Desert Sunlight 250, LLC v. United States

CORRECTED

In the United States Court of Federal Claims
No. 17-1826 T
Filed: October 8, 2021
Re-issued: November 2, 20211

)
DESERT SUNLIGHT 250, LLC and DESERT )
SUNLIGHT 300, LLC, )
)
Plaintiffs, )
)
v. )
)
THE UNITED STATES, )
)
Defendant. )
)

Steven J. Rosenbaum, Covington & Burling LLP, Washington, D.C., for Plaintiff.
Dennis B. Auerbach, Sean M. Akins, Alexis N. Dyschkant, and Seth A. Mohney, of counsel.

Matthew D. Lucey, United States Department of Justice, Tax Division, Court of Federal
Claims Section, Washington, D.C., with whom were Richard E. Zuckerman, Principal Deputy
Assistant Attorney General, David I. Pincus, Chief, Court of Federal Claims Section, G. Robson
Stewart, Assistant Chief, Court of Federal Claims Section, and Jason S. Selmont and Katherine
Powers, Trial Attorneys, of counsel, for Defendant.

OPINION AND ORDER

MEYERS, Judge.

To encourage investment, Section 1603 of the American Recovery and Reinvestment Act
of 2009 (“Section 1603”) provided cash grants to reimburse the cost of placing in service certain
renewable energy properties. Here, Plaintiffs acquired and applied for Section 1603 grants for
the Desert Sunlight solar energy project in California (the “Project”), which the Government
largely approved and awarded the Plaintiffs $550 million in grants. But the Government did not
approve $59 million of Plaintiffs’ application, and Plaintiffs sued to recover this $59 million.
The Parties have cross moved for summary judgment and the Government has filed a motion in
limine to strike certain statements in Plaintiffs’ declarations.

1
The Court issued this opinion under seal and directed the Parties to confer and propose any
redactions pursuant to the protective order. Because the Parties advise that no redactions are
necessary, the Court re-issues this opinion in full.
 The Government moves for summary judgment arguing that the Plaintiffs have not
produced the required allocation of the consideration they paid for Desert Sunlight among all the
assets Plaintiffs acquired. While the Plaintiffs did not submit their allocation on an IRS Form
8594, whether they provided sufficient information to support their claim is a factual issue that is
not amenable to resolution on summary judgment. Therefore, the Court denies the
Government’s motion insofar as it seeks judgment ending this case.

In the alternative, the Government seeks summary judgment on several legal questions
about how assets should be categorized. Plaintiffs also seek summary judgment on these
questions and several more. For the reasons explained below, these motions are granted-in-part
and denied-in-part.

Finally, the Government has moved to strike certain statements in Plaintiffs’ witness
declarations because they are either legal conclusions, hearsay, or the declarant has not laid the
foundation for the challenged statements. Because none of the challenged statements are
necessary for the resolution of the summary judgment motions, the Court does not consider them
and denies the Government’s motion in limine as moot without prejudice to raise these
objections at trial.

I. Background

A. The Desert Sunlight Facility, the LGIA, and the PPAs

The Desert Sunlight Facility (the “Facility”) is a massive solar energy power plant
located in California’s Mohave Desert. Brannen Decl. ¶ 10 (ECF No. 86-3 at App’x 5).2 Sitting
on a site roughly 20% of the area of Manhattan, the Facility occupies an area approximately 3.2
miles by 2.6 miles. Id. ¶ 11. Prior to the Facility, nobody had undertaken building a solar
energy facility of similar size. Id. ¶ 20. As a solar photovoltaic (“PV”) electricity plant, the
Facility converts sunlight into electricity. Id. ¶ 12. It does so using solar panels, or “modules,”
to convert sunlight into direct current (“DC”) electricity. Id. ¶ 32. The Facility then uses
inverters to convert the DC electricity into alternating current (“AC”) electricity for delivery to
the utility transmission system. Id. (at App’x 21). As designed, the Facility’s DC generating
capacity was projected to be 724 megawatts of DC (“MWdc”). As built, the Facility has a
slightly greater 740.7 MWdc capacity. Id. ¶ 16. Its total AC generating capacity is
approximately 550 megawatts of alternating current (“MWac”). Id. The Facility is comprised of
two segments, a 250 MWac-capacity plant and a 300 MWac-capacity plant. Id. ¶ 17. Plaintiffs
Desert Sunlight 250, LLC and Desert Sunlight 300, LLC own the 250 MWac plant and the 300
MWac plant, respectively. ECF No. 1 ¶¶ 10-11. The Facility entered service in 2014. Brannen
Decl. ¶ 10.

2
As detailed more below, Mr. Brannen was NextEra’s Senior Director of Project Engineering
and Due Diligence. He served as a lead contract negotiator and oversaw the Plaintiffs’ team that
was on-site for the Desert Sunlight Facility’s construction.

2
 Before the Facility’s construction, Plaintiffs were owned by First Solar, Inc. (“First
Solar”).3 ECF No. 1 ¶ 40. First Solar is a leading manufacturer of “thin film” solar modules,
which are widely used in solar PV facilities throughout the world. Brannen Decl. ¶ 15. First
Solar is also a leader in developing, financing, engineering, constructing, operating, and selling
many of the world’s largest grid-connected solar PV power plants. Id.; see also ECF No. 81-2 at
App’x 26. Before construction begins on a solar PV facility, the developer generally “execut[es]
an interconnection agreement” and “enter[s] into a power purchase agreement” (“PPA”). ECF
No. 81-2 at App’x 31. An interconnection agreement “is a contract in which a utility scale
electricity producer obtains the right to interconnect its facility to the electricity grid,” through
which it can then transport its electricity to the purchaser. Charles Decl. ¶ 9 (ECF No. 86-6 at
App’x 1127).4 “A PPA is a long-term contract in which an electricity producer agrees to sell
electricity to a utility or other customer pursuant to an agreed pricing formula[] . . . .” Id. First
Solar achieved both milestones as Plaintiffs’ owner while developing the Facility. Id.

In 2010, Plaintiffs entered into a Standard Large Generator Interconnection Agreement
(“LGIA”) with the Southern California Edison Company (“SCE”) and the California
Independent System Operator Corporation (“CAISO”). ECF No. 81-29 at App’x 2650, 2718-20.
Under the LGIA, Plaintiffs secured the right to interconnect the Facility to CAISO’s electricity
grid by way of SCE’s transmission system once the Facility became operational. Id. at App’x
2650, 2656-57, 2661-62. Also, Plaintiffs were responsible for building everything necessary to
connect the Facility to a substation that SCE was responsible for building. Id. at App’x 2722-23.
In other words, the SCE substation served as the handoff point of the Facility’s electricity to the
power grid; Plaintiffs were responsible for delivering electricity to the substation and SCE was
responsible for transporting it to the grid.

Plaintiffs also executed two PPAs for the output of the Facility. In 2009, Plaintiffs
executed a PPA with SCE. ECF No. 81-21; ECF No. 81-20 at App’x 2030. Under this PPA,
SCE agreed to purchase the total electrical output from DS 250 for 20 years. ECF No. 81-21 at
App’x 2062, 2071. And in 2010, Plaintiffs executed a PPA with Pacific Gas and Electric
Company (“PG&E”). ECF No. 81-22 at App’x 2299. Here, PG&E agreed to purchase the total
electrical output from DS 300 for 25 years. Id. at App’x 2324.

B. First Solar Sells the Facility to GE and NextEra

In 2011, First Solar began negotiating with NextEra Energy Resources, LLC (“NextEra”)
and GE Energy Financial Services (“GE”) to sell First Solar’s interest in Plaintiffs. Brannen
Decl. ¶ 13. NextEra builds and operates solar power facilities. Id. ¶¶ 7-8. GE is a financial
services provider with experience investing in the renewable energy industry. Id. ¶ 13.

3
The corporate entities to the transactions in this case acted through various subsidiaries.
Because the specific subsidiary entities are not relevant to the issues before the Court, the Court
refers to the parent entities for ease of reference unless otherwise indicated.
4
Prior to his retirement, Mr. Charles was a Senior Principal Consultant at Sargent & Lundy,
where he worked extensively on solar projects. Plaintiffs have retained Mr. Charles as an expert
witness.

3
 On September 29, 2011, First Solar, NextEra, and GE executed several agreements,
including: (1) three Engineering, Procurement, and Construction (“EPC”) Agreements; (2) three
Operating and Maintenance (“O&M”) Agreements; (3) financing agreements; (4) a Department
of Energy (“DOE”) Loan Guarantee; and (5) a Membership Interest Purchase and Sale
Agreement (“MIPSA”).

1. The EPC Agreements and O&M Agreements

The three fixed-price EPC Agreements are contracts to engineer, procure, and construct
the Facility. Id. ¶ 14. The first EPC Agreement is between Plaintiff DS 250 and First Solar.
ECF No. 81-12. In this agreement, Plaintiff DS 250 hired First Solar to develop and construct
the 250 MWac portion of the Facility for a fixed price of $807,932,028. Id. at App’x 858, 904.
The second agreement is between Plaintiff DS 300 and First Solar, in which Plaintiff DS 300
hired First Solar to develop and construct the 300 MWac portion of the Facility for a fixed price
of $967,742,759. ECF No. 81-13 at App’x 1011, 1056. The third agreement is between both
Plaintiffs and First Solar, in which Plaintiffs hired First Solar to develop and construct the
common areas of the Facility for a fixed price of $175,132,299. ECF No. 81-15 at App’x 1666,
1702. The total price of all three EPC Agreements is $1,950,807,086. Plaintiffs contend that
this total includes approximately $104 million in sales tax. Brannen Decl. ¶¶ 21, 62.

The EPC Agreements’ terms underwent changes over the course of negotiations between
First Solar, NextEra, and GE, according to both NextEra’s lead negotiator, William Brannen, Id.
¶¶ 22-23, and First Solar’s Kent Draper, Draper Dep. Tr. at 28:7-8, 29:1-7, 265:13-17 (ECF No.
81-63 at App’x 4149, 4208). In the beginning of the negotiations, First Solar sought a total price
of over $2.36 billion for all three EPC Agreements, but NextEra and GE negotiated the final
price down by over $400 million to reach the final $1,950,807,086 total price. Brannen Decl. ¶
22; see also Draper Dep. Tr. 264:21-265:12. NextEra and GE also convinced First Solar to take
on greater risk than it originally accepted, such as the risk of increased cost of equipment,
material, and labor during the construction period, even though acceptance of greater risk
typically results in higher prices. Brannen Decl. ¶ 23; Draper Dep. Tr. 270:25-271:19, 273:1-7,
274:10-15. Mr. Brannen attributes the final deal to “the hard bargain that NextEra and GE
drove.” Brannen Decl. ¶ 23. Mr. Draper agreed. Draper Dep. Tr. 274:22-275:6. Indeed, Mr.
Draper testified that the negotiation “was . . . very arm’s length . . . it was an extremely hard
negotiation with GE and NextEra.” Id. at 265:13-17. Mr. Draper further testified that “they [GE
and NextEra] were tough negotiators. They’re, you know, extremely well-versed in . . . the
issues that we were dealing with, both on the technical side and risk allocation under EPC
agreements and . . . they drove a hard bargain.” Id. at 266:24-267:5.

Moreover, the negotiations concerned not just the total EPC Agreement prices but also
the price for specific components under each EPC Agreement. Brannen Decl. ¶¶ 18, 44. The
specific component prices were set forth in Exhibit I to all three EPC Agreements. ECF No. 88
at App’x 285 ¶ 12; see id. at App’x 353-360.

Exhibit I effectively has two parts. See Brannen Dep. Tr. at 75:9-76:8, 117:5-118:13
(ECF No. 81-6 at App’x 230, 240-41) (Mr. Brannen referring to the first part as the “schedule of
values” and the second part as the “milestone payment schedule.”). The first part contains
schedules of component pricing for each EPC Agreement. ECF No. 88 at App’x 353-56; see

4
also Pitale Decl. ¶¶ 12-13 (ECF No. 88 at App’x 285-86).5 In relevant part, as shown in the
excerpt below of the schedule of the DS 300 EPC Agreement’s component pricing, the schedules
list: (1) the EPC Agreement’s components (Milestones); (2) the quantity of components at which
they are priced (Unit of Measure); (3) the percentage of each unit’s price within the EPC
Agreement’s price (Milestone Value (% of Sale Price)); (4) each unit’s price (Milestone Value);
(5) the total number of units (Project Quantities); (6) the total cost for all of the units (Total
Project Value); and (7) the percentage of the total EPC Agreement price that each component
represents (Total Project %):

Id. ¶ 13; ECF No. 88 at App’x 355.

Exhibit I’s second part contains a payment schedule for each EPC Agreement. Pitale
Decl. ¶ 14. As shown in the excerpt of the DS 300 EPC Agreement’s payment schedule below,
the schedules list each monthly payment Plaintiffs would make to First Solar (Total EPC
Payment – Monthly), as well as the aggregate amount Plaintiffs have paid to First Solar up to
each month (Total EPC Payment – Cumulative):

5
Ms. Pitale is NextEra’s Senior Director of State Tax Policy. At the time of the Desert Sunlight
Transaction, she was a Senior Director of Tax at NextEra. She supervised the preparation of the
Section 1603 grant at issue in this case.

5
Id.; ECF No. 88 at App’x 359. Each monthly payment amount was simply calculated by
multiplying the number of component units expected to be completed that month by the price of
each component unit. See Brannen Dep. Tr. at 76:3-8 (ECF No. 81-6 at App’x 230).

In negotiating the EPC Agreements, Mr. Brannen states that he “was careful to ensure
that the price for the three EPC contracts standing alone was a fair-market price for the EPC
work that First Solar performed.” Brannen Decl. ¶ 24 (ECF No. 86-3 at App’x 12). Mr.
Brannen explains that he had multiple reasons to do so beyond pursuing NextEra’s normal
business objective of paying only fair market value for EPC agreements. Id. First, Mr. Brannen
states he knew that NextEra would require loans to pay for the EPC Agreements, and that banks
would not finance the EPC Agreements if they were above market. Id. Second, due in part to
the magnitude of the Facility’s construction, Mr. Brannen states he feared that First Solar may
not complete the Project, and that NextEra did not want to have paid First Solar an above-market
amount for any work already completed. Id. Third, Mr. Brannen states that he had a personal
incentive to reach the best price and lowest risk EPC Agreements because a significant portion of
his salary was an incentive payment tied to the deal reached. Id. ¶ 25. Mr. Draper also states
that First Solar was “trying to stay cash neutral or cash positive throughout the course of
construction for the project.” Draper Dep. Tr. at 252:8-12 (ECF No. 81-63 at App’x 4205).

The Parties also took several steps in assessing the EPC Agreement prices.

First, NextEra performed its own “cost build up,” which was reviewed by an independent
engineering firm, BEW. Brannen Decl. ¶ 26e. Excluding the cost of the EPC Agreements’
module and cartridge components, NextEra estimated the cost of the remaining components,
known as the balance of system (“BOS”) component cost, under the EPC Agreements to be
$1.04/Wdc, which BEW believed was low. Id. NextEra then compared its estimated $1.04/Wdc
BOS price to First Solar’s proposed BOS price of $1.21/Wdc and found that the prices were
similar. Id. Next Era viewed the $0.17/Wdc difference between the two prices as merely “a
reasonable markup for sales tax and for the overhead, profit and risk contingency that an EPC
contractor like First Solar would charge.” Id. Based on its knowledge of the solar modules
market, NextEra further determined that the EPC Agreement prices for modules and cartridge
were also fair market value, and that their percentage of the cost of the EPC Agreements was
standard as well. Id. ¶¶ 26e, 53-56. Accordingly, NextEra concluded that the EPC Agreement
prices were fair market value. Id. ¶ 26e.

6
 Second, First Solar, at the direction of Goldman Sachs (one of NextEra and GE’s lenders
funding the EPC Agreements), requested that an independent engineering firm, Black & Veatch
Corporation (“Black & Veatch”), assess the Project. ECF No. 88 at App’x 257. Black & Veatch
concluded that “the total construction cost is consistent with the total construction cost of other
PV solar generating facilities utilizing the same technology and with our understanding of
current market pricing in the solar PV industry.” Id. at App’x 258.

Third, NextEra consulted Treasury guidance “reflect[ing] that a cost of $4/W to build
the . . . property for a utility-scale solar PV facility was presumptively reasonable.” Brannen
Decl. ¶ 26d (ECF No. 86-3 at App’x 14). Specifically, the guidance provided that, as of the first
quarter of 2011, its benchmark price to construct such facilities was $4 per watt (“/W”).6 Id.; see
also ECF No. 88 at App’x 250. According to the guidance, the benchmarks reflect a property’s
fair market value and “are predicated on an open-market, arm’s-length transaction between two
entirely unrelated parties with adverse economic interests.” ECF No. 88 at App’x 249-50. In
contrast to the $4/W price, Mr. Brannen calculated that the EPC Agreements’ total price of over
$1.9 billion “translates to a price of $2.69 per watt of DC capacity based on the facility’s original
planned capacity of 724 MWdc, . . . $2.63/Wdc based on actual as-built capacity of 740.7
MWdc[, and,] [i]ncluding net early completion bonuses . . . , $2.73/Wdc based on Desert
Sunlight’s as-built capacity of 740.7 MWdc.” Brannen Decl. ¶ 21. In contrast to Mr. Brannen’s
calculations, Ms. Angela Pitale, NextEra’s Senior Director of Tax from 2011-2015, calculated a
price of $2.77/Wdc. Pitale Decl. ¶ 59 (ECF No. 88 at App’x 315).

Fourth, Mr. Brannen was aware that various public reports estimated the EPC agreement
prices of typical utility-scale solar PV facilities, and that his estimated prices of $2.63/Wdc or
$2.73/Wdc for the Facility’s EPC Agreements was lower than all of them. Brannen Decl. ¶ 26c.
For example, the Solar Energy Industry Association’s (“SEIA”) 2011 First Quarter U.S. Solar
Market Insight publication estimated the average utility-scale solar PV facility system price in
the first quarter of 2011 to be $3.85/W. ECF No. 88 at App’x 222. And a National Renewable
Energy Laboratory’s (“NREL”) February 2011 report estimated a utility-scale solar PV facility
system price at the end of 2010 to be approximately $4/Wdc. Id. at App’x 239.

Apparently satisfied that the Facility’s total cost was reasonable, the Parties signed the
EPC Agreements. Plaintiffs and First Solar also entered into three O&M Agreements, under
which Plaintiffs hired First Solar to provide operations and maintenance services for the 250
MWac portion, the 300 MWac portion, and the common facilities of the Facility, respectively.
See ECF Nos. 81-16, 81-17, & 81-18.

2. Financing Agreements and the DOE Loan Guarantee

As part of its development of the Project, First Solar arranged Project financing terms
with several private banks. ECF Nos. 91-3, 91-4, 95-2 at App’x 2892-2900. At the time that
Plaintiffs acquired the Project, they also entered into financing agreements with these banks on
the terms First Solar arranged for loans of $1.461 billion. Brannen Decl. ¶ 94; Scarff Decl. ¶ 2

6
Mr. Brannen clarifies that “[i]t is unclear whether the Treasury benchmark number is $/Wdc or
$/Wac, but, either way, the Desert Sunlight number was lower.” Brannen Decl. ¶ 26d.

7
(ECF No. 95-2 at App’x 2865).7 According to Plaintiffs, they paid slightly less than $87 million
in interest on the loans during the construction of the Facility. Pitale Decl. ¶ 88. NextEra and
GE used equity to finance more than $600 million that they did not finance through the loans.
Brannen Decl. ¶ 94; ECF No. 81-60 at App’x 3982.

The Department of Energy also provided a loan guarantee8 covering 80% of the loans—
$1,168,800,000—in case Plaintiffs defaulted. Brannen Decl. ¶ 94. The DOE Loan Guarantee
was made pursuant to Section 1705 of Title XVII of the Energy Policy Act of 2005, Pub. L. No.
109-58, § 1705, 119 Stat. 594, which the American Recovery and Reinvestment Act of 2009
added to Title XVII to create a temporary program permitting DOE to issue loan guarantees in
support of renewable energy projects. ECF No. 81-65 at App’x 2571.

Before providing the guarantee, DOE took multiple steps to assess First Solar’s total EPC
price, which was then over $2.36 billion. DOE conducted its own due diligence and determined
that there was a reasonable assumption that, based on the proposed total EPC Agreement price,
the loan would be repaid. Walker Dep. Tr. at 99:8-14 (ECF No. 88-1 at App’x 2816).9 It did so
by comparing the proposed total EPC Agreement price to other projects in its internal databases.
Id. at 103:16-17. DOE then engaged in interagency discussions with Treasury, which was
concerned that the proposed total EPC Agreement price was too high in comparison to several
other projects. Id. at 44:18-45:11, 59:2-8. Specifically, in June 2011—before the more than
$400 million reduction from First Solar’s proposed total EPC Agreement price and before the
Facility was built with a 740.7/MWdc generating capacity—the total EPC Agreement price was
proposed by First Solar at $2.97/Wdc (including labor adjustments), and Treasury indicated that
a comparable EPC agreement price would be $2.80/Wdc without labor adjustments and
$2.92/Wdc with labor adjustments. Id. at 154:21-158:3; see also ECF No. 88-1 at 2863. To
assuage Treasury’s concerns, DOE consulted numerous experts for their assessment of the
proposed total EPC Agreement price. Walker Dep. Tr. at 46:17-47:17.

The Loan Programs Office (“LPO”), which provides DOE’s loan guarantees, requested
the analysis from DOE’s Solar Energy Technology Program (“SETP”). ECF No. 88-1 at App’x
2425; Ramesh Dep. Tr. at 23:21-25:12 (ECF No. 88-1 at App’x 2364).10 In response to the
LPO’s request, SETP coordinated with the NREL, one of the seventeen national labs within
DOE, to conduct its review using the NREL’s solar PV cost estimating tool—bottom-up cost
modeling. Ramesh Dep. Tr. at 17: 13-16, 25:1-17. Bottom-up cost modeling is a method used to
determine a project’s cost based on its various components, which in turn is based on market

7
During the negotiations regarding the Desert Sunlight acquisition, Mr. Scarff was a project
manager and was involved in those negotiations.
8
In fact, there were multiple loan guarantees that cover various loans. Scarff Decl. ¶ 6. Because
the disputes involving the guarantees only address them in the aggregate, the Court does so as
well.
9
Mr. Walker is an employee of the U.S. Department of Energy in the Loan Programs Office. He
was involved in the Department’s due diligence regarding the loan guarantee for Desert Sunlight.
10
Dr. Ramesh formerly served in the Department of Energy and performed or oversaw the due
diligence of the Desert Sunlight Transaction.

8
surveys in consultation with industry stakeholders. Ramesh Dep. Tr. at 19:24-20:12, 34:18-35:3.
Upon completion of the SETP/NREL review, the SETP’s Program Manager, Dr. Ramesh, issued
a memorandum to the LPO that concluded that First Solar’s proposed total EPC Agreement price
is “below the current estimated average total project cost for comparable projects and [is] within
a reasonable range of the expected system price, as estimated using NREL’s internal solar PV
cost estimating tool” (i.e., the bottom-up cost model). ECF No. 88-1 at App’x 2425; Ramesh
Dep. Tr. at 25:13-17; see also ECF No. 88-1 at App’x 2420 (another DOE memorandum
concluding that “[b]ased on this analysis the EPC price falls within an acceptable range of
expected system costs based on technology selection, system size and configuration.”).
Specifically, the NREL determined that First Solar’s then-proposed total EPC Agreement price
of $3.43/Wdc was within a reasonable range of the NREL’s estimated EPC system price for a
project similar to the Desert Sunlight Facility of $2.93/Wdc. ECF No. 88-1 at App’x 2427-28;
Ramesh Dep. Tr. at 129:16-21. Dr. Ramesh explains that the memorandum compared the
proposed total EPC Agreement price to the NREL’s 2010 Q4 Benchmark of $4.13/Wdc, not the
NREL’s estimated price of $2.93/Wdc. Ramesh Dep. Tr. at 40:18-41:5. Further, Dr. Ramesh’s
memo states that “[w]e have compared a detailed break out of the pricing for the elements of the
EPC contract against our model and have found the pricing to be within our expected range.”
ECF No. 88-1 at App’x 2426.

In addition to its internal resources, DOE consulted three outside, independent
engineering firms to assess the proposed total EPC Agreement price. Walker Dep. Tr. 46:17-
47:17 (ECF No. 88-1 at App’x 2803). The first firm was Luminate, LLC (“Luminate”), which
concluded that both First Solar’s initial EPC Agreement price and the final EPC Agreement price
were reasonable. Grover Dep. Tr. at 17:20-18:7, 79:20-81:18 (ECF No. 81-36 at App’x 3181-82,
3197).11 The second firm, Burns and Roe Enterprises, Inc., also concluded that the proposed
EPC Agreement prices were reasonable. Walker Dep. Tr. at 53:16-22, 58:16-59:18 (ECF No.
88-1 at App’x 2804, 2806). The third firm DOE consulted was Shaw Consultants International,
Inc. Id. at 69:9-71:6, 75:23-76:4; ECF No. 88-1 at App’x 2556-76. Shaw defined a “fair market
value” price as “one where the price is comparable to an EPC price that was obtained through a
competitive third party EPC solicitation for comparable PV projects.” Id. at App’x 2560. Based
on its analyses, Shaw determined that the Facility’s total EPC Agreement price was the cheapest
of seventeen solar facility EPC prices considered. Id. at App’x 2573. Shaw thus concluded that
the EPC Agreement prices were fair market values. Id. at App’x 2576. In addition, Shaw
concluded that the Facility’s EPC Agreement price was “comparable to ‘arm’s length’ or third
party EPC prices.” Id.; see also Walker Dep. Tr. at 92:4-94:18.

After consulting these experts, DOE ultimately issued the Loan Guarantee.

3. The MIPSA

Lastly, First Solar, NextEra, and GE also executed the MIPSA, through which First Solar
sold its entire interest in Plaintiffs to NextEra and GE. ECF No. 81-10 at App’x 333. Included
in the sale was the LGIA, the two PPAs, and the DOE Loan Guarantee. Id. Under the MIPSA,

11
Ms. Grover is a senior vice president of Luminate.

9
NextEra and GE each paid First Solar $7,222,803, for a total price of $14,445,606. Id. at App’x
351.

The EPC Agreement and MIPSA prices were arrived at using a “Project Model,” an
intricate, multi-tab Excel spreadsheet aimed at calculating the Transaction’s prices based on the
Facility’s expected return on equity (“ROE”) for the purchasers (NextEra and GE). See id. at
App’x 351-52; see also ECF No. 81-25 (Project Model Cover Sheet) and related CD (Project
Model). As explained in NextEra’s summary of the Facility’s transaction, the Project Model
calculates the expected ROE by considering various factors, including “expected debt terms,
plant performance (energy output, availability during the O&M phase), interconnection
energization date, energization dates for the 20 blocks (ramp revenues), substantial completion
date and PPA commercial operation dates (COD), future operating, certain owner costs and
future maintenance expenses.” ECF No. 81-19 at App’x 2013. The MIPSA sets forth the
limited conditions and procedures warranting revising the input into the Project Model, see ECF
No. 81-10 at App’x 351-53, as otherwise “the calculations embedded in the Project Model will
not be changed or revised for any reason, including the discovery of a mathematical or formulaic
error in the Project Model,” id. at App’x 352.

Following the execution of all the agreements, First Solar began constructing the Facility.
Given the scale of the Project, First Solar constructed the Facility in phases, known as “blocks.”
Pitale Decl. ¶ 16. The Facility was constructed in twenty blocks. Id. Blocks 1-11 comprised the
300MWac portion and Blocks 12-20 comprised the 250 MWac portion. Id.

C. The American Recovery and Reinvestment Act of 2009

“Congress has long used tax incentives to promote investment in new renewable energy
projects.” Alta Wind I Owner Lessor C v. United States, 897 F.3d 1365, 1368 (Fed. Cir. 2018).
“Initially, these incentives came in the form of tax credits—specifically the production tax
credit . . . under I.R.C. § 45 and the investment tax credit . . . under I.R.C. § 48.” Id. By 2009,
however, the Great Recession had reduced the benefit of these tax credits. Id. In response to
economic problems, Congress passed the American Recovery and Reinvestment Act of 2009
(“ARRA”). Pub. L. No. 111-5, 123 Stat 115 (2009). “The ARRA created a temporary program
that offered a cash payment in lieu of a tax credit to certain investors for certain qualified
investments in clean energy property.” WestRock Virginia Corp. v. United States, 136 Fed. Cl.
267, 270 (2018), aff'd, 941 F.3d 1315 (Fed. Cir. 2019). Specifically, Section 1603(a) provides
that “[u]pon application, the Secretary of the Treasury shall . . . provide a grant to each person
who places in service specified energy property to reimburse such person for a portion of the
expense of such property.” “Specified energy property” includes certain solar property. §
1603(d)(3). And the amount of the cash grant is the “applicable percentage of the basis of such
property,” which is 30% for solar property. § 1603(b)(1)-(2). “It is intended that the grant
provision mimic the operation of the credit under [IRC] section 48.” WestRock Virginia Corp.,
941 F.3d at 1316 (quoting H.R. Rep. No. 111-16, at 620–21 (2009)) (alteration in original).

D. Plaintiffs’ Section 1603 Applications

As First Solar constructed and placed into service the blocks of the Facility, Plaintiffs
submitted applications (fifteen in total) to Treasury for cash grants under Section 1603. Pitale

10
Decl. ¶ 16. Plaintiffs’ fifteen applications concerned a total of approximately $2.13 billion in
expenses incurred to place into service the Facility, of which Plaintiffs contend $2,049,419,165
is Section 1603-eligible. Id. ¶ 17. According to Plaintiffs, the almost $2.05 billion was mainly
comprised of the costs under the EPC Agreements, including sales tax. See id. ¶ 22.

In addition to the EPC costs, Plaintiffs concluded that certain other costs were Section
1603-eligible. First, Plaintiffs assert that the almost $87 million in interest paid to lenders during
the construction of the Facility is Section 1603-eligible. Id. ¶ 89. Second, they also contend that
approximately $72 million in net bonuses paid to First Solar for early completion of certain
blocks of the Project is 1603-eligible. Brannen Decl. ¶¶ 63-64. And third, Plaintiffs argue that
additional miscellaneous costs including reimbursements to First Solar under the MIPSA for
development costs it incurred before transacting with NextEra and GE are Section 1603-eligible.
Brannen Decl. ¶ 67. In sum, Plaintiffs’ applications claimed entitlement to a Section 1603 cash
grant of $614,825,750 (i.e., the 30% applicable percentage of the $2,049,419,165 eligible basis).
ECF No. 1 ¶ 45.

With each of their applications, Plaintiffs submitted a cost segregation of that
application’s expenses. Pitale Decl. ¶ 20. A cost segregation is a document that assigns
expenses to distinct categories. O’Connell Decl. ¶ 6 (ECF No. 86-6 at App’x 1249).12 Plaintiffs’
cost segregations categorized their application’s expenses among direct costs and indirect costs
(costs necessary to acquire and install tangible property) and between claimed Section 1603
grant-eligible and grant-ineligible expenses. Pitale Decl. ¶ 34. In addition, after submitting the
fifteen applications and accompanying cost segregations, Plaintiffs submitted to Treasury a
single document in Excel format, the Total Cost Spreadsheet, which comprises all the cost
segregations on individual tabs. Id. ¶¶ 46-47. The Total Cost Spreadsheet also has a tab called
the Total Project Cost Summary, which is the summary page of the Total Cost Spreadsheet. Id. ¶
48. For each application, the Total Project Cost Summary presents the total cost reported on
each application and the total eligible cost Plaintiffs claimed on each application. Id. ¶¶ 48-49.
Moreover, the last row of the tab provides the total reported Project cost of $2,129,201,444.90
and the total claimed Section 1603-eligible cost for the Project of $2,049,419,162.17. Id. ¶ 49.

Plaintiffs worked with KPMG, LLP, and principally with Mark O’Connell of KPMG, to
create the cost segregations. O’Connell Decl. ¶ 4 (ECF No. 86-6 at App’x 1249). Mr.
O’Connell explained that KPMG’s work preparing the cost segregations involved numerous
steps. KPMG first categorized costs between direct and indirect costs. Id. ¶¶ 26-33. Mr.
O’Connell segregated which direct costs were, in his opinion, Section 1603-eligible costs and
which were Section 1603-ineligible costs. Id. ¶ 34. But before determining whether the costs for
certain asset components were Section 1603-eligible, Mr. O’Connell had to first determine the
component costs because only the larger asset cost was known. For example, Mr. O’Connell
calculated the prices of components falling under the larger “substation” cost (the costs incurred
to construct the onsite substation at the Facility). Id. ¶ 41. Mr. O’Connell specifically allocated
the total price of the substation to its individual components based on benchmarking information
from other projects. Id. Mr. O’Connell similarly calculated the prices of components in the
larger cost category of “FNTP/Site Improvements,” which he divided into sub-categories labeled

12
Mr. O’Connell worked at KPMG and prepared the cost segregations for the Project.

11
“General Conditions” and “Buildings/Roads.” Id. ¶ 49. Primarily utilizing the Marshall & Swift
manual, which “contains price figures for various construction related expenses,” Mr. O’Connell
allocated the total price of the “FNTP/Site Improvements” to its individual components. Id. ¶¶
52-53. Reflecting on these analyses, Mr. O’Connell concluded that “[t]he cost analyses that my
team performed on the substation, the General Conditions site improvements category, and the
Buildings/Roads site improvements category covered over 99% of the ineligible direct EPC
costs, after excluding the gen-tie line, that Plaintiffs had incurred at the time of the analyses.” Id.
¶ 61. In addition, Mr. O’Connell determined that the “Gen-Tie” costs (for the construction of the
generation tie line, or transmission line, from the Facility’s on-site substation to a utility
company’s offsite substation) were Section 1603-ineligible costs. Id. ¶ 38.

Plaintiffs also hired Deloitte & Touche LLP to issue the accountant reports attesting to
the accuracy of the applications, which Treasury required. Pitale Decl. ¶¶ 30, 41. After
receiving all requested documentation, Deloitte examined the cost segregations. Id. ¶¶ 41-42.
Specifically, Deloitte tested “whether the claimed costs were in fact incurred and paid, and
whether the costs were properly classified as direct or indirect, and eligible or ineligible” in its
opinion. Id. ¶ 42. Deloitte concluded its examination by issuing an independent accountant’s
report for each application verifying it was “fairly stated, in all material respects.” Id.

After reviewing the applications, Treasury awarded Plaintiffs a Section 1603 cash grant
of approximately $555,506,250—$59,319,500 less than Plaintiffs’ requested amount of almost
$615 million, as broken down in the following table:

Block Installation Owner Grant Sought Grant Issued Difference
(Pre-Sequestration) (Pre-Sequestration)

Blocks 1, 2, 10, 11 DS 300 $148,798,318 $129,309,750 $19,488,568

Blocks 16, 17, 20 DS 250 $95,021,918 $77,347,500 $17,674,418

Block 3 DS 300 $41,149,028 $38,436,750 $2,712,278

Block 4 DS 300 $30,906,856 $29,014,500 $1,892,356

Block 12 DS 250 $41,478,228 $38,343,000 $3,135,228

Block 13 DS 250 $41,448,054 $38,530,500 $2,917,554

Block 14 DS 250 $28,149,884 $26,564,250 $1,585,634

Block 19 DS 250 $26,720,334 $26,160,000 $560,334

Block 18 DS 250 $30,329,825 $28,875,000 $1,454,825

Block 15 DS 250 $15,250,665 $14,392,500 $858,165

Block 5 DS 300 $25,935,287 $24,390,000 $1,545,287

Block 6 DS 300 $26,060,838 $24,397,500 $1,663,338

12
 Block 7 DS 300 $20,828,189 $19,680,000 $1,148,189

Block 8 DS 300 $19,475,260 $18,337,500 $1,137,760

Block 9 DS 300 $ 23,273,066 $21,727,500 $1,545,566

Total $614,825,750 $555,506,250 $59,319,500

See id. ¶ 122; see also ECF No. 1 ¶ 49.

II. Standard of Review

“The [C]ourt shall grant summary judgment if the movant shows that there is no genuine
dispute as to any material fact and the movant is entitled to judgment as a matter of law.” RCFC
56(a). “By its very terms, this standard provides that the mere existence of some alleged factual
dispute between the parties will not defeat an otherwise properly supported motion for summary
judgment; the requirement is that there be no genuine issue of material fact.” Anderson v.
Liberty Lobby, Inc., 477 U.S. 242, 247-48 (1986) (emphasis in original). A “genuine” dispute of
material fact exists where a reasonable factfinder “could return a verdict for the nonmoving
party.” Id. at 248. “Material” facts are those “that might affect the outcome of the suit under the
governing law,” as opposed to “disputes that are irrelevant or unnecessary.” Id. “When both
parties move for summary judgment,” moreover, “the [C]ourt must evaluate each party’s motion
on its own merits, taking care in each instance to draw all reasonable inferences against the party
whose motion is under consideration.” AT&T Advertising, L.P. v. United States, 147 Fed. Cl.
478, 482 (2020) (citing Mingus Constructors, Inc. v. United States, 812 F.2d 1387, 1391 (Fed.
Cir. 1987)).

In addition, as in tax refund cases, the Court reviews claims against the Government for
Section 1603-cash-grant reimbursements de novo. W.E. Partners II, LLC v. United States, 119
Fed. Cl. 684, 690 (2015).

III. Discussion

A. The Government’s Motion in Limine

Before reaching the motions for summary judgment, the Court must address the
Government’s motion in limine to exclude various testimonial evidence Plaintiffs submitted to
support their summary judgment briefing. ECF No. 92. Because none of the challenged
statements are necessary for the resolution of the cross motions for summary judgment, the Court
does not consider them.13 Therefore, the Government’s motion in limine is denied as moot
without prejudice to objecting at trial.

13
To be clear, the Court does cite a few paragraphs in which the Government challenges portions
of those paragraphs. The Court only refers to the unchallenged statements in those paragraphs.

13
 B. The Government’s Motion for Summary Judgment

The Government moves for summary judgment on the ground that Plaintiffs cannot meet
their burden of proof to establish entitlement to the Section 1603 grant they seek because they
have not produced (and cannot now produce) the required allocation of the price of the Desert
Sunlight Transaction14 under 26 U.S.C. § 1060 on IRS Form 8594. ECF No. 80 at 32. Plaintiffs
counter that they produced their allocation of the Transaction price, albeit not on IRS Form 8594,
which they argue contains far more information about their allocation than a Form 8594. ECF
No. 86-1 at 67. Plaintiffs also explain that they did not file a Form 8594 because at the time of
their filing they did not understand that Section 1060 governed the Desert Sunlight Transaction
when they originally applied for a Section 1603 grant. Id. at 67 n.9. Based on Alta Wind, which
was decided years later in 2018, all now agree that Section 1060 applies. See 897 F.3d at 1376.
And Plaintiffs note that the penalty for failing to file a Form 8594 is a mere $250 fine, making a
forfeiture of their claim now unreasonable. ECF No. 86-1 at 67 (citing 26 U.S.C. §§ 6721(a)(1),
6724(d)(1)(B)(xvii)).

Putting this aside, there are significant, material issues in dispute, which preclude
summary judgment for the Government.

1. The Legal Framework of an I.R.C. § 1060 Allocation.

I.R.C. § 1060 governs the purchaser’s cost basis in “applicable asset acquisition[s].”
I.R.C. § 1060(a). An “applicable asset acquisition” refers to any transfer “of assets which
constitute a trade or business.” I.R.C. § 1060(c)(1). As relevant here, Treasury Regulations
define a “trade or business” as a group of assets whose “character is such that goodwill or going
concern value could under any circumstances attach to such group.” 26 C.F.R. § 1.1060-
1(b)(2)(i)(B). “Goodwill is the value of a trade or business attributable to the expectancy of
continued customer patronage.” 26 C.F.R. § 1.1060-1(b)(2)(ii). “Going concern value is the
additional value that attaches to property because of its existence as an integral part of an
ongoing business activity.” Id. The regulations further provide that, in determining whether
goodwill or going concern value could attach, “all the facts and circumstances surrounding the
transaction are taken into account,” including:

(A) [t]he presence of any intangible assets . . . ; (B) [t]he existence
of an excess of the total consideration over the aggregate book
value of the tangible and intangible assets purchased . . . ; and (C)
[r]elated transactions, including lease agreements, licenses, or
other similar agreements between the purchaser and seller . . . in
connection with the transfer.

26 C.F.R. § 1.1060-1(b)(2)(iii).

14
The term “Desert Sunlight Transaction” or “Transaction” used in this section refers
specifically to the EPC Agreements and the MIPSA, as all the relevant Section 1603-eligible
assets were transferred from First Solar to NextEra and GE through these two agreements.

14
 For transfers of such applicable asset acquisitions, “the transferee’s basis in such assets is
determined wholly by reference to the consideration paid for such assets.” I.R.C. § 1060(c)(2).
Specifically, “the consideration received for such assets shall be allocated among such assets
acquired in such acquisition in the same manner as amounts are allocated to assets under section
338(b)(5).” I.R.C. § 1060(a). I.R.C. § 338(b)(5) provides that allocation is to be done according
to Treasury regulations. These regulations direct purchasers to “allocate the consideration under
the residual method as described in [26 C.F.R.] §§ 1.338-6 and 1.338-7 in order to determine . . .
the basis in[] each of the transferred assets.” 26 C.F.R. § 1.1060-1(a)(1); see also 26 C.F.R. §
1.1060-1(c)(2) (providing that the purchaser’s basis in the assets purchased in an applicable asset
acquisition requires allocating the consideration to all the assets purchased under the residual
method of 26 C.F.R. §§ 1.338-6 and 1.338-7).

Section 1.338-6 “set[s] out a method of allocation—the residual method—in which the
consideration is distributed among seven asset classes, some classes for tangible assets and
others for intangible assets.” Alta Wind, 897 F.3d at 1376. Those asset classes are:

Class I: Cash and general deposit accounts.

Class II: Actively traded personal property, certificates of deposits, U.S.
government securities and publicly traded stock.

Class III: Debt instruments [(including accounts receivable)].

Class IV: Inventory and other property held for sale to customers.

Class V: Assets that do not fit within any other class, including tangible property.

Class VI: I.R.C. § 197 intangibles, including contract rights, but not goodwill and
going concern value.

Class VII: Goodwill and going concern value.

Id. (citing 26 C.F.R. § 1.338-6(b)). “The consideration is allocated among these classes in the
order they are listed in a ‘waterfall’ fashion, using the fair market value of the assets within each
class.” Id. Also, “[t]he amount . . . allocated to an asset (other than Class VII assets) cannot
exceed the fair market value of that asset . . . .” 26 C.F.R. § 1.338-6(c)(1). Thus, the
consideration is first allocated to Class I assets up to their fair market value, then to Class II
assets up to their fair market value, and so on until the consideration is fully allocated. 26 C.F.R.
§ 1.338-6(b)(1)-(2).

Here, the Parties agree that I.R.C. § 1060 and the residual method apply to determine the
amount of a Section 1603 grant. ECF No. 80 at 21-23; ECF No. 86-1 at 15. The Circuit also
made this clear in Alta Wind. 897 F.3d at 1376. Specifically, the Transaction was an “applicable
asset acquisition” under I.R.C. § 1060 because the MIPSA and EPC Agreements transferred
assets (i.e., the Facility’s tangible assets, the PPAs, and the LGIA) that constitute a “trade or
business” whose “character is such that goodwill or going concern value could under any
circumstances attach to such group.” 26 C.F.R. § 1.1060-1(b)(2)(i)(B).

15
 Two of the three non-exhaustive factors identifying goodwill and going concern value
support this conclusion. See 26 C.F.R. § 1.1060-1(b)(2)(iii). First is the presence of intangible
assets. 26 C.F.R. § 1.1060-1(b)(2)(iii)(A). The PPAs and the LGIA are intangible assets, as
discussed below. See infra III.B.; see also Alta Wind, 897 F.3d at 1373-74 (recognizing that at
least some portion of the PPAs and the interconnection agreement are intangible assets). Second
is the presence of related transactions. 26 C.F.R. § 1.1060-1(b)(2)(iii)(C). Here, the O&M
Agreements are related transactions between the purchaser (NextEra and GE) and seller (First
Solar) in connection with the transferred assets via the EPC Agreements and the MIPSA because
the O&M Agreements provide for operations and maintenance services to the Facility by First
Solar. See ECF Nos. 81-16, 81-17, & 81-18; see also Alta Wind, 897 F.3d at 1373 (finding that
leaseback agreements for the seller to operate the assets are related transactions).

The fact that the Facility was not yet operational at the time of the Transaction makes no
difference. See Alta Wind, 897 F.3d at 1373 (citing 26 C.F.R. § 1.1060-1(b)(2)(iii)) (“There is
no need to show that a transaction had actual, accrued goodwill or going concern value at the
time of the transaction.”). That is particularly true here given the Parties’ expected ROE from
the Facility as reflected in the “Project Model” that formed the basis of the EPC Agreement and
MIPSA negotiated prices. See ECF No. 81-10 at App’x 351-53; see also ECF No. 81-25
(Project Model Cover Sheet) and related CD (Project Model); Alta Wind, 897 F.3d at 1373, 1375
(finding that goodwill could attach to assets yet to be in operation that were negotiated based on
anticipated cash flows that were valued with reference to intangible assets such as PPAs).
Accordingly, because the Desert Sunlight Transaction is an applicable asset acquisition, the
I.R.C. § 1060 allocation applies in this case.

2. Plaintiffs’ I.R.C. § 1060 Allocation.

Given I.R.C. § 1060’s application to this case, the Government moves for summary
judgment on the ground that Plaintiffs have not provided the required I.R.C. § 1060 allocation of
the purchase price of the Desert Sunlight Transaction in accordance with the fair market value of
the assets acquired. ECF No. 80 at 28-29, 32, 36-37. The Government primarily focuses on the
fact that Plaintiffs did not complete an IRS Form 8594 for their I.R.C. § 1060 allocation, and that
their failure to do so precludes their eligibility for a further Section 1603 grant. Id. at 32; ECF
No. 91 at 35. The Court disagrees.

a) Plaintiffs’ failure to submit an IRS Form 8594 is not fatal to their
claim.

The contention that IRS Form 8594 is the sole method to submit an I.R.C. § 1060
allocation for Section 1603 purposes is without merit. The Government has cited no authority to
support this proposition, and the Government itself has even implied that an I.R.C. § 1060
allocation could take other forms than Form 8594. According to the Government, “plaintiffs
have not presented a copy of a filed Form 8594, or otherwise provided the government with the
required allocation of the purchase price in the Desert Sunlight Transaction in accordance with
the fair market value of the assets acquired.” ECF No. 80 at 32 (emphasis added). Indeed, it is
particularly difficult to accept that IRS Form 8594 is the only way to provide an I.R.C. § 1060
allocation given the limited information it requires. The Form merely contains boxes to fill in:
(1) each asset class’s aggregate fair market value; (2) each asset class’s allocation of the sales

16
price; (3) the assets’ total aggregate fair market value; and (4) the assets’ total allocation of the
sales price. See ECF No. 81-32 at App’x 2841-42. Should an applicant for a Section 1603 grant
provide the same or more detailed information as its I.R.C. § 1060 allocation, the Government
has offered no reason why it should be rejected for failing to use IRS Form 8594. Indeed, the
penalty for failing to submit a Form 8594 is $250, not forfeiture of a claim. See 26 U.S.C. §§
6721(a)(1), 6724(d)(1)(B)(xvii).

It is also difficult to accept the Government’s argument that it cannot determine whether
Plaintiffs were entitled to any Section 1603 grant, much less the amount they seek in this action,
because they did not submit an IRS Form 8594. ECF No. 80 at 32. If it truly thought Plaintiffs’
Section 1603 applications were fatally deficient, it is unfathomable that the Government would
have approved and paid more than $555 million of Plaintiffs’ grant requests already. The
Government conducts an extensive review of each Section 1603 application before approving
them. See ECF No. 88 at App’x 249-51. Therefore, the Government’s approving most of
Plaintiffs’ applications suggests that the Government was, and presumably is, able to determine
Plaintiffs’ Section 1603 eligibility.

During oral argument, the Government attempted to reconcile this contradiction,
explaining that, under Section 1603, “the idea was to get this money out quick because the
financing for alternative energy products -- projects was drying up” and “Treasury, you know,
had to use different rubrics to determine which cases they were going to pay more attention to
and which cases they weren’t, [asking] . . . are we comfortable enough making some sort of
payment in the neighborhood without actually making the taxpayer go through the allocation that
is needed to be done?” ECF No. 106 at 17:21-23, 17:25-18:2, 18:6-9.

But it appears that there was back-and-forth between the Government and Plaintiffs in
which the Government had the opportunity to, and did, seek additional information from
Plaintiffs while it was considering their Section 1603 applications. See ECF No. 81-28
(Plaintiffs’ responses to Government questions and requests for additional documentation). The
Government’s requests included: “We are unable to find this [document] . . . please provide
another copy”; “Please provide the following original Excel financial models”; and “Using best
efforts, please provide a current estimate and cost segregation of the expected total cost of the
entire Desert Sunlight project . . . [and] an excel version of the September 20, 2011, Cost
Breakdown Study prepared by KPMG.” Id. at App’x 2637-38. Notably missing from the
Government’s requests for additional information is any request for a completed IRS Form 8594,
which indicates the Government did not need that form to determine Plaintiffs’ eligibility for a
Section 1603 grant. Nor does the Court.

The cases the Government relies upon do not hold to the contrary. As an initial matter,
nobody disputes that the Plaintiffs bear the burden of proof to show that they are entitled to an
additional § 1603 payment. ECF 80 at 28 (citing W.E. Partners II, LLC v. United States, 119
Fed. Cl. 684, 690 (2015); WestRock Va. Corp., 136 Fed. Cl. at 284); see also ECF No. 87 at 32
(Plaintiffs acknowledging what they must show). The Government also relies heavily on
WestRock for the proposition that “[i]f plaintiffs fail to put forward sufficient evidence of an
appropriate allocation of the overall purchase price for the Desert Sunlight enterprise as between
eligible and ineligible property that would demonstrate entitlement to an additional § 1603
entitlement, then plaintiffs should not recover any additional payment.” ECF No. 80 at 28 (citing

17
WestRock Va. Corp., 136 Fed. Cl. at 283-84) (emphasis in original). True enough, but the
Government attempts to stretch this proposition to support summary judgment because
“plaintiffs have not presented a copy of a filed Form 8594, or otherwise provided the government
with the required allocation of the purchase price in the Desert Sunlight Transaction in
accordance with the fair market value of the assets acquired.” ECF No. 80 at 32. Whether
Plaintiffs provided a sufficient § 1060 allocation is a factual matter for trial.

b) A genuine dispute of material fact exists whether Plaintiffs
provided a sufficient I.R.C. § 1060 allocation.

Summary judgment is not appropriate here because there are disputed material facts
regarding whether Plaintiffs provided a sufficient I.R.C. § 1060 allocation. Plaintiffs contend
that they provided a proper and sufficient allocation in the cost segregations and Total Cost
Spreadsheet they submitted to Treasury with their Section 1603 applications. E.g., ECF No. 86-
1 at 63-64. The Government argues that Plaintiffs’ cost segregations and Total Cost Spreadsheet
are not a proper I.R.C. § 1060 allocation. ECF No. 91 at 7. The Government also argues that
Plaintiffs have failed to otherwise produce evidence regarding the fair market value of the assets,
which dooms their claim. Id. at 13-14, 18-21. The Court disagrees. The Government’s
argument that Plaintiffs failed to produce any evidence of the fair market value of the Facility
ignores all the evidence that Plaintiffs produced on that very point. Indeed, the Government
spends nearly a dozen pages explaining why various pieces of Plaintiffs’ evidence of the
Facility’s fair market value are insufficient, standing alone, to establish the fair market value.
ECF No. 80 at 32-37; ECF No. 91 at 13-14, 18-21. But there is no requirement that the Court
view each piece of information in isolation. And Plaintiffs have clearly provided enough
information to create a factual dispute as to the fair market value of the Facility.

(1) The Government’s argument regarding Class III assets
fails.

The Government points to the financing agreements as showing that Plaintiffs did not
provide a sufficient I.R.C. §1060 allocation. Specifically, the financing agreements that the
Plaintiffs secured from private banks are Class III assets, according to the Government, yet the
Plaintiffs never valued or allocated any portion of the consideration to these assets. ECF No. 91
at 8-9. The Government is mistaken, however, for at least two reasons.

First, I.R.C. § 1060 allocates only the consideration for assets acquired in a transaction.
I.R.C. § 1060(a)(2) (providing that “the consideration received for such assets shall be allocated
among such assets acquired in such acquisition”) (emphasis added). Plaintiffs, however, did not
acquire the loan agreements from First Solar in the Desert Sunlight Transaction. Rather,
Plaintiffs executed them separately with private financial institutions. ECF No. 95 at 5.
Plaintiffs executed the “A-2 Loan Agreement” with the U.S. Department of Energy and
Deutsche Bank on September 29, 2011, contemporaneously with the Transaction. See ECF No.
91-5. First Solar is not party to it to this agreement. The same is true of the other financing
agreements. See Scarff Decl. ¶ 3 (ECF No. 95-2 at App’x 2866). Therefore, the Transaction’s
consideration is not allocated to the loan agreements under I.R.C. § 1060.

18
 Second, even if Plaintiffs acquired the agreements in the Transaction, the fact is that they
are not assets. Rather, they are liabilities reflecting amounts owed to private banks. I.R.C. §
1060, however, only allocates the consideration for assets acquired in a transaction. See I.R.C. §
1060(a) (providing that “the consideration received for such assets shall be allocated among such
assets acquired in such acquisition in the same manner as amounts are allocated to assets under
section 338(b)(5).”) (emphasis added). Indeed, Class III property is specifically referred to as
assets. See 26 C.F.R. § 1.338-6(b)(2)(iii) (“Class III assets. Class III assets are assets . . . .”)
(emphasis added). Thus, the financing agreements are not included in the Desert Sunlight
Transaction for allocation under I.R.C. § 1060 as Class III assets or otherwise, and evidence of
their fair market value is unnecessary for the requisite I.R.C. § 1060 allocation for Plaintiffs’
sought Section 1603 relief.

The Government also argues that Plaintiffs failed to provide a sufficient I.R.C. § 1060
allocation because they have not valued or allocated consideration to another purported Class III
asset: the financing-related agreements First Solar entered with private banks. ECF No. 91 at 8-
9. But the regulations defining Class III assets as including “debt instruments” does not define
that term. 26 C.F.R. § 1.338-6(b)(2)(iii)(A). While undefined in the regulation, a Class III “debt
instrument” appears to mean an instrument under which a debt is owed because the Code defines
a “debt instrument” as “a bond, debenture, note, or certificate or other evidence of indebtedness.”
I.R.C. § 1275(a)(1)(A) (emphasis added). But neither First Solar nor Plaintiffs owe any debts
under First Solar’s financing-related agreements. Instead, one of these documents is a 2009
letter from Goldman Sachs to First Solar confirming that Goldman would serve as the structuring
agent for future financing agreements. ECF No. 91-4. The Government also points to ECF No.
91-3 as a Class III debt instrument, but that document is merely an engagement letter between
First Solar, Desert Sunlight, Goldman, and Citigroup under which Goldman and Citigroup would
serve as “joint lead arrangers and joint bookmakers,” and provide “structuring assistance” for
future financing. Id. at App’x 4341, 4343. Finally, the Government points to a conditional loan
commitment letter between Citibank and First Solar in which Citibank offered to provide
financing if certain conditions precedent were met. ECF No. 95-2 at App’x 2892-96. Among
the conditions precedent is “the preparation, execution and delivery of mutually acceptable
Credit Documents,” First Solar’s provision of all “know your customer” documents, and
Citibank’s verification of the information received. See id. at App’x 2893. Here too, there is
nothing in this document creating an obligation or indebtedness. Therefore, First Solar’s
financing-related agreements are not debt instruments and thus not Class III assets for purposes
of Plaintiffs’ I.R.C. § 1060 allocation.

This comports with the apparent understanding the parties, including the Government,
had of liabilities in the Alta Wind litigation. In Alta Wind, the parties stipulated that one of the
transactions involved the purchaser’s assumption of $307,147,911.69 of liabilities from the
seller. Stipulation of Facts ¶ 22, Alta Wind I Owner Lessor C v. United States, No. 13-402 (Fed.
Cl. Apr. 15, 2016) (ECF No. 115). Yet, when Alta Wind was on appeal before the Federal
Circuit, “the parties agree[d] that none of the assets at issue in this case fits within Class I, II, III,
or IV.” 897 F.3d at 1376. While there may be distinctions between those liabilities and the ones
before this Court, it does not appear that the common understanding of Class III assets includes
assumed liabilities (assuming, counterfactually, that the financing agreements were acquired
through the Transaction).

19
 As an alternative, the Government argues that if the financing agreements are not Class
III assets, the agreements are intangible assets for which Plaintiffs paid First Solar under the
MIPSA. ECF No. 106 at 37:10-16, 39:11-25. Further, the Government argues that, if not Class
III assets, the agreements have no asset class to fall under except Class V. Id. at 37:23-38:1,
39:15-25. Asserting that Plaintiffs did not value the agreements, the Government argues that
Plaintiffs’ I.R.C. § 1060 allocation fails. Id. at 37:23-38:1. The Court disagrees. First Solar’s
financing-related agreements are intangible assets, they are not Class V assets but rather
supplier-based intangible assets under I.R.C. § 197 (“Section 197”), and Section 197 assets fall
under Class VI. 26 C.F.R. § 1.338-6(b)(2)(vi). “A supplier-based intangible is the value
resulting from the future acquisition, pursuant to contractual or other relationships with suppliers
in the ordinary course of business, of goods or services that will be sold or used by the taxpayer.”
26 C.F.R. § 1.197-2(b)(7)(i). First Solar’s financing-related agreements are supplier-based
intangible assets because they form a contractual relationship with private banks in the ordinary
course of business regarding potential future loans. See also I.R.S. Tech. Adv. Mem. 199909002
(March 5, 1999) (concluding that an agreement under which a taxpayer pays a fee for the
subsequent sale of customer notes to company is a supplier-based Section 197 intangible asset).
As discussed below, Class VI assets need not be valued to perform the requisite I.R.C. § 1060
allocation unless there is a portion of the transaction price remaining unallocated after allocation
to Classes I-V. See infra III.B.3. Thus, even if Plaintiffs did not value First Solar’s financing-
related agreements, that would not affect Plaintiffs’ I.R.C. § 1060 allocation if Plaintiff is correct
that the entire Transaction price is allocated to Class V assets.

Accordingly, the Transaction had no Class III assets to which any of the Transaction’s
consideration had to be allocated under I.R.C. § 1060.

(2) A genuine dispute of material fact exists whether the cost
segregations and total cost spreadsheet contain the Class V assets’
fair market values.

Whether Plaintiffs provided a sufficient I.R.C. § 1060 allocation hinges on whether they
submitted the fair market values of the Class V assets to Treasury. Here, there is a genuine
dispute of material fact because Plaintiffs submitted detailed cost segregations to Treasury with
their Section 1603 applications, which they claim contain the fair market values of the Class V
assets. Whether these prices are, in fact, the fair market prices “is a question of pure fact . . . .”
Pabst Brewing Co. v. Comm’r, 69 T.C.M. (CCH) 2773 (1995), 1995 WL 325875, at *22.

Courts have long recognized that “[f]air market value is the price at which property
would change hands between a willing buyer and a willing seller, neither being under any
compulsion to buy or sell, and both reasonably informed as to all relevant facts.” Solitron
Devices, Inc. v. Comm’r, 80 T.C. 1, 20-21 (1983), aff’d 744 F.2d 95 (11th Cir. 1984); see also,
e.g., Innovair Aviation, Ltd. v. United States, 83 Fed. Cl. 498, 501 (2008) (“The Supreme Court
has held that fair market value is the amount that ‘a willing and informed buyer . . . under no
compulsion to buy, will pay, and what a willing and informed seller, under no compulsion to sell,
will accept.’”) (emphasis omitted) (citation omitted); United States v. Cartwright, 411 U.S. 546,
551 (1973) (“The willing buyer-willing seller test of fair market value is nearly as old as the
federal income, estate, and gifts taxes themselves[.]”). Courts generally accept that in “a sale
between unrelated parties at arm’s length[,] . . . the sales price is the best evidence of fair market

20
value.” Solitron Devices, 80 T.C. at 20; see also, e.g., Suitum v. Tahoe Reg’l Plan. Agency, 520
U.S. 725, 741-42 (1997) (recognizing the benefit of arm’s length prices in determining market
value); Schonfeld v. Hilliard, 218 F.3d 164, 178 (2d Cir. 2000) (“[I]t is well-established that a
recent sale price for the subject asset, negotiated by parties at arm’s length, is the ‘best evidence’
of its market value.”) (citations omitted). Thus, Plaintiffs contend that the actual amounts they
paid for each asset conveyed in the Transaction are the fair market values for each of those
assets.

The Government does not dispute that the willing-buyer, willing-seller standard
determines fair market value, but it contends that the standard has a particular meaning. Namely,
the Government maintains that “this standard requires an analysis of the property from the
perspective of a hypothetical buyer and a hypothetical seller, not the actual parties to the
transaction.” ECF No. 91 at 13 (emphasis added and emphasis removed). There are several
issues with this argument.

First, the Government questions the applicability of Plaintiffs’ cases to the present case
because they supposedly do not involve an “applicable asset acquisition” under I.R.C. § 1060.
Id. at 14. Even assuming that is a meaningful distinction, however, there are cases supporting
this standard that involve applicable asset acquisitions. See., e.g., East Ford, Inc. v. Comm’r, 67
T.C.M. (CCH) 3068 (1994), 1994 WL 243713, at *3-4 (finding that the sale of a truck-leasing
business was an “applicable asset acquisition” under I.R.C. § 1060). In fact, East Ford
undermines the Government’s argument because there the Court had to determine the reasonable
values of the assets because the two parties to the applicable asset transaction assigned different
values to the same assets to maximize their respective tax treatment. According to the Tax
Court: “Although the parties could have avoided this dispute by agreeing at the time of the sale
to a reasonable allocation of the purchase price, their failure to do so leaves it to this Court to
assign reasonable values to the assets transferred.” Id. at *1. If the parties could have avoided
the dispute by assigning reasonable values to assets at the time of their transfer, the Government
cannot be correct that the actual transaction is not sufficient to overcome summary judgment.

Second, Treasury itself reviews Section 1603 grant applications under the willing buyer –
willing seller paradigm. According to its 2011 guidance, “[t]he IRS generally defines fair market
value . . . as ‘the price at which property would change hands between a buyer and a seller,
neither having to buy or sell, and both having reasonable knowledge of all necessary facts.’”
ECF No. 88 at App’x 251. In reviewing Section 1603 grant applications, moreover, Treasury
assumes that “a Section 1603 applicant’s claimed cost basis reflects the eligible property’s fair
market value,” so long as the property was purchased through an arm’s-length transaction
involving no unusual circumstances. Id. at App’x 249. “Once a plaintiff has produced such
evidence, the burden is on the defendant to demonstrate ‘special circumstances which would
negate [the relevance] of a prior arm’s-length purchase price.’” Schonfeld, 218 F.3d at 179
(citation omitted). Even if the Government had come forward with evidence that the Transaction
was not arm’s length or that there was compulsion to buy or sell, those would be issues for trial,
not summary judgment.

Third, the cases that the Government relies upon to assert that the actual sales price does
not reflect the Plaintiffs’ basis do not help the Government here. For example, in Estate of
Bright v. United States, the court interpreted the estate tax regulation, which provides that “[t]he

21
fair market value is the price at which the property would change hands between a willing buyer
and a willing seller, neither being under any compulsion to buy or to sell and both having
reasonable knowledge of relevant facts.” 658 F.2d 999, 1005 (5th Cir. 1981) (en banc) (citing 26
C.F.R. § 20.2031-1(b)). In doing so, the court commented that “[i]t is apparent from the
language of the regulation that the ‘willing seller’ is not the estate itself, but is a hypothetical
seller.” Id. The Government also relies on Pabst Brewing Company v. Commissioner, as
holding that the willing buyer – willing seller “standard requires the court ‘to analyze the
property from the perspective of a hypothetical buyer and a hypothetical seller; not the actual
parties to the transaction.’” ECF No. 91 at 22-23 (citing Pabst, 69 T.C.M. (CCH) 2773, 1995
WL 325875, at *22). The Government maintains that it is “[c]ontrary to the hypothetical buyer
and seller standard of fair market valuation [for] plaintiffs [to] rely on the purchase price to
which they—the actual parties to the transaction—agreed to attempt to establish fair market
value.” ECF No. 91 at 13. According to the Government, “[t]he Transaction purchase price is
only probative of how the parties agreed to divvy up the total consideration for the purchase and
sale of a going concern business enterprise between the MIPSA and EPC Agreements,” and that
“[n]o inference may be drawn that the prices ascribed in these agreements by themselves are
indicative of the fair market value of the tangible solar property.” Id. “Thus,” the Government
maintains, “plaintiffs’ reliance on their own purchase price as evidence of fair market value is
insufficient to overcome summary judgment in favor of the government.” Id. at 14. The Court
disagrees.

To the extent the cases the Government relies upon deal with transfers of assets upon the
death of the owner without an actual transaction, they do not help the Government here. It goes
without saying that when there was no arms-length transaction, the Court must resort to a
hypothetical buyer and seller. As the court stated in Bank of California, National Ass’n v.
Commissioner, “we are required to assume the existence of a willing buyer and a willing seller,
regardless of whether they actually existed or not, and to assume that the property could and
would change hands, even though such a change could not in fact occur.” 133 F.2d 428, 433
(9th Cir. 1943) (emphasis added). In fact, the emphasis on the hypothetical willing buyer and
seller in such cases is not intended to the exclusion of an actual transaction—which, again, did
not occur—but instead to the exclusion of the decedent’s interests before death or the estate’s
interests after death. As the court in Estate of Bright explained:

The notion of the “willing seller” as being hypothetical is also
supported by the theory that the estate tax is an excise tax on the
transfer of property at death and accordingly that the valuation is to
be made as of the moment of death and is to be measured by the
interest that passes, as contrasted with the interest held by the
decedent before death or the interest held by the legatee after
death.

658 F.2d at 1006 (emphasis added). Stated differently:

Defining fair market value with reference to hypothetical willing-
buyers and willing-sellers provides an objective standard by which
to measure value. . . . The use of an objective standard avoids the

22
 uncertainties that would otherwise be inherent if valuation methods
attempted to account for the likelihood that estates, legatees, or
heirs would sell their interests together with others who hold
undivided interests in the property. Executors will not have to
make delicate inquiries into the feelings, attitudes, and anticipated
behavior of those holding undivided interests in the property . . . .

Propstra v. United States, 680 F.2d 1248, 1252 (9th Cir. 1982) (citations omitted). Accordingly,
Estate of Bright and similar estate tax dispute cases do not support minimizing the import of the
price at which an actual buyer and seller transferred property at apparent arm’s length in favor of
evidence of what a hypothetical buyer and seller would rather do to value such property.

The same is true for Whitehouse Hotel Limited Partnership v. Commissioner of Internal
Revenue, 615 F.3d 321 (5th Cir. 2010). See ECF No. 106 at 111:9-11. As an initial observation,
Whitehouse deals with the tax implications of the granting of a historic preservation easement to
a nonprofit entity that was treated as a charitable contribution for tax purposes. 615 F.3d at 325-
26. As with the estate cases, there was no arm’s length transaction at all, requiring the parties
and the court to examine (1) the “comparable-sales” method of valuating real property, which
looks to the ‘“sales from a willing seller to a willing buyer of similar property in the vicinity at or
about the same time’ as the property being valued[,]”; and (2) the inquiry of “what a hypothetical
willing buyer would consider in deciding how much to pay for the property[]” in determining
fair market value. Id. at 333, 335. Like in Estate of Bright, the court in Whitehouse had no
choice but to think in hypothetical terms because there was no actual buyer and seller in that
case. Therefore, Whitehouse also does not suggest overlooking the price at which an actual
buyer and seller transferred property at arm’s length, as the Government posits.

As for the Government’s reliance on Pabst, a non-estate tax case, the Government
overstates that case’s significance. Pabst involved a transaction between companies in which
various assets were priced and transferred between the parties. 69 T.C.M. (CCH) 2773 (1995),
1995 WL 32587, at *8-9. In relevant part, one of the issues before the court concerned the
proper valuation of the transferred assets. Id. at *19. The petitioner moved for partial summary
judgment that the fair market value of the transferred assets was the amount at which they were
priced because the transfer was at arm’s length. Id. The court denied the petitioner’s partial
summary judgment motion because the question of fair market value “is a question of pure
fact . . . .” Id. at *22. And while the Court does state that its analysis requires the hypothetical
seller and buyer, it did so because the Government “raises a genuine issue of material fact as to
the arm’s-length nature of the transaction and the fair market value of the Transferred Assets.”
Id. at *24. Accordingly, Pabst does not support the Government’s proposition that an actual
transaction’s purchase price is necessarily insufficient, and that further evidence must be shown
in valuing assets under the hypothetical willing-buyer, willing-seller standard. Rather, Pabst (at
most) suggests that the purchase price should not be the sole evidence consulted in valuing assets
and that the price could potentially be as important as other evidence in determining assets’ fair
market value.

The Government further argues that adopting the non-hypothetical willing-buyer, willing-
seller standard “would render § 1060 nugatory because § 1060 was enacted to ensure that buyers
and sellers of assets constituting a trade or business did not manipulate tax laws to their own

23
benefit.” ECF No. 91 at 14. According to the Government, “[i]f plaintiffs could rely on their
own agreement as substantive evidence of the fair market value of the assets conveyed in an
applicable asset acquisition, the ability of the government to challenge an erroneous allocation
and insist on compliance with the tax laws would be severely hampered.” Id. Not so. The
Government is free to challenge the Plaintiffs’ valuation of the Transaction assets and whether
the Transaction was truly negotiated at arm’s length. That does not mean that the Plaintiffs
cannot meet their burden for summary judgment based largely on their own allocation of prices.
In short, the hypothetical buyer and seller reflect that the Government and the Court are not
bound by the Plaintiffs’ allocation of assets, not that their allocation is irrelevant.

And Plaintiffs have come forward with evidence that, if true, would show that the
Transaction qualifies as an arms-length transaction such that their assets’ prices are indicative of
their fair market values. Specifically, there is no indication that NextEra or GE were under any
compulsion to buy the Facility. It also does not appear that First Solar was under any
compulsion to sell the Facility. At oral argument, the Government briefly argued that First Solar
was compelled to agree to the MIPSA’s terms because it otherwise would lose out on signing the
EPC Agreements. ECF No. 106 at 97:18-21. Even if true, this is an issue for trial and not
summary judgment.

Plaintiffs’ evidence also suggests that First Solar, NextEra, and GE were fully
knowledgeable about the Transaction given their respective experience in the renewable energy
industry. First Solar describes itself as the world’s leading manufacturer of solar PV facility
“thin film” solar modules, and is a leader in developing, financing, engineering, constructing,
operating, and selling many of the world’s largest grid-connected solar PV power plants.
Brannen Decl. ¶ 15; see also ECF No. 81-2 at App’x 26. NextEra builds and operates solar
power facilities and has been significantly involved in the solar industry since 2009. Id. ¶ 7.
And GE is a financial services provider, ECF No. 81-4 at App’x 88, with substantial experience
investing in the renewable energy industry, Brannen Decl. ¶ 13. Mr. Draper testified that GE and
NextEra “were . . . extremely well-versed in . . . the issues that we were dealing with, both on the
technical side and risk allocation under EPC agreements . . . .” Draper Dep. Tr. 266:24-267:5.
In addition, the Project Model further supports that First Solar, NextEra, and GE were well-
informed about the Desert Sunlight Transaction, as shown by the intricate formulas used to
calculate the Transaction’s prices based on the Facility’s expected ROE. See ECF No. 81-10 at
App’x 351-53.

Plaintiffs also assert that the negotiations over the EPC Agreements demonstrate that this
was an arm’s length transaction. NextEra and GE negotiated down the total EPC Agreement
price by over $400 million from First Solar’s initial offering of over $2.36 billion to the final
price of $1,950,807,086. Brannen Decl. ¶ 22; Draper Dep. Tr. at 264:21-12. NextEra and GE
also convinced First Solar to take on greater risk than they originally proposed, even with a
lower total EPC Agreement price, as a result of “the hard bargain that NextEra and GE drove.”
Brannen Decl. ¶ 23; Draper Dep. Tr. at 270:25-271:19, 273:1-7, 274:10-15, 274:22-275:6.
Indeed, Mr. Draper testified that “it was very arm’s length . . . it was an extremely hard
negotiation with GE and NextEra . . . .” Draper Dep. Tr. at 265:13-17.

Moreover, each party had reason to negotiate fair market value EPC Agreements. Mr.
Brannen explains that he had multiple reasons to do so beyond pursuing NextEra’s normal

24
business objective of paying only fair market value for EPC agreements. Brannen Decl. ¶ 24.
First, Mr. Brannen states he knew that NextEra would require loans to pay for the EPC
Agreements, and that banks would not finance the Agreements if they were above market. Id.
Second, due in part to the magnitude of the Facility’s construction, Mr. Brannen states he feared
that First Solar may not complete the Project, and that NextEra did not want to have paid First
Solar an above-market amount for any work already completed. Id. Third, Mr. Brannen states
that he had a personal incentive to reach the best price and lowest risk EPC agreements because a
significant portion of his salary was an incentive payment tied to the deal reached. Id. ¶ 25. Mr.
Draper similarly testified that First Solar was “trying to stay cash neutral or cash positive
throughout the course of construction for the project.” Draper Dep. Tr. at 252:10-12 (ECF No.
81-63 at App’x 4205). In addition to these indicia of arm’s length, Shaw concluded that the EPC
Agreement prices were comparable to prices in third-party, arm’s length transactions, meaning
that First Solar being the seller and construction contractor did not distort the prices in Shaw’s
opinion. ECF No. 88-1 at App’x 2576; Walker Dep. Tr. 92:4-94:18.

As a result, there is a genuine dispute of material fact whether the Transaction’s asset
prices as contained in the cost segregations and the Total Cost Spreadsheet—and as further
detailed by Exhibit I to the EPC Agreements—amount to their fair market values.

(3) There is a dispute of material fact regarding the fair market
value of the MIPSA assets.

The Government also argues that Plaintiffs have not provided evidence of the MIPSA
assets’ fair market value based on their prices alone because of the disparity between the
approximately $35 million First Solar spent to develop the assets and the far lower price of just
over $14 million that Plaintiffs paid under the MIPSA. ECF No. 91 at 11-12. At trial, the
Government may be proven correct. But this is a pure question of fact not amenable to
resolution on summary judgment because there remains a genuine dispute of material fact
whether the MIPSA assets’ purchase prices reflect their fair market value. The fact that the
MIPSA contract price is far lower than the amount First Solar spent developing its assets does
not mean that the price is not the fair market value. It is certainly possible for things to sell for
less than the cost of producing them, and a seller is willing to accept less than its cost to produce
the items for a number of reasons (e.g., to avoid further costs and “cut its losses”). Therefore,
the prices for the Desert Sunlight Transaction’s assets could indicate their fair market value, and
there is thus a genuine dispute of material fact whether the cost segregations and Total Cost
Spreadsheet containing the prices suffice to serve as Plaintiffs’ required I.R.C. § 1060 allocation.

(4) Plaintiffs Rely on More than just the Transaction Price to
Show the Assets’ Fair Market Value.

In addition to the evidence regarding the Transaction negotiations, Plaintiffs also rely on
several analyses of the Project and other benchmarks to support their contention that the
Transaction price allocations represent the fair market values of the Transaction assets. Taken
together, these analyses further bolster the conclusion that there is a genuine dispute of material
facts—making summary judgment inappropriate.

25
 Several firms with significant experience in the construction of power generation
facilities found that the Transaction price was reasonable. Shaw determined that First Solar’s
proposed total EPC Agreement price reflected a fair market value. ECF No. 88-1 at App’x 2576.
Burns and Roe similarly found that the proposed EPC Agreement price was reasonable. Walker
Dep. Tr. at 53:16-22, 58:16-59:18. Luminate also concluded that both First Solar’s initially
proposed total EPC Agreement price and the final total EPC Agreement price were reasonable.
Glover Dep. Tr. at 17:20-18:7, 79:20-81:18 (ECF No. 81-36 at App’x 3181-82, 3197). And
Black & Veatch concluded that the final total EPC Agreement price was “consistent with the
total construction cost of other PV solar generating facilities utilizing the same technology and
with our understanding of current market pricing in the solar PV industry.” ECF No. 88 at
App’x 258.

In addition to the independent engineering firms, the SETP/NREL concluded that the
originally proposed total Project price was “below the current estimated average total project
cost for comparable projects and [is] within a reasonable range of the expected system
price . . . .” ECF No. 88-1 at App’x 2425; see also, id. at App’x 2420 (an earlier SETP/NREL
memorandum concluding that “[b]ased on this analysis the EPC price falls within an acceptable
range of expected system costs based on technology selection, system size and configuration.”).
The SETP/NREL specifically determined that the projected EPC price of $3.43/Wdc was within
a reasonable range of NREL’s estimated EPC system price for a facility like the Desert Sunlight
Facility of $2.93/Wdc. Id. at App’x 2427-28. Further, Dr. Ramesh stated in his memorandum
that “[w]e have compared a detailed break out of the pricing for the elements of the EPC contract
against our model and have found the pricing to be within our expected range.” Id. at App’x
2426.

Plaintiffs also point to other valuations and benchmarks as further evidence of the EPC
Agreement assets’ fair market values. For one, NextEra made its own determination that the
final total EPC Agreement price was fair market value via its “cost build up,” which was
reviewed by BEW. Brannen Decl. ¶ 26e. In addition, DOE conducted its own due diligence and
determined that there was a reasonable assumption that, based on the originally proposed total
EPC Agreement price, the loan would be repaid by comparing the proposed total EPC
Agreement price to other projects in its internal databases. Walker Dep. Tr. at 99:8-100:12.
Moreover, all estimated total EPC Agreement prices were below Treasury’s 2011 first quarter
benchmark of fair market value for large commercial solar PV, which was $4/W. Brannen Decl.
¶ 26d; ECF No. 88 at App’x 249-50. All the estimated total EPC Agreement prices were
similarly less than publicly reported estimates of EPC agreement prices of utility-scale solar PV
facilities, including SEIA’s estimate of $3.85/Wdc, ECF No. 88 at App’x 222, and NREL’s
approximate estimate of $4/Wdc, Id. at App’x 239. Further, Mr. Brannen’s estimated total EPC
Agreement price of $2.69/Wdc for the originally planned 724 MWdc-capacity Facility, Brannen
Decl. ¶ 21, fell below the $2.80/Wdc or $2.92/Wdc prices that Treasury indicated at the time
were comparable EPC agreement prices. Walker Dep. Tr. at 157:13-158:3. And Deloitte
verified that Plaintiffs’ Section 1603 applications were “fairly stated, in all material respects.”
Pitale Decl. ¶¶ 41-42.

In addition to the above evidence from the events culminating in the Desert Sunlight
Transaction, Plaintiffs cite two additional pieces of evidence of the EPC Agreement assets’ fair
market values.

26
 First, Plaintiffs cite an April 2013 report from the U.S. Energy Information
Administration (“EIA”), entitled “Updated Capital Cost Estimates for Utility Scale Electricity
Generating Plants.” ECF No. 86-1 at 51-52; ECF No. 88-1 at App’x 2588-2604. The purpose of
the report in part was to provide information on the EPC costs of various types of utility-scale
power facilities in 2012 dollars. ECF No. 88-1 at App’x 2593-94, 2628. Its authors cautioned,
however, that all estimates in the report were “broad in scope” and “a more in-depth cost
assessment would require a more detailed level of engineering and design work, tailored to a
specific site.” Id. at App’x 2594. One of the facilities studied was a 150 MWac solar PV facility,
id., for which the report indicates that the facility’s total EPC costs would be approximately
$3.49 per kilowatt (“/kW”) (excluding owner costs), id. at App’x 2622, 2730. And this report
concluded that in California, the cost would be substantially higher. Id. at App’x 2784. Based
on a typical ratio of DC module capacity to AC inverter capacity being 1.3:1, id. at App’x 2727-
28, the generic cost would be $2.68/Wdc, Spence Decl. at 155:4-156:15 (ECF No. 88-1 at App’x
2583). And the lowest cost in California would be $2.88/Wdc. Id. at 165:19-166:8.
Accordingly, Mr. Brannen’s estimated Desert Sunlight EPC Agreement price of either
$2.63/Wdc or $2.73/Wdc, Brannen Decl. ¶ 21, was less than the lowest EPC price estimate in the
EIA report. ECF No. 87 at 51-52; ECF No. 88-1 at App’x 2588-2604.

Second, Plaintiffs rely on the opinion of Mr. Robert Charles. ECF No. 87 at 54-57. Mr.
Charles is an independent engineer Plaintiffs retained as an expert witness who states that “the
price that plaintiffs paid First Solar under the EPC contracts was a fair market value price to
engineer, procure and construct the Desert Sunlight Facility’s physical assets.” Charles Decl. ¶
7. Mr. Charles “reached that conclusion based on a comparison of the Desert Sunlight EPC price
with the prices of other contemporaneous EPC contracts for utility-scale solar PV facilities
employing exactly the same solar technology as Desert Sunlight.” Id. According to Mr. Charles,
“[i]ndependent engineers like me regard this market comparables approach as the best way to
conduct an independent assessment of whether an EPC price reflects fair market value when
appropriate comparables are available, as they were here.” Id. In addition, Mr. Charles
evaluated the solar module prices specifically, which he maintains are the most important and
costly component of a solar PV facility. Id. ¶¶ 11, 16. He concluded they were priced at fair
market value as well. Id. ¶ 11.

In response to the evidence of the EPC Agreement assets’ fair market values that
Plaintiffs muster, the Government raises several arguments why, actually, “none of
[it] . . . establishes the fair market value of tangible solar property.” ECF No. 91 at 18.
Specifically, the Government contends that the SETP/NREL, Luminate, and Shaw
determinations that the EPC Agreement prices fall within a certain price range is different from
determining their assets’ fair market value. See ECF No. 80 at 33-34. The Government also
asserts that Luminate’s finding of a price’s reasonableness is not equivalent to its fair market
value. Id. at 33 (citation omitted). Further, the Government argues that Shaw, Mr. Charles, and
Black & Veatch’s methodological approach of determining fair market value in comparison to
similar projects’ EPC prices is improper. See ECF No. 80 at 35; ECF No. 91 at 19. In addition,
the Government contends that Luminate’s and Shaw’s reviews are irrelevant because they
concern estimations at a later point in time than the Desert Sunlight EPC Agreement prices set in
2011. See ECF No. 80 at 33, 35. As for the other evidence outlined above, the Government
either does not address it specifically or makes a conclusionary statement that they do not
establish fair market value. See ECF No. 91 at 20 n.14 (commenting on NextEra’s cost-buildup

27
that, “[e]ven disregarding the self-serving nature of this document, it is insufficient to establish
fair market value of tangible solar property.”); ECF No. 80 at 36 n.14 (after reviewing Deloitte’s
review process, commenting that “Deloitte did not determine the fair market value of the grant-
eligible assets.”).

Whether or not the Plaintiffs’ evidence establishes the EPC Agreement assets’ fair market
values, however, is a matter for determination at trial. The above evidence establishes that there
is a genuine dispute of material facts regarding the EPC Agreement assets’ fair market value and
whether Plaintiffs provided a sufficient I.R.C. § 1060 allocation in the form of Plaintiffs’ cost
segregations and Total Cost Spreadsheet containing the prices of the Transaction’s assets.

Accordingly, the Government’s summary judgment request is denied.

C. The Parties’ Motions for Partial Summary Judgment

Both Parties move for partial summary judgment. As an alternative to its summary
judgment motion, the Government seeks partial summary judgment that the DOE Loan
Guarantee and the LGIA are intangible assets and that, “as a matter of law, any consideration
that is properly allocable to the acquisition of these assets is not eligible for a § 1603 payment.”
Id. at 37. In contrast, Plaintiffs seek partial summary judgment that the DOE Loan Guarantee is
not a separable asset capable of reducing the basis of the Section 1603-eligible property, and thus
not included as part of the I.R.C. § 1060 allocation. ECF No. 86-1 at 2, 10-11; ECF No. 95 at
10. Further, Plaintiffs agree that the LGIA is an intangible asset ineligible for a grant under
Section 1603. See ECF No. 86-1 at 19-20. But Plaintiffs request summary judgment that the
LGIA is a Class VI asset, Id. at 20, while the Government contends that the LGIA is a Class V
asset, ECF No. 80 at 48. In addition, Plaintiffs ask the Court to rule that, for purposes of the
I.R.C. § 1060 allocation, there is no need to value Class VI or Class VII assets to determine the
appropriate allocation to Class V assets per their fair market value. ECF No. 86-1 at 2-3, 21.
Lastly, Plaintiffs seek that, as a matter of law, sales tax, interest during construction, and early
completion bonuses are included in the basis of the Section 1603-eligible assets. Id. at 3. The
Court will address each request in turn.

1. Alta Wind Resolves Two Arguments.

There are two issues for which the Parties seek summary judgment where the Federal
Circuit has squarely decided the issue. First, the Government asks the Court to determine that
the PPAs are intangible assets. ECF No. 80 at 37. Plaintiffs similarly ask the Court to rule that
the PPAs are intangible assets, but Plaintiffs also ask the Court to determine that the PPAs are
Class VI rather than Class V intangible assets. ECF No. 86-1 at 20. The Federal Circuit has
clearly held that “the PPAs . . . may be characterized as customer-based intangible assets under
I.R.C. § 197.” Alta Wind, 897 F.3d at 1373-74. As explained above, the Treasury Regulations
define I.R.C. § 197 intangible assets as Class VI assets. 26 C.F.R. § 1.338-6(b)(2)(vi) (“Class VI
assets are all section 197 intangibles, as defined in section 197, except goodwill and going
concern value.”); Alta Wind, 897 F.3d at 1376 (citing 26 C.F.R. § 1.338-6(b)). The Parties’
motions for summary judgment on this issue are granted.

28
 Second, Plaintiffs seek the Court’s determination that, for purposes of the I.R.C. § 1060
allocation, the consideration in the Desert Sunlight Transaction must first be allocated to the
Class V assets to the extent of their fair market value, and only afterwards should any remaining
consideration be allocated to the Class VI and VII assets. ECF No. 86-1 at 3, 20. Here, too, the
Federal Circuit has spoken clearly. “The consideration is allocated among these classes in the
order they are listed in a ‘waterfall’ fashion, using the fair market value of the assets within each
class.” Alta Wind, 897 F.3d at 1376 (citing 26 C.F.R § 1.338-6(b)). Therefore, “[t]he purchase
price must . . . be allocated to Class V, then to Class VI, and finally to Class VII, if any value
remains.”) Id. (emphasis added). Plaintiffs’ motion for summary judgment on this issue is
granted.

2. The DOE Loan Guarantee’s Asset Status

The Government moves the Court to find that the DOE Loan Guarantee is an intangible
asset and that any allocable consideration to it is not eligible for a Section 1603 grant. ECF No.
80 at 37. Plaintiffs, however, move the Court to hold that the DOE Loan Guarantee is not a
separable asset capable of reducing the basis of the Section 1603-eligible property as part of the
I.R.C. § 1060 allocation. ECF No. 86-1 at 2, 10-11; ECF No. 95 at 10. The Court agrees with
Plaintiffs.

Plaintiffs largely rely on a March 2012 IRS Notice that “provides guidance in a question-
and-answer format on tax-related issues involving cash payments for specified energy property
in lieu of tax credits under Section 1603 of the American Recovery and Reinvestment Tax Act of
2009.” IRS Notice 2012-23, 2012-11 I.R.B. 483, 2012 WL 759615. Specifically, the Notice
states, in pertinent part:

Q-2. What are the federal income tax consequences to a taxpayer
who receives a Section 1603 payment and a Department of Energy
loan guarantee or an energy conservation subsidy from a public
utility?

A-2: Receipt of an incentive in addition to a Section 1603 payment
may reduce the eligible basis used in calculating the Section 1603
payment. (See FAQ1, Eligible Basis.) Receipt of a Department of
Energy loan guarantee does not reduce the basis of specified
energy property.

Id. ¶ Q-2 (emphasis added). According to Plaintiffs, the Notice “is squarely on point and
unambiguous: a DOE loan guarantee does not reduce eligible basis under ARRA Section 1603,
and thus none of the payments made with respect to the facility at issue are allocated to such loan
guarantees.” ECF No. 86-1 at 11.

Plaintiffs buttress their argument with the history of the energy tax credit. Id. at 11-13.
Specifically, Plaintiffs note that, prior to Section 1603’s amendment in 2009, the energy tax
credit provided that the eligible basis for the credit would be reduced by any “subsidized energy
financing.” Id. at 12 (citing I.R.C. § 48(a)(4) (2008)). “Subsidized energy financing”
encompassed any government financing for the purpose of subsidizing projects designed to

29
conserve or produce energy. Id. (citing I.R.C. § 48(a)(4)(C) (2008)). According to Plaintiffs, the
reason for the reduction was explained in the law’s legislative history: to prevent double-dipping
by investing in energy products with the aid of both government financing and a tax credit. Id. at
12 (citing H.R. Conf. Rep. No. 817, 96th Cong., 2d Sess. 1980, 1980 U.S.C.C.A.N. 642). And
equally clear, Plaintiffs contend, in the legislative history was Congress’ explicit carve-out:
“SUBSIDIZED FINANCING DOES NOT INCLUDE, HOWEVER, LOAN
GUARANTEES.” Id. (emphasis in original).

The Court need not wade into the legislative history on this point because the statute’s
language is clear. But Plaintiffs also point out that the IRS has interpreted the Code to exclude
loan guarantees from “subsidized energy financing.” The IRS adopted this understanding in
private letter rulings based on the same legislative history. ECF No. 86-1 at 12 (citing, e.g., PLR
8428035 (“The cited legislative history . . . clearly indicates that loan guarantees are not taken
into account in determining the credit reduction provided in [the ITC] and that loan guarantees
are not considered grants.”)). Plaintiffs add that this is even more true since 2009, when
Congress amended the energy tax credit so that subsidized financing would no longer reduce the
eligible basis, much less a government loan guarantee. ECF No. 86-1 at 13 (citing, e.g., I.R.C. §
1060 § 48(a)(4)(D)). Because Section 1603’s operation was designed to “mimic” the energy tax
credit, WestRock Virginia Corp., 941 F.3d at 1316 (“It is intended that the grant provision mimic
the operation of the credit under [IRC] section 48.”) (quoting H.R. Rep. No. 111-16 at 620–21),
Plaintiffs conclude that the energy tax credit’s non-reduction of an eligible basis by a
government loan guarantee is particularly relevant in concluding the same for a Section 1603
basis, ECF No. 86-1 at 13. While these IRS determinations are not binding, the Court may
consider them as persuasive authority. E.g., Glass v. Comm’r, 471 F.3d 698, 709 (6th Cir. 2006)
(“Although under I.R.C. § 6110(k)(3), a Private Letter Ruling cannot be used as precedent, a
recent ruling provides persuasive authority for refuting the Commissioner's argument.”). In this
case, the Court considers these letter rulings insofar as they shed light on the IRS’s
understanding of the Code and regulations at issue here.

In response, the Government argues that Plaintiffs’ cited authorities are irrelevant to the
Government’s position and the relevant issue surrounding the DOE Loan Guarantee. See ECF
No. 91 at 30-31. The Government does not argue that the DOE Loan Guarantee reduces a
Section 1603-eligible basis, as addressed in the above sources. See id. at 30-32. Rather, the
Government argues that the DOE Loan Guarantee must be included as an asset to which an
asset’s consideration is allocated under I.R.C. § 1060 to determine the Section 1603-eligible
basis. See id. at 30-31. In other words, “the issue is not whether the amount of the Loan
Guarantee, or the amount of guaranteed debt, is subtracted from basis; the issue is how § 1060
allocates the Transaction purchase price to establish § 1603-eligible basis in the first instance.”
Id. at 30. Thus, the Government maintains that neither the IRS Notice nor the energy tax credit
history are relevant to the allocability of the DOE Loan Guarantee under I.R.C. § 1060. See id.
at 30-31.

The Government’s argument draws a distinction without a difference. Even assuming the
DOE Loan Guarantee would be a Class V intangible asset under I.R.C. § 1060 as the
Government contends, ECF No. 80 at 48, including the DOE Loan Guarantee among the assets
to which consideration is allocated reduces a Section 1603-eligible basis.

30
 Persuaded by the IRS Notice and the energy tax credit history, the Court agrees with
Plaintiffs that the DOE Loan Guarantee is not a separable asset to be included in the I.R.C. §
1060 allocation and reduce the Section 1603-eligible basis. Accordingly, the Court grants
Plaintiffs’ partial summary judgment request and denies the Government’s partial summary
judgment request as to the DOE Loan Guarantee’s asset status.

3. The LGIA’s Asset Class

As with the DOE Loan Guarantee, the Government requests that the Court find the LGIA
is an intangible asset and that any allocable consideration to it is not eligible for Section 1603
relief. Id. at 37. Partially agreeing with the Government’s request, Plaintiffs ask the Court to
determine that the LGIA is a Class VI asset specifically, ECF No. 86-1 at 20, in contrast to the
Government’s view that the LGIA is a Class V asset, ECF No. 80 at 48. The Court again
concurs with Plaintiffs.

Plaintiffs maintain that the LGIA is an intangible, Class VI asset under Section 197. ECF
No. 86-1 at 19. According to Plaintiffs, “[t]he LGIA is a contract wherein the CAISO and SCE
provide Plaintiffs the right to interconnect to the utility grid at a specified point,” which
“constitutes a ‘supplier-based intangible’ under IRC Section 197 – i.e., ‘the value resulting from
the future acquisition, pursuant to contractual or other relationships with suppliers in the ordinary
course of business, of goods or services that will be sold or used by the taxpayer.’” Id. (citing 26
C.F.R. § 1.197-2(b)(7)(i)). “Specifically,” Plaintiffs reason, “the LGIA reflects the value
resulting from Plaintiffs’ future acquisition, pursuant to contract, of interconnection services
from suppliers (CAISO and SCE) that will be used by Plaintiffs.” Id. Thus, Plaintiffs contend
that the LGIA is a Class VI intangible asset. Id. at 19-20.

The Government disagrees that the LGIA is a Class VI Section 197 asset, and specifically
not a supplier-based intangible. ECF No. 91 at 38 (citing 26 C.F.R. § 1.197-2(b)(7)(i)). The
Government reasons that “the LGIA is not a supplier-based intangible because there is no ‘future
acquisition’ occurring under the LGIA.” Id. Rather, according to the Government, “[t]he LGIA
represents the legal right for the Facility to interconnect to the grid,” and “plaintiffs acquired that
right in September 2011” when it “was contractually secured.” Id. “There was nothing left for
Desert Sunlight to acquire in the future from the [CAISO] under the LGIA (or from SCE, the
other party to the LGIA),” which the Government argues means that the LGIA is thus not a
supplier-based intangible asset “intended to capture value that results from a contract or other
arrangement that itself will permit the taxpayer to buy goods and services from a supplier in the
future.” Id.

The Government argues that the LGIA is not a supplier-based intangible because the
“example of a supplier-based intangible provided in the relevant regulation” of “an arrangement
that a taxpayer has with a retail store that allows the taxpayer to purchase in the future shelf
space in that store.” Id. (citing 26 C.F.R. § 1.197-2(b)(7)(i)) (emphasis in original). In that
example, according to the Government, “[t]he intangible asset is the ability to acquire the shelf
space in the future, which could have value because of uncertainty in the future of the
availability, or price of, shelf space.” Id. (emphasis in original). The Government contrasts that
to the LGIA where “plaintiffs are not acquiring . . . the ability to buy any good or service in the
future from CAISO, the relevant LGIA counterparty, or from any other party.” Id. “Plaintiffs

31
have simply secured a contract right that allows them to connect to the grid upon completion of
the Facility,” which “is entirely disconnected from the ‘future acquisition’ of a good or service
that is contemplated by the definition of a supplier-based intangible under Treas. Reg. § 1.197-
2(b)(7).” Id. “Therefore,” the Government concludes, “because the LGIA is not a supplier-
based intangible, it is not a section 197 intangible, and, thus is not Class VI asset. Consequently,
the LGIA falls within the general catch-all category of Class V asset.” Id. at 39.

The Government, however, misconstrues the definition of a Section 197 supplier-based
intangible. The federal regulation defines it as “the value resulting from the future acquisition,
pursuant to contractual or other relationships with suppliers in the ordinary course of business, of
goods or services that will be sold or used by the taxpayer.” 26 C.F.R. § 1.197-2(b)(7)(i). In
arguing that the LGIA is not a supplier-based intangible, the Government apparently assumes
that a “future acquisition . . . of goods or services” requires a new transaction in the future. But
that is too strict an interpretation. Rather, the future acquisition merely refers to a future transfer
of goods or services. Thus, even though the LGIA does not provide for any entirely new
transaction in the future, it does provide for the right to obtain a service in the future, i.e., the
right to interconnect to the grid. ECF No. 81-29 at App’x 2650, 2659, 2661-62. This is
precisely the definition of a supplier-based intangible under Section 197. And even if “future
acquisition” in the regulation encompassed actually acquiring something new, the LGIA meets
that definition as well, as it also requires SCE to construct a new substation to which Plaintiffs
will connect the Facility. Id.

Indeed, the example in the regulation that the Government cites supports the Court’s
understanding. The regulation specifically states that “the amount paid or incurred for supplier-
based intangibles includes, for example, any portion of the purchase price of an acquired trade or
business attributable to the existence of a favorable relationship with persons providing
distribution services (such as favorable shelf or display space at a retail outlet), or the existence
of favorable supply contracts.” 26 C.F.R. § 1.197-2(b)(7)(i). As Plaintiffs correctly note, in the
case of “shelf space in a store, the purchaser (a seller of consumer products) acquires from the
supplier (a grocery store or other retail outlet) – on the acquisition date itself – the right to use
designated shelf space in the future,” which “constitutes a supplier-based intangible . . . because
the agreement encompasses the right to receive a service in the future: the use of the shelf space
on an ongoing basis.” ECF No. 95 at 15. “Similarly,” they conclude, “while Plaintiffs received
the right to connect to the utility grid when First Solar conveyed the LGIA in September 2011,
the service itself – connectivity to the grid – was to be provided in the future, i.e., if and when the
Desert Sunlight facility began to operate, and continuously thereafter.” Id.

In fact, Plaintiffs correctly observe that “the LGIA is an even clearer example of a
supplier-based intangible[]” than the shelf space example. Id. at 16. “In the case of shelf space
in a retail outlet,” Plaintiffs continue, “a party receives the contractual right to use shelf space in
the future, but the regulations suggest that the party also may begin using shelf space on Day 1,
immediately after it acquires the right.” Id. “Given that the contract[] . . . ha[s] been deemed to
convey ‘value resulting from the future acquisition’ of a good or service, and thus constitute [a]
supplier-based intangible[], then a fortiori, the LGIA – under which Plaintiffs would only be
interconnecting to the grid years after acquiring the contract in September 2011 – is
a . . . supplier-based intangible.” Id. (emphasis omitted).

32
 Therefore, the Court agrees with Plaintiffs that the LGIA is a Class VI Section 197
supplier-based intangible asset. Alternatively, even if the Court had not agreed with the
Plaintiffs that the LGIA is a supplier-based intangible asset, the LGIA would still presumably be
a Class VI Section 197 intangible asset under 26 C.F.R. § 1.197-2(b)(12), which provides that
“Section 197 intangibles include any other intangible property that is similar in all material
respects to the property specifically described in . . . paragraphs (b)(3) through (7) of this
section.” 26 C.F.R. § 1.197-2(b)(12). As Plaintiffs correctly note, the LGIA is materially similar
to supplier-based intangible assets because it too “confers the right to receive something in the
future on specified terms.” ECF No. 95 at 17. Thus, the LGIA would still be a Class VI Section
197 intangible asset. As a result, the Court grants the Government’s summary judgment motion
insofar as it seeks judgment that the LGIA is an intangible asset, as well as Plaintiffs’ partial
summary judgment motion insofar as it seeks judgment that the LGIA is a Class VI Section 197
intangible asset.

4. Operation of I.R.C. § 1060’s Allocation

Plaintiffs’ next move for summary judgment that “[u]nder the I.R.C. § 1060 waterfall,
there is no need to value Class VI or Class VII intangibles in order to determine the appropriate
allocation to Class V assets.” ECF No. 86-1 at 21. For purposes of allocating consideration to
Class V’s assets, Plaintiffs explain, “it is only necessary under IRC § 1060 to determine the fair
market value of the Class V assets, and to allocate to Class V in accordance with that fair market
value. That determination is made without regard to any purported value that Class VI or VII
intangibles might have.” Id.

In response, the Government asserts that, while it is true that there may be no
consideration left to be allocated to the Class VI or VII assets after allocation to the Class V
assets, “[t]hat factual determination[] . . . does not alter the taxpayer’s legal requirement to set
forth its determination of the fair market value associated with each asset class under § 1060.”
ECF No. 91 at 40 (emphasis in original). “A taxpayer’s belief that its purchase price should be
allocated in a way that renders the value of Class VI intangible assets irrelevant does not excuse
the taxpayer from the requirement to inform the government of the fair market of the Class VI
intangible assets.” Id. (emphasis omitted). Noting that Plaintiffs’ failure to value all of the
Desert Sunlight Transaction’s assets subjected them to penalties under the Internal Revenue
Code, the Government explained that “[t]here is a good reason that plaintiffs were required to
provide this information to the government. It provides the government with the ability to
review the totality of the circumstances and determine whether to accept or challenge the
taxpayer’s stated fair market values and related allocation. . . .” Id. However, the IRS “may
challenge the taxpayer’s determination of the fair market value of any asset by any appropriate
method and take into account all factors . . . .” Id. (quoting 26 C.F.R. § 1.338-6(a)(2)(iii)).
“Without the taxpayer’s view on the valuation of all the assets acquired, neither the government
nor the court can evaluate the totality of the circumstances,” which the Government argues “is
especially important when the application of § 1060 involves a consideration of [‘]all the facts
and circumstances.[’]” Id. at 41 (quoting 26 C.F.R. § 1.1060-1(b)(2)(iii)).

The Government’s argument is unpersuasive. As an initial matter, the Government’s
argument that Plaintiffs’ failure to provide fair market values for Class VI and VII property
deprived it of the ability to determine whether to challenge the Plaintiffs’ asset valuations is

33
belied by the fact that the Government is clearly challenging the Plaintiffs’ asset valuations. But
as explained above, those challenges are factual ones for trial, not summary judgment. Among
those factual issues will be the fair market values of the assets Plaintiffs acquired in the
Transaction.

Because this is a Section 1603 cash grant case, what matters most is the value of the
Class V tangible assets in which the Section 1603-eligible tangible assets fall, not the value of
the Class VI or VII intangible assets. As already analyzed at length, Plaintiffs mustered
sufficient evidence concerning the fair market value of the Class V tangible assets to survive
summary judgment. See supra III.A.2(b)(2)-(3). If Plaintiffs prove these fair market values at
trial, all the Transaction consideration would be allocated to Class V and the fair market value of
the Class VI and VII assets would not be necessary. Of course, if the Government shows errors
in Plaintiffs’ valuations, there may be consideration left over to allocate to Class VI or VII assets.
This too is a trial issue. And while the Government appears to disagree, the Plaintiffs have come
forward with analyses of the Class VI assets by KPMG and their proposed expert Mr. Charles
that both found these assets to have no or negligible value. See ECF 87 at 68-70 (summarizing
evidence). This too shall be resolved at trial.

The Court grants Plaintiffs’ third request for partial summary judgment that, for purposes
of the I.R.C. § 1060 allocation, there is no need to value Class VI or Class VII intangibles in
order to determine the appropriate allocation to Class V assets per their fair market value.

5. Sales Tax, Interest During Construction, and Early Completion Bonuses

Plaintiffs’ fourth request for partial summary judgment is for the Court to determine that,
as a matter of law, sales tax, interest during construction, and early completion bonuses are
included in the basis of the Section 1603-eligible assets. Id. at 3. In making their request,
Plaintiffs are clear that they do not seek the Court to decide the amount of the sales tax, interest
during construction, and early completion bonuses that should be included in the basis, as that
determination is an issue for trial. Id. at 21. Rather, Plaintiffs merely request that the Court find
that these cost categories are included in the basis as a matter of law. Id.

The Government “does not dispute that in general sales tax and interest are subject to the
capitalization rules of § 263A, and in general are § 1603-eligible costs if incurred in connection
with the acquisition of specified energy property under § 1603 and Treasury Guidance.” ECF
No. 91 at 41. The Government similarly does not dispute that early completion bonuses, as a
matter of law, may be included in the eligible basis. See id. at 42-45 (only addressing why the
early completion bonuses in this case should not be included in the basis, but not disputing that
they may be included as a matter of law).

Indeed, that sales tax, interest during construction, and early completion bonuses are
included in the basis of Section 1603-eligible assets as a matter of law is generally not a matter
of dispute. See ECF 88 at App’x 249 (“Basis . . . may also include the capitalized portion of
certain other costs related to buying or producing the property (e.g., permitting, engineering, and
interest during construction).”) (citing 26 C.F.R. § 1.263A-1). Section 263A of the U.S. Code,
Title 26, requires that certain costs incurred in producing certain property must be capitalized to
that property, I.R.C. § 263A(a)(1)(B), meaning “charge[d] to a capital account or basis,” 26

34
C.F.R. § 1.263A-1(c)(3). Section 263A applies to “[r]eal or tangible personal property produced
by the taxpayer,” I.R.C. § 263A(b)(1), and specifically “property produced by the taxpayer for
use by the taxpayer . . . in a trade or business or an activity conducted for profit,” I.R.C. §
263A(c)(1). “Produc[ing]” property refers to constructing, building, installing, manufacturing,
developing, or improving property, I.R.C. § 263A(g)(1), and “[t]he taxpayer shall be treated as
producing any property produced for the taxpayer under a contract with the taxpayer . . . ,” I.R.C.
§ 263A(g)(2). The costs to be capitalized with such property include both “direct costs” and
“indirect costs.” I.R.C. § 263A(a)(2). Indirect costs “directly benefit or are incurred by reason
of the performance of production . . . .” 26 C.F.R. § 1.263A-1(e)(3)(i).

There are many examples of indirect costs that are capitalized to the costs to produce
property under Section 263A. One example is taxes, I.R.C. § 263A(a)(2)(B), “to the extent such
taxes are attributable to labor, materials, supplies, equipment, land, or facilities used in
production . . . .” 26 C.F.R. § 1.263A-1(e)(3)(ii)(L). Indirect costs also include interest costs
“paid or incurred during the production period, and [] allocable to property . . . which has (i) a
long useful life, (ii) an estimated production period exceeding 2 years, or (iii) an estimated
production period exceeding 1 year and a cost exceeding $1,000,000.” I.R.C. § 263A(f)(1); see
also 26 C.F.R. § 1.263A-1(e)(3)(ii)(V) (“Interest includes interest on debt incurred or continued
during the production period to finance the production of real property or tangible personal
property to which section 263A(f) applies.”). Lastly, early completion bonuses have been
determined to be capitalized under Section 263 of the U.S. Code, Title 26, which generally
requires the capitalization of “[a]ny amount paid out for new buildings or for permanent
improvements or betterments made to increase the value of any property or estate.” I.R.C. §
263(a)(1); see Rev. Rul. 70-332, 1970-1 C.B. 31, 1970 WL 20683, at *2 (concluding that “the
premium time or overtime . . . paid or incurred solely to expedite the installation of additional
stands in the rolling mill is not an ordinary and necessary business expense to be deducted” but
instead, “under section 263(a) of the Code, such amount is added to the primary cost of the
installation.”); see also, e.g., Sears Oil Co. v. Comm’r, 359 F.2d 191, 197 (2d. Cir. 1966)
(holding in part that payments for early delivery of oil barges cannot be deducted as ordinary and
necessary business expenses but were capital expenditures).

Accordingly, the Court grants Plaintiffs’ fourth ground for summary judgment, that, as a
matter of law, the cost categories of sales tax, interest during construction, and early completion
bonuses are capitalized to the costs included in the basis of the Section 1603-eligible assets.

IV. Conclusion

For the foregoing reasons, the Court hereby rules:

1. The Government’s Motion for Summary Judgement, ECF No. 80, is GRANTED-IN-
PART and DENIED-IN-PART;

2. Plaintiffs’ Cross Motion for Partial Summary Judgment, ECF No. 86, is GRANTED;
and

35
 3. The Government’s Motion in Limine to Exclude Inadmissible Testimony Offered by
Plaintiffs’ Witnesses, ECF No. 92, is DENIED AS MOOT without prejudice to object
at trial.

The Court will schedule a status conference in the week of October 18, 2021, to
determine the date and location of trial.

IT IS SO ORDERED.

s/ Edward H. Meyers
Edward H. Meyers
Judge

36