Nos. 81-1986, 81-2037.United States Court of Appeals, Fourth Circuit.Argued May 4, 1982.
Decided August 27, 1982.
James R. Trotter, Rocky Mount, N.C., for A. C. Monk Co., Inc.
William A. Whitledge, Tax Div., Dept. of Justice, Washington, D.C. (Samuel T. Currin, U.S. Atty., Raleigh, N.C., Glenn L. Archer, Jr., Asst. Atty. Gen., Michael L. Paup, Robert T. Duffy, Tax Div., Dept. of Justice, Washington, D.C., on brief), for the United States.
Appeal from the United States District Court for the Eastern District of North Carolina.
Before WINTER, Chief Judge, PHILLIPS, Circuit Judge, and CLYDE H. HAMILTON, United States District Judge for the District of South Carolina, sitting by designation.
JAMES DICKSON PHILLIPS, Circuit Judge:
 The Internal Revenue Code allows an investment tax credit for tangible property used as an integral part of manufacturing or used for the bulk storage of fungible commodities, but specifically excludes “a building and its structural components” from the tax credit. I.R.C. § 48(a)(1)(B). The issues in this case require us to determine whether various parts of a tobacco processing plant of the taxpayer A.C. Monk Company (Monk or taxpayer) are properly to be considered buildings or their structural components (giving Monk no tax credit); or, alternatively, other kinds of structures that are used as integral parts of manufacturing or for bulk storage (giving the credit). On cross-appeals by the Government and the taxpayer we affirm in part, reverse in part, and vacate and remand for further proceedings in part.
 During 1971-72, Monk, a major tobacco exporter, designed and built in Farmville, North Carolina, a new plant in which to store and process tobacco. It included, inter alia, a large room used to receive, regrade and store the tobacco bought at auction (green storage room); a raised, concrete railroad dock to load and unload railroad cars; electrical systems providing power for both lighting and machinery; and a section of raised roof to accommodate the height of the hogshead loader and press machinery (the high bay).
 On its 1972-75 tax returns, Monk claimed investment tax credits for numerous portions of the new facility, including those above identified. The IRS disallowed certain of the claims and assessed and collected a larger tax. Monk then sued for refund of the disputed amount. Following a bench trial, the district court determined that Monk was entitled to the tax credit in respect of nine items disallowed by the IRS, and was not entitled to the credit in respect of three others. The Government’s appeal challenges only three of the nine determinations adverse to it: those related, respectively, to the high bay, the railroad dock, and a portion of the electrical system. Monk’s appeal challenges only one of the three determinations adverse to it: that related to the green storage room.
 For reasons that follow, we affirm the district court’s determination that Monk was not entitled to the credit for the green storage room; we reverse its determinations that Monk was entitled to the credit for the high bay, and the railroad dock; and we remand for reconsideration its determination that Monk was entitled to the credit for a portion of the electrical system. With
respect to each of the latter three determinations we conclude that the district court erred as a matter of law in its application of the controlling statute and regulations to the facts as found.
 Congress enacted the investment tax credit provisions to create incentives that would increase employment and spur productivity and output by the formation of capital equipment. In not allowing credits for investments in buildings and their structural components, Congress felt that businesses would respond more greatly to net price reductions in integral manufacturing and production machinery than in buildings and that such investments would further the economy-spurring goals of the tax credit more than would investments in buildings. See H.R. Rep. No. 1447, 87th Cong., 2d Sess. (1962), 1962-3 C.B. 405, 413.
 In exempting buildings and their structural components from the investment tax credit, Congress intended that the term “building” be accorded its “commonly accepted meaning, that is, a structure or edifice enclosing a space within its walls, and usually covered by a roof.” H.R. Rep. No. 1447, 87th Cong., 2d Sess. (1962-3 C.B. 405, 516). Accord, S. Rep. No. 1881, 87th Cong., 2d Sess. (1962-3 C.B. 707, 858). The house and senate reports gave as examples of the purposes of buildings, “to provide shelter or housing or to provide working . . . space.” Id.
 Treasury Department regulation § 1.48-1(e)(1) uses this Congressional language
in defining the term “building.” In applying the definition, courts have consistently rejected a simple appearance test, see, e.g., Brown-Forman Distillers Corp. v. United States, 499 F.2d 1263, 1269-71 (Ct.Cl. 1974), but rather insist that a structure both appear and function like a building to be a building for purposes of IRC § 48(a)(1)(B). The appearance prong comes from the sentence in the regulation (tracking language in the house and senate reports) that a building “generally means any structure or edifice enclosing a space within its walls, and usually covered by a roof. . . .” Treas. Reg. § 1.48-1(e)(1). The function prong comes from language in the continuation of that sentence: “the purpose of which is, for example, to provide shelter or housing, or to provide working, office, parking, display, or sales space.” Id.
 The division among the reported cases comes in determining the scope of this function test. Some courts have determined that any structure providing shelter or work space, etc., functions as a building. See Consolidated Freightways, Inc. v. United States, 620 F.2d 862 (Ct.Cl. 1980) (trucking docks are buildings); Yellow Freight System, Inc. v. United States, 538 F.2d 790 (8th Cir. 1976) (same). Other courts have reasoned that a structure, even if it provides shelter or work space, is nevertheless not necessarily a building if it is a specialized structure not reasonably adaptable to other possible uses. See Thirup v. Commissioner, 508 F.2d 915 (9th Cir. 1974) (greenhouse not a building); Brown Williamson Tobacco Corp. v. United States, 369 F. Supp. 1283 (W.D.Ky. 1973), aff’d per curiam on reasoning of district court, 491 F.2d 1258 (6th Cir. 1974) (tobacco storage sheds not buildings). See also Catron v. Commissioner, 50 T.C. 306
(1968) (refrigerated cold-storage area of hut not a building) Palmer Olson, T.C. M. ¶ 70,296 (1970) (P-H) (quonset structures storing grain not available for credit because can store seed, fertilizer, or farm equipment).
 We conclude that Congress intended that the adaptability of a structure to other uses should be a factor — along with its appearance and the purpose it presently serves — in determining whether the structure is a building and therefore ineligible for the investment tax credit. Congress intended to encourage investment in productive equipment — to which highly specialized structures can be linked — as being most important in increasing productivity and output. But cf. Consolidated Freightways, Inc. v. United States, 620 F.2d at 872 (rejecting this argument as not providing “much assistance” in drawing line between buildings and other structures). More importantly, Treasury Regulation §1.48-1(e)(1), from which the appearance and function tests are derived, seems to limit the definition of building in this way. Although including as buildings structures such as “apartment houses, factory and office buildings, warehouses, barns, garages, railway or bus stations, and stores,” the Regulation excludes structures so closely related to productive property that they “could not be economically used for other purposes.” The Regulation gives as examples of non-buildings “oil and gas storage tanks, grain storage bins, silos, fractionating towers, blast furnaces, basic oxygen furnaces, coke ovens, brick kilns, and coal tipples.” The distinction between the two types of structures is
best seen in the relatively general purpose nature of the former.
 Additionally, Congress has extended the investment credit t “single purpose agricultural or horticultural structures,” I.R.C. § 48(a)(1)(D) (p) (as amended by the Revenue Act of 1978, Pub.L. No. 95-600, § 314, 92 Stat. 2763, 2827-28 (applicable to tax years ending after August 15, 1971)). This indicates a Congressional intent that single purpose structures not be viewed as buildings under I.R.C. § 48(a)(1)(B).
 Applying these general definitions and policies to the “high bay” at issue here, we conclude that the high bay is a structural component of Monk’s factory building. To accommodate the hogshead press and loader machinery, Monk designed an 80 by 90 foot section of its roof to be raised 15 feet above the 20 foot high roof covering the rest of the factory. The four sides and roof of this addition are made of sheet metal, the roof being covered with tar and gravel and the sides being bolted to the rest of the factory frame. Parts of this machinery not needing protection from the weather extend even higher than the raised roof.
 Following the outline of Regulation 1.48-1(e)(1), the district court reasoned that the high bay obviously was specifically designed to house the hogshead machinery and could not be used for any other purpose. The court recognized that it might not be absolutely necessary to dismantle the high bay if Monk ceased to use the machinery, but opined that the energy costs of heating and cooling the extra space may make it desirable to dismantle it. It therefore concluded that the high bay was not a structural component of Monk’s factory building but rather an integral part of its manufacturing process — and thus eligible for the investment tax credit.
 In our view, the district court insufficiently emphasized the Regulation’s requirement that a structure housing machinery, in order not to be considered part of a building, must be “so closely related to the use [of the machinery] that the structur clearly can be expected to be replaced when [the machinery] is replaced.” [Emphasis added.] The import is that the taxpayer has the distinct burden of showing the structure has no feasible alternate use than housing the specific machinery it currently houses.
 Here, the record indicates without contradiction that the hogshead machinery could be removed without damaging the high bay, and vice versa. Although the raised roof would be unnecessary for most alternate manufacturing processes which the building could reasonably house, neither would it interfere with other uses. The record reveals no estimate of the energy costs needed to heat and cool the air space of the high bay, and certainly suggests no comparative estimate between energy costs and the cost of replacing the high bay with a flat roof. In short, Monk has not met its burden of demonstrating the integral link between the high bay and the machinery it covers.
 Monk suggests that the high bay is simply the counterpart to the sunken “press pit” in which the hogshead machinery rests, and notes that the I.R.S. allowed an investment tax credit for the pit. However, the distinction between a raised roof and a sunken section of the floor is that the next user of the factory building will almost certainly have to cover the pit, for most manufacturing operations require a smooth and level floor. The raised roof, on the contrary, could easily be used in alternate operations and thus is not so closely related to the specific use of the press machinery.
 Monk’s suggested analysis would allow a tax credit whenever a company builds a nonstandard building. For example, adding a kink to the side of a square building (to accommodate some piece of machinery) would entitle the company to an investment tax credit for the kink, even though it does not hinder alternate uses. Indeed, such tax incentives may perversely lead to the building of nonstandard buildings not easily adaptable to other uses, a result clearly incompatible with the general Congressional policy of encouraging an efficient, productive economy. We hold, therefore, that
the high bay is a structural component of the factory and thus not eligible for the investment tax credit.
 Monk’s railroad dock consists of a concrete platform, columns supporting a roof, and a wall extending down approximately three feet from the edge of the roofline. Along one side of the dock is the factory’s west wall, which has openings to allow tobacco to be unloaded directly off railroad cars into the factory. The district court determined without discussion that the dock was property other than a structural component of a building used as an integral part of Monk’s processing of tobacco, and thus allowed Monk’s investment tax credit claim.
 Monk concedes that the railroad dock furnishes shelter for employees loading and unloading railroad cars and thus functions as a building, but argues that it does not look like a building and thus does not pass the appearance portion of the two-part building test.
 We disagree. Although the dock’s workplace is not enclosed by full walls on all sides, it is an obvious extension of the factory structure much as a porch is an extension of a house. I Yellow Freight Systems, Inc. v. United States, 538 F.2d 790
(8th Cir. 1976), the court determined that truck docks sharing a common wall with office buildings were themselves buildings within the meaning of I.R.C. § 48(a)(1)(B), even though at least some of the docks had no walls or heating or plumbing. Accord, Consolidated Freightways, Inc. v. United States, 620 F.2d 862
(Ct.Cl. 1980). The railroad dock would be useful in many other manufacturing processes, for loading and unloading material in and out of the factory is a necessary step in the production process. Whether termed a building in its own right, as i Yellow Freight and Consolidated Freightways, or a structural component of the adjoining factory, we are satisfied that Congress intended a railroad dock of this type to be ineligible for the investment tax credit.
 The green storage room is a large room at the northern end of the factory building. It shares the same sidewalls and roof as the factory and is separated from the factory by a nonloadbearing firewall. Except for the more numerous skylights in its roof (so the stored tobacco can be viewed in natural light), the green storage room is structurally indistinguishable from the remainder of the factory. Monk uses the northwest corner of the room to receive and regrade the tobacco. The remaining space (about 85-90%) is used to store the tobacco, by grade, on large metal racks. When needed, the tobacco is taken through doors in the firewall for processing in the factory.
 Monk concedes, as it must, that the green storage room looks like a building, but argues that it does not function as a building but rather constitutes a facility “for the bulk storage of fungible commodities” eligible for tax credit under I.R.C. §48(a)(1)(B)(iii).
 One difficulty with Monk’s argument is that § 48(a)(1)(B) apparently excepts buildings from its definition of credit-eligible property even if they would be storage facilities under § (B)(iii). Congress added § (B)(iii) in 1971 in an attempt to clarify the distinction between buildings and storage facilities.
 The Senate Report explained:
The committee’s bill specifically provides that property eligible for the investment credit includes a facility, used in connection with any of the activities referred to above (specified in sec. 48(a)(1)(B)(i)) for the bulk storage of fungible commodities. . . .
For a “storage” facility to be eligible for the credit, it must be used principally as a storage facility. Thus, if the facility has a work area in which more than a de minimis amount of processing and handling of the stored commodities can be carried on, it will not be considered to be used principally as a storage facility.
 S.Rep. No. 92-437, 92d Cong., 1st Sess., reprinted in
1971 U.S. Code Cong. Ad.News 1918, 1936.
 We believe the appropriate analysis to resolve the conflicting themes is, again, to focus on whether the structure provides alternate uses. If the structure can only be used to store one type of commodity, and has no other feasible use, it does not function as a building. Thus, as the Treasury Regulation mandates, structures such as oil and gas storage tanks and silos are not buildings, even though they enclose a space within walls and a roof and provide shelter.
 Here, the district court found that the green storage room afforded working space for some 60 to 80 people during the tobacco season and is used to receive, import, and re-grade tobacco as well as store it. Certainly this is more than de minimis activity.
 Monk urges that, even if the entire room is not a storage facility, we consider the 85-90 percent of the room in which these other activities do not occur as a storage facility eligible for the credit. Monk argues that it could have built a separate building to conduct its receiving and regarding activities (or even partitioned the existing green storage room) and thereby received a tax credit for the remaining storage area; and suggests that Congress did not intend taxpayers to take economically unwise actions simply to get the credit.
 We agree that Congress should ordinarily be presumed to have intended economically rational effects from its tax laws. The flaw in Monk’s argument is that the green storage room would be ineligible for a tax credit even if no other activities occurred there. Monk likens the green storage room, with its “unique rack system” for storing tobacco, to a “rectangular, horizontal silo.” However, the green storage room clearly is adaptable to other uses. It is a large room which the district court found to be simply a large warehouse. Not only are receiving, regrading and inspection activities possible (as shown by current activity), but testimony in the record suggests that it would be possible to convert the room for use in factory operations or general processing. Thus, unlike a silo which is not economically adaptable to other uses, the green storage room must be considered a building, and thus ineligible for the investment tax credit.
 The power company provides electricity to the exterior of Monk’s factory. Under the system designed and purchased by Monk, the power is split at two “switchgear” panel boards — one supplying power to the factory and the other to the office structure containing laboratory and computer rooms. Power to the factory is conducted by a large, copper bus duct running the length of the factory, designed to be
simply tapped to provide power at any location. Tap-ins feed the electricity through several panel boards and transformers to motor control boxes (essentially on-off switches for the machinery). Power for machinery and for the building itself (e.g., lights) comes through the same electrical system. Power to the office flows through a similar system.
 The I.R.S. allowed a tax credit for the motor controls and tap-in lines by which machinery is directly connected to the factory electrical system, on grounds they were integrally associated with the machinery, but denied a credit for the other elements of the factory system and for the entire office system. The I.R.S. bases this on the definition of structural component in Treas. Reg. § 1.48-1(e)(2) which includes “electric wiring and lighting fixtures; . . . and other components relating to the operation or maintenance of a building.” See note 2, supra.
 The district court rejected the government’s contention that the entire electrical systems constitute nonqualifying structural components of the building. The court emphasized the “relating to the operation or maintenance of a building” language of the Regulation, reasoning that this indicates that electrical systems used to operate machinery should not be considered a structural component of a building. The court felt that Monk should not be penalized for efficiently combining its building and operational electrical needs in a single system, and therefore allowed as a credit the portion of the systems’ costs that could be allocated to machinery operations. Thus, for example, it allowed the investment tax credit for 85 percent of the cost of panel boards containing electrical switch gear, 100 percent of the cost of certain transformers, 30 percent of the cost of other transformers, and 54 percent of the cost of the computer room and laboratory room wiring.
 This allocation approach was used by the Tax Court in Scott Paper Co. v. Commissioner, 74 T.C. 137 (1980). The government asserts its rationale is erroneous, and cites two appellate cases rejecting a “function” test for determining whether an item is a structural component. Those cases determined that removable wall partitions were not structural components of buildings, even though they functioned as walls, because they were not permanently installed. The government urges that the regulations require all electrical systems to be considered per se structural components, or at least that all permanently installed (i.e., non-removable) electrical systems be so classified.
 We agree that a narrow function test (i.e., does the system service machinery?) is inappropriate for determining whether an electrical system is a structural component. Further, we find no justification for allocating a portion of a single system as structural. The regulations never speak of such an allocation, and it seems difficult to conceive of a single bus duct, transformer, or piece of wiring being both a structural component and other property. However, we believe the government overstates the regulation’s intent in arguing all structural wiring must be considered a structural component of a building.
 We believe that the proper approach is, again, to determine whether the system has more general uses than simply operating specific items of machinery. Thus, if the wiring and other components of the electrical system could be adapted to other operations, they are structural components of the building. For example, wiring providing electricity for lighting is certainly a structural
component, because lighting is a general need of many manufacturing operations. Similarly, an electrical system providing power to machinery is a structural component if the system can feasibly be adapted to uses other than the specific machine it was designed to serve.
 One method of analyzing the issue would be to determine whether a manufacturer converting the building to an alternate process would be able, with reasonable alterations, to use the existing system, or whether he would have essentially to scrap the system and install another. Of course, to be considered a structural component the electrical system need not be adaptable to all conceivable uses, but only flexible enough that it is not inextricably linked to the present, specific machinery.
 Monk’s electrical systems were designed to meet its own operational needs, and the government, in an effort to show it is a permanent installation, emphasizes that the systems cannot be removed and used elsewhere. The relevant inquiry, however, is whether they can be reasonably adapted in the present building to more general uses. If so, they are structural components of the building.
 We are unable to determine on the present record whether Monk’s electrical systems have this adaptability and thus remand for initial determination by the district court. Without constricting the factfinding process should our assumptions prove unfounded, it seems, for example, that the bus duct (and prior elements in the system) is a structural component because it can be tapped to provide power for machinery at any point in the building. The panel boards and transformers, further down the line, may likewise be adaptable (after economically feasible alterations) to other uses, and thus also be structural components of the building.
 For the foregoing reasons, the judgment of the district court is affirmed in part, reversed in part, and remanded in part, for further proceedings consistent with this opinion.
 AFFIRMED IN PART; REVERSED IN PART; REMANDED IN PART.
§ 48(a) Section 38 Property —
(1) In general. — Except as provided in this subsection, the term “section 38 property” means —
(A) tangible personal property, or
(B) other tangible property (not including a building and its structural components) but only if such property —
(i) is used as an integral part of manufacturing, production, or extraction or of furnishing transportation, communications, electrical energy, gas, water, or sewage disposal services, or
(ii) constitutes a research facility used in connection with any of the activities referred to in clause (i), or
(iii) constitutes a facility used in connection with any of the activities referred to in clause (i) for the bulk storage of fungible commodities (including commodities in a liquid or gaseous state) . . . .
§ 1.48-1 Definition of section 38 property
(e) Definition of building and structural components. (1) Buildings and structural components thereof do not qualify as section 38 property. The term “building” generally means any structure or edifice enclosing a space within its walls, and usually covered by a roof, the purpose of which is, for example, to provide shelter or housing, or to provide working, office, parking, display, or sales space. The term includes, for example, structures such as apartment houses, factory and office buildings, warehouses, barns, garages, railway or bus stations, and stores. Such term includes any such structure constructed by, or for, a lessee even if such structure must be removed, or ownership of such structure reverts to the lessor, at the termination of the lease. Such term does not include (i) a structure which is essentially an item of machinery or equipment, or (ii) a structure which houses property used as an integral part of an activity specified in section 48(a)(1)(B)(i) if the use of the structure is so closely related to the use of such property that the structure clearly can be expected to be replaced when the property it initially houses is replaced. Factors which indicate that a structure is closely related to the use of the property it houses include the fact that the structure is specifically designed to provide for the stress and other demands of such property and the fact that the structure could not be economically used for other purposes. Thus, the term “building” does not include such structures as oil and gas storage tanks, grain storage bins, silos, fractionating towers, blast furnaces, basic oxygen furnaces, coke ovens, brick kilns, and coal tipples.
(2) The term “structural components” includes such parts of a building as walls, partitions, floors, and ceilings, as well as any permanent coverings therefor such as paneling or tiling; windows and doors; all components (whether in, on, or adjacent to the building) of a central air conditioning or heating system, including motors, compressors, pipes and ducts; plumbing and plumbing fixtures, such as sinks and bathtubs; electric wiring and lighting fixtures; chimneys; stairs, escalators and elevators, including all components thereof; sprinkler systems; fire escapes; and other components relating to the operation or maintenance of a building. However, the term “structural components” does not include machinery the sole justification for the installation of which is the fact that such machinery is required to meet temperature or humidity requirements which are essential for the operation of other machinery or the processing of materials or foodstuffs. Machinery may meet the “sole justification” test provided by the preceding sentence even though it incidentally provides for the comfort of employees, or serves, to an insubstantial degree, areas where such temperature or humidity requirements are not essential. For example, an air conditioning and humidification system installed in a textile plant in order to maintain the temperature or humidity within a narrow optimum range which is critical in processing particular types of yarn or cloth is not included within the term “structural components”.