Reply Brief — Kinder Morgan, Inc., et al., Applicants v. Environmental Protection Agency, et al.

Supreme Court briefNov 1, 2023

Ask Donna

What actually matters in this document.

Text

No. 23A350

IN THE

Supreme Court of the United States

KINDER MORGAN, INC.; ENBRIDGE (U.S.) INC.;

TRANSCANADA PIPELINE USA LTD.;

INTERSTATE NATURAL GAS ASSOCIATION OF AMERICA;

AMERICAN PETROLEUM INSTITUTE

Applicants,

UNITED STATES ENVIRONMENTAL PROTECTION AGENCY AND

MICHAEL S. REGAN, ADMINISTRATOR,

UNITED STATES ENVIRONMENTAL PROTECTION AGENCY, ET AL.,

Respondents.

REPLY BRIEF IN SUPPORT OF EMERGENCY APPLICATION FOR STAY OF

FINAL AGENCY ACTION DURING PENDENCY OF PETITIONS FOR REVIEW

Ana Maria Gutiérrez

Michael D. Miller

WOMBLE BOND DICKINSON (US) LLP

2001 K St NW

Washington, DC 20006

CATHERINE E. STETSON

Counsel of Record

HOGAN LOVELLS US LLP

555 Thirteenth Street, N.W.

Washington, DC 20004

(202) 637-5600

cate.stetson@hoganlovells.com

Counsel for Kinder Morgan, Inc.

Additional counsel listed on signature block

November 1, 2023

Eric D. McArthur

SIDLEY AUSTIN LLP

1501 K Street, NW

Washington, DC 20005

Counsel for the Interstate

Natural Gas Association

of America and American

Petroleum Institute

Brittany M. Pemberton

BRACEWELL LLP

2001 M Street, NW,

Suite 900

Washington, DC 20036

Counsel for TransCanada

PipeLine USA Ltd.

Laura K. McAfee

BEVERIDGE & DIAMOND, PC

2091 North Charles Street,

Suite 2200

Baltimore, MD 21201

Counsel for Enbridge (U.S.) Inc.

TABLE OF CONTENTS

Page

TABLE OF AUTHORITIES ................................................................................................................................... ii

INTRODUCTION ..................................................................................................................................................... 1

ARGUMENT ............................................................................................................................................................. 3

APPLICANTS ARE LIKELY TO SUCCEED ON THE MERITS ....................................... 5

The Rule Is Irretrievably Broken, Given The Stays Of EPA’s

State Plan Disapprovals ........................................................................................... 5

EPA Required Emission Reductions At Costs Vastly In Excess

Of Its Cost Threshold ................................................................................................ 7

EPA’s Response Demonstrates The Paucity Of Its Reliability

Analysis ........................................................................................................................11

Hundreds Of 1,000-Horsepower Engines Emit Less Than

EPA’s Significant-Contribution Threshold Of 100 Tons Per

Year Of NOx ................................................................................................................15

EPA’s Purported Compliance Flexibilities Cannot Remedy Its

Flawed Rule ................................................................................................................17

II.

PIPELINE APPLICANTS FACE SUBSTANTIAL, UNRECOVERABLE

HARM ..........................................................................................................................................20

III.

THE BALANCE OF HARMS AND THE PUBLIC INTEREST WEIGH

HEAVILY IN FAVOR OF A STAY .........................................................................................23

CONCLUSION ........................................................................................................................................................24

TABLE OF AUTHORITIES

Page(s)

Cases:

Ala. Ass’n of Realtors v. HHS,

141 S. Ct. 2485 (2021) .......................................................................................................................... 4, 23

Ass’n of Oil Pipe Lines v. FERC,

281 F.3d 239 (D.C. Cir. 2002) ..................................................................................................................20

Balt. Gas & Elec. Co. v. FERC,

954 F.3d 279 (D.C. Cir. 2020) .......................................................................................................... 16, 17

Balt. Gas & Elec. Co. v. NRDC,

462 U.S. 87 (1983) .......................................................................................................................................15

Burlington Truck Lines, Inc. v. United States,

371 U.S. 156 (1962) ....................................................................................................................................13

CSL Plasma Inc. v. U.S. Customs & Border Prot.,

628 F. Supp. 3d 243 (D.D.C. 2022).........................................................................................................23

Does 1–3 v. Mills,

142 S. Ct. 17 (2021)....................................................................................................................................... 4

EPA v. EME Homer City Generation, L.P.,

572 U.S. 489 (2014) .................................................................................................. 4, 7, 8, 9, 10, 11, 16

FCC v. Fox Television Stations, Inc.,

556 U.S. 502 (2009) ...................................................................................................................................... 4

FCC v. Prometheus Radio Project,

141 S. Ct. 1150 (2021) ................................................................................................................................. 4

Hollingsworth v. Perry,

558 U.S. 183 (2010) ....................................................................................................................................20

In re NTE Conn., LLC,

26 F. 4th 980 (D.C. Cir. 2022) ..................................................................................................................23

MD/DC/DE Broadcasters Ass’n v. FCC,

236 F.3d 13 (D.C. Cir.), aff’d on reh’g, 253 F.3d 732 (D.C. Cir. 2001) ......................................... 7

Memphis Light, Gas & Water Div. v. Craft,

436 U.S. 1 (1978)..........................................................................................................................................23

Michigan v. EPA,

213 F.3d 663 (D.C. Cir. 2000) ................................................................................................................2, 7

Michigan v. EPA,

576 U.S. 743 (2015) ............................................................................................................................... 4, 11

Minnesota v. Mille Lacs Band of Chippewa Indians,

526 U.S. 172 (1999) ...................................................................................................................................... 7

Motor Vehicle Mfrs. Ass’n v. State Farm Mut. Auto Ins. Co.,

463 U.S. 29 (1983) .......................................................................................................................................14

Nat’l Fed. of Indep. Bus. v. Dep’t of Labor,

595 U.S. 109 (2022) ...................................................................................................................................... 3

Nken v. Holder,

556 U.S. 418 (2009) ............................................................................................................................... 3, 23

Scripps-Howard Radio v. FCC,

316 U.S. 4, 10 (1942) ............................................................................................................................. 3, 24

Sierra Club v. Ga. Power Co.,

180 F.3d 1309 (11th Cir. 1999)..............................................................................................................23

Small Refiner Lead Phase-Down Task Force v. EPA,

705 F.2d 506 (D.C. Cir. 1983) .................................................................................................................... 5

U.S. Telecom Ass’n v. FCC,

359 F.3d 554 (D.C. Cir. 2004) ..................................................................................................................20

Utility Air Regulatory Grp. v. EPA,

573 U.S. 302 (2014) ...................................................................................................................................... 4

West Virginia v. EPA,

142 S. Ct. 2587 (2022) ................................................................................................................................. 4

Window Covering Mfrs. Ass’n v. Consumer Prod. Safety Comm’n,

82 F. 4th 1273 (D.C. Cir. 2023) .......................................................................................................... 6, 11

Winter v. Natural Resources Defense Council,

555 U.S. 7 (2008)............................................................................................................................................ 4

Wis. Gas Co. v. FERC,

758 F.2d 669 (D.C. Cir. 1985) ..................................................................................................................23

Wisconsin v. EPA,

938 F.3d 303, 311 (D.C. Cir. 2019) .......................................................................................................... 8

Statutes & Other Authorities:

5 U.S.C. § 705 ........................................................................................................................................................... 3

42 U.S.C. § 7607(b)(1) ......................................................................................................................................... 6

40 C.F.R. § 52.40(d)(3) .............................................................................................................................. 19, 22

40 C.F.R. § 52.40(e)(2)(i)(A) ...........................................................................................................................19

iii

40 C.F.R. § 52.40(e)(2)(B) ................................................................................................................................19

40 C.F.R. § 52.41(a).............................................................................................................................................20

40 C.F.R. § 52.41(d) ............................................................................................................................................18

5 Code Colo. Regs. § 1001-30:B.I.D.5.b.(v)(B) ..........................................................................................22

5 Code Colo. Regs. § 1001-30:B.I.D.5.c ........................................................................................................22

81 Fed. Reg. 74,504 (Oct. 26, 2016) ............................................................................................................... 8

87 Fed. Reg. 20,036 (Apr. 6, 2022) ...................................................................................................8, 10, 15

Commissioner James Danly, Response to Questions for the Record for June 13, 2023

House Energy & Commerce Oversight Hearing, https://perma.cc/C757-3DD3

(637a–638a) ..................................................................................................................................................15

EPA, Non-EGU Facilities and Units.xlsx (Mar. 2023), https://perma.cc/UDK9-LRKU ..............16

EPA, Pipeline Natural Gas - Engines Analysis data, (June 4, 2023),

https://perma.cc/4HGB-JDQH ...............................................................................................................18

EPA, Transport Proposal, Screening Assessment Non-EGU Facility and Emissions

Unit Lists - 03-18-2022.xlsx, https://www.regulations.gov/document/EPA-HQOAR-2021-0668-0191 ................................................................................................................................. 9

Federal ‘‘Good Neighbor Plan’’ for the 2015 Ozone National Ambient

Air Quality Standards, 88 Fed. Reg. 36,654 (June 5, 2023) ...................... 6, 8, 9, 10, 13, 19, 20

NOx Emissions Control Technology Installation Report Timing for Non‐EGU Sources,

Final Report, EPAHQ-OAR-2021-0668-1077, ES-8 (Mar. 14, 2023) .......... 12, 13, 14, 18, 21

Order, Arkansas v. EPA, No. 23-1320, ECF 5280996 (8th Cir. May 25, 2023) ............................... 6

Order, Missouri v. EPA, No. 23-1719, ECF 5281126 (8th Cir. May 26, 2023) ................................. 6

Order, Texas v. EPA, No. 23-60069, ECF 269-1 (5th Cir. May 1, 2023) ............................................. 6

Response to Public Comments on Proposed Rule (Mar. 2023) (RTC),

https://perma.cc/6DY8-Y5G4 ................................................................................................................13

Sup. Ct. R. 10(c) ...................................................................................................................................................... 4

INTRODUCTION

EPA’s opposition to the Pipeline Applicants’ stay request is more notable for what it

does not say. The agency’s opposition fails to answer three fundamental issues: (1) the

agency’s unexplained and patently unlawful divergence from its approach to defining the

emissions “amounts” authorized to be eliminated by the ozone transport provision; (2) the

utter infeasibility of its compliance timeline; and (3) the resulting precariousness of natural

gas service to homes and businesses and irreparable harm to pipeline operators. The stay

application should be granted.

Applicants have demonstrated a likelihood of success on the merits. EPA contends

that Pipeline Applicants collaterally attack state plan disapprovals and suggests this Court is

“without authority” to consider the impact on the Rule of the subsequent judicial stays of

many of those state plan disapprovals. EPA Resp. 20. That is not so. The stays of EPA’s state

plan disapprovals issued by multiple circuits have a substantive and profound impact on the

agency’s federal plan, and this Court can certainly consider that in deciding whether the portions of the federal plan applicable to pipeline engines should be stayed.

EPA also maintains that it “appropriately” applied the cost threshold to determine the

“amounts” of emission reductions required for pipeline engines. Id. at 2. That is also wrong.

The agency realized in the Rule that its cost threshold did not reflect the full range of costs

for pipeline engines—yet failed to correct course, instead relying on that rejected cost

threshold to apply stringent emissions requirements to thousands of pipeline engines, many

of which would require much higher costs-per-ton for controls. Garbage in, garbage out.

EPA argues its 1,000-horsepower applicability criterion was a reasonable “starting

point” for identifying pipeline engines emitting 100 tons per year—the “largest emitters.” Id.

at 32. Yet, when this “starting point” turned out to be wrong—less than 10 percent of the

engines covered by the Rule emit more than 100 tons per year—EPA again failed to coursecorrect, sweeping into the Rule thousands of pipeline engines whose individual contributions to downwind nonattainment are, on the agency’s own reckoning, insignificant.

And perhaps of utmost importance at this stage, EPA ignored natural gas pipeline reliability in establishing the May 2026 compliance date—a date that requires operators to

start retrofitting pipeline engines now in order to have any chance of coming close to meeting

the less-than-three-year timeline. The only report EPA relies on for the compliance date’s

feasibility expressly disavows having evaluated natural gas pipeline reliability and presents

data that a less-than-three-year timeline is not in fact feasible for the pipeline industry. The

too-soon compliance date leads directly to the irreparable injury pipeline companies—and

downstream customers—will face while litigation is pending. It is simply not possible to

retrofit pipeline engines on EPA’s timeframe without starting while the litigation is pending

and without risking service interruptions, leaving pipeline companies only two choices:

(1) be out of compliance with EPA’s Rule; or (2) restrict transportation of natural gas, at

grave costs to the public.

EPA has engaged in numerous rulemakings pursuant to the ozone transport provision. Through this Rule, EPA is directly regulating pipeline engines under the ozone

transport provision for the first time with per-engine emissions-rate limits.1 Such a first-of-

EPA states that in 1998, the agency “limited the emissions of nitrogen oxides [(NOx)]—a

precursor to ozone—for both power plans and non-electric generating units (non-EGUs),

including pipeline engines . . .” EPA Resp. 6. Notably, however, the 1998 rule merely established state-wide budgets for affected states, and emissions reductions analyzed from pipeline engines were a factor going into the budgets. See Michigan v. EPA, 213 F.3d 663, 681–

82 (D.C. Cir. 2000) (discussing how EPA calculated budgets for states based on what “highly

its-kind action, considering the liberties EPA took in applying and interpreting the Clean Air

Act, should not be allowed to become a fait accompli before any court has reviewed it.

This Court should restore the status quo ante for pipeline operators while the D.C.

Circuit—and potentially this Court—consider the Rule’s legality. See Scripps-Howard Radio

v. FCC, 316 U.S. 4, 10 (1942) (“If the administrative agency has committed errors of law . . .

judicial review would be an idle ceremony if the situation were irreparably changed before

the correction could be made.”).

ARGUMENT

First things first: A heightened standard does not apply to Applicants’ stay request.

EPA Resp. 16. Applicants seek a stay of the Rule’s effectiveness pending judicial review, as

expressly authorized by 5 U.S.C. § 705. The standard for granting a stay—not an injunction—thus applies here. In Nken v. Holder, 556 U.S. 418 (2009), this Court held that the

standard for a stay applied in putting on hold a Board of Immigration Appeals’ order pending

judicial review because a stay “temporarily suspend[s] the source of the authority to act . . .

not by directing an actor’s conduct” as an injunction would. Id. at 428–29; see also Nat’l Fed.

of Indep. Bus. v. Dep’t of Labor, 595 U.S. 109, 120 (2022) (per curiam) (staying a Department

of Labor rule). Here, Pipeline Applicants similarly request “the temporary setting aside of”

EPA’s Rule, not “a coercive order against the government.” Nken, 556 U.S. at 429.

cost-effective controls” could be implemented). Regarding pipeline engines, the D.C. Circuit

reversed and remanded, holding that EPA did not provide adequate notice of its assumed

level of emissions reductions to be achieved from the sector when calculating the budgets.

Id. at 693.

Even if the standard for granting an injunction applied here—if the standard even

differs materially—this Court should still stay the Rule provisions applicable to pipeline engines. Pipeline Applicants have made a “clear showing” that the Rule’s provisions for Pipeline Engines are unlawful and that Applicants will suffer irreparable harm absent a stay.

Winter v. Natural Resources Defense Council, 555 U.S. 7, 22 (2008).

EPA also contends that this Court weighs the likelihood of granting certiorari when

deciding stay applications. EPA Resp. 41–43 (citing Does 1–3 v. Mills, 142 S. Ct. 17, 18 (2021)

(Barrett, J., concurring). However, this Court routinely grants stay applications without discussing the probability of granting certiorari. See, e.g., Ala. Ass’n of Realtors v. HHS, 141 S. Ct.

2485 (2021) (per curiam). In any case, given their national import and federalism impacts,

see S. Ct. R., 10(c), this Court frequently grants certiorari—and reverses—lower court rulings

on the lawfulness of EPA rules promulgated under the Clean Air Act. See EPA v. EME Homer

City Generation, L.P., 572 U.S. 489 (2014); Utility Air Regulatory Grp. v. EPA, 573 U.S. 302

(2014); Michigan v. EPA, 576 U.S. 743 (2015); West Virginia v. EPA, 142 S. Ct. 2587 (2022).

EPA argues that Pipeline Applicants’ motion presents mere “fact-dependent arbitrary-andcapricious challenges.” EPA Resp. 43. That is again not so; Pipeline Applicants challenge the

Rule in part as “run[ning] afoul of the Clean Air Act—and this Court’s prior precedent[.]”

Emergency Application 14. Regardless, this Court regularly grants and decides cases based

on the Administrative Procedure Act’s arbitrary-and-capricious standard as well. E.g., FCC v.

Prometheus Radio Project, 141 S. Ct. 1150 (2021); FCC v. Fox Television Stations, Inc., 556 U.S.

502 (2009).

Because of the Rule’s substantial legal flaws, there is also a fair prospect of reversal.

EPA was clearly in a hurry to promulgate the Rule. But when Pipeline Applicants identified

grave concerns with EPA’s proposed rule, the agency plowed ahead anyway. As a result, the

Rule for pipeline engines is unlawful multiple times over, and it causes substantial irreparable harm and detriment to the public interest.

APPLICANTS ARE LIKELY TO SUCCEED ON THE MERITS.

The Rule Is Irretrievably Broken, Given The Stays Of EPA’s State Plan Disapprovals.

The Rule EPA originally proposed, received comment on, and promulgated looks

nothing like the Rule now. Currently, 78 percent of the Rule’s estimated collective emissions

reductions would be realized in states where a judicial stay of the state plan disapproval is

in effect. Emergency Application 12.

EPA paints Applicants’ argument as a collateral attack on EPA’s individual state plan

disapprovals. EPA Resp. 21. That is wrong. The state plan disapprovals are subject to their

own challenges; Applicants’ point here is that the legal predicate supporting the Rule has

dissolved in a majority of states the Rule would cover. As a result, the Rule no longer accomplishes what EPA states it was designed to do, and yet EPA refuses to put it on hold until the

legality of EPA’s actions disapproving state plans are resolved across the courts of appeals.

The agency also asserts that Applicants should have filed a petition for reconsideration of the Rule. EPA Resp. 19. But “key assumptions” underlying EPA’s rulemakings do not

need to be challenged through requests for reconsideration. Small Refiner Lead Phase-Down

Task Force v. EPA, 705 F.2d 506, 535 (D.C. Cir. 1983) (“[A]ggregate analysis is a vital assumption underlying the [EPA’s] model. Thus, EPA must justify that assumption even if no one

objects to it during the comment period . . . .”). The number of states—and their respective

emission reductions—is one of the most critical assumptions underlying the Rule. From a

practical standpoint, requiring reconsideration would make challengers chase a moving target. Pipeline Applicants challenge the Rule because EPA failed to conduct the most basic

analyses or provide record support for its conclusions—fundamental tenets of administrative law. Requiring operators to file a new petition for reconsideration after every judicial

stay would create an endless reconsideration loop. In other words, it would be an exercise

in futility.

EPA also argues that the Rule “was finalized” on March 15, 2023, prior to any of EPA’s

state plan disapprovals being stayed, implying that it could do nothing to correct course. See

EPA Resp. 20 (citing press release). But EPA did not publish the Rule in the Federal Register

until June 5, 2023. See 42 U.S.C. § 7607(b)(1) (contemplating that EPA’s rules will be published and final via the Federal Register); Federal ‘‘Good Neighbor Plan’’ for the 2015 Ozone

National Ambient Air Quality Standards, 88 Fed. Reg. 36,654 (June 5, 2023). By that time,

four state plan disapprovals had already been stayed by the Fifth and Eighth Circuits, and the

writing was on the wall for several others.2 EPA’s reliance on a press release is a thinly veiled

attempt to avoid the import of the state plan disapproval stays that were entered before EPA

even published the Rule. “[T]he culmination of [EPA’s] . . . process was the Final Rule, and

[pipeline companies] had no new compliance obligations until the Final Rule was promulgated.” Window Covering Mfrs. Ass’n v. Consumer Prod. Safety Comm’n, 82 F. 4th 1273, 1292

(D.C. Cir. 2023).

See Order, Texas v. EPA, No. 23-60069, ECF 269-1 (5th Cir. May 1, 2023) (Texas and Louisiana plans); Order, Arkansas v. EPA, No. 23-1320, ECF 5280996 (8th Cir. May 25, 2023)

(Arkansas plan); Order, Missouri v. EPA, No. 23-1719, ECF 5281126 (8th Cir. May 26, 2023)

(Missouri plan).

Next, EPA claims that the Rule is severable and is therefore implementable—even

without 78 percent of projected emissions reductions. Id. at 24–28. But key aspects of the

Rule are based on the inclusion of those 12 now-stayed states. See Emergency Application

12–13 (noting EPA’s intention to apply the Rule on a “national scale” and relying on the “collective” contribution of all the states). Because the Rule “embodies . . . one coherent policy,”

Minnesota v. Mille Lacs Band of Chippewa Indians, 526 U.S. 172, 194 (1999), its applicability

to individual states is inseverable. That is especially true here; when 78 percent of the estimated emissions reductions are no longer attainable, that “severely distort[s EPA’s] program

and produce[s] a rule strikingly different from any the [agency] has ever considered or promulgated.” MD/DC/DE Broadcasters Ass’n v. FCC, 236 F.3d 13, 23 (D.C. Cir.), aff’d on reh’g, 253

F.3d 732, 734 (D.C. Cir. 2001) (en banc).

EPA Required Emission Reductions At Costs Vastly In Excess Of Its Cost

Threshold.

EPA both violated the Clean Air Act and departed from its past practice by failing to

identify the emissions “amounts” that natural gas pipeline engines in upwind States contribute to downwind States’ nonattainment. In EME Homer, this Court upheld EPA’s methodology for identifying and allocating the emissions “amounts” to be eliminated among regulated

States. See EME Homer, 572 U.S. at 492. The methodology: “As EPA interprets the statute,

upwind emissions rank as ‘amounts [that] . . . contribute significantly to nonattainment’ if

they . . . can be eliminated under the cost threshold set by the Agency.” Id. at 518 (emphasis

added). EPA states that it applied this same EME Homer “regulatory framework” to the Rule.

EPA Resp. 8 (citing EME Homer, 572 U.S. at 524; Michigan, 213 F.3d at 674‒679). Applying

this analysis at the proposal stage, EPA identified $7,500/ton as the marginal cost threshold—a threshold below which EPA anticipated emissions could be reasonably eliminated.

See 87 Fed. Reg. 20,036, 20,083, 20,096 (Apr. 6, 2022). In the final Rule, however, EPA concluded that that cost threshold was “not reflect[ive of] the full range of cost-effectiveness

values that are likely present” for non-EGUs. 88 Fed. Reg. at 36,740. But rather than reassessing the EME Homer methodology, EPA kept the same control technologies and associated emissions rate thresholds. That was a fundamental mistake.

In its response, EPA purports to have applied its “longstanding” approach and arrived

at an “average cost-per-ton” of $4,981 for available pipeline controls. EPA Resp. 29‒30. To

arrive at $4,981/ton, EPA applied a simple average of outdated cost data (from the proposed

rule) against the number of pipeline engines EPA estimates will be subject to the Rule. See

88 Fed. Reg. at 36,739. Contrary to EPA’s assertion, an “average representative cost[ ]” does

not equate to a marginal cost threshold and is not consistent with EPA’s prior approach in

ozone transport rules. EPA Resp. 30. The EME Homer methodology is no simple averaging

exercise. Rather, as Applicants explained in their Application, the agency “measure[s] the

emissions reductions that would occur at various levels of cost controls” and runs a model

at various control levels expressed in costs-per-ton. Wisconsin v. EPA, 938 F.3d 303, 311

(D.C. Cir. 2019) (citing 81 Fed. Reg. 74,504, 74,540–41 (Oct. 26, 2016)). For each cost-control

level, EPA then “estimate[s] the air quality improvements that would occur” downwind and

then compares the levels with the resulting emission reductions to find “the cost-control

level” that “represent[s] the point at which upwind ‘NOx reduction potential and corresponding downwind ozone air quality improvements are maximized with respect to marginal

cost.’” Id. (quoting 81 Fed. Reg. at 74,550). EPA did nothing of the sort here.

Setting aside the fact that a simple average in no way comports with the EME Homer

methodology, it’s easy for EPA to arrive at a cost-per-ton of $4,981/ton when it first throws

out costs above $7,500/ton and only averages costs less than $7,500/ton.3 88 Fed. Reg. at

36,740.4 EPA now calls this simple averaging of data an “updated analysis.” EPA Resp. 12‒

13. In reality it’s no “analysis” at all.

The Applicants explained to EPA that its estimated average costs are far below the

costs that industry commenters—armed with actual, concrete data rather than stale estimates—assessed. See TC Energy Comment 5 (noting total costs of Rule of $900 million for

engines that operate infrequently) (603a); Kinder Morgan Comment 21–26 (noting costs

above $100,000/ton and even above $684,169/ton) (543a–548a). EPA paid no heed. Even

taking EPA’s average cost at face value, moreover, it is irrational for EPA to assume that because the average engine can comply with the Rule at a cost below the threshold, all engines

can—particularly when the record shows that the cost of controls to meet the emissions limits is much higher for many engines. Emergency Application 14.

EPA also argues that the average cost of emissions controls for pipeline engines “compares favorably” with the average assessed for power plants. EPA Resp. 13, 30. That is not

the test. The EME Homer methodology does not ask whether the average cost is “comparable” to the costs for other regulated entities. Even if it were the test, EPA itself acknowledges

Recall that at the draft rule stage, EPA thought $7,500/ton was the marginal cost threshold representative of the control costs across non-EGUs industries. On this basis, at the

draft rule stage, EPA’s cost-per-ton data points did not exceed $7,500 for pipeline engines.

That means the underlying dataset EPA used to arrive at the simple average of $4,981 for

pipeline engines considered costs of $585/ton at the low end and $5,457/ton at the high

end. EPA, Transport Proposal, Screening Assessment Non-EGU Facility and Emissions Unit

Lists - 03-18-2022.xlsx, https://www.regulations.gov/document/EPA-HQ-OAR-2021-06680191.

This now-central value of $4,981/ton only shows up once in the Rule preamble. Applicants were required to scour the record to otherwise discern EPA’s consideration of

$4,981/ton as applied to pipeline engines. In comparison, the marginal cost-per-ton

threshold of $7,500 shows up not fewer than thirteen times in the Rule preamble alone.

that there is “far greater variation” in emissions levels and control technologies for pipeline

engines than for power plants, which are “relatively homogenous,” 88 Fed. Reg. at 36,720,

and where power plants may trade credits with other sources.

EPA also tries to recast its proposed $7,500/ton threshold as a mere screening

threshold, or “starting point,” not intended to determine the emissions that “contribute significantly.” EPA Resp. 12, 31. The threshold may have been included in a document called a

“Screening Assessment,” but nowhere in the proposed or final Rule does EPA characterize

the cost value as a screening threshold or a “starting point.” In fact, consistent with the result

EPA is supposed to reach when applying it’s longstanding framework, EPA’s proposal states

it “believes that four stroke lean burn engines subject to this proposed [rule] can achieve the

emissions limit of 1.5 g/hp-hr with the installation and operation” of specified “control technologies at the marginal cost threshold of $7,500 per ton.” 87 Fed. Reg. at 20,142–43 (emphasis added); see also id. at 20,143 (similar statement for two stroke lean burn engines).

Finally, EPA now contends its cost threshold—whatever the mark, wherever it came

from—is not a cost cap, but a general “comparison of different available controls in different

industries, which allows EPA to determine which controls would optimally yield significant

reductions with down-wind benefits.” EPA Resp. 30. There are several replies to this alone.

First, EPA’s litigation-driven attempt to recharacterize the purpose of its judicially approved

methodology is out of place. EME Homer, 572 U.S. at 521 (EPA may not require emissions

reductions “at odds with the . . . threshold the Agency has set”).

Second, in prior ozone

transport rules where EPA found a cost threshold unrepresentative, it did not regulate that

source. Emergency Application 15. Finally, if EPA itself determined $7,500/ton is not representative of a “range” of costs, 88 Fed. Reg. at 36,740, there is no rational basis for a lower

threshold of $4,981/ton to be considered “cost-effective,” EPA Resp. 29. The result, as applied: because EPA did not select a Step 3 cost threshold, the Rule requires all affected pipeline engines comply regardless of the cost to that unit.

In the end, EPA argues that the case-by-case alternative emissions limits—reserved

for entities that can establish extreme economic hardship—solve any infirmities in the Rule.

EPA Resp. 30-31. Again, that discretionary, one-off alternative does not excuse the Rule’s

fundamental flaws. See infra at 17–20.

When properly executed, this Court has determined that “using costs in the Transport

Rule calculus is an efficient and equitable solution to the allocation problem.” EME Homer,

572 U.S. at 504, 519. EPA states it “applie[d] that same regulatory framework” to the Rule.

EPA Resp. 8. It most certainly did not.

EPA’s Response Demonstrates The Paucity Of Its Reliability Analysis.

In developing the compliance date for pipeline engines, EPA “fail[ed] to consider” natural gas reliability impacts, no doubt “an important aspect of the problem.” Michigan, 576

U.S. at 752. Insofar as EPA did analyze reliability, its conclusion was irrational.

EPA contends that it considered the reliability impacts of the May 2026 compliance

date for pipeline engines, and determined there would be no issue. EPA Resp. 34–37.5 However, the record material EPA cites demonstrates just how little supporting analysis exists

EPA also argues that non-EGU sources have “at least three years before compliance obligations begin.” Id. at 34. That is wrong. The Rule was published in the Federal Register on

June 5, 2023 and did not take effect until August 4, 2023. Non-EGU compliance obligations

start May 1, 2026, leaving less than three years before compliance obligations begin. The

pre-publication draft did not begin the clock. See Window Covering Mfrs. Ass’n, 82 F. 4th at

1292 (published rule is culmination of rulemaking process).

on this critical question. The only record evidence EPA cites for support on a purported reliability finding, see id. at 35–36, is a few sentences on average pipeline engine capacity utilization and generalized discussions of time required for “equipment installation” in the NOx

Emissions Control Technology Installation Report Timing for Non‐EGU Sources, Final Report,

EPAHQ-OAR-2021-0668-1077, ES-8, 8, 22, 32 (Mar. 14, 2023) (Timing Report) (385a,

394a).6

But the Timing Report expressly disclaims any determination about reliability; it instead states that its authors “were not able to complete an evaluation of” reliability concerns.

Timing Report ES-8 (385a). EPA largely concedes this point in its response, while attempting

to frame the Timing Report’s disclaimer as only relating to any Federal Energy Regulatory

Commission certification process required for pipeline engine retrofits. EPA Resp. 35. But

entirely aside from the process FERC requires, Pipeline Applicants also explained to EPA that

the compliance date had not considered reliability issues and that engines must be addressed

“one-by-one” at a compressor station. Kinder Morgan Comment 28–29 (550a–551a); INGAA

Comment 41–42 (505a–506a). To the extent the Timing Report only focuses on the FERC

process, as EPA suggests, then it failed entirely to consider comments on the need to phase

retrofits.

EPA’s reliance on superficial average capacity utilization also assumes that engine capacity can be pooled and borrowed widely across the industry. EPA Resp. 35. But such borrowing is physically impossible; pipeline capacity is location- and pipeline-specific. Emer-

The Timing Report was also only made available in parallel with the early (March 2023)

pre-publication of the Rule.

gency Application 26. And the Timing Report’s assessment of capacity utilization and coordination produced the qualifier-ridden statement that reliability concerns “may not present

a substantial basis for assuming much if any delay in control installation timing on this basis.”

Timing Report ES-8 (emphasis added). EPA leaps from that wobbly assessment in the record

to claiming in its response that it “reasonably determined that individual unit outages may

be staggered and need not interrupt natural-gas supply.” EPA Resp. 35; see also id. at 4 (stating that EPA “analysis indicates that near-term capital expenditures to achieve compliance

need not be extensive and will not endanger natural-gas supply”). But EPA never made such

a determination, not in any other report or in the Rule, and its attempt to clean up the record

now is prohibited “post hoc rationalization.” Burlington Truck Lines, Inc. v. United States, 371

U.S. 156, 168 (1962).7

EPA’s failure to consider natural gas pipeline reliability is all the more stark when

compared to the analysis the agency performed—and now touts, see EPA Resp. 47 (citing 88

Fed. Reg. at 36,770– 75)—for power plants. In the published Rule, EPA devotes six pages to

electric sector reliability, see 88 Fed. Reg. at 36,770–75, and EPA even prepared a stand-alone

Resource Adequacy and Reliability Analysis Final Rule Technical Support Document for the

electric sector, see id. at 36,772 n.301. Pipeline Applicants do not maintain that EPA needed

While EPA cites 88 Fed. Reg. at 36,759–60 and its Response to Public Comments on Proposed Rule (Mar. 2023) (RTC), https://perma.cc/6DY8-Y5G4, neither source shows any

reasoned determination about natural gas reliability. See 88 Fed. Reg. at 36,759–60 (discussing permitting delays, supply chain delays, and purported flexibilities in the Rule, but

not discussing pipeline reliability and the capacity to serve demand during peak times);

RTC 877 (stating—without providing rationale—that “the rule does not affect the ability of

the natural gas industry or FERC to ensure the delivery of adequate natural gas supplies to

consumers.”).

to follow that precise script here, but the lack of an administrative record for pipeline engines, compared to that thorough effort, is glaring. See Motor Vehicle Mfrs. Ass’n v. State Farm

Mut. Auto Ins. Co., 463 U.S. 29, 43 (1983) (agency decision is arbitrary and capricious if

agency “entirely failed to consider an important aspect of the problem”).

Even if the Timing Report reached the definitive conclusion that EPA now says it did,

it would still be irrational. Id. (agency must articulate “a rational connection between the

facts found and the choice made”). As Pipeline Applicants highlighted, average capacity level

says little to nothing about available capacity at peak demand, especially given the undisputed physical limitations in substituting one engine’s capacity with capacity somewhere

else on the pipeline and the inability to substitute capacity across different pipelines. Emergency Application 19. 8 The discussion on the “equipment installation” times of a single

month or between three and seven months for pipeline engines in the Timing Report (at 32

(418a)) also does not provide a reasoned basis for finding that pipeline reliability would not

be impacted. Peak demand for natural gas consumption extends from May to September

(summer) and November to March (winter), and pipelines must maximize capacity during

these periods to meet customer demands. Even with a one-month installation timeframe,

then, accomplishing more than a single engine retrofit per compressor station in a shoulder

season would be challenging. And with the longer installation timeframe that is required for

some types of controls, the outage is bound to run up against a peak season. When pipeline

companies are required to retrofit thousands of engines with only five lower-demand seasons between now and May 1, 2026, the only way for pipelines to come close to achieving

Moreover, only two vendors are available for pipeline engine retrofits. Id. at 17 n.9.

that compliance date is taking more units offline for retrofit projects, including in peak-demand months. Under the circumstances, EPA’s case-by-case compliance flexibilities, EPA

Resp. 45‒46, cannot save its deeply flawed Rule. See infra at 17–20.

In the end, EPA asks this Court to defer to its “technical, predictive judgment” about

natural gas reliability. EPA Resp. 37. But the very case cited by EPA states that such deference is afforded when the agency acts “within its area of special expertise.” Balt. Gas & Elec.

Co. v. NRDC, 462 U.S. 87, 103 (1983). EPA is not an expert in the transportation of natural

gas and pipeline reliability, as illustrated by its altogether lacking record and the call-out by

FERC Commissioner Danly, who noted that “EPA [did] not ever consider the impacts that the

timeline for compliance for non-EGUs would have on electric reliability or residential uses.”

Commissioner James Danly, Response to Questions for the Record for June 13, 2023 House Energy & Commerce Oversight Hearing 23–24, https://perma.cc/C757-3DD3 (637a–638a).

Hundreds Of 1,000-Horsepower Engines Emit Less Than EPA’s Significant-Contribution Threshold Of 100 Tons Per Year Of NOx.

At the proposal stage, EPA proposed to regulate all pipeline engines with a design

capacity of 1,000 horsepower or greater, finding that this criterion “reasonably approximates the selection of 100 [tons per year] used within the non-EGU screening assessment.”

87 Fed. Reg. at 20,142. Even after commenters demonstrated that EPA’s projection of the

number of pipeline engines this would sweep into the Rule was off by a factor of ten, and that

the vast majority of them emit less than 100 tons per year, EPA refused to adjust course.

Emergency Application 21–22.

EPA now claims that its 100-tons-per-year threshold was yet another “starting point”

and “did not reflect an EPA determination that particular pipeline engines make a significant

contribution to downwind air pollution.” EPA Resp. 32. But EPA does not dispute that it

excluded many sources from regulation in this Rule because they emit less than 100 tons per

year, which necessarily means that EPA found those sources do not “contribute significantly.” See Emergency Application 22. EPA thus exceeded its statutory authority by requiring emissions reductions “at odds with the [significant-contribution] threshold the Agency

has set.” EME Homer, L.P., 572 U.S. at 521. And, at minimum, EPA failed to justify its disparate

treatment of pipeline engines. See, e.g., Balt. Gas & Elec. Co. v. FERC, 954 F.3d 279, 285 (D.C.

Cir. 2020) (the “duty to explain inconsistent treatment is incumbent on the agency”).

EPA also persists in claiming that its 1,000-horsepower criterion “allowed EPA to focus its analysis on the largest emitters with the most significant potential reductions.” EPA

Resp. 32. But EPA does not dispute that less than ten percent of the engines subject to the

Rule emit more than 100 tons per year. Emergency Application 22–23. Indeed, more than

half of the engines subject to the Rule emit less than 10 tons per year. See EPA, Non-EGU

Facilities and Units.xlsx (Mar. 2023), https://perma.cc/UDK9-LRKU (downloads file). EPA’s

finding that its horsepower-proxy “reasonably approximates” the 100-tons-per-year threshold, 88 Fed. Reg. at 36,820, is patently arbitrary and capricious.9

Nor did EPA offer a rational explanation for its grossly overinclusive Rule. See Emergency Application 22–23. EPA claims that if low-emitting units were exempted, pipelines

could “otherwise shift emissions between controlled and uncontrolled units, thereby evading limits.” EPA Resp. 33. EPA fails to explain why any such maneuvering could not be prevented through monitoring and reporting requirements, as EPA did for low-use boilers. See

Notably, EPA’s screening “analysis focused on actual emissions rather [than] ‘potential to

emit’ emissions.” RTC 109.

Emergency Application 23; EPA Resp. 33. Here again, EPA failed to explain its “inconsistent

treatment” of pipelines. Balt. Gas, 954 F.3d at 285.

Likewise, it is no answer to say that while EPA’s 1,000-horsepower criterion “captures more units than the agency estimated at the time of proposal,” “that threshold still allows for cost-effective emissions reductions.”10 EPA Resp. 34. EPA never determined that it

would be cost-effective to control the 2,500-plus low-emitting pipeline engines it swept into

the Rule, let alone that it would be more cost-effective to control those pipeline engines than

it would be to control the many other sources that EPA excluded from regulation in the Rule.

At bottom, EPA asserts authority to regulate all pipeline engines, regardless of

whether their individual contributions to downwind nonattainment are “significant,” on the

ground that pipeline engines’ “aggregate contribution to nonattainment may be significant.”

EPA Resp. 33-34 (emphasis added); see also RTC at 627 (recognizing that low-use engines’

individual contributions “may be insignificant”). That approach exceeds EPA’s authority under the statute, which nowhere authorizes EPA to regulate individual sources that do not

“contribute significantly” by lumping them together with other sources that do.

The 1,000-horsepower criterion is unlawful.

EPA’s Purported Compliance Flexibilities Cannot Remedy Its Flawed

Rule.

According to EPA, its Rule contains some flexibilities that will lessen its impact on

pipeline engines. These purported flexibilities do not have the benefits EPA claims, and they

do not save the Rule.

Notably, in this section of its response EPA provides a cost-effectiveness threshold of

$4,921/ton for pipeline engines. EPA Resp. 34. This value differs from the cost-per-ton

threshold of $4,981 offered by EPA earlier in its response as satisfying Step 3. So, which is

it – $7,500/ton, $5,339/ton, $4,981/ton, $4,921/ton? The likely answer is: none of them.

First, EPA claims that facility-wide averaging will allow operators to reduce the number of engines needing controls by approximately two-thirds (from 3,005 to 905). 88 Fed.

Reg. at 36,760 (“EPA estimates that the facility-wide emissions averaging provision would,

in many cases, allow facilities to install controls on only one-third of their engines.”), 36,824

(“EPA estimates that a total of 3,005 . . . engines are subject to the final rule.”). It is important

to recognize that this assumption is built into the Timing Report, meaning the Timing Report

is already based on this supposed flexibility. Timing Report ES-8 (385a) (“Given that the

total number of [pipeline engines] that may require retrofits in response to this final rule is

estimated at about 905, we estimate that the maximum length of control installation time for

all sources in this category may potentially be as long as 905/150 = 72 months.”).

Beyond that point, EPA’s facility-wide averaging concept fails as safety a valve. EPA’s

dataset in support of the concept was nowhere close to “statistically significant,” as the

agency now suggests. EPA Resp. 46.11 Indeed, EPA relied on a sample of ten compressor

stations out of 713, and even those were not geographically diverse, nor did they reflect representative engine counts. See EPA, Pipeline Natural Gas - Engines Analysis data, (June 4,

2023), https://perma.cc/4HGB-JDQH (downloads file) (e.g., EPA’s ten selected facilities average 8.3 engines per facility, while record data shows that over 40% of the facilities subject

to the Rule have only one engine); Emergency Application 26. On top of all this, facility-wide

averaging still requires case-by-case EPA approval, at EPA’s discretion. See 40 C.F.R.

§ 52.41(d).

The source EPA cites for that proposition only mentions that EPA selected a “subset”

without ever discussing statistical significance. See Final Non-EGU Sectors TSD 19.

Second, EPA argues that discretionary compliance date extensions should avoid natural gas reliability concerns. EPA Resp. 34–36. Notably, however, that was not why EPA

included the provision for case-by-case compliance date extensions, see 88 Fed. Reg. 36,760

(basing availability of extension on “aforementioned supply chain delays or other circumstances entirely beyond the owner or operator’s control”), and EPA did not assess how its

extension provision would impact natural gas reliability. After all, pipelines still must take

“all steps possible to install controls for compliance with the applicable requirements,” 40

C.F.R. § 52.40(d)(3) (emphasis added), before any extension even will be considered. Operators cannot reasonably rely on this provision to delay implementing emissions controls

projects in the near term because EPA has sole discretion to approve an extension, and because EPA has stated it “anticipates that the majority of industrial sources . . . will not qualify

for a compliance extension.” 88 Fed. Reg. at 36,760. Further, the compliance extension provision, tellingly, does not list reliability impacts as a criterion that would warrant consideration in an extension request.

Third, EPA points to discretionary case-by-case exceptions based on either technical

impossibility or extreme economic hardship. EPA Resp. 14, 31. But EPA intends for these

exceptions to be used be for “unique” and “limited” circumstances. 88 Fed. Reg. at 36,818.

The discretionary exception for technical impossibility requires a showing that a unit cannot

meet emissions limits even with available control technology. 40 C.F.R. § 52.40(e)(2)(i)(A).

And the discretionary exception for extreme economic hardship requires an operator to establish, to EPA’s “satisfaction,” 40 C.F.R. § 52.40(e)(2)(B), that “the costs of compliance for

the source seeking the case-by-case limit would significantly exceed the highest representa-

tive end of the range of estimated cost-per-ton figures identified for any source in the relevant industry,” id. at 36,819 (emphasis added). This standard, and the administrative record,

leave operators guessing what EPA means both by “significantly exceed,” and also, “highest

representative end of the range of estimated cost-per-ton figures.” EPA’s ambiguous, caseby-case approach makes it very unlikely that widespread exemptions will be granted across

the many hundreds of high-cost engines that are required to install controls. Emergency

Application 17.

Fourth, EPA maintains that exempting emergency engines from the Rule helps address pipeline reliability concerns. EPA Resp. 36, 46. But EPA never made this finding in the

Rule, and pipeline companies urged EPA to exempt emergency engines because of their overall low emissions, not the need for reliability. See 88 Fed. Reg. at 36,820–21. Emergency

engines are primarily used as backup power generation, not for transporting natural gas.

See, e.g., 40 C.F.R. § 52.41(a) (providing specific examples of emergency engines, none of

which include transportation of natural gas).

A “safety valve” cannot rescue an irrational rule from “systemic errors, for then the

exception would swallow the rule.” Ass’n of Oil Pipe Lines v. FERC, 281 F.3d 239, 244 (D.C.

Cir. 2002); accord U.S. Telecom Ass’n v. FCC, 359 F.3d 554, 571 (D.C. Cir. 2004). The various

one-off and discretionary exceptions EPA tacked on do not rescue the Rule.

II.

PIPELINE APPLICANTS FACE SUBSTANTIAL, UNRECOVERABLE HARM.

The Pipeline Applicants' Application (at 23–28) explains that irreparable harm from

service outages and high compliance costs “will likely result from the denial of a stay.” Hollingsworth v. Perry, 558 U.S. 183, 195 (2010) (per curiam). Pipeline Applicants explained

that each pipeline engine retrofit installation takes months. Emergency Application 24. To

retrofit thousands of affected engines in less than three years means—as a matter of arithmetic—that a considerable number of engines will be offline during peak-demand periods

occurring in winter and summer. Id. at 18.

In response, EPA claims that “the Rule’s compliance timeline does not threaten reliability.” EPA Resp. 45. That will be cold comfort in the cold months. As explained above, EPA

did not even attempt to analyze the capacity utilization of pipeline engines during peak demand, nor did EPA assess the risk of natural gas shortages during peak demand periods. See

Timing Report at ES-8 (385a) (“[W]e were not able to complete an evaluation of” the reliability concerns raised during the comment period.).

EPA also has no response to the Pipeline Applicants’ concrete evidence of the likely

supply impacts in two example markets. Emergency Application 25; Grubb Decl. ¶¶ 64–67

(687a–691a). As the Pipeline Applicants explained, a pipeline serving 60 percent of the Chicago market would experience a 20 percent capacity shortfall during a peak-demand winter

day, the equivalent of more than a million homes going unserved. Emergency Application

25. A pipeline in the Gulf Coast area would experience shortfalls of natural gas deliveries

during both winter and summer peak periods, equating to hundreds of thousands of homes

going unserved. Id.

The Public Interest Groups, for their part, accuse Applicants of trying to “walk back”

their statements about some engines being used relatively infrequently (which informs the

cost-per-ton of emissions reductions). Public Interest Resp. 26–27. But there is no inconsistency. An engine can have low usage on an annual basis while nevertheless being critically

important during peak demand. Emergency Application 26 (discussing difference between

annual utilization and peak demand). And once a pipeline company chooses to take an engine offline for retrofits, that engine is out of commission until the retrofit is complete and

cannot be readily or quickly brought back online to meet peak demand. The Public Interest

groups also cite experience in other states, see Public Interest Resp. 27, but those other states

have state-wide averaging and longer phase-in periods over a much smaller number of affected engines.12 At least one Applicant recommended such flexibility, see Kinder Morgan

Comment 38 (560a), but EPA declined to adopt the recommendation, without explanation.

As for the Applicants’ substantial and unrecoverable financial injury, EPA has little to

say. The agency contends that pipeline companies do not face immediate enough injury because their compliance date is not until May 1, 2026. EPA Resp. 44. But as they have explained, given the compressed timeframe for compliance, Pipeline Applicants have already

begun spending on the retrofit process—well before a judicial decision has issued. See Emergency Application 27–28 (INGAA members estimate spending several hundred million dollars on retrofits in 12 to 18 months after Rule’s effective date). The Public Interest Groups

contend that costs do not need to be incurred in the short term because of compliance extensions. As explained above, that is no panacea; the standard EPA set for approving a compliance extension requires operators to “take[] all steps possible to install controls for compliance with the applicable requirements.” 40 C.F.R. § 52.40(d)(3) (emphasis added).

The nongovernmental groups also contend that financial injury does not merit a stay

because the compliance costs are not sufficiently “certain” or “great.” Public Interest Resp.

See, e.g., 5 Code Colo. Regs. § 1001-30:B.I.D.5.b.(v)(B) & tbl. 3 (six-year phase-in of requirements); id. § 1001-30:B.I.D.5.c. (providing for state-wide, company-wide emissions

averaging). At least one Applicant recommended such flexibility, see Kinder Morgan Comment 38 (560a), but EPA declined to adopt the recommendation, without explanation.

20 (citing Wis. Gas Co. v. FERC, 758 F.2d 669, 674 (D.C. Cir. 1985)). Pipeline Applicants have

shown they are certain to spend hundreds of millions of dollars on retrofits in the near term.

Emergency Application 27–28. See In re NTE Conn., LLC, 26 F. 4th 980, 990 (D.C. Cir. 2022)

(loss of “millions of dollars” is indisputably “significant”); CSL Plasma Inc. v. U.S. Customs &

Border Prot., 628 F. Supp. 3d 243, 264 (D.D.C. 2022) (“between $2.5 and $4.0 million” is a

“certain and great expense”). EPA, for its part, nowhere disputes that this level of compliance

costs constitutes irreparable injury.

The Pipeline Applicants have shown that harm is “likely” absent a stay, Hollingsworth,

558 U.S. at 195—far more than a “possibility of irreparable injury.” Nken, 556 U.S. at 434.

III.

THE BALANCE OF HARMS AND THE PUBLIC INTEREST WEIGH HEAVILY IN FAVOR OF A STAY.

Because of the likelihood of natural gas shortages at critical times, the public interest

clearly tips in favor of a stay. See Memphis Light, Gas & Water Div. v. Craft, 436 U.S. 1, 18

(1978) (recognizing “utility service” as a “necessity of modern life”); Sierra Club v. Ga. Power

Co., 180 F.3d 1309, 1311 (11th Cir. 1999) (“[A] steady supply of electricity[,]” which is often

generated at natural gas-fired power plants, “during the summer months, especially in the

form of air conditioning to the elderly, hospitals and day care centers, is critical”).

EPA and the nongovernmental groups argue that benefits of reduced NOx emissions

(across the states not currently subject to judicial stays, one supposes) outweigh harms from

energy outages. EPA Resp. 48‒49; Public Interest Resp. 31–35. That is an odd argument. To

begin with, this Court should not weigh benefits resulting from an illegal rule. See Ala. Ass’n

of Realtors, 141 S. Ct. at 2490 (“[O]ur system does not permit agencies to act unlawfully even

in pursuit of desirable ends.”). Second, it is unsurprising that an action EPA takes is intended

to improve the environment; EPA is, after all, the Environmental Protection Agency. If the

mere fact that a rule purports to benefit the public can overcome a stay, EPA’s rules would

all be immune to stays pending judicial review. That would make “judicial review [of EPA’s

rules] an idle ceremony” if emissions controls are installed by the time a judicial decision

issues. Scripps-Howard Radio, 316 U.S. at 10. Third, the Pipeline Applicants only request that

this Court stay the portion of the Rule applicable to pipeline engines, meaning the aggregation of total emissions impacts is not the right metric to balance against. Further, emissions

reductions that EPA estimates will come from pipeline engines by May 1, 2026, are grossly

overstated; as the Pipeline Applicants have explained, that compliance date is impossible to

meet across the thousands of regulated engines. See Emergency Application 17–20.

EPA also argues that the Clean Air Act requires “all feasible reductions” must be attained urgently, by May 2026. EPA Resp. 50. But it cannot now claim urgency in obtaining

emissions reductions where it was the source of delay in disapproving state plans. Emergency Application 29.

The balance of equities and public interest weigh strongly in favor of a stay.

CONCLUSION

For the foregoing reasons, and those in the Application, the Court should stay the Rule

as applicable to pipeline engines and postpone the compliance date for such engines pending

review.

Respectfully submitted,

/s/ Catherine E. Stetson