From Casetext: Smarter Legal Research

Northern States Power Company v. United States

United States Court of Appeals, Eighth Circuit
Aug 12, 1998
151 F.3d 876 (8th Cir. 1998)

Summary

holding that a power company effectively had committed a posting error and did not change its method of accounting when it filed a refund claim seeking to deduct contract losses as it consistently had other losses

Summary of this case from O'Shaughnessy v. C.I.R

Opinion

No. 97-2914, 97-2915

Submitted March 12, 1998

Filed August 12, 1998

Appeals from the United States District Court for the District of Minnesota.

Counsel who presented argument on behalf of the appellant was Kenneth W. Rosenberg, U.S. Department of Justice, Washington, DC. Appearing on the brief were Richard Farber and Loretta C. Argrett.

Counsel who presented argument on behalf of the appellee was Steven Zane Kaplan of Minneapolis, Minnesota. Appearing on the brief was Richard D. Snyder.

Before WOLLMAN and HANSEN, Circuit Judges, and GOLDBERG, Judge.

The Honorable Richard W. Goldberg, Judge, United States Court of International Trade, sitting by designation.


This tax refund dispute raises two independent issues. First, when did the taxpayer, Northern States Power Co. ("NSP"), place two shipments of nuclear fuel assemblies into service, and thereby become eligible to claim a depreciation deduction and an investment tax credit? And second, when NSP filed refund claims currently deducting certain losses that it had previously capitalized, did NSP change its method of accounting? NSP filed this action to recover a refund for taxes levied and collected by the Internal Revenue Service ("IRS") for the years 1985 and 1986. With respect to the first issue, involving the nuclear fuel assemblies, NSP argued that the fuel assemblies were placed in service when NSP acquired them because they were fully assembled and destined for use in an existing nuclear power plant. With respect to the second issue, involving the method of accounting, NSP argued that it did not change its method of accounting when it filed refund claims changing how it treated certain contract losses; rather, it was merely correcting mistakes on its 1985 and 1986 tax returns.

After referring the matter to a magistrate judge, the district court granted summary judgment in favor of NSP on the first issue, and in favor of the United States on the second issue. The parties cross-appeal. Appellant/cross-appellee-defendant United States contends that NSP could not claim a depreciation deduction and an investment tax credit for the nuclear fuel assemblies until NSP inserted them into the reactor core and used them to generate salable electric power. Appellee/cross-appellant-plaintiff NSP, in turn, continues to argue that it was correcting a mistake, and not changing its method of accounting, when it filed refund claims seeking to deduct certain contract losses instead of capitalizing them. We affirm in part and reverse and remand in part.

I.

The material facts are not in dispute. Because the pertinent facts for each issue are entirely separate, we first summarize the facts relating to the nuclear fuel assemblies issue and then the facts relating to the method of accounting issue.

A. The Nuclear Fuel Assemblies

NSP owns and operates a nuclear electric power plant in Prairie Island, Minnesota. The Prairie Island plant has two nuclear reactors housed in separate containment buildings, Prairie Island I and Prairie Island II. This case involves nuclear fuel assemblies that NSP purchased for Prairie Island I. Prairie Island I has a core composed of 121 individual fuel assemblies which must be replaced periodically. As a result, NSP operates Prairie Island I in ten to sixteen month cycles, shutting it down in between cycles for refueling and routine maintenance. These operating cycles are sequentially numbered. Hence, the first ten to sixteen months that Prairie Island I produced power is referred to as Cycle 1, the next ten to sixteen month period following refueling is referred to as Cycle 2, and so forth. Each fuel assembly generally lasts for three cycles before it must be removed and replaced. The removal and replacement of the fuel assemblies is staggered: that is, at the end of each operating cycle, NSP replaces the oldest-third of the reactor's fuel assemblies with new fuel assemblies. Thus, for each cycle the core is composed of one-third new fuel assemblies, and two-thirds partially-used fuel assemblies. This case involves fuel assemblies that NSP purchased for Cycle 11 and Cycle 12.

NSP schedules the shut downs well in advance, i.e., at least one year. The shut downs usually last anywhere from five to seven weeks. Once the reactor is shut down, the refueling process can begin. Briefly, it includes the following steps: new fuel assemblies are moved from the "new fuel pit" to the "spent fuel pool"; the reactor is cooled; the reactor head is removed; one-third of the spent (used) fuel assemblies are removed and stored in the spent fuel pool; the core is reconfigured using the remaining fuel assemblies according to a predetermined design; the new fuel assemblies are inserted into the core at specified locations; and the reactor head is replaced. Once this process and any needed maintenance are completed, the reactor is then ready to undergo "startup physics testing." Startup physics testing verifies that the fuel assemblies, both new and old, are properly positioned in the core and that the core design meets certain technical specifications. Once everything is verified, the reactor returns to full operating status and a new cycle begins.

Between December 9 and 19, 1985, NSP received 40 new fuel assemblies from Westinghouse to be used in Cycle 11. The assemblies were fabricated according to detailed technical specifications provided to Westinghouse by NSP, and were delivered fully assembled. After NSP received the fuel assemblies, it stored them in the new fuel pit before moving them to the spent fuel pool. They were inserted into the reactor in the middle of March 1986, and the reactor reached full power on April 14, 1986. Because NSP viewed the fuel assemblies as placed in service in December 1985, NSP claimed both a depreciation allowance and an investment tax credit for them in 1985.

Between December 12 and December 22, 1986, NSP received 44 new fuel assemblies from Westinghouse for use in Cycle 12. Again, NSP initially stored the fuel assemblies in the new fuel pit, and later moved them to the spent fuel pool. They were inserted into the reactor between April 9 and April 19, 1987. The reactor reached full power on June 1, 1987. Just as it had done the preceding year, NSP claimed both a depreciation allowance and an investment tax credit for the assemblies in the year that it received them from Westinghouse, this time in 1986.

The IRS disallowed both the depreciation deductions and the investment tax credits for 1985 and 1986 that related to the nuclear fuel assemblies. Specifically, the IRS disallowed $2,061,819 of the total depreciation deduction and $1,446,891 of the investment tax credit for 1985, and $2,953,915 of the total depreciation deduction and $1,394,975 of the investment tax credit for 1986. NSP paid the additional tax and interest due, and timely filed claims with the IRS for a refund. After the IRS denied the claims, NSP commenced suit in the district court. There, the parties cross-moved for summary judgment. The district court referred the matter to a magistrate judge, who recommended granting summary judgment in favor of NSP on this issue. Specifically, the magistrate judge concluded that the fuel assemblies were placed in service when NSP acquired them from Westinghouse because at that point, the fuel assemblies were fully assembled, inspected, and ready to be inserted into the reactor core. After reviewing the magistrate judge's report and recommendations, the district court agreed and granted summary judgment in favor of NSP on this issue. See Northern States Power Co. v. United States, 952 F. Supp. 1346, 1347 (D.Minn. 1997). The United States appeals.

B. The Method of Accounting for Contract Losses

Nuclear power plants produce electricity by using nuclear fuel assemblies to generate a fission reaction. Nuclear fuel assemblies are composed of fuel rods containing individual uranium pellets. Uranium pellets are the end result of a complex process in which natural uranium is mined, converted into a gas, and enriched to increase the U235 isotope. The enriched gas is then piped into containers and shipped to the fuel assembly fabricator. The fuel assembly fabricator converts the enriched gas into uranium pellets, which are then used to make the fuel rods. The enrichment process is critical because natural uranium does not contain enough of the U235 isotope to create and sustain a fission reaction.

Uranium is enriched at plants in the United States and Europe. These plants sell enriching services at a specified dollar amount for each "separative work unit" of service ("SWU"). A SWU reflects the processing work needed to enrich a unit of gas to a specified concentration of the U235 isotope. Beginning in the mid-1970s, NSP contracted with the Department of Energy's ("DOE") plant in Oak Ridge, Tennessee for its enrichment needs. These contracts were designed to secure NSP's projected enrichment needs well into the future at a fixed price per SWU. They were structured as "take or pay" contracts. That is, NSP was bound to pay for the SWUs whether it used them or not. In the mid-1980s, it became apparent that NSP had overestimated its enriched uranium needs, primarily because it abandoned plans to complete another nuclear power plant. At the same time, the worldwide price for enriched uranium fell below the contract price. Thus, in 1984, NSP found itself bound to take or pay for SWUs that it no longer needed at a price that greatly exceeded the prevailing market price. Consequently, NSP began to sell or assign various amounts of the DOE SWUs to third parties in order to mitigate its losses. Nevertheless, NSP still lost $1,286,634 on these contracts in 1985 and $811,647 in 1986. This case involves, how, for tax purposes, NSP treated these losses.

NSP also incurred substantial losses in 1984. These losses are not at issue in this case, however, because NSP failed to file a refund claim for the 1984 losses before the filing deadline expired.

When the tax department at NSP prepared and filed NSP's tax returns for 1984, 1985, and 1986, it was apparently unaware that NSP had sold or assigned the DOE SWUs to third parties. Therefore, it included the amount of the contract losses in the cost basis of the SWUs that Westinghouse used to fabricate the fuel assemblies acquired by NSP in those years. As a result, NSP depreciated the amount of the DOE contract losses over the life of the fuel assemblies that NSP used to generate nuclear power instead of currently deducting them as an ordinary and necessary business expense, an ordinary loss, or a capital loss. When, in 1994, the NSP's tax department learned that the DOE SWUs had been sold or assigned to third parties at a loss in 1984, 1985, and 1986, it timely filed refund claims for the tax years 1985 and 1986. In these refund claims, NSP sought to recover the difference between the taxes that it paid while depreciating the DOE contract losses and the lower taxes it would have paid had it currently deducted them. That difference, with interest, is approximately $1,800,000.

The IRS denied the refund claims. It construed the refund claims as a veiled attempt by NSP to change its method of accounting for the contract losses without first obtaining the consent of the Commissioner of the IRS as required by law. See 26 U.S.C. § 446(e). Since NSP had not obtained the Commissioner's consent, and the time for doing so had lapsed, the IRS concluded that NSP could not now change how it treated the contract losses. NSP challenged this ruling by filing this action, arguing that it was merely attempting to correct a mistake on its earlier returns and that it had never adopted a method of accounting for these losses.

The parties cross-moved for summary judgment. As with the previous issue, the district court referred the matter to a magistrate judge. The magistrate judge recommended finding in favor of the United States. The district court reviewed the magistrate judge's report and agreed with its recommendations. Specifically, the district court concluded that because NSP's mistake implicated the proper timing of taking a deduction, NSP sought to change its method of accounting, and was therefore required to obtain the Commissioner's consent before doing so. Consequently, the district court granted summary judgment in favor of the United States. See Northern States Power, 952 F. Supp. at 1348-49. NSP appeals.

II.

We review the district court's grant of summary judgment de novo. See, e.g., Bremen Bank Trust Co. v. United States, 131 F.3d 1259, 1264 (8th Cir. 1997). Summary judgment is appropriate when there is no genuine issue of material fact and the moving party is entitled to judgment as a matter of law. Fed.R.Civ.P. 56(c). Because the parties have stipulated to the relevant facts, the questions on appeal solely concern whether the district court's judgment should be affirmed as a matter of law.

A. Each Group of Nuclear Fuel Assemblies Qualified for a Depreciation Deduction and an Investment Tax Credit in the Year It Was Acquired by NSP.

Under the Internal Revenue Code ("I.R.C."), sections 167 and 168, a taxpayer is allowed to claim a depreciation deduction for property it uses in its trade or business. I.R.C. §§ 167, 168. Furthermore, during the years in question, if that property also qualified as "section 38 property," a taxpayer could also claim a one-time investment tax credit ("ITC") based on the acquisition cost of the property. See I.R.C. §§ 38(b)(1), 46(c). Importantly, neither the depreciation deduction nor the ITC are available until a taxpayer places the property in service. See Treas. Reg. §§ 1.167(a)-10(b) (relating to the depreciation allowance) and 1.46-3(a) (relating to the ITC). The United States concedes that the cost of the nuclear fuel assemblies is a capital expenditure and that each group of assemblies is a unit of depreciable property with a five-year life. See Rev. Rul. 72-507, 1972-2 C.B. 198, as clarified by Rev. Rul. 74-237, 1974-1 C.B. 70; Rev. Proc. 74-29, 1974-2 C.B. 482. Likewise, the United States concedes that any nuclear fuel assemblies placed in service on or before December 31, 1986 qualified for an ITC in the year in which they were first placed in service. Hence, the issue before us is primarily one of timing: namely, when were the two groups of nuclear fuel assemblies first placed in service by NSP?

As relevant here, the ITC is not available to property placed in service after December 31, 1986. See I.R.C. §§ 49(b)(1) and (e)(1), as enacted by the Tax Reform Act of 1986, Pub.L. No. 99-514, § 211, 100 Stat. 2085, 2166 and § 203(b)(1)(A), 100 Stat. at 2144.

The United States argues that the district court wrongly concluded that each group of nuclear fuel assemblies were placed in service in the taxable year NSP acquired them from Westinghouse. It contends that the fuel assemblies were not placed in service until NSP inserted them into the reactor core and used them to generate salable electric power. Thus, according to the United States, NSP is only entitled to begin depreciating the two groups of fuel assemblies in the year that they were inserted into the reactor core, 1986 and 1987, respectively, and may only claim an ITC for the group of fuel assemblies that it inserted in 1986. We find the United States' contention unpersuasive.

We begin by noting that "placed in service" is not synonymous with "used." Rather, under the Treasury regulations governing both depreciation deductions and the ITC, see Treas. Reg. §§ 1.167(a)-11(e)(i) and 1.46-3(d)(1)(ii), respectively, property is deemed to be placed in service when it is "first placed in a condition or state of readiness and availability for a specifically assigned function." Treas. Reg. § 1.167(a)-11(e)(i); accord Treas. Reg. § 1.46-3(d)(1)(ii). While the United States correctly observes that we have previously construed "a condition or state of readiness and availability" to require property to be "productive on a fairly consistent basis," United States v. Tierney, 947 F.2d 854, 866 (8th Cir. 1991) (citation omitted), the United States forgets the context in which that statement was made. In Tierney, we discussed when a number of ethanol plants were first placed in service. In so doing, we observed that before the taxpayer could claim tax credits for these plants, the plants had to be complete and actually producing ethanol. Id. Here, the issue is not when was NSP's nuclear power plant placed in service, but instead when were replacement fuel assemblies placed in service? Prairie Island I has been consistently producing electrical power since 1973, and NSP seeks a deduction and an ITC not for the entire plant, but for an entirely separate unit of property, the replacement nuclear fuel assemblies. Consequently, we find the United States' reliance on Tierney, and other cases involving entire plants, facilities, or stores not yet open for business to be misplaced. See, e.g., Siskiyou Communications, Inc. v. Commissioner, 60 T.C.M. (CCH) 475 (1990) (new telephone switching system not placed in service until it could process calls); Consumers Power Co. v. Commissioner, 89 T.C. 710 (1987) (hydroelectric plant not placed in service until it consistently generated electrical power); Piggly Wiggly Southern, Inc. v. Commissioner, 84 T.C. 739 (1985), aff'd on other grounds, 803 F.2d 1572 (11th Cir. 1986) (air conditioner and humidifier units purchased for unopened stores not placed in service until stores opened for business).

Importantly, as illustrated by Treasury Regulation section 1.46-3(d), property can be in a "condition or state of readiness and availability" and yet not be "in use." It provides as follows:

In the case of property acquired by a taxpayer for use in his trade or business . . ., the following are examples of cases where property shall be considered in a condition or state of readiness and availability for a specifically assigned function:

(i) Parts are acquired and set aside during the taxable year for use as replacements for a particular machine (or machines) in order to avoid operational time loss.

(ii) Operational farm equipment is acquired during the taxable year and it is not practicable to use such equipment for its specifically assigned function in the taxpayer's business of farming until the following year.

(iii) Equipment is acquired for a specifically assigned function and is operational but is undergoing testing to eliminate any defects.

Treas. Reg. § 1.46-3(d)(2).

All of these examples depict functional property acquired by a taxpayer for use in a going concern on one date, but not used by the taxpayer until a later date. Thus, the examples closely parallel the facts of this case. When the nuclear fuel assemblies arrived at the Prairie Island plant, a facility that was already producing electricity, they were fully constructed and tested according to detailed technical specifications and quality assurance plans. Before it could use the assemblies, NSP merely had to verify that they had not been damaged when they were shipped to the plant. NSP did so in the same year that it acquired them. Thus, on a fundamental level, the assemblies were ready and available for their assigned function, i.e., to refuel a nuclear reactor, in the same year that NSP acquired them. And, as such, the assemblies were "in a condition or state of readiness and availability" within the meaning of Treasury regulations sections 1.167(a)-1(e)(i) and 1.46-3(d).

The United States attempts to avoid this obvious conclusion by characterizing the examples set forth in the Treasury regulation as narrow exceptions to its basic premise that a taxpayer must actually use property before he or she may depreciate it or claim an ITC. Specifically, it contends that fuel assemblies are not like the spare parts in the first example because NSP did not acquire them to avoid operational time loss. It then attempts to limit the second example to those situations in which a taxpayer acquires property in one year, intends to use it in that year, but is prevented from doing so by forces beyond the taxpayer's control. Since NSP acquired the fuel assemblies for use in a scheduled shut down, the United States asserts that this case cannot parallel the example of the farmer's tractor because the timing of the shut down is within NSP's control. Finally, the United States asserts that the third example demonstrates that the fuel assemblies were not placed in service until after refueling because (1) any equipment that requires such a complex installation process cannot be considered operational; and (2) the fuel assemblies had to undergo start-up physics testing before the reactor could return to full power. The United States' arguments miss their mark.

We turn first to the spare parts example. The record demonstrates that manufacturing nuclear fuel assemblies is complex and time consuming, requiring approximately two-year lead times and detailed engineering. The lead time is necessary because Westinghouse manufactures fuel assemblies for other parties in addition to NSP, and each fuel assembly must be fabricated according to technical specifications that are unique to each reactor. Early delivery of the fuel assemblies is designed to ensure that the assembly is fabricated and delivered to the nuclear power plant in time for post-delivery inspection and the scheduled shut down. Early delivery also ensures that if NSP needs to shut down the reactor ahead of schedule, it can refuel the reactor without incurring the cost of an additional shut down. Even so, the record indicates that financing and storing fuel assemblies is extremely expensive. Thus, the final delivery date for fuel assemblies reflects a balance between two competing concerns: on the one hand, NSP needs to have the nuclear fuel assemblies at the reactor early enough to reduce the risk that the normal operating schedule of the plant will be delayed; but on the other hand, NSP needs to avoid incurring unnecessary carrying costs. In this way, the nuclear fuel assemblies resemble the "parts acquired and set aside during the taxable year for use as replacements" described by the first example set forth in Treasury regulation section 1.46-3(d)(2).

Furthermore, we do not see a distinction between the operational farm equipment described in the second example and the nuclear fuel assemblies at issue here. While some courts have suggested that this example is limited to circumstances in which the taxpayer is prevented from using property by external factors, see, e.g., Siskiyou, 60 T.C.M. (CCH) at 477-78 (discussing SMC Corp. v. United States, 675 F.2d 113 (6th Cir. 1982); Sears Oil Co. v. Commissioner, 359 F.2d 191 (2d Cir. 1966) and Schrader v. Commissioner, 34 T.C.M. (CCH) 1572 (1975), aff'd, 582 F.2d 1374 (6th Cir. 1978)), we do not. First, the language of the example contains no such limitation. Rather, it describes a situation in which it is not "practicable" for the taxpayer to use the property in the same year that he or she has acquired it. Here, it is not practicable for NSP to refuel the reactors until spring when demand for electricity is lower.

Second, we decline to engage in the potentially endless exercise of delineating circumstances within the taxpayer's control from circumstances beyond the taxpayer's control. Take, for instance, the circumstances involved in Sears Oil. In Sears Oil, the taxpayer purchased a barge in the fall of 1957 and towed it to a canal in upstate New York to finish outfitting it. Even though the taxpayer finished outfitting the barge in 1957, it was unable to use the barge until 1958 because the canal froze over. 359 F.2d at 198. The Sears Oil court rejected the IRS's argument that the barge was placed in service in 1958, and held that it was placed in service in 1957. Id. The United States posits that Sears Oil demonstrates that the second example of property in a "condition or state of readiness and availability" only applies when external factors limit the use of the property in the acquisition year because Sears Oil involved a frozen canal, and a frozen canal is clearly beyond the taxpayer's control.

We think it less clear. After all, the taxpayer in Sears Oil decided to tow a barge to upstate New York in the late fall, and it is hardly surprising that canals in upstate New York freeze in the winter. Thus, we think that the taxpayer's predicament in Sears Oil was as much a product of external factors as it was the taxpayer's decision to outfit the barge in upstate New York in the late fall. The situation described in the example itself makes our analysis more concrete. The example allows a farmer who acquires operational farm equipment to depreciate it in the acquisition year when it is "not practicable to use [it] . . . until the following year." Treas. Reg. § 1.46-3(d)(2)(ii). Notwithstanding the recent effects of El Niño, a farmer living in North Dakota who decides to purchase an assembled tractor in November, does so knowing full well that he or she will not use it to plow a field until the following spring. We see little difference between that farmer and NSP. Like the farmer, NSP acquired the fuel assemblies in December, planning to refuel the reactor in the spring. And like the tractor that could (in theory) have been used to plow a field in November, the fuel assemblies were assembled and ready to refuel the reactor when NSP acquired them. Therefore, as in the second example, though the fuel assemblies were acquired in 1985 and 1986, it was not practicable for NSP to use them until the following years.

Finally, we are equally unconvinced by the United States' argument based on the third example, which involves equipment that is operational but still undergoing testing to eliminate any defects. See Treas. Reg. § 1.46-3(d)(2)(iii). Briefly, the United States contends that the fuel assemblies were not operational when NSP acquired them because the refueling process is too complex and because start-up testing must be completed before the reactor can return to full power. This contention is flawed on two levels. First, when the fuel assemblies arrived at NSP's power plant they were fully completed and ready to be inserted into the reactor. Except for inspecting them for possible damage incurred in shipping, there was nothing further that NSP had to do, or indeed could do, to make the assemblies more ready and available for refueling. Since the parts example set forth in Treasury regulation section 1.46-3(d)(2)(i) does not contain a caveat for parts that are extremely difficult to install, we decline to create one.

Second, the United States misapprehends the purpose of start-up physics testing. Start-up physics testing does not make the fuel assemblies operational. Instead, it verifies that the core design conforms to certain technical specifications and that the fuel assemblies were properly installed. Furthermore, as discussed previously, the issue here involves the nuclear fuel assemblies, not the power plant. Thus, the United States reliance on cases discussing when a new facility becomes operational (and is thereby placed in service) is misplaced. Accordingly, we conclude that NSP is entitled to the depreciation deductions and investment tax credits that it claimed on its federal income tax returns for 1985 and 1986, respectively, for the nuclear fuel assemblies that it acquired in each of those years.

B. NSP Did Not Change Its Accounting Method When It Filed A Refund Claim Seeking to Deduct Contract Losses That It Had Previously Capitalized.

On cross appeal, NSP argues that the district court erred when it held that NSP attempted to impermissibly change its method of accounting by filing refund claims seeking to currently deduct the DOE contract losses that NSP had previously capitalized. NSP contends that it did not change its method of accounting within the meaning of I.R.C. section 446(e), but instead merely corrected an error in its pre-existing method of accounting. Because we agree with NSP, we reverse and remand on this issue.

Under I.R.C. section 446(e), a taxpayer may not change "the method of accounting on the basis of which he regularly computes his income in keeping his books" without first obtaining the Secretary's consent. The applicable regulations define "method of accounting" operationally. It includes both the "overall method of accounting" that a taxpayer uses and "the accounting treatment of any item." Treas. Reg. § 1.446-1(a)(1). Hence, a taxpayer changes his or her method of accounting when he or she changes either the "overall plan of accounting for gross income or deductions or . . . the treatment of any material item used in such overall plan." Treas. Reg. § 1.446-1(e)(2)(ii)(a). Under Treasury regulation section 1.446-1(e)(2)(ii)(a), a material item is "any item which involves the proper time for the inclusion of the item in income or the taking of a deduction."

Importantly, a taxpayer does not change his or her method of accounting when he or she seeks to correct mathematical or posting errors (e.g., recording a figure in the wrong account), errors in the computation of tax liability (e.g., errors in the calculation of net operating loss), a change in treatment resulting from a change in underlying facts or any other "adjustment of any item of income or deduction which does not involve the proper time for the inclusion of the item of income or the taking of a deduction." Treas. Reg. § 1.446-1(e)(2)(ii)(b). Notwithstanding the foregoing, taxpayers who attempt to depreciate, rather than deduct, the cost of a class of depreciable assets are to be treated as though they seek to change their method of accounting if they have consistently treated the cost as an expense in the year of purchase on prior returns. Id.

Drawing upon these provisions, the United States argues that NSP seeks to change its method of accounting without first obtaining consent as required by law. At first blush, the United States' argument seems compelling: NSP originally capitalized the DOE contract losses by including them in the cost basis of the nuclear fuel assemblies that it used, thereby depreciating the losses over the assemblies' useful lives; now NSP, through its refund claims, seeks to deduct the contract losses in the year in which it incurred them. Hence, on a fundamental level, NSP's refund claims implicate the timing of a deduction, which may lead one to believe that the refund claims change NSP's method of accounting. Nevertheless, we conclude that NSP does not seek to change its method of accounting for contract losses, but rather seeks to correct an error akin to a posting error.

Both parties agree that as a public utility, see 16 U.S.C. § 824(e), NSP uses two distinct methods of accounting, one for financial accounting purposes, and one for federal income tax purposes. For financial accounting purposes, the Federal Energy Regulatory Commission ("FERC") requires NSP to use a uniform system of accounting. See 18 C.F.R. § 101. This system of accounting is used in ratemaking. It seeks to match the depreciation expense for the fuel assemblies with the actual British Thermal Units of energy that they produce. Hence, under the FERC system, NSP uses specific procedures to account and record costs related to the production of electricity. Specifically, NSP uses a capital account called a "work order" to reflect the total acquisition costs of each group of fuel assemblies that NSP acquires for refueling. The work order capital account includes the cost of the SWUs used to fabricate the assemblies. Pursuant to FERC accounting rules, NSP accounted for the DOE contract liabilities by including them in the relevant work order capital accounts with the cost of the enriched uranium actually used to produce electricity, less any amount that NSP was able to recover by selling the DOE SWUs to third parties.

For federal income tax purposes, NSP reports its taxable income under the accrual method of accounting. That is, NSP reports its income in the taxable year in which it has a fixed right to receive that income; its business expenses in the taxable year in which its liability is fixed and the amount of that liability is reasonably certain; and any ordinary losses in the taxable year in which the loss has been incurred and any further reimbursement for that loss is no longer reasonably foreseeable. See Treas. Reg. § 1.446-1(c)(1)(ii).

The parties have stipulated that during the relevant taxable years, NSP, using the accrual method of accounting, reported any gains from the sales of unneeded or unwanted coal or oil as ordinary income, and any losses from these sales as current expenses. The parties have also stipulated that NSP's tax department was unaware that the figures in the work order capital accounts for 1984, 1985, and 1986, which reflected FERC accounting rules, included the net unrecouped losses from the DOE enrichment contracts. Consequently, instead of subtracting these losses from the work order capital accounts and reporting them separately, NSP's tax department mistakenly included them in the capital cost of the fuel assemblies that NSP acquired and put in service for those years. Finally, the parties have stipulated that if NSP's tax department had known that the figures in the work order capital accounts for 1984, 1985, and 1986 included the net DOE contract losses, it would have subtracted them from the relevant capital accounts and currently deducted them as ordinary and necessary business expenses, and/or ordinary losses, or capital losses.

Under these facts, we are unwilling to conclude that NSP's mistake constituted anything more than a type of posting error. NSP's tax department filed the refund claims when it learned that the work order account included currently deductible contract losses. In so doing, it sought to treat the DOE contract losses in the same manner that it has consistently treated similar types of losses, i.e., the coal and oil contract losses. In this sense, NSP was not attempting to change its accounting method for a class of assets or expenses as contemplated by Treasury regulation section 1.446-1(e)(2)(ii)(b) (emphasis added), and it was not required to seek the Commissioner's consent. See, e.g., Diebold, Inc. v. United States, 891 F.2d 1579, 1582 (Fed. Cir. 1989) (distinguishing cases in which a taxpayer seeks "to account for [an item] in the same manner that it accounts for other similar items or to correct the omission of an item from a method of accounting that it otherwise consistently applies to a single category of related items" from cases in which the taxpayer seeks to abandon its consistently applied method of accounting for a class of items); Gimbel Bros. v. United States, 535 F.2d 14 (Ct. Cl. 1976) (holding that taxpayer did not change its method of accounting when it sought a tax refund for two years in which it mistakenly reported sales using the accrual method when taxpayer had previously elected the installment method of accounting); Korn Indus. v. United States, 532 F.2d 1352 (Ct.Cl. 1976) (holding that taxpayer did not change its method of accounting when it included three previously omitted items in finished goods inventory, even though it affected the timing of a deduction, because taxpayer's actions were consistent with how taxpayer treated similar items in that class of expenditures); Evans v. Commissioner, 55 T.C.M. (CCH) 902 (1988) (holding that taxpayers did not change their method of accounting when they sought to correct the mistaken use of the accrual method of accounting for bonuses when they used the cash method of accounting for all items of income and expense).

The United States counters that the aforementioned cases are inapposite. It argues that because NSP treats the costs of coal and oil as a deferred asset for both tax and financial accounting purposes, NSP's treatment of the losses on the sales of coal and oil is irrelevant. According to the United States, the losses are apparently distinguishable because when NSP decided to capitalize the DOE contract losses for financial accounting purposes, NSP necessarily decided to capitalize them for tax accounting purposes. We find this contention unpersuasive.

Although NSP does not treat the costs of coal and oil that it uses in its power plant as capital expenditures, it does not follow that losses on contracts for the sale of coal and oil are somehow different from losses on contracts for the sale of SWUs. First, the issue here is not how NSP accounts for the underlying subject matter of the contract whether it be coal, oil, or SWUs. The issue is how NSP treats losses on sales for fuel that it does not need. The United States does not explain why this common denominator for tax purposes should not hold. Second, the United States forgets that NSP does not treat the SWUs as an independent capital expenditure. Rather, it treats the fuel assemblies containing uranium pellets manufactured from enriched gas as a capital expenditure. Because the DOE contract losses involve SWUs that NSP never used, the SWUs at issue never became capital assets. And finally, NSP's decision to capitalize the DOE contract losses for financial accounting purposes was driven by FERC accounting rules. Since NSP entered the DOE contracts in good faith to secure a stable supply of SWUs, the losses that stemmed from these contracts were passed on to NSP's customers. The manner that NSP chose to pass on the losses had to be consistent with FERC accounting rules, and the United States has stipulated that it was. The United States has also stipulated that NSP's method of accounting for ratemaking and financial purposes differs from NSP's method of accounting for tax purposes on numerous levels (e.g., the length of the useful life of a fuel assembly and the rate of depreciation). Against this backdrop, we find that NSP's accounting method for the DOE contract losses for ratemaking purposes should not dictate NSP's accounting method for tax purposes. Thus, we find that NSP's mistakes were akin to posting errors and did not constitute a method of accounting. Accordingly, NSP did not change its method of accounting when it filed refund claims seeking to deduct the DOE contract losses on its 1985 and 1986 tax returns. We therefore remand to the district court so that it can determine the proper amounts of the refund for each of these years.

Because we find that NSP did not change its method of accounting, we do not address the parties' arguments relating to whether taxpayers must obtain the Commissioner's approval before they may depart from an impermissible method of accounting.

III.

We affirm the district court's opinion in part and reverse and remand in part, with instruction to proceed in accordance with this opinion.


Summaries of

Northern States Power Company v. United States

United States Court of Appeals, Eighth Circuit
Aug 12, 1998
151 F.3d 876 (8th Cir. 1998)

holding that a power company effectively had committed a posting error and did not change its method of accounting when it filed a refund claim seeking to deduct contract losses as it consistently had other losses

Summary of this case from O'Shaughnessy v. C.I.R

holding that a taxpayer need not obtain consent where it seeks to correct its treatment of an item to bring it into conformity with the manner in which it has consistently treated other similar items

Summary of this case from Cargill Inc. v. U.S.

holding that a power company corrected an error and did not change its method of accounting when it changed from depreciating to deducting contract losses as it deducted other kinds of contract losses

Summary of this case from H. E. Butt Grocery Company v. U.S.
Case details for

Northern States Power Company v. United States

Case Details

Full title:Northern States Power Company, Cross-Appellant/Appellee, v. United States…

Court:United States Court of Appeals, Eighth Circuit

Date published: Aug 12, 1998

Citations

151 F.3d 876 (8th Cir. 1998)

Citing Cases

O'Shaughnessy v. C.I.R

See notes 1, 2 supra. By shifting assets, Cardinal argues, the IRS did not change Cardinal's accounting…

Cargill Inc. v. U.S.

Cargill's final argument is that it does not need to obtain the Commissioner's consent because under its…