Slip Op. 26-
UNITED STATES COURT OF INTERNATIONAL TRADE
ARCHER DANIELS MIDLAND COMPANY,
Plaintiff, JOINT STOCK COMPANY APATIT,
Plaintiff-Intervenor, Before: Jane A. Restani, Judge v.
Consol. Court No. 23-00239 UNITED STATES,
Defendant, THE MOSAIC COMPANY, Defendant-Intervenor.
OPINION AND ORDER
[Commerce’s final results of its redetermination pursuant to court remand in the countervailing duty order review of phosphate fertilizers from the Russian Federation are partially sustained and partially remanded for reconsideration consistent with this opinion.]
Dated: February 6, 2026
Warren E. Connelly, Trade Pacific PLLC, of Washington, DC, for plaintiff, Archer Daniels Midland Company. With him on the brief were Jonathan M. Freed, Kenneth Neal Hammer, and Robert George Gosselink.
Harold Deen Kaplan, Hogan Lovells US LLP, of Washington, DC, for plaintiff-intervenor, Joint Stock Company Apatit. With him on the brief were Jared Rankin Wessel, Jonathan Thomas Stoel, and Maria Alejandra Arboleda Gonzalez.
Sosun Bae, Lead Attorney, Commercial Litigation Branch – Civil Division, U.S. Department of Justice, of Washington, DC, for the defendant. With her on the brief was Meen Geu Oh. Of counsel on the brief were Kenneth Garrett Kays and Samuil Oshri Agranovich, Office of the Chief Counsel for Trade Enforcement and Compliance, U.S. Department of Commerce, of Washington, DC.
Consol. Court No. 23-00239 Page 2
Alexandra S. Maurer, Wilmer, Cutler, Pickering, Hale & Dorr, LLP, of Washington, DC, for defendant-intervenor, The Mosaic Company. With her on the brief were David J. Ross, Lindsey A. Ricchi, and Stephanie Ellen Hartmann.
Restani, Judge: Before the court is the United States Department of Commerce’s (“Commerce”) final remand redetermination pursuant to the court’s remand order, see generally Archer Daniels Midland Co. v. United States, 779 F. Supp. 3d 1349 (CIT 2025) (“Archer Daniels Midland I”), on Commerce’s final determination in its countervailing duty order review of phosphate fertilizer from the Russian Federation (“Russia”) covering the period from November 30, 2020 through December 31, 2021. See generally Final Results of Redetermination Pursuant to Court Remand, ECF No. 90-1 (Aug. 4, 2025) (“Remand Results”). In Archer Daniels Midland I, the court remanded in part to Commerce to either provide additional record evidence supporting its phosphate rock and natural gas benchmarks or to reconstruct the benchmarks. See Archer Daniels Midland I at 1360, 1367. For the following reasons, the court sustains in part and remands in part Commerce’s Remand Results.
BACKGROUND
The court presumes familiarity with the facts of this case as set out in the court’s previous opinion ordering remand to Commerce, see Archer Daniels Midland I, and recounts only those facts relevant to the issues currently before the court. On April 7, 2021, Commerce issued a countervailing duty order on imported phosphate fertilizer from Morocco and Russia. See Phosphate Fertilizers From the Kingdom of Morocco and the Russian Federation, 86 Fed. Reg. 18,037 (Dep’t Commerce Apr. 7, 2021). On June 9, 2022, Commerce initiated its review of the order for Period of Review (“POR”) from November 30, 2020, to December 31, 2021. See Initiation of Antidumping and Countervailing Duty Administrative Reviews, 87 Fed. Reg. 35,165 (Dep’t Commerce June 9, 2021). Commerce selected Joint Stock Company Apatit (“JSC Apatit”),
Consol. Court No. 23-00239 Page 3
a producer of phosphate fertilizer in Russia, as a mandatory respondent. Decision Memorandum for the Preliminary Results and Partial Rescission of the Countervailing Duty of Administrative Review; 2020–2021: Phosphate Fertilizers from the Russian Federation at 1, P.R. 188 (Apr. 27, 2023) (“PDM”).
Two components of the subject phosphate fertilizer are at issue here. First, Commerce assessed the Government of Russia’s (“GOR”) provision of phosphate ore1 mining rights to JSC Apatit through a tier-three benchmark2 pursuant to 19 C.F.R. § 351.511(a)(2)(iii), comparing JSC Apatit’s phosphate rock cost buildup to world market igneous phosphate rock export prices. See Issues and Decision Memorandum for the Final Results of the Countervailing Duty Administrative Review of Phosphate Fertilizers from the Russian Federation; 2020-2021 at 16, 31, P.R. 242 (Dep’t Commerce Oct. 31, 2023) (“IDM”). Second, Commerce assessed the GOR’s provision of natural gas to JSC Apatit through a tier-two benchmark pursuant to 19 C.F.R. § 351.511(a)(2)(ii) using Kazakh natural gas export prices. Id. at 48. On November 6, 2023, Commerce published its final results and determined the total countervailable subsidy rate to be 28.50 percent ad valorem. See Phosphate Fertilizers From the Russian Federation: Final Results of Countervailing Duty Administrative Review; 2020-2021, 88 Fed. Reg. 76,182, 76,183 (Dep’t Commerce Nov. 6, 2023) (“Final Results”).
On May 6, 2025, the court remanded in part Commerce’s final results as unsupported by substantial evidence. Archer Daniels Midland I at 1360, 1367. The court first held that Commerce
1 Phosphate ore is transformed into beneficiated phosphate rock and then into phosphate fertilizer. Letter from Hogan Lovells, JSC Apatit Benchmark, Appx. 7 at 15, C.R. 179, 182–191, P.R. 143, 146–155 (Mar. 15, 2023) (the “Davis Report”). 2 As the court will explain in further detail, Commerce must set benchmarks that reflect “prevailing market conditions.” 19 U.S.C. § 1677(5)(E)(iv). 19 C.F.R. § 351.511(a)(2) offers three methodological tiers for Commerce’s benchmark calculations.
Consol. Court No. 23-00239 Page 4
unreasonably limited its tier-three phosphate rock benchmark to phosphate rock from igneous ore reserves because Commerce failed to demonstrate that the phosphate rock market price is significantly driven by the difference in beneficiation processes of sedimentary and igneous phosphate rock. Id. at 1360. The court instructed Commerce to either provide evidence to support its tier-three benchmark or to reset the benchmark. Id.
Second, the court held that Commerce unreasonably constructed its tier-two natural gas benchmark because Commerce had not demonstrated that the benchmark third-party sales it used were comparable to sales of gas into Russia or that sales to a government entity that distorts the natural gas market were within the intended meaning of “purchaser” in the regulation. Id. at 1367. The court instructed Commerce to either address these issues or construct a tier-three benchmark for JSC Apatit’s natural gas purchases. Id.
On August 4, 2025, Commerce filed its Remand Results. See generally Remand Results.
Commerce maintained its tier-three phosphate rock benchmark and cited record evidence attempting to demonstrate that the costs to produce phosphate rock from igneous and sedimentary ore reserves differ significantly. Id. at 6–7. As to the second component, Commerce reconsidered its benchmark for JSC Apatit’s natural gas purchases and found that the natural gas purchased by third parties was not comparable to the gas purchased by JSC Apatit. Id. at 13. Accordingly, Commerce constructed a tier-three natural gas benchmark, id. at 14, and calculated a revised subsidy rate of 22.13 percent ad valorem for the gas input. Id. at 17. Commerce calculated a total subsidy rate of 49.64 percent ad valorem for JSC Apatit. Id. at 2.
On September 25, 2025, plaintiff-intervenor and consolidated plaintiff JSC Apatit filed comments on the Remand Results, arguing that Commerce unreasonably constructed the phosphate rock and natural gas benchmarks. JSC Apatit’s Comments on Remand Results at 3–29,
Consol. Court No. 23-00239 Page 5
ECF No. 96 (Sep. 25, 2025) (“JSC Apatit Cmts.”). On September 25, 2025, plaintiff Archer Daniels Midland (“ADM”) filed its comments, arguing that Commerce unreasonably constructed the phosphate rock benchmark. Archer Daniels Midland Company’s Comments on Remand Results at 8–35, ECF No. 97 (Sep. 25, 2025) (“ADM Cmts.”). On December 15, 2025, consolidated defendant-intervenor The Mosaic Company (“Mosaic”) filed its comments in support of Commerce’s Remand Results. The Mosaic Company’s Comments on Remand Results, ECF No. 102 (Dec. 15, 2025) (“Mosaic Cmts.”). On December 15, 2025, the government filed its comments in support of Commerce’s Remand Results. Gov.’s Comments on Remand Results, ECF No. 104 (Dec. 15, 2025) (“Gov. Cmts.”).
JURISDICTION AND STANDARD OF REVIEW The court has jurisdiction pursuant to 19 U.S.C. § 1516a(a)(2)(B)(i) (2020) and 28 U.S.C.
§ 1581(c) (2020). The court will uphold Commerce’s determinations in a countervailing duty proceeding unless they are “unsupported by substantial evidence on the record, or otherwise not in accordance with law[.]” 19 U.S.C. § 1516a(b)(1)(B)(i). “The results of a redetermination pursuant to court remand are also reviewed for compliance with the court’s remand order.” U.S. Steel Corp. v. United States, 219 F. Supp. 3d 1300, 1307 (CIT 2017) (citation modified) (citation omitted).
DISCUSSION
I. Commerce’s Phosphate Rock Benchmark is not Supported by Substantial Evidence ADM and JSC Apatit argue that Commerce failed to demonstrate that the production cost of igneous and sedimentary phosphate rock differs significantly or that the distinction between the two kinds of rock is a significant driver of prices in the phosphate rock market. ADM Cmts. at 4– 5, 25–30; JSC Apatit Cmts. at 21, 28. Specifically, ADM and JSC Apatit argue that Commerce
Consol. Court No. 23-00239 Page 6
cherry-picked record evidence by relying on excerpts from two sources in the record: a report by Dr. Graham A. Davis and a book by Professor Petr Ptáček. See generally ADM Cmts. at 1–2, 20– 22; JSC Apatit Cmts. at 22–23 (citing Letter from Hogan Lovells, JSC Apatit Benchmark, Appx. 7, C.R. 179, 182–191, P.R. 143, 146–155 (Mar. 15, 2023) (the “Davis Report”); Letter from WilmerHale, Mosaic Benchmark, Exs. 19–21, P.R. 132–33, 136, 138 (Mar. 15, 2023) (the “Ptáček Report”)). They argue further that the Ptáček Report does not support Commerce’s conclusion that igneous and sedimentary ore categorically have different beneficiation processes. JSC Apatit Cmts. at 22 (adding that the Ptáček Report notes that “calcareous” sedimentary rock has a different beneficiation process than igneous rock but that other types of sedimentary phosphate rock have the same beneficiation processes as igneous rock); see also ADM Cmts. at 14–20 (noting that the Ptáček Report demonstrates that there are many types of igneous and sedimentary phosphate rock with complex and varying beneficiation processes). ADM argues that India, Japan, Brazil, and New Zealand import phosphate rock above a certain bone phosphate of lime (“BPL”)3 content derived from both types of ore, which indicates that BPL content, not the source of the rock, drives the market price.4 ADM Cmts. at 31–32. JSC Apatit and ADM conclude that Commerce should recalculate the phosphate rock benchmark by including the volumes and values of exports from
3 The fertilizer industry uses the terms “BPL content” and “P2O5 content” (phosphorous pentoxide) interchangeably to refer to the grade of phosphate rock and its suitability for processing into phosphate fertilizer. Letter from WilmerHale, Mosaic Benchmark, Ex. 21 at 436–37, P.R. 132– 33, 136, 138 (Mar. 15, 2023). 4 ADM also argues that Commerce improperly attempts to shift the burden of proof to ADM to demonstrate that a significant cost-of-production difference does not exist. ADM Cmts. at 30–31. Because the court holds that Commerce unreasonably constructed the phosphate rock benchmark, the court does not reach this argument.
Consol. Court No. 23-00239 Page 7
Togo and Iran, countries that export sedimentary phosphate rock with a similar BPL content to Russian phosphate rock.5 JSC Apatit Cmts. at 5; ADM Cmts. at 5, 35.
Mosaic and the government respond that Commerce articulated a “thorough explanation of how production costs impact market forces of supply and demand.” Mosaic Cmts. at 8; see also Gov. Cmts. at 14–15. The government notes that, in conducting its tier-three analysis, Commerce’s aim is to isolate JSC Apatit’s costs for phosphate ore mining and beneficiation activities, so Commerce reasonably focused on JSC Apatit’s cost to mine and process phosphate ore into phosphate rock. Gov. Cmts. at 22. The government argues that the court did not require Commerce to quantify the cost differences between the beneficiation processes for sedimentary and igneous ore, and that Commerce has sufficiently explained why the difference in production costs would significantly drive price differences on the world market. Id. at 22.
To find a countervailable subsidy, Commerce must establish that an authority provided a financial contribution and thereby conferred a benefit. 19 U.S.C. § 1677(5)(B) (2020). In cases where goods or services are provided, a benefit is considered conferred if the goods or services are provided for less than adequate remuneration (“LTAR”). Id. § 1677(5)(E)(iv). Adequacy of remuneration is determined in relation to prevailing market conditions for the good or service being provided in the country subject to the review. Id. Prevailing market conditions include price, quality, availability, marketability, transportation, and other conditions of purchase or sale. Id. Commerce determines the amount of the subsidy by comparing remuneration actually paid
5 JSC Apatit argues that the court should direct Commerce to incorporate either the additional data from the Eurostat database, which Commerce rejected because Eurostat did not identify whether the data was sedimentary or igneous rock, or the Global Trade Atlas (“GTA”) data from Togo and Iran for comparable phosphate rock. JSC Apatit Cmts. at 5, 21. ADM does not address which database Commerce should use. Rather, it argues only that Commerce should include export data from Togo and Iran. See generally ADM Cmts.
Consol. Court No. 23-00239 Page 8
with a market-determined price for the goods under a “three-tiered hierarchy” employed by Commerce “to determine the appropriate remuneration benchmark.” Changzhou Trina Solar Energy Co. v. United States, 352 F. Supp. 3d 1316, 1332 (CIT 2018); see also 19 C.F.R. § 351.511(a)(2)(i)–(iii).
Commerce constructs a tier-one benchmark “by comparing the government price to a market-determined price for the good or service resulting from actual transactions in the country in question.” 19 C.F.R. § 351.511(a)(2)(i). If the actual market-determined price is unavailable, Commerce constructs a tier-two benchmark “by comparing the government price to a world market price where it is reasonable to conclude that such price would be available to purchasers in the country in question.” Id. § 351.511(a)(2)(ii). If the world market price is unavailable to purchasers in the country at issue, Commerce constructs a tier-three benchmark and measures the adequacy of remuneration by “assessing whether the government price is consistent with market principles.” Id. § 351.511(a)(2)(iii). If Commerce concludes that the government price is inconsistent with market principles, it constructs an external benchmark. Canadian Solar Inc. v. United States, 537 F. Supp. 3d 1380, 1389 n.6 (CIT 2021).
Because the GOR owns subsoil resources, companies such as JSC Apatit must obtain a license from the GOR to extract subsoil resources. IDM at 14. Commerce determined that the GOR’s provision of mining rights to JSC Apatit constitutes a financial contribution in the form of a provision of a good within the meaning of Section 771(5)(D)(iii) of the Trade Act of 1930,19 U.S.C. § 1677(5)(D)(iii). PDM at 18–20. Commerce constructed a tier-three benchmark and stated it compared the actual per-unit cost buildup plus profit for JSC Apatit’s beneficiated phosphate rock to a world market price for comparable phosphate rock. PDM at 20; IDM at 30; USDOC Preliminary Calculation Memo at 7–8, C.R. 218–19, P.R. 189–90 (May 1, 2023); USDOC
Consol. Court No. 23-00239 Page 9
Final Calculation Memo – JSC Apatit at 2–3, C.R. 226–27, P.R. 243–44 (May 1, 2023). Commerce identified BPL content and the type of ore deposit (i.e., igneous or sedimentary) as qualities relevant for selecting “comparable phosphate rock.” IDM at 30. Commerce’s benchmark data accounted for a small percentage of total global exports of phosphate rock.6 The court held that Commerce unreasonably limited the benchmark to phosphate rock from igneous ore deposits. See Archer Daniels Midland I at 1360. The court reasoned that “while Commerce reasonably found that the record indicated that sedimentary and igneous rock have different production processes, it extrapolated that the cost to produce each kind of rock differs and that therefore, the price of sedimentary and igneous rock is not comparable, even if BPL content does not differ.” Id. The court remanded for Commerce “to either present record evidence to show that the phosphate rock market is significantly driven by the distinction between sedimentary and igneous rock or to reconstruct the tier three benchmark.” Id.
In its Remand Results, Commerce maintained its benchmark of phosphate rock exports from countries with igneous ore deposits and comparable BPL grades to Russia. Remand Results at 9–11. Specifically, Commerce relied on POR export prices from Global Trade Atlas (“GTA”) for Finland, Brazil, and South Africa under Harmonized System (“HS”) codes 2510.10 and 2510.20. Id. at 3. Commerce cited record evidence to support its argument that the production processes for phosphate rock from igneous and sedimentary ore reserves differ and, accordingly,
6 According to JSC Apatit, Commerce’s benchmark relies on data accounting for 0.02 percent of the total global production of phosphate rock and less than 0.12 percent of the total global exports of phosphate rock in 2021. JSC Cmts. at 21 (citing Letter from Hogan Lovells, JSC Apatit Benchmark, Appx. 9, C.R. 179, 182–191, P.R. 143, 146–155 (Mar. 15, 2023)). According to ADM, Commerce’s benchmark accounts for 0.63 percent of total export volumes in 2021 of phosphate rock with BPL content of 78 percent or higher. Archer Daniels Midland Company’s Mot. for J. on the Agency R., ECF No. 41-1 (Aug. 17, 2024). Neither Mosaic nor the government meaningfully dispute these characterizations.
Consol. Court No. 23-00239 Page 10
the cost to produce each kind of rock differs. Id. at 5–6. Commerce pointed to evidence from the Davis and Ptáček Reports that differences in development costs of phosphate rock could result in significant differences in lease bonuses,7 that Russian mining lease bonuses “cannot be compared with” those of Jordan because Russian deposits are igneous and Jordanian deposits are “a completely different type of mineralization (sedimentary),” and that calcination and organic acid leaching beneficiation processes are applicable to sedimentary with calcareous gangue8 ore but not to igneous ores. Id. at 6–7 (citation modified). From this evidence, Commerce extrapolated that “major cost items such as capital costs, costs for developing reserves, and energy costs” differ between the igneous and sedimentary production processes. Id. at 7. Commerce concluded that this cost difference is a “condition of purchase or sale,” id. at 11, that “will logically impact” world market prices for phosphate rock because differences in production costs of each kind of rock would hypothetically lead to one kind of rock being “profitable at lower world market prices relative to the other type[.]” Id. at 7–8.
To the contrary, costs and profitability are not relevant for a tier-three, price-based benchmark if the price is driven by another factor. Thus, these are not relevant conditions of purchase or sale. Prices, not costs per se, are what is at issue. If the BPL content of beneficiated rock drives the world market prices, then only difference in BPL content matters in selecting a benchmark. Commerce has not cited evidence that the difference between ore from igneous or sedimentary sources significantly impacts the world market price of beneficiated phosphate rock. Much of Commerce’s Remand Results and the parties’ subsequent briefing focuses on the relative costs to beneficiate igneous and sedimentary ore, but never get to the core issue. Assuming,
7 A lease bonus is an “up-front cash payment for access to the lease area.” Davis Report at 4. 8 Gangue is “the worthless rock or vein matter in which valuable metals or minerals occur.” Gangue, MERRIAM-WEBSTER.COM, https://www.merriam-webster.com/dictionary/gangue.
Consol. Court No. 23-00239 Page 11
arguendo, that the cost to produce igneous and sedimentary phosphate rock differs, Commerce has not pointed to evidence on this record that these purported cost differences significantly drive the export prices of phosphate rock on the world market. Specifically, in its Remand Results, Commerce did not point to any significant record evidence that phosphate rock with similar BPL levels produced from igneous and sedimentary ore reserves have different export prices. Rather, Commerce extrapolated that a higher cost of production necessarily means that a good will sell at a higher price on the market.9 See Remand Results at 9. This is not a logical conclusion from the record evidence. The record evidence indicates that the BPL content of phosphate rock, not whether the rock is produced from sedimentary or igneous ore, drives the prices in the phosphate rock market from whence the benchmark comes.10 If the product at issue were phosphate ore, then
9 There may be an implication which can be drawn from the Remand Results that because Commerce is looking at JSC Apatit’s cost buildup, the difference in cost to beneficiate igneous and sedimentary ore into beneficiated phosphate rock is significant because lower beneficiation costs would make JSC Apatit’s underlying mining rights more valuable than mining rights for sedimentary ore. Remand Results at 8–9 (“[E]ven if . . . the final marketable product (i.e., phosphate rock) produced from igneous and sedimentary reserves is the same, differences in production costs impact the value of the underlying good conveyed via the mining rights.”). Here, Commerce chose to use JSC Apatit’s beneficiated phosphate rock per-unit cost buildup plus profit and compared that to a world market price for comparable phosphate rock. PDM at 20; IDM at 30; USDOC Preliminary Calculation Memo at 7–8; USDOC Final Calculation Memo – JSC Apatit at 2–3. At this stage, however, there is no dispute about how to calculate JSC Apatit’s cost buildup. The dispute is over the benchmark to be compared to the results of the cost buildup method. Having decided to use world price as a tier-three benchmark, Commerce had to use a real world price for beneficiated phosphate rock. Accordingly, because Commerce used the price of phosphate rock to compare to JSC Apatit’s mining rights cost plus profit (i.e., price) substitute, the relative costs to beneficiate igneous and sedimentary ore are not relevant to the current dispute, even if one could determine them. Assuming, arguendo, that ore used by respondent costs less to beneficiate, it would be accounted for in the cost buildup already selected, resulting in a higher potential subsidy than using greater production costs would generate. Further, whatever the value is of the ore input, it is accounted for in the cost buildup. If it is valuable because of lower beneficiation costs, it results in an overall smaller cost buildup thereby increasing the subsidy determination after comparison of cost plus profit to world price. 10 Chapter 8 of the Ptáček Report notes that “igneous phosphate ores are often low in grade (less than 5% P2O5) but can be upgraded to high-grade products (from about 35% to over 40% P2O5).” Letter from WilmerHale, Mosaic Benchmark, Ex. 20 at 337, P.R. 133 (Mar. 13, 2023). It does not
Consol. Court No. 23-00239 Page 12
the costs of beneficiating igneous and sedimentary rock could impact the price for the ore. But the benchmark product is not ore—it is phosphate rock. This, coupled with the fact that the benchmark selected by Commerce reflects only a tiny percentage of global phosphate rock exports, see supra note 6, leads the court to conclude that Commerce unreasonably limited the benchmark to the price of phosphate rock known to be from igneous ore deposits. Accordingly, on this record, Commerce’s only remaining option in constructing the tier-three benchmark is to include world phosphate rock price data that was previously excluded solely on the distinction between igneous and sedimentary ore.
II. Commerce Reasonably Constructed its Natural Gas Benchmark JSC Apatit asserts that Commerce’s use of a tier-three benchmark for natural gas was unreasonable and inconsistent with Commerce’s past practice.11 JSC Apatit Cmts. at 9 n.4. Specifically, JSC Apatit argues that Kazakh natural gas, whether raw or refined, is comparable to Russian gas.12 Id. at 4, 10. JSC Apatit points to evidence that Kazakh natural gas was exported to
suggest, however, that this leads to different prices for beneficiated rock on the world market based on anything other than BPL content. 11 JSC Apatit argues that Commerce impermissibly relied on data from calendar year 2021 and should have included data from December 2020. JSC Apatit Cmts. at 19. The court ruled on this issue in its previous opinion, see Archer Daniels Midland I at 1354–55, and will not revisit it now. JSC Apatit also argues that Commerce impermissibly relied on a simple, rather than a weighted, average to calculate the natural gas benchmark in the Remand Results. JSC Apatit Cmts. at 19– 20. JSC Apatit notes that applying a simple average is contrary to Commerce’s practice and adds that JSC Apatit provided the weighted-average European Organization for Economic Cooperation and Development natural gas end-use price. Id. at 20. The government responds that where there is more than one commercially available market price to construct a benchmark price, Commerce’s practice is to use a simple average for the prices. Gov. Cmts. at 13. JSC Apatit has not demonstrated how a weighted average would lead to a more reasonable result on this record. Accordingly, the court concludes that Commerce reasonably applied a simple average. 12 JSC Apatit contends that Commerce has repeatedly found that Kazakh-origin natural gas is sufficiently comparable to Russian-origin natural gas to serve as a tier-two benchmark and that Commerce must explain when it changes a longstanding practice. JSC Apatit Cmts. at 11. Here, Commerce explained that it found that Kazakh natural gas is not comparable to Russian refined gas because of the additional processing needed for Kazakh natural gas to be fit for consumption.
Consol. Court No. 23-00239 Page 13
Russia through a pipeline that terminates in Russia, that pipeline maps demonstrate that the Republic of Kazakhstan’s (“Kazakhstan”) pipelines transport both raw and processed gas to domestic and international markets, and that Kazakh exported gas falls under the same HTS code as Russian gas.13 Id. at 7–10. JSC Apatit argues that the fact that raw gas requires further processing is not a basis for categorical exclusion from consideration as a benchmark price because such processing is a common feature in global gas markets and routinely reflected in market pricing.14 Id. at 11–12. JSC Apatit also argues that, for a tier-two benchmark, Commerce need only apply a benchmark that would be available to purchasers in the country in question and need not show that the respondents would have purchased the same good being employed in constructing the benchmark.15 Id. at 13–14.
Remand Results at 13–14. This explanation is sufficient. Further, each administrative review is a “separate exercise of Commerce’s authority that allows for different conclusions based on different facts in the record.” Qingdao Sea-Line Trading Co. v. United States, 766 F.3d 1378, 1387 (Fed. Cir. 2014) (citation omitted). 13 JSC Apatit contends that the court has “upheld Commerce’s use of data based on the HTSUS subheading corresponding to the inputs consumed by a respondent.” JSC Apatit Cmts. at 10 (citing Archer Daniels Midland Co. v. United States, 968 F. Supp. 2d 1269, 1278 (CIT 2014); RZBC Grp. Shareholding Co. v. United States, 2016 WL 3880773, at *10 (CIT 2016)). In these cases, however, the court held that Commerce reasonably considered HTS headings and subheadings as part of its analysis. The court did not rule that a common HTS code is always definitive evidence of comparability. See Archer Daniels Midland Co., 968 F. Supp. 2d at 1278–79; RZBC Grp. Shareholding Co., 2016 WL 3880773, at *10. 14 JSC Apatit notes that the court has held that differences in grade or processing do not preclude use as a benchmark unless the products are “so dissimilar” as to make comparison unfair. JSC Cmts. at 12 (citing Essar Steel Ltd. v. United States, 721 F. Supp. 2d 1285, 1294 (CIT 2010)). 15 JSC Apatit argues that Commerce should use its tier-two Kazakh gas benchmark because Kazakh gas is available to Russian purchasers despite Gazprom’s (a GOR-owned entity) monopoly over gas distribution. JSC Apatit Cmts. at 17. JSC Apatit adds that Commerce failed to address whether “sales to a government entity that distorts the market satisfy the meaning of ‘purchaser,’” id. at 14, and argues that Kazakh gas is available to Russian “purchasers.” Id. at 17. The government responds that the court did not instruct Commerce to address both the issues of comparability and availability to purchasers. Gov. Cmts. at 7 (citing Archer Daniels Midland I at 1367). The court instructed that “if the sales are comparable, Commerce must address why sales to a government entity that distorts the market was within the intended meaning of ‘purchaser’ in the regulation and how that fits into the logic of the three-level benchmark scheme.” Archer
Consol. Court No. 23-00239 Page 14
Mosaic and the government respond that the gas exported from Kazakhstan to Russia is “raw high-sulfur and unrefined” natural gas that is a byproduct of oil production that is not commercially usable without additional costly processing. Gov. Cmts. at 8 (citing Remand Results at 13); see also Mosaic Cmts. at 20–21. The government concludes that Commerce reasonably found immediately commercially usable and not immediately commercially usable products not to be comparable for tier-two benchmark purposes. Gov. Cmts. at 10. The government asserts that JSC Apatit cannot establish that HTS subheadings are definitive markers of comparability, particularly when other record evidence suggests that the two types of gas are not comparable. Id. Mosaic adds that Commerce sufficiently responded to JSC Apatit’s argument that requiring further processing before end-use is a common feature in global gas markets. Mosaic Cmts. at 22. Mosaic notes that this fact actually underscores its conclusion that refined natural gas is a more expensive product than raw gas. Id.
As discussed above, Commerce measures the adequacy of remuneration by using a three-
tiered hierarchy. 19 C.F.R. § 351.511(a)(2). Relevant here is the choice between tier-two or tier- three analyses. Commerce conducts a tier-two analysis by comparing the government price to a “world market price where it is reasonable to conclude that such price would be available to purchasers in the country in question.” Id. § 351.511(a)(2)(ii). If a specific world market price is unavailable to purchasers in the country in question, Commerce proceeds to a tier-three analysis. See id. § 351.511(a)(2)(iii).
Commerce originally constructed a tier-two benchmark to assess Russia’s provision of natural gas to JSC Apatit using the export prices of natural gas under HS code 2711.21.0000 from
Daniels Midland I at 1367 (emphasis added). Commerce, therefore, properly did not reach this second issue once it decided that the sales in question were not comparable.
Consol. Court No. 23-00239 Page 15
Kazakhstan based on export data from the Bureau of National Statistics (“BNS”) of the Agency for Strategic Planning and Reforms of the Republic of Kazakhstan. IDM at 48; Remand Results at 3 (citing IDM at 44–49). The court remanded to Commerce to consider two issues: “the question of comparability of the benchmark third-party sales and sales into Russia” and “if the sales are comparable, . . . why sales to a government entity that distorts the market was within the intended meaning of ‘purchaser’ in the regulation and how that fits into the logic of the three-level benchmark scheme.” Archer Daniels Midland I at 1367.
On remand, Commerce found that “the raw natural gas exported by Kazakhstan is not comparable to the refined natural gas available in the Russian market that JSC Apatit purchased from GOR authorities during the POR.” Remand Results at 13. Commerce noted that the high- sulfur content in natural gas renders Kazakh gas unusable in commercial and industrial settings. Id. at 13–14. Commerce adds that the bulk of Kazakhstan’s natural gas output requires expensive additional processing to be fit for consumption, and the unprocessed “sour” or “raw” gas is reinjected into wells to support oil production or disposed of. Id. Commerce concluded that the raw Kazakh natural gas price is not an appropriate tier-two benchmark because such gas is not comparable to the natural gas used to produce the subject good, and that there are no viable tier- two benchmarks to compare with JSC Apatit’s purchases of natural gas. Id. at 14. Commerce noted—as the court noted in its prior opinion, see Archer Daniels Midland I at 1366–67—that Gazprom’s prices are not set in a manner consistent with market principles as required by the next step of the analysis. Remand Results at 15; see also 19 C.F.R. § 351.511(a)(2)(iii). Accordingly, Commerce applied the regional European Organization for Economic Cooperation and Development natural gas price from the International Energy Agency (“IEA”) to construct a tier-
Consol. Court No. 23-00239 Page 16
three benchmark. Id. at 15–16. Commerce calculated a subsidy rate of 22.13 percent ad valorem for this input. Id. at 17.
Commerce reasonably concluded that the raw natural gas imported from Kazakhstan was not comparable to JSC Apatit’s natural gas purchases. The court finds Commerce’s explanations set forth above to be reasonable and well supported. The record evidence substantially supports the conclusion that the products are not comparable. Thus, Commerce reasonably constructed a tier-three benchmark for JSC Apatit’s natural gas purchases.
CONCLUSION
The court sustains Commerce’s natural gas benchmark. For the foregoing reasons, the Court remands to Commerce for reconsideration of the phosphate rock benchmark consistent with this opinion. The remand shall be issued within 75 days hereof. Comments may be filed 30 days thereafter and any response 15 days thereafter.
/s/ Jane A. Restani
Jane A. Restani, Judge
Dated: February 6, 2026 New York, New York