United States Court of Appeals
         FOR THE DISTRICT OF COLUMBIA CIRCUIT




Argued September 7, 2022          Decided February 14, 2023

                       No. 21-1126

         SOLAR ENERGY INDUSTRIES ASSOCIATION,
                     PETITIONER

                             v.

       FEDERAL ENERGY REGULATORY COMMISSION ,
                    RESPONDENT

   NEWSUN ENERGY LLC AND BROADVIEW SOLAR, LLC,
                  INTERVENORS




   Consolidated with 21-1136, 21-1142, 21-1149, 21-1175



            On Petitions for Review of Orders of
        the Federal Energy Regulatory Commission


    Jeremy C. Marwell argued the cause for petitioners The
Edison Electric Institute and Northwestern Corporation. With
him on the briefs were Sarah N. Norcott and James T. Dawson.
                              2
    Heather Curlee argued the cause for petitioner Solar
Energy Industry Association. With her on the briefs was Todd
G. Glass.

    Adam Lowney and Christopher Jones were on the brief for
amicus curiae Pacificorp d/b/a/ Pacific Power and Rocky
Mountain Power in support of petitioners.

     Jared B. Fish, Attorney, Federal Energy Regulatory
Commission, argued the cause for respondent. With him on
the brief were Matthew R. Christiansen, General Counsel, and
Robert H. Solomon, Solicitor. Anand Viswanathan, Attorney,
entered an appearance.

    Robert M. Loeb argued the cause for intervenors NewSun
Energy LLC and Broadview Solar, LLC. With him on the brief
were Gregory M. Adams, Adam Wenner, and Jeremy R.
Peterman. Peter Richardson entered an appearance.

    Kip D. Nelson, Nick Jimenez, and Irion A. Sanger were on
the brief for amici curiae Carolinas Clean Energy Business
Association, et al. in support of respondent.

   Before: PILLARD and WALKER, Circuit Judges, and
SENTELLE , Senior Circuit Judge.

   Opinion for the Court filed by Senior Circuit Judge
SENTELLE .

    Opinion concurring in part and dissenting in part filed by
Circuit Judge WALKER.

      SENTELLE, Senior Circuit Judge: The Edison Electric
Institute and NorthWestern Corporation, d/b/a NorthWestern
Energy, (collectively, “Utilities”) petition for review of an
                                3
order by the Federal Energy Regulatory Commission
(“Commission”) in which the Commission granted Broadview
Solar’s application to become a qualifying facility under the
Public Utility Regulatory Policies Act of 1978 (“PURPA”).
The Solar Energy Industries Association (“SEIA”) petitions for
review of the Commission’s denial of its motion to intervene
in the adjudication of Broadview’s application.

      Because we conclude that the Commission’s interpretation
of the statute is entitled to deference and that the Commission
did not act arbitrarily or capriciously, we deny the Utilities’
petitions. We dismiss SEIA’s petitions because it lacks Article
III standing.

                          Background

     Section 210 of PURPA was enacted with the goal of
promoting the creation and use of alternative energy. See Am.
Paper Inst., Inc. v. Am. Elec. Power Serv. Corp., 

,
404–05 (1983). It does so, in part, by directing the Commission
to prescribe rules affording “qualifying small power production
facilities,” also commonly known as “qualifying facilities,”
certain benefits. See 16 U.S.C. § 824a-3(a)–(b). To be a
qualifying facility under the Act, a facility must use “biomass,
waste, renewable resources, geothermal resources, or any
combination thereof” to produce energy and have “a power
production capacity which, together with any other facilities
located at the same site . . . , is not greater than 80 megawatts.”
Id. § 796(17)(A)(i)–(ii). Facilities may self-certify that they
meet these requirements, or they may apply for certification
from the Commission. See 

(a)–(b). One
notable benefit to being a qualifying facility is the mandatory
purchase obligation. Under it, electric utilities are required to
purchase the energy generated by qualifying facilities,
                               4
providing those facilities with a guaranteed market. See 16
U.S.C. § 824a-3(a)(2); 

(a).

     In September 2019, Broadview applied for certification
from the Commission that its Montana facility was a qualifying
facility. That facility consists of a 160 MW solar array and a
50 MW battery storage system, both of which produce or store
direct current, or DC, power. Because the nation’s electric grid
runs on alternating current, or AC, power, solar facilities must
also have devices known as inverters to convert DC power into
grid-usable AC power. Broadview’s Montana facility has
inverters with a total net capacity of 80 MW.

     In its application, Broadview noted its intent to
interconnect with and sell energy to NorthWestern Energy, as
it would be entitled to do under the mandatory purchasing
requirement as a qualifying facility. The Edison Electric
Institute, a trade association representing investor-owned
electric companies across the United States subject to
mandatory purchasing requirements, and NorthWestern
Energy filed motions to intervene in the Broadview docket,
objecting to certification of Broadview’s facility. Both
motions were timely filed by the October 2, 2019, deadline.

     The Commission denied Broadview’s application for
certification in a September 2020 Order, determining that
Broadview’s facility exceeded the statute’s maximum “power
production capacity” of 80 MW. See Broadview Solar, LLC,

 (2020), set aside, 

(2021), reh’g denied and modified, 

 (2021).
In reaching this conclusion, the Commission determined that
the relevant “capacity” was that of the solar array, which was
160 MW of DC power, and not the inverters’ “conversion
limit” of 80 MW of AC power. 

. The Commission
acknowledged it was departing from its previous approach set
                               5
out in Occidental Geothermal, Inc., 

 (1981),
which focused on the facility’s net output, or “send-out,”
capacity. It determined, however, that the send-out approach
was inconsistent with the statute’s text. Broadview filed a
request for rehearing. After the Commission issued its
September 2020 Order, SEIA also filed a motion to intervene,
nearly one year after the original deadline.

     In March 2021, the Commission issued a new Order
granting Broadview qualifying facility status and setting aside
its September 2020 Order. Broadview Solar, LLC, 

 (2021). After determining that § 796(17)(A) was
ambiguous as to the proper measure of a facility’s “power
production capacity,” the Commission determined that its
former send-out approach was the best interpretation because
it takes into account all of the facility’s components working
together, not just the maximum capacity of one subcomponent,
and focuses on grid-usable AC power. Broadview Solar, LLC,

, at 61,797. Because Broadview’s send-out
capacity at any single point in time is capped by the inverters’
net output capacity of 80 MW of power, the Commission
determined that Broadview’s facility met the statutory
requirements and granted it qualifying facility status. 

,801–02. In the same March 2021 Order, the
Commission also determined SEIA failed to establish good
cause for its untimely motion to intervene and denied that
motion. 

.

     The Utilities and SEIA filed requests for rehearing. The
Commission issued its June 2021 Order, reaffirming that
Broadview was a qualifying facility and modifying its March
2021 Order to reject the Utilities’ arguments that Broadview’s
facility represented a novel subversion of the statute and that
the battery’s capacity had to be calculated separately from the
                                6
capacity of the solar array. Broadview Solar, LLC, 

 (2021). This appeal followed.

                            Analysis
   A. PURPA

       i.      Chevron Challenge

     The Utilities argue that the Commission exceeded its
statutory authority because, in their view, the “power
production capacity” of Broadview’s facility is the total
amount of DC power generated by the solar array and not the
grid-usable AC power produced by the inverters working in
conjunction with the solar array and battery. The Commission
argues that the statute is ambiguous as to the proper measure of
a facility’s “power production capacity” and that its
interpretation, focusing on the amount of AC power being sent
out to the grid, is reasonable. We agree with the Commission.

    In interpreting the statute, this Court’s analysis is governed
by the two-step framework set out in Chevron U.S.A. Inc. v.
Natural Resources Defense Council, Inc., 

 (1984).
Under step one, the court asks “whether Congress has directly
spoken to the precise question at issue.” 

. If it has,
“the court, as well as the agency, must give effect to the
unambiguously expressed intent of Congress.” 

 at 842–43.
But “if the statute is silent or ambiguous with respect to the
specific issue,” the court moves to step two and must uphold
any agency interpretation that is “reasonable.” 

 at 843–44.

    The parties’ dispute in this case turns on the meanings of
“facility” and “power production capacity” in 

(17)(A). PURPA does not define these terms. In plain
language, a facility’s “power production capacity” is the
maximum amount of power that the facility can produce. But
                                 7
the statute does not state whether the relevant capacity is that
of the individual subcomponent generating DC power, i.e., the
solar array, or of all the facility’s components working together
to produce grid-usable AC power, which would include the
inverters. Because Congress has not spoken to the issue, we
move to step two and must defer to any reasonable agency
interpretation.

    To determine whether the Commission’s interpretation
was reasonable, we look to see if it “is based on a permissible
construction of the statute in light of its language, structure, and
purpose” and consistent with the legislative history. Nat’l
Treasury Emps. Union v. Fed. Lab. Rels. Auth., 

 (D.C. Cir. 2014) (quoting Am. Fed’n of Labor v. Chao,

 (D.C. Cir. 2005)); see also Bell Atl. Tel. Co.
v. FCC, 

 (D.C. Cir. 1997).

     We start with the text. On appeal, the Commission raised
for the first time the argument that “capacity” has an industry-
specific definition meaning the maximum amount of power
that can be supplied to the power grid, i.e., for end-user
demand. Because this was not a basis for the Commission’s
decision, we do not consider it here. See Secs. & Exch.
Comm’n v. Chenery Corp., 

 (1943). Even
so, the Commission’s interpretations of “power production
capacity” as “the facility’s net output to the electric utility,” and
of “facility” as “all of the putative [qualifying facility’s]
component parts as they work together as a whole,” were
eminently reasonable. See 

, at 62,316–17
(internal quotation marks and citation omitted). As discussed,
the statute is ambiguous on the meanings of “power production
capacity” and “facility.” The only grid-usable “power” that
Broadview produces is AC power, and Broadview’s inverters
work with the solar array and battery as an integral component
in producing that power.
                                 8

     The Commission’s interpretation was further guided, and
is amply supported, by the statute’s structure and purpose.
Determining qualifying facility status by the facility’s net
output brings various provisions of PURPA into harmony. One
of the main benefits of being a qualifying facility is the
mandatory purchasing requirement. But the mandatory
purchasing requirement only applies to grid-usable power—
meaning AC power. The Commission’s interpretation of
“power production capacity” similarly focuses on net output of
grid-usable AC power. Thus, the measure used to determine
whether a facility is eligible for qualifying facility status is the
same used to determine benefits available to those qualifying
facilities.

     The Commission’s focus on net output is likewise
“consistent with the statutory purpose” of PURPA. Troy Corp.
v. Browner, 

 (D.C. Cir. 1997) (citing
Chevron, 

). Title II of PURPA was intended
“to encourage the development of . . . small power production
facilities” and promote the use of alternative energy sources,
such as solar. Conn. Valley Elec. Co. v. FERC, 

 (D.C. Cir. 2000) (quoting FERC v. Mississippi, 

 (1982)). Excluding facilities from qualifying facility
status because their component parts have individual
production capacities over 80 MW, even though the overall
facility cannot send out more than 80 MW to the grid, would
be inconsistent with that goal.

     Compared to facilities that rely on other energy sources,
solar facilities are relatively inefficient at generating power. A
solar array needs sunlight; cloud cover and nighttime hinder its
production capabilities. Broadview addressed this by installing
a solar array with a capacity of 160 MW and a battery, enabling
it to produce extra power to be stored in the battery while
                                9
conditions are optimal and then release that power to the grid
when conditions prevent the array from producing enough
power to meet the inverters’ 80 MW limit. The Utilities
complain that this allows Broadview to circumvent the
statutory restrictions on qualifying facilities. But viewed in
light of the statute’s purpose, this arrangement is a feature, not
a bug: Broadview is able to more consistently produce, send
out, and sell the maximum amount of renewable energy
permitted under the statute.

     The Commission’s interpretation is also consistent with
the legislative history. See City of Cleveland v. U.S. Nuclear
Reg. Comm’n, 

, 1367–68 (D.C. Cir. 1995). The
Utilities rely on one sentence from a House Committee Report
stating that “[t]he power production capacity of the facility
means the rated capacity of the facility.” H.R. Rep. No. 95-
1750, at 89 (1978) (Conf. Rep.). While neither the legislative
history nor PURPA defines “rated capacity,” it is most
frequently used to refer to the performance anticipated under
“standard operating conditions.” Occidental, 

, at 61,444–45. The Utilities adopt this definition in
their briefing but fail to apply that definition to the House
Committee’s full quote, which referred to the “rated capacity
of the facility.” H.R. Rep. No. 95-1750, at 89 (1978) (Conf.
Rep.) (emphasis added). Broadview’s facility consists of a
solar array, battery, and inverters that can regularly produce 80
MW of grid-usable power. As the Commission previously
recognized, “a facility’s power production capacity is not
necessarily determined by the nominal rating of even a key
component of the facility. . . . [I]t is not uncommon for smaller
facilities to find it most economic to employ commercially
available components[,] some of which have individual
capabilities significantly exceeding the overall facility
capabilities.” Occidental, 

, at 61,445.
                               10
     The Commission’s determination that Broadview is a
qualifying facility with a “power production capacity . . . not
greater than 80 megawatts,” 

(17)(A)(ii),
because its component parts, working together, produce no
more than 80 MW of grid-usable AC power was reasonable and
well-supported by the statute’s text, structure, purpose, and
legislative history.

       ii.     Arbitrary and Capricious Challenges

     The Utilities raise several other arguments, none of which
compels a different result than their first. First, the Utilities
claim the Commission acted arbitrarily and capriciously by
granting Broadview’s application and ignoring errors on one of
Broadview’s form submissions. The Commission requires that
all qualifying facility applicants complete its Form 556. See 

(a)(1), (b)(2). That form provides a formula
for calculating the facility’s maximum net power production
capacity, starting with the “maximum gross power production
capacity at the terminals of the individual generator(s)” and
subtracting out certain enumerated figures, including electrical
losses and power used to run the facility’s equipment. FERC
Form No. 556. When asked for the “maximum gross power
production capacity at the terminals of the individual
generator(s),” Broadview, in one submission, reported a value
of approximately 82.5 MW, while the Utilities claim the
correct value was 160 MW. Because of that error, the Utilities
claim the Commission could not grant Broadview’s
application.

      This argument fails because it treats an applicant’s
completion of Form 556—a tool meant to aid the Commission
in its eligibility determination—as itself determinative. As the
Commission explained in its March 2021 Order, “Form No.
556 was always intended to be a flexible tool . . . to submit
                                11
information relevant to whether a facility meets the
requirements to be considered a [qualifying facility].” 

, at 61,800. Even assuming the correct input on
the form was 160 MW, Broadview explained its facility’s novel
setup and why its “maximum net power production capacity”
was 80 MW. The Commission’s decision to treat Broadview’s
Form 556 submissions as helpful for determining, but not
dispositive of, the facility’s eligibility was not arbitrary or
capricious.

     The Utilities also argue the Commission’s decision to treat
the solar array and battery as a single facility was arbitrary and
capricious. Because the Commission’s decision to do so was
not inconsistent with the statutory text nor the Commission’s
own precedent, this argument also fails.

     When determining whether a facility is eligible for
qualifying facility status, the Commission must look at the
combined power production capacity of “facilities located at
the same site.” 

(17)(A)(ii); see also 

(a)(1)–(2). As we have discussed, the Commission’s
interpretation of “facility” to encompass all the components
working together to produce grid-usable AC power was
reasonable. But standing on its own, Broadview’s battery can
store only DC power and cannot deliver any usable power to
the grid. Accordingly, the battery is not a separate “facility”
under the Commission’s reasonable interpretation of the
statutory text.

     Citing Luz Development & Finance Corp., 

 (1990), the Utilities argue that Broadview’s battery
must be considered a separate facility and its capacity
aggregated with that of the solar array or inverters. But Luz
merely recognized that a battery can be a standalone qualifying
facility, 

; that possibility does not compel the result
                               12
that it must be a separate facility. The battery in Luz was used
to store energy purchased from the grid until it was later resold
during periods of higher demand, 

, and is easily
distinguishable from Broadview’s battery that stores DC power
until it can be sent through the inverters and transformed into
grid-usable AC power.

     Finally, the Utilities challenge the Commission’s decision
to look at Broadview’s instantaneous net power output and not
its power output over time. The statute measures “power
production capacity” in “megawatts.” But power production
over time is measured in “megawatt-hours.” Rather than being
arbitrary and capricious, the Commission’s focus on
instantaneous power production adhered to the statutory
language.

   B. SEIA’s Petitions

    Turning now to SEIA’s petitions for review of the
Commission’s denial of its motion to intervene, “[o]ur analysis
begins and ends with consideration of our jurisdiction.”
Swanson Grp. Mfg. LLC v. Jewell, 

 (D.C. Cir.
2015).

     For this Court to have jurisdiction, the plaintiff must have
standing.     “The ‘irreducible constitutional minimum of
standing contains three elements’: (1) the plaintiff must have
suffered injury in fact, an actual or imminent invasion of a
legally protected, concrete and particularized interest; (2) there
must be a causal connection between the alleged injury and the
defendant’s conduct at issue; and (3) it must be ‘likely,’ not
‘speculative,’ that the court can redress the injury.” Ctr. for
Law & Educ. v. Dep’t of Educ., 

 (D.C. Cir.
2005) (quoting Lujan v. Defs. of Wildlife, 

, 560–
61 (1992)).
                                13

      SEIA fails on the first requirement as it has not suffered an
Article III injury-in-fact. SEIA’s claimed injury is that it was
“effectively precluded” from defending the net output, or send-
out, approach in the Commission’s adjudication of
Broadview’s application. Pet. Br. at 9–10. According to SEIA,
any reconsideration of that approach was likely to occur, if at
all, during the Commission’s contemporaneous rulemaking or
the ensuing Ninth Circuit litigation. Because SEIA failed to
anticipate FERC’s decision to reconsider the send-out
approach in the Broadview adjudication, it also failed to timely
intervene in that proceeding and thus could not participate to
defend the approach.

     At the outset, it should be noted that agencies have “very
broad discretion to decide whether to proceed by adjudication
or rulemaking.” Conf. Grp., LLC v. FCC, 

(D.C. Cir. 2013). The Commission’s decision to consider the
send-out approach in the Broadview adjudication, rather than
through the rulemaking process, was within the bounds of its
discretion. SEIA’s claimed injury presupposes that it had a
right to participate in any proceedings regarding the send-out
approach. It did not. “[T]he mere fact that an adjudication
creates a precedent that could harm a non-party does not create
the injury-in-fact required for Article III standing.” 

.

     SEIA’s failure to timely intervene is the result of its own
mistaken judgment. The effect of that mistake—SEIA’s
inability to participate in the Commission’s proceedings—does
not give rise to an Article III injury. Accordingly, its petitions
are dismissed.
                           14
                    CONCLUSION

     For the reasons stated above, we deny the Utilities’
petitions and dismiss SEIA’s petitions.
WALKER, Circuit Judge, concurring in part and dissenting in
part:

    The Public Utility Regulatory Policies Act gives lucrative
benefits to small facilities that produce solar power. It defines
them as facilities with a “power production capacity” of no
more than 80 megawatts. 

(17)(A)(ii).

     Broadview is a solar-power facility. At its peak, it can pro-
duce up to 130 megawatts of useful power. So it is not a “small
facility.”

    Because the Federal Energy Regulatory Commission con-
cluded otherwise, I would grant the petitions for review and
vacate FERC’s decision.

                        I. Background

A. The Public Utility Regulatory Policies Act

   The Public Utility Regulatory Policies Act encourages
companies to produce renewable energy. See 16 U.S.C.
§ 824a-3(a); see generally FERC v. Mississippi, 

 (1982) (describing the Act’s history).

     To achieve that goal, the Act gives extraordinary benefits
to “small power production facilit[ies].”            

(17)(A). Those facilities produce electricity from “bio-
mass, waste, renewable resources, [or] geothermal resources.”

 § 796(17)(A)(i). The Act exempts them from several regu-
latory burdens. Id. § 824a-3(e)(1) (directing FERC to make
rules exempting “small power production facilities” from reg-
ulation under various statutes). And it guarantees them a viable
market by forcing public utilities to buy power that small facil-
ities produce. 16 U.S.C. § 824a-3(a)(2), (b).
                               2
    Requiring public utilities to purchase all the power pro-
duced by small facilities is strong medicine. It can force them
to buy power that they do not need or to buy power at an above-
market price. That cost is passed on to consumers. Powering
America: Reevaluating the Public Utility Regulatory Policies
Act’s Objectives and its Effects on Today’s Consumers: Hear-
ing Before the H. Subcomm. on Energy & Commerce, 115th
Cong. 84 (2017) (testimony of Terry L. Kouba, Vice President,
Alliant Energy).

     Thus, the Act’s definition of “small facility” plays a key
role in the statutory scheme: It keeps the mandatory-purchasing
regime within bounds. The broader the definition of “small fa-
cility,” the greater the number of power plants that get special
regulatory treatment under the Act.

     The Act defines “small facility” as a “facility” with a
“power production capacity” of no more than 80 megawatts.
Id. § 796(17)(A)(ii).

B. Broadview’s Design

    Broad Reach Power makes solar and wind energy in Cali-
fornia, Montana, Texas, Utah, and Wyoming. Its complex in
Yellowstone County, Montana cost at least $2 billion to build.
In 2019, the Montana Complex could deliver 620 megawatts of
power. That is only slightly less than the amount of power pro-
duced by the Hoover Dam in 1939, when it became the world’s
largest hydroelectric facility. The Story of the Hoover Dam,
Bureau       of     Reclamation        (July    13,     2022),
https://perma.cc/6JWN-BY77.

   In 2019, the Montana Complex contained four separate but
similar solar-power projects. One of them is called Broad-
view I.
                                  3

    Broadview includes a solar array, a battery, and inverters.
With 470,000 solar panels, its solar array produces up to 160
megawatts of direct-current power. The battery stores some of
those megawatts. And the inverters convert up to 80 mega-
watts from DC power to alternating-current power. Because
the electric grid accepts only AC power, inversion makes the
power ready for the grid to receive it.

    Depending on the time of day, Broadview’s components
serve different purposes. During the day, the solar array sends
80 megawatts of power to the inverters and charges the battery.
But at night, it can’t generate power. That’s when the battery
matters most. At night, it sends stored power to the inverters
and then on to the grid. With the battery, Broadview can de-
liver more power to the grid than it could without it.

C. FERC’s Decision

    In 2019, Broadview asked FERC to certify it as a “small
facility.” It argued that its “power production capacity” was
not greater than 80 megawatts because its inverters can send
only 80 megawatts to the grid at once.             

(17)(A)(ii). 1

    FERC initially denied Broadview’s application, but it re-
versed course on rehearing. According to FERC, the Public
Utility Act’s definition of “small facility” is ambiguous

1
  Because Broadview is more than one mile apart from the other fa-
cilities in the Montana Complex, FERC analyzes it separately under
the small-facility rule. 

(a)(1)-(2); see also Order
re. Broadview Solar III, 

 (Aug. 13, 2021) (accept-
ing withdrawal of an application for small-facility status for another
plant in Montana Complex).
                                4
because the statute “neither defines the terms ‘facility’ and
‘power production capacity,’ nor explains how the Commis-
sion is supposed to ascertain the ‘power production capacity’
of any particular ‘facility.’ ” JA 200. FERC decided to inter-
pret “power production capacity” to mean the “maximum out-
put that the facility can produce for the electric [grid].” JA 201.

    Two intervenors, Northwestern Energy and the Edison
Electric Institute, petitioned for this Court’s review. If Broad-
view is a small facility, the Public Utility Act’s mandatory-pur-
chasing rule will force Northwestern and some of Edison’s
members to buy Broadview’s power — even if they don’t need
it.

                           II. Chevron

     The majority opinion captures the central issue: “The par-
ties’ dispute in this case turns on the meanings of ‘facility’ and
‘power production capacity’ in 

(17)(A). [The
Public Utility Act] does not define these terms. In plain lan-
guage, a facility’s ‘power production capacity’ is the maximum
amount of power that the facility can produce. But the statute
does not state whether the relevant capacity is that of the indi-
vidual subcomponent generating DC power, i.e., the solar ar-
ray, or of all the facility’s components working together to pro-
duce grid-usable AC power, which would include the invert-
ers.” Majority Op. 6-7.

     I agree with that summary. The statute does not expressly
state whether “power production capacity” includes “all the fa-
cility’s components working together.” But a lack of express
language does not mean that the statute has no answer to the
question presented. I would not so quickly conclude, as the
Court’s next sentence does, that “Congress has not spoken to
the issue” and so we “must defer to any reasonable agency
                                5
interpretation” under Chevron v. Natural Resources Defense
Council, 

 (1984). Majority Op. 7.

    That is the path of “Chevron maximalism.” Buffington v.
McDonough, 

, 21 (2022) (Gorsuch, J., concurring
in denial of certiorari). When no express text makes the answer
immediately obvious, some maximalists make a beeline to
agency deference — before any inquiry into statutory struc-
ture, cross-references, context, precedents, dictionaries, or can-
ons of construction. Then, they use the tools of statutory inter-
pretation not to find the best reading of the text but instead to
test whether the agency’s interpretation is “reasonable.” 

.

    On the D.C. Circuit, Chevron maximalism is alive and well.
See, e.g., Loper Bright Enterprises, Inc. v. Raimondo, 

 (D.C. Cir. 2022) (“some question” about the meaning
of a statute is enough to trigger Chevron deference); American
Hospital Association v. Azar, 

 (D.C. Cir. 2020)
(relying heavily on Chevron), rev’d sub nom American Hospi-
tal Association v. Becerra, 

, 1906 (2022) (not
mentioning Chevron).

    But the Supreme Court’s recent decisions repudiate maxi-
malism. Indeed, the Court has not deferred to an agency under
Chevron since 2016. See, e.g., Becerra v. Empire Health Foun-
dation, 

 (2022) (not mentioning Chevron); Na-
tional Federation of Independent Business v. OSHA, 

 (2022) (same); BNSF Railway Co. v. Loos, 

(2019) (same). Instead, the Court has policed the limits of def-
erence to agencies. See, e.g., West Virginia v. EPA, 

 (2022).

   The most important limit is found in Chevron itself: “If a
court, employing traditional tools of statutory construction,
                                6
ascertains that Congress had an intention on the precise ques-
tion at issue, that intention is the law and must be given effect.”
Chevron, 

 n.9. In other words, courts must try
every tool of statutory construction before declaring the text
ambiguous and proceeding to agency deference. If they do,
they “will almost always reach a conclusion about the best in-
terpretation” of the statute, thus resolving any ambiguity. Ki-
sor v. Wilkie, 

 (2019) (Kavanaugh, J.,
concurring); see also SAS Institute Inc. v. Iancu, 

 (2018) (quoting Chevron, 

 n.9)
(“Even under Chevron, we owe an agency’s interpretation of
the law no deference unless, after ‘employing traditional tools
of statutory construction,’ we find ourselves unable to discern
Congress’s meaning.”).

    True, Congress may leave “a gap for the agency to fill.”
Chevron, 

. “For example, Congress might [di-
rect] an agency to issue rules to prevent companies from dump-
ing ‘unreasonable’ levels of certain pollutants. In such a case,
what rises to the level of ‘unreasonable’ is a policy decision.”
Brett M. Kavanaugh, Fixing Statutory Interpretation, 

, 2152 (2016). Where an agency uses its
expertise to fill such a gap, courts should not second guess the
agency’s decision. 

    But today’s case is different. The Public Utility Act does
not invite FERC to fill a policy gap. Instead, as FERC recog-
nizes, the meaning of the statute’s technical language “turns on
legal principles of the sort that a court usually [applies] — i.e.,
principles of statutory interpretation — and not determinations
specifically entrusted to an agency’s expertise.” FERC Br. 40
n.9 (cleaned up). And courts should not defer when a statute’s
meaning can be resolved using normal interpretive tools. “The
judiciary is the final authority on issues of statutory construc-
tion.” Chevron, 

 n.9.
                                  7

     So here there is every reason to resist the temptation “ha-
bitua[lly] to defer to the interpretive views of [the] agenc[y].”
Valent v. Commissioner of Social Security, 

(6th Cir. 2019) (Kethledge, J., dissenting). Instead, we can de-
cide this case by applying, in FERC’s words, the “legal princi-
ples of the sort that a court usually [applies] — i.e., principles
of statutory interpretation.” FERC Br. 40 n.9 (cleaned up).
That approach follows the Supreme Court’s recent Chevron
caselaw and avoids further entrenching a vertical split between
how the Supreme Court and lower courts apply Chevron. 2

           III. Broadview Is Not a “Small Facility”

   Applying the normal tools of statutory interpretation,
Broadview is not a “small facility” under the Public Utility Act
because its “power production capacity” is greater than 80
megawatts.

A. “Facility”

    Start with the term “facility.” 

(17)(A). A
facility is “something . . . that is built, installed, or established
to serve a particular purpose.” Facility (def. 4b), Merriam-

2
  Though the Supreme Court has given up on Chevron maximalism
(and perhaps on Chevron altogether), lower courts have not. Be-
tween 2003 and 2013, lower courts applied Chevron in 74.8% of stat-
utory interpretation cases involving agencies and reached step two
65.7% of the time. Kent H. Barnett & Christopher J. Walker, Chev-
ron in the Circuit Courts, 

, 29, 33 (2017). That
trend has continued since then. In 2020 and 2021, circuit courts ap-
plied 

.5% of the time and reached step two in 59.2% of
those cases. See Brief of the Cato Institute and Liberty Justice Center
as Amicus Curiae in Support of Petitioners at 21, Loper Bright v.
Raimondo, No. 22-451 (2022) (supporting petition for certiorari).
                                8
Webster (2023). The statute’s focus on a “facility” suggests
that we should assess the production capacity of a power plant
as a whole, not the capacity of an individual component.

    That rules out a few possibilities.

    First, it tells us that we should not look only at the capacity
of Broadview’s 160-megawatt solar array. That approach
would ignore the facility’s other components — for instance,
the inverters that limit the array’s output to the grid.

    Second, it tells us that we should not exclude the power
used to charge the facility’s battery. The battery is part of the
facility. So refusing to count power that the solar array sends
to the battery fails to give full meaning to the word “facility.”

     FERC says we shouldn’t count power sent to the battery
because it is “not useful to anybody.” See Oral Arg. Tr. 31.
But a battery like Broadview’s lets a solar facility send power
to the grid at times when it otherwise could not. By allowing
the facility to deliver power at night, the battery “increase[s]
[Broadview’s] ability to provide reliable and/or timely service
to . . . customers.” JA 54 (Pasley Affidavit).

    The battery also makes Broadview more efficient. A solar-
power facility without a battery sends to the grid “approxi-
mately 25 to 30 percent” of the maximum power its array could
theoretically generate each day. 

 With the battery, Broad-
view sends “approximately 35 to 40 percent,” 

 because it is
“capable of sustaining its maximum output for additional hours
in the day,” JA 23. That increased efficiency makes the facility
more profitable. See Christopher Cerny, A Broad View of
Broadview Solar: How FERC’s Whiplash-Inducing Orders Ex-
pand the Scope of PURPA, 23 Minn. J.L. Sci. & Tech. 363, 406
(2022).
                               9

  In short, the battery is useful. It lets Broadview make more
money by prolonging its maximum output.

B. “Power Production Capacity”

   Turn next to the phrase “power production capacity.”

   1. “Power”

    Power means “a source or means of supplying energy, es-
pecially[ ] electricity.” Power (def. 6), Merriam-Webster
(2023). “Power” includes both DC power and AC power. See
Chemeheuvi Tribe of Indians v. Federal Power Commission,

 (D.C. Cir. 1973) (discussing history of
power transmission). So both the DC power used to charge the
battery and the AC power sent directly to the grid count as
“power.”

    Yet FERC claims that only the 80 megawatts of AC power
sent to the grid should count as Broadview’s power-production
capacity. That adds an atextual limit that Congress didn’t
adopt. The Public Utility Act says “power production capac-
ity,” not “AC power production capacity.” And Congress is
perfectly capable of saying “AC” when it wants to. See, e.g.,
26 U.S.C. § 48E(a)(2)(A)(ii) (defining a “qualified facility” as
one “with a maximum net output of less than 1 megawatt (as
measured in alternating current))” (emphasis added).

   2. “Production”

    After “power” comes “production.” To “produce” some-
thing is to “create” it, or to “cause [it] to accrue.” Produce
(defs. 6 & 7), Merriam-Webster (2023). Another apt synonym
is to “generate.” See Facebook, Inc. v. Duguid, 

1163, 1171-72 (2021) (noting the “close[ ] connect[ion]” be-
tween the verb “produce” and the noun “generator”).

    Power sent to a battery like Broadview’s is created and does
accrue. Before the sun’s rays hit Broadview’s array, the battery
is empty. It is charged when the facility converts solar energy
into useful power. If Broadview did not “produce” the power
used to charge the battery, what did? 3

    Consider what happens when the battery charges. Broad-
view uses a lithium-ion battery. Charging that battery prompts
a chemical reaction, causing lithium ions to move within the
battery. How Does a Lithium-Ion Battery Work?, Energy.gov
(Sept. 14, 2017), https://perma.cc/CUA8-Y9UK (during charg-
ing “[l]ithium ions are released by the cathode and received by
the anode”). Without power, that chemical reaction could not
happen. So Broadview must “produce” the power used to
charge the battery.

    3. “Capacity”

    In the statute’s context, “capacity” means “the maximum
amount of power that the facility can produce.” Majority Op.
6-7; see also Capacity (def. 5), Merriam-Webster (2023) (de-
fining “capacity” as “maximum output”).

    But here, FERC rewrites the statute. It says “capacity” in-
cludes only the power that a facility supplies to the electric grid.

3
 Some power at facilities like Broadview is lost to inefficiencies dur-
ing production. FERC allows power plants to deduct those “electri-
cal losses” from their power production capacity. See JA 210. So if
Broadview had a 160-megawatt array, 80-megawatt inverters, and no
battery, it would count as a “small facility” — albeit an inefficient
one that loses half of its potential output during production.
                                 11
Yet that changes “power production capacity” to “power deliv-
ery capacity.” And the word “production” means something
different from “delivery.” See Deliver (def. 5), Merriam-Web-
ster (2023) (“[T]o send . . . to an intended target or destina-
tion.”). 4

    To its credit, FERC conceded at oral argument that “power
production capacity” would likely include power never deliv-
ered to the grid if it is used “on site” for a “useful” purpose like
powering an on-site factory. Oral Arg. Tr. 30. But that con-
cession just highlights the problem with FERC’s approach:
Charging a battery like Broadview’s is a useful purpose.

C. Broadview’s “Power Production Capacity”

     Broadview has the capacity to produce 130 megawatts of
power. It produces 80 megawatts of inverted AC power that is
delivered to the grid while producing 50 megawatts of not-yet-
inverted DC power to charge its battery. 5 Because “power”




4
  FERC conflated “production” and “delivery” in its rehearing order,
although its counsel wisely retreated from that approach on appeal.
Compare JA 201 (FERC: “ ‘ production’ and ‘delivery’ . . . are over-
lapping”), with Oral Arg. Tr. 33 (FERC: “we’re not talking about
delivery”); 

 (“[Y]ou’re not depending on a conflation of the
words production and delivery — right? [FERC:] Correct.”).
5
  The record is unclear on the amount of power the battery can re-
ceive from the array. But the parties agree that the battery can take
in up to 50 megawatts. Compare Edison Br. 10 n.3 (“The Broadview
Project’s battery can be charged at the same rate as it dis-
charges — i.e., it can receive and send out 50 megawatts of energy
each hour.”), with FERC Br. 14 (“[U]p to 50 megawatts of power is
diverted to battery storage for later release.”).
                               12
includes AC and DC power, Broadview’s power production ca-
pacity is the sum of the two:

                         80 + 50 = 130

    Consider an analogy. Every weekday, a lumberjack cuts
down two trees and chops them into sellable timber. But he
has a small truck and can take only one tree’s worth of timber
to market daily. What is the lumberjack’s daily timber “pro-
duction capacity”? Two trees. Every day he works, he can turn
two trees into sellable timber. (Maybe he delivers some of the
other trees on the weekends.)

     Broadview is similar. When the sun is out, Broadview pro-
duces 80 megawatts of power for the inverters and 50 mega-
watts of power for the battery — the equivalent of the lumber-
jack’s two trees. Like the lumberjack’s second tree, the 50
megawatts of power sent to the battery is still produced even
though it isn’t immediately delivered to the market for use on
the grid. The key is that the 50 megawatts produced by the
solar array and sent first to the battery is not wasted by the fa-
cility. Those 50 megawatts end up on the grid — just like the
80 megawatts sent from the solar array directly to the inverters.

    That gives Broadview a power production capacity of 130
megawatts. And because the power production capacity of a
“small facility” cannot exceed 80 megawatts, Broadview is not
a “small facility.” 

(17)(A).
                                 13
                         IV. Conclusion

    The following three facts are uncontested:

    1. When the Public Utility Act says “power,” it does not
       specify between AC power and DC power.

    2. Broadview can send 80 megawatts of AC power di-
       rectly to the grid for sale via the inverters.

    3. At the exact same moment, up to 50 megawatts of DC
       power goes straight to the battery, then later to the in-
       verters, and then on to the grid for sale.

    Because Broadview can produce 80 megawatts for its in-
verters while it simultaneously produces 50 megawatts for its
battery, Broadview’s facility is capable of producing more than
80 megawatts of power. So it is too large to be a “small facil-
ity.”

   For that reason, I would grant the petitions, vacate the re-
hearing orders, and remand to FERC for reconsideration. 6




6
  I agree with the majority that Solar Energy lacks standing to chal-
lenge FERC’s denial of its motion to intervene.