The United States and seven third-party defendants appeal the decision of the Court of Federal Claims holding that the methodology adopted by the Western Area Power Authority ("Western") to refund excess revenues collected under energy sales contracts was unreasonable, and ordering Western to calculate the appropriate refund to which plaintiffs Southern California Edison ("SCE") and Los Angeles Department of Water and Power ("LADWP") were entitled. In addition, the third-party defendants contend that the Court of Federal Claims improperly exercised jurisdiction over them. We conclude that the Court of Federal Claims had jurisdiction over the third-party defendants, and thus affirm-in-part. However, because the Court of Federal Claims erred in determining that Western's refund methodology was unreasonable, we reverse-in-part.

In 1928, Congress passed legislation authorizing the construction of a dam and hydroelectric power plant at Boulder Canyon, located on the Colorado River at the Nevada/Arizona border. See Boulder Canyon Project Act ("BCPA"), Pub. L. No. 70-642, 45 Stat. 1057 (codified as amended at 43 U.S.C. §§ 617-617v (1994)). The dam was later named the Hoover Dam. In order to fund this mammoth construction project, Congress decided to take advantage of the revenue generating potential of the power plant. To this end, Congress granted the Secretary of the Interior ("Secretary") the authority to enter into contracts for the sale of energy produced at the Hoover Dam, and required these contracts to generate sufficient revenue to ensure "payment of all expenses of operation and maintenance of said works [i.e., the dam and power plant] incurred by the United States, and the repayment, within fifty years from the date of completion of said works, of all amounts advanced [by the United States] for such works." 43 U.S.C. § 617c(b).

The Boulder Canyon Project Adjustment Act ("Adjustment Act"), Pub. L. No. 76-756, 54 Stat. 774 (codified as amended at 43 U.S.C. §§ 618-618p (1994)), enacted twelve years later, specified additional details surrounding the sale of energy produced at the Hoover Dam, such as the allocation of revenues collected from energy contracts. Additionally, Congress authorized the Secretary to "promulgate such regulations . . . as he may find necessary or appropriate for carrying out the purposes of [the Adjustment Act] and the [BCPA]." Id. § 618g. Finally, because the Hoover Dam began producing energy in 1937, the Adjustment Act specified that the duration of the contracts would be the fifty-year period from June 1, 1937 to May 31, 1987, and mandated that energy rates be set to repay construction, operation, and maintenance costs by the end of that contract period. See 43 U.S.C. § 618.

Pursuant to the Adjustment Act, the Secretary promulgated General Regulations ("Regulations") governing contracts for the sale of energy generated at the Hoover Dam. These Regulations established two categories of energy--firm energy and secondary energy. See Regulations, art. 3. Firm energy was described a pre-defined minimum amount of energy that the Secretary reasonably expected the Hoover Dam to generate each year. See id. For example, in the first year of operation, the Regulations defined the firm energy as the first 4,330,000,000 kilowatt-hours generated at the site. See id. In each subsequent year, the amount of energy designated as firm energy decreased by 8,760,000 kilowatt-hours per year. See id. Over the fifty-year duration of the contracts, the total amount of firm energy was to be 205 billion kilowatt-hours. Secondary energy was defined as the amount of energy generated in any given year in excess of firm energy. See id. Unlike the firm energy, the yearly availability of secondary energy was speculative and depended upon the vicissitudes of the Colorado River's hydrology. While the availability of secondary energy in any given year could not be predicted, the Secretary assumed, based upon expert analysis, that 40 billion kilowatt-hours of secondary energy would be produced over the contract period. See id.

Under the Regulations, the Secretary had the authority to set the rates for firm and secondary energy. See id., art. 6. These rate-setting determinations were to be made at five-year intervals, subject to annual adjustments to take into account the actual expenditures and costs incurred each year. See id., art. 14. In making rate-setting determinations, the Secretary assumed that the full amount of firm energy would be generated each year, and that over the fifty-year term of the contracts, the entire 40 billion kilowatt-hours of secondary energy would be available for sale. See id., art. 3. Based on these output assumptions, rates were set so as to ensure that the revenues collected from the sale of energy would "be sufficient, but not more than sufficient" to cover the costs associated with the construction and operation of the Hoover Dam and the appurtenant power plant. Id., art. 6.

Aside from the revenue requirements, a number of additional requirements had to be satisfied in setting energy rates. For example, the Regulations established a fixed mathematical relationship between the rates for firm and secondary energy, and a 0.2 mill per kilowatt-hour minimum charge for secondary energy. See id., art. 8. Also, the Regulations required uniformity of rates among all customers, that is, the rate for firm energy was required to be the same for all purchasers of firm energy, and the rate for secondary energy was required to be the same for all purchasers of secondary energy.

Finally, at the end of the fifty-year contract period, the Regulations provided for a final accounting of project costs and revenues collected. See id., art. 14. If revenues collected did not exactly offset the costs associated with the project, the Regulations mandated that "adjustments shall be made, through refunds or collections as the Secretary may find necessary to achieve consummation of the principle that the revenues shall be sufficient but not more than sufficient to" cover project costs. Id., art. 14(e).

Between 1941 and 1960, all of the parties to this litigation entered into contracts with the United States for the purchase of energy generated at the Hoover Dam. These contracts incorporated the terms and conditions of the Regulations described above, which were made a part of the contracts "as fully and completely as though set forth herein in length." Under these contracts, each customer was allocated a certain percentage of firm energy for purchase each year. See id., art. 4(a). In addition, three customers, SCE, LADWP, and Metropolitan Water District of Southern California ("Metropolitan") were granted options to purchase secondary energy. See id., art. 4(c). Metropolitan had the first opportunity to buy any or all secondary energy, with the balance made available for purchase by SCE and LADWP. For most of the contract period, SCE had the right to 55 percent and LADWP had the right to 45 percent of the secondary energy not bought by Metropolitan.

Following the mandate of the Regulations, the Secretary set the rates for energy at five-year intervals, and made annual adjustments to account for fluctuations in project operation and maintenance costs. In making these determinations, the Secretary relied on the output assumptions for firm and secondary power embodied in the Regulations. See id., art. 3. In other words, the Secretary assumed that in each year, revenues for the entire predicted amount of firm energy would be generated. Additionally, the Secretary assumed that over the fifty-year duration of the contracts, revenues for the entire 40 billion kilowatt-hours would be generated. During the first two decades of operation, the output assumptions proved remarkably accurate, and thus, the revenues collected kept pace with the project expenses. However, between 1954 and 1983, the amount of secondary energy generated was well below expectations, with no secondary energy at all produced in twenty-five of those twenty-nine years. By 1982, only 17.5 billion kilowatt-hours of the estimated 40 billion kilowatt-hours of secondary energy had been produced. Recognizing that the project would experience significant shortfalls if prevailing diminished water-flow conditions continued, Western1 altered its rate determination methodology. In particular, Western revised the output assumptions for secondary energy over the remaining five years of the contracts. Thus, Western set the energy rates for the final five-year cycle under substantially reduced assumptions for secondary energy output. This, in turn, increased rates for energy over this time period.

Beginning in 1983, however, water flow of the Colorado River watershed significantly changed. Rain and runoff from excessive snowfall resulted in widespread flooding in the western states. Consequently, the energy produced at the Hoover Dam increased dramatically, far surpassing the revised predictions for secondary energy output. This turned an expected shortfall into a large surplus. In fact, in 1986, one year prior to the expiration of the contract, the entire cost of the project had been met, and at the end of the contract period, Western had collected a $25 million surplus. Thus, pursuant to Article 14(e) of the Regulations, Western was required to make "adjustments" so that the revenues collected were not "more than sufficient" to cover the project costs.

Western considered a number of methodologies for allocating the refund to the various customers, including methodologies taking into consideration revenue generated from secondary energy purchases. Ultimately, Western decided to refund the excess to its customers based only on each customers' firm energy entitlement percentage as defined by Article 4(a) of the Regulations. Refund checks reflecting this methodology were distributed, and were accepted and negotiated by all customers.

In March 1991, SCE and LADWP formally notified Western that they disputed the refund methodology. They claimed that the refund should have been based on both the firm and secondary energy purchases, rather than firm energy purchases alone. Upon receiving this complaint, Western notified the other seven customers of SCE's and LADWP's objections, because in Western's view, any increase in the share of SCE and LADWP would come from the other customers. SCE, LADWP, and Western exchanged correspondence and met on several occasions in an effort to resolve the controversy. On January 6, 1995, Western attempted to get all interested parties together for a settlement meeting. However, only SCE and LADWP agreed to participate in the meeting.

In response to this impasse, the United States filed suit in the United States District Court for the District of Nevada on February 14, 1996, seeking a judicial declaration upholding the reasonableness of Western's refund methodology. A few days later, SCE and LADWP filed suit against the government in the Court of Federal Claims, alleging breach of contract. The government then joined the seven other customers as third-party defendants, so that it could seek reimbursement from them for any amounts paid to SCE and LADWP. The third-party defendants moved for dismissal of the contingent claims against them, alleging that the Court of Federal Claims lacked subject matter jurisdiction over these claims. This motion was rejected. See Southern Cal. Edison Co. v. United States, 38 Fed. Cl. 54 (1997). After this decision, the parallel district court proceedings were dismissed without prejudice.

All parties filed motions for summary judgment following discovery. SCE and LADWP argued that the allocation method adopted by Western breached their contracts and violated the Regulations, and that Western's actions were not entitled to judicial deference. The government, on the other hand, argued that the allocation method represented a reasonable interpretation of unclear regulatory language, and should be given deference by the reviewing court. The third-party defendants supported the government's argument for deference, and also argued that SCE and LADWP should be barred by laches, equitable estoppel, or the statute of limitations from asserting their claims. The Court of Federal Claims determined that while a reasonable allocation method would have been entitled to deference, the particular scheme adopted by Western was an unreasonable construction of the Regulations. See Southern Cal. Edison Co. v. United States, 43 Fed. Cl. 107, 118-119 (1999). In the court's view, it was improper to refund the over-collection only to the firm energy customers. The court determined that the refund should be allotted to those who paid for energy after the date at which the contract revenues fully met the project expenses, which occurred in mid-1986. Thus, the court ordered Western to refund the surplus based on both firm and secondary energy purchases made after this "break-even" date, and remanded the case to Western to determine the amount each customer was due under this methodology. The court also rejected the laches, equitable estoppel, and statute of limitations arguments of the third-party defendants.

Before reaching the merits of this appeal, we must determine whether the decision of the Court of Federal Claims, in which the court remanded this case to Western for a calculation of damages, is final and appealable. See 28 U.S.C. § 1295(a)(3) (1994); see also Palmer v. Barram, 184 F.3d 1373, 1377 (Fed. Cir. 1999) ("[I]t is always the duty of the court to determine its jurisdiction and satisfy itself that the appeal is properly before it."). In some circumstances, this court has held that an order remanding a matter to an administrative agency is not a final judgment, and thus is outside the scope of this court's jurisdiction. See Teledyne Continental Motors v. United States, 906 F.2d 1579, 1583 (Fed. Cir. 1990). In Teledyne, the Armed Services Board of Contract Appeals ("ASBCA") decided certain legal issues and then, while maintaining jurisdiction over the matter, remanded the case to the agency for further fact-finding and proceedings. See id. at 1582-83. Because these issues were still pending before the ASBCA, we held that the remand order was not final. See id. On the other hand, when a court determines that an agency's statutory or regulatory interpretation is unreasonable or contrary to law, and remands the matter to the agency to implement the court's alternative interpretation, we have held that the remand order is final and appealable. See Travelstead v. Derwinski, 978 F.2d 1244, 1248-49 (Fed. Cir. 1992);see also Sullivan v. Finkelstein, 496 U.S. 617, 619 (1990). Such orders are final and appealable because they represent "the final disposition of the proceeding respecting the Secretary's practice" and "compel[] action . . ., on remand, contrary to the Secretary's prior ruling." Travelstead, 978 F.2d at 1248. In this case, after rejecting Western's refund methodology, the court remanded the case to Western to "refund the surplus funds in proportion to each contractor's dollar contribution to that surplus as measured by energy purchases made [after mid-1986]." Southern Ca. Edison, 43 Fed. Cl. at 121. Western's only remaining task on remand was essentially ministerial--calculating the proper amount of damages based on the Court of Federal Claims' refund methodology. Like Travelstead, the remand order requires the agency to implement the court's regulatory interpretation, which is contrary to the agency's prior interpretation. Thus, we hold that the judgment of the Court of Federal Claims is final, and we can properly exercise jurisdiction over this appeal.

We must also determine whether the Court of Federal Claims properly exercised jurisdiction over the contingent claims filed by the government against the third-party defendants, an issue we review de novo. See Shearin v. United States, 992 F.2d 1195, 1195 (Fed. Cir. 1993). The third-party defendants assert that this exercise of jurisdiction was improper on two grounds. First, they claim that the statute on which the court relied as a jurisdictional basis for the third-party complaint, 41 U.S.C. § 114(b) (1994), does not apply to this type of contingent claim. Second, they contend that the contingent claims are time-barred because they were not filed within the six-year limitations period of 28 U.S.C. § 2415(a) (1994). Neither argument is persuasive.

Under § 114(b), the Court of Federal Claims has jurisdiction over contingent claims asserted by the government against third parties "for the recovery of money . . . paid by the United States in respect of the transaction or matter which constitutes the subject matter of such case." This section provides for jurisdiction in situations where the government is seeking the recovery of a sum of money disbursed to the wrong party. See Maryland Cas. Co. v. United States, 135 Ct. Cl. 428, 431 (1958). Thus, when the government, under a mistake of fact or law, erroneously pays one party money that rightfully should have been paid to another, § 114(b) allows the government to seek reimbursement from the erroneously paid party. See United States v. Rush, 804 F.2d 645, 647 (Fed. Cir. 1986).

The third-party defendants claim that because each customer has a separate contract with the government, the contingent suits against them do not constitute the same "subject matter" as the suits filed by SCE and LADWP, and thus are outside the scope of § 114(b). In support of their position, the third-party defendants point to Oliver-Finnie Co. v. United States, 137 F. Supp. 719 (Ct. Cl. 1956). In Oliver-Finnie, the government was sued for breach of contract because it supplied infested food under an agreement with a private party. The government asserted a contingent claim against its food supplier in an attempt to shift responsibility for the breach of contract to that supplier. The Court of Claims held that this type of indemnification claim was outside the scope of its jurisdiction because the government's rights against its supplier were not "derived through the contract or claim upon which the plaintiff instituted the suit." Id. at 721. The situation in Oliver-Finnie, however, is entirely distinct from the current dispute. While each of parties in this case has a separate contract with the government, the subject matter of this dispute is the proper distribution of the over-collection of excess revenues. If SCE and LADWP prevail on their claim that the government under-paid them, then the government necessarily over-paid the third-party defendants. It is irrelevant that these rights arise under separate contracts with the government because the claims all relate to the same disbursement of funds arising under the contracts. Under Rush and Maryland Casualty, this is precisely the type of situation that Congress intended to include under § 114(b). Thus, § 114(b) was properly invoked by the Court of Federal Claims.

The third party defendants also assert that the government's contingent claim is time-barred under § 2415. This argument is specious. Section 2415 specifically provides that the six-year statute of limitations does not apply to the government when it is seeking to recover improperly disbursed funds. See § 2415(f) ("[T]his section shall not prevent the assertion . . . of any claim of the United States . . . against . . . a third party, that arises out of a transaction or occurrence that is the subject matter of the opposing party's claim . . . ."). Thus, the Court of Federal Claims properly exercised jurisdiction over the third-party defendants.

The primary dispute in this case centers around the refund methodology adopted by Western to distribute surplus funds that it held at the end of the contract period. The regulatory basis for this refund comes from Article 14(e), which reads: "[A]djustments shall be made, through refunds or collections as the Secretary may find necessary to achieve consummation of the principle that the revenues shall be sufficient but not more than sufficient to" cover the construction, operating, and maintenance costs of the Hoover Dam project. Western, after considering various refund methodologies, decided that the refunds should be based on the percentage of firm power each contractor was entitled to purchase under the contracts. We must first determine the proper standard of review to apply to this determination. Thus, we must decide whether the Court of Federal Claims was correct in reviewing this regulatory interpretation only for reasonableness, or whether this interpretation must be reviewed independently and without deference.

Courts are, in certain circumstances, required to give deference to a regulation that embodies an agency's reasonable interpretation of a statute under its administrative authority. See Chevron, U.S.A., Inc. v. Natural Resources Defense Council, Inc., 467 U.S. 837, 843-44 (1984). To determine whether such deference is appropriate, the court must initially decide "whether Congress has directly spoken to the precise question at issue." Id. at 842. If so, no deference is warranted and the court must give effect to the express statutory intent of Congress. See id. at 842-43. If, on the other hand, Congress is silent on a particular matter or the statutory language is ambiguous, the court must defer to the agency's interpretation as long as it "is based on a permissible construction of the statute." Id. at 843. Similarly, an agency's interpretation of its own regulations must be given effect "so long as the interpretation sensibly conforms to the purpose and wording of the regulations." Martin v. Occupational Safety and Health Review Comm'n, 499 U.S. 144, 150 (1991); see also Thomas Jefferson Univ. v. Shalala, 512 U.S. 504, 512 ("We must give substantial deference to an agency's interpretation of its own regulations."); Princess Cruises, Inc. v. United States, 201 F.3d 1352, 1359-60 (Fed. Cir. 2000) (same). Deference is particularly appropriate when the agency is applying its regulations to a complex or changing circumstance, thus requiring the agency to bring to bear its unique expertise and policy-making prerogatives. See Martin, 499 U.S. at 151. When such judicial deference is appropriate, a court must accept the agency's reasonable interpretation of a regulation, even if there may be other reasonable interpretations to which the regulation is susceptible, and even if the court would have preferred an alternative interpretation. See Sharp Corp. v. United States, 63 F.3d 1092, 1096 (Fed. Cir. 1995).

While regulatory interpretations are generally afforded substantial deference by reviewing courts, the regulation at issue in this case is incorporated by reference into contracts entered into between the government and several private energy customers. According to SCE and LADWP, it is inappropriate to give deference to an agency's interpretation of a contractual provision when the government is itself a party to the contract. This argument is not without force. When a party enters into a contract with the government, that party should reasonably expect to be on equal legal footing with the government should a dispute over the contract arise. If courts were generally to adopt a principle of deference to an agency's contract interpretation, such deference could lead the courts to endorse self-serving post-hoc reinterpretations of contracts that an agency might offer in the context of a litigation. See National Fuel Gas Supply v. Federal Energy Reg. Comm'n, 811 F.2d 1563, 1571 (D.C. Cir. 1987) (noting that while an agency's contract interpretation may be entitled to deference, such deference may be inappropriate where the agency itself is a party to the contract). It would be unfair to give the government such a distinct advantage during an ordinary breach of contract litigation. Nevertheless, under the unique circumstances of this case, reviewing the agency's regulatory interpretation under a deferential standard is appropriate and does not unfairly prejudice SCE and LADWP.

First, despite the fact that the government is a party to these contracts, it had no economic stake in the excess revenue that was to be distributed. Western's only function under the Regulations was to adopt a methodology to refund the over-collection to the energy customers. In such a circumstance, there was no motivation for Western to adopt a self-interested regulatory interpretation because under any refund allocation method, all excess revenues would be distributed from its coffers to the customers. In other words, the agency was acting as a neutral arbiter resolving the customers' rights to the over-collected funds, rather than as an interested party to a contract. Because Western had no interest in the particular methodology adopted, the primary rationale against giving deference to an agency's contract interpretations is diffused. SeeNational Fuel, 811 F.2d at 1571 (recognizing that deference could lead to the endorsement of "self-serving . . . post hoc reinterpretation[s]"). SCE and LADWP dispute the assertion that the government is not an "interested party," noting that the government could be required to pay additional money depending on the outcome of this litigation. However, they ignore the fact that the refund allocation decision was made prior to this litigation, when there was no taint of self-interest, and when the government was not adverse to any of the energy customers. That the government defends this interpretation in a subsequent litigation does not call into question the integrity of the initial decision. Thus, merely because the government was a party to these contracts does not preclude the application of deference to its initial allocation determination.

In addition, the statutory and regulatory framework supports the application of judicial deference in this case. With the BCPA and the Adjustment Act, Congress delegated significant responsibility for the administration of these energy contracts to the Secretary. See, e.g., 43 U.S.C. § 617c(b) (authorizing the Secretary to "promulgate such regulations . . . as he may find necessary or appropriate" to carry out the purposes of the BCPA and Adjustment Act). Neither the BCPA nor Adjustment Act provide any guidance as to how excess funds should be distributed at the conclusion of the fifty-year contract period. See Chevron, 467 U.S. at 837 (holding that deference to an agency interpretation is proper when the relevant statute is silent or ambiguous on the issue at hand). The Regulations, while clearly requiring the agency to make an "adjustment" at the end of the contract period, do not provide detailed guidance as to the specific method for refunding the over-collected funds. See Regulations, art. 14(e). Thus, the agency is put to the task of determining, from a range of reasonable alternatives, the details of the refund allocation method. In the face of such statutory and regulatory silence, the role of the judiciary is not to determine which among these reasonable alternatives is most reasonable. See Sharp, 63 F.3d at 1096. Rather, the proper role of this court, given the expertise the agency developed over the fifty years in interpreting these Regulations, is to ensure that the methodology selected by the agency is reasonable. Thus, assuming that the interpretation adopted by Western was reasonable under the Regulations, we will give it effect.

This holding is not inconsistent with prior cases from this court which have held that a de novo standard of review is appropriate for matters of regulatory interpretation when regulations are incorporated into contracts. See United States v. Boeing Co., 802 F.2d 1390, 1393 (Fed. Cir. 1986) ("The interpretation of regulations which are incorporated into government contracts is a question of law which this court is free to resolve."); Brown v. United States, 195 F.3d 1334, 1340 (Fed. Cir. 2000) (same). In both of these cases, the government's interests were directly adverse to the contractor at the time the interpretation was made. Thus, these cases are merely applications of the general rule against deference in an ordinary contract dispute. As described above, this general rule is inapplicable to the current dispute because here, the government was not adverse to the contractors when the regulatory interpretation was made. Additionally, Boeing is an appeal taken from the ASBCA under the Contract Disputes Act, 41 U.S.C. §§ 601-613 (1994). In appeals taken directly from such agency adjudicatory tribunals, this court is required to review the substantive regulatory interpretation issues de novo. See 41 U.S.C. § 609(b). Here, where the agency interpretation was not rendered in an adversarial adjudicatory setting, judicial deference is appropriate. Thus, Boeing and Brown do not control the current dispute.

Having determined that the agency's refund allocation method is entitled to deference, we turn to the question of its reasonableness. As described above, at the end of the fifty-year contract period, Western was faced with a $25 million surplus, which it was required to refund to the energy customers. See Regulations, art. 14(e). After considering several alternatives, Western decided to refund the excess revenue to the firm energy customers in proportion to each customer's yearly allotment of firm energy. Western justified this decision on several grounds. First, Western noted that throughout the fifty-year duration of the contracts, the adjustments that were made to account for fluctuations in annual operating costs were accomplished through collections from or refunds to firm energy customers only. Western believed that it would be reasonable to follow a consistent methodology for the final adjustment. In addition, any revenue shortfall that was experienced at the end of the contract period would have been borne by the firm energy customers alone, in proportion to the firm energy share of each customer. Because the firm energy purchasers would have borne the risk of any shortfall, Western felt it was appropriate that any excess revenues should be refunded to them.

The Court of Federal Claims determined, however, that Western's methodology was unreasonable. According to the court, Western's refund methodology was improper because it failed to satisfy the rate determination precepts set forth in the Regulations. See id., arts. 6-8. Because Article 14 refers to "adjustments to energy rates," the court reasoned that adjustments must be, in effect, retroactive rate-settings that maintain the fixed relationship between firm and secondary energy rates. See id., art. 8. By refunding money only to the firm energy customers, Western effectively lowered firm energy rates while leaving secondary energy rates the same. In the court's judgment, this violated Article 8, rendering the refund methodology unreasonable.

While the interpretation by the Court of Federal Claims represents a reasonable reading of the Regulations, we cannot say that Western's alternative interpretation was unreasonable. The regulatory language expressly grants broad discretion to the Secretary. See id., art. 14(e) ("[A]djustments shall be made, through refunds or collections as the Secretary may find necessary . . . .") (emphasis added). Consistent with this grant of discretion, the Regulations do not include specific guidance regarding the details of the refund methodology to be adopted. This regulatory silence supports the conclusion that the Secretary was to have wide latitude in devising the details of the refund methodology, so long as it was consistent with the purpose and wording of the Regulations. See Thomas Jefferson Univ., 512 U.S. at 512.

Western's decision to base the refund only on the firm energy allotments found in Article 4(a) was within the broad discretion granted under this regulatory scheme. As described above, throughout the course of these contracts, the periodic adjustments required under Article 14 to ensure sufficient annual revenue were made by collections from or refunds to firm energy customers only. It was reasonable for Western to base the final adjustment under Article 14(e) on a similar scheme. Also, because any revenue shortfall that would have occurred at the end of the contract period would have been covered by the firm energy customers only, it was reasonable to refund the excess revenues to them exclusively as well.

Western's interpretation also appropriately reflects other fundamental differences between firm and secondary energy purchasers. Firm energy customers were required under the contracts to purchase a fixed quantity of energy each year. Thus, revenues generated from firm energy customers guaranteed the payments of the project's costs. Secondary energy customers were not required to purchase any energy under the contracts. Rather, they were granted options to purchase secondary energy at rates that were at a fraction of the rates for firm energy. Although revenues generated from the sale of secondary energy made a modest contribution to the overall project costs, firm energy purchasers alone assumed the financial risk of the project. Western's decision to refund the excess revenues only to firm energy customers is consistent with the allocation of risk between the firm and secondary energy customers.

The trial court's primary rationale for faulting Western's methodology was its failure to satisfy the rate-setting precepts of Articles 6 through 8. However, this does not render Western's methodology unreasonable. First, the plain language of Article 14(e) does not require the interpretation adopted by the trial court. While Article 14 refers to "adjustments to rates," there is no express requirement that these adjustments comply with the rate-setting requirements found elsewhere in the Regulations. In fact, such a requirement would seem particularly inappropriate with respect to the adjustment contemplated by Article 14(e), which occurs after the conclusion of the contract period when there are no longer forward looking rate determinations to be made. Also, Article 14(e) does not itself expressly refer to an adjustment to "rates" but simply to "adjustments . . . through refunds or collections." Because the Regulations do not contain an express requirement that the final adjustment satisfies the rate-setting precepts found in the Regulations, we hold that Western was not unreasonable in implementing a refund methodology that does not satisfy these precepts. See Thomas Jefferson Univ., 512 U.S. at 512 ("[W]e must defer to the Secretary's interpretation unless an alternative reading is compelled by the regulation's plain language.").

The correctness of this conclusion is underscored by the difficulty of implementing such a retroactive rate-setting determination. Article 14(e) clearly contemplates an adjustment that reconciles project costs with collected revenues over the entire fifty-year contract term. To develop a refund methodology that would retroactively adjust rates over this entire contract term would be an extremely complex undertaking. We will not presume, based simply on the reference in Article 14 to "adjustment to rates," that the Regulations require such a complex undertaking. SCE and LADWP contend that the appropriate time period to consider is not the entire fifty-year term, but only the final four years of the contract. While this would certainly simplify the determination, it has no support in the Regulations and thus must be rejected.

Finally, as described above, adjustments throughout the course of the contract were made with respect to firm energy customers alone, and any shortfall at the end of the contract period would have been covered by the firm energy customers alone. Taken to its logical extreme, these adjustments would be unreasonable under the trial court's analysis because they would also violate the rate-setting precepts, particularly the requirement of Article 8 to maintain a fixed relationship between firm and secondary energy rates. However, these adjustments are clearly proper under the Regulations. It would be anomalous to suggest that these adjustments, which do not follow Article 8, are permissible, but Western's refund methodology is not. Therefore, we hold that Western's refund methodology represents a reasonable interpretation of the Regulations and must be given effect.

The third-party defendants assert a number of other defenses against SCE's and LADWP's claims. They contend that SCE and LADWP should be barred from asserting their claims against the government by the one-year statute of limitations found in the Hoover Power Plant Act, 43 U.S.C. § 619a(h)(1) (1994). In addition, they argue that SCE's and LADWP's claims should be barred by equitable estoppel and laches. Because we determine that Western's refund methodology was reasonable, and thus entitled to finality, we need not address these additional defenses.

We affirm the decision of the Court of Federal Claims to exercise jurisdiction over the third-party defendants. However, because the Court of Federal Claims failed to give deference to the reasonable regulatory interpretation of Western, we reverse that portion of the court's determination, and hold that Western's refund methodology was entitled to finality.

Each party shall bear its own costs.

AFFIRMED-IN-PART AND REVERSED-IN-PART.