the escalator for specific expenses. He posited that increased tree trimming could result in fewer power outages, thus decreasing outage-related expenses. Defenders of the generic escalators argue that Reliant increased its tree-trimming expenses in anticipation of the new reliability rules, but they rely on evidence from the Reliant-specific proceeding. Commission employee Brian Al-mon testified that demand-side management expenses and transmission access charges should be excluded from the generic escalator because both were modified by statutes effective after the end of the 1999 test year and would have a “significant impact on the expenses of the utilities.”

Reliant argues that the Commission’s failure to exclude the increased vegetation-control costs from application of the generic escalator is arbitrary and capricious because it fails to consider the utility’s actual operating expenses and thereby fails to allow Reliant to recover those expenses. Reliant dismisses the explanation that the escalator accounts for some costs being overrecovered and others being underre-covered, noting that the Commission excluded demand-side management expenses, transmission access charges, and corporate restructuring costs-the first two of which, at least, also were among Reliant’s expenses before unbundling. Reliant contends that the Commission acted arbitrarily and capriciously by excluding those expenses, but not the vastly increased vegetation control expenses.

We conclude that the record supports the Commission’s use of generic escalators. There was evidence that generic escalators could account generally for TDUs’ increased expenses. There was evidence that increased expense of one type that exceeded the escalator rate could nevertheless be balanced by increases in productivity or decreases in other types of related expense. We are cited to no evidence from the generic proceeding supporting Reliant’s argument that the Commission acted arbitrarily and capriciously by applying the escalator to tree-trimming costs, but excluding demand-side management expenses. The Commission is the sole judge of the weight to be accorded the testimony of each witness. CPL, 36 S.W.3d at 561. The agency may accept or reject the testimony of witnesses or may accept part of a witness’s testimony and disregard the remainder. Id. Therefore, the agency was free to accept its staff witness’s testimony regarding what expenses should be exempted from the generic escalation factor and reject Reliant’s witness’s testimony regarding its increased tree-trimming costs and the necessity that those expenses be exempted from the generic escalation factor. Substantial evidence supports the Commission’s order. We overrule Reliant’s third issue.

NEIL Surplus

Gulf Coast complains that the Commission erred by rejecting the ALJ’s recommendation and failing to reduce the Reliant TDU’s rates based on its share of surplus funds from the Nuclear Electric Insurance Limited (“NEIL”) funds. In the order, the Commission describes NEIL as “a mutual enterprise founded, controlled, and operated by utilities owning nuclear generating capacity that insures against losses related to nuclear generation plants.” See Reliant Order at 59. The Commission went on to write the following:

In a typical year, NEIL pays out a portion of its underwriting and investment income in the form of distributions to its members. These distributions are essentially rebates against the prior year’s premiums. The amount that is not paid out to members as distributions accrues as surplus and is retained by NEIL. NEIL is required to have a reserve in the amount necessary to cover two full policy limit losses in one year. At the end of 1999, the NEIL surplus stood at a total of $4.1 billion. While the surplus is not actually paid out to the members, NEIL tracks each member’s share, which is known as the member account balance (‘MAB’). Reliant’s MAB at the end of 1999 was $4,850,718.

Id. The ALJ recommended that the Commission require Reliant to calculate its MAB at the end of 2001, establish a regulatory asset that remains with the TDU, and use those amounts to moderate rates in future transmission and distribution rate proceedings regardless of whether NEIL distributes the MAB to Reliant. Although Reliant did not file an exception to the recommendation, the Commission reversed the ALJ. The Commission opted to keep the NEIL assets with the generation company because they are primarily generation-related, and use them to increase the value of the generation plant and reduce stranded costs at the 2004 true-up proceeding. Id. at 60.

Gulf Coast complains that the Commission should have followed the ALJ’s proposal to credit ratepayers with Reliant’s MAB because Reliant’s NEIL premiums have been paid for and supported by ratepayer-derived funds. Gulf Coast argues that NEIL assets are not transferrable, are not an asset affiliated with the generation plant, and will not be accounted for in the true-up. Gulf Coast contends that the Commission’s contrary decision is without support in the record.

The record provides support for the Commission’s decision. There was conflicting testimony. Gulf Coast’s witness, Michael Arndt, testified that Reliant’s MAB should be credited to ratepayers because Reliant’s NEIL premiums have been paid for and supported by ratepayer-derived funds. Arndt testified that ratepayers should benefit from the surplus accrued before generation became unregulated. Therefore, he recommended that Reliant’s share of the surplus be amortized over a short period and returned to ratepayers through transmission and distribution rates. He opined that, if NEIL would not release the funds, Reliant’s generation company should provide the funds. Reliant witness Brian testified that the surplus, did not actually belong to the constituent companies except as they remained members until the dissolution of NEIL. Brian explained that, after unbundling, whatever interest Reliant had in the MAB would be owned by Texas GENCO, Reliant’s generation company, not Reliant’s TDU. He also testified that ratepayers had benefitted from rates attributable to the premiums because the utility had been insured. But ratepayers of the TDU would be relieved of those costs in rates. Further, were there any NEIL benefit to Reliant, in excess of risk coverage, that benefit would show up in the stock valuation that would be used to determine the Reliant generation company’s actual stranded costs under PURA. Therefore, ratepayers still had a chance to reap a benefit from the MAB.

In its order, the Commission stated that it agreed with Reliant that ratepayers have received benefits from the NEIL premiums through reduction of financial risk from catastrophic losses at the nuclear plant. Ratepayers have also received credits for these rate expenses through NEIL distributions. The Commission stated that the value of the MAB will be taken into account when the generation plant is valued and accounted for in the 2004 true-up proceeding for stranded costs; the Commission stated that the benefits from NEIL insurance should increase the value of the nuclear plant, thereby decreasing any stranded costs. Therefore, the Commission determined that NEIL MAB assets should remain with the generation company. Although reasonable minds could differ, the Commission’s decision is reasoned and supported by substantial evidence in the record. Gulf Coast’s fourth issue is overruled.

RATE DESIGN

OPC challenges three elements of the rate design. OPC complains about the transmission cost recovery factor, a rate design feature that allows TDUs to pass through to retail electric providers any changes to wholesale transmission rates that TDUs charge retail electric providers as billing agents for all interconnected wholesale transmission providers in Texas. OPC also complains that the Commission included an amount in rates for a minimum cost necessary to distribute some electricity to each customer. OPC also challenges the rate of escalation for coal fuel cost estimates and the rate of reduction of the estimated capacity of generators.

Transmission Cost Recovery Factor

OPC argues by its first point of error that the Commission exceeded its authority in approving a transmission cost recovery factor (“TCRF”) containing an automatic flow-through mechanism. OPC contends that, although the Commission may authorize an electric utility to pass along wholesale rate increases automatically, the Commission’s approval of the TCRF instead allows the automatic adjustment of retail rates. OPC further challenges the implementation of the TCRF because it is inconsistent with the rate freeze instituted during the transition to deregulation. Last, OPC argues that the TCRF is bad policy.

Although the Commission cannot authorize an electric utility to pass through changes in fuel or other costs automatically (see PURA § 36.201), the Commission may provide for periodic adjustment of wholesale rates “to ensure timely recovery of transmission investment.” See id. § 35.004(d). The Commission may even approve “a change in wholesale transmission service rates during the freeze period.” See id. § 39.052(h).

The TCRF is a means to allow a distribution service provider (DSP) to adjust the rate it charges retail electric providers to account for changes in the transmission costs of the electricity it distributes to the retail electric providers. See 16 Tex. Admin.Code § 25.193(b). Consistent with the Commission’s authority to permit periodic adjustments of wholesale rates to account for changes in transmission rates (see PURA § 35.004(d)), the administrative code permits a transmission service provider (TSP) to “periodically revise its transmission service rates to reflect changes in the cost of providing such services” subject to the Commission’s filing requirements. See 16 Tex. Admin. Code § 25.192(g). “A TSP’s transmission rate shall remain in effect until the commission approves a new rate.” See id. § 25.192(b)(1). The Commission rules allow a DSP

to include within its tariff a TCRF clause which authorizes the distribution service provider to charge or credit its customer for the cost of wholesale transmission cost changes approved or allowed by the commission service to the extent that such costs vary from the transmission service cost utilized to fix the rates of the distribution provider.”

See id. § 25.193(b).

OPC argues that the TCRF is an impermissible automatic pass-through of transmission costs to retail customers, but the TCRF does not apply to retail sales. OPC’s argument that changes in the TCRF directly affect the retail price of electricity does not overcome the fact that the TCRF expressly applies only to the transaction between the DSP and the retail electric provider, which is not a retail transaction.

More important to the protection of retail customers, the adjustment of the transmission rates is not entirely automatic. While the TCRF is formulaic, it is derived from transmission rates that are approved by the Commission. See 16 Tex. Admin. Code §§ 25.192-.193 (2004). There are procedural prerequisites to the pass-along. The DSP must have a TCRF rate increase clause in its tariff. See id. The DSP’s rate increase is also subject to a Commission-approved formula. See id. § 25.193(c). The formula consists of many variables, one of which is the “new wholesale transmission rate approved by the commission by order or pursuant to commission rules.” See id. The Commission’s formula appears within the statutory authorization to facilitate wholesale rate in creases to “ensure the timely recovery of transmission investment.” PURA § 35.004(d) (emphasis added).

The record indicates that the Commission considered alternatives to the TCRF. It discussed options presented by Commission staff. See Reliant Order at 37-38. It noted the testimony of City of Houston witness Daniel, who contended the TCRF violated PURA because transmission costs were not specifically mentioned as an exception to the rule prohibiting the automatic pass through of a utility’s costs. Id. The Commission’s decision to follow its staffs recommendations is within its discretion. Despite OPC’s argument that the Commission’s decision is bad policy, we may reverse the agency’s decision only if the decision is “not reasonably supported by substantial evidence.” See Tex. Gov’t Code Ann. § 2001.174(2)(E). An agency’s decision is supported by substantial evidence if it is reasonable or rational. See City of El Paso v. Public Util. Comm’n, 883 S.W.2d 179, 185 (Tex.1994). The Commission was authorized to adopt the TCRF to adjust wholesale rates to ensure the timely recovery of transmission investments, and it did so in a reasoned- manner. We overrule OPC’s first point of error.

Reliant’s Minimum Plant Methodology

In point of error four, OPC argues that the Commission erred by rejecting the ALJ’s recommendation and adopting Reliant’s methodology for calculating its “minimum plant” — the investment needed to connect customers and provide for minimum usage — as part of its allocation of costs of service. The ALJ recommended that the Commission reject Reliant’s minimum plant classification of poles, towers, overhead lines, and transformers. OPC complains that the Commission rejected the ALJ’s recommendation and adopted Reliant’s calculations even though Reliant failed to file an exception to the ALJ’s recommendation and otherwise failed to explain what OPC describes as statistical and logical problems in the analysis underlying Reliant’s calculations. OPC asserts that the Commission did so impermissibly without explanation or support in the record.

The Commission can change, modify or vacate an ALJ’s order when the Commission determines that the ALJ misapplied or misinterpreted Commission rules or policies, applicable law or prior administrative decisions, or issued a finding of fact not supported by a preponderance of the evidence. See Tex. Gov’t Code Ann. §§ 2001.058(e)(1), 2003.049(g). The Commission must “state in writing the specific reason and legal basis for its determination” to depart from the ALJ’s order. See id. § 2003.049(h).

In its order, the Commission noted that the ALJ rejected Reliant’s minimum-plant methodology “because no other utility had employed it, and because the underlying plant study was ‘questionable.’ ” Reliant Order at 63. The Commission explained its rejection of the methodology as follows:

The Commission notes, however, that P.U.C. Subst. R. 25.344(h)(2)(B) requires that costs be allocated based on the methodology used in the utility’s last cost of service study unless determined otherwise by the Commission. Reliant used the minimum plant methodology in Docket No. 12065. Therefore, the Commission finds that the use of minimum plant methodology is appropriate.

Id. In its brief, the Commission points to evidence that supports its rejection of the ALJ’s recommendation. The Commission points to evidence from Reliant’s witness James Purdue that some costs should be allocated to consumers because those costs are specific to the provision of power to consumers irrespective of the amount of their usage. He testified that those costs included distribution equipment such as poles, conductors, and transformers. He noted that these costs were calculated in Reliant’s application. TIEC witness Jeffry Pollock agreed with Purdue. He stated that, although some portion of the distribution network costs could fairly be allocated based on usage or demand because increased demand would require more capacity, that portion of the basic infrastructure needed to supply power should be allocated to customers. He opined that assessing 14% of the net distribution plant costs as customer-related was not excessive, and flatly rejected the assertion'by OPC witness Clarence Johnson that all of the distribution network costs should be allocated as demand-related rather than customer-related.

In its reply brief, OPC does not challenge the Commission’s assertion that Purdue’s evidence provides substantial evidence on which the Commission’s could decide to reject the ALJ’s recommendation, but insists that the Commission’s explanation in its opinion for rejecting the ALJ’s recommendation is inadequate.

The Commission plainly rejected the ALJ’s recommendation because it was inconsistent with the Commission’s previous ruling approving Reliant’s minimum plant methodology. Although the Commission (and the ALJ) could depart from the previous ruling if there was reason to depart from the previous methodology, the record contains substantial evidence to support the Commission’s decision that no such reason existed. Accordingly, we affirm the Commission’s rejection of the ALJ’s recommendation. We overrule OPC’s fourth point of error.

3% Escalation Rate

In point of error five, OPC argues that the Commission erred in applying a 3% escalation rate to the price of coal for the years 2010 through 2026. OPC complains that the ALJ did not provide support for the finding that Powder River Basin coal is in significantly higher demand than coal from other regions and will escalate in price more rapidly than other coal. OPC complains that no evidence supported either the Commission’s use of 3% as the escalation rate or its change of the start date for the rate escalation to 2010 from the ALJ’s recommended 2011.

In the February 7, 2001 PFD, the ALJ recommended a 3% escalation rate. The ALJ’s finding was supported by Reliant witness Janie Mitcham, who testified that the popularity of low-sulfur coal from the Powder River Basin used by Reliant would likely increase regardless of overall coal demand, and that transportation bottlenecks would nevertheless limit the accessible supply. She attached to her testimony an excerpt from the Coal News, which predicted a 3%-5% increase in the cost of Powder River Basin coal. Several witnesses, including Randall Falkenberg, challenged Mitcham’s methodology, arguing for lesser escalation. Commission staff witness Jay Curtis testified that a 3% escalation rate was “appropriate because it is consistent with general price forecasts and is generally consistent with the escalation Reliant used for coal prices for the 2000-2009 period and for lignite prices for the post-2009 period.”

OPC’s assertion that the Commission changed the ALJ’s recommendation for the start date is mistaken. In Finding of Fact No. 20 of the PFD, the ALJ recommended that the 3% annual escalation begin in 2010. The record indicates that Reliant’s fuel contracts expire in 2009, thus making the year 2010 the logical year in which to begin applying a forecasted rate. The Commission adopted the ALJ’s rec ommendation that the escalation begin in 2010.

Rather than changing the ALJ’s PFD, the Commission adopted the ALJ’s decision, which was supported by substantial evidence. We overrule OPC’s fifth point of error.

0.5% Reduction in the South Texas Nuclear Power Plant and Coal Plant Capacity Factors

In points of error six and seven, OPC asserts that no evidence supports the ALJ’s decisions to adopt a 0.5% annual reduction in the capacity factor 19 for Reliant’s nuclear and coal power plant. In point of error six, OPC contends that no party requested that the Commission begin applying the reduction factor beginning in 2010 instead of 2011 as the ALJ proposed.

The ALJ’s decision to adopt a reduction factor for both the nuclear power plant and the coal powered plants has evidentiary support. Reliant witness William G. Rice testified that Reliant’s “older units” would begin to experience some degradation in their performance capability beginning in 2011. He estimated that the decline for coal and nuclear plants would be 1% annually. Reliant witness Ernie McWilliams agreed with Rice that degradation would occur due to increasing rates of random component failures. He asserted that the prime productive period of a generator ranges between 10 and 40 years. Rice testified that the relevant generators would be between 20 and 25 years old in 2011. Rice and McWilliams disagreed with OPC’s Falkenberg and the City of Houston’s Scott Norwood, who asserted that maintenance and advanced technology could offset any normal degradation in the capacity of power plants. There is substantial evidence for the adoption of a reduction rate.

The Commission’s decision to set the reduction rate at 0.5% is also supported by substantial evidence. OPC asserts that no witness advocated a 0.5% reduction rate. That is strictly correct. But some witnesses’ testimony supported a 1% reduction rate, which by definition encompasses the 0.5% rate adopted. The contrary evidence that degradation was avertable did not require a choice of between 0% or 1%; rather, the evidence that degradation would occur but was possibly avoidable supports the choice of any figure in that range. See City of El Paso, 883 S.W.2d at 186. Substantial evidence supports the Commission’s adoption of the ALJ’s proposed 0.5% degradation rate.

The Commission’s decision to change the beginning date for applying the degradation rate to the nuclear generator was consistent with evidence in the record and with a prior decision by the Commission. The Commission may depart from an ALJ’s recommendation if it does not properly apply previous Commission decisions. See Tex. Gov’t Code Ann. §§ 2001.058(e)(1), 2003.049(g). The ALJ recommended applying the degradation rate beginning in 2011. Despite the fact that this start date was derived from Reliant witnesses, Reliant excepted to the ALJ’s proposal because the 2011 start date differed from the 2010 start date that the Commission used for the Same degradation rate for the same generator in another proceeding. See Tex. Pub. Util. Comm’n, Application of Cent. Power and Light Co. for Approval of Unbundled Cost of Service Rate Pursuant to PURA 89.201 and Pub. Util. Comm’n Substantive R. 25.344, Dock et No. 22352 at 112, 2001 Tex. PUC LEXIS 110, at *205 (Oct. 2, 2001). In its order in this case dated the next day, October 3, 2001, the Commission agreed with Reliant’s exception and noted that, for the sake of consistency, the 0.5% reduction factor should begin in 2010 as it did in the CPL case, not in 2011 as the ALJ recommended in this case. See Reliant Order at 54-55. The change is permissible because it is supported by substantial evidence and brings the decision in this case into line with a prior agency decision. We overrule OPC’s points of error six and seven.

CONCLUSION

We conclude that the district court correctly affirmed the Commission’s order except as it concerns the costs of interconnecting Merchant Plant 4. We reverse that finding and related conclusions and remand to the Commission for further proceedings based on the existing administrative record.

1 . During proceedings before the Commission, Reliant Energy Houston Lighting and Power changed its name to Reliant. Although some documents filed with the Commission refer to the company as "Houston Lighting and Power” or “HLP,” we will refer to the company as "Reliant.”

2 . Appellants Magic Valley Electric Cooperative, Inc., Medina Electric Cooperative, Inc., Rayburn Country Electric Cooperative, Inc., and the City of Bryan are collectively referred to as "Consumer Owned Power Systems.”

3 . Tex. Util.Code Ann. §§ 11.001-64.158 (West 1998 & Supp.2004) (hereinafter "PURA”).

4 . This Court has discussed the components of the "rate base”:

The term "invested capital” is not made the subject of a specific definition in PURA although the term is said to be synonymous with the term "rate base,” see 16 Tex. Admin. Code § 25.231(c)(2) (2003); and, the "components” of invested capital are described broadly as "property used by and useful to the utility in providing service,” appraised based on original cost less depreciation. PURA § 36.053(a). In fixing an electric utility’s rates, the Commission exercised a statutory authority to separate and allocate "costs of facilities, revenues, expenses, taxes, and reserves" in arriving at rates that were just and reasonable. Id. § 36.055.

American Elec. Power Co. v. Public Util. Comm’n, 123 S.W.3d 33, 35 (Tex.App.-Austin 2003, no pet.).

5 . Stranded costs are "the positive excess of the net book value of generation assets over the market value of those assets....” PURA § 39.251(7). Under regulation, a utility could recover over time its prudently incurred costs of acquiring power-generation assets through rates approved by the Commission and paid by captive customers. See Reliant Energy, Inc. v. Public Util. Comm’n, 101 S.W.3d 129, 134 (Tex.App.-Austin 2003), rev'd in part sub nom. CenterPoint Energy, Inc. v. Public Util. Comm’n, 47 Tex. S.Ct. J. 672, 2004 Tex. LEXIS 540 (Tex. June 18, 2004). The Commission facilitated this cost recovery by incorporating depreciation expenses into approved rates. See Central Power & Light Co. v. Public Util. Comm’n, 36 S.W.3d 547, 553 (Tex.App.Austin 2000, pet. denied) (''CPL In a deregulated environment, however, competition could drive rates to levels so low that a formerly regulated utility would be unable to recoup its investments. Stranded costs represent that portion of the net book value of a utility’s generation assets not yet recovered through depreciation that has become unrecoverable in a deregulated environment. See City of Corpus Christi v. Public Util. Comm'n, 51 S.W.3d 231, 237-38 (Tex.2001); see also PURA § 39.251(7).

6 . PURA section 36.051 provides as follows:

In establishing an electric utility's rates, the regulatory authority shall establish the utility’s overall revenues at an amount that will permit the utility a reasonable opportunity to earn a reasonable return on the utility’s invested capital used and useful in providing seivice to the public in excess of the utility’s reasonable and necessary operating expenses.

(Emphasis added.) PURA section 36.053 provides as follows:

(a)Electric utility rates shall be based on the original cost, less depreciation, of property used by and useful to the utility in providing service.

(b) The original cost of property shall be determined at the time the property is dedicated to public use, whether by the utility that is the present owner or by a predecessor.

(c) In this section, the term "original cost” means the actual money cost or the actual money value of consideration paid other than money.

(Emphases added.)

7 . Brian’s testimony is dated December 2000. Whether the retirement plan remained over-funded into 2002 is not apparent from the record and is beyond the scope of this review.

8 . Though there apparently is no dispute that the Commission in other cases deducted underfunded amounts from the rate base, our review is restricted to the record before us. See Tex. Gov't Code Ann. § 2001.175(e) (West 2000). “We might accurately compare the circumstances in the company’s case with those in the earlier contested cases referred to only if we had before us the agency record compiled in those other cases.” Public Util. Comm’n v. GTE-SW, 833 S.W.2d 153, 159 (Tex.App.-Austin 1992), rev’d on other grounds, 901 S.W.2d 401 (Tex.1995).

9 . Although the quoted passage may be ambiguous in isolation regarding which "project” is estimated to cost $50.2 million, the context shows that the relevant project is the Merchant Plant 4 interconnection. Houston stated that new technology had increased the capacity of existing transmission facilities, rendering building the Greens Bayou-White Oak circuit (which he estimated would cost $42.5 million) unnecessary. Reducing the revised cost estimate of $92.7 million for Merchant Plant 4 by the $42.5 million cost of the unnecessary circuit leaves $50.2 million.

10 . To the extent that Reliant believed the Commission erred by not raising the general allocation of interconnection costs, it has failed to raise that error for our review.

11 . In order to develop a list of issues to be considered in this proceeding, the Commission directed all parties to file a list of issues to be addressed and to separately identify those issues generic to all unbundled cost of service filings. Parties were also encouraged to identify any issues that should not be addressed in the generic proceeding.

12 . There is some criticism that Hinkle testified in November, well after the Commission decided to determine the generic return on equity, and therefore could not supply evidence to support the decision to use the generic proceeding. However, the interim order authorizing the generic proceeding left open the possibility of individualized determinations. Because Hinkle’s evidence apparently helped dissuade the Commission from conducting such individual determinations, we can consider whether it supports the Commission’s decision to develop a generic rate.

13 . The methodology at issue is the time-value-of-tax-shield methodology. See CPL, 36 S.W.3d at 555-56. This formula initially calculates fifteen years’ worth of savings enjoyed by the parent company as a result of using its profitable affiliates’ gains to utilize the tax benefits of its unprofitable affiliates’ losses. Then, looking at the gains of each subsidiary standing alone, the Commission calculated which of those losses would have been offset by the subsidiary’s own gains during that fifteen-year period. The remaining losses— losses that would have had no value to the parent during the fifteen-year period if one of its affiliates had not earned profits against which those losses could be applied — were multiplied by the percentage that represents the affiliate's share of the parent company's overall profits during those years. The resulting figure was multiplied by the affiliate's long-term debt rate, resulting in the consolidated tax savings to the affiliate. Id. at 556.

14 . Reliant explains in its appellant’s brief that the gross-up factor is an adjustment made to negate the effect of taxes: "For example, a party who receives $100 will have to pay approximately $35 in taxes, and thus will have net revenues of only about $65 absent a gross-up. To ensure a net amount of $100 after taxes, the party must be allowed to recover approximately $154.”

15 . No party challenges Reliant's assertion that the gross-up factor was not used in CPL; the term "gross-up” does not appear in our CPL opinion. See generally, 36 S.W.3d at 547-72.

16 . Compare 16 Tex. Admin. Code § 25.52(f)(2)(A) (2004) (no distribution feeder shall be among system’s worst 10% for service interruption in consecutive years), with 23 Tex. Reg. 11921, 11930 (1998) (setting out former 16 Tex. Admin. Code § 25.51(g)(2)(C)) (no distribution feeder shall be among system's worst 2% for service interruption in consecutive years).

17 . This determination excludes from our review testimony in the Reliant-specific proceeding from witnesses such as Georgianna Nichols, Alan Ahrens, and Reginald Comfort. We note further that the testimony of Nichols and Comfort was admitted by the ALJ in the Reliant-specific proceeding as an offer of proof only, and that we have not been asked to review the correctness of the exclusion of that evidence.

18 . Evidence of the tree-trimming expenses was contained in the evidence from the Reliant-specific proceeding but, as discussed, such evidence is outside the scope of our review.

19 . The ALJ defined capacity factor as "the percentage of time that a plant (1) is available to run (i.e., is not undergoing a planned outage or forced outage); and (2) would be anticipated to run under economic dispatch principles.”