Sacramento Municipal Utility District v. United States

566 F. App'x 985

• Court: Court of Appeals for the Federal Circuit
• Decided: June 20, 2014
• Docket: 2013-5086, 2013-5087
• Status: Unpublished

Opinion

RADER, Circuit Judge.

Under the Nuclear Waste Policy Act of 1982 (NWPA), 42 U.S.C. §§ 10101-10270, Sacramento Municipal Utility District (SMUD), a California public utility, entered into a standard contract with the Department of Energy (Department) in which SMUD agreed to pay $40 million into the Nuclear Waste Fund (NWF) and, in return, the Department promised to begin accepting and disposing of SMUD’s spent nuclear fuel (SNF) and high-level radioactive waste (HLW) in 1998. 10 C.F.R. § 961.11 (2010). To date, the Department has yet to begin performance. SMUD filed its first suit for damages in 1998 and, despite prevailing in a complex series of adjudications, SMUD has yet to recover any damages owed by the Department.

Because the United States Court of Federal Claims erred by reducing SMUD’s award in a first suit by inflated offsets improperly assessed in a second suit, this court reverses the combined judgment of $38,845,898 for the period of 1992-2009 and reinstates a prior $53,159,863 award for the period of 1992-2003. That prior judgment shall be executed immediately. Further, the trial court erred by barring SMUD’s exchange theory in the second suit on the basis of collateral estoppel, and by re-assessing a prior wet pool offset determination that is the law of the case. Accordingly, this court vacates the award with respect to 2004-2009 and remands for consideration of the exchange theory and to determine a fuel-out date.

I.

This appeal is one of many arising from a longstanding series of contract disputes between the nuclear power industry and the Government. These disputes arise from damages owed to the nation’s nuclear utilities resulting from the Government’s failure to fulfill contractual obligations to accept and dispose of radioactive waste under the NWPA. Several decisions of this court address the broader circumstances surrounding these cases. See, e.g., Dairyland Power Co-op. v. United States, 645 F.3d 1363 (Fed.Cir.2011); Pac. Gas & Elec. v. United States, 536 F.3d 1282 (Fed.Cir.2008); Yankee Atomic Elec. Co. v. United States, 536 F.3d 1268 (Fed.Cir.2008); Ind. Mich. Power Co. v. United States, 422 F.3d 1369 (Fed.Cir.2005).

In 1988, like many nuclear utilities across the nation, SMUD entered into a standard contract under which SMUD paid tens of millions of dollars into the NWF in exchange for the Department’s promise to begin acceptance and disposal of SNF and HLW beginning in 1998. J.A. 2455-88; see also Standard Contract, 10 C.F.R. § 961.11 (2010). On December 20, 1988, the Department proposed a schedule for waste acceptance and removal beginning by the statutorily-mandated date of January 31, 1998. See Pac. Gas, 536 F.3d at 1286.

Instead of providing a firm rate for SNF and HLW acceptance and disposal, the standard contract required the Department to issue a series of annual capacity reports (ACRs) beginning no later than July 1, 1987. Id. at 1285. These reports set forth projected annual receiving capacities for the Department’s facilities and annual acceptance rankings. Id. In a later series of suits brought by affected utilities, this court concluded that the 1987 Annual Capacity Report (1987 ACR) provided the best available pre-breach snapshot of the parties’ intended acceptance rate. Id. at 1292; see also Yankee Atomic, 536 F.3d at 1274.

Accordingly, the standard contract executed by SMUD did not include a specific pickup schedule for its SNF and HLW transfer. Rather, it provided that the utility would submit delivery requests for the SNF/HLW that it wished to deliver to the Department for a given year. J.A. 2465. If the requests exceeded the Department’s capacity, potential acceptance slots would be allocated based on the date fuel was discharged from the reactor — a scheme known as “oldest fuel first” (OFF) ranking. Id. at 2468-69. Beyond this, the standard contract included a provision prioritizing acceptance of fuel originating from decommissioned or shutdown utilities, id. at 2469, as well as an “Exchanges” provision giving utilities the option to exchange or swap acceptance slots to adjust delivery schedules, albeit subject to approval by the Department. Id. at 2466-67.

The Department’s inability to begin disposal services became apparent many years before January 31, 1998. As early as 1987, the Department began announcing delays in the expected opening of its repository. Id. at 2821-22. In a 1989 report to Congress, the Department announced that the repository would not operate until “approximately 2010.” Id. at 2822 (quoting Dep’t of Energy, Report to Congress by the Secretary of Energy on Reassessment of the Civilian Radioactive Waste Management Program, Nov. 29, 1989). In 2005, the Department announced further delays in opening a repository, estimating a start date of 2012, with more delay likely. Id.

SMUD sued for costs incurred to mitigate from January 1,1992 to December 31, 2003, and sued separately for the period of January 1, 2004 to December 31, 2009.

II.

SMUD operated the Rancho Seco nuclear power plant throughout the 1980s until it was shut down in response to a voter referendum in 1989. See Sacramento Mun. Util. Dist. v. United States, 70 Fed.Cl. 332, 339 (2006) (SMUD I). At the time of its proposed decommissioning, SMUD stored 493 SNF and HLW assemblies at Rancho Seco in wet pool installations having an estimated annual operating cost of between $6-12 million. J.A. 2822-23.

Knowing that the Department’s performance would be many years late, SMUD decided to follow through with construction of a dual-purpose, dry storage facility called an Independent Spent Fuel Storage Installation (ISFSI). Id. at 2828. Specifically, beginning in 1992, SMUD pursued construction of an ISFSI for long term dry storage of its spent fuel — one that would require fewer personnel and carry a much lower estimated annual operating cost of between $1.5-4.5 million. Id. at n. 4. Between 1992 and 2001, SMUD encountered various regulatory and technical hurdles, causing it to reevaluate its storage strategy many times. Id. at 2828. However, in view of the Department’s then-announced, and seemingly inevitable, breach SMUD decided to continue pursuing construction of an ISFSI. Id. SMUD completed its ISFSI in 2001, transferred all of its waste to dry storage by August of 2002, and shut down its wet pool that year. SMUD I, 70 Fed.Cl. at 375. SMUD incurred $78,558,212 in costs related to development of its ISFSI from 1992-2003. J.A. 2824.

In 1998, SMUD brought its first action against the Department for breach of the standard contract, seeking damages of $76,603,412 in mitigation costs to compensate for constructing the ISFSI and storing SNF and HLW. That action covered the period spanning January 1, 1992, to December 31, 2003. Determining that the breach was a substantial factor in SMUD’s decision to complete construction of its ISFSI, the Court of Federal Claims awarded SMUD $39,796,234. Sacramento Mun. Util. Dist. v. United States, 74 Fed.Cl. 727, 735 (2006). That award included various offsets, among them a $4,146,360 offset that SMUD realized in 2003 from decommissioning the wet pool and transferring all of its fuel to dry storage. See SMUD I, 70 Fed.Cl. at 374-75 (explaining that SMUD shut down systems and reduced its staff by 30 employees).

SMUD disputed the wet pool offset on grounds that it would have achieved a “fuel-out” date before 2003 in the non-breach world, a fact that would preclude any wet pool “savings.” Specifically, SMUD argued that if the Department began accepting fuel by January 31, 1998, all of Rancho Seco’s SNF and HLW would have been accepted for disposal and transported off-site no later than 2006. This assumed an acceptance rate of 3,000 metric tons of uranium (MTUs) per year and OFF ranking. SMUD further asserted that another utility in the queue may have been willing to swap places with SMUD to enable all of SMUD’s fuel to be removed prior to 2003. In this regard, SMUD accepted that, if the Court of Federal Claims were to conclude that SMUD would not have off-loaded all its fuel before 2003, a deduction in its costs of $4.2 million in wet pool offset for 2003 would be appropriate.

The trial court rejected SMUD’s 2003 fuel-out date, finding that “[t]o accept SMUD’s position, the court would need to speculate about whether SMUD would have been successful in trading SMUD’s acceptance priority with another utility or convincing DOE to accept SMUD’s SNF early.” Id. at 375. The trial court further explained, the law simply did “not permit the court to do so.” Id. Thus, the Court of Federal Claims reduced SMUD’s award for the period of 1992-2003 by $4.2 million in wet pool offset, in addition to various other offsets. For example, the trial court granted the Department an offset for SMUD’s internal expenses related to time and labor that 256 SMUD employees charged to the ISFSI project. SMUD, 74 Fed.Cl. at 735. The trial court also granted an offset to the Department for costs arising from SMUD’s storage of various nonfuel assembly components. Id. at 733.

On appeal, this court affirmed-in-part and reversed-in-part. First, this court found that the trial court did not use the required 1987 ACR rate, instructing the trial court on remand to “apply the Stan dard Contract acceptance rate identified in [Pac. Gas, 536 F.3d at 1291-92] to assess causation ... in evaluating the Government’s partial breach of contract as a substantial factor in causing SMUD to pursue dual-purpose storage.” Sacramento Mun. Util. Dist. v. United States, 293 Fed.Appx. 766, 768, 771-72 (Fed.Cir.2008). This court also found error in the trial court’s damages calculation with respect to certain offsets granted for labor and nonfuel component storage, totaling $13.4 million. Id. at 771-74. This court did not disturb the $4.2 million in wet pool offset.

On remand, the Court of Federal Claims held an additional evidentiary hearing and recalculated SMUD’s damages. Taking into account the correct acceptance rate, and restoring certain offsets, the trial court arrived at an award of $53,159,863 for 1992-2003. Sacramento Mun. Util. Dist. v. United States, 91 Fed.Cl. 9, 18-19 (Fed.Cl.2009). The Court of Federal Claims confirmed the $4.2 million offset for wet pool savings, explaining in its final opinion and order of judgment “that [the wet pool offset] ruling was not reversed on appeal.” Id. at 18.

At this juncture, the Department sought to stay execution of judgment in SMUD I pending resolution of a separate suit in which SMUD sought damages for the period of 2004-2009. Id. Specifically, in 2009, SMUD filed a second action for damages in accordance with precedent dictating that a SNF plaintiff must seek recovery for post-breach damages in successive suits that cover damages as they are incurred over time. Ind. Mich., 422 F.3d at 1376-78 (“[Subsequent claims ... accrue ... at the time such damages are incurred.”); Yankee Atomic, 536 F.3d at 1281 (“[Utilities cannot now collect damages not yet incurred under the ongoing contract.... [T]he Government’s own refusal to timely perform cannot serve as a basis for accelerating plaintiffs’ performance obligations. The Yankees’ obligations under the contractual scheme have not matured.” (citations omitted)).

Initially, the Department argued that the trial court should account for wet pool savings allegedly realized from 2004 to 2008 in SMUD I. SMUD, 91 Fed.Cl. at 18. And, at first, the Court of Federal Claims rejected this request on grounds that the scope of SMUD I extended only through December 31, 2003. Id. at 19 (“SMUD has requested an unspecified amount of damages ‘from January 1, 2004 forward’ in a separate proceeding.... Accordingly, the court has decided it is proper to account for any 2004-2008 SNF wet pool savings offset in that proceeding.” (citations omitted)). But, the Court of Federal Claims ultimately granted the Department’s request to stay execution of judgment in SMUD I in order to determine whether wet pool savings would apply to 2004-2009:

The Government, however, legitimately is concerned that the amount of costs SMUD may seek in the next proceeding will not be sufficient to allow the Government to recoup the amount of SNF wet pool savings realized from 2004-2008. Therefore, the court has decided that the interests of justice require a stay of the execution of this judgment, pending the court’s resolution of the damages to be asserted [in the second action].

Id. (citations omitted).

In response, SMUD sought a writ of mandamus from this court to compel the trial court to lift the stay and allow for entry of the judgment for the period of 1992-2003. In re Sacramento Mun. Util. Dist., 395 Fed.Appx. 684, 685 (Fed.Cir.2010). Because the stay would not last for a protracted or indefinite amount of time, this court denied SMUD’s mandamus petition, finding that the Court of Federal Claims did not abuse its discretion. Id. at 688. .

III.

SMUD’s second suit, concerning mitigation costs owed for the period from January 1, 2004 to December 31, 2009, proceeded to trial on October 24, 2011. Sacramento Mun. Util. Dist. v. United States, 109 Fed.Cl. 660, 669 (2013) (SMUD II). Relevant to this appeal, SMUD contended that, in light of certain economic incentives, the standard contract’s “Exchanges” provision would have been utilized such that SMUD’s fuel would have been removed before January 1, 2004 in the non-breach world.

SMUD offered testimony from various Department officials and government experts explaining that exchanges would readily have occurred. Id. at 671-72. For example, SMUD’s economic expert, Mr. Frank Graves, confirmed that using OFF ranking would have imposed an additional $1.6 billion in storage costs on the nuclear power utility industry that could have been avoided if exchanges were utilized. Id. Mr. Graves noted that the utilities do not act as competitors and, historically, have engaged in mutually beneficial exchanges of personnel, material, and even contract rights based on collective self-interest. Id. at 671. Moreover, since the Department stood to benefit from these cost-saving exchanges, it would have been reasonable to assume that the Department would approve them. Id. Under Mr. Graves’ analysis, the last of Rancho Seco’s SNF and HLW would have been removed in 1999. Id. Mr. Graves further estimated that the exchange costs would total $8.4 million— an amount dramatically less than the $20 million in wet pool operating costs incurred during the same period in the breach world. Id. at 673.

In addition to these economic incentives, SMUD established that fuel removal absent a breach would not likely have proceeded under a strict OFF schedule. SMUD offered testimony by experts and Department officials explaining that the OFF schedule would have been both inefficient and inconsistent with historical practices of shipping and managing spent fuel. Id. For example, Mr. Graves testified that dozens of utilities would have had OFF allocation rights in excess of their needs in the early years of spent fuel removal. Id. at 671. And those utilities would have been able to exchange allocation rights without disrupting their own fuel storage needs. Id. To verify these assumptions, Mr. Graves analyzed discharge and capacity data, variations in spent fuel storage costs, exercise of market power by utilities possessing early acceptance allocations, relevant transaction costs, barriers to trade, and factors introduced by greater-than-Class-C waste. Id. at 672. Ultimately, Mr. Graves testified that “Rancho Seco’s fuel would have been removed by 2003, and in fact likely well before then, even under unduly pessimistic assumptions about the non-breach world.” Id. at 671-72. Mr. Graves’ analysis concluded that, based on the acceptance rate of the 1987 ACR, within a few years after the Department’s removal of SNF commenced in 1998, the Department’s total acceptance capacity would have exceeded the total amount of must move fuel held by other utilities. Id. These circumstances would have allowed for a fuel-out date anywhere from 1999 to 2003, depending on the inputs. Thus, absent evidence SMUD would have been somehow prevented from exchanging a 2008 allocation for an earlier one, OFF scheduling and projected 2008 fuel-out date did not reflect “a plausible view of the non-breach world for SMUD IT Id.

While the Department contested Mr. Graves’ testimony, it declined to take a position on what it considered the proper model inputs or to offer an alternative model of the non-breach world. Id. Rather, the Department argued that the law of the case and collateral estoppel precluded SMUD from claiming that its spent fuel would have been removed earlier than 2008. The Department argued that SMUD did not appeal the earlier ruling rejecting use of its exchanges theory in SMUD I before the introduction of Mr. Graves’ testimony in SMUD II. Id. (citing SMUD I, 70 Fed.Cl. at 375 (rejecting the acceptance rate and possibility of exchange as too speculative)). Thus, the Department sought to bar any fuel-out theory premised on accelerated removal through the use of exchanges.

SMUD responded that nothing barred the Court of Federal Claims from determining a fuel-out date based on what would have occurred in the non-breach world from 2004-2009. Id. at 674. Given its mandatory separation from the 1992-2003 damages period, all issues regarding the 2004-2009 period should have been preserved for subsequent adjudication. Id. (citing SMUD, 91 Fed.Cl. at 19 (“SMUD has requested an unspecified amount of damages ‘from January 1, 2004 forward’ in a separate proceeding. Accordingly the court has decided it is proper to account for any 2004-2008 SNF wet pool savings offset in that proceeding.” (citations omitted))). Thus, the trial court’s rejection of exchange evidence should only work to offset damages through December 31, 2003. Id. at 674-75.

While rejecting a proposed fuel-out date before 2003 as speculative, the trial court had yet to determine a fuel-out date. Indeed, in SMUD I, the Court of Federal Claims specifically declined to establish one, because any determination on that issue depended on whether SMUD could have exchanged one or more of its 2004, 2005, or 2008 allocations. As such, the trial court expressly deferred issues falling within the scope of this later time period to SMUD II. Id. at 675 (citing SMUD, 91 Fed.Cl. at 18-19).

Moreover, SMUD argued that applying collateral estoppel would be improper given an intervening change in the law that altered the legal framework for evaluating Mr. Graves’ exchange model. Id. (citing Yankee Atomic Electric Company v. United States, 679 F.3d 1354, 1359 (Fed.Cir.2012) (upholding application of an exchange model); Dairyland, 645 F.3d at 1369-71 (affirming use of an exchange model offered by the same Mr. Graves); Pac. Gas, 536 F.3d at 1292 (establishing the 1987 ACR as the acceptance rate and applying an exchange model)).

On collateral estoppel, the trial court decided that it had already rejected SMUD’s argument with respect to the use of “exchanges” in SMUD I. Specifically, that “[SMUD] might have been able to convince another utility in the queue to switch places to enable SMUD’s SNF to be removed in 2003.” Id. at 676 (quoting SMUD I, 70 Fed.Cl. at 375). The Court of Federal Claims noted that SMUD did not appeal that ruling. Rather, it “chose to live with this 2003 wet pool deduction.” Id. at 677.

Moreover, the trial court concluded that SMUD presented an “exchanges” argument in SMUD II identical to the one decided in SMUD I, i.e., whether SMUD could have used exchanges to have its SNF removed earlier than 2008, as estimated by the OFF schedule. Id. Thus, although the two suits involved mutually exclusive time periods for purposes of damages, the Court of Federal Claims saw no significant change in the evidence or the circumstances affecting SMUD’s proposed use of exchanges in SMUD II that would distinguish the period of 2004-2009 from 1992-2003. Id. With respect to establishing a 2008 fuel-out date, the trial court concluded that the additional proffer of Mr. Graves’ testimony in SMUD II was at best cumulative to evidence asserted in SMUD I.

The Court of Federal Claims also rejected SMUD’s contention that issue preclusion cannot apply because of a significant change in the legal atmosphere governing exchange models. Id. at 679. In this regard, the trial court distinguished Yankee Atomic, Dairyland, and Pacific Gas as all limited to situations where this court merely affirmed the trial court’s fact findings based on the standard of review. Id. (citing Dairyland, 645 F.3d at 1370 (“We find no error in [the trial court’s conclusion regarding exchanges], as it appears to have been grounded in proper weighing of the evidence.”); Yankee Atomic, 679 F.3d at 1359 (“The Government did not identify any record evidence to support a finding that the trial court committed clear error in adopting an exchanges model.”); Pac. Gas, 668 F.3d at 1355 (“The record supports the trial court’s award of damages [based on use of the exchanges model] and this court discerns no error.”)). In sum, the trial court concluded that none of these cases affected the legal landscape on exchange models, which may at one time have been considered speculative. See SMUD II, 109 Fed.Cl. at 679. Accordingly, it imposed collateral estoppel.

But, at the same time, the trial court acknowledged: “Of course, in [SMUD I ] no fuel-out date for SMUD was established.” Id. Rather, by barring SMUD from litigating an earlier fuel-out date based on an exchange theory, the Court of Federal Claims determined that the fuel-out date must be 2008. Id. The trial court emphasized:

[T]he application of collateral estoppel is particularly important given the ongoing nature of the damage claims in the SNF cases. To date, the Government has been in breach of the Standard Contract for fourteen years. Barring some other resolution, the United States Court of Federal Claims faces decades of determining ongoing damages in six-year intervals. Therefore, these cases must be managed and adjudicated with finality.... [Thus,] SMUD is precluded from relitigating whether SNF would have been removed from Rancho Seco earlier than under the OFF schedule, [i.e., 2008,] by exchanges or otherwise.

Id. at 679-80.

IV.

In the non-breach world, SMUD would have continued to operate a wet pool until the Department removed SMUD’s fuel. However, SMUD I established that SMUD actually did not operate the wet pool during the period of 2004-2009, having transferred all its fuel to dry storage in 2002. Thus, assuming a fuel-out date of 2008, the Court of Federal Claims considered whether the Department was entitled to wet pool offset for costs SMUD saved from 2004 to 2008 by transferring all of its fuel to dry storage.

In SMUD I, the Court of Federal Claims held that “the Government, not SMUD, is responsible for operation and maintenance costs associated with storage of SNF.” SMUD I, 70 Fed.Cl. at 371. However, in SMUD II, the Department sought a deduction not only for SMUD’s wet pool savings of $4.2 million per year, but additionally for costs associated with dry storage. These dry storage costs contributed an additional $14.2 million to the Department’s offset spanning 2004-2008. SMUD II, 109 Fed.Cl. at 704-05. Specifically, the Department wanted to start with an annual $4.2 million in savings, then add $50,000 to cover spare parts and other materials as well as $2.84 million for the transition to, and cost of, dry storage, Id. at 707. This resulted in SMUD’s achieving annual savings, or avoided costs, of approximately $8 million per year, $38 million in total, to be deducted from its award of mitigation costs. The Department’s offset theory, as advanced in SMUD II, imposed a “double deduction,” Appellant’s Br. 40, by combining SMUD’s wet and dry storage expenses.

SMUD responded that, even if SMUD’s fuel would not have been removed prior to 2004 (as SMUD still maintained), $4.2 million and not $8 million represented a reasonable estimate of any avoided costs. SMUD argued that the Department’s proposed $8 million annual offset lacked evidentiary support and ran counter to analysis supporting the original $4.2 million annual offset determined in SMUD I. See SMUD II, 109 Fed.Cl. at 704-05.

The trial court agreed with the Department, reasoning that, at a minimum, SMUD’s claimed mitigation costs must be reduced by $4.2 million for each year that SMUD would have stored its own fuel (i.e., $20,981,800 for 2004, 2005, 2006, 2007 and 2008) because SMUD “chose to live with this 2003 wet pool deduction.” Id. at 677, 703-04. But, despite having previously concluded that the Department would bear the costs of storage, SMUD I, 70 Fed.Cl. at 371, the trial court revisited the $4.2 million wet pool offset and concluded that SMUD incurred costs associated with both wet storage and dry storage. SMUD II, 109 Fed.Cl. at 706. This resulted in additional offset for the Department totaling $34,987,913 for 2004 to 2008.

Ultimately, the trial court concluded that SMUD was entitled to $20,703,595 in mitigation costs resulting from the Department’s partial breach for the period of 2004-2009. Id. And the Department received the offset of $34,987,913, which the trial court deducted from SMUD’s award. However, instead of stopping at zero, the trial court decided to carry over a negative value, -$14,284,318, reducing SMUD’s award of $53,139,863 in SMUD I (for 1992-2003) by that amount. On March 18, 2013, the trial court entered combined judgment for SMUD I and SMUD II, awarding SMUD a net $38,845,39 for the period spanning January 1, 1992, to December 31, 2009. Id. at 708.

SMUD timely appealed the combined judgment. This court has jurisdiction under 28 U.S.C. § 1295(a)(3).

V.

This court reviews the trial court’s application of collateral estoppel without deference. See Dureiko v. United States, 209 F.3d 1345, 1355 (Fed.Cir.2000). This court reviews the trial court’s choice of a damages theory for an abuse of discretion and its factual findings concerning damages for clear error. See Golden Blount, Inc. v. Robert H. Peterson Co., 438 F.3d 1354, 1371 (Fed.Cir.2006). This court may reverse a damages award that rests upon “an erroneous conclusion of law, clearly erroneous factual findings, or a clear error of judgment.” Juicy Whip, Inc. v. Orange Bang, Inc., 382 F.3d 1367, 1370 (Fed.Cir.2004).

At the outset, this court finds that the Court of Federal Claims erred in applying the doctrine of collateral estoppel to bar consideration of SMUD’s exchange theory.

A party seeking to invoke collateral es-toppel must establish four preconditions: (1) identity of the issues in a prior proceeding; (2) actual litigation of those is sues; (3) necessity of the prior determination to the resulting judgment; and (4) full and fair opportunities to litigate issues for the party defending against preclusion. See Levi Strauss & Co. v. Abercrombie & Fitch Trading Co., 719 F.3d 1367, 1371 (Fed.Cir.2013); see also Stephen Slesinger, Inc. v. Disney Enters., Inc., 702 F.3d 640, 646 (Fed.Cir.2012). And even if all these conditions are met, application of collateral estoppel “must be confined to situations where the matter raised in the second suit is identical in all respects with that decided in the first proceeding and where the controlling facts and applicable legal rules remain unchanged.” Comm’r of Internal Revenue v. Sunnen, 333 U.S. 591, 599-600, 68 S.Ct. 715, 92 L.Ed. 898 (1948). Specifically, it should not be applied to foreclose full consideration of an issue “when there ha[s] been a significant change in decisional law between the first case and the second.” Bin gaman v. Dep’t of Treasury, 127 F.3d 1431, 1438 (Fed.Cir.1997).

In light of this court’s intervening adoption of the 1987 ACR rate as the required acceptance rate in SNF cases, Pac. Gas, 536 F.3d at 1292, and its acceptance of damages theories relying on the use of exchange models, e.g., Dairyland, 645 F.3d at 1370, the trial court should have allowed the parties to litigate the fuel-out date in SMUD II. Moreover, the trial court should have determined a specific fuel-out date in SMUD II, particularly given its decision in SMUD I to reserve that precise issue for later adjudication. SMUD, 91 Fed.Cl. at 18-19.

With respect to the acceptance rate deemed “speculative” by the trial court in SMUD I, this court’s holding in Pacific Gas represented a significant development in the law. In fact, this development in the law directly contributed to reversal of the trial court’s damages calculation in SMUD I on appeal. SMUD, 293 Fed.Appx. at 771-72.

Moreover, the Court of Federal Claims appears to have relegated this court’s holding in Dairyland to merely affirming the trial court’s factual findings based on the standard of review. See SMUD II, 109 Fed.Cl. at 679. But, in Dairyland, this court substantially addressed the viability of exchange models to prove what would have happened in the non-breach world in SNF cases. 645 F.3d at 1370. Specifically, in Dairyland, this court approved a damages theory relying on evidence and expert testimony that a market would have existed in the non-breach world for exchanging acceptance slots to adjust delivery schedules, and that such exchanges were foreseeable in view of the standard contract’s “Exchanges” provision. Id.

Turning to the doctrine of collateral estoppel, this court disagrees that the issues between SMUD I and SMUD II were so clearly identical. Under SMUD’s fuel-out theory in SMUD I, SMUD argued that it could have unloaded all of its fuel prior to 2003. SMUD I, 70 Fed.Cl. at 375. SMUD focused on the period before 2003 because the trial court properly limited the scope of SMUD I to damages and offset available between 1992 and 2003. In SMUD II, SMUD argued that it would have achieved a fuel-out date before 2004 and, in any event, before 2008. This second contention relied on analysis of likely behavior by utilities, keyed to determine the amount of excess capacity in the system in every year through 2008 — i.e., comparing excess capacity to amounts of SNF that SMUD would need to exchange. J.A. 1836-37. Therefore, whether SMUD would have been out of fuel prior to 2003 is not identical to the question of whether SMUD would have been out of fuel in 2003, 2004, 2005, 2006, 2007, or 2008. Also, in contrast to SMUD I, where the trial court rejected the exchange model as “speculative” based on SMUD’s estimated acceptance rate, SMUD’s later exchange model establishing a fuel-out date prior to 2008 distinctly applied the 1987 ACR rate dictated by Pacific Gas.

This court also disagrees that the parties actually litigated a fuel-out date in SMUD I. First, the trial court expressly declined to determine a fuel-out date in SMUD I, preserving that issue for SMUD II. SMUD, 91 Fed.Cl. at 18-19. Second, while SMUD ultimately decided to accept the $4.2 million offset for 2003 in SMUD I, that acquiescence does not make the fuel-out itself date actually litigated with respect to the period of 2004-2009. This court’s mandate to apply the 1987 ACR displaced the trial court’s underlying determination that the exchange theory requires improper speculation. Therefore, the trial court should have permitted litigation of these issues and determined a specific fuel-out date in SMUD II.

VI.

The trial court’s determination of a $4.2 million annual wet pool offset, as upheld by this court on appeal, became the law of the case. Thus, it should not have been supplanted by the Department’s inconsistent offset theory in SMUD II.

The law of the case doctrine “posits that when a court decides upon a rule of law, that decision should continue to govern the same issues in subsequent stages in the same ease.” Christianson v. Colt Indus. Operating Corp., 486 U.S. 800, 815-816, 108 S.Ct. 2166, 100 L.Ed.2d 811 (1988) (citations omitted). Also, the mandate rule, encompassed by the broader law of the case doctrine, furthermore dictates that “an inferior court has no power or authority to deviate from the mandate issued by an appellate court.” Briggs v. Pa. R. Co., 334 U.S. 304, 306, 68 S.Ct. 1039, 92 L.Ed. 1403 (1948). Thus, once an issue has been decided by an appellate court, that question may not be reconsidered at any subsequent stage of the litigation, save on appeal. See In re Sanford Fork & Tool Co., 160 U.S. 247, 255, 16 S.Ct. 291, 40 L.Ed. 414 (1895) (“Whatever was before [the Supreme Court], and disposed of by its decree, is considered as finally settled. ... If the circuit .court mistakes or misconstrues the decree of this court, and does not give full effect to thé mandate, its action may be controlled ... upon a new appeal.”).

Under the mandate rule, a court below must adhere to issues settled in a prior appeal unless one of three “exceptional circumstances” exist: (1) evidence presented later is substantially different from the original evidence; (2) controlling authority has made a contrary and applicable decision of the law; or (3) the prior decision was clearly erroneous “and would work a manifest injustice.” Gindes v. United States, 740 F.2d 947, 950 (Fed.Cir.1984) (citations omitted). The mandate rule, however, is limited to issues “actually decided, either explicitly or by necessary implication” in the prior litigation. See Toro Co. v. White Consol. Indus., Inc., 383 F.3d 1326, 1335 (Fed.Cir.2004).

The Court of Federal Claims erred in abandoning the prior determination of wet pool offset as $4.2 million:

[T]he Government, not SMUD, is responsible for operation and maintenance of costs associated with storage of SNF, until DOE accepts the last of SMUD’s SNF. Accordingly, the court has determined not to offset SMUD’s damage award by the costs incurred to operate and maintain the ISFSI in 2003.

SMUD I, 70 Fed.Cl. at 371.

On remand, following this court’s reversal-in-part on the 1987 ACR rate, the trial court recalculated SMUD’s damages and arrived at an award of $58,159,863 for the period of 1992-2008. SMUD, 91 Fed.Cl. at 18-19. It also confirmed the $4.2 million offset for wet pool savings. Id. at 18. Thus, by necessary implication, this court’s mandate affirmed the trial court’s decision to exclude costs incurred to operate and maintain the ISFSI. Thus, a later offset theory adding back ISFSI storage costs to the base of $4.2 million — a theory the trial court correctly rejected in SMUD I— should have been foreclosed. Through SMUD I, the $4.2 million calculation became the law of the case.

Only true exceptions to the mandate rule warrant deviation from an appellate court’s determination on issues within, or subsumed by, its mandate. Throughout SMUD I the parties litigated the wet pool offset and further offset based on ISFSI transition and storage. As between SMUD I and SMUD II, the facts did not differ substantially and no intervening authority contradicts the outcome. Indeed, the Department never challenged the $4.2 million offset. SMUD, 91 Fed.Cl. at 18-19 (“[T]he Government did not argue entitlement to this offset at trial, on appeal, nor did the Federal Circuit’s remand direct the court to do so.”). Thus, no exception to the mandate rule applies that would permit reassessment of this prior determination.

Finally, the trial court erred in combining the awards in the two cases for purposes of reducing SMUD’s damages from 1992-2003 with a negative value determined with respect to SMUD’s damages-minus-offset for the period of 2004-2009. Here, the Court of Federal Claims applied the $34,987,913 offset against the damages proven in SMUD II, i.e., $20,703,595, resulting in a negative dollar amount, i.e., - $14,284,388. And, instead of merely awarding SMUD zero dollars in SMUD II, it erroneously carried over the negative amount to reduce damages awarded on remand in SMUD I, i.e., $53,139,863, for 1992-2003.

According to the procedure set forth in Indiana Michigan Power, 422 F.3d at 1376-78, SMUD litigated damages in successive suits, as it incurred mitigation costs over time. This court dictated this litigation structure to limit a SNF plaintiffs ability to collect future damages. By that same token, this precedent also limits the Department’s ability to obtain future offsets. Id. at 1377; Yankee Atomic, 536 F.3d at 1281. By staying execution of judgment in SMUD I pending an offset determination in SMUD II, the trial court constructively awarded the Department future offsets in a manner prohibited by this court’s precedent.

Because the trial court erred in SMUD II by barring SMUD’s exchange theory and by declining to determine a specific fuel-out date, this court vacates the amount calculated as the Department’s offset from 2004-2008, i.e., $34,987,913, and remands that set of issues for further consideration consistent with this opinion. As for arriving at negative mitigation costs, this outcome violates the general principle that a breaching party’s offset cannot exceed a non-breaching party’s damages. In other words, SMUD cannot avoid more damages than it claims. Thus, even if the trial court had not erred with respect to the offset determination, the award should have stopped at zero for the period covered by SMUD II. See generally Myers v. Hurley Motor Co., 273 U.S. 18, 27, 47 S.Ct. 277, 71 L.Ed. 515 (1927) (“[T]he amount allowed defendant shall not exceed the amount of plaintiff’s claim.”). Thus, the trial court erred by awarding any resulting difference to the Department.

As for combining a negative value for a later damages period with an earlier final judgment covering a separate and distinct damages period, the $53,139,863 awarded for mitigation costs incurred by SMUD from 1992 to 2003 should have remained untouched. In the end, the trial court’s decision to combine the judgments in SMUD I and SMUD II netted the Department $14.3 million in this series of cases. At best, this result violates the settled principle that a breaching party should never be placed in a better position as a result of its breach. Cf. LaSalle Talman Bank, F.S.B. v. United States, 317 F.3d 1363, 1371 (Fed.Cir.2003). At worst, the decision to stay execution of a prior award pending a determination of any offset available in the future creates a perverse incentive for the Department to stall pending SNF actions with the hope of eviscerating past claims with future offsets.

As the trial court correctly noted, the Department has been in continuous breach of the standard contract for decades. And, barring some other resolution, the Court of Federal Claims likely “faces decades of determining ongoing damages.” SMUD II, 109 Fed.Cl. at 679-80. However, this alone does not compel a resolution such as the one reversed and remanded today. This court appreciates the trial court’s imperative to adjudicate this class of cases with finality, but this must be taken on balance with the burden imposed on SNF plaintiffs, like SMUD, who have prevailed at trial, appeal, and remand, but have yet to recover any damages. SMUD has done so while paying out over $110 million for building, operating, and maintaining an ISFSI. Appellant’s Br. 2.

Accordingly, this court further reinstates the award to SMUD of $53,139,863 for costs incurred to mitigate from January 1, 1992 to December 31, 2003, and directs the trial court to execute that prior judgment immediately.

REVERSED, VACATED, REMANDED, and REINSTATED,