POWDER RIVER BASIN RESOURCE COUNCIL and Sierra Club, Appellants (Petitioners),
v.
WYOMING DEPARTMENT OF ENVIRONMENTAL QUALITY, and Basin Electric Power Cooperative, Appellees (Respondents).
No. S-09-0037.
Supreme Court of Wyoming.
March 5, 2010.*811 Representing Appellants: James S. Angell and Robin Cooley of Earthjustice, Denver, Colorado. Argument by Robin Cooley.
Representing Appellee Wyoming Department of Environmental Quality: Bruce A. Salzburg, Wyoming Attorney General; Jay A. Jerde, Deputy Attorney General; Nancy E. Vehr, Senior Assistant Attorney General; Lucas J. Esch, Assistant Attorney General. Argument by Ms. Vehr.
Representing Appellee Basin Electric Power Cooperative, Inc.: Patrick R. Day and Mark R. Ruppert of Holland & Hart LLP, Cheyenne, Wyoming. Argument by Mr. Day.
Before VOIGT, C.J., and GOLDEN, HILL, KITE, and BURKE, JJ.
KITE, Justice.
[¶ 1] The Wyoming Department of Environmental Quality (DEQ) issued an air quality permit to Basin Electric Power Cooperative (Basin Electric) for a new coal-fired *812 electric power plant, called the Dry Fork Station, to be built in Wyoming's Powder River Basin. The Powder River Basin Resource Council and the Sierra Club (collectively the PRBRC) challenged that air quality permit before the Wyoming Environmental Quality Council (Council). After hearings on the PRBRC's different claims, the Council upheld the DEQ's issuance of the permit. The PRBRC appealed the Council's decision to the district court, which certified the appeal directly to this Court pursuant to W.R.A.P. 12.09(b). The Northern Cheyenne Tribe was granted leave to file an amicus curiae brief. We will affirm the issuance of the air quality permit.
ISSUES
[¶ 2] The PRBRC presents these issues for our consideration:
1. Whether the Wyoming Environmental Quality Council ("Council") and the Wyoming Department of Environmental Quality ("DEQ") violated the law by authorizing construction of the Dry Fork Station coal-fired power plant despite modeled violations of Class I air quality standards at the Northern Cheyenne Indian Reservation;
2. Whether the Council and DEQ violated the law by finding that DEQ was not required to consider more efficient supercritical technology as part of its best available control technology ("BACT") analysis for the Dry Fork Station and that the permit applicant alone defines the emission source DEQ may consider; and
3. Whether the Council and the DEQ violated the law by finding that DEQ did not have to consider greenhouse gas emissions as part of the BACT analysis for the Dry Fork Station.
FACTS
[¶ 3] Because the facts in this case are largely undisputed, we will present a very general background here, and provide more detailed facts as they arise in the discussion below. On November 10, 2005, Basin Electric submitted an application to the DEQ for an air quality permit for the Dry Fork Station, a new 422 megawatt coal-fired electric power plant proposed to be constructed near the Dry Fork coal mine, approximately seven miles north of Gillette, Wyoming. To obtain this permit, Basin Electric was required to demonstrate, among other things, that emissions from Dry Fork will not cause significant deterioration of existing air quality, and the power plant will use the best available control technology for each regulated pollutant.
[¶ 4] The DEQ reviewed the permit application, asking Basin Electric to provide additional information on several issues. The DEQ also considered public comments from interested parties, including the PRBRC and the Northern Cheyenne Tribe. The DEQ issued the permit on October 15, 2007. The PRBRC appealed to the Council. The Council granted a motion to dismiss the PRBRC's claim regarding greenhouse gases (listed above as Issue 3). It granted motions for summary judgment on PRBRC's claims regarding increment consumption and best available control technology (listed above as Issues 1 and 2). The PRBRC's appeal has now made its way before us for review, and the amicus curiae brief filed by the Northern Cheyenne Tribe supports the PRBRC on Issue 1.
STANDARD OF REVIEW
[¶ 5] Our standard of review for appeals from administrative agency decisions is governed by the Wyoming Administrative Procedure Act, which provides in pertinent part:
To the extent necessary to make a decision and when presented, the reviewing court shall decide all relevant questions of law, interpret constitutional and statutory provisions, and determine the meaning or applicability of the terms of an agency action. In making the following determinations, the court shall review the whole record or those parts of it cited by a party and due account shall be taken of the rule of prejudicial error. The reviewing court shall:
. . . .
*813 (ii) Hold unlawful and set aside agency action, findings and conclusions found to be:
(A) Arbitrary, capricious, an abuse of discretion or otherwise not in accordance with law;
(B) Contrary to constitutional right, power, privilege or immunity;
(C) In excess of statutory jurisdiction, authority or limitations or lacking statutory right;
(D) Without observance of procedure required by law; or
(E) Unsupported by substantial evidence in a case reviewed on the record of an agency hearing provided by statute.
Wyo. Stat. Ann. § 16-3-114 (LexisNexis 2007). As we have further explained:
When reviewing a case certified to us from district court pursuant to W.R.A.P. 12.09(b), we apply the appellate standards applicable to a reviewing court of the first instance. Williams Prod. RMT Co. v. State Dep't of Revenue, 2005 WY 28, ¶ 7, 107 P.3d 179, 182-183 (Wyo.2005). We review factual determinations for substantial evidence, meaning we consider whether there is relevant evidence in the entire record which a reasonable mind might accept in support of the agency's conclusions. Dale v. S & S Builders, LLC, 2008 WY 84, ¶ 21, 188 P.3d 554, 561 (Wyo.2008). Importantly, our review of any particular decision turns not on whether we agree with the outcome, but on whether the agency could reasonably conclude as it did based upon all of the evidence presented. Id., ¶ 23, 188 P.3d at 561.... We review an agency's conclusions of law de novo, and will affirm an agency's legal conclusion only if it is in accordance with the law. Dale, ¶ 27, 188 P.3d at 562.
Kennedy Oil v. Dep't of Revenue, 2008 WY 154, ¶ 7, 205 P.3d 999, 1002 (Wyo.2008).
[¶ 6] While the interpretation of statutes and their implementing regulations is a question of law that we review de novo, it is also settled that we defer to an agency's interpretation of its own rules and regulations unless that interpretation is clearly erroneous or inconsistent with the plain language of the rules. Pinther v. Wyoming Dep't of Admin. and Info., 866 P.2d 1300, 1302 (Wyo.1994); RME Petroleum Co. v. Wyoming Dep't of Revenue, 2007 WY 16, ¶ 44, 150 P.3d 673, 689 (Wyo.2007). Accordingly, when we review the DEQ's interpretations of regulations promulgated under Wyoming's Environmental Quality Act, we apply the same standard the Council was required to use: we accept those interpretations unless they are clearly erroneous or inconsistent with the plain language of the rules.
DISCUSSION
[¶ 7] The DEQ administers and enforces the Wyoming Environmental Quality Act, Wyo. Stat. Ann. §§ 35-11-101 through XX-XX-XXXX (LexisNexis 2009). The DEQ's Air Quality Division is responsible for the air quality program, and it operates under the Wyoming Air Quality Standards and Regulations (WAQSR). The federal Environmental Protection Agency has approved of Wyoming's air quality regulatory program, 40 C.F.R. part 52, subpart ZZ, and so the DEQ is the primary regulatory authority for air quality in Wyoming. See 42 U.S.C. § 7410(a). However, because the state program is intended to be compatible with, and at least as stringent as, the federal Clean Air Act, federal precedent and regulatory guidance is persuasive authority in Wyoming air quality cases.
Issue 1. Increment Protection
[¶ 8] Before reaching the heart of this issue, it is helpful to review the applicable law and introduce some key terms. Wyoming's Environmental Quality Act states that "No person shall cause, threaten or allow the discharge or emission of any air contaminant in any form so as to cause pollution which violates rules, regulations and standards adopted by the council." Section 35-11-201. The rules and regulations require the DEQ to review permit applications for proposed major sources of air emissions. It issues a permit only if the proposed source demonstrates that its emissions will not cause significant deterioration of ambient air quality. 6 WAQSR §§ 2 and 4.
*814 [¶ 9] More specifically, the regulations provide that the DEQ's review must
include analysis of the predicted impact of the allowable and secondary emissions from the stationary source.... Such analysis shall identify and quantify the impact on the air quality in the area of all emissions not included in the baseline concentrations including, but not limited to, those emissions resulting from the instant application and all other permits issued in the area. The purpose of this analysis is to determine the total deterioration of air quality from the baseline concentrations.... A permit to construct ... shall be issued only ... if the predicted impact (over and above the baseline concentration) of emissions defined above is less than the maximum allowable increment shown in Table 1 for the classification of the area in which the impact is predicted.
6 WAQSR § 4(b)(i)(A)(I). To predict whether the impacts of a proposed source's emissions will exceed the increments shown in Table 1, one tool available to the DEQ is a computer model that estimates what the impacts will be. The results of the computer model are based on information that includes the emissions from the proposed source and other sources in the area, air dispersion over time and distance, varying terrain, and meteorological data such as wind direction, wind speed, and temperatures.
[¶ 10] The DEQ's review includes two separate phases to determine whether emissions from the proposed source and other area sources will cause or contribute to increment exceedances. In the first phase, which we will refer to as the screening phase, the computer model is run to estimate the impacts of emissions from the proposed source alone. Results from this model run are compared to Significant Impact Levels, which are a very small percentage (generally 4%) of the increments. If the computer model indicates that the estimated impacts of emissions from the proposed source alone are below the Significant Impact Levels, then the DEQ can determine that the cumulative impacts of emissions from the proposed source and other area sources will not exceed the increments, and no further analysis is necessary. In this case, the computer model indicated that the estimated impacts of emissions from Dry Fork were below the Significant Impact Levels for particulate matter and nitrogen dioxide, but above the Significant Impact Levels for sulfur dioxide. The DEQ therefore required Basin Electric to proceed to the second phase of review for sulfur dioxide emissions.
[¶ 11] In the second phase, which we will refer to as the cumulative phase, the computer model is run to estimate the impacts of the combined emissions from the proposed source along with emissions from other area sources. The model results are compared to the increments. As the regulation provides, a permit is issued for the proposed source only if the estimated impacts are less than the maximum allowable increments.
[¶ 12] The other area sources of particular significance in this case are the coal-fired electric power plants known as Colstrip Units 3 and 4, located in Montana approximately 120 miles north of the proposed Dry Fork location. The Colstrip Units are only about fifteen miles from the Northern Cheyenne Indian Reservation (Reservation). The Reservation is a Class I area which, in air quality terms, is an area where existing air quality is considered pristine. See 42 Fed. Reg. 40,695 (Aug. 5, 1977). Air quality in Class I areas is more stringently protected than air quality in Class II areas. For example, the 24-hour increment for sulfur dioxide is 5 micrograms per cubic meter in Class I areas, and 20 micrograms per cubic meter in Class II areas. 6 WAQSR § 4(b)(i)(A)(I), Table 1. Given the proximity of the Colstrip Units to the Reservation, their emissions can have a significant influence on the Reservation's air quality. As a result, when the computer models are run to estimate the impacts of combined emissions from Dry Fork and other area sources, the emissions from the Colstrip Units have a significant influence on the results.
[¶ 13] In the cumulative phase, Basin Electric ran the computer model using different assumptions about the sulfur dioxide emissions from the Colstrip Units. For our purposes, it is sufficient to consider only two of these computer model runs. As further *815 detailed below, in the first model run Basin Electric used the maximum actual daily emission rate for the Colstrip Units. In the second model run, Basin Electric used the maximum allowable daily emission rate for the Colstrip Units.[1] The regulations do not specify that maximum allowable emissions rates from existing sources must be used in this analysis.
[¶ 14] In the first computer model run, Basin Electric obtained the actual reported sulfur dioxide emissions from Colstrip Units 3 and 4 for the years 2004 and 2005. It combined the emissions from the two units, and determined the day on which emissions from the two units were at their maximum. It then ran the computer model assuming that the Colstrip Units would emit sulfur dioxide continuously at this maximum actual one day rate. In reality, the Colstrip Units emitted sulfur dioxide at this maximum rate only one day during this two year period, and at lower rates the other 730 days. Accordingly, because the model used this worst case maximum actual daily emission rate for the Colstrip Units, the estimated impacts were necessarily higher than the actual impacts will be. Even so, the results of this first model run estimated that the impacts of combined actual emissions from the Colstrip Units, the Dry Fork Station, and all other area sources would not exceed the Class I increments for sulfur dioxide at any location within the Reservation.
[¶ 15] In the second computer model run, Basin obtained information on the maximum emissions the Colstrip Units are allowed under their air quality permits. These are known as the allowable emissions rates, and unless the Colstrip Units are violating their air quality permits, the allowable emissions are always higher than actual emissions. Basin Electric then ran the computer model assuming that the Colstrip Units would continuously emit sulfur dioxide at the maximum allowable rate. Because actual emissions from the Colstrip Units are well below the highest allowable rates, this second computer model run overestimated impacts to an even greater degree than the first computer model run. The results of the second run showed that the impacts of sulfur dioxide emissions from Dry Fork and other area sources could exceed the Class I increments within the Reservation.
[¶ 16] More specifically, the model estimated that there would be, over a three year period, forty-seven violations of the 24-hour increment of 5 micrograms per cubic meter. For eighteen of these modeled increment exceedances, emissions from the Dry Fork Station had no impact at all, meaning that the exceedances were caused entirely by other area sources, chiefly the Colstrip Units. For the remaining twenty-nine modeled increment exceedances, the estimated impact of Dry Fork's emissions was between 0.0002 and 0.0009 micrograms per cubic meter.[2]
[¶ 17] That brings us to the heart of this dispute. The PRBRC contends that, because the second computer model run indicated exceedances of the increment, the DEQ could not legally issue the air quality permit for the Dry Fork Station. The DEQ and Basin Electric point out that the first computer model run using maximum actual emissions projected no increment exceedances. Even in the second computer model run using maximum allowable emissions, Dry Fork's contributions to the increment exceedances were so exceedingly small that the DEQ treated them as non-existent. Under these circumstances, the DEQ and Basin Electric assert, the agency had discretion to determine that Dry Fork will not cause or contribute to any actual exceedances of the increment, and the air quality permit was properly issued.
[¶ 18] The PRBRC's position, supported by the Northern Cheyenne Tribe, is based squarely on the language of the regulation: *816 "A permit to construct ... shall be issued only ... if the predicted impact ... is less than the maximum allowable increment...." 6 WAQSR § 4(b)(i)(A)(I). This language is plain and unambiguous, the PRBRC contends, and because the second computer model run indicated exceedances of the increment, the permit could not be issued.
[¶ 19] The DEQ's position, supported by Basin Electric, is that the DEQ is allowed a certain amount of flexibility in administering Wyoming's air quality program. It asserts that the regulatory language relied upon by the PRBRC must be read together with the provision of 6 WAQSR § 2(c)(iii) that no permit shall be granted "unless the applicant shows, to the satisfaction of the Administrator of the Division of Air Quality that ... [t]he proposed facility will not cause significant deterioration of existing ambient air quality." The DEQ contends that Basin Electric showed, to the satisfaction of the agency, that Dry Fork would not cause significant deterioration of air quality within the Reservation. For the increment exceedances indicated by the second run of the computer model, the contributions of Dry Fork's emissions were extremely small, ranging from 0.0002 to 0.0009 micrograms per cubic meter. The DEQ contends that it properly exercised its discretion to overlook these modeled increment exceedances because Dry Fork's contributions were well below the Significant Impact Levels used by the DEQ.
[¶ 20] Under the DEQ's interpretation of its own rules and regulations, this use of Significant Impact Levels is proper. We defer to that interpretation unless it is clearly erroneous or inconsistent with the plain language of the regulations. Pinther, 866 P.2d at 1302. The DEQ contends that its interpretation cannot be considered clearly erroneous because it is well established by the agency's longstanding practices, and consistent with the EPA's interpretation.
[¶ 21] The DEQ explains that the use of Significant Impact Levels in air quality regulation arose from a leading federal air quality case in which the court stated that an agency should have "authority to provide exemption when the burdens of regulation yield a gain of trivial or no value." Alabama Power Co. v. Costle, 636 F.2d 323, 360-61 (D.C.Cir. 1979). This, the DEQ says, led the federal EPA to develop Significant Impact Levels below which the impacts of a proposed source's emissions are treated as de minimis rather than as a disqualification from receiving permit. The EPA later proposed a regulation allowing the use of Significant Impact Levels. See 61 Fed.Reg. 38250, 38292 (July 23, 1996). The DEQ and Basin Electric admit that the EPA never finalized or promulgated the regulation, but point out that the EPA "is aware that many States have been using these proposed [Significant Impact Levels] for ... screening tools since 1996." 72 Fed.Reg. 54,112 (Sept. 21, 2007). The DEQ and Basin Electric also cite federal environmental cases supporting the use of Significant Impact Levels. See, e.g., In re Prairie State Generating Co., 2006 WL 2847225, P.S.D. Appeal No. 05-05, slip op. (EAB Aug. 24, 2006).
[¶ 22] The difficulty we have with the DEQ's position is one emphasized by the Northern Cheyenne Tribe. As discussed earlier, the DEQ employs a two-phase review of air quality impacts. The screening phase analyzes emissions from the proposed source by itself. If the screening phase indicates that further analysis is required, the cumulative phase proceeds to analyze emissions from the proposed source and other area sources. After careful consideration of the authorities cited by the DEQ, we agree that they support the use of Significant Impact Levels in the screening phase. It is less clear that these authorities support the use of Significant Impact Levels in the cumulative phase, as the DEQ did in this case.
[¶ 23] In the preamble to its proposed regulation, the EPA explained that Significant Impact Levels "would be used to determine whether a new major source or major modification, due to the predicted ambient concentration from its own emissions, would be required to conduct a comprehensive Class I increment analysis for a given pollutant." 61 Fed.Reg. at 38292. This clearly establishes the EPA's position that Significant Impact Levels may be employed in the screening phase, but there is no indication that the EPA intended them to be used in *817 the cumulative phase as well. Accordingly, while it is true that the EPA recognizes that states use Significant Impact Levels as "screening tools," 72 Fed.Reg. at 54,140, it is also true that these screening tools were meant to be used "for determining when a new major source ... must conduct a more extensive air analysis to demonstrate that it will not cause or contribute to a violation of the ... increment." Id. at 54,138. In other words, these regulatory materials support the DEQ's use of Significant Impact Levels in the initial screening phase, but provide no support for their use in the subsequent cumulative phase.
[¶ 24] Other authorities suggest that Significant Impact Levels may be used in the cumulative phase. In Prairie State, for example, the Environmental Appeals Board (the administrative body authorized to hear appeals from EPA decisions) appears to have approved of the EPA's use of Significant Impact Levels in both the screening phase and the cumulative phase. Prairie State, slip op. at 134, 137-38.[3] Similar support appears in the Environmental Protection Agency's "New Source Review Workshop Manual: Prevention of Significant Deterioration and Nonattainment Area Permitting" (Draft, Oct. 1990) (NSR Manual).[4] The manual indicates how Significant Impact Levels are to be used in the screening phase:
The EPA does not require a full impact analysis for a particular pollutant when emissions of that pollutant from a proposed source or modification would not increase ambient concentrations by more than prescribed significant ambient impact levels.
Id. at C.24. It also indicates how Significant Impact Levels are to be used in the cumulative phase:
When a violation of any ... increment is predicted at one or more receptors in the impact area, the applicant can determine whether the net emissions increase from the proposed source will result in a significant ambient impact at the point (receptor) of each predicted violation, and at the time violation is predicted to occur. The source will not be considered to cause or contribute to the violation if its own impact is not significant.... In such a case, the permitting agency, upon verification of the demonstration, may approve the permit.
Id. at C.52 (emphasis in original). Thus, the manual does seem consistent with the DEQ's application of Significant Impact Levels in both the screening phase and the cumulative phase.
[¶ 25] On the other hand, the EPA has made it clear that, in at least some circumstances, Significant Impact Levels cannot be used in the cumulative phase. In 2002, the EPA's comments on North Dakota's State Implementation Plan included this statement:
We have recently consulted with our Headquarters office and it is EPA's position (as we stated in an August 30, 2001 letter to the North Dakota Department of Health) that it is not appropriate to establish Class I significance levels when an increment violation already exists. We believe any impact (not just one that is "significant") on a receptor in a Class I area that shows a violation of the PSD increment would be considered to contribute to that violation. Furthermore, we believe that, even if some of the impacts are relatively small they are still contributing to an existing problem.
Under current EPA policy, the PSD Class II significant impact levels are used primarily as a threshold in new source permitting to determine the scope of the modeling analysis.
Letter from Richard R. Long, EPA, to Terry L. O'Clair, North Dakota Department of Health, April 12, 2002.
*818 [¶ 26] Unfortunately, we are unable to determine how this statement from the EPA might apply in our current case, because the record does not establish whether the Class I area on the Reservation has already experienced increment violations. The Northern Cheyenne Tribe's brief describes the air quality monitoring program established on the Reservation, but does not say whether any actual increment exceedances have been monitored. At oral argument, counsel for the parties suggested that there have been no measured exceedances, but the record is insufficient to establish or refute that fact. However, while we cannot determine the applicability of the EPA's statement, we can perceive its generally disapproving tone as to the use of Significant Impact Levels in the cumulative phase of review. This, together with the other EPA materials considered, leaves the federal agency's interpretation uncertain, or at least unclear. The EPA's inconsistency undermines the DEQ's argument that its interpretation cannot be clearly erroneous because it is supported by the EPA.
[¶ 27] The DEQ and Basin Electric further assert that, over the past several years, the DEQ has routinely used Significant Impact Levels when reviewing air quality permit applications. Conceding that the use of Significant Impact Levels is not explicitly authorized by any particular Wyoming regulation, the DEQ maintains that they are commonly used by many air quality regulatory agencies, and that their use is well within the DEQ's discretion. However, this assertion by the DEQ suffers from the same flaw as its previous argument. The record indicates that the DEQ has used Significant Impact Levels in at least ten permit reviews, but in almost every instance, it did so in the screening phase. There appears to be only one instance in which the DEQ employed Significant Impact Levels in the cumulative phase. A single, recent example is insufficient to establish that the DEQ has a settled regulatory interpretation entitled to deference. See RME Petroleum, ¶ 44, 150 P.3d at 689.
[¶ 28] At this point, it seems appropriate to quote the applicable regulation again. It requires the DEQ's permit review to
include analysis of the predicted impact of the allowable and secondary emissions from the stationary source.... Such analysis shall identify and quantify the impact on the air quality in the area of all emissions not included in the baseline concentrations including, but not limited to, those emissions resulting from the instant application and all other permits issued in the area. The purpose of this analysis is to determine the total deterioration of air quality from the baseline concentrations.... A permit to construct ... shall be issued only ... if the predicted impact (over and above the baseline concentration) of emissions defined above is less than the maximum allowable increment shown in Table 1 for the classification of the area in which the impact is predicted.
6 WAQSR § 4(b)(i)(A)(I).
[¶ 29] The regulation clearly does not provide authority for the DEQ to treat small exceedances as de minimis and issue the permit anyway. If it were meant to provide such discretion, we would expect it to say, "A permit to construct ... shall be issued only... if the predicted impact ... is not significantly greater than the maximum allowable increment shown in Table 1," or words to that effect. The regulation does not say that, or anything like that. Moreover, the increments shown in Table 1 are precise numbers, not subject to interpretation. The 24-hour increment for sulfur dioxide, for example, is 5 micrograms per cubic meter, not approximately 5 micrograms per cubic meter. Given the language of this regulation, we see no room for the DEQ to waive application of the increment through the use of Significant Impact Levels, and no authority for the DEQ to invoke Significant Impact Levels to issue a permit despite modeled exceedances of the increment, no matter how small those exceedances might be or how small the proposed source's contribution may be. We therefore conclude that the DEQ's reliance solely on Significant Impact Levels is not consistent with the language of the regulations.
[¶ 30] We are mindful of the argument presented by the DEQ and Basin Electric that this regulation must be read in *819 context. "The rules of statutory interpretation also apply to the interpretation of administrative rules and regulations." Powder River Coal Co. v. Wyoming State Bd. of Equalization, 2002 WY 5, ¶ 6, 38 P.3d 423, 426 (Wyo.2002). Accordingly, "all portions of an act [or regulation] must be read in pari materia, and every word, clause and sentence of it must be considered so that no part will be inoperative or superfluous." KP v. State, 2004 WY 165, ¶ 22, 102 P.3d 217, 224 (Wyo. 2004). We are guided by the "full text of the statute [or regulation], paying attention to its internal structure and the functional relation between the parts and the whole." Hede v. Gilstrap, 2005 WY 24, ¶ 6, 107 P.3d 158, 163 (Wyo.2005).
[¶ 31] The internal structure of the applicable regulations is explained in the regulations themselves:
Chapter 6 establishes permitting requirements for all sources constructing and/or operating in the State of Wyoming. Section 2 covers general air quality permitting requirements for construction and modification as well as minor source permits to operate. Section 3 is the state operating permit program required under Title V of the Clean Air Act. Section 4 is the prevention of significant deterioration (PSD) program.
6 WAQSR § 1(a). Section 2, covering general permitting requirements, provides that no permit may be issued "unless the applicant shows, to the satisfaction of the Administrator of the Division of Air Quality," that:
The proposed facility will not cause significant deterioration of existing ambient air quality in the Region as defined by any Wyoming standard or regulation that might address significant deterioration.
WAQSR § 2(c)(iii). Section 4, the prevention of significant deterioration section, includes the regulation providing that a permit shall be issued only if the predicted impact of emissions is less than the increment. 6 WAQSR § 4(b)(i)(A)(I).
[¶ 32] The DEQ and Basin Electric repeatedly point out the provision of Section 2 that a permit applicant must show compliance with permitting requirements "to the satisfaction of the Administrator of the Division of Air Quality." This language, they contend, provides sufficient discretion to justify using Significant Impact Levels in the cumulative phase of permit review. We can hardly disagree that this language provides discretionary authority to the agency, but that authority is limited. Section 2 provides that a permit applicant must demonstrate to the agency's satisfaction that the proposed source "will not cause significant deterioration of existing ambient air quality in the Region as defined by any Wyoming standard or regulation that might address significant deterioration." This gives the agency discretion to determine whether or not the proposed source will cause significant deterioration. It does not give the agency discretion in determining what significant deterioration is. Rather, significant deterioration is as "defined by any Wyoming standard or regulation that might address significant deterioration."
[¶ 33] We must therefore disagree with the Council's conclusion that the DEQ properly used Significant Impact Levels to determine that the Dry Fork Station would not cause or contribute to increment exceedances on the Reservation. However, we will ultimately affirm the Council's ruling on summary judgment that the DEQ properly issued the air quality permit to Basin Electric, because there is another valid basis for that decision. In the administrative law context, we have indicated that "we review an agency's order granting a summary judgment in the same manner as in the civil context by employing our de novo standard of review and utilizing the same standards and reviewing the same materials as the agency." Rollins v. Wyo. Tribune-Eagle, 2007 WY 28, ¶ 7 n. 7, 152 P.3d 367, 370 (Wyo.2007). In the civil context, it is well established that "we can affirm a district court's summary judgment order on any basis apparent in the record." Stewart Title Guaranty Co. v. Tilden, 2005 WY 53, ¶ 22, 110 P.3d 865, 874 n. 7 (Wyo.2005). The combination of these two concepts suggests that we can affirm an administrative agency's summary judgment decision on any basis apparent in the record.
*820 [¶ 34] The basis on which we sustain the Council's summary judgment decision in this case is mentioned, but not pressed, in the DEQ's brief. The regulation provides that the permit shall be issued only if the predicted impact is less than the increment. One definition of predict is "to declare in advance; [to] foretell on the basis of observation, experience, or scientific reason." Webster's Ninth New Collegiate Dictionary 926 (1991). In applying its experience and scientific reason, the DEQ must exercise its discretion to make a sound prediction of whether the impact of emissions from a proposed source will be less than the increment.
[¶ 35] It is appropriate to conclude that the DEQ has discretion in determining the predicted impacts of pollution emissions, but does not have discretion to depart from the increments by the use of Significant Impact Levels. When the regulations provide specific numbers, such as the increments, the regulated community and other interested parties expect the DEQ to apply those numbers strictly as written. If the DEQ is going to use Significant Impact Levels to vary from those specific numbers, then it should incorporate the Significant Impact Levels into its regulations in order to provide notice of that practice to the regulated community and other interested parties. In contrast, when the regulation requires the DEQ to predict impacts of pollution emissions, the agency can be expected to apply its judgment and expertise to make a sound prediction. Some discretion is inherent in the exercise of judgment and expertise.
[¶ 36] In the analysis of emissions from the Dry Fork Station, exceedances were modeled only when the maximum allowable emissions for the Colstrip Units were used, but not when the maximum actual emissions were used. There is nothing in the regulations requiring the DEQ to consider only maximum allowable emissions from existing sources in its analysis, or to ignore models using the maximum actual emissions. Indeed the NSR Manual at C.48-49 indicates that it is maximum actual emissions that should be modeled for area sources such as the Colstrip Units. The DEQ could reasonably determine that the computer modeling based on maximum actual emissions provided a better prediction than the computer modeling based on maximum allowable emissions.
[¶ 37] We therefore conclude that the regulations provide the DEQ with some discretion in predicting the impacts of emissions from a proposed source. In this appeal, the DEQ does not explicitly contend that it reached its decision on this basis, nor did it defend its decision on this basis before the Council. However, that is the logical underpinning of its decision. We could remand this case to allow the DEQ to explain its decision on the proper basis, but we fail to see what that would accomplish. See, State ex rel. Arnold v. Ommen, 2009 WY 24, ¶ 32, 201 P.3d 1127, 1136 (Wyo.2009). In this case, the facts are not in dispute, "making it possible for this Court to review without re-weighing disputed evidence." Wells Fargo Bank v. Hodder, 2006 WY 128, ¶ 32, 144 P.3d 401, 413 (Wyo.2006). Based on the DEQ's analysis as reflected in the record, there is no doubt that the fundamental position of the DEQ was that no increment exceedances could be predicted to result from the operation of the Dry Fork facility. Based on common sense and notions of judicial economy, it is appropriate for us to decline to remand a case if further proceedings "would serve no useful purpose." Cellers v. Adami, 2009 WY 120, ¶ 2, 216 P.3d 1134, 1136 (Wyo. 2009); State v. Homar, 798 P.2d 824, 826 (Wyo.1990).
[¶ 38] The first computer model run was based on maximum actual emissions from Colstrip, rather than typical actual emissions, so it necessarily overpredicted the impacts. Even so, it indicated no increment exceedances on the Reservation. The second computer run was based on maximum allowable emissions, so it overpredicted the impact of emissions to an even greater degree. Although it did model theoretical increment exceedances on the Reservation, the estimated contribution of the Dry Fork emissions was exceedingly small. Moreover, according to the expert who performed the computer modeling, in every instance for which this second model run indicated increment exceedances on the Reservation, "the winds were blowing the Dry Fork emissions away from *821 the [Reservation]. It was not physically possible for Dry Fork to contribute to this problem." (Emphasis added.)
[¶ 39] The computer model does not make predictions, but rather, it is a tool for the DEQ to use to make predictions. Given the DEQ's and the Council's position that the regulations provided the agency some discretion, and their ultimate conclusion that there would be no actual increment violation, it seems obvious the DEQ, based on its experience, expertise, and judgment, gave greater weight to the results of the first computer model run, and discounted the results of the second. On that basis, the DEQ essentially decided that the predicted impacts of emissions from the Dry Fork Station will be less than the maximum allowable increment. We therefore affirm the Council's summary judgment in favor of Basin Electric on the increment consumption issue.
Issue 2. BACTControl Technology versus Redesign
[¶ 40] In its application for the air quality permit for the Dry Creek Station, Basin Electric was required to demonstrate that "[t]he proposed facility will utilize the Best Available Control Technology with consideration of the technical practicability and economic reasonableness of reducing or eliminating the emissions resulting from the facility." 6 WAQSR § 2(c)(v). Best Available Control Technology, commonly referred to as BACT, is defined as
an emission limitation (including a visible emission standard) based on the maximum degree of reduction of each pollutant subject to regulation under these Standards and Regulations or regulation under the Federal Clean Air Act, which would be emitted from or which results [from] any proposed major stationary source or major modification which the Administrator, on a case-by-case basis, taking into account energy, environmental, and economic impacts and other costs, determines is achievable for such source or modification through application [of] production processes and available methods, systems, and techniques, including fuel cleaning or treatment or innovative fuel combustion techniques for control of such pollutant.
6 WAQSR § 4(a). In simplified terms, this regulation requires the DEQ to consider a broad range of available pollution control and reduction options, determine which can be achieved reasonably, and impose those as permit requirements.
[¶ 41] The PRBRC asserts that the DEQ did not consider a sufficiently broad range of pollution control options when it reviewed the Dry Fork permit application. As proposed by Basin Electric, the Dry Fork Station will incorporate a "subcritical" boiler, which will operate at temperatures and pressures below the critical point of water. The PRBRC contends that, as part of its BACT analysis, the DEQ should also have considered a "supercritical" boiler, which would operate at temperatures and pressures above the critical point of water.[5] The PRBRC maintains that supercritical boiler technology is, in the terms used in the regulation, one of the "production processes" or "available methods, systems, [or] techniques" by which pollution from the Dry Fork Station could be reduced.
[¶ 42] The first step in resolving this issue is to interpret the BACT regulations. As noted above, we defer to an agency's interpretation of its own rules and regulations unless that interpretation is clearly erroneous or inconsistent with the plain language of the rules. Pinther, 866 P.2d at 1302; RME Petroleum, ¶ 44, 150 P.3d at 698. With that in mind, we consider the DEQ's interpretation of its BACT regulations.
[¶ 43] The DEQ begins by pointing out that the BACT regulations obligate it to analyze the pollutants emitted by "any proposed major stationary source." The source is proposed in the permit application, and the DEQ must analyze other "production processes and available methods, systems, and techniques" that can be applied to or adapted by the proposed source. However, the DEQ *822 is not required to analyze options that would require substantial changes to the basic design of the proposed source. This interpretation by the DEQ is consistent with the regulatory language, and the DEQ has apparently applied this interpretation since at least 1993. In a case before the Council that year, the permit applicant had proposed a pulverized coal boiler. The Council upheld the DEQ's decision that the BACT regulations did not require analysis of "a different type of boiler, such as a circulating fluidized bed boiler" because that would require the DEQ "to redefine the source." In re Permit Issued to Black Hills Power & Light Co., Neil Simpson Unit # 2, Docket No. 2476-93 (Council 1993).
[¶ 44] Wyoming's regulatory definition of BACT is nearly identical to the federal statutory definition found at 42 U.S.C. § 7479(3). It is therefore persuasive that the DEQ's interpretation is also consistent with the EPA's interpretation:
Historically, EPA has not considered the BACT requirement as a means to redefine the design of the source when considering available control alternatives. For example, applicants proposing to construct a coal-fired electric generator, have not been required by EPA as part of a BACT analysis to consider building a natural gas-fired electric turbine although the turbine may be inherently less polluting per unit product (in this case electricity).
NSR Manual at B.13.
[¶ 45] The EPA's interpretation has been upheld repeatedly in litigation. For example, in Sierra Club v. United States Environmental Protection Agency, 499 F.3d 653, 654 (7th Cir.2007), the court accepted the EPA's interpretation that BACT "does not include redesigning the plant proposed by the permit applicant." More recently, in Blue Skies Alliance v. Texas Comm'n on Environmental Quality, 283 S.W.3d 525, 535 (Tex.App.2009), the court concluded "that a BACT analysis must consider any control technology that may be applied to the proposed facility, but does not need to consider any control technology that would require such a redesign of the facility that it would constitute an alternative proposal." See also, In re Knauf Fiber Glass, 8 E.A.D. 121 (E.A.B.1999); In re Old Dominion Elec. Coop., 3 E.A.D. 779, 793 (Adm'r 1992).
[¶ 46] Because we conclude that the DEQ's interpretation is not inconsistent with the plain language of the regulation, and is not clearly erroneous, we accept that interpretation. That, however, is only the first step in resolving the issue. As stated in Sierra Club, 499 F.3d at 655, the next question is "where control technology ends and a redesign of the `proposed facility' begins." In other words, we must decide whether imposing supercritical boiler technology on Basin Electric would redefine the basic design of the proposed Dry Fork Station. The Council's findings of fact on this question are clear and concise:
12. Supercritical boilers operate at temperatures and pressures above the "critical point" of water, while subcritical boilers operate at temperatures and pressures below the critical point of water. As a result of these different pressure and temperature conditions, changing from subcritical to supercritical technologies would require a different boiler made with different steel alloys, different water wall tubing, different valves, different turbines, different reheaters, different boiler feed pumps, and a different economizer.
. . . .
16. DEQ did not consider ... supercritical technologies to be control technologies that had to be evaluated as part of the BACT process.... As a consequence, although DEQ did require Basin Electric to explain the reasons for its decision not to employ these technologies, DEQ did not submit these technologies to a separate "BACT" analysis as potential pollution control options when issuing Basin Electric's permit.
17. DEQ did not do so because it considered these technologies to be fundamentally different emission source technologies than the one proposed by Basin Electric and, if applied, would require Basin Electric to "redefine" its proposed emissions source, a subcritical pulverized coal boiler, contrary to DEQ's interpretation of [the regulations].
*823 (Internal citations omitted.) Our review of the record revealed substantial evidence to support these findings of fact, including reports from experts and affidavits from DEQ personnel. Under the applicable standard of review, these findings must be affirmed.
[¶ 47] As found by the Council, requiring the Dry Fork Station to adopt supercritical boiler technology would force it to make substantial changes to its proposed design. Such substantial changes seem less like a control technology option and more like a redesign of the proposed source. That is the decision reached by the DEQ. Because this decision was made "in a technically complex field with limited statutory guidance," it is better "entrusted to the judgment of the agency that administers the regulatory scheme rather than to courts of generalist judges." Sierra Club, 499 F.3d at 656.
[¶ 48] The PRBRC disagrees with the DEQ's judgment, and contends that the DEQ was required to perform a BACT analysis of supercritical boiler technology because it is a potentially lower-polluting production process or method. In support of its contention, the PRBRC cites In re Prairie State Generating Co., in which an air quality permit was issued for a pulverized coal-fired power plant. As part of its BACT analysis, the permitting agency considered an option known as integrated gasification combined cycle technology (IGCC).[6] The agency required this analysis even though "selection of IGCC would have required extensive design changes to Prairie State's proposed facility." Prairie State, slip op. at 36. This decision, the PRBRC maintains, indicates that the DEQ was required to consider supercritical boiler technology despite the fact that it would require extensive design changes to the Dry Fork Station.
[¶ 49] We disagree that Prairie State supports the PRBRC's position. The specific issue in that case was whether the permitting agency's BACT analysis should have included the use of low-sulfur coal from the western United States in place of the high-sulfur coal from Illinois proposed by the permit applicant. The Board upheld the agency's decision not to require consideration of low-sulfur coal, and the Board's decision was ultimately upheld on judicial review. Sierra Club, 499 F.3d at 657. The Board was not asked to decide whether the agency was required to consider IGCC technology. It appears to have noted that the agency considered IGCC technology only to show how broad a range of options the agency had considered in its BACT analysis. The Board did not hold that the agency was required to perform BACT analysis on a technology that would require extensive design changes. Rather, it suggested in dicta that the agency may have done even more than the law required.
[¶ 50] The PRBRC further contends that supercritical boiler technology would not actually redefine or redesign the Dry Fork Station as proposed by Basin Electric. This argument rests on the concept that a facility's basic design is defined by the raw materials it uses and the product it produces. On this basis, the PRBRC concedes that a permit applicant who proposes a coal-fired power plant is not required to consider a natural gas-fired power plant, because that would require the use of a different raw material as input. In contrast, the PRBRC contends, an applicant who proposes a coal-fired power plant can be forced to use any technology options available, so long as the raw material is coal and the product is electricity. Characterizing subcritical and supercritical boiler technology as "different flavors" of coal-fired power plants, the PRBRC insists that the DEQ must consider the option of supercritical boiler technology in its BACT analysis.
[¶ 51] It is too simplistic to say that a proposed source is defined solely by the raw materials it uses and the product it makes. The PRBRC's contention that subcritical and supercritical boiler technology represent the same basic design is contrary to the fact, as found by the Council, that "changing from subcritical to supercritical technologies would require a different boiler made with different steel alloys, different water wall tubing, different valves, different turbines, different reheaters, *824 different boiler feed pumps, and a different economizer." Moreover, the PRBRC's contention is contrary to the decision in Sierra Club, 499 F.3d 653, which held that the agency did not have to consider the option of low-sulfur coal from the western United States as a replacement for high-sulfur coal from a nearby Illinois mine. Even though both would use the same raw material to make the same product, the court held that analysis of low-sulfur coal was not required because "[t]o burn low-sulfur coal, Prairie State would have to arrange for it to be transported from mines more than a thousand miles away and would have to make changes in the design of the plant-specifically, the design of the plant's facilities for receiving coal." Id. at 654. In the case before us now, to employ supercritical boiler technology, Basin Electric would have to make changes in the design of the plant-specifically, the design of the plant's boiler. For that reason, supercritical boiler technology is beyond the scope of options the DEQ was required to consider in its BACT analysis.
[¶ 52] In summary, we have determined that the DEQ's interpretation of the BACT regulations is not clearly erroneous or inconsistent with the plain language of the regulations, and is supported by persuasive federal authority. Accordingly, we agree that BACT analysis does not have to include options that would require the proposed source to redefine its basic design. The record contains substantial evidence to support the Council's findings that imposing supercritical boiler technology on the Dry Fork Station would require extensive changes to its basic design. These determinations lead to the conclusion that the DEQ was not required by the BACT regulations to consider supercritical boiler technology as an alternative to Basin Electric's proposed subcritical boiler technology. We affirm the Council's decision on this issue.
Issue 3. Greenhouse Gas Emissions
[¶ 53] The Dry Fork Station is predicted to emit 3.7 million tons per year of carbon dioxide, along with lesser amounts of other gases that the PRBRC characterizes as greenhouse gases.[7] The PRBRC asserts that the DEQ was required to impose BACT requirements forcing Dry Fork to control its emissions of carbon dioxide. Indeed, the PRBRC claims that the DEQ's failure to do so is "indefensible."
[¶ 54] Under Wyoming's regulations, BACT is "an emission limit ... based on the maximum degree of reduction of each pollutant subject to regulation under these Standards and Regulations or regulation under the Federal Clean Air Act." 6 WAQSR § 4(a). The PRBRC does not maintain that carbon dioxide is subject to regulation under the Wyoming Air Quality Standards and Regulations. Rather, it contends that carbon dioxide is subject to BACT analysis and control because it is subject to regulation under the federal Clean Air Act. Accordingly, we consider this issue solely under federal law.
[¶ 55] Federal regulations define which pollutants are considered subject to regulation under the federal Clean Air Act:
(i) Any pollutant for which a national ambient air quality standard has been promulgated...;
(ii) Any pollutant that is subject to any [new source review] standard promulgated under section 111 of the [Clean Air] Act;
(iii) Any Class I or II substance subject to [an ozone protection] standard promulgated under or established by title VI of the Act; or
(iv) Any pollutant that otherwise is subject to regulation under the Act.
40 C.F.R. §§ 51.166(b)(49), 52.21(b)(50). While conceding that carbon dioxide does not fall within the first three categories, the PRBRC contends that carbon dioxide is a "pollutant that otherwise is subject to regulation under the Act." It bases this contention on the fact that, in 1993, the federal EPA promulgated regulations requiring specified sources to monitor and report emissions of *825 carbon dioxide. See 40 C.F.R. §§ 75.10, 17.13, 75.64(a)(10).
[¶ 56] The DEQ and Basin Electric point out that there are no limits, standards, or control requirements for carbon dioxide. The EPA requires only monitoring and reporting of carbon dioxide emissions. The DEQ and Basin Electric contend that monitoring and reporting requirements alone do not make carbon dioxide subject to regulation.
[¶ 57] The position taken by the DEQ and Basin Electric is fully consistent with the EPA's longstanding position. The definition quoted above of what pollutants are subject to regulation was promulgated in 2002. 67 Fed.Reg. 80,186 (Dec. 31, 2002). In the preamble to this regulation, the EPA also provided a list of all pollutants it considered subject to regulation. Id. at 80,240. Carbon dioxide was not on that list, even though the EPA had imposed carbon dioxide monitoring and reporting requirements in 1993. The EPA did not consider monitoring and reporting requirements, by themselves, sufficient to make carbon dioxide emissions subject to regulation under the federal Clean Air Act.
[¶ 58] According to the PRBRC, the EPA was forced to change this interpretation by the United States Supreme Court's decision in Massachusetts v. EPA, 549 U.S. 497, 127 S. Ct. 1438, 167 L. Ed. 2d 248 (2007). In that case, the Court said that greenhouse gases, including carbon dioxide, are "air pollutants" as defined in the Clean Air Act. It ruled that the EPA has the authority to regulate carbon dioxide emissions from new motor vehicles, but it did not hold that the EPA is required to do so. In fact, the case was remanded to allow the EPA to decide whether or not to regulate carbon dioxide emissions from new motor vehicles. Contrary to the PRBRC's assertion, the Court's ruling established that carbon dioxide is potentially subject to regulation, but not that it is subject to regulation. See Longleaf Energy Assoc. v. Friends of the Chattahoochee, Inc., 298 Ga.App. 753, 681 S.E.2d 203, 207 (2009) ("The United States Supreme Court decision in Massachusetts v. EPA does not mandate the Superior Court's ruling [that carbon dioxide is subject to regulation under the Clean Air Act].")
[¶ 59] Following the decision in Massachusetts, the Sierra Club challenged a PSD permit issued by the EPA to a coal-fired power plant in Utah. In re Deseret Power Elec. Coop., 2008 WL 5572891 (E.P.A. Nov. 13, 2008). As in the case before us now, the Sierra Club asserted that the EPA was required to apply BACT analysis and controls to the plant's carbon dioxide emissions. The EPA responded that it had historically interpreted the term "subject to regulation" to include only those air pollutants subject to statutory or regulatory emissions controls, not to pollutants such as carbon dioxide that are subject only to monitoring and reporting requirements.
[¶ 60] The Environmental Appeals Board rejected the Sierra Club's argument that the term "subject to regulation" was so clear and unambiguous as to require the EPA to include carbon dioxide emissions. Id. at 26. It also ruled, however, that the EPA's stated reason for not including a BACT limit for carbon dioxide in the permitthat it was bound by the historical interpretation of the term subject to regulationwas not sufficiently supported in the administrative record of the permitting decision. Id. at 37. On that basis, the Board remanded the permit to the EPA to reconsider whether carbon dioxide should be considered subject to regulation. Id. at 63. The Deseret decision, much like the Massachusetts decision, establishes only that carbon dioxide is potentially subject to regulation at some future time.
[¶ 61] Shortly after the Deseret decision, the EPA issued a memorandum reaffirming its historical interpretation that the term "subject to regulation" includes those pollutants for which a statute or regulation "requires actual control of emissions of that pollutant." Memorandum from Stephen L. Johnson, Administrator, EPA, to Regional Administrators (Dec. 18, 2008). The PRBRC tries to minimize the significance of this memorandum by pointing out that the EPA later granted a petition to reconsider. See Letter from Lisa P. Jackson, EPA Administrator, to David Bookbinder, Sierra Club (Feb. 17, 2009). In granting reconsideration, however, the EPA expressly refused to stay *826 the effectiveness of the interpretation set forth in the memorandum.
[¶ 62] On reconsideration, the EPA might change its interpretation, and begin to consider carbon dioxide subject to regulation. Such a change would not affect Dry Fork, however, because its air quality permit has already been issued. While the PRBRC has made a persuasive argument that carbon dioxide may be regulated in the future, it has not shown that carbon dioxide was subject to regulation when the Dry Fork permit was pending. We therefore agree with the DEQ and Basin Electric that the DEQ was not required to subject the Dry Fork Station's carbon dioxide emissions to BACT analysis and control. On this basis, we affirm the Council's decision to dismiss the PRBRC's claim on this issue.
CONCLUSION
[¶ 63] We affirm the Council's decision that the DEQ properly issued an air quality permit to Basin Electric for the Dry Fork Station.
NOTES
[1] All of the computer model runs used maxim allowable emissions from the Dry Fork Station. Because Dry Fork was not yet in operation, its actual emissions could not be documented.
[2] Exceedances of the 3-hour increment were also modeled. Because these were comparable in number and degree to the exceedances of the 24-hour increment, and because the regulatory framework is the same for the 3-hour and 24-hour increments, our analysis is the same for both. To avoid confusion and duplication, our discussion will focus on the 24-hour increment.
[3] To be precise, in Prairie State the Significant Impact Levels were used to determine compliance with the National Ambient Air Quality Standards, not to determine whether emissions from the proposed source would cause or contribute to increment exceedances. None of the parties have addressed this distinction or suggested that it makes a difference in this case.
[4] While this manual was never finalized, and has remained in draft form for twenty years, it is widely recognized as an authoritative source on air quality regulation. It has been cited as persuasive by the parties on both sides of this case.
[5] The record includes technical definitions of the critical point of water, but for our purposes, it is sufficient to note simply that supercritical boilers operate at higher temperatures and pressures than subcritical boilers.
[6] The key difference is that a pulverized coal-fired power plant is fueled by pulverized coal, while a power plant employing IGCC technology is fueled by synthetic gas converted from coal. See Prairie State, slip op. at 35.
[7] We will focus on carbon dioxide because it is recognized as "the most important species" of greenhouse gases. Massachusetts v. EPA, 549 U.S. 497, 505, 127 S. Ct. 1438, 1446, 167 L. Ed. 2d 248 (2007). Our analysis pertaining to carbon dioxide also applies to other greenhouse gas emissions.