Original proceeding in this court
¶1 In 2004, the Utah Division of Air Quality granted Sevier Power Company (the Power Company) an approval order to construct a coal-fired, circulating fluidized bed power plant. Sierra Club appeals the decision of the Utah Air Quality Board (the Board) denying its Request for Agency Action, which challenged the approval order for failing to comply with both federal statutes and the Utah State Implementation Plan's requirements for the prevention of significant deterioration (PSD) of air quality. We reverse in part, and affirm in part the Board's decision.
¶2 The Clean Air Act, codified at 42 U.S.C. §§ 7401 to 7515, aims to "protect and enhance the quality of the Nation's air resources" by prescribing national ambient air quality standards, which state and regional authorities are required to either maintain or progress toward. 42 U.S.C. § 7401 (2006). Each state enforces both the federal air quality requirements, and their own air quality requirements, via a state implementation plan.1
¶3 Critical to the maintenance of the national air quality standards is the prevention of significant deterioration (PSD) program. Areas that meet the national standards are considered to be in attainment. This status is maintained by, among other things, requiring the owner or operator of a "new major source" to apply for a new source permit. An application for a new source permit must show that the owner will construct the facility in a manner that applies the best available pollution control technology for regulated pollutants and complies with the PSD program's limitations on pollution increases. In Utah, the new source review permit, also called the PSD permit, is part of the approval order process. See Utah Admin. Code r. 307-401 (2004).
¶4 All PSD programs, whether federal or state, require a new source to undergo a best available control technology (BACT) review. This review is often conducted using the five-step "top-down method," which in essence requires the applicant to adopt the most stringent control technology, unless it can show that the technology is not achievable due to energy, environmental, or fiscal impacts.2 EPA, New Source Review Workshop Manual:
Prevention of Significant Deterioration and Nonattainment Area Permitting B.2 (1990), http://www.epa.gov/region07/programs/artd/air/hsr/hsrmemos/ 1990wman.pdf. Once the BACT is selected for a new facility, an emission limitation based on that control technology is also imposed as part of BACT.
¶5 In addition to the BACT analysis, a new source seeking a PSD permit must also complete an air quality analysis. The purpose of this analysis is to ensure that the emissions from the proposed facility, in conjunction with the emissions from other sources, will not cause or contribute to a violation of the national ambient air quality standards or the PSD increment limits. The national air standards place a ceiling on the total concentration of certain pollutants in the atmosphere. The PSD increments, on the other hand, place a limitation on the amount a source may increase the concentration of a pollutant over the baseline. The amount of increase allowed varies based on the location of the proposed source. Locations are divided into three categories. Class I allows the smallest increase in pollution levels and typically covers state and national parks and other wilderness and recreation areas. Class II allows a moderate increase and applies to areas considered normal growth areas. Class III areas can have the largest increment and are areas where the state or local authority foresees a greater amount of industrial development.
¶6 On October 12, 2004, Richard W. Sprott, executive secretary (the Secretary) for the Utah Air Quality Board and director of the Division of Air Quality (the Division)3 issued an approval order to the Power Company to build a coal-fired, circulating fluidized bed power plant. The approval order indicated that the Division found the proposal for the new electric power facility to comply with the Utah State Implementation Plan and authorized the Power Company to begin construction so long as the conditions of the approval order were met. The conditions itemized in the approval order included a requirement that the Power Company, in writing, notify the Secretary of the status of the facility's construction or installation if it had not been completed within eighteen months of the date of the approval order, October 12, 2004. "At that time, the Executive Secretary shall require documentation of the continuous construction and/or installation of the operation and may revoke the AO [approval order] in accordance with R307-401-11." Approval Order, DAQE-AN2529001-04, at *5, ¶ 9 (Oct. 12, 2004).
¶7 The Sierra Club challenged the approval order by submitting a Request for Agency Action. In its Request for Agency Action, the Sierra Club argued that the Board's approval order was invalid because it failed to comply with the Clean Air Act, the Utah Air Conservation Act, and various provisions of the Utah Administrative Code enforcing the federal and state Acts. Among the numerous grounds for challenging the approval order, relevant to this appeal the Sierra Club argued (1) that the Division failed to evaluate the emission of carbon dioxide and other greenhouse gases in its BACT analysis, (2) that the Division improperly excluded integrated gasification combined cycle technology as an available control technology, (3) that the emission limits set as part of the BACT review were in error, (4) that the Division wrongfully adopted a significant impact levels policy for determining the cumulative impact of a new source on Class I areas, and (5) that the Division failed to adequately complete a Class I increment analysis for sulfur dioxide.
¶8 The Board challenged the Sierra Club's standing, but lost on appeal to this court. See Utah Chapter of the Sierra Club v. Utah Air Quality Bd., 2006 UT 73, ¶ 11, 148 P.3d 975. PacifiCorp was then allowed to intervene on two issues--whether a BACT evaluation was required for carbon dioxide and other greenhouse gases and whether integrated gasification combined cycle technology was properly excluded from the BACT evaluation.
¶9 The Sierra Club then amended its complaint to include a claim that, in issuing the approval order, the Division failed to follow federal procedures for the invalidation or extension of a PSD permit.4 The Sierra Club moved for summary judgment on this issue. The Board denied this motion and instructed the Secretary to formalize in writing the Division's decision regarding the Power Company's request for an extension.
¶10 The Power Company, PacifiCorp, and the Division also filed motions for judgment on the pleadings asking the Board to dismiss the Sierra Club's claims that the Division did not properly administer the BACT review as a matter of law. The Board granted the motion for judgment on the pleadings as to the evaluation of carbon dioxide emissions, noting that neither the EPA nor the Utah Air Quality Board had promulgated rules for limiting or otherwise controlling the emission of greenhouse gases. The Board, however, denied summary judgment on the consideration of integrated gas combined cycle technology as part of the BACT evaluation because the Board determined that the issue presented contested factual issues that would require an evidentiary hearing.
¶11 The Board then held three days of hearings to address the integrated gas combined cycle technology issue, as well as the Sierra Club's remaining challenges. In a written order, the Board denied all of the Sierra Club's challenges and approved the granting of the approval order. Specifically, the Board held as follows:
(1) A BACT review is only required for pollutants that are regulated, not pollutants that could be regulated; carbon dioxide and other greenhouse gases are not currently regulated.
(2) The Division did not err by excluding integrated gas combustion cycle technology from the BACT analysis because adopting the process would redefine the design of the power plant. Additionally, the Board found that the technology was still in development and, therefore, was not available.
(3) The Division did not err in accepting selective noncatalytic reduction and an emission limitation of 0.1 lb/MMBtu on a twenty-four-hour basis as the BACT for sulfur dioxide because other control technologies were not feasible and a lower emission limitation had not been demonstrated for a similar plant.
(4) The Division did not err in adopting the significant impact levels as a screening method for whether the Class I increment would be violated. No rulemaking was required because this is a technical tool for making a PSD determination.
(5) The Division did not err in allowing the exclusion of IPP Unit 3 and Hunter 1 from the Power Company's cumulative analysis for the Class I increment. Additionally, the use of long-term averages for modeling did not violate the state's rules for PSD.
(6) The Division did not err by not requiring a new BACT analysis when construction on the proposed facility did not begin within eighteen months of the approval order. First, the approval order was still valid because the automatic-revocation language of the federal rules had not been adopted by the state. Second, the Division properly interpreted and enforced the state implementation plan and the approval order's requirements that a review take place and that revocation was based on the discretion of the Secretary.
¶12 The Sierra Club appeals this order.5 We have jurisdiction pursuant to Utah Code sections 63G-4-403 (Supp. 2008) and 78A-3-102(3)(b) (2008).
STANDARD OF REVIEW
¶13 The Utah Administrative Procedures Act governs our review of final determinations made by state administrative agencies, such as the Utah Air Quality Board. Utah Code Ann. § 63G-4-102(1)(b) (Supp. 2008). The Administrative Procedures Act provides relief to a party challenging a formal adjudicative procedure if the agency "erroneously interpreted or applied the law," or based an action "upon a determination of fact... that is not supported by substantial evidence when viewed in light of the whole record before the court," or is "otherwise arbitrary or capricious." Id. § 63G-4-403(4)(d), (g), (h). When reviewing an agency's interpretation of law, we review for correctness, "granting little or no deference to the agency's determination." Utah Chapter of the Sierra Club v. Utah Air Quality Bd., 2006 UT 74, ¶ 9, 148 P.3d 960. We review administrative rules in the same manner as statutes, focusing first on the plain language of the rule. Burns v. Boyden, 2006 UT 14, ¶ 19, 133 P.3d 370. "In our inquiry, we seek to give effect to the intent of the body that promulgated the rule." Id. When such intent is not clear from the plain language, we may rely on the administrative history of the rule to guide our interpretation. See R&R Indus. Park, L.L.C. v. Utah Prop. & Cas. Ins. Guar. Ass'n, 2008 UT 80, ¶ 25, 199 P.3d 917 ("When a statute is ambiguous, we use extrinsic interpretive tools such as policy and legislative intent to guide our analysis.").
¶14 In contrast, "questions of 'ultimate fact, mixed findings of fact and law, and [the agency's] interpretation of the operative provisions of statutory law it is empowered to administer'" are reviewed under an intermediate standard that considers whether the agency's determination was rational. Sierra Club, 2006 UT 74, ¶ 9 (alteration in original) (quoting Associated Gen. Contractors v. Bd. of Oil, Gas & Mining, 2001 UT 112, ¶ 18, 38 P.3d 291). We set aside these determinations "'only if they are imposed arbitrarily and capriciously or are beyond the tolerable limits of reason.'" Id. (quoting Associated Gen. Contractors v. Bd of Oil, Gas & Mining, 2001 UT 112, ¶ 18).
¶15 The Sierra Club challenges the Board's holdings as to the approval order's compliance with both state and federal PSD requirements. In addition, the Sierra Club argues that together the Board's errors constitute cumulative error that requires reversal. We address each of the Sierra Club's arguments, beginning with whether the approval order was invalidated by the expiration of an eighteen month deadline for construction.
I. AN APPROVAL ORDER MUST BE REVIEWED IN A MANNER THAT FURTHERS THE GOALS OF THE PSD PROGRAM
¶16 The Clean Air Act requires states with regions in attainment of the national ambient air quality standards to guard their attainment standards by implementing the PSD provisions found in sections 160 to 165 of the Clean Air Act. See 42 U.S.C. §§ 7470-79 (2006). To ensure proper enforcement of these provisions, the EPA promulgated a federal PSD program, which is found in rule 52.21. See Approval and Promulgation of Implementation Plans; PSD, 68 Fed. Reg. 74483, 74484 (Dec. 24, 2003). The federal program serves as a default program for states that have not had their own PSD program approved or states that choose to rely instead on the federal rules.6 As part of the federal PSD program, rule 52.21(r)(2) provides that [a]pproval to construct shall become invalid if construction is not commenced within 18 months after receipt of such approval, if construction is discontinued for a period of 18 months or more, or if construction is not completed within a reasonable time. The Administrator may extend the 18-month period upon a satisfactory showing that an extension is justified.
Approval and Promulgation of Implementation Plans, 40 C.F.R. § 52.21(r)(2) (2009). In contrast, Utah's corresponding rule, which is part of the approval order process, does not provide for automatic expiration of an approval order, but instead grants the Secretary discretion to revoke the approval order. Utah Admin. Code r. 307-401-11.7
¶17 The Sierra Club argues that federal rule 52.21(r)(2) applies to Utah's PSD permit process and supersedes the state administrative rule relating to an approval order's deadline for construction. According to the Sierra Club, the Environmental Protection Agency (EPA) incorporated the federal rule in Utah's PSD program when approving amendments to the state implementation plans. Alternatively, the Sierra Club argues that the federal rule should be applied because Utah's rule is less stringent than the federal rule. We hold that while the federal rule has not been incorporated and while the EPA accepted the state rule as part of the state implementation plan, the state rule must still be interpreted in a way that serves the purpose of the Clean Air Act and the PSD program.
A. Federal Rule 40 C.F.R. 52.21(r)(2) Has Not Been Incorporated into Utah's Air Quality Permitting Procedures
¶18 The Sierra Club argues that federal rule 52.21(r)(2) is part of the state implementation plan and therefore is binding on the Division and applicants for air quality permits. To support its position, the Sierra Club points to 40 C.F.R § 52.2346 to argue that the federal rule was incorporated into Utah's state implementation plan. Looking at plain language and, where necessary, the legislative history of the rule on which the Sierra Club relies, we hold that the federal rule was not incorporated.
¶19 Section 52.2346 of the Code of Federal Regulations approves the Utah State Implementation Plan (SIP) as meeting the requirements of the PSD program, but provides that "[t]he provisions of § 52.21... are hereby incorporated and made a part of the Utah State Implementation Plan and are applicable to proposed major stationary sources or major modifications to be located on Indian Reservations." On its face, the plain language of the rule seems to be in conflict with itself, and is therefore, ambiguous. For this reason, we turn to the rule's legislative history.
¶20 The administrative history, as reflected in the Federal Fegister, manifests an intent to accept the PSD provisions in Utah's plan and incorporate the federal provisions only in relation to the enforcement of the PSD provisions on tribal lands. 47 Fed. Reg. 6427. The Sierra Club makes numerous arguments relating to the administrative history of a 2003 amendment of rule 52.2346. However, as this amendment was only a change to ensure automatic updating of internal references, it is not the most relevant administrative history. See 68 Fed. Reg. 74483, 74484-85. Additionally, any arguments referencing the language adopted for other states provides little assistance, as we do not know the status of their PSD programs. We instead rely on the administrative history of the approval of Utah's PSD program. As explained by the EPA when first amending rule 52.2346, the agency found the Utah and the federal PSD provisions to be "practically identical" except for one major difference: "the State's regulation does not necessarily apply on Indian Reservations." Approval and Promulgation of State Implementation Plans; Utah, 47 Fed. Reg. 6427 (Feb. 12, 1982) (codified at 40 C.F.R. § 52.2346). The language of 52.2346 was amended with the intent to "approve the State regulation and remove the federal regulation except as it applies on Indian Reservations." Id. Thus, while the language of the rule may be conflicting, it is clear that by adopting the language, the EPA intended to incorporate rule 52.21 to apply only to Indian Reservations. Accordingly, we hold that federal rule 52.21 has not been incorporated into Utah's State Implementation Plan for land within the state's jurisdiction; therefore, Utah's provisions govern.
B. Utah's Eighteen-month Review Provision Must Be Interpreted and Applied in a Manner that Furthers the Purpose of the PSD Program
¶21 The Sierra Club argues that even if the federal rule has not been incorporated into Utah's SIP, the Division should have applied the federal rule because the state rule was less stringent than the federal. We hold that while the EPA's approval of Utah's PSD program is an implicit approval of the stringency of its provisions, the Division must interpret Utah's provisions in a way that furthers the purpose of the state rule and its enforcement of the Clean Air Act.
¶22 State PSD provisions, like all other air quality protections, must comply with and adequately enforce the Clean Air Act. The Act requires states to adopt emission control and maintenance plans that implement standards and limitations as stringent as those set by the EPA. See, e.g., 42 U.S.C. § 7416. Additionally, in line with setting local standards, the Clean Air Act requires states to adopt enforcement standards and procedures as part of their SIP in order to ensure that emitting facilities comply with the standards and requirements of both the Clean Air Act and state air quality statutes. 42 U.S.C. § 7410 (state plan must "include a program to provide for the enforcement of... a permit program as required in part C," i.e., a PSD program); 40 C.F.R. § 51 app. V. If this requirement and other requirements are met, the EPA approves a state implementation plan. The EPA accepted Utah's PSD program, including its enforcement provisions, as part of Utah's SIP in 1982. 47 Fed. Reg. 6427.8
In so doing, we must assume that the EPA found the enforcement provisions of Utah's PSD program to be adequate. Therefore, we cannot agree with the Sierra Club that on its face, Utah's eighteen-month review, which we interpret to be an enforcement provision, is not adequately stringent.
¶23 We must also assume that the EPA intended the provision to be interpreted and applied in a way that enforced the purpose and requirements of the PSD program. 42 U.S.C. § 7410(a)(2)(J) (SIP must meet the applicable requirements of "part C," prevention of significant deterioration program). The "basic purpose" of the PSD program is "simply put, to keep clean air clean." Roosevelt Campobello Int'l Park Comm'n v. EPA, 684 F.2d 1034, 1036 (1st Cir. 1982). Specifically the authors of the Clean Air Act intended the program to "protect public health and welfare from any... adverse effect[s]... from air pollution," "preserve, protect, and enhance the air quality" in ecologically sensitive areas, "insure that economic growth... occur[s] in a manner consistent with the preservation of existing clean air resources," and "assure that any decision to permit increased air pollution in any area... is made only after careful evaluation of all the consequences of such a decision."
42 U.S.C. § 7470. In accordance with these purposes, both the federal and Utah PSD programs have an enforcement provision that requires a review and possible revocation of a PSD permit if construction has not begun within eighteen months of the issuance of an approval order. The EPA's Environmental Appeals Board has explained that the purpose of the enforcement rule is "to ensure that major emitting facilities... are constructed in accordance with reasonably current pollution control standards" and "to ensure that the available PSD increments... allotted to a permittee are not tied up for indefinite periods of time...." N.Y. Power Auth., 1 E.A.D. 825 (1983). Because the EPA accepted Utah's PSD program with the understanding that it was "practically identical" to the federal program, 47 Fed. Reg. 6427, Utah's eighteen-month enforcement provision must be interpreted in a way that also achieves this outcome.
¶24 The Sierra Club and the appellants in a companion case, Kennon v. Air Quality Board, argue that if the Board was not required to revoke the approval order, which we hold it was not, then it should have required a new BACT analysis and a period for public comment. 2009 UT __ ¶ 24, __ P.3d __. We decline to determine the specific procedures required under Utah's enforcement statute9 but, as discussed more fully in Kennon, we hold first, that the review performed by the Board was inadequate to ensure that the Power Company's facility would utilize the most up-to-date pollution control technologies and second, that the Board's failure to set a new deadline or any requirements for establishing a construction timeline failed to prevent an unwarranted hold of PSD increments for an indefinite period. As ordered in Kennon, we remand to the Division to apply enforcement rule 307-401-11 in a way that ensures that the most up-to-date control technology has been adopted and that increment limits are not tied up indefinitely.
II. A BACT ANALYSIS WAS NOT REQUIRED FOR CARBON DIOXIDE
¶25 The Sierra Club argues that the Board erred by not requiring the Division to perform a BACT analysis for greenhouse gases such as carbon dioxide. The Board, the Sierra Club urges, should have relied on Utah's 2006 BACT regulation, which requires a BACT analysis for all "air contaminants," rather than the 2005 regulation, which requires a BACT analysis for all "pollutants subject to regulation." Acknowledging that it failed to preserve this argument, the Sierra Club argues that applying the 2005 BACT rule was plain error. Alternatively, the Sierra Club asserts that under the 2005 regulation the Division was still required to perform a BACT analysis for carbon dioxide because it is a "pollutant subject to regulation."
A. It Was Not Plain Error for the Board to Rely on the 2005 BACT Regulation
¶26 The Board did not commit plain error by applying the 2005 regulation after a 2006 amendment renumbered the regulation and changed the language "pollutant subject to regulation" to "air contaminant." We do not review an unpreserved issue unless exceptional circumstances are present or the error was plain. Plain error requires the showing of a harmful error that should have been obvious to the district court, or in this case, the Board. State v. Holgate, 2000 UT 74, ¶ 13, 10 P.3d 346. "But under the doctrine of invited error, we have declined to engage in even plain error review when 'counsel, either by statement or act, affirmatively represented to the [trial] court that he or she had no objection to the [proceedings].'" State v. Winfield, 2006 UT 4, ¶ 14, 128 P.3d 1171 (alterations in original) (quoting State v. Hamilton, 2003 UT 22, ¶ 54, 70 P.3d 111).
¶27 In this case, the Board applied the very administrative rule that the Sierra Club presented to it in both the 2004 Request for Agency Action and the 2007 Amended Request for Agency Action. "We have held repeatedly that on appeal, a party cannot take advantage of an error committed at trial when that party led the trial court into committing the error." State v. Dunn, 850 P.2d 1201, 1220 (Utah 1993). In this case, we see no reason to apply the plain error doctrine to a claimed error introduced by the appealing party.
¶28 The Sierra Club argues, however, that even though it presented its argument to the court under the 2005 regulation, it should have been obvious to the Board that the 2006 regulation applied because the Board itself adopted the new regulation. We disagree. The retroactive application of a new regulation is not obvious. Instead, a controversy is typically "determined on the basis of the [statutory or administrative] law as it existed at the time of the occurrence...." Foil v. Ballinger, 601 P.2d 144, 151 (Utah 1979) (quoting Okland Constr. Co. v. Industr. Comm'n, 520 P.2d 208, 211 (Utah 1974)). Retroactive application of an administrative rule is an exception to this approach that requires thorough analysis. So, while there is no indication that the Sierra Club intentionally withheld this argument from the Board, we decline to review an error that was not obvious and that the Sierra Club should have first presented the Board. We therefore decline to consider the merits of whether the 2006 regulation should have been applied retroactively and instead consider the Board's review of the Division's actions under the 2005 regulation.
B. The Board's Interpretation of the BACT Regulation to Exclude Carbon Dioxide Was Reasonable
¶29 The Sierra Club argues that a BACT analysis should have been completed because the Clean Air Act regulates carbon dioxide by requiring facilities to monitor and report carbon dioxide emissions.10 In contrast, Respondents and PacifiCorp argue that to be regulated, the emission of a pollutant must actually be restricted or limited in some way, not merely measured. We conclude that the Board did not err in narrowly interpreting the phrase "subject to regulation" when reasonable policy concerns support this interpretation.
¶30 Rule 307-101-2 required a new source to adopt the BACT for each "pollutant subject to regulation."11 Recently the United States Supreme Court determined that carbon dioxide and other greenhouse gases are pollutants under the Clean Air Act, Massachusetts v. EPA, 549 U.S. 497, 532 (2007) ("[G]reenhouse gases fit well within the Clean Air Act's capacious definition of 'air pollutant.'"); therefore, our review focuses on whether carbon dioxide is "subject to regulation under the Clean Air Act and/or the Utah Air Conservation Act." Utah Admin. Code r. 307-101-2 (2004).
¶31 We begin our analysis by determining the proper meaning of the term "subject to regulation." This issue has recently been addressed by two state courts and by the Environmental Appeals Board, with no consensus on the meaning of the phrase. It is clear, however, from the analysis completed by these courts that the language itself does not compel a particular interpretation. Therefore, the meaning of the phrase must be determined by reference to administrative intent and policy.
¶32 We conclude that the phrase "subject to regulation" is ambiguous because there does not appear to be an agreed upon ordinary meaning for the phrase. As the Environmental Appeals Board recently concluded, the "language is broad enough to embrace different meanings, or shades of meaning." Inre: Deseret Power Elec. Coop., __ E.A.D. __ (2008), 2008 EPA App. LEXIS 47, *62.
¶33 Both sides direct our attention to the parallel federal regulation, which uses a similar phrase in its BACT definition but also defines New Source Review Pollutants. We find the definition of little help, however, as it only directs us once again to our starting point--is carbon dioxide a pollutant "subject to regulation under the Act"? As defined in 40 C.F.R. § 52.21 (b)(50), regulated pollutants as used in the PSD permit rules include pollutants for which a national ambient air quality standard has been set, a new source performance standard has been set, those specifically listed and governed by the ozone protections of the Clean Air Act, and "[a]ny pollutant that otherwise is subject to regulation under the Act." To date, carbon dioxide is not governed by the national ambient air quality standards, the new source performance standards, or the ozone protection standards. Thus, it is only a regulated pollutant for purposes of federal PSD permitting if it is regulated somewhere else in the Clean Air Act. To date, the EPA has not settled on a definition of what it means to be "regulated" by the Clean Air Act.12
¶34 We focus next on policy concerns underlying the adoption of a broad or narrow definition of the "subject to regulation" language. Respondents and PacifiCorp argue that requiring BACT review for carbon dioxide undermines the rulemaking process for controlling pollution emissions. Further, the Board argues that it would be impractical to complete a BACT analysis for carbon dioxide without any "governing standards or rules." This argument was accepted recently in a Wyoming administrative adjudication, which upheld the Wyoming Department of Environmental Quality's interpretation of "subject to regulation" as not applying to carbon dioxide. The Wyoming Environmental Quality Council was greatly troubled by the fact that a broader interpretation would require the state agency to "regulate greenhouse gases... without the scientific and policy resources available to the EPA and the United States Congress." Basin Elec. Power Coop., Dry Fork Station, EQC Docket No. 07-2801 2008 Wyo. ENV LEXIS 3 at *13 (Wyo. Envtl. Quality Council Aug. 21, 2008). We agree and conclude that, given these concerns, the Board's interpretation of the BACT rule was reasonable.
¶35 The negative impact of carbon dioxide concentrations in the atmosphere is an issue of national and global concern. See Massachusetts, 549 U.S. 497, 521-24 (chronicling the development of climate change research). We do not disregard the import of this issue. We also do not disregard the broad wording and preventative purpose of the BACT regulation. Still, when reviewing the decisions of an administrative agency, we must afford proper deference to the expertise and discretion of the Board and decline to overturn its interpretation unless it was irrational or unreasonable. In this case we conclude that the Board's interpretation of the BACT regulation as a supplement to already existing pollutant control regulations rather than a stand-alone control regulation is reasonable in light of the need for a thorough research and rulemaking process to determine the appropriate standards for controlling or limiting carbon dioxide. The EPA (spurred by the United States Supreme Court) has recently undertaken this process. See Regulating Greenhouse Gas Emissions Under the Clean Air Act, 73 Fed. Reg. 44354 (July 30, 2008). We hold that the Board's determination to defer to this formal process rather than "preempt[ing] ongoing Congressional and EPA efforts to formulate a CO2 emissions policy" by instituting an emission limitation as part of BACT review was reasonable. Longleaf Energy Assocs., LLC v. Friends of the Chattahoochee, Inc., 681 S.E. 2d 203, 207 (Ga. Ct. App. 2009).
III. THE BACT ANALYSIS SHOULD HAVE INCLUDED INTEGRATED GASIFICATION COMBINE CYCLE TECHNOLOGY
¶36 Next, the Sierra Club argues that the Board erred in affirming the Division's decision to exclude Integrated Gasification Combine Cycle (IGCC) as a control technology in the BACT analysis. The Respondents and PacifiCorp argue that the PSD program did not require consideration of IGCC as an alternative control technology because it would have required the Power Company to redesign its facility and, alternatively, because IGCC is not an available technology. We treat each of these arguments in turn.
A. The Plain Language of the BACT Analysis Requires Consideration of IGCC
¶37 Utah Administrative Code rule 307-101-2 defines the BACT as an emission limitation and/or other controls to include design, equipment, work practice, operation standard or combination thereof,... which the Air Quality Board, on a case-by-case basis taking into account energy, environmental and economic impacts and other costs determines is achievable for such installation through application of production processes and available methods, systems and techniques, including fuel cleaning or treatment or innovative fuel combustion techniques for control of each such pollutant.
¶38 The plain language of this definition indicates that IGCC technology should be evaluated as part of the BACT review. First, the definition indicates that the application of production processes should be considered when determining what the best control technology will be for a proposed facility. Additionally, the definition includes innovative fuel combustion techniques. PacifiCorp argues that fuel combustion is distinct from IGCC, which is a fuel conversion technology. However, we do not interpret combustion as limited to only burning fuels; combustion can also involve a rapid change in a fuel, Webster's II New College Dictionary 223 (1995); a fuel combustion technique could also involve burning the gas created by the fuel. Regardless, IGCC is a production process and therefore falls under the plain language of the rule. Acknowledging that IGCC is a production process, PacifiCorp points to the Board's factual finding that "IGCC is a power generation technology, not an emission control technology." Findings of Fact, Conclusions of Law, and Final Order, DAQE-AN2529001-04, at *7, ¶ 10 (Jan. 9, 2008). This fact does not exclude IGCC from the BACT review, however. As a power generation technology, IGCC can still act as an emission control technology. The use of IGCC technology can reduce the emission of criteria pollutants such as nitrogen oxides, sulphur oxides, carbon monoxide, and particulate matter. U.S. Dep't of Energy, Major Environmental Aspects of Gasification-based Generation Technologies ES-2 (2002) http:/www.netl.doe.gov/technologies/coalpower/gasification/pubs/p df/final%20env.pdf. Therefore, we determine it is a control technology.
¶39 We also reject the Board's factual finding 11, which implies that control technologies are only those that can be added onto the proposed facility. This factual finding, however, also implies that control technologies are processes and systems that can be "designed into" a proposed installation. Along the same lines, the Respondents and PacifiCorp vigorously argue that IGCC should not have been part of the BACT review because its adoption could not be designed into the Power Company's proposed facility but would require a redesign.
¶40 The notion that a control technology cannot require the redesigning of a facility originates from EPA guidelines, which indicate that the BACT review should not be used "as a means to redefine the design of the source." EPA, New Source Review Workshop Manual: Prevention of Significant Deterioration and Nonattainment Area Permitting B.13 (1990). The EPA illustrated the rule by explaining that an applicant proposing a coal-fired electric generator is not required to consider, under BACT review, building a natural gas-fired electric turbine. We agree that changing a fuel source would drastically redesign a proposed facility and therefore production processes that involve a completely different fuel source need not be considered.
¶41 IGCC and other production processes are more difficult to analyze where they involve the same fuel source but different forms of production. Several courts have interpreted this guidance and concluded that the design of a facility is redefined when the adoption of a control technology changes the objective or purpose of the facility. See Sierra Club v. EPA, 499 F.3d 653, 657 (7th Cir. 2007) (upholding the Environmental Appeals Board's decision that requiring a proposed mine-mouth plant to consider a different fuel source would redefine the "'fundamental purpose or basic design of [the] proposed Facility.'"); In re: Old Dominion Elec. Coop., 3 E.A.D. 779, 793 n.38 (1992) ("EPA does not require a PSD applicant to change the fundamental scope of its project."). Deciphering which control strategies would result in the redefinition of a proposed facility requires a case-by-case analysis in which the difference between the strategies accepted and those rejected is often a matter of degree. For example, a proposed waste combustion facility was not required to consider as part of its BACT analysis disposing of the waste by burning in existing electricity-generating facilities. In re: Pensauken County Res. Recovery Facility, 2 E.A.D. 667 (1988). The adoption of this alternative process would not only redefine the proposed facility but would negate the need for the waste combustion facility altogether. In comparison, the use of clean fuels, a control strategy identified by the BACT definition, presents a closer question. The Seventh Circuit recently determined that a mine-mouth electricity plant was not required to consider the use of clean, Western coal as a control technology in its BACT analysis. Sierra Club, 499 F.3d 656-57. Exploring the scope of the redefinition test, the court acknowledged that a stand alone coal-fired electricity plant would be required to consider clean coal as part of BACT, even though "[s]ome adjustment in the design of the plant would be necessary." Id. at 656. In contrast, the court determined that because the proposed electricity plant was a mine-mouth power plant, "the use of [the] particular coal supply [was] an inherent aspect of the proposed project" and requiring clean coal would excise this aspect from the project. Id. (citing In re Prairie State Generating Co., 189808AAB, 2006 EPA App. LEXIS 38, at *39 (EAB Aug. 24, 2006)).
¶42 As illustrated by the Illinois mine-mouth cases, the purpose of a proposed facility is determined by the description in the application submitted for the proposed facility, so long as the "purpose or design is objectively discernable." Prairie State, 2006 EPA App. LEXIS 38, at *54. In the Seventh Circuit case, the applicants proposed to build a mine mouth power plant, and that description guided the permit issuers' demarcation of what technologies would require a redefinition of the design of the proposed facility. We emphasize that the purpose of the project must be objective and must focus on the overall business purpose for the proposed facility. We are wary of the risk of applicants describing a project in such a limited manner that they are able to circumvent the goals of BACT, which include encouraging the use of new technologies. See In re: Knauf Fiber Glass, GmbH, 8 E.A.D. 121, 127 (1999) ("[T]he purpose of BACT... is to promote use of the best control technologies as widely as possible. If a company can claim that the only facilities similar to a proposed project are its own facilities, this objective of the BACT program would not be fulfilled."). Thus, when considering what design elements are inherent to the project, cost and avoidance of the risks associated with adopting new technologies cannot support what is considered fundamental to a project's design. Prairie State, 2006 EPA App. LEXIS 38, at *58 n.23. Instead, the fundamental aspects must relate to the basic business purpose of the proposed facility.
¶43 Applying our holding to the facts of this case, we conclude that considering IGCC would not require the Power Company to redefine the design of its proposed facility. Like the plant in Prairie State/Sierra Club, the basic design of the Power Company's proposed facility is an electric power generating plant fueled by coal. With this purpose, it is evident that the Power Company was not required to consider wind generation for electric power as an alternative process. However, as in the Prairie State BACT analysis, Prairie State, 2006 EPA App. LEXIS 38, at * 68-71, the Power Company should have included IGCC in its BACT review. IGCC is a control technology that can reduce the emissions of several criteria pollutants. The adoption of this standard would not require the Power Company to redefine the design of its proposed facility. The facility would still remain an electric power generating plant fueled by coal. We note that the consideration of IGCC in the BACT review does not compel its adoption; instead, it only requires the Power Company to subject IGCC to the five-step top down analysis used to determine the best available technology.
B. There Was Insufficient Evidence to Support the Board's Determination That IGCC Is Not an Available Technology
¶44 Respondents and PacfiCorp argue, however, that even if IGCC would not redefine the purpose of the Power Company's proposed facility, its inclusion in the BACT review was not required because it was not an available technology. The Board agreed with this argument and included in its order a factual finding that IGCC was not an available control technology. We review this not as a factual finding but as an application of law to fact, which is a mixed question of law and fact. Therefore, we review the record to determine whether the Board's determination was rational.
¶45 The term "available" is undefined in Utah's rules. The EPA and its appeals board have defined available technologies as those that have a "practical potential for application... includ[ing] those employed outside of the United States." EPA, New Source Review Workshop Manual: Prevention of Significant Deterioration and Nonattainment Area Permitting B.5 (1990). For a technology to be practically applied, it must have been applied to or permitted for a full-scale operation. Id. at B.13. Thus, "the term available is used in its broadest sense under the first step and refers to control options with a 'practical potential for application to the emissions unit' under evaluation." Knauf Fiber Glass, GmbH, 8 E.A.D at 130 (emphasis in original)(quoting New Source Review Workshop Manual at B.5). This definition is not binding on the Division or the Board, but in our view, because it is included in a manual created to guide air-permitting authorities throughout the nation, it is very persuasive. Further, while the Board has discretion to interpret its own regulations, as we discussed in Part I, it must do so with an eye to furthering the goals of the PSD program. As indicated in its title, the purpose of the BACT review is to ensure that the best available control technology is adopted. Implicit in this purpose is a goal to encourage the adoption of new technologies.
¶46 In this case, the Board seems to have used the term "available" in a narrow sense. After receiving testimony that IGCC is currently being used in coal-based electric plants, two in the United States in addition to fourteen other plants operating worldwide, and that more IGCC plants are proposed in the United States, the Board determined that the technology was not available but still in the developmental stage. This determination appeared to be based on the testimony of engineer Stephen Jenks, who testified that the two coal-based IGCC plants in the United States had not reached their operational goals. This testimony was given in the context of his testimony that circulating fluidized bed was a better production process because it better reached its operational capability. While relevant to the BACT process, this testimony does not relate to determining which technologies are available; instead it relates to Step II, which compares the available control strategies based on their technological feasibility. Therefore, it was unreasonable to rely on this testimony to determine that IGCC was unavailable. To the contrary, if adequate evidence is presented that a control technology, including production processes, is operating or permitted for similar operations, the permitting authority should consider the technology available. In this case, such evidence was presented. Therefore, we conclude that the Board's determination that IGCC was unavailable was unreasonable. Because IGCC was available and the plain language of the BACT definition indicates it should be considered in the BACT analysis, we vacate the Board's BACT analysis conclusions.
IV. THE BACT EMISSION LIMITATION FOR NITROGEN OXIDES WAS NOT REASONABLE WHEN THERE WAS EVIDENCE THAT LOWER EMISSIONS WERE ACHIEVABLE
¶47 The Sierra Club argues that the Board erred by concluding that the Division's BACT determination for nitrogen oxides was sufficient when the factual findings supporting the Division's BACT emission limitation were insufficient and lower emission rates for nitrogen oxides were possible. Specifically, the Sierra Club argues that the Division failed to set an appropriate BACT emission limitation because it relied on the Power Company's suggested emission limitation and failed to adopt lower emission limitations achieved by other facilities using the control technology selective noncatalytic reduction. The Respondents argue that the lower rate achieved by the other facilities--a rolling average of.07 pounds per million Btu in a thirty-day period--is equivalent to the emission limitation--0.1 pounds per million Btu in a twenty-four hour period--imposed on the Power Company. Further, even if these are not equivalent, the Division and Power Company argue that the Division's selection of the 0.1 emission limitation was proper because it was the most stringent limitation for nitrogen oxides. We hold that the imposition of the nitrogen oxides emission rate was unreasonable because no evidence was presented to show that this emission limitation would achieve the same overall nitrogen oxides reduction as a lower thirty-day average emission limitation.
¶48 As defined, BACT is "an emission limitation and/or other controls... based on the maximum degree or reduction of each pollutant subject to regulation." Utah Admin. Code r. 307-101-2 (2004). Further, the EPA has described the goals of BACT emission limitations in three-parts: (1) to achieve the lowest percent reduction, (2) to protect short-term ambient standards, and (3) to be enforceable as a practical matter. See EPA, New Source Review Workshop Manual: Prevention of Significant Deterioration and Nonattainment Area Permitting B.6-.9 (1990). As presented to the Board, the emission limitation set by the Division achieves two of the three goals. First, evidence was presented to the Board that the manufacturer guaranteed a maximum 0.1 twenty-four-hour average emission. This was sufficient evidence that the twenty-four-hour average emission limitation was achievable. Additionally, the Division presented evidence that the lowest twenty-four-hour emission limitation was 0.1 pounds per million. Although there was evidence that lower rates were achievable, this was sufficient evidence for the Board to rely on in determining that 0.1 pounds per million was the greatest short-term emission reduction achievable. The Division, however, did not provide sufficient evidence that this same limitation achieved the maximum reduction of nitrogen oxides possible. The Division and Power Company argued at length that a twenty-four-hour average and thirty-day average were comparable. This argument, however, was supported with scant evidence. Respondents argued that the thirty-day averages were more variable, but were unable to show that the overall actual nitrogen oxide emissions from plants complying with the 0.1 twenty-four-hour limitation was the same as the emission from plants that complied with the lower thirty-day average. That is, the Division could not show that the Power Company's facility would operate at thirty percent or more below its twenty-four hour emission limit. Without such evidence, we hold that it was unreasonable for the agency to adopt the 0.1 twenty-four hour emission limitation when there was evidence that a lower overall emission limitation was achievable; therefore we set aside the Division's determination that this standard was BACT for nitrogen oxides.
V. THE DIVISION DID NOT NEED TO UNDERTAKE FORMAL RULEMAKING PROCEDURES TO ADOPT THE SIGNIFICANT IMPACT LEVEL POLICY
¶49 In the midst of reviewing the Power Company's application for a PSD permit, the Secretary adopted, via an interoffice memorandum, a significant impact levels policy for Class I increment analyses required by the PSD program. The significant impact level policy excludes from the regularly required cumulative Class I increment analysis emissions from a proposed source that model below a value identified by the Division as de minimis. The Sierra Club argues that this policy is invalid because it was not adopted pursuant to the Utah Rulemaking Act, Utah Code Ann. § 63G-3-101 to -702 (Supp. 2008). We hold that formal rulemaking was not required when the agency's action did not constitute a fundamental policy change but instead served as technical guidance for the implementation of an already existing administrative rule.
¶50 The Utah Administrative Rulemaking Act requires an agency to undertake formal rulemaking procedures when "issu[ing] a written interpretation of state or federal mandate" but not when "an agency issues policy or other statements that are advisory, informative, or descriptive." Utah Code Ann. § 63G-3-201(3), (4). When determining whether rulemaking is required under the statute, we focus our attention on whether an agency action amounts to a rule. Williams v. Pub. Serv. Comm'n, 720 P.2d 773, 776 (Utah 1986).13 Defined both statutorily and in case law, a rule is a policy or statement that is generally applicable, implements or interprets law, and results in a change in clear law. Id. at 776-77; see also Utah Code Ann. § 63G-3-102(16). Conversely, an agency action is not a rule when it provides informal guidelines for implementing agency rules or answers a technical question within the agency's expertise. Gray v. Dep't of Employment Sec., 681 P.2d 807, 815-16 (Utah 1984).
¶51 In this case, the policy adopted via interoffice memorandum is not easily categorized. There is little room to debate that the policy is generally applicable. As described in the memorandum, the significant impact levels analysis is a "matter of policy" that will apply to all new sources subject to the PSD program. Whether it is an interpretation or implementation of law is a much closer question. We conclude that it is a technical guideline for implementing existing interpretation and implementation of the Utah Conservation Act and the Clean Air Act and is not subject to formal rulemaking procedures. Critical to our conclusion is the fact that the significant impact levels policy does not by itself impose new requirements on applicants for PSD permits or the Division. Rule 307-405-6(2)14 requires the Secretary to determine the impact of the new source on the air quality in the area. Where practicable, this review should "take into account" the cumulative effect of all sources and growth in the affected area. Id. The significant impact levels policy does not interpret or alter this rule. Instead, it provides internal guidance for the Division for determining compliance with existing administrative rules. Although it appears that it will be applied in a standard fashion, it is not binding on the Division. Should the Division determine that a complete cumulative analysis is warranted, the significant impact levels analysis in no way prevents it from conducting such analysis.
¶52 Still, the Sierra Club argues that this new guidance is a clear change in the law because previously the Division conducted a complete cumulative analysis for all new sources affecting Class I areas. A change in how compliance is measured as opposed to what is required of a party is not a clear change in law requiring rulemaking but instead an internal governance issue on which we defer to the agency. The difference between the facts in Williams and this case illustrates the distinction. In Williams, the Public Service Commission exercised jurisdiction over mobile telephone paging services for an extended period of time and as part of that process required operators to obtain certificates of public service and convenience. 720 P.2d at 774. Then the Federal Communications Commission deregulated radio frequencies for use in paging services. Id. In correspondence replying to a service provider asking whether it was required to obtain a certificate, the Commission indicated that it no longer had jurisdiction over paging services using radio frequencies and therefore the certificates were no longer required. Id. We concluded that this complete reversal of the requirements imposed on service providers was a "fundamental policy change" that required formal rulemaking. Id. at 777. The action taken by the Division in this case is quite different. Rather than reversing a longstanding position, the memorandum issued by the Secretary instead changed a procedure for applying a longstanding rule.
The requirements imposed on parties and the Division remained the same--new sources could not exceed the increment limits and the Division was required to determine the impact of the new source on the air quality and whether the increments would be exceeded. Therefore, we conclude that the Board did not err in concluding that the adoption of the significant impact levels did not require formal rulemaking.
¶53 Finally, the Sierra Club argues that, even if rulemaking was not required, the significant impact levels policy is invalid because it fails to satisfy the requirement of rule 307-405-6 that the Division's review "shall take into account all allowable emissions of approved sources or modifications whether constructed or not." The significant impact levels policy as presented to this court does not violate this rule. The rule requires a review to determine whether a source will violate the allowable incremental increases or the national ambient air quality standards. Because we have determined that the significant impact levels policy was properly adopted we need not consider the Sierra Club's arguments relating to the cumulative Class I analysis.
VI. CUMULATIVE ERROR DOCTRINE DOES NOT APPLY TO THE ARGUMENTS RAISED BY THE SIERRA CLUB
¶54 Finally, the Sierra Club argues that the cumulative error doctrine should be applied to the Board's decision. Again, this was an unpreserved argument. Under the cumulative error doctrine, we reverse only when "'the cumulative effect of the several errors undermines our confidence... that a fair trial was had.'" State v. Kohl, 2000 UT 35, ¶ 25, 999 P.2d 7 (quoting State v. Dunn, 850 P.2d 1201, 1229 (Utah 1993). The Sierra Club presented numerous arguments as to why the errors it alleged the Board committed, only some of which we have determined to have been actual error, were prejudicial. However, none of these arguments go to the fairness of the proceedings below. Therefore, we decline to address the preservation issue when on the face of the Sierra Club's arguments, the cumulative error doctrine does not apply.
¶55 The Sierra Club asserted numerous errors that occurred in the Board's review of the approval order the Division issued to the Power Company. We affirm in part, and reverse in part. First, we agree with the Board that the federal enforcement provision found in 40 C.F.R. § 52.21 was not incorporated into the Utah PSD rules at the time the Power Company's PSD permit was issued. However, we hold that the Board's interpretation of Utah's rules was unreasonable when, despite the discretion given to the Secretary, the Division and Board interpreted Utah's PSD enforcement provision in a way that failed to achieve the purposes of the PSD program. Therefore, we reverse and remand to the Division to conduct a new review. This review must ensure that the approval order encompasses the most current control technology and must assign a reasonable deadline for the construction of the Power Company's facility that avoids tying up the PSD increments longer than necessary.
¶56 Second, we hold that the Board did not err in concluding that carbon dioxide was not subject to BACT when reasonable policy concerns supported their interpretation of the BACT analysis requirement. Third, we hold that the Board erred in determining that IGCC need not be considered as part of BACT. We conclude that IGCC is an available control technology that the plain language of the BACT definition indicates should be considered. Therefore, we set aside the BACT review and remand to the Division to conduct a BACT analysis that considers IGCC as an available control strategy. Additionally, we hold that the Board erred in affirming the emission limitation determined by the BACT analysis for nitrogen oxides when the Division failed to provide sufficient evidence that the emission limitation level set was equivalent to lower emission limitations determined by a different modeling period. We set aside this decision.
¶57 Fifth, we hold that the Board did not err in affirming the Division's adoption of the significant impact levels policy, which was an internal guidance policy that did not require formal rulemaking. For this reason, we decline to review the Sierra Club's arguments regarding the cumulative increment analysis performed for the Class I areas affected the Power Company's proposed facility.
¶58 Finally, we decline to address the Sierra Club's argument that sufficient errors were committed below to constitute cumulative error. The Sierra Club's argument on its face does not support this argument because the Sierra Club did not argue that anything in the proceedings below suggested that the Board's review of the Power Company's approval order was unfair.
¶59 Associate Chief Justice Durrant, Justice Wilkins, Justice Parrish, and Justice Nehring concur in Chief Justice Durham's opinion.
1. Utah's State Implementation Plan is incorporated into the Utah Air Conservation Act, Utah Code sections 19-2-101 to -127 and Utah Administrative Code rule 307.
2. Beginning with step one, the top-down method requires the applicant to identify all available control technology options for the proposed facility for each regulated pollutant. Under the BACT definition, these technologies include "production processes, or available methods, systems, and techniques, including fuel cleaning, clean fuels, or treatment or innovative fuel combustion techniques." Clean Air Act § 169, 42 U.S.C. § 7479(3); Utah Admin. Code r. 307-101-2. In effect, the reviewer must consider lower emitting processes and practices, add-on controls, or a combination of the two. Under step two, the reviewer eliminates technically infeasible options. This requires a documented demonstration that technical difficulties, such as physical, chemical or engineering principles, would prevent successful use of the control technology at the proposed facility. Step three of the top-down review involves a ranking of the control technologies by their effectiveness, which is determined based on efficiency, emission rate, and emission reductions. Finally, in step four, the reviewer analyzes the economic, environmental, and energy impacts, both beneficial and adverse, beginning with the first ranked technology. If an objective evaluation of the impacts proves the top technology to be inappropriate for the facility, that technology is eliminated and the next most stringent technology is similarly evaluated. This process continues until a technology cannot be eliminated based on environmental, energy, or economic impacts, and that technology is then proposed as the BACT for the pollutant and emission under review. EPA, New Source Review Workshop Manual: Prevention of Significant Deterioration and Nonattainment Area Permitting B.6-.9 (1990).
3. Because Mr. Sprott as director and executive secretary of the Board acts on behalf of the Division, we use the terms "the Division" and "the Secretary" interchangeably.
4. In a separate case issued today, Appellants James O. Kennon and Dick Cumiskey also challenge this issue and present additional arguments. Where their arguments overlap, we address them in this opinion.
5. The Board considered two additional issues, which the Sierra Club does not appeal.
6. In 2006, the Utah Air Quality Board did incorporate all of rule 52.21 in order to comply with the new source reform of 2002. This action, however, occurred after the approval order in this case was granted and even after the approval order would have expired under the federal rule.
7. Renumbered as r. 307-401-18 in 2006. See 23 Utah Bull. 14, 22 (Dec. 1, 2005).
8. Utah's SIP has been revised and subsequently approved numerous times by the EPA since 1982. The approval in 1982, however, marks the acceptance of Utah's PSD program.
9. We note that the Board has now approved this procedure, but did so pursuant to its 2006 incorporation of the federal PSD program, including the federal enforcement provision.
10. Respondents argue that the Sierra Club did not preserve this issue for appeal. We disagree. The Sierra Club adequately raised the issue by addressing it in its Second Request for Agency Action and the Board ruled on the issue in its Final Order.
11. Renumbered as rule 307-401-2 (2008). See 23 Utah Bull. 14, 22 (Dec. 1, 2005).
12. The EPA's Region 8 office recently adjudicated this question under the federal PSD program and concluded that carbon dioxide was not a regulated pollutant because the EPA's historic interpretation of the phrase excluded carbon dioxide and other uncontrolled pollutants. The EPA Appeals Board reversed this decision, concluding that the Region failed to prove the EPA had an established interpretation. Inre: Deseret Power Elec. Coop., __ E.A.D. __ (2008), 2008 EPA App. LEXIS 47. The EPA Administrator then issued a formal interpretation of regulated pollutants that excluded carbon dioxide. Memorandum from Stephen L. Johnson, Administrator, EPA, to Regional Administrators (Dec. 18, 2008). This interpretation was based on both historic treatment of carbon dioxide in terms of the PSD program and the policy and practical implications of applying BACT. Id. Two months later, a new EPA Administrator agreed to reconsider the interpretation of "regulated pollutant" and explained that the formal interpretation "does not bind States issuing permits under their own State Implementation Plans" and that "other PSD permitting authorities should not assume that the [interpretation] is the final word on the appropriate interpretation of Clean Air Act requirements." Letter from Lisa P. Jackson, Administrator, EPA, to Mr. David Bookbinder, Chief Climate Counsel, Sierra Club (Feb. 17, 2009).
13. In Williams we interpreted the predecessor to the Utah Rulemaking Act. However, as we recognized then and in subsequent cases, the analysis is the same under either version of the Act. C.P. v. Utah Office of Crime Victims' Reparations, 966 P.2d 1226, 1230 n.5 (Utah 1998).
14. Amended in 2006 to incorporate, by reference, the provisions of 40 C.F.R. § 52.21(e), effective March 3, 2003.