This is an appeal of the denial of a tax credit under the Crude Oil Windfall Profit Tax Act, Pub. L. 96-223, 94 Stat. 229 (1980) (codified in scattered sections of 7, 19, 26, and 42 U.S.C.) (repealed in part 1988) ("COWPTA") which grants oil producers an income tax credit of $3.00 for each barrel-of-oil equivalent of "oil produced from . . . tar sands" extracted through wells drilled between January 1, 1980 and December 31, 1992. The resolution of this case turns on the proper definition of "oil produced from tar sands," which the Act does not define. Shell contends the definition should be derived from commonly accepted usage in the petroleum industry and asserts that when the statute was enacted in 1980, "tar sand oil" meant "oil so viscous1 that it cannot be recovered economically through `primary production methods'2-- i.e., methods used in ordinary oil fields." The government maintains the proper definition may be found in Department of Energy Ruling 1976-4, which defines "tar sands" as "[t]he several rock types that contain an extremely viscous hydrocarbon which is not recoverable in its natural state by conventional oil well production methods including currently used enhanced recovery techniques." Department of Energy Ruling 1976-4, 10 C.F.R. ch. II Rulings 371, 372 (1980).3 Shell concedes that it is not entitled to a refund if we accept the DOE Ruling definition. Following a bench trial, the District Court adopted the definition of tar sands crafted by the DOE Ruling and denied the refund. See Shell Petroleum, Inc. v. United States, 996 F. Supp. 361, 372 (D. Del. 1997). Because we substantially agree with the District Court's well reasoned opinion, we will affirm.

Beginning in 1973, world oil prices quadrupled in less than a year after the Organization of Petroleum Exporting Countries embargoed oil sales to Western nations and then fixed prices above market levels. See Gary D. Allison, Energy Sectionalism: Economic Origins and Legal Responses, 38 Sw. L.J. 703, 705 (1984). Congress responded with the 1973 Emergency Petroleum Allocation Act (EPAA), Pub. L. No. 93-159, 87 Stat. 627 (1973) (codified as amended at 15 U.S.C. §§ 751-760h) (expired 1981), authorizing the President to regulate prices and allocate supplies of crude oil and related petroleum products.

In 1976, the Federal Energy Agency, which administered EPAA, issued Ruling 1976-4 in response to "inquiries with respect to the applicability of the [EPAA price and supply controls] to the so-called synthetic fuels (or crude oil substitutes) processed from oil shale, tar sands, coal, and other natural deposits that must be mined4 before the crude oil substitute can be extracted." Department of Energy Ruling 1976-4, 10 C.F.R. ch. II Rulings at 371. According to FEA,

at the time of enactment of the EPAA, domestic production of crude oil substitutes derived from oil shale, coal and tar sands was, as it is now, undertaken only for experimental purposes, and the synthetic products obtained thereby were not commercially available for use as refinery or petro-chemical feedstocks and were not expected to become commercially available for several years.

Id. at 372. Therefore, FEA concluded that synthetic fuels processed from tar sands were not subject to the EPAA regulatory scheme. See id. at 373.

The Emergency Petroleum Allocation Act proved counterproductive. Price controls reduced incentives to produce oil domestically and increased domestic oil consumption overall, making America more dependent on expensive imported oil. See Atlantic Richfield Co. v. United States Dep't of Energy, 655 F.2d 1118, 1121 (Temp. Emer. Ct. App. 1981). Oil prices fell in real terms after 1975, but world oil prices doubled in 1979, when the Iranian revolution severely curtailed oil exports. See Allison, supra, at 705-06. The Carter administration announced in 1979 it would phase out price controls, dramatically increasing domestic producers' short-term profits. See H.R. No. 96-304, at 5-7 (1979), reprinted in 1980 U.S.C.C.A.N. 587, 592-94. For this reason, Congress responded to the price decontrols with the Crude Oil Windfall Profit Tax Act. Title I ("Windfall Profit Tax on Domestic Crude Oil") imposed an excise tax on revenue from producer sales of crude oil, designed to capture the windfall profit. See 26 U.S.C. §§ 4986-98 (1982) (repealed 1988). To avoid suppressing the domestic oil supply, new production and other production considered sensitive to price fluctuations were taxed at lower rates or exempted. See id.§§ 4987(b), 4991-93; S. Rep. No. 96-394, at 2 (1979), reprinted in 1980 U.S.C.C.A.N. 410, 414. More importantly for our purposes, Title II ("Energy Conservation and Production Incentives") sought to reduce American dependence on imported oil by various means, including a tax credit "for producing fuel from a nonconventional source." 26 U.S.C.A. § 29 (West Supp. 1999) (as amended 1981, 1983, 1984, 1986, 1988, 1990, 1992, and 1996) (originally designated § 44D).5 Among the fuels qualified to receive the credit was "oil produced from shale and tar sands" through wells drilled between January 1, 1980 and December 31, 1992. Id. §§ 29(c)(1)(A), (f)(1)(A).

In 1983 and 1984, the tax years in question, Shell Petroleum, Inc. and certain subsidiary corporations owned working interests in properties in the Midway Sunset Field, Kern County, California. Part of the Midway Sunset Field lies over a reservoir known as "the Potter Sand formation" or "Potter Sands." For over a decade Shell had been extracting oil from Potter Sands through two steam injection techniques (steam soak, begun in 1963, and steam flood, begun in 1971). By 1980, both were well established and widely accepted enhanced recovery techniques in the petroleum industry.6

In its 1983 and 1984 tax returns, Shell did not seek the nonconventional source tax credit for Potter Sands oil, but in 1991 Shell filed amended tax returns claiming credits of $5,351,150 for Potter Sands oil produced during 1983 and 1984 from wells drilled after December 31, 1979. The tax credits were denied and Shell timely filed suit. Relying in part on the "instructive" analysis of Texaco Inc. v. Commissioner, 101 T.C. 571 (1993),7 the District Court, as noted, entered judgment for the United States. See Shell, 996 F. Supp. at 371. Shell now appeals.8

As we will discuss, Title I of COWPTA establishes that Shell's Potter Sands oil is crude oil. But the structure of the statute and the legislative history establish that crude oil is not tar sand oil. Therefore, Shell's oil cannot be tar sand oil. Furthermore, the legislative history establishes that the nonconventional source tax credit was intended to foster new energy technologies, whereas Shell's Potter Sands oil was produced in 1983 and 1984 using widely available means of production.

" `Where . . . resolution of a question of federal law turns on a statute and the intention of Congress, we look first to the statutory language and then to the legislative history if the statutory language is unclear.' " Murphy v. Dalton, 81 F.3d 343, 350 (3d Cir. 1996) (quoting Blum v. Stenson, 465 U.S. 886, 896 (1984)). As noted, COWPTA does not define "oil produced from tar sands," but "where Congress has used technical words or terms of art, it is proper to explain them by reference to the art or science to which they are appropriate." Corning Glass Works v. Brennan, 417 U.S. 188, 201 (1974) (internal quotation marks and alterations omitted). We ordinarily look to the meaning of a statutory term at the time the statute was adopted. See MCI Telecomm. Corp. v. American Tel. & Tel. Co., 512 U.S. 218, 228 (1994) (explaining that "the most relevant time for determining a statutory term's meaning" is the time when the statute became law); see also McDermott Int'l, Inc. v. Wilander, 498 U.S. 337, 342 (1991) (interpreting statute as of the date of its passage); Perrin v. United States, 444 U.S. 37, 42 (1979) (same). In this case, the District Court did not rely on the industry understanding of "tar sand oil," in part because "[t]he testimony offered at trial suggests that, in fact, there was no industry consensus regarding a definition of tar sands in 1980 when Section 29 was enacted." Shell, 996 F. Supp. at 369.9 Shell contests the court's factual finding and argues that, even in the absence of an industry consensus, the court should have limited itself to selecting among definitions that were acceptable to some petroleum experts in 1980.

Contending the District Court misunderstood its theory of the case, Shell contests the court's finding there was no consensus definition of "tar sand oil" in 1980. The District Court described Shell's definition of tar sand oil as

[a]ny consolidated or unconsolidated rock (other than coal, oil shale, or gilsonite) that either: (1) contains a hydrocarbonaceous material with a gas-free viscosity, at original reservoir temperature, greater than 10,000 centipoise,10 or (2) contains a hydrocarbonaceous material and is produced by mining or quarrying.

Shell, 996 F. Supp. at 369. At trial, Shell referred to this definition as the "viscosity standard" and on appeal calls it the "quantitative standard." Shell concedes this standard was not generally accepted until 1983 but asserts that, in 1980, petroleum engineers agreed that "tar sand oil" was oil too viscous to be produced using primary production methods (the "primary production definition"). Furthermore, Shell maintains that petroleum experts in 1980 were able to identify tar sand and tar sand oil by inspection. Shell calls this experience-based identification of tar sand oil the "qualitative standard." According to Shell, the viscosity standard was a conservative quantification of the earlier primary production and qualitative definitions.11 Because the trial court made no findings with respect to the other proposed definitions or their relationships to the viscosity standard,12 Shell seeks a remand to determine whether the qualitative or the primary production definition was widely accepted by the industry in 1980 and whether the Potter Sands oil meets those definitions.

But Shell did not properly assert these additional industry definitions at trial. In the Pretrial Order, entered three days before trial, Shell listed only two "Issues of Law Which Remain to Be Litigated":

"(1) Whether the consensus definition of tar sand oil as having a viscosity greater than 10,000 centipoise, measured gas-free, at original reservoir temperature, identifies "oil produced from . . . tar sand" under 26 U.S.C. § 29 in 1983 and 1984?"

"(2) Whether FEA ruling 1976-4 . . . has any utility in identifying "oil produced from . . . tar sand" under 26 U.S.C. § 29?"

Shell Petroleum, Inc. v. United States, No. 93-508, slip op. at 15 (D. Del. Aug. 24, 1995) (Pretrial Order) (first and third ellipsis in original). The pretrial order controls the subsequent course of the trial unless modified to prevent manifest in Justice. See Fed. R. Civ. P. 16(e); Hassan v. Stafford, 472 F.2d 88, 95 (3d Cir. 1973). This order was never modified, nor did Shell request a modification. Indeed, Shell's opening statement confirmed to the court its theory of the case:

"[W]e will provide testimony that will show that the 10,000 centipoise definition of tar sand oil was the subject of considerable Discussion, considerable thought among industry experts and knowledgeable Government officials prior to 1980 and in the--throughout the year, the early 1980's."

"And the point is here that, although tar sand experts knew what tar sands were, there was no precise and generally acceptable demarcation between tar sand oil and other heavy oils."

The statement does not advance the theories Shell now proposes.

Shell has directed our attention to passages in the pretrial order, to passages in briefs submitted to the District Court, and to certain testimony elicited without objection at trial13 in which it claims it raised its other definitions. Generally, pretrial orders are liberally construed to permit the parties to advance their cases, see United States Gypsum Co. v. Schiavo Bros., 668 F.2d 172, 181 n.12 (3d Cir. 1981), and an issue can be tried by consent when testimony relevant to the issue is introduced without objection, see Douglas v. Owens, 50 F.3d 1226, 1236 (3d Cir. 1995). But a party still must unequivocally put its position before the trial court at a point and in a manner that permits the court to consider its merits. See Keenan v. City of Philadelphia, 983 F.2d 459, 471 (3d Cir. 1992) ("[T]he crucial question regarding waiver is whether defendants presented the argument with sufficient specificity to alert the district court"); Portis v. First Nat'l Bank, 34 F.3d 325, 331 (5th Cir. 1994) ("The raising party must present the issue so that it places the opposing party and the court on notice that a new issue is being raised."). We believe that Shell did not raise its primary production or qualitative definitions in an unequivocal manner.14

We have frequently noted "the well-established rule that absent compelling circumstances an appellate court will not consider issues that are raised for the first time on appeal." Patterson v. Cuyler, 729 F.2d 925, 929 (3d Cir. 1984).15 "This general rule applies with added force where the timely raising of the issue would have permitted the parties to develop a factual record," because we cannot know on appeal what evidence the adverse party would have presented or brought out through cross-examination. Harris v. City of Philadelphia, 35 F.3d 840, 845 (3d Cir. 1994) (internal quotation marks omitted). Finding no "compelling circumstances" requiring consideration of Shell's theories, we decline to exercise our discretion to do so. See American Bearing Co. v. Litton Indus., 729 F.2d 943, 949-50 (3d Cir. 1984).

Furthermore, our review of those portions of the record Shell has brought to our attention leads us to conclude that the possibility the District Court would find the facts as Shell describes them is remote. The strongest evidence Shell has cited comes from a report by its expert, Dr. Khalid Aziz, who wrote,

"Typically described [prior to 1980] as `extremely viscous,' the tar sand oil was further depicted as oil that could not be produced commercially on primary production. The literature most commonly describes `tar sand' as a naturally occurring rock formation containing a hydrocarbon so viscous that it cannot be economically produced through wells using only primary recovery methods."

But Dr. Aziz testified at trial,

"[P]rior to 1980, I did not see that there was a consensus in the industry as far as a precise definition of tar sand was concerned . . . ."

"But by 1980 a consensus did start to build--did not reach a consensus point, but did start to build around the use of 10,000 centipoise. . . ."

"In 1983, the definition of tar sands oil was well-established as the viscosity base definition that we have been talking about, viscosity of greater than 10,000 centipoise at reservoir conditions on a gas-free basis. . . ."

". . . ."

"[Before 1980, e]veryone knew that oils from tar sands were highly viscous. And I think people simply looked at the extreme and--looked at them as highly viscous, but they didn't know where to draw the line between tar sand oils and heavy oils. . . ."

Dr. Aziz also confirmed at trial the accuracy of a statement he had written in his report: "In 1980 the phrase `oil produced from tar sands' was commonly understood to mean all crude oil more viscous than heavy oil. However, there was no universally-accepted precise distinction between heavy oil and tar sand oil."

Similarly, Shell has called our attention to this testimony of a former DOE petroleum engineer:

"The general description that [DOE] used of tar sand [in 1979/1980] . . . was any rock containing a very viscous oil. And you could expand on that. In many cases we said it was too viscous to flow through the reservoir or to the well bore, or too viscous to be produced by primary recovery methods, for instance."

The witness was also asked how he identified tar sand when he first began working with it in 1970 and replied, "I looked at it and I knew it was tar sand." But on cross-examination, the witness agreed that "in the industry, as of April of 1980, there was no consensus definition of [sic] distinguishing heavy oil from tar sand oil." Because Shell's own experts testified there was no consensus definition of tar sand oil in 1980, it is unlikely Shell could prevail on remand. Where the possibility of a different outcome is remote, remand for consideration of factual issues not raised at trial is inappropriate. See Hormel v. Helvering, 312 U.S. 552, 558 (1941).

Assuming no definition was universally accepted in the petroleum industry in 1980, Shell insists the District Court should have limited itself to selecting one of the definitions used by petroleum engineers at that time. In that regard, Shell argues the congressional record should have been consulted only for that limited purpose.

Reliance on expert definitions of terms of art is a sound "general rule of construction," Massachusetts v. Blackstone Valley Elec. Co., 67 F.3d 981, 986 (1st Cir. 1995), but Shell has cited no authority supporting the proposition that, where no definition of a term of art is broadly accepted in the relevant technical field, a court is permitted only to choose among rival expert definitions. Where technical experts differ in their use of a term, the presumption that Congress has adopted an industry definition grows weaker and the courts will rely more heavily on other tools to ascertain Congress's meaning. Thus, in Atlantic Mutual Insurance Co. v. Commissioner, 523 U.S. 382, 118 S. Ct. 1413, 140 L. Ed. 2d 542 (1998), aff'g 111 F.3d 1056 (3d Cir.1997), the Supreme Court considered a property and casualty insurer's challenge to the IRS's administrative interpretation of "reserve strengthening" as used in the Tax Reform Act of 1986 1023(e)(3)(B), reprinted in note following 26 U.S.C.A. § 846 (West 1988). After concluding, based on a review of the expert testimony at trial, that the term had no established meaning in the field of property and casualty insurance, the Court accorded Chevron deference to the agency regulation without further reference to insurance industry usage. See Atlantic Mutual, 118 S.Ct at 1417 18. Here, there is no agency interpretation to defer to, but we rely on traditional tools of statutory interpretation in adopting the DOE definition.16

We believe the definition of "tar sand oil" in DOE Ruling 1976-4 is the one most compatible with congressional intent. As noted, Title I of COWPTA defines oil produced using enhanced extraction techniques as crude oil. Shell's definition, on the other hand, would categorize oil produced using enhanced extraction techniques as tar sand oil. Congress clearly distinguished tar sand oil from crude oil and considered tar sand oil a crude oil substitute. Because oil produced using enhanced extraction techniques, such as Shell's Potter Sands oil, is not a crude oil substitute, it could not qualify for the nonconventional source tax credit. Furthermore, Shell's definition would categorize crude oil produced with technologies widely available when COWPTA was enacted as tar sand oil. But Congress enacted the nonconventional source credit to encourage the development of new technologies that would produce crude oil alternatives.

As noted, Shell's Potter Sands oil was produced through steam injection, an enhanced (or tertiary) recovery method. As the District Court's reading of the statute demonstrated, oil produced using tertiary recovery methods (which the statute calls "tertiary oil"), is crude oil subject to the windfall profit tax imposed in Title I. See Shell, 996 F. Supp. at 370.17 Shell does not dispute this Conclusion. But Shell contends that tertiary oil can be classified as both crude oil and tar sand oil, arguing that its Potter Sands oil, extracted through enhanced (tertiary) recovery methods, is also tar sand oil eligible for the Title II tax credit. Relying on the legislative history, the District Court found that Congress did not consider crude oil to be tar sand oil, nor would Congress have imposed an excise tax and granted a tax credit to the same type of oil in the same statute. Because Shell's Potter Sands oil was indisputably crude oil subject to the windfall profit tax, we hold, like the District Court, it cannot also be tar sand oil eligible for the tax credit.

On this point, the legislative history is instructive. In its report on Title I (imposing the windfall profit tax), the Senate Finance Committee specified, "The term `crude oil' . . . applies only to natural crude petroleum and does not include synthetic petroleum, such as oil from shale or tar sands." S. Rep. No. 96-394, at 56, reprinted in 1980 U.S.C.C.A.N. at 465. The Conference Committee similarly explained, "The term `crude oil' . . . does not apply to synthetic petroleum such as oil production from shale or tar sands." H.R. Conf. Rep. No. 96-817, at 114 (1980), reprinted in 1980 U.S.C.C.A.N. 642, 667. Based on this legislative history, the District Court concluded, "Congress explicitly distinguished crude oil from tar sands, and understood tar sands to be a crude oil substitute." Shell, 996 F. Supp. at 370; accord Texaco, 101 T.C. 579. Because Congress specified that oil extracted using enhanced production techniques is crude oil, oil extracted using enhanced production techniques cannot be tar sand oil, a crude oil substitute.

Shell contends this legislative history should be interpreted in light of expert testimony that tar sand oil in its natural state is not synthetic oil but rather crude oil that can be processed into synthetic oil. But the clear import of the Committees' Reports is that all oil produced from tar sands is synthetic petroleum, not crude oil. Furthermore, because only crude oil is subject to the windfall profit tax, see 26 U.S.C. §§ 4986(a), 4991(a), the intent behind the Committee Reports is to exempt tar sand oil from the tax by specifying that it is not crude oil. But according to Shell, the processing of tar sand oil into synthetic oil is generally unnecessary in the continental United States.18 Shell has proposed no reason why Congress would exempt from the windfall profit tax the production of tar sand oil that is later processed into synthetic oil if the processing serves no useful purpose in important oil-producing parts of the country. Finally, it is worth noting that the Federal Energy Agency, in Ruling 1976-4, spoke of tar sand oil as a "so-called synthetic fuel[ ] (or crude oil substitute)." 10 C.F.R. ch. II Rulings at 371. This language supports the view that, when COWPTA was enacted, tar sand oil was not generally considered to be crude oil, subject to the windfall profit tax.

The District Court concluded that Congress did not regard oil capable of extraction through enhanced production methods as tar sand oil when COWPTA was enacted. We agree. The statute establishes that tertiary oil (oil extracted with enhanced production methods) is a type of crude oil, but Congress, like the FEA, did not consider tar sand oil to be crude oil. Therefore, tertiary oil cannot be tar sand oil. As the District Court explained, under Shell's interpretation, tertiary oil (oil produced by enhanced recovery methods)

"would be taxed as crude oil and, therefore, excluded from the category of tar sands. That same oil, however, would be defined as tar sands under Section 29 according to Shell's definition. This would result in two different definitions of tar sands under Title I and Title II. Clearly, this was not the intent of Congress."

Shell, 996 F. Supp. at 370-71 (footnote omitted).19 Because Shell does not contest the District Court's finding that its Potter Sands oil is tertiary oil, Shell's oil cannot be tar sand oil eligible for the tax credit.

The logic and structure of the act also demonstrate that crude oil is not tar sand oil. If oil producible using enhanced recovery methods, including the oil produced at Potter Sands, were classified both as crude oil and as tar sand oil, it would be subject to the windfall profit tax and also eligible for the nonconventional source tax credit. The District Court declined to interpret the statute this way "[a]bsent some evidence that Congress intended such an anomalous result." Shell, 996 F. Supp. at 371. We agree. See United States v. McKie, 112 F.3d 626, 631 (3d Cir. 1997) (discussing courts' "obligation to construe statutes sensibly and avoid constructions which yield absurd or unjust results").

Shell defends its interpretation by pointing out that under COWPTA the windfall profit tax was phased out as the price of oil fell whereas the nonconventional fuels credit was phased in under the same conditions. Shell suggests that the tax collected in years when oil prices were high was to be used to finance the subsidy when oil prices fell. Essentially, Shell suggests that the tax and credit together functioned as a rough-and-ready producer price stabilization scheme. But the legislative record is devoid of any suggestion that Congress intended such a result. Instead, the Senate Finance Committee explained that the Act "uses a large part of the revenue from the windfall profit tax to finance tax incentives for a wide range of alternate sources of energy." S. Rep. No. 96-394, at 8, reprinted in 1980 U.S.C.C.A.N. at 419. It appears Congress wanted to transfer windfall crude oil profits to producers of a distinct, "alternate," class of energy sources. Furthermore, Shell concedes that, on its reading of the statute, oil produced from Potter Sands would have received both a tax and a credit in 1981 and 1982. If Congress had sought to stabilize producer prices for tar sand oil, we do not see why Congress would have imposed a tax and provided a subsidy in the same year.20 Therefore, we believe the District Court correctly found that Congress did not intend to classify the same oil as both taxable crude oil and tar sand oil eligible for the tax credit. Since oil produced using enhanced recovery methods, such as Shell's, was subject to the windfall profit tax, it cannot have been eligible for the nonconventional source tax credit.

In summary, Shell's proposed definition, unlike DOE's, would classify oil producible through enhanced recovery techniques as tar sand oil. Because this classification would be contrary to the statutory scheme and to congressional intent, the District Court properly adopted the definition from DOE Ruling 1976-4.

Congress's explanation of the purpose of COWPTA provides further reason to reject Shell's proposed definition. As noted, Shell obtained crude oil from Potter Sands using technologies already well established in 1980. But the alternative fuel production credit was added by the Senate Finance Committee "to encourage energy conservation and production of alternate energy sources." S. Rep. No. 96-394, at 6, reprinted in 1980 U.S.C.C.A.N. at 417. "These alternative energy sources," the Committee explained, "typically involve new technologies, and some subsidy is needed to encourage these industries to develop to the stage where they can be competitive with conventional fuels. The information gained from the initial efforts at producing these energy sources will be of benefit to the entire economy." Id. at 87, reprinted in 1980 U.S.C.C.A.N. at 496. Because continued use of a widely available means of crude oil production is not an "initial effort[ ]" that will stimulate the development of new energy technologies, we agree with the District Court that oil producible with tertiary extraction methods available in 1980, such as steam injection, does not qualify for the nonconventional source credit. See Shell, 996 F. Supp. at 367 ("[D]omestic crude oil would not logically be categorized as an alternative energy source. . . . Rather, [Congress] endeavored to credit the production of alternative fuels or so-called crude oil substitutes."); accord Texaco, 101 T.C. 577 ("[S]section [29] was not intended to subsidize the production of crude oil but, to the contrary, was intended to encourage the development of crude oil substitutes.").

Shell maintains the District Court imposed a "new technologies" requirement not found in the statute on those seeking the tax credit. We disagree. The energy sources eligible for subsidy are those specified in 26 U.S.C.A. § 29(c)(1). Many of those sources were defined by Congress. See id. § 29(c)(2)-(3); H.R. Conf. Rep. No. 96-817, at 138-41, reprinted in 1980 U.S.C.C.A.N. at 689-93; S. Rep. No. 96-394 at 87-88, reprinted in 1980 U.S.C.C.A.N. at 496-97. So long as the energy produced is of one of the specified types, the taxpayer is eligible for the credit, whether or not the taxpayer used new technology. But where the definition of an energy source is unclear, Congress's directive that the eligible energy sources to be subsidized "typically involve new technologies" assists us in interpreting the provision. Congress's intent is best implemented through the DOE definition, which excludes energy sources produced with technologies already widely used or exploited when the Act was passed.

In this case, as the District Court noted, "[t]he definition of `tar sands' contained in the [DOE] Ruling is completely congruous with Congress' intent to both encourage the development of new technologies and limit the Section 29 credit to crude oil substitutes that could not be obtained using conventional methods." Shell, 996 F. Supp. at 367. Shell points out that the DOE Ruling was issued in response to "inquiries with respect to the applicability of the [EPAA price and supply controls] to the so-called synthetic fuels (or crude oil substitutes) processed from . . . tar sands . . . and other natural deposits that must be mined before the crude oil substitute can be extracted." DOE Ruling 1976-4, 10 C.F.R. ch. II Rulings at 371. Shell interprets this language to mean that FEA had received inquiries involving only tar sand oil that was both extracted from mined tar sand and processed into synthetic crude oil. Shell argues that FEA adopted an artificially narrow definition of "tar sands" in order to ensure that its Ruling would exclude only such oil from EPAA regulation. We are not convinced. It seems unlikely that FEA would have knowingly propounded an inaccurate definition, even for a limited purpose, rather than simply specifying that its Ruling applied only to tar sand oil that is extracted from mined tar sand and processed into synthetic crude oil. We note that the Ruling's definition does not limit tar sands oil to oil extracted from tar sands after mining. Therefore we agree with the District Court in adopting the DOE Ruling's definition of tar sands for purposes of the nonconventional fuels tax credit.21

For the foregoing reasons, we will affirm the judgment of the District Court.