United States Court of Appeals
FOR THE DISTRICT OF COLUMBIA CIRCUIT
Argued March 17, 2016 Decided August 30, 2016
No. 15-1009
PETRO STAR INC.,
PETITIONER
v.
FEDERAL ENERGY REGULATORY COMMISSION AND UNITED
STATES OF AMERICA,
RESPONDENTS
STATE OF ALASKA, ET AL.,
INTERVENORS
On Petition for Review of Orders of the
Federal Energy Regulatory Commission
Eric F. Citron argued the cause for petitioner. With him
on the briefs was Thomas C. Goldstein. Michael Diamond,
Jonathan D. Simon, Angela K. Speight, and Lawrence G.
Acker entered appearances.
Bradley S. Lui, Joseph R. Palmore, Marc A. Hearron,
and Craig W. Richards, Attorney General, Office of the
Attorney General for the State of Alaska, were on the briefs
for intervenor State of Alaska in support of petitioner.
2
Susanna Y. Chu, Attorney, Federal Energy Regulatory
Commission, argued the cause for respondents. With her on
the brief were William J. Baer, Assistant Attorney General,
Robert B. Nicholson and Robert J. Wiggers, Attorneys, and
Robert H. Solomon, Solicitor, Federal Energy Regulatory
Commission.
James F. Bendernagel, Jr. argued the cause for
intervenors. With him on the brief were Eugene R. Elrod,
Robin O. Brena, Kelly M. Helmbrecht, Jeffrey G. DiSciullo,
Andrew T. Swers, Matthew W.S. Estes, John A. Donovan,
Glenn S. Benson, Barbara S. Jost, Dean H. Lefler, and
Deborah R. Repman.
Before: TATEL and SRINIVASAN, Circuit Judges, and
EDWARDS, Senior Circuit Judge.
Opinion for the Court filed by Circuit Judge SRINIVASAN.
SRINIVASAN, Circuit Judge: The Trans Alaska Pipeline
System is the sole means of transporting oil from Alaska’s
North Slope to the shipping terminal at Valdez, Alaska,
roughly 800 miles to the south. Oil companies deposit crude
oil extracted from their fields on the North Slope into the
pipeline at its northern point. Although the companies’ crude
oil deposits differ in ways that affect their respective market
values, the deposits necessarily become commingled in the
pipeline. At the southern end of the pipeline in Valdez, the oil
companies receive the same proportion of oil they initially
contributed to the common stream. Because of the
commingling, however, the companies generally will not
receive the same quality of oil at Valdez that they initially
delivered into the pipeline at the North Slope.
3
Absent monetary adjustments to compensate for the
difference in quality between inputs and outputs, companies
depositing relatively higher-value crude oil into the pipeline
would unfairly suffer a financial loss, while those depositing
lower-value crudes would secure a financial windfall. To
avoid that result, the Federal Energy Regulatory Commission
oversees a mechanism for calibrating payments known as the
Quality Bank. The Quality Bank assigns each company’s
crude oil a value based on the quality of its components or
“cuts.”
This case concerns the formula used to value one of those
cuts, called Resid. In 2013, the Commission initiated an
investigation into Resid pricing. During this investigation,
Petro Star argued that the Quality Bank methodology
undervalues Resid in an unjust and unreasonable manner.
The Commission rejected Petro Star’s argument and declined
to change the Resid valuation formula.
We conclude that the Commission failed to respond
meaningfully to evidence presented by Petro Star, rendering
its decision arbitrary and capricious, and that Petro Star’s
purported failure to provide a viable methodology does not
provide an independent ground for the Commission’s
decision. We thus grant the petition for review and remand
for the Commission to reconsider the methodology used to
value Resid or to provide a more reasoned explanation for its
approach. We also find that Alaska lacks standing to
intervene in this matter.
4
I.
A.
Since 1984, the Federal Energy Regulatory Commission
(FERC) has relied upon the Quality Bank to calculate
monetary adjustments between oil companies that use the
Trans Alaska Pipeline System (TAPS) to transport oil in a
commingled stream. See Trans Alaska Pipeline System, 29
FERC ¶ 61,123 (1984). The Quality Bank “charges shippers
of relatively low-quality petroleum who benefit from
commingling and distributes the proceeds to shippers of
higher quality petroleum whose product is degraded by
commingling.” OXY USA, Inc. v. FERC, 6 F.3d 679, 684-85
(D.C. Cir. 1995). The Quality Bank is thus a zero-sum
transfer mechanism: the goal is to “place each [company] in
the same economic position it would enjoy if it received the
same petroleum at Valdez that it delivered to [the pipeline] on
the North Slope.” Id.
Since 1993, the Quality Bank has used the “distillation
method” to calculate the monetary adjustments. See Trans
Alaska Pipeline System, 65 FERC ¶ 61,277, 62,282 (1993).
Distillation is the initial step in the oil refining process. It
involves the separation of crude oil into different components
or “cuts” through heating and boiling. From lightest to
heaviest, the nine Quality Bank cuts are: (1) Propane,
(2) Isobutane, (3) Normal Butane, (4) Light Straight Run,
(5) Naphtha, (6) Light Distillate, (7) Heavy Distillate,
(8) Vacuum Gas Oil, and (9) Resid. The heavier cuts at the
end of the list are of lower quality.
The Quality Bank assigns a value to each of the nine
distillation cuts and determines how much of each cut makes
up the crude oil streams deposited by an oil company into the
5
TAPS. It then calculates the value of each company’s crude
oil contribution based on the volume-weighted value of its
component cuts. The same formula determines the value of
the commingled common stream.
Before calibrating payments, the Quality Bank must also
account for another variable (in addition to commingling):
the impact of refineries connected to the pipeline along the
route to Valdez. Those refineries divert portions of the
common stream, refining the oil for their own purposes and
processing other petroleum products out of the stream. The
refiners then return the remaining, unused oil to the pipeline.
Because the oil returned generally contains a higher
percentage of lower-quality cuts (like Resid) than the
common stream withdrawn by the refiners, the refining
process reduces the value of the common stream.
Accordingly, at Valdez, the Quality Bank again
calculates the value of the common stream. Oil companies
make payments into or receive payments from the Quality
Bank based on the difference in value between the oil they
deliver into the pipeline and the common stream they
ultimately receive at Valdez. In order to account for the
impact of the refiners, the Quality Bank “compares the value
of the diverted portion of the common stream to that of the
[refinery] return stream, charging the refiners and
compensating other [companies] for the reduction in the
common stream’s value caused by the removal of the refinery
products.” OXY, 64 F.3d at 685. In keeping with the zero-
sum methodology, the charge paid by refiners is distributed to
oil companies who receive lower-quality oil at Valdez than
that which they initially contributed.
6
B.
Because payments under the Quality Bank scheme are
based on the difference in value between different oil streams,
the proper functioning of the Quality Bank depends on
assigning accurate relative values to the nine distillation cuts.
“FERC must accurately value all cuts—not merely some or
most of them—or it must overvalue or undervalue all cuts to
approximately the same degree.” Id. at 693.
Under its current approach, the Quality Bank aims to
achieve that goal by assigning a value to each cut reflecting
its actual market price as closely as possible. Six of the cuts
can be sold following distillation without any additional
processing, and they thus have published market prices. The
Quality Bank uses those prices to value the six “marketable”
cuts. The published market prices for those six cuts are
assumed to include the refining cost of producing the cut—
i.e., distilling the individual cut out of commingled oil.
The remaining three cuts—Light Distillate, Heavy
Distillate, and Resid—cannot be sold without additional
processing following distillation. Those “pre-market” cuts
thus have no published market prices. The current Quality
Bank methodology requires the Commission to set a value for
pre-market cuts, like marketable cuts, after simple distillation
but prior to any further processing. See id. at 694. Because
there is no market for those three cuts without additional
processing, however, there are no published market prices.
In order to determine the hypothetical market price of those
cuts, the Quality Bank starts with the published market prices
for finished products that could be developed from the pre-
market cuts with additional refining. It then ascertains the
value of the pre-market cuts by deducting the additional
processing costs required to produce the finished products.
7
Determining the amount of the deduction requires estimating
the costs associated with operating a hypothetical refinery.
As relevant here, Resid, with additional refining, can be
developed into “coke,” which has a published market price.
Under the current Quality Bank methodology, Resid’s value
equals the market price of coke minus the processing cost
required to convert Resid into coke. In other words, “Resid’s
value is the value of the products from the coking less the cost
of the apparatus and material used in coking.” Order on
Initial Decision and Request for Rehearing, 149 FERC
¶ 61,149, at ¶ 4 (2014) (“Order”).
Of particular significance, the cost deduction for Resid
includes a 20% capital recovery factor (also referred to as a
capital investment allowance). The capital recovery factor
accounts for the capital investment that would be required to
build a hypothetical refinery capable of processing the pre-
market cut into a marketable product, i.e., coke. By
increasing the estimated processing costs of coking, the
capital recovery factor has the effect of reducing the Quality
Bank valuation of Resid. The Quality Bank methodology also
includes a similar 20% capital recovery factor in valuing the
other pre-market cuts, Light Distillate and Heavy Distillate.
(Coking facilities are specific to Resid processing, but
analogous refineries process Light and Heavy Distillate into
their respective finished products.)
This case presents a challenge to the Quality Bank’s
valuation formula for Resid. The current formula was
adopted in a 2004 agency hearing. Trans Alaska Pipeline
Sys., 108 FERC ¶ 63,030 (2004). The Commission affirmed
the Administrative Law Judge’s decision, 113 FERC ¶ 61,062
(2005), and this court upheld the Commission’s Order in its
8
entirety, Petro Star, Inc. v. FERC, 268 F. App’x 7 (D.C. Cir.
2008).
C.
Flint Hills Resources Alaska (Flint Hills) operated a
refinery along the TAPS pipeline. Flint Hills diverted the
common stream for use in its facility, returned an oil stream to
the pipeline that included a greater proportion of Resid, and
made payments into the Quality Bank. The greater the value
of Resid, the more credit Flint Hills would receive for the oil
it returned to the pipeline, and the lower its payments would
be. Resid’s valuation therefore was a particular concern for
Flint Hills.
In August 2013, Flint Hills brought a complaint to the
Commission under the Interstate Commerce Act, 49 U.S.C.
App. § 15(1) (1988), questioning whether the Quality Bank
valuation method remained “just and reasonable,” as required
by the Act. Flint Hills suggested that, as a result of the capital
recovery factor included in Resid’s processing cost
adjustment, the Quality Bank undervalued Resid relative to
the other cuts.
The Commission decided that the complaint should be
dismissed on timeliness grounds. But it initiated its own
investigation into the Quality Bank methodology, explaining
that “a sufficient showing has been made as to whether the
existing Q[uality] B[ank] formula is just and reasonable
insofar as it values Resid” and that the “Commission is not
barred from seeking to determine whether a rate is no longer
just and reasonable no matter how long [ago] it may have
become unjust and unreasonable.” Order Dismissing
Complaint, Initiating an Investigation and Establishing
Hearing, 145 FERC ¶ 61,117, 61,620 ¶ 47 (2013). The
9
investigation focused on “the lawfulness of the existing
Quality Bank methodology”—particularly, its “valuation of
Resid.” Id. at 15.
FERC set the matter for a hearing before an
administrative law judge (ALJ). Petro Star, another refiner
along the TAPS, intervened in the proceeding to support Flint
Hills’ position that the Quality Bank formula undervalued
Resid. Petro Star and Flint Hills argued for removing the
20% capital investment allowance from the Quality Bank’s
formula for Resid. In their view, the capital allowance
resulted in a valuation incommensurate with the prices of the
six marketable cuts. Several major oil companies intervened
in the proceeding to argue in favor of maintaining the existing
formula.
The ALJ rejected Flint Hills’ and Petro Star’s argument
for two independent reasons: (i) they had failed to propose a
just and reasonable alternative to the existing Quality Bank
method, and (ii) they had failed to demonstrate that it was
unjust and unreasonable to include a capital investment
allowance in Resid’s processing cost adjustment. See Initial
Decision, 147 FERC ¶ 63,008 (2014) (Initial Decision). Petro
Star filed exceptions to both parts of the ALJ’s decision. The
Commission affirmed the ALJ’s decision in its entirety, and
Petro Star filed a timely petition for review. Flint Hills had
terminated operations at its North Pole refinery by that time,
so it does not join the appeal.
II.
Section 15(1) of the Interstate Commerce Act authorizes
FERC to prescribe just and reasonable rates if it finds, after a
hearing, that existing rates are unjust or unreasonable. 49
U.S.C. App. § 15(1) (1988). We review the Commission’s
10
determination in order to assess whether it is “arbitrary,
capricious . . . or otherwise not in accordance with law.” 5
U.S.C. § 706(2)(A). When, as here, the Commission’s
analysis “requires a high level of technical expertise,” we
“must defer to the informed discretion of the responsible
federal agencies.” Exxon Co., USA v. FERC, 182 F.3d 30, 37
(D.C. Cir. 1999) (quoting Marsh v. Or. Nat’l Res. Council,
490 U.S. 360, 377 (1989)). In all events, however, “we
require the Commission to engage in rational
decisionmaking.” OXY, 64 F.3d at 690.
In prior cases, that requirement has prompted remands
from our court instructing the Commission to reconsider its
valuation of particular cuts or to provide a more detailed
justification for its existing approach. See Tesoro Alaska
Petroleum Co. v. FERC, 234 F.3d 1286, 1294 (D.C. Cir.
2000); Exxon, 182 F.3d at 34; OXY, 64 F.3d at 701. Most
recently, in Tesoro, Exxon and Tesoro filed complaints
challenging aspects of the prevailing formula. The
Commission dismissed those complaints, finding that they did
not establish “changed circumstances” and that reexamination
of the Quality Bank methodology therefore was unnecessary.
234 F.3d at 1289. We reversed and remanded the case for the
Commission to reconsider the contested formulas or explain
why it need not do so.
We relied on the understanding that a “rate order must be
modified where ‘new evidence warrants the change.’” Id. at
1288 (quoting Tagg Bros. & Moorhead v. United States, 280
U.S. 420, 445 (1930)). Both parties had “offered evidence
that [wa]s new in relation to what was before the Commission
in its earlier determinations and sufficiently compelling to
require reconsideration of the earlier resolution.” Id. In such
circumstances, we concluded, the “Commission’s failure to
respond meaningfully to the evidence renders its decisions
11
arbitrary and capricious.” Id. at 1294. “Unless an agency
answers objections that on their face appear legitimate, its
decision can hardly be said to be reasoned.” Id.
We reach the same result here. In doing so, we note that
the parties dispute Tesoro’s applicability in this context in one
respect. Tesoro requires that the Commission respond
meaningfully only to “new evidence” in evaluating whether
its methodology continues to be just and reasonable. Id. at
1288. The Commission adopted that requirement in the
Order. Order ¶¶ 57-59. Petro Star argues that such a
limitation has no place where, as here, the Commission has
itself initiated the investigation. In that circumstance, Petro
Star contends, the Commission should be required to evaluate
its methodology in light of all of the evidence before it,
regardless of whether that evidence is new. We need not
resolve that dispute in this case, because we find that even if
Tesoro’s new-evidence requirement does apply, Petro Star has
offered indisputably new evidence in support of its argument
that the Quality Bank methodology for Resid valuation is
unjust and unreasonable.
Petro Star’s claim that the Quality Bank methodology’s
inclusion of a capital recovery factor results in undervaluation
of Resid relative to the other cuts is rooted in theoretical
economic principles. Petro Star contends that the published
market prices for the six marketable cuts are short-run, spot-
market prices that do not reflect long-run considerations such
as capital investment returns, which are regarded as sunk
costs. By contrast, the Quality Bank calculates the post-
distillation value of Resid based on a capital investment
allowance that assumes a long-term return of 20% on capital.
According to Petro Star, Resid’s valuation thus is
incommensurate with the valuation of the six marketable cuts,
infringing the essential requirement that the Quality Bank
12
“assign accurate relative values” to the cuts. OXY, 64 F.3d at
693.
We conclude that Petro Star “establish[ed] a prima facie
case that new evidence warrants re-examination” of the
Quality Bank formula used to value Resid. Tesoro, 234 F.3d
at 1293. Accordingly, the Commission was obligated to offer
a meaningful response to Petro Star’s arguments. It failed to
do so. And although we may affirm on the basis of an ALJ’s
reasoning when the agency adopts his or her decision (as the
Commission did here), see Cities of Bethany v. FERC, 727
F.2d 1131, 1144 (D.C. Cir. 1984), the ALJ also failed to
provide a sufficient response to Petro Star’s arguments. We
thus find that the Commission’s decision was arbitrary and
capricious.
Of course, the Commission might reasonably find on
remand that the existing formula used to value Resid
continues to be just and reasonable, such that Petro Star’s
claim will ultimately fall short. But Petro Star has raised a
facially legitimate objection to the inclusion of the capital
recovery factor in the Quality Bank’s processing cost
adjustment for Resid. In response, the Commission must
either answer that objection or change its formula.
A.
We first consider the “less-than-a-barrel” anomaly Petro
Star identifies, which served as the initial impetus for the
proceedings below. Petro Star argues that the Commission
failed to provide a meaningful explanation for how the
Quality Bank methodology can function correctly in light of
that purported anomaly. We agree.
13
The anomaly is premised on the theory that the sum of
the value assigned to each cut under the Quality Bank
methodology should exceed (or at least equal) the real-world
market price for a barrel of Alaska North Slope (ANS) crude.
That is because the Quality Bank, under its current approach,
seeks to assign each cut a value reflecting its market price.
The sum of the Quality Bank price for all nine cuts, Petro Star
contends, thus should approximate the market price for a
barrel of crude oil plus the added value of distillation.
That result in fact prevailed from the time of the current
Resid formula’s adoption in 2005 to 2008. From 2009 to
2012, however, the relationship reversed. During that period,
the market price for a barrel of ANS crude exceeded the
calculated Quality Bank barrel price based on the assigned
value of the nine cuts. See Revised Prepared Direct
Testimony of Philip K. Verleger, Jr. at 28-29 (Feb. 14, 2014)
(J.A. 210-11) (“Verleger Testimony”).
Petro Star argues that the reversal reveals a flaw in the
Quality Bank methodology—specifically, its calculation of
the hypothetical market price of Resid. Because the value of
the six marketable cuts reflects published market prices, the
Quality Bank’s valuation of those cuts is necessarily correct.
Any flaw in the methodology therefore must come from the
valuation of the three pre-market cuts, for which the Quality
Bank estimates hypothetical market prices. Of those three
cuts, the refining costs for Resid substantially exceed those
for Light and Heavy Distillate. On that basis, Petro Star
claims that the most likely explanation for the anomaly is an
error in the Quality Bank formula for Resid, which results in
its systematic undervaluation.
In light of Petro Star’s showing concerning the purported
less-than-a-barrel anomaly, we conclude that Petro Star offers
14
“sufficiently compelling” evidence that warrants a reasoned
response. Tesoro, 234 F.3d at 1288. The Commission’s
decision, however, does not address the alleged anomaly.
Assuming, as we must, that the Commission found Petro
Star’s argument about the anomaly unpersuasive, there is no
explanation for wholly disregarding it. The Commission
described the ALJ’s discussion of the issue, but did not
expressly endorse it or otherwise give any opinion on the
merits. See Order ¶ 74. Even assuming that the
Commission’s silence amounted to an implicit affirmation of
the ALJ’s analysis in light of its ultimate decision on the
matter, the ALJ’s opinion also gave no adequate response to
Petro Star’s argument.
The ALJ asserted that the premise of Petro Star’s
theory—i.e., “that the composite value of the Quality Bank
cuts always should exceed the . . . published price for ANS
common stream crude oil”—is “simply wrong.” Initial
Decision ¶ 136. That may (or may not) be true, but the ALJ’s
explanation falls short regardless. The ALJ stated that the
“QB methodology’s objective is to assign accurate relative
values among the various Quality Bank cuts/ANS crude oil
streams,” not “to determine the actual market values of the
cuts or streams for comparison purposes.” Id. ¶ 137. Because
the expert testimony about the anomaly rested on the
assumption that the Quality Bank cut values reflected actual
market prices, the ALJ rejected the expert’s reasoning as
“unsound” and dismissed the anomaly as insignificant. Id.
That analysis suffers from an important defect. The ALJ
was correct that the Quality Bank cut valuations do not
necessarily have to reflect market prices. After all, the “goal
of the Quality Bank methodology . . . is to assign accurate
relative values to the petroleum that is delivered to TAPS.”
OXY, 64 F.3d at 693 (emphasis added). Under its current
15
approach, however, FERC has chosen to achieve that goal by
assigning values to each cut reflecting their actual market
price as closely as possible. As we have described, the
Quality Bank derives values for the six marketable cuts
entirely from their published market prices, and the formulas
for the pre-market cuts similarly aim to reflect their post-
distillation value in market-price terms. See id. at 694; see
also Exxon, 182 F.3d at 42. Because the Quality Bank
methodology—as constructed—seeks to mirror market prices,
the ALJ failed to give a reasoned response in dismissing the
anomaly on the ostensible ground that the Quality Bank
composite cut valuation and the ANS barrel market price
“have no meaningful connection for Quality Bank valuation
purposes.” Initial Decision ¶ 137.
That explanation, moreover, stands in tension with other
parts of the ALJ’s Initial Decision. Elsewhere, the decision
reflects the understanding that market prices substantiate the
accuracy of Quality Bank valuations. For instance, the ALJ
agreed with Petro Star that, if the evidence demonstrated that
“Resid has a higher market value (vis-à-vis its coker
feedstock Quality Bank valuation) as an FO-380 blendstock,”
that would suggest undervaluation of Resid. Id. ¶ 140. That
acknowledgement rests on the notion that market prices
inform Quality Bank valuations. Whether Resid’s market
value as a blendstock exceeded its value as coker feedstock
would be irrelevant if the two figures were truly independent,
as the ALJ asserted in dismissing the less-than-a-barrel
anomaly.
Of course, Petro Star’s theory concerning the anomaly
assumes that the distillation process adds enough value such
that the composite value of the nine Quality Bank cuts must
always exceed the price of a barrel of ANS crude oil. That
premise may be oversimplified or incorrect. For instance, if
16
some of the post-distillation cuts (such as Resid) effectively
have no value until converted into marketable products
through further refining, distillation, in isolation, could
actually reduce the value of the barrel, because the process of
distillation necessarily involves expenditures—e.g., the costs
associated with construction and operation of the distillation
machinery. Those costs may not be recovered when
distillation produces, in part, a cut requiring additional
refining more costly than its ultimate value. That could
potentially result in a Quality Bank composite value lower
than the ANS crude barrel price.
But we cannot discern any such explanation in the ALJ’s
Initial Decision. The statements that come closest are the
ALJ’s observations that the “Quality Bank composite cut
valuation is based on simple distillation,” id. ¶ 136, and that
the “record indicates there are no simple distillation refineries
operating on the U.S. West Coast,” id. ¶ 136 n.71. The latter
statement includes a citation to the testimony of an expert
witness for Exxon. The expert noted that, “[a]lthough
presumably distillation normally adds value, there are no
distillation refineries operating on the West Coast—
presumably because simple distillation refineries cannot,
without further refining, cover the costs of distillation.”
Prepared Testimony of Michael C. Keeley, Ph.D at 20 (Feb.
21, 2014) (J.A. 87). But neither the ALJ nor the Commission
adopted or expanded upon that reasoning in its response to the
purported anomaly. In light of that silence, we find no
sufficient answer to Petro Star’s argument that the formula for
Resid valuation is flawed because the composite value of the
Quality Bank cuts should exceed the market price of an ANS
barrel.
On appeal, the Commission argues that, regardless of the
significance of the less-than-a-barrel anomaly, the
17
Commission acted reasonably in declining to eliminate the
capital recovery factor from the Resid valuation based on
temporary market conditions. It is true that the anomaly
lasted for less than three years (based on the record before us).
See Initial Decision ¶ 137 & n.73. But whatever the merits of
the Commission’s argument that the anomaly was merely a
temporary phenomenon reflecting no underlying
methodological flaw, the Commission did not offer that
rationale in the proceedings below. The Commission
therefore cannot rely on it here. See Chenery Corp. v. SEC,
318 U.S. 80, 95 (1943). We thus conclude that the
Commission failed to respond meaningfully to Petro Star’s
argument and evidence about the less-than-a-barrel anomaly.
See Tesoro, 234 F.3d at 1294.
B.
We next consider information offered by Petro Star about
recent conditions in the West Coast coking market. Petro
Star’s argument for excluding the capital investment
allowance from the Resid valuation formula rests on the
theory that short-run, spot-market prices for Quality Bank
cuts do not reflect capital investment returns, which instead
are considered sunk costs and are ignored in purchasing
decisions. In addition to the less-than-a-barrel anomaly, Petro
Star presents several items of evidence about the West Coast
coking market aimed to show that its theory is in fact borne
out in the real world.
According to Petro Star, its evidence demonstrates that
“permanent market changes” brought about by the 2008
recession have “compel[led] West Coast refiners to abandon
any reasonable expectation they ever again will realize capital
investment returns on their cokers.” Initial Decision ¶ 143.
Petro Star contends that conditions in the West Coast coking
18
market, at least since 2009, reveal that cokers do not in fact
reap consistent 20% returns on capital, and thus that coke
prices in reality do not include capital recovery costs. If so,
the Quality Bank formula undervalues Resid by nonetheless
subtracting those costs as part of its processing adjustment.
Although the Commission made some effort to respond to
those arguments, we find that the responses failed sufficiently
to address Petro Star’s evidence.
First, Petro Star presents evidence that there has been
effectively no investment in new coking capacity on the West
Coast in recent years and that new coking projects have been
cancelled because they no longer meet rate-of-return goals.
See Verleger Testimony at 61-63 (J.A. 223-25); see also J.A.
199. The Commission affirmed the ALJ’s finding that “the
record contradicts the claim that there has been no significant
new investment in West Coast coking capacity,” but provided
no additional thoughts on the issue. Order ¶ 78. That is
inadequate.
Petro Star argues that there has been no new coker
investment on the West Coast in particular—where cuts
derived from an ANS barrel are actually used and where
market conditions thus would best inform the valuation of
Resid. The ALJ, however, seemingly ignored that
specification. He concluded that the record contradicts Petro
Star’s claim on the basis of witness testimony focused on
coker investment elsewhere in the country. Of the eight new
coker projects mentioned in the testimony, only one is on the
West Coast. See Revised Answering Testimony of John B.
O’Brien at 42-44 (Feb. 3, 2014) (J.A. 171-73); see also J.A.
183. The Commission noted that Petro Star had filed
exceptions disputing the ALJ’s finding on that basis. Order
¶ 79. But it failed to rebut that point or explain why it might
be immaterial.
19
Second, Petro Star offers evidence that existing coking
facilities have been sold at depressed prices, hundreds of
millions of dollars below what would be expected if the 20%
capital returns assumed by the Quality Bank were possible.
See Verleger Testimony at 63-65 (J.A. 225-27). Neither the
Commission nor the ALJ directly addressed the evidence
concerning depressed refining asset values. The Commission
did, however, express disagreement with the inference Petro
Star seeks to draw from that evidence—i.e., that coking
facilities are no longer profitable. The Commission explained
that the “record confirms that refiners still receive significant
margins for investment in new coker facilities” and observed
generally that the “evidence does not demonstrate that refiners
have abandoned any expectation of return on or of investment
from cokers.” Order ¶ 80.
But the ALJ’s findings supporting those conclusions,
which the Commission summarily affirmed, suffer from an
important shortcoming. The ALJ explained that the “record
establishes that while U.S. West Coast coking margins varied
widely over the period from 2004 through 2013, they were
never negative.” Initial Decision ¶ 144. His conclusion
rested upon data showing that coking refiners earned $8-$15
above operating costs per barrel over the past few years. See
Revised Answering Testimony of John B. O’Brien at 38-42
(Feb. 3, 2014) (J.A. 167-71). That data excluded capital
costs, however, as Petro Star pointed out in the proceedings
below. Brief on Exceptions of Petro Star Inc. (June 9, 2014)
(J.A. 508). The profit margins highlighted by the ALJ might
still disprove Petro Star’s supposition about coker
profitability. But the Commission’s failure to acknowledge or
address the apparent limitations of the data leaves its
conclusions largely unsubstantiated. In conjunction with its
failure to address directly the most concrete evidence put
20
forth by Petro Star (the depressed asset prices), its explanation
here is, at best, incomplete.
Finally, Petro Star asserts that there is extra refining
capacity in existing cokers, i.e., that coker facilities are
underutilized. See Verleger Testimony at 60 (J.A. 222). If
cokers could refine with 20% capital recovery, such gaps
would not exist, Petro Star contends. The Commission found
that the record contradicted that argument, as “coker facility
utilization remains at historic levels.” Order ¶ 80. On that
count, unlike the others, the Commission’s answer satisfies
the requirement of reasoned decisionmaking. As the ALJ
explained, on average, the “U.S. West Coast coker
utilization/capacity rates have not fallen materially below . . .
the 87% utilization rate adopted in” the prior proceedings.
Initial Decision ¶ 144. The current utilization rates thus do
not suggest that capital recovery has decreased among coker
facilities.
On the whole, though, the Commission’s analysis
nonetheless falls short. Petro Star presents “sufficiently
compelling” evidence, based on recent conditions in the West
Coast coking market, that refineries no longer expect to
recover capital investment returns on cokers and that coke
prices thus exclude capital costs. Tesoro, 234 F.3d at 1288.
Although the Commission made some effort to respond to
that evidence, its responses contain marked deficiencies. As a
result, we conclude that the Commission failed adequately to
address the evidence before it. In doing so, we recognize that
there may be evidence in the record or elsewhere
contradicting Petro Star’s claims about West Coast market
conditions or undercutting the inferences Petro Star seeks to
draw. But the Commission’s decision fails to contain such an
explanation.
21
In support of its claim that refiners valuing oil streams do
not factor sunk capital costs into their short-run purchasing
decisions, Petro Star additionally points to linear
programming models and “Platts net-back yields.” Refiners
use linear programming models to make crude oil purchasing
decisions. The models predict what refiners can earn by
refining a given crude oil, using the market prices for the
finished products and the variable costs of additional, post-
distillation processing. Similarly, the net-back yields are
designed to reflect the “net-back” that a refiner would earn
from processing a particular crude oil. Like the linear
programming models, the yield estimates published in
Platts—a trade publication relied on by the Quality Bank to
price the six marketable cuts—do not account for capital costs
associated with processing equipment. Petro Star makes
much of the Commission’s (and the ALJ’s) failure to address
the linear programming models and net-back yields.
Insofar as the Commission’s silence on those matters
may have stemmed from an assumption that they do not
constitute “new evidence” which requires a response, see
Order ¶ 59, we need not delve into the merits of the
Commission’s understanding of “new evidence” in that
regard, as noted earlier. Even assuming that the Commission
was obligated to respond only to “new evidence,” taking into
account all of the evidence presented by Petro Star, we
conclude Petro Star “establish[ed] a prima facie case that new
evidence warrants re-examination of how [Resid] should be
valued.” Tesoro, 234 F.3d at 1293. On remand, the
Commission, in responding to the less-than-a-barrel anomaly
and the data from the West Coast coking market, presumably
will also address the linear programming models and net-back
yields given the intertwined nature of that evidence in the
context of Petro Star’s argument that Resid’s valuation should
not include any capital costs deduction.
22
III.
The Commission maintains that its order rests on another,
independent ground: Petro Star was required to propose a just
and reasonable alternative methodology, and its suggestion to
remove the capital recovery factor from the Quality Bank
Resid valuation did not meet that standard. That ground, the
Commission contends, suffices to uphold its decision,
notwithstanding any deficiencies in its analysis of Petro Star’s
evidence. We disagree.
We assume without deciding that, under Tesoro, Petro
Star’s alleged failure to offer a viable proposal would obviate
the Commission’s responsibility to answer Petro Star’s
objections to the existing methodology. See Tesoro, 294 F.3d
at 1294; Reply Br. 3. But even if that were the case as a
general matter, here, the Commission’s basis for rejecting
Petro Star’s proposal is not “independent” at all. Rather, it
rests on an implicit rejection of Petro Star’s argument that
including a capital recovery factor in the Quality Bank Resid
valuation is unjust and unreasonable. That circular rationale
fails to satisfy the requirement of reasoned decisionmaking.
The Commission found that Petro Star could not prevail
because it had “failed to meet its burden by proposing an
inconsistent valuation methodology for Resid.” Order ¶ 72.
The order largely echoed, and then affirmed, the ALJ. See id.
¶ 71. The ALJ in turn noted that, for all three pre-market
cuts, the Quality Bank’s processing cost adjustment includes a
capital investment allowance of 20%. See Initial Decision
¶ 123 & n.61. According to the ALJ, because Petro Star
argues that the capital recovery factor be removed from the
Quality Bank Resid formula but not the Light and Heavy
Distillate formulas, its suggested approach would create an
inconsistency in the valuation of the three pre-market cuts.
23
More specifically, the “disparity necessarily will overvalue
Resid vis-à-vis Light Distillate and Heavy Distillate.” Id.
¶ 124. That result, the ALJ concluded, was impermissible in
light of the requirement that the Quality Bank methodology
“assign accurate relative values” to the cuts. OXY, 64 F.3d at
693. The Commission echoed that rationale, explaining that
“it is the goal of the QB methodology to assign accurate
values to the petroleum that is delivered into TAPS, and it
must accurately value all cuts to achieve this goal.” Order
¶ 71.
We agree with the Commission that methodological
consistency is key in valuing the Quality Bank cuts. But we
cannot see how that affords an independent basis for rejecting
Petro Star’s argument. If Petro Star’s theory is correct—that
the capital investment allowance makes the formula for
valuing Resid incommensurate with the short-run, spot-
market prices used to value the six marketable cuts—
removing the capital investment allowance from Resid’s
valuation would improve the Quality Bank’s methodological
consistency by better aligning Resid’s valuation with that of
the six marketable cuts. Even assuming, as the ALJ does, see
Initial Decision ¶ 131, that Petro Star’s proposal would
overvalue Resid relative to Light and Heavy Distillate, the
proposal still would correct one of the three existing
distortions in the Quality Bank methodology. Petro Star
emphasized that point to the Commission, explaining that the
ALJ’s focus on whether the Resid and distillate cuts were
valued “in lock-step” reflected “the untenable assumption that
three cuts falling short of the OXY standard is somehow more
acceptable than two.” Brief on Exceptions of Petro Star Inc.
(June 9, 2014) (J.A. 514-15).
Intervenor Exxon’s argument manifests the same defect.
Exxon contends that valuing eight of the cuts at the point of
24
simple distillation, while valuing Resid at a downstream point
after additional processing in the refinery (i.e., the coker),
would distort the relative cut values under the Quality Bank
methodology. But to the extent that Petro Star’s theory is
correct, removing the capital investment allowance would not
result in Resid valuation at a downstream point following
additional processing; rather, it would be necessary to
determine the hypothetical market price for Resid at the point
of simple distillation. In short, the Commission’s conclusion
that Petro Star’s proposal would create a methodological
inconsistency follows only if we assume that the capital
recovery factor should be used to derive an accurate market
price for Resid’s post-distillation value. Its analysis thus rests
on rejecting Petro Star’s core contention with respect to the
new evidence.
Moreover, the Commission initiated the proceedings
below as “an investigation . . . into the lawfulness of the
existing Quality Bank methodology”—particularly, its
“valuation of Resid.” Order Dismissing Complaint, Initiating
an Investigation and Establishing Hearing, 145 FERC
¶ 61,117, 61,620 (2013). Yet the Commission then faulted
Petro Star for failing to propose an alternative that addressed
any corresponding deficiencies in the valuation of Light and
Heavy Distillate. In doing so, the Commission effectively
required Petro Star to address matters outside the scope of its
own investigation. The Commission’s only explanation—that
there is “no merit to the argument that this investigation was
limited in scope to the value of Resid without any reference to
the interrelation between valuations of other cuts within the
common stream”—is plainly inadequate. Order ¶ 71. Petro
Star has never argued that its Resid proposal should be
evaluated without reference to the valuation of other cuts. In
fact, the goal of its proposal is to better reflect that
interrelationship. The Commission thus offered no answer to
25
Petro Star’s more nuanced argument that the proceeding, and
therefore its proposal, was focused on the accuracy of the
Resid valuation.
For those reasons, we find that Petro Star’s alleged
failure to suggest a viable alternative proposal cannot serve as
an independent ground for the Commission’s decision. It
follows that that the Commission must, under Tesoro, provide
a meaningful response to the new evidence presented by Petro
Star.
IV.
Finally, we consider the argument made by the State of
Alaska as intervenor. Alaska takes no position on the
appropriate formula for Resid valuation, the focus of Petro
Star’s challenge. Alaska instead seeks to raise an entirely
different issue concerning the showing that parties must make
when challenging the Quality Bank methodology. Alaska
claims that the Commission failed to meaningfully respond to
its argument that a party challenging the Quality Bank
methodology “should be allowed to suggest an alternative,
superior, pro-competitive methodology to replace the existing
. . . methodology”—in other words, that a party should not be
required to demonstrate that the existing methodology is
unjust or unreasonable. Pet. Intervenor Br. 17.
We do not reach the merits of that argument because we
conclude that Alaska lacks standing to bring its claim. In
order to establish standing, a party must demonstrate that it
has suffered an “injury in fact” that is “fairly traceable” to the
defendant’s action and that can likely be “redressed by a
favorable decision.” Lujan v. Defs. of Wildlife, 504 U.S. 555,
560-61 (1992) (quotations and internal alterations and
quotations marks omitted). Here, Alaska’s alleged injury
26
“flows from the legal rationale employed by the Commission
. . . not from the denial of relief actually sought by [the state]
before the agency.” Shell Oil Co. v. FERC, 47 F.3d 1186,
1201 (D.C. Cir. 1995). We have previously found such an
unfulfilled desire insufficient to confer Article III standing in
the absence of any concrete harm. Id. at 1201-02; see
Crowley Caribbean Transp., Inc. v. Pena, 37 F.3d 671, 674
(D.C. Cir. 1994). That understanding equally applies here.
Additionally, Alaska seeks to present an issue not raised
by Petro Star. As a general matter, however, “[i]ntervenors
may only argue issues that have been raised by the principal
parties; they simply lack standing to expand the scope of the
case to matters not addressed by the petitioners in their
request for review.” NARUC v. ICC, 41 F.3d 721, 729 (D.C.
Cir. 1994); see Cal. Dep’t of Water Res. v. FERC, 306 F.3d
1121, 1126 (D.C. Cir. 2002). We thus conclude that Alaska
lacks standing to challenge the “unjust or unreasonable”
standard applicable in proceedings concerning the Quality
Bank methodology.
* * * * *
For the foregoing reasons, we grant the petition for
review and remand the matter to the Commission.
So ordered.