Consumers Power Co. v. Commissioner , 89 T.C. 710 ( 1987 )


Menu:
  • SWIFT, Judge:

    In these ponsolidated cases, respondent determined deficiencies in petitioners’ Federal income tax liabilities as follows:

    Docket No. 997-81
    Petitioner.- Consumers Power Co.
    Year Deficiency
    1968 . $4,477,005.67
    1969 . 10,985.94
    1970 . 362,292.50
    1971 . 8,988,165.07
    1972 . 991,768.02
    Docket No. 1002-81
    PETITIONER: CONSUMERS POWER CO. AND SUBSIDIARIES
    Year Deficiency
    1973 . $1,829,691.88
    1974 . 5,300,982.94
    Docket No. 1005-81
    PETITIONER: MICHIGAN GAS STORAGE CO.
    Year Deficiency
    1970 . $289.19
    1972 . 4,007.76

    After concessions, the issues for decision are: (1) Whether petitioner Consumers Power Co. used a proper method of accruing income earned with respect to natural gas and electrical power provided to customers, and (2) whether, for purposes of depreciation and the investment credit, petitioner’s Ludington Pumped Storage Hydroelectric Plant was placed in service in 1972. The two issues remaining for decision pertain only to Consumers Power Co., and all references herein to “petitioner” will be to Consumers Power Co.

    FINDINGS OF FACT

    Many of the facts have been stipulated and are so found. Petitioner Consumers Power Co. is a regulated public utility company with its principal place of business in the State of Michigan. Consumers Power Co. is the parent company of the following subsidiary compánies: Michigan Gas Storage Co., Northern Michigan Exploration Co., and Michigan Utility Collection Co. For 1968 through 1972, petitioner Consumers Power Co. and its subsidiaries timely filed separate Federal corporate income tax returns. For 1973 and 1974, petitioner Consumers Power Co. and its subsidiaries timely filed consolidated Federal income tax returns.

    Accrual of Utility Income

    Petitioner is the largest utility company in Michigan and is engaged in the business of providing natural gas and electrical power to customers throughout the State. During 1974, petitioner served approximately 2 million customers.

    Petitioner is subject to regulation by the Michigan Public Service Commission (MPSC), the Federal Energy Regulatory Commission (and its predecessor, the Federal Power Commission), and the Nuclear Regulatory Commission. The MPSC regulates, among other things, petitioner’s methods of accounting, the rates petitioner charges its retail customers for natural gas and electrical power, and the issuance of securities by petitioner. The Federal Energy Regulatory Commission regulates the rates petitioner charges its wholesale customers for electrical power and is the licensor of petitioner’s hydroelectric generating plants. The Nuclear Regulatory Commission regulates the construction and operation of petitioner’s nuclear generating plants. As required by the MPSC, petitioner maintains its books and records in accordance with the Uniform System of Accounts for Electric and Gas Utilities.

    During the years in issue, petitioner filed financial and operational reports with the MPSC, the Federal Power Commission, and the Securities and Exchange Commission. Petitioner also submitted certified financial statements to its shareholders. Petitioner is an accrual basis taxpayer for both financial and Federal income tax purposes.

    Charges made by petitioner to its customers are predicated on monthly readings of the customers’ gas and electric meters.2 Petitioner issues a monthly bill to its customers, and petitioner is prohibited by the rules and regulations of the MPSC from issuing more than one bill to its customers in any one month.

    On October 24, 1974, the MPSC adopted specific rules governing billing practices of public utility companies located in Michigan. Those billing rules provide, in part, as follows:

    R 460.2102. Definitions.
    Rule 2. (1) “Billing month” means a utility service comsumption period of not less than 26 nor more than 35 days.
    *******
    (6) “Cycle billing” means a system employed by a utility which results in the rendition of bills for utility service to various customers on different days of any one calendar month.
    *******
    R 460.2111. Billing frequency.
    Rule 11. A utility shall render a bill once during each billing month to every residential customer in accordance with approved rate schedules.
    *******
    R 460.2115. Cycle billing.
    Rule 15. A utility may bill its customers on a cyclical basis if the individual customer receives each billing on or about the same day of each billing month. If a utility changes meter reading routes or schedules, billing cycles may be altered upon 10 days written notice to the affected customer.

    For Federal income tax purposes, petitioner, since its organization in 1910 through the years in dispute, has utilized the “meter reading and billing cycle” method of accruing income earned from its utility business. The meter reading and billing cycle method of accruing utility income complies with the above requirements of the MPSC and is in accordance with generally accepted accounting principles applicable to public utility companies.

    Under petitioner’s meter reading and billing cycle method of accounting, the amount of utility income accrued throughout the year is based, generally, on the quantity of natural gas and electrical power that is used by its customers. That amount is determined by employees of petitioner who each month visit the premises of all of petitioner’s customers, physically inspect and read the customers’ utility meters, andt record the level of gas and electrical power used by its customers as indicated on the meters.3

    As indicated, under the rules of the MPSC, petitioner may bill each customer only once a month. Therefore, each year is divided by petitioner into 12 meter reading and billing cycles. Each cycle is represented by 1 month. The utility meters of petitioner’s customers are read 12 times a year (or once a month), and, based on such readings, petitioner bills its customers 12 times a year (or once a month). Understandably, petitioner’s staff of “meter readers” cannot read all of its customers’ utility meters on 1 day of the month. Accordingly, petitioner’s utility service area is divided into 21 districts, representing 21 geographic areas, and all of petitioner’s customers are divided among the 21 districts. The 21 districts are a function of the average 21 working days that exist in each month (after taking into account weekends and holidays). Each district is assigned a particular working day of the month on which the utility meters within that district are read by petitioner’s meter readers.

    Within a few days after the meters have been read in a particular district and based on the level of power used by each customer for each monthly period as determined by the meter readers, bills are prepared and mailed to all of the customers assigned to that district. In the above manner, petitioner’s meter readers and billing personnel are occupied each working day of the year — each day reading meters and preparing and mailing bills to customers in 1 of the 21 districts, and each month repeating the process for each of the 21 districts. Under this method, there are 252 separate, district-wide days of meter readings in each year.4

    Under petitioner’s method of accruing utility income, based on the information obtained each month from its customers’ gas and electric meters, each year petitioner accrues into taxable income the utility income attributable to 250 of the 252 meter-reading days that occur. This represents 99.2 percent of the total meter-reading days that actually occur in a year.

    The reason that utility income attributable to 2 of petitioner’s meter-reading days is not included in petitioner’s taxable income in the same year as the year in which those 2 meter reading days occur is because those 2 days occur late in the month of December. Early in the following January, the bills with respect to those 2 meter-reading days are prepared and mailed to customers, and utility income attributable thereto is accrued by petitioner. Thus, in January of each year petitioner accrues utility income with respect to 2 meter-reading days that occurred late in the prior December. In each year, however, utility income with respect to 12 months — and the same 12 months — is accrued into petitioner’s taxable income (i.e., in every year petitioner accrues utility income with respect to 250 of the 252 districts whose meters are read within the year and with respect to the last 2 districts whose meters were read in late December of the prior year).

    Until the mid-1970’s, the meter reading and billing cycle method of accounting for utility income, or some variation thereof, was the predominent method of accounting used by utility companies to accrue utility income for financial purposes and for Federal income tax purposes. Petitioner used this method for financial accounting purposes from its formation in 1910 through 1973. Beginning in 1974, however, petitioner changed it method of accounting for utility income solely for financial accounting purposes. The new method petitioner used for financial purposes is referred to as the full-accrual method.

    Under petitioner’s full-accrual method, utility income is accrued for financial purposes based upon estimates of the amount of gas and electrical power provided to customers each day. Rather than wait until customer’s meters are read and bills are prepared reflecting actual use of natural gas and electrical power, under the full-accrual method, petitioner accrues utility income from January 1 to December 31 of each year based on estimates of the natural gas and electrical power used by its customers during the year.

    The change in petitioner’s method of accruing utility income for financial purposes resulted in the accrual by petitioner in 1974 of $70,666,000 in utility income that was not accrued by petitioner for Federal income tax purposes.5 It is this change that occurred in 1974 in petitioner’s method of accruing utility income for financial accounting purposes that triggered respondent’s audit adjustment on this issue. Upon audit, respondent determined that petitioner’s continued use in 1974 of the meter reading and billing cycle method of accruing utility income for Federal income tax purposes resulted in an improper accrual of petitioner’s utility income, and respondent placed petitioner on the full-accrual method for Federal income tax purposes.6

    Disregarding the prior years’ and the section 481 adjustments that were made respectively by petitioner and by respondent for financial reporting and tax purposes, petitioner’s total utility income for 1974 under the different methods of accounting was as follows:

    Utility Income for 1974
    Financial Tax Percent
    accrual accrual Difference difference
    $1,103,790,000 $1,085,813,075 7$17,976,925 1.66%

    Ludington Pumped Storage Hydroelectric Plant

    In 1969, petitioner and the Detroit Edison Co., as joint owners, began construction of the Ludington Pumped Storage Hydroelectric Plant (the Ludington Plant) on the eastern shore of Lake Michigan. The Ludington Plant is the largest pumped storage plant in the United States. A pumped storage plant is not designed to provide electrical power continuously, but generates electrical power only during periods of high or peak electrical power demand. Upon completion, electrical power generated by the plant was to be made available to the Michigan Power Pool, an association of Michigan public utility companies of which petitioner is a member.8

    As we recently explained—

    A pumped storage plant provides-supplemental electrical power during periods of peak energy demand in the following manner. During evenings or weekends when electrical power demand is low, water is pumped from a lower reservior — such as a river or lake — up to a storage reservoir at a significantly higher elevation than the lower reservoir. When supplemental electrical power is needed during periods of peak energy demand, water is released from the upper reservoir. As water flows back down to the lower reservoir it passes through tunnels wherein it turns hydraulic turbine generators, thereby producing electrical power. The electrical power produced as the water flows through the turbines obviates the need during peak-demand periods to purchase or to produce a similar amount of electrical power by other, more expensive methods. The size of the upper and lower water reservoirs and the difference in their elevations determine the amount of electrial power that can be produced by a pumped storage plant.
    The construction of pumped storage plants requires a particular site or land topography because the location of the upper reservoir must be at a sufficiently high elevation in relation to the lower reservoir, but not be too distant laterally from the lower reservoir, to permit efficient operation. Pumped storage plants also must be built adjacent to large natural sources of water that readily can supply the water that is pumped to the upper reservoirs.
    [Stanley Works v. Commissioner, 87 T.C. 389, 391 (1986).]

    The upper reservoir at the Ludington Plant consists of a man-made lake with a perimeter of approximately 5.5 miles and a depth of approximately 110 feet. It covers an area of 842 acres and has a capacity of approximately 27 billion gallons. The water level of the reservoir must be maintained within certain levels in order properly to operate the turbine generators. Those levels are known as low-pond (the minimum acceptable level for operation), mid-pond, and high-pond (the maximum allowable water level for operation). Low-pond is 875 feet above sea level, mid-pond is 910 feet above sea level, and high-pond is 942 feet above sea level. Lake Michigan serves as the lower reservoir for the Ludington Plant.

    The turbines and other equipment associated with the physical facilities at the Ludington Plant consist primarily of six tunnels or penstocks, the pumphouse, and six reversible pump-turbine generators which are designated unit 1 through unit 6. The penstocks connect the upper reservoir with the pumphouse in which the pump-turbine generators are located. The plant was designed so that upon the successful completion of preoperational testing of unit 1, the Ludington plant would become operational and would begin selling electrical power to the Michigan Power Pool even though the other five generating units were still under construction.

    In September of 1972, petitioner received final approval from the Federal Power Commission to fill the upper reservoir and to begin regular operation of the plant, subject to the successful completion of preoperational testing on unit 1. On October 23, 1972, as part of the preoperational testing, unit 1 began pumping water from Lake Michigan into the upper reservoir. Tests were conducted in the pumping and generating modes at the low-pond, mid-pond, and high-pond levels of the reservoir.

    From October 23 through December 7, 1972, unit 1 pumped water into the reservoir on the dates, for the lengths of time, and to the pond elevations indicated below:

    Elapsed pumping time Date (hours: minutes) Highest pond elevation in feet above sea level
    10/23/72 0:40 838
    10/24/72 5:35 839
    10/25/72 5:05 847
    10/26/72 3:54 8{>2
    11/09/72 4:57' 852
    11/10/72 7:24 864
    11/11/72 5:30 871
    11/12/72 8:18 882
    11/18/72 1:41 879
    11/20/72 0:43 879
    11/21/72 6:54 885
    11/23/72 23:27 908
    11/24/72 1:55 910
    11/28/72 0:58 907
    11/29/72 7:57 913
    11/30/72 7:37 920
    12/01/72 4:03 920
    12/02/72 0:30 921
    12/03/72 10:05 930
    12/05/72 8:00 936
    12/06/72 4:40 936
    12/07/72 0:52;, 937
    Total 120:45,

    On November 18, 1972, the generating mode of unit 1 was initially synchronized (i.e., placed on line) with petitioner’s electrical transmission system and produced electrical power for 1 hour and 5 minutes. From November 18 through December 5, 1972, still as part of the preoper-ational testing, unit 1 generated electrical power as follows:

    Generating time Total megawatt hours Average megawatts generated Average percentage of rated
    Date (hours: minutes) generated per hour capacity 1
    11/18/72 1:05 60 55 18%
    11/19/72 0:51 130 153 49
    11/20/72 4:12 490 117 38
    11/21/72 11/27/72 3:48 3:11 510 550 134 173 43 55
    11/28/72 2:50 370 131 42
    11/30/72 4:18 1190 277 89
    12/05/72 6:16 1280 204 66
    Total 26:31 4580
    which is 312 megawatts per hour. Rated capacity, in general, is the maximum continuous output of electrical power that a generator is capable of producing.

    Of the 4,580 megawatt hours of electrical power unit 1 generated during the preoperational testing, 4,447 megawatt hours were sold to petitioner’s customers.

    During testing, unit 1 had operated successfully at its maximum rated capacity of 312 megawatts at the mid-pond level. By December 7, 1972, the preoperational testing of unit 1 was completed at the low-pond and mid-pond levels of the reservoir, but a number of tests were not yet completed at the high-pond level.

    On the evening of December 7, 1972, in conjunction with one of the final preoperational tests in the pumping mode, electrical power to unit 1 was temporarily disrupted. When this happened, unit 1 did not automatically shut down as designed, but, due to a mechanical failure, reversed into the generating mode, damaging certain parts of the turbine generator. As a result, further preoperational testing was suspended, and unit 1 was shut down for repairs. The repairs to unit 1 were completed on January 9, 1973, and preoperational testing was resumed. On January 17, 1973, testing of unit 1 was completed, and unit 1 formally was accepted by petitioner from the general contractor. On January 18, 1973, electrical power generated from unit 1 was available for transmission into the Michigan Power Pool.

    In his notice of deficiency, respondent determined that petitioner’s Ludington Plant was not placed in service in 1972. Accordingly, respondent disallowed with respect to the Ludington Plant depreciation deductions totaling $1,601,411 and $1,562,032 claimed on petitioner’s 1972 and 1973 Federal income tax returns, and investment credits totaling $2,625,264 claimed on petitioner’s 1972 Federal income tax return.

    OPINION

    Accrual of Utility Income

    The proper method of accruing utility company income for Federal income tax purposes has been the subject of much litigation. A number of cases have addressed methods of accruing such income not dissimilar from the meter reading and billing cycle method at issue in this case. See, e.g., Orange & Rockland Utilities, Inc. v. Commissioner, 86 T.C. 199 (1986); Public Service Co. of N.H. v. Commissioner, 78 T.C. 445 (1982); Bay State Gas Co. v. Commissioner, 75 T.C. 410 (1980), affd. 689 F.2d 1 (1st Cir. 1982). The opinions in those cases generally have approved the meter reading and billing cycle method consistently used by the utility companies and have done so regardless of whether a different method was used for financial accounting purposes. Orange & Rockland Utilities, Inc. v. Commissioner, supra at 207; Public Service Co. of N.H. v. Commissioner, supra at 457-458. In Orange & Rockland, we noted that—

    It has long been considered a generally accepted accounting principle for utilities to accrue revenues based on either cycle meter readings or bills rendered basis with no accounting recognition of unbilled revenue, and certified public accountants issue unqualified financial statements prepared so as to accrue revenue when meters are read or the consumer is billed. * * * [Orange & Rockland Utilities, Inc. v. Commissioner, supra at 209.]

    In the Tax Reform Act of’JL986, Pub. L. 99-514, 100 Stat. 2085, 2373, however, Congress provided that for taxable years beginning after December 31, 1986, utility income must be accrued into income on the basis of the full-accrual method of accounting. Sec. 451(f)(1), I.R.C. 1986.9 With regard to taxable years beginning before August 16, 1986, Congress provided that the “meters-read” method “shall be deemed to be proper,” but did not specify what other methods of accruing utility income will be allowed. Sec. 821(b)(3) of the Tax Reform Act of 1986, supra,10 provides as follows:

    (3) Special rule for certain cycle billing. — If a taxpayer for any taxable year beginning before August 16, 1986 * * * took into account income from [utility] services * * * on the basis of the period in which the customers’ meters were read, then such treatment for such year shall be deemed to be proper.

    We emphasize that for years prior to 1987, Congress in section 821(b)(3) did not state what other methods of accruing utility income were to be allowed or disallowed. The allowability of such other methods was left to the courts.

    The reason Congress, in the Tax Reform Act of 1986, provided some retroactive relief to utility companies from respondent’s audit adjustments is explained in the Conference Committee report as follows:

    The conferees are aware that the proper accounting for utility services is presently a matter of controversy between taxpayers and the Internal Revenue Service. In order to minimize disputes over prior taxable years, the conference agreement provides that, for any taxable year beginning before August 16, 1986, a method of accounting which took into account income from the providing of utility services on the basis of the period in which the customers’ meters were read shall be deemed to be proper for Federal income tax purposes. No inference is intended as to methods of accounting for utility services not described in the preceding sentence (e.g., a method of accounting which takes income into account on the basis of the date the customer is billed for utility services). [H. Rept. 99-841 (Conf.) (1986), 1986-3 C.B. (Vol. 4) 323. Emphasis added.]

    Petitioner argues that its method of accruing utility income for tax purposes qualifies for the retroactive relief provision of section 821(b)(3). Petitioner argues that if section 821(b)(3) is construed narrowly to protect only those taxpayers who utilize a “pure” meters-read method of accruing utility income, the remedial nature of the statute will be defeated. Petitioner also argues that the language of section 821(b)(3) does not refer specifically to a method of accruing utility income but only to the “treatment” of utility income and that effectively all of its utility income was reported, or treated, in the same year in which petitioner’s meters were read. Petitioner thus contends that it qualifies for the relief provision of section 821(b)(3). Respondent argues that only those taxpayers who utilize the pure meters-read ‘ method of accruing utility income qualify for the retroactive relief provision of section 821(b)(3).

    After carefully considering this issue, we conclude that petitioner qualifies for the relief provided in section 821(b)(3). Under petitioner’s method of accruing utility income, utility income earned by petitioner from more than 99 percent of its customers is accrued into income in the same year its customers’' meters are read. The amount of utility income accrued is determined by reading the meters, and the overwhelming majority of days on which the income is accrued occur in the same year the meters are read. Only with respect to two of petitioner’s meter-reading districts and then only with respect to the late December readings in those two districts (thus, for only two out of 252 meter readings each year) is the income attributable to utility services not accrued in the same year as the year in which the respective meters are read.

    We also note that the deferral each December of the utility income attributable to the last two meter readings in December is effectively offset in each year by the accrual in January of that year of the, utility income attributable to the last two meter readings of the prior December.

    Our conclusion herein is based in part on the remedial nature of section 821(b)(3), which justifies a broad construction of the statutory language to effectuate its purpose. See Miller v. Robertson, 266 U.S. 243, 248 (1924); Morristown Magnavox Former Employees v. Marshall, 671 F.2d 194, 197 (6th Cir.), cert. denied 459 U.S. 1041 (1982); National Butane Gas Co. v. Commissioner, 11 T.C. 593, 597-598 (1948). The evidence in this case indicates that few utility companies accrue utility income on the basis of a pure meters-read method of accounting and that the majority of utility companies utilize some variation of the meter reading and billing cycle method. If we were to adopt respondent’s narrow interpretation of section 821(b)(3), we would limit the relief provided thereunder to such an extent that the statutory purpose would be undermined.

    Under the facts of this case, the deviation in petitioner’s method of accruing utility income from a pure meters-read method is not significant. For the years in dispute, we conclude that petitioner qualifies for the relief provided by section 821(b)(3).

    Ludington Pumped Storage Hydroelectric Plant

    The parties agree that petitioner is allowed a deduction for depreciation and an investment credit with respect to the Ludington Plant in the year the plant was placed in service. Secs. 38(a),11 46(a)(1), 46(a)(2), 46(c)(1); secs. 1.46-3(a)(1), 1.167(a)-10(b), 1.167(a)-ll(e)(l)(i), Income Tax Regs. The regulations provide that property will be regarded as placed in service when it is “placed in a condition or state of readiness and availability for a specifically assigned function.” Secs. 1.46-3(d)(l)(ii) and 1.167(a)-ll(e)(l)(i), Income Tax Regs. Section 1.46-3(d)(2)(iii), Income Tax Regs., provides that property will be considered to be in a condition or state of readiness and available for a specifically assigned function when it is “operational but is undergoing testing to eliminate any defects.”

    This Court applied the above placed-in-service rules to an asset constructed by the taxpayer in Noell v. Commissioner, 66 T.C. 718 (1976). The taxpayer in Noell constructed an airport runway, which included grading, supplying a rock base, laying asphalt pavement, and installing sod. After the rock base was laid, but before the runway was completed, pilots occasionally used the runway for landing and takeoff. Despite such usage, it was concluded that the runway was not placed in service until the runway was paved and available for full service. Noell v. Commissioner, supra at 729. In Madison Newspaper, Inc. v. Commissioner, 47 T.C. 630, 637 (1967) (Court-reviewed), we held that three units of an eight-unit printing press were placed in service when the units were installed and publishing newspapers on a regular basis.

    Petitioner argues that unit 1 was placed in service in 1972 because it actually pumped water from Lake Michigan to the upper reservoir from October 23 to December 7, 1972, and because in 1972, unit 1 generated 4,580 megawatt hours of electrical power, most of which were sold to customers. Petitioner emphasizes that during 1972, unit 1 was operated successfully at its rated capacity at the mid-pond level of the reservoir, that synchronization of the unit’s electrical power generating system with petitioner’s transmission system had occurred, and that the reservoir had been filled to within 5 feet of the high-pond level.

    Respondent contends that unit 1 was not placed in service in 1972 because preoperational testing was not completed until January of 1973, because responsibility and control of unit 1 remained with the contractor until petitioner formally accepted unit 1 in January of 1973, and also because sustained, regular generation of electrical power did not occur until January of 1973. For the reasons stated below, we agree with respondent.

    Although unit 1 pumped water into the reservoir and generated electrical power during preoperational testing in 1972, unit 1 was not available in 1972 to provide electrical power on a regular basis in 1972. The amount of electrical power generated in 1972 is insufficient to establish that the Ludington Plant was available for full operation on a regular basis in 1972. The generation of electrical power and the pumping of water into the upper reservoir were both necessary parts of preoperational testing. Accordingly, the production of some electrical power (even the sale thereof) and the filling of the upper reservoir to within 5 feet of the high-pond level do not establish that the Ludington Plant was available for use in 1972. Not until January 17, 1973, after unit 1 successfully had completed all phases of preoperational testing, thereby demonstrating that it was available for service on a regular basis, was the unit in a state of readiness and availability for its specifically assigned function within the meaning of sections 1.46-3(d)(l)(ii) and 1.167(a)-ll(e)(l)(i), Income Tax Regs.

    Petitioner, relying on Sears Oil Co. v. Commissioner, 359 F.2d 191 (2d Cir. 1966), and SMC Corp. v. United States, an unreported case (E.D. Tenn. 1980, 46 AFTR2d 80-5827, 80-2 USTC par. 9642), affd. per curiam 675 F.2d 113 (6th Cir. 1982), argues that the accident which occurred on December 7, 1972, and which prevented the completion of preoperational testing in 1972, was a circumstance beyond its control, and that notwithstanding the accident, unit 1 was operational in 1972. In Sears Oil Co. and SMC Corp., however, the assets in question were held to have been placed in service during the taxable year because they were ready and available to perform their specifically assigned functions, even though the taxpayers were precluded from using the assets during the taxable year due to circumstances beyond their control. See also Schrader v. Commissioner, 582 F.2d 1374 (6th Cir. 1978), affg. a Memorandum Opinion of this Court. Although the mechanical failure that caused the shutdown of unit 1 on December 7, 1972, may have been beyond the control of petitioner, unit 1 was not yet ready and available for regular operation as of that date.

    Petitioner alternatively argues that if we conclude the Ludington Plant was not placed in service in 1972, the upper reservoir was available for use in 1972 and depreciation and investment credit are allowable in 1972 with respect to the costs of the reservoir. Respondent maintains that the reservoir cannot be regarded as placed in service until the other physical facilities were placed in service.

    Neither the statute, nor the regulations specifically address what is to be regarded as a single property for purposes of depreciation and the investment credit. In Public Service Co. of N.M. v. United States, 431 F.2d 980 (10th Cir. 1970), the 10th Circuit considered whether component assets of an electrical power plant could be placed in service before the entire plant was placed in service. The component assets included a turbine-generator, a steam generating unit, a cooling tower, a transformer, and a main power plant building. The court held that the various assets were placed in service when the entire plant became operational, stating that—

    No one of these [component assets] would serve any useful purpose to [Public Service Co.] but all of them properly fitted together by the contractor, together with the building, constituted a complete unit which was operational and served the purpose intended by [Public Service Co.]. It was this complete operational unit that [the contractor] agreed to construct from the many components and deliver to [Public Service Co.] as an electrical power generating plant. [Public Service Co. of N.M. v. United States, supra at 984. Emphasis added.]

    In Hawaiian Independent Refinery, Inc. v. United States, 697 F.2d 1063 (Fed. Cir.), cert. denied 464 U.S. 816 (1983), the court addressed the placed-in-service date of an oil refinery complex. The complex consisted of the refinery facility, an offshore tanker-mooring facility located 2 miles from the refinery and connected to the refinery by a pipeline system, and two pipelines used to transport finished products from the refinery to various storage facilities. The court agreed with the trial court’s reasoning that the component assets “functionally [formed] a single property” and held that the entire refinery complex was to be treated as a single asset for purposes of the investment credit. Hawaiian Independent Refinery, Inc. v. United States, supra at 1069. In our opinion, based on the foregoing, the Ludington Plant must be viewed as one integrated unit because the physical plant and the reservoir operate simultaneously and as a unit in order to produce electrical power.

    We conclude that the placed-in-service date of the reservoir must be deferred until the date on which unit 1 also was placed in service. Based on the record before us, we hold that for purposes of depreciation and the investment credit, the Ludington Plant was not placed in service in 1972. *

    Decisions will be entered under Rule 155.

    Where customers’ meters cannot be read in a particular month, estimates of the gas and electrical power used during the month are made and the charges are based on such estimates. Estimates are required each month for approximately 5 percent of petitioner’s customers.

    By prearrangement, in a limited number of instances {e.g., where customers prepay for 3 months at a time) a customer’s meter may be read less frequently than once a month.

    21 districts times 12 months equals 252.

    Of the $70,666,000, only $18,806,000 was attributable to 1974 and the balance of $51,860,000 was attributable to prior years. The after-tax increase in petitioner’s 1974 net income as a result of the change in petitioner’s method of accruing utility income was $33,880,000.

    Respondent’s audit adjustment resulted in an increase in petitioner’s utility income, for Federal income tax purposes, of $18,806,000 attributable to 1974, and $42,641,000 attributable to the required adjustment under sec. 481, I.R.C. 1954, as in effect in 1974. Respondent’s total increase for tax purposes in petitioner’s utility income for 1974 was $61,447,000.

    The minor difference between the $17,976,925 increase in utility income attributable to 1974 in the above schedule and the $18,806,000 figure therefor mentioned in note 6 supra, is not explained in the record.

    Utility companies that belong to the Michigan Power Pool buy and sell electrical power from each other as needed to assure that electrical power is available to customers of each utility company during peak-demand power periods.

    The average percent of rated capacity is computed by dividing the average megawatts generated per hour by the rated capacity of unit 1,

    Sec. 451(f)(1) provides as follows:

    SEC. 451(f). Special Rules for Utility Services.—
    (1) In general. — In the case of a taxpayer the taxable income of which is computed under an accrual method of accounting, any income attributable to the sale or furnishing of utility services to customers shall be included in gross income not later than the taxable year in which such services are provided to such customers.

    Hereinafter unless otherwise indicated, all section references are to the Tax Reform Act of 1986. Pub. L. 99-415. 100 Stat. 2085.

    Hereinafter unless otherwise indicated, all section references are to the Internal Revenue Code of 1954 as in effect during the years in dispute.