ZHEJIANG SANMEI CHEMICAL IND CO LTD v. HONEYWELL INTERNATIONAL INC (2024)

OPINION

[U.S. Department of Commerce's final determination is sustained.]

Eaton, Judge: The subject merchandise in this case is pentafluoroethane (“R-125”) from the People's Republic of China (“China”). R-125 is a colorless, odorless gas used in refrigerants.¹

Following the filing of a petition by Defendant-Intervenor Honeywell International Inc. (“Honeywell”), the U.S. Department of Commerce (“Commerce” or the “Department”) investigated imports of R-125 from China and found that the subject gas was sold in the United States at less than fair value during the period from July 1, 2020, to December 31, 2020. See Pentafluoroethane (R-125) From the People's Republic of China, 87 Fed. Reg. 1,117 (Dep't of Commerce Jan. 10, 2022) (“Final Determination”) and accompanying Issues and Decision Mem. (Dec. 30, 2021) (“Final IDM”), PR 272; see also Pentafluoroethane (R-125) From the People's Republic of China, 87 Fed. Reg. 12,081 (Dep't of Commerce Mar. 3, 2022) (orders).

Before the court is the motion for judgment on the agency record of Zhejiang Sanmei Chemical Industry Co., Ltd. (“Sanmei”), Shandong Dongyue Chemical Co., Ltd., and Huantai Dongyue International Trade Co., Ltd.² (collectively, “Plaintiffs”). See Pls.' Mem. Supp. J. Agency R. (Oct. 25, 2022) (“Pls.' Br.”), ECF No. 29; Pls.' Reply Br. (Apr. 3, 2023), ECF No. 41.

Sanmei, a producer and exporter of R-125 from China, was the sole mandatory respondent ³ in Commerce's investigation. As will be seen, Sanmei's affiliated toller,⁴ Fujian Qingliu Dongying Chemical Ind. Co., Ltd. (“Qingliu”), produced the only subject R-125 sold in the United States during the period of investigation. See Final IDM at 30 (noting that Qingliu's “sales are the only reviewable sales in the [period of investigation]”).

Neither Shandong Dongyue Chemical Co., Ltd. nor Huantai Dongyue International Trade Co., Ltd. was selected for individual examination, but both were found eligible for the all-others rate.⁵ See Compl. ¶¶ 8, 11, ECF No. 8.

By their motion, Plaintiffs challenge Commerce's determination of the final estimated weighted-average dumping margin of 277.95% for Sanmei, and its use of this margin as the all-others rate. See Pls.' Br. at 2. Specifically, Plaintiffs argue that substantial evidence does not support Commerce's (1) employment of the “intermediate-input method,” i.e., the use of the value of an intermediate input when constructing the normal value of the subject R-125, i.e., anhydrous hydrofluoric acid (“AHF”), instead of the value of the upstream raw materials that went into making AHF; (2) denial of offsets for the by-products that were generated in the production of the subject R-125, and (3) calculation of the surrogate inland freight rate. See id. at 5-24.

Defendant the United States (“Defendant”), on behalf of Commerce, opposes Plaintiffs' motion and asks the court to sustain the Final Determination. See Def.'s Resp. Br. (Feb. 21, 2023) (“Def.'s Br.”), ECF No. 38. Defendant-Intervenor Honeywell, a U.S. producer of R-125, also asks the court to sustain Commerce's Final Determination. See Def.-Int.'s Resp. Br. (Feb. 21, 2023), ECF No. 36.

Jurisdiction lies under 28 U.S.C. § 1581(c) and 19 U.S.C. § 1516a(a)(2)(B)(i). For the following reasons, the court sustains Commerce's direct valuation of the intermediate input AHF, its denial of Sanmei's claimed by-product offsets, and its calculation of the surrogate inland freight rate.

BACKGROUND

On February 1, 2021, Commerce initiated an antidumping duty investigation of R-125 from China, covering the period from July 1, 2020, to December 31, 2020. See Pentafluoroethane (R-125) From the People's Republic of China: Initiation of Less-Than-Fair-Value Investigation, 86 Fed. Reg. 8,583 (Dep't of Commerce Feb. 8, 2021). Commerce selected the Russian Federation (“Russia”) as the surrogate country.⁶ Preliminary Surrogate Value Mem. (Aug. 10, 2021) at 1, PR 223.

I. Antidumping Questionnaires

As a part of its investigation, between March and July 2021, Commerce issued initial and supplemental antidumping questionnaires to Sanmei. See, e.g., Initial Quest. Secs. A, B, C, and D (Mar. 12, 2021) (“Initial Quest.”), PR 94; Suppl. Secs. A, C, and D Quest. (July 6, 2021), PR 177. Sanmei filed timely responses on behalf of itself and its affiliated toller, Qingliu. See, e.g., Sanmei's Sec. D Resp. (May 11, 2021), PR 149, CR 102; Sanmei's Suppl. Sec. D. Resp. (July 26, 2021), PR 198, CR 127.

As it turned out, only R-125 that was produced by Qingliu was sold in the United States during the period of investigation. See Final IDM at 30 (noting that Qingliu's “sales are the only reviewable sales in the [period of investigation]”).

Sanmei's initial and supplemental Section D questionnaire responses included factors of production information for Qingliu. As Commerce would later find, however, some information for Qingliu was missing. Specifically, “information regarding Qingliu's production process and its consumption of water as a direct input [was] missing from the record.” Id. at 35. To fill in the missing information, Commerce found that it would “rely on Sanmei's production process and reported consumption of water as a direct input as neutral [facts available].” Id.; see also 19 U.S.C. § 1677e(a)(1) (providing that if “necessary information is not available on the record” Commerce must use “facts otherwise available”). Notably, Qingliu's water consumption data was not available because the company “used water pumped from the nearby river for production,” and its “water usage was not measured.” Sanmei's Sec. D Resp. at 16.

As described in Sanmei's questionnaire responses, the production process used to make R-125 is comprised of two stages. In the first stage, the intermediate input AHF is either self-produced by the R-125 manufacturer or procured from a supplier. See id. at 12. Here, Sanmei reported that “Qingliu used their self-produced AHF ․ to produce R-125.” Id. at 3. The upstream raw materials that Qingliu used to produce AHF in-house included purchased materials, i.e., fluorite powder, 105% sulfuric acid, and 98% sulfuric acid. Id. at 12 (“For the first processing stage, [Sanmei] ․ and [Qingliu] produce AHF with purchased materials such as fluorite powder, 105% sulfuric acid and 98% sulfuric acid.”).

In the second stage, AHF is combined with another input, perchloroethylene (“PCE”), resulting in R-125. Id. Sanmei reported that it purchased the PCE (and reported the purchase price), from market economy suppliers, which was then provided to Qingliu. See id. at 8, Ex. D-8. Qingliu then combined the AHF that it produced in-house and the PCE purchased by Sanmei to produce the R-125 covered by the investigation.

Regarding water consumption, neither Qingliu nor Sanmei reported water as a direct input in the production of the intermediate input AHF and subject R-125. For its part, Qingliu did not track its water consumption. As to Sanmei, the company tracked its water consumption, but only reported water as an input for energy production, not as a direct input used to produce AHF and R-125. See Sanmei's Sec. D Resp. Ex. D-7 (reporting 98% sulfuric acid, 105% sulfuric acid, and fluorite powder as material inputs, but reporting water under “Energy”) and Ex. D-8 (list of inputs). That is, Sanmei only reported water in response to the “Energy Inputs” section of Commerce's questionnaire. See id. Ex. D-8. Sanmei later argued in its administrative case brief, however, that “[w]ater is an important factor in the production of both AHF and R-125 and has a tremendous impact on yield loss and by-product production.” Sanmei's Case Br. (Oct. 19, 2021) at 12, PR 268; see also Final IDM at 23 (quoting case brief).

Sanmei, as the mandatory respondent, claimed offsets for each of four by-products that were generated during Qingliu's in-house production of the intermediate input AHF and subject R-125: (1) the fluosilicic acid and (2) fluorine gypsum that were by-products generated in the production of AHF, and in addition (3) the hydrochloric acid and (4) R-134a that were by-products generated in the production of R-125. See Sanmei's Sec. D Resp. at 17, Ex. D-12.

Regarding freight expenses within China, Commerce's questionnaire asked for the “distance in kilometers from the plant to the nearest port where the plant can receive supplies shipped in international containers.” Sanmei's Sec. D Resp. at 19. Sanmei reported that “[t]he distance[ ] from the Respondent [Sanmei] to Wenzhou port, the nearest seaport, is 187.2 kilometers.” Id.

II. Preliminary Determination

On August 10, 2021, Commerce issued its preliminary determination that imports of R-125 from China were sold in the United States at less than fair value during the period of investigation. See Pentafluoroethane (R-125) From the People's Republic of China, 86 Fed. Reg. 45,959 (Dep't of Commerce Aug. 17, 2021) (“Preliminary Determination”) and accompanying Preliminary Decision Mem. (Aug. 10, 2021) (“PDM”), PR 220.

Importantly, in the Preliminary Determination, Commerce found that the information on the record was inadequate to value the upstream raw materials used to make the intermediate input AHF. Thus, Commerce preliminarily found that it would value AHF directly, using the intermediate-input method, i.e., the use of the value of AHF when constructing the normal value of the subject R-125, instead of the value of the upstream raw materials that went into making AHF. See PDM at 34-35.

In addition, Commerce found that the information on the record was inadequate to support Sanmei's claims for offsets, either for by-products generated during Qingliu's in-house production of AHF, i.e., fluosilicic acid and fluorine gypsum, or for by-products generated during Qingliu's production of subject R-125, i.e., hydrochloric acid and R-134a. See id. at 35.

To value freight expenses within China, Commerce used data from the World Bank's publication, Doing Business 2020: Russian Federation. See PDM at 31. Commerce calculated a U.S. dollars per metric ton per kilometer (USD/MT/KM) rate for transport within Russia using export and import data for two cities, Moscow and St. Petersburg. See Preliminary Surrogate Value Mem. at 10 (stating that the Doing Business report “gathers information concerning the distances and costs to transport products in a container for export [from Moscow and the city of St. Petersburg] to the border crossing at the St. Petersburg Port and import from the border crossings at Krasnaya Gorka ․ [to Moscow], and [from] St. Petersburg Port [to the city of St. Petersburg].”).

Following issuance of the Preliminary Determination, Commerce verified Sanmei's reported information. In lieu of on-site, in-person verification,⁷ Commerce issued a verification questionnaire to which Sanmei timely responded. See Sanmei's Verification Resp. (Sept. 20, 2021), PR 255.

III. Final Determination

On December 30, 2021, Commerce issued the Final Determination. As it had in the Preliminary Determination, Commerce valued the intermediate input AHF directly (instead of valuing the upstream raw materials) based on two main findings: (1) “Sanmei failed to demonstrate that it accurately reported or substantiated the consumption of the upstream material inputs used in the production of AHF,” and (2) “Sanmei failed to provide sufficient evidence to support its claim that it can differentiate between self-produced AHF and AHF purchased from other sources.” Final IDM at 21.

Commerce also denied Sanmei's claimed offsets for the by-products generated in the production of the intermediate input AHF and the subject R-125. Commerce found that, though “[t]he evidence on the record demonstrates that Sanmei and Qingliu sold fluosilicic acid, fluorine gypsum [by-products of the intermediate input, AHF], hydrochloric acid, and R-134a [by-products of the subject R-125]” during the period of investigation, which demonstrated that the by-products had commercial value, “there [was] insufficient evidence to corroborate [1] the volume of by-products generated in the production of subject merchandise and [2] whether the sales of by-products included quantities generated in the production of non-subject merchandise or from other purchases.” Id. at 29.

As to freight expenses, Commerce left its surrogate inland freight calculation unchanged from the Preliminary Determination. Id. at 42-43.

Commerce calculated a final estimated weighted-average dumping margin of 277.95% for Sanmei, which it assigned as the all-others rate to the eligible non-individually examined companies. See Final Determination, 87 Fed. Reg. at 1,118. This action followed.

STANDARD OF REVIEW

The court will sustain a determination by Commerce unless it is “unsupported by substantial evidence on the record, or otherwise not in accordance with law.” 19 U.S.C. § 1516a(b)(1)(B)(i).

LEGAL FRAMEWORK

In an antidumping case, Commerce must determine whether goods are being sold, or are likely to be sold, in the United States at less than fair value. See 19 U.S.C. § 1673. Commerce generally makes this determination by comparing export price and normal value, as adjusted. See id. §§ 1677a, 1677b.

Export price, or U.S. price, is “the price at which the subject merchandise is first sold (or agreed to be sold) before the date of importation by the producer or exporter of the subject merchandise outside of the United States to an unaffiliated purchaser in the United States or to an unaffiliated purchaser for exportation to the United States.” Id. § 1677a(a).

Normal value is “the price at which the foreign like product is first sold (or, in the absence of a sale, offered for sale) for consumption in the exporting country.” Id. § 1677b(a)(1)(B)(i). Where the exporting country is a nonmarket economy, such as China, the statute directs Commerce to determine normal value using surrogate values for the factors of production used to make the subject merchandise and for general expenses and profit. See id. § 1677b(c)(1); Fujian Yinfeng Imp. & Exp. Trading Co. v. United States, 46 CIT __, __, 607 F. Supp. 3d 1301, 1307 (2022).

I. Determining Normal Value in the Nonmarket Economy Context

Where, as here, the exporting country is a nonmarket economy, Commerce determines the value of each of the factors of production using surrogate data from a market economy country. Factors of production include, but are not limited to, the “hours of labor required,” the “quantities of raw materials employed,” and “amounts of energy and other utilities consumed.” 19 U.S.C. § 1677b(c)(3)(A)-(C). The statute directs Commerce to use the “best available information” to value the factors of production. Id. § 1677b(c)(1). When valuing these factors, Commerce must “utilize, to the extent possible, the prices or costs of factors of production in one or more market economy countries that are—(A) at a level of economic development comparable to that of the nonmarket economy country, and (B) significant producers of comparable merchandise.” Id. § 1677b(c)(4). Additionally, Commerce considers and “selects ․ surrogate values that are publicly available, are product-specific, reflect a broad market average, and are contemporaneous with the period of review.” Qingdao Sea-line Trading Co. v. United States, 766 F.3d 1378, 1386 (Fed. Cir. 2014) (citation omitted). To determine what constitutes the best available information, Commerce must act according to the statute's purpose: “to obtain the most accurate dumping margins possible.” Shandong Huarong Gen. Grp. Corp. v. United States, 25 CIT 834, 838, 159 F. Supp. 2d 714, 719 (2001) (citation omitted), aff'd, 60 F. App'x 797 (Fed. Cir. 2003).

II. Direct Valuation of Intermediate Input Instead of Upstream Raw Materials

An “intermediate input” is an input made from upstream raw materials. Where a respondent in a nonmarket economy country self-produces an intermediate input and uses the intermediate input to produce the subject merchandise, Commerce has developed a practice, under certain circumstances, of valuing the intermediate input directly, instead of the upstream raw materials. See Anshan Iron & Steel Co. v. United States, 28 CIT 1728, 1730, 358 F. Supp. 2d 1236, 1238 (2004) (stating that intermediate inputs were “produced [by a respondent] from various purchased materials” and were used “[i]n the process of producing [subject merchandise]”).

The method of directly valuing an input is called the “intermediate-input method.” See CP Kelco US, Inc. v. United States, No. 13-00288, 2015 WL 1544714, at *10 (Ct. Int'l Trade Mar. 31, 2015) (“Under the intermediate-input method, Commerce will occasionally treat a self-produced product as an input even though it has been made in house.”). Commerce will apply the intermediate-input method, for example, under circumstances where “it is clear that attempting to value the factors used in a production process yielding an intermediate product would lead to an inaccurate result because a significant element of cost would not be adequately accounted for in the overall factors buildup.” Jining Yongjia Trade Co. v. United States, 34 CIT 1510, 1516 (2010) (not reported in the Federal Supplement); see also Final IDM at 21 (stating circumstances where intermediate-input method is employed). Generally, Commerce does not apply the intermediate-input method “unless there are questions about the accuracy and validity of reported factors of production,” i.e., the upstream raw materials. Linyi Chengen Imp. & Exp. Co. v. United States, 44 CIT __, __, 487 F. Supp. 3d 1349, 1355 (2020).

In addition to valuing directly any inputs that are self-produced (again, under certain circumstances, Commerce will employ the intermediate-input method), Commerce will also value directly any inputs that are purchased. See CP Kelco US, Inc., 2015 WL 1544714, at *10 (“[I]f a producer buys a necessary product readymade, then Commerce will value the product itself as an input.”).

III. Adjusting Normal Value Through By-Product Offsets

Neither the antidumping statute nor Commerce's regulations address by-product offsets.⁸ Commerce has a practice, however, of adjusting ⁹ normal value by providing offsets for by-products generated during the production of subject merchandise when not all raw materials are included in the final product. NTSF Seafoods Joint Stock Co. v. United States, 44 CIT __, __, 487 F. Supp. 3d 1310, 1322 (2020) (“As not all raw materials are incorporated into the final product, Commerce provides offsets for byproducts generated during the production process.” (citations omitted)).

“Generally, ․ the Department's practice has been to grant an offset to normal value, for sales of by-products generated during the production of subject merchandise, if the respondent [i.e., producer] can demonstrate that the by-product is either resold or has commercial value and re-enters the respondent's production process.” Arch Chems., 33 CIT at 956 (footnote omitted); see also NTSF Seafoods, 44 CIT at __, 487 F. Supp. 3d at 1322.

When deciding whether to grant a respondent producer's claim for a by-product offset, Commerce looks at “whether the respondent's production process for subject merchandise actually generated the amount of [by-product] claimed as a by-product offset.” Arch Chems., Inc. v. United States, 35 CIT 424, 428 (2011) (not reported in the Federal Supplement) (citation omitted). Thus, the information on the record regarding the quantity of the inputs used, and the by-products generated, is important to the Department's decision of whether to grant an offset. “Commerce values byproduct offsets based on the best available information.” NTSF Seafoods, 44 CIT at __, 487 F. Supp. 3d at 1322 (citation omitted). The party claiming the by-product offset bears the burden of substantiating the offset and “must present Commerce with sufficient information to support its claims.” Id. (citing Arch Chems., 33 CIT at 956).

DISCUSSION

I. Commerce's Direct Valuation of the Intermediate Input AHF Is Supported by Substantial Evidence and Otherwise in Accordance with Law

In the Final Determination, Commerce applied the intermediate-input method to determine the value of one input, i.e., AHF, directly instead of valuing the upstream raw materials reportedly used to make it (including fluorite powder, 105% sulfuric acid, and 98% sulfuric acid). The Department's decision to use the value of the intermediate input AHF (when constructing normal value) was based on several factual findings.

First, Commerce found that it could not accurately value the upstream raw materials because the record lacked any information regarding how Sanmei and Qingliu accounted for yield loss, i.e., the amounts of raw materials consumed but not incorporated into the intermediate input AHF or R-125, since neither company tracked yield loss. See Final IDM at 23 (“[W]e continue to find that Sanmei did not provide adequate evidence for how it accounts for the yield loss from the fluorite powder, 105 percent sulfuric acid, and 98 percent sulfuric acid inputs in the consolidated [factors of production] database.”).

Next, Commerce found that the record lacked sufficient information on water consumption for either Sanmei or Qingliu, so it could not accurately value water as an upstream raw material. That is, Sanmei failed to report water as a direct input, and Qingliu did not measure its water consumption. See PDM at 34 (“Water is ․ a significant input in the production of AHF and R-125, but Sanmei failed to report water as a direct material input ․”); see Final IDM at 23 (“Qingliu does not measure its water consumption ․”).

As noted, Commerce filled the gap in the record regarding water consumption, when producing the subject R-125, using Sanmei's reported consumption of water as an energy input ¹⁰ as neutral facts available in the Final Determination. Final IDM at 35 (finding “that information regarding Qingliu's production process and its consumption of water as a direct input [for R-125] is missing from the record. Accordingly, we continue to rely on Sanmei's production process and reported consumption of water as a direct input [for R-125] as neutral [facts available] for the final determination.”). Commerce found that this same data was “insufficient to justify valuing the upstream inputs of AHF.” Id. (finding that “while [Commerce] will use Sanmei's reported water consumption as [facts available] for the final determination, we continue find that the reported water consumption is insufficient to justify valuing the upstream inputs of AHF or granting a by-product offset” as discussed in earlier comments). In other words, Commerce found Sanmei's reported water consumption data usable as a direct input to construct the normal value of R-125, but not to value the upstream raw materials used to make AHF.

Commerce's stated reason for not using Sanmei's water consumption data for purposes other than constructing the normal value of R-125 was a concern for accuracy, i.e., to “avoid using inaccurate data stemming from the fact that Qingliu does not measure its water consumption.” Id. at 23-24 (finding that “by valuing AHF directly, we are avoiding a burdensome analysis that would require more information, which the record does not contain, and we avoid using inaccurate data stemming from the fact that Qingliu does not measure its water consumption, which is a significant cost element with a tremendous impact on yield loss and by-product production”).

Finally, Commerce found it could not accurately value the upstream raw materials used to make AHF because the record was unclear regarding the extent to which different types (self-produced or purchased) of AHF were used in the workshops that produced subject and non-subject merchandise. Commerce found that though Sanmei and Qingliu reported that they both self-produced and purchased AHF, “there is no record evidence demonstrating how Sanmei differentiates between the production of AHF and purchases of AHF.” Final IDM at 22. In so finding, Commerce rejected Sanmei's claim that there was sufficient evidence to show that only self-produced AHF was used in subject merchandise workshops: “We disagree that a simple comparison of production, consumption, and warehouse-out quantities of AHF is conclusive on its own merits.” Id. Commerce found that the lack of record evidence documenting the extent of the use of self-produced and purchased AHF was relevant to its analysis, stating by way of explanation:

Sanmei reported that it produces and purchases AHF from affiliated and unaffiliated Chinese suppliers and that this AHF is either resold and shipped directly to customers or is used in the production of non-subject merchandise. However, the record remains unclear as to how Sanmei is able to distinguish whether self-produced AHF or purchased AHF is used in the various subject and non-subject merchandise producing workshops. Therefore, even if Sanmei produced more AHF than it consumed in the production of R-125, the record still indicates that there were other types of AHF available and used on Sanmei's premises, but the extent of their use is not documented.

Id. In other words, for Commerce, the lack of clarity in the record regarding the extent of the use of self-produced and purchased AHF to make subject and non-subject merchandise justified using the value of AHF (instead of the value of upstream raw materials) when constructing normal value.

Plaintiffs argue that substantial evidence does not support Commerce's finding that it could not accurately value the upstream raw materials that Qingliu used to make AHF. First, Plaintiffs argue that the record contained enough information to calculate yield loss for AHF. While Plaintiffs acknowledge that “neither [Sanmei] nor [Qingliu] tracked loss yields in the ordinary course of business during the period of investigation,” they insist that Sanmei “submitt[ed] accurate unit consumption rates and output rates for the[ ] inputs” used to make AHF (fluorite powder, 105% sulfuric acid, and 98% sulfuric acid), and so it was possible to calculate yield loss regarding Qingliu's production of AHF using the formula they had proposed to Commerce.¹¹ Pls.' Br. at 17-18.

Next, Plaintiffs insist that Sanmei did, in fact, report its water consumption as a direct material input: “Sanmei reported the water factor of production in Field 5.3 [for (“Energy”)] of its [factors of production] databases, which included the entirety of its water usage, including both water consumed as a raw material input and water consumed as an energy component.” Id. at 22. Plaintiffs further argue that Sanmei properly “reported its own water consumption factor as ․ Qingliu's per unit water consumption in the [factors of production] calculations” because Qingliu did not track its water usage.¹² Id. at 23.

Finally, Plaintiffs insist that whether the record shows how Sanmei differentiated between self-produced or purchased AHF was not relevant to Commerce's normal value determination because Qingliu reported that it consumed only self-produced AHF when producing the subject R-125. Id. at 16-17.

Based on the record here, in particular in the absence of yield loss data, Commerce's use of the intermediate-input method was justified. This Court has sustained Commerce's use of the intermediate-input method when “it could not achieve an accurate result in constructing normal value if it used the [upstream factors of production] reported by respondent.” Jining Yongjia Trade Co., 34 CIT at 1517 (sustaining Commerce's direct valuation of the intermediate input).

The parties do not dispute that yield loss information is required to calculate the amounts of raw materials, including fluorite powder, 105% sulfuric acid, and 98% sulfuric acid, that ultimately were not incorporated into the AHF that was used by Qingliu to make subject R-125.¹³ Zhejiang Sanhua Co. v. United States, 39 CIT __, __, 61 F. Supp. 3d 1350, 1353 (2015) (stating that yield loss is “the percentage of inputs neither incorporated into the final product nor recovered and sold as scrap”). Nor is there any dispute that neither Sanmei nor Qingliu measured yield loss in the ordinary course of business. In the absence of actual yield loss data, it was reasonable for Commerce to find that it could not determine, with accuracy, the amounts of raw materials (including fluorite powder, 105% sulfuric acid, and 98%sulfuric acid) consumed in the production of AHF and that, thus, directly valuing the intermediate input AHF would lead to a more accurate result.

The court cannot credit Plaintiffs' counterarguments. First, it was not unreasonable for Commerce to decline to use the formula proposed by Sanmei to derive yield loss, when the formula itself requires information that is not found on the record—the amount of water consumed as a direct input in the production of AHF. See Pls.' Br. at 18 (including in the proposed yield loss formula an amount for “calculated water at ․ Qingliu” (emphasis added)). As Sanmei acknowledged in its administrative case brief, water is an important cost element in the production of AHF, but neither Sanmei nor Qingliu reported it as a direct input. Moreover, without yield loss information Commerce could not calculate the amounts of raw materials, including fluorite powder, 105% sulfuric acid, and 98% sulfuric acid, that ultimately were not incorporated into the AHF that was used by Qingliu to make subject R-125.

Next, the court is not convinced, as Plaintiffs now argue, that the amount of water reported in Field 5.3 (“Energy”) of Sanmei's Section D factors of production database “included the entirety of its water usage, including both water consumed as a raw material input and water consumed as an energy component.” Pls.' Br. at 22. A review of the record supports Commerce's finding that Sanmei did not report water as a direct material input. See, e.g., Sanmei's Sec. D Resp. Ex. D-7 (reporting 98% sulfuric acid, 105% sulfuric acid, and fluorite powder as material inputs, but reporting water under “Energy”) and Ex. D-8 (same list of inputs). Rather, Sanmei only reported water in response to the “Energy Inputs” section of Commerce's questionnaire. See id. Ex. D-8. In Sanmei's calculation worksheets of energy, including water, it did not specify that water was a direct material input. See id. Ex. D-11 (providing a “[calculation] worksheet of [Sanmei]'s water,” including “[a]llocation of water for cooling” and “[w]ater of each workshop”). Thus, it is hardly clear that, on this record, using Plaintiffs' proposed formula would lead to an accurate calculation of yield loss particularly with respect to upstream production of AHF. See Jining Yongjia Trade Co., 34 CIT at 1520 (sustaining application of intermediate-input method where reliance on the respondent's reported factors of production “would lead to an inaccurate result because the Department would not be able to account for a significant element of cost” (citation omitted)).

Finally, the absence of information on the record regarding how self-produced AHF was differentiated from purchased AHF is not, as Plaintiffs argue, irrelevant to Commerce's decision whether to value AHF directly. The lack of clarity in the record as to the extent self-produced AHF and purchased AHF were used in warehouses where subject merchandise and non-subject merchandise were made provided reasonable grounds for Commerce to question (1) how much of any chemical was used to make the AHF that was ultimately used to make R-125 (rather than non-subject merchandise), and (2) whether only self-produced AHF was used to make subject R-125. Commerce considered the record evidence, including “production, consumption, and warehouse-out quantities of AHF,” as advocated by Sanmei, and reasonably found that the absence of record evidence as to where the purchased AHF was used (subject or non-subject workshops) favored valuing AHF directly under the intermediate-input method. Accordingly, Commerce's direct valuation of the intermediate input AHF is sustained.

II. Commerce's Denial of Sanmei's Claimed By-Product Offsets Is Supported by Substantial Evidence and Otherwise in Accordance with Law

The court next turns to Commerce's denial of Sanmei's claimed offsets for by-products generated during the production of AHF (i.e., fluosilicic acid and fluorine gypsum) and R-125 (i.e., hydrochloric acid and R-134a).

In the Final Determination, Commerce found:

Consistent with our practice, we continue to deny Sanmei's claims for a by-product offset for its production of AHF, which produces fluosilicic acid and fluorine gypsum, and R-125, which produces hydrochloric acid and R-134a, because Sanmei has not provided sufficient evidence to substantiate the quantity of fluosilicic acid, fluorine gypsum, hydrochloric acid, and R-134a generated from the production of subject merchandise during the [period of investigation].

Final IDM at 28. Commerce stated, by way of explanation, that “providing the production quantity is important because, in considering a by-product offset, Commerce examines whether the by-product was produced from the quantity of the [factors of production] reported and whether the respondent's production process for the merchandise under consideration actually generated the amount of the by-product claimed as an offset.” Id. For Commerce, the evidence on the record was insufficient to (1) establish “the volume of by-products generated in the production of subject merchandise,” and (2) show “whether the sales of by-products included quantities generated in the production of non-subject merchandise or from other purchases.” Id. at 29.

Plaintiffs insist that Sanmei's reporting of the by-products produced and sold by Sanmei and Qingliu provided sufficient information to grant the claimed offsets:

[Sanmei] first reported the monthly quantities of each of the four by-products produced separately by both [Sanmei] and [Qingliu], and separately reported the monthly quantity of each of the four by-products sold by both [Sanmei] and [Qingliu], respectively. Further, in separate worksheets, [Sanmei] further provided the inventory-in records for [Sanmei] and [Qingliu], respectively, that identified which of the four by-products ․ entered into inventory and specified the workshop that generated the entered by-product. Finally, these figures reconciled to the production reports that included the By-Products produced during the production of AHF and the subject merchandise.

Pls.' Br. at 20. In other words, Plaintiffs maintain that they have provided Commerce with sufficient data to track the quantity of by-products, by simple addition and subtraction, through the different production stages and that denying the claimed offsets was unreasonable on this record.

Additionally, to address apparent inaccuracies in Sanmei's reported data, namely that two by-products (fluorine gypsum and hydrochloric acid) weighed more than their respective “main products,” Plaintiffs provided to Commerce, by way of explanation, certain chemical formulas.¹⁴ By “main product” Plaintiffs appear to refer to the product whose manufacture generated each by-product: AHF in the case of the by-product fluorine gypsum and R-125 in the case of the by-product hydrochloric acid.

For Plaintiffs, “[t]he large volume of documentation regarding the production, inventory and sales records of the by-products resulting from the production of AHF and the subject merchandise demonstrate[s] that Sanmei's reported by-product quantities are accurate and that Commerce's denial of the claimed by-product offsets should be reversed.” Pls.' Br. at 20.

For the following reasons, the court finds that Commerce's denial of each of the claimed offsets is supported by the record.

A. AHF By-Product Offsets: Fluosilicic Acid and Fluorine Gypsum

First, Commerce found that some of the same record shortcomings that prevented it from valuing the upstream raw materials that went into making AHF, also justified its decision to deny the claimed offsets for fluosilicic acid and fluorine gypsum. Significantly, Commerce found that the absence from the record of actual yield loss data meant that Commerce could not accurately determine the volume of by-products generated. Final IDM at 29 (“Sanmei and Qingliu do not track yield loss at each stage of production.”).

The absence from the record of yield loss data provided reasonable grounds for Commerce to find that it could not determine, with accuracy, the volume of by-products generated during production of AHF. That is, because there was no actual data on the record regarding yield loss for each of the raw materials that went into making AHF (including fluorite powder, 105% sulfuric acid, and 98% sulfuric acid), Commerce reasonably found that it could not determine whether the quantities of the claimed by-product offsets were accurate and, thus, could provide the basis for an accurate adjustment to normal value. See Arch Chems., 35 CIT at 428-30 (noting that the question when determining whether to grant a by-product offset is “whether the respondent's production process for subject merchandise actually generated the amount of [by-product] claimed as a by-product offset” (citation omitted); see also 19 C.F.R. § 351.401(b)(1) (2021) (when seeking to adjust normal value, “[t]he interested party that is in possession of the relevant information has the burden of establishing to the satisfaction of the Secretary the amount and nature of a particular adjustment”).

The presence on the record of Plaintiffs' proposed chemical formulas fails to convince the court that Commerce's denial of the offset was unreasonable. Plaintiffs acknowledge that the record information they urge Commerce to rely on, in the absence of actual by-product data, contained apparent inaccuracies that had to be accounted for. For Plaintiffs these inaccuracies could be explained through the use of formulas they proposed. Commerce disagreed:

[T]he chemical formulas that Sanmei provided may indeed be the standard equations for AHF and R-125, but, as Sanmei is aware, the output is a function of the input. Therefore, while Sanmei and Qingliu can demonstrate the amount of AHF and R-125 produced, both formulas remain unsubstantiated regarding the by-products. Specifically, for AHF, Sanmei has not demonstrated that it can adequately separate the production of AHF from sales [i.e., purchases] of AHF.

Final IDM at 29 (emphasis added). In other words, even taking into account these formulas, Commerce found that Sanmei's failure to adequately differentiate self-produced AHF from purchased AHF in its records meant that Commerce could not accurately determine the amounts of by-products generated from the self-production of AHF.

So, because it was not possible, using record evidence, to determine if the amount of each of the two principal inputs, i.e., AHF and PCE, consumed in the production process resulted in (1) the production of subject R-125, (2) the generation of by-products, or (3) yield loss, it was not actually possible to calculate a by-product offset.

In addition, other factors that led Commerce to deny the offsets for fluosilicic acid and fluorine gypsum were (1) the lack of data regarding the consumption of water as a direct input; and (2) lack of clarity in the record as to how self-produced AHF and purchased AHF were differentiated in workshops that produced subject and non-subject merchandise. See Final IDM at 35, 29. These record deficiencies also reasonably support Commerce's decision to deny the claimed offsets because, as Sanmei acknowledged in its administrative case brief, “[w]ater is an important factor in the production of both AHF and R-125 and has a tremendous impact on yield loss and by-product production.” Sanmei's Case Br. at 12; see also Final IDM at 29-30. It seems generous for Commerce to allow all of the water reported as energy to be used as an input in making R-125; asking Commerce to make an assumption that an exact proportion of the water was used to make AHF seems to be overreach. Moreover, as the court noted in Section I, the lack of clarity in the record as to what extent self-produced AHF and purchased AHF were used in workshops where subject merchandise and non-subject merchandise were made provided reasonable grounds for Commerce to question whether only self-produced AHF was used to make the subject R-125. That being the case, Commerce could not be sure that the claimed offsets for fluosilicic acid and fluorine gypsum were, in fact, generated from the production of subject merchandise during the period of investigation. Accordingly, the court finds Commerce's denial of offsets for fluosilicic acid and fluorine gypsum reasonable.

B. R-125 By-Product Offsets: Hydrochloric Acid and R-134a

The court next turns to Commerce's denial of offsets for by-products generated during Qingliu's production of subject R-125, i.e., hydrochloric acid and R-134a.

1. Hydrochloric Acid

Unlike other by-products for which Sanmei claimed offsets, hydrochloric acid is generated during the production of both subject R-125 itself and non-subject merchandise. Pls.' Br. at 22 (“The hydrochloric acid by-product is the exception [among the other by-products fluosilicic acid, fluorine gypsum, and R-134a] because it is a by-product that is generated from both the production of R-125, the subject merchandise, and non-subject merchandise.”).

In the Final Determination, Commerce found that Sanmei's reporting failed to “adequately track production and sales of hydrochloric acid” that was generated during Qingliu's production of subject R-125:

Sanmei states that “the generation of hydrochloric acid from different workshops is recorded in the production reports and the inventory records,” but the information provided in the warehouse-in records and the screen shots of the warehouse management system do not appear to account for and demonstratively show a difference between production records in the R-125 and non-subject merchandise workshops, and there is no supporting documentation for distinguishing warehouse-out records for hydrochloric acid. Given that the “hydrochloric acid by-product is just mixed together without distinction of source,” [as reported by Sanmei] we find Sanmei's arguments fail to demonstrate that it can adequately track production and sales of hydrochloric acid.

Final IDM at 30. As with the AHF by-product fluorine gypsum, Plaintiffs claim that a formula that Sanmei provided to Commerce, adequately accounted for the amount of hydrochloric acid generated by the production of R-125. Commerce was unconvinced:

Sanmei's argument that the chemical formula is sufficient to demonstrate weight ratios is incongruous with the fact that water, a significant input in the production of R-125 that Qingliu does not measure, may comprise as much as 70 percent of hydrochloric acid. This fact introduces an element of variability, which Sanmei recognizes in its statement that “taking account of water and impurities contained in [the] by-product as well as any possible variance in reaction conditions and facility environment, a deviation to [a] certain extent from the theoretical ratio may also take place in the actual industrial production process.” Without adequate tracking of water at Qingliu, whose sales are the only reviewable sales in the [period of investigation], a theoretical calculation without adequate supporting documentation does not produce an accurate result in the production of hydrochloric acid.

Id. (emphasis added).

The court finds Commerce's denial of the claimed offset for hydrochloric acid reasonable based on this record. Commerce addressed the shortcomings it found in the record evidence: “information provided in the warehouse-in records and the screen shots of the warehouse management system do not appear to account for and demonstratively show a difference between production records in the R-125 and non-subject merchandise workshops, and there is no supporting documentation for distinguishing warehouse-out records for hydrochloric acid.” Final IDM at 30. Without clear production records distinguishing the amount of hydrochloric acid generated in the subject R-125 workshops from that generated in non-subject workshops, Commerce reasonably found it could not “substantiate the quantity of ․ hydrochloric acid ․ generated [by Qingliu] from the production of subject merchandise during the [period of investigation].” Id. at 28. In other words, for Commerce, Sanmei failed to adequately distinguish between hydrochloric acid generated in the production of subject R-125 and that generated in the production of non-subject merchandise. As a result, Sanmei could track neither the volume nor the value of self-generated hydrochloric acid and hydrochloric acid generated from the manufacture of non-subject merchandise because apparently the acid from each source was mixed together.

Additionally, missing information regarding Qingliu's consumption of water as a direct input in the production of subject R-125 further supports Commerce's decision to deny the offset for hydrochloric acid. Water “may comprise as much as 70 percent of hydrochloric acid.” Final IDM at 30 (observing that the fact that water may comprise as much as 70% of hydrochloric acid “introduces an element of variability ․ Without adequate tracking of water at Qingliu, whose sales are the only reviewable sales in the [period of investigation], a theoretical calculation without adequate supporting documentation does not produce an accurate result in the production of hydrochloric acid”). So, without water consumption data, it was not unreasonable for Commerce to find that it could not calculate accurately the quantity of the hydrochloric acid generated as a by-product of subject R-125. Thus, the court finds that the record supports Commerce's decision to deny the claimed offset for hydrochloric acid.

2. R-134a

R-134a is also a by-product generated in the production of R-125. With respect to R-134a, Commerce found that “Sanmei has not justified a by-product offset,” stating, by way of explanation, certain flaws in the record evidence:

First, the record remains unclear as to the processing that R-134a undergoes prior to introduction in non-subject merchandise or when it is sold. Second, Sanmei failed to provide adequate records for tracking production and sales of R-134a. We note that, while there are warehouse-out records, there are no supporting documents for warehouse-in records. Third, Sanmei notes that it purchases R-134a from Qingliu and sells R-134a to affiliated and unaffiliated parties. However, without proper warehouse-in documentation, it is unclear if Sanmei or Qingliu are able to distinguish R-134a resulting from the production of R-125 or purchases from affiliated or unaffiliated parties.

Final IDM at 30-31. Thus, for these reasons, Commerce denied the claimed offset for R-134a.

Plaintiffs do not make any specific argument with respect to Commerce's denial of the offset for R-134a, except to repeat that, as with the other claimed by-product offsets, Sanmei's reporting was sufficient to grant the offset. Pls.' Br. at 19-20. In other words, Plaintiffs do not demonstrate that Commerce's cited reasons for denying the offset lack the support of record evidence, but rather they claim that other evidence on the record supports granting it. The evidence relied upon by Plaintiffs, however, is lacking because, as Commerce stated: “while there are warehouse-out records, there are no supporting documents for warehouse-in records.” Final IDM at 30. “[W]ithout proper warehouse-in documentation, it is unclear if Sanmei or Qingliu are able to distinguish R-134a resulting from the production of R-125 or purchases from affiliated or unaffiliated parties.” Id. at 31. Thus, Plaintiffs have not convinced the court that Commerce's finding that Sanmei did not satisfy its “burden to demonstrate [the respondent's] eligibility for a requested by-product offset” was unreasonable. Id.

III. Commerce's Calculation of the Surrogate Freight Rate Is Supported by Substantial Evidence and Otherwise in Accordance with Law

Finally, the court turns to Commerce's calculation of a surrogate freight rate. In the Final Determination, Commerce relied on data from the Doing Business report for Russia to calculate the surrogate freight rate. See Final IDM at 43. The report was included as an exhibit to the petition and in Sanmei's surrogate value submission. See Pet. for the Imposition of Antidumping and Countervailing Duties on Behalf of Honeywell International Inc., Crowell & Moring LLP, Vol. II Ex. II-9a (Jan. 11, 2021), PR 9; see also Sanmei's Initial Surrogate Value Submission Ex. 3 (June 14, 2021), PR 170.

Commerce stated how it determined the surrogate freight rate using a simple average:

The Doing Business Russia report provides costs to both import and export a standardized cargo of 15 MT in a 20-foot container in Russia by truck for two locations, Moscow and St. Petersburg. Specifically, Doing Business Russia provides two distances for export: (1) from [the city of] St. Petersburg to St. Petersburg Port of 8 km; and (2) from Moscow to the port in St. Petersburg of 724 km. Doing Business Russia also provides two distances for import: (1) from [the city of] St. Petersburg to St. Petersburg Port of 8 km; and (2) from Moscow to the Krasnaya Gorka border crossing in Smolenskaya Oblast of 500 km. Using these data, we calculated a USD/MT/KM [U.S. dollars per metric ton per kilometer] rate for each transaction and then used the simple average of those two rates to calculate an average USD/MT/KM inland freight rate. Then, because the data used are not contemporaneous with the [period of investigation], we adjusted the simple-average rate using the Russian producer price index (PPI) to find the average USD/MT/KM inland freight [surrogate value]s for import and export. The resultant truck freight rates were $1.039/MT/KM for export and $1.038/MT/KM for import.

Final IDM at 43.¹⁵ All cost data related to domestic transport included the cost of loading and unloading. See Sanmei's Initial Surrogate Value Submission Ex. 3 at 82 (according to the Doing Business report, the indicators in the report measure “[l]oading or unloading of the shipment at the warehouse or port/border”).

For its part, Sanmei reported that the distance between its factory and the nearest port was 187.2 kilometers. See Sanmei's Sec. D Resp. at 19. Plaintiffs contest (1) Commerce's inclusion of the St. Petersburg export and import freight data in its inland freight calculation, and (2) Commerce's use of a simple average of the Moscow and St. Petersburg freight data.

For the following reasons, the court finds that substantial evidence supports Commerce's calculation of the surrogate freight rate. It should be noted that Sanmei would benefit from the exclusion of the St. Petersburg data because it would lower Sanmei's calculated antidumping duty margin. Pls.' Br. at 8 (“Inclusion of the short-distance intra-city freight data not only vastly overstated [Sanmei's] calculated antidumping duty margin, but failed to reflect [Sanmei's] actual freight experience during the period of investigation.”).

A. Commerce Reasonably Included St. Petersburg Data in Its Freight Calculation

In the Final Determination, Commerce found that the Doing Business report, and in particular, the Moscow and St. Petersburg data, were the best available information:

The broad market average provided by both the short- [St. Petersburg] and long-haul [Moscow] freight distances in Doing Business Russia continues to be the “best available information” on the record of this investigation. Therefore, we continue to rely upon all the data for truck freight found in Doing Business Russia to calculate the import/export truck freight [surrogate value]s for purposes of this final determination.

Final IDM at 45. In other words, Commerce found that both the long-distance Moscow freight data and the short-distance St. Petersburg freight data were the best available information because they represented broad market averages.

Plaintiffs contest Commerce's inclusion of the St. Petersburg (short haul) data in the surrogate freight rate calculation. As noted, Sanmei reported that the distance between its factory and the nearest port was 187.2 kilometers, which it claims is a “long-distance.” For Plaintiffs, “[i]nclusion of the short-distance intra-city St. Petersburg import and export data points in the simple-averaged freight calculations severely distorted the surrogate USD/MT/KM freight rate.” Pls.' Br. at 8. Plaintiffs insist that the St. Petersburg data is aberrational and does not reflect Sanmei's experience:

Due to the great disparity between 8 KM distance travelled in the in the [sic] intra city datapoint [i.e., St. Petersburg] compared to the 724 KM or 500 KM distance travelled in the long-distance data points [i.e., Moscow], the calculated St. Petersburg intra-city freight rate is aberrational when applied to a to [sic] [Sanmei]'s long-distance freight factor. The intra-city freight rate was more than 23 times greater than either of the two calculated Moscow long-distance freight rates. Given the great disparity between the intra-city and long-distance freight rates evidence in the World Bank data, for purposes of the final determination Commerce should calculate the surrogate freight rate using only the long-distance data points that reflect the manner in which [Sanmei]'s [sic] incurred its trucking freight costs. Inclusion of the short-distance intra-city freight data not only vastly overstated [Sanmei]'s calculated antidumping duty margin, but failed to reflect [Sanmei]'s actual freight experience during the period of investigation.

Id. As a result, although Sanmei placed on the record the Doing Business report (and no party contests that it is the best available information), Plaintiffs argue that the use of the St. Petersburg data from the report is unsupported by substantial evidence because the actual distance Sanmei's product traveled between its factory and the nearest port is much longer than the 8-kilometer figure used as a part of the average when determining the freight rate to and from the city of St. Petersburg.

The court finds that Commerce reasonably included, in its surrogate, the value calculations of truck freight data for the import and export of merchandise to and from Moscow (long-haul) and St. Peterburg (short-haul) to approximate a broad market average of freight prices. See Qingdao Sea-line Trading Co., 766 F.3d at 1386; see also Jacobi Carbons AB v. United States, 38 CIT 932, 940, 992 F. Supp. 2d 1360, 1368 (2014) (“When making a ‘best available information’ finding, this Court, among other things, has repeatedly confirmed the importance that the information used to value the factors of production (1) represents a broad market average of prices for the input in question, and (2) be exclusive of taxes and duties.”).

Here, for freight costs related to the importation of merchandise, Commerce calculated (1) a short-haul freight rate of $2.050 per kilogram per kilometer, and (2) a long-haul truck freight rate of $0.087 per kilogram per kilometer. See Preliminary Surrogate Value Mem. Ex. 8. The simple average of these two rates was $1.068 per metric ton per kilometer (USD/MT/KM). Id. After applying an inflator, the surrogate inland freight rate for import was 1.038 USD USD/MT/KM. Id. For freight costs related to the exportation of merchandise, Commerce calculated (1) a short-haul truck freight rate of $2.050 per kilogram per kilometer, and (2) a long-haul truck freight rate of $0.088 per kilogram per kilometer. Id. The simple average of these two rates was $1.069 per metric ton per kilometer (USD/MT/KM). Id. After applying an inflator, the surrogate inland freight rate for import was 1.039 USD USD/MT/KM. Id.

The short-haul freight rate ($2.050 per kilogram per kilometer), that Plaintiffs would have Commerce exclude from its calculation, is higher than the long-haul freight rate ($0.087 per kilogram per kilometer, and $0.088 per kilogram per kilometer) because each trip includes loading and unloading that increases the cost per kilometer. Under the facts here, the inclusion of the short-haul freight rates was not unreasonable. Sanmei's reported distance between its factory and the nearest port (187.2 kilometers) exceeds the distance between the city of St. Peterburg and the port (8 kilometers) and is shorter than the distance between Moscow and the border (500- and 724-kilometers) that Sanmei argues Commerce should rely on to make its freight calculation. Commerce reasonably found that using both the Moscow and St. Petersburg data more closely represents a broad market average than only using the Moscow freight data. Based on the available record information, substantial evidence supports Commerce's inclusion of the St. Petersburg freight data in its calculation of the inland freight surrogate values.

Plaintiffs' arguments to the contrary are unconvincing. For Plaintiffs, the “short-distance intra-city” St. Petersburg freight data is aberrational because (1) the distances for St. Petersburg are much smaller than those for Moscow, and (2) the St. Petersburg data does not reflect the “long-distance” freight expenses Sanmei incurred during the period of investigation—only the “long-distance” Moscow data does. The problem with Plaintiffs' argument is that its proposal of using only the long-haul rate would undoubtedly end up being aberrational.

In the Final Determination, Commerce stated why the “short-distance” St. Petersburg rate is higher than the “long-distance” Moscow rate: “[I]t is the fixed costs related to loading, unloading, and traveling within the urban environment that increase the cost per kilometer of short-haul trucking.” Final IDM at 45 (quotation marks and citation omitted). Thus, for Commerce, neither the short distance nor long distance values are necessarily aberrational. Id. Rather, Commerce found, “they are representative of the separate experiences they reflect.” Id. (quotation marks and citation omitted). In other words, the freight rate for long distances is lower than for short distances on a per kilometer basis because on a long-haul trip the expense of loading and unloading is spread over a longer distance than is the case for short distances. Plaintiffs have failed to demonstrate that this finding lacks record support or that their argument that the long-haul rate alone would be reasonable when Sanmei's actual distances were far shorter than the Doing Business long-haul rates.

Plaintiffs have failed to show that using the St. Petersburg freight data resulted in an inaccurate surrogate value for Sanmei's freight expenses. Plaintiffs assume that Sanmei incurred “long-distance” expenses for shipping freight and thus that St. Petersburg's “short-distance intra-city” freight was unrepresentative. See Pls.' Br. at 8. Commerce considered, and rejected, this argument in the Final IDM, finding that “[t]he record evidence does not demonstrate that the inclusion of a short freight distance is unrepresentative.” Final IDM at 45. As support for its finding, Commerce noted: “The record shows that the distance between Sanmei and Wenzhou port (i.e., the nearest seaport to Sanmei) is 187.2 km.” Id. at 45 n.296. This distance is closer to the distance of the St. Petersburg freight (8 kilometers) than it is to that of the Moscow freight (500 kilometers for import or 724 kilometers for export). Excluding the St. Petersburg data, then, would tend to produce a less accurate result than including data from both cities (and using the simple average of this data). Commerce reasonably concluded that using data from both cities results in a freight rate that most accurately represents Sanmei's freight factor of production. Yangzhou Bestpak Gifts & Crafts Co. v. United States, 716 F.3d 1370, 1379 (Fed. Cir. 2013) (“An overriding purpose of Commerce's administration of antidumping laws is to calculate dumping margins as accurately as possible.” (citing Rhone Poulenc, Inc. v. United States, 899 F.2d 1185, 1191 (Fed. Cir. 1990))).

Commerce reasonably concluded that the shorter St. Petersburg data is not aberrational or unrepresentative of Sanmei's freight expenses when averaged with the Moscow data. Commerce's inclusion of the St. Petersburg data in its surrogate freight calculation for Sanmei is supported by substantial evidence. Had Plaintiffs proposed something other than the use of the long-haul rate alone, the court might have reached a different conclusion, but given the record evidence, Commerce's simple average is reasonable.

B. Substantial Evidence Supports Commerce's Use of a Simple Average to Calculate Freight Rate

Plaintiffs also challenge Commerce's use of a simple average to calculate a surrogate freight rate for Sanmei. See Pls.' Br. at 12.

Instead of a simple average, Plaintiffs propose a method that involves making two “weighted-average” calculations to determine the surrogate freight rate: one using the export freight data for Moscow and St. Petersburg, and the other using import freight data for Moscow and St. Petersburg. For each calculation, Plaintiffs insist, Commerce should have averaged each individual component of the Moscow and St. Petersburg data (i.e., cost, weight, and distance) and divided the average cost by the average weight. Then, Commerce should have divided the result by the average distance.¹⁶ In other words, Plaintiffs propose averaging the components of the freight rates (cost, weight, and distance), instead of averaging the St. Petersburg and Moscow export and import U.S. dollars per metric ton per kilometer freight rates to reach an average export freight rate and an average import freight rate.

Plaintiffs claim that a simple average is an “incorrect mathematical formula” because it “is not equal to the ‘average USD per MT per KM['] freight rates for import and export transactions.” Pls.’ Br. at 11-12. In other words, for Plaintiffs, a simple average does not result in an average U.S. dollars per metric ton per kilometer freight rate. Plaintiffs insist that “[o]nly [their proposed] weight averaging methodology, will result in the intended ‘average USD per MT per KM’ freight rate (i.e., a weighted average of the three variables that are incorporated into the calculation of the freight factor).” Id. at 12.

In the Final Determination, Commerce found that Plaintiffs had failed to demonstrate that their proposed method would increase the accuracy of the surrogate freight rates (and, thus, the dumping margin):

Sanmei fails to explain why the simple-average methodology is logically or mathematically incorrect, other than to point out that the freight rates and dumping margins are higher than if we were to use Sanmei's preferred methodology. We do not find this argument convincing․ [T]he fact that [a surrogate value] is larger or smaller than others on the record does not render it distorted or aberrational. By the same logic, a methodology is not mathematically incorrect simply because it produces a larger or smaller result. Instead, it must be demonstrated to be incorrect on logical or mathematical principles, which Sanmei has not done. In addition, just because the end result using Commerce's preferred simple-average methodology leads to a higher margin for Sanmei, and a higher portion of the margin attributable to truck freight, does not render Commerce's calculation improper. [Normal value] and U.S. price are necessarily a function of their parts. In an NME context, when certain [surrogate value]s (such as truck freight) are high and are used on both sides of the dumping calculations (i.e., truck freight is frequently incorporated in the buildup of [normal value] when calculating the per-unit value of inputs and deducted from U.S. price), the effect on the margin may be significant. However, this does not mean that the [surrogate value]s are distorted, nor does it mean [the] resultant dumping margin is overstated. It only indicates that Sanmei's dumping of subject merchandise is more attributable to the costs incurred for transportation of materials and finished merchandise than to the direct material costs incurred to produce the merchandise.

Final IDM at 46-47. Thus, Commerce found that the rate that resulted from using a simple average was not aberrational just because it was less advantageous to Sanmei.

In addition, Commerce points out that the addition of weight as a factor when calculating the freight rate may have the effect of understating the cost of loading and unloading. As Commerce stated, each freight shipment carried certain fixed costs for loading, unloading, and traveling. See id. at 46. Thus, by including cost, weight, and distance all together for the Moscow shipments, and then again for the St. Petersburg shipments (i.e., calculating surrogate freight rates in USD/MT/KM for both Moscow and St. Petersburg), Commerce retained the fixed costs of loading, unloading, and traveling that are tied to each freight shipment. For Commerce, it would be unreasonable to separately average the cost, then the weight, and then the distance of the shipments for Moscow and St. Petersburg, as Plaintiffs propose, because these fixed costs would be discounted in each shipment, and therefore would not be accurately accounted for. Id. (“While the calculation of the truck freight [surrogate value]s is not capable of taking into account every variable affecting truck freight based on the record information, when averaging the component parts fixed costs (i.e., related to loading, unloading, and traveling) are disproportionally spread out over longer average distances rather than being taken into account in their respective transactions, such as with the simple-average methodology.”).

Because Plaintiffs neither explain why Commerce's use of a simple average was unreasonable, nor show how using their proposed method would lead to a more accurate calculation than Commerce's chosen method (particularly when the undervaluation of loading and unloading costs is taken into account), the court is not persuaded that Commerce has erred in its calculations. Plaintiffs' other arguments are similarly unpersuasive.¹⁷

CONCLUSION

Based on the foregoing, the court denies Plaintiffs' motion for judgment on the agency record and sustains the Final Determination. Judgment will be entered accordingly.

New York, New York

FOOTNOTES

1.   See Pet. for the Imposition of Antidumping and Countervailing Duties on Behalf of Honeywell International Inc., Crowell & Moring LLP, Vol. I at 6 (Jan. 11, 2021), PR 4.

2.   After the filing of the motion for judgment on the agency record, Shandong Dongyue Chemical Co., Ltd. and Huantai Dongyue International Trade Co., Ltd. were voluntarily dismissed from this action, leaving Sanmei as the sole plaintiff. See Order (Sept. 3, 2024), ECF No. 51. For purposes of consistency with the motion and other filings made prior to dismissal, the court will retain the reference to “Plaintiffs,” which shall mean Sanmei.

3.   Commerce initially selected two mandatory respondents, but Zhejiang Quzhou Juxin Fluorine Chemical Co., Ltd. withdrew from the investigation, leaving only Sanmei. The statute permits Commerce to limit its examination to a “reasonable number of exporters or producers” when there is a “large number of exporters or producers involved in the investigation or review.” 19 U.S.C. § 1677f-1(c)(2). Though the number of mandatory respondents has not been challenged in this case, the court observes that under the statute a “ ‘reasonable number’ is generally more than one.” YC Rubber Co. (N. Am.) LLC v. United States, No. 21-1489, 2022 WL 3711377, at *4 (Fed. Cir. Aug. 29, 2022).

4.   “ ‘Toll manufacturing,’ also called ‘toll processing,’ is ‘[a]n arrangement under which a customer provides the materials for a manufacturing process and receives the finished goods from the manufacturer ․ The same party owns both the input and the output of the manufacturing process. This is a specialized form of contract manufacturing.’ ” Wind Tower Trade Coal. v. United States, 46 CIT __, __, 569 F. Supp. 3d 1221, 1226 n.3 (2022) (quoting Toll Manufacturing, BLACK'S LAW DICTIONARY (11th ed. 2019)).

5.   The “all-others” rate is the rate assigned to all exporters and producers of the subject merchandise in an investigation who were granted separate rate status, but which Commerce did not select for individual investigation. See Shanxi Hairui Trade Co. v. United States, 39 F.4th 1357, 1360 (Fed. Cir. 2022).

6.   The selection of Russia as the surrogate country is not in dispute. See Preliminary Decision Mem. (Aug. 10, 2021) at 7, PR 220 (“Sanmei and the petitioner both agree that the Russian Federation (Russia) is suitable to serve as the primary surrogate country.”).

7.   On-site verification by Commerce officials was suspended during the COVID-19 global pandemic. See Preliminary Determination, 86 Fed. Reg. at 45,961 (“Commerce is currently unable to conduct on-site verification of the information relied upon in making its final determination in this investigation. Accordingly, we intend to take additional steps in lieu of on-site verification.”). In lieu of on-site verification, Commerce issued a verification questionnaire. See Sanmei Verification Questionnaire (Sept. 9, 2021), PR 247.

8.   It is worth noting that, from as far back as 2006, this Court has observed that, in the absence of any statutory law on the treatment of by-product offsets, “Commerce has not filled the statutory gap with a regulation.” Arch Chems., Inc. v. United States, 33 CIT 954, 956 (2009) (not reported in the Federal Supplement) (citing Guangdong Chems. Imp. & Exp. Corp. v. United States, 30 CIT 1412, 1422, 460 F. Supp. 2d 1365, 1373 (2006)). This remains the case today.

9.   The antidumping statute provides for the adjustment of both export price and normal value. See, e.g., 19 U.S.C. § 1677a(c); id. § 1677b(a)(6)-(7); see also 19 C.F.R. § 351.401(b)(1)-(2) (2021) (“In making adjustments to export price, constructed export price, or normal value, the Secretary will adhere to the following principles: (1) The interested party that is in possession of the relevant information has the burden of establishing to the satisfaction of the Secretary the amount and nature of a particular adjustment; and (2) The Secretary will not double-count adjustments.”).

10.   In the Preliminary Surrogate Value Memorandum, Commerce stated:Although Sanmei reported water as an energy input, we are treating it as a direct material because it appears that it is actually incorporated into the by-products produced during the manufacture of R-125. To the extent that Sanmei may have also used water for energy purposes, because it is impossible to break out the water used as a direct material from the reported quantities, we are treating the entirety of Sanmei's water input as a direct material input.Preliminary Surrogate Value Mem. at 6 (emphasis added).

11.   Plaintiffs have proposed the following calculation, or formula, to derive yield loss for AHF in the absence of actual yield loss information maintained in the ordinary course of business:The administrative record confirms that the total inputs for AHF were equal to 250,046 MT, including 92,273.976 MT of fluorite powder, 73,111.711 MT of 98% sulfuric acid, 40,049.549 MT of 105% sulfuric acid, 16,492.35 MT of water at ․ Sanmei, and 28,118 MT of calculated water at ․ Qingliu. The total reported output was equal to 217,730 MT, including 44,089.461 MT of AHF, 6,272.37 MT of fluosilicic acid by product and 167,367.8 MT of fluorine gypsum by-product. The total yield loss [for AHF] was thus 250,046 – 217,730 = 32,316, or about 13%.Pls.' Br. at 18.

12.   It is worth noting that, contrary to Commerce's Section D questionnaire instructions, Sanmei did not contact Commerce regarding its decision to calculate a water consumption amount for Qingliu. See Final IDM at 23 (“In an attempt to rectify the record, Sanmei reported its own water usage as a surrogate value for Qingliu's water consumption without contacting Commerce per the initial questionnaire.”); see also Initial Quest. at D-1 (“If you have any questions regarding how to compute the factors of the merchandise under consideration, please contact the official in charge before preparing your response to this section of the questionnaire.”).

13.   While yield loss includes the quantity of a raw material lost during production, it also includes the weight of a raw material lost during production. See An Giang Fisheries Imp. & Ex. Joint Stock Co. v. United States, 41 CIT __, __, 203 F. Supp. 3d 1256, 1277 (2017) (stating that, where fingerlings (small fish) were a factor of production for frozen fish fillets, the yield loss was the number of fingerlings that died during a certain period); Jining Yongjia Trade Co., 34 CIT at 1515, 1519 (indicating that, in the production of fresh garlic, yield loss was the amount the garlic shrunk during production, occurring when garlic lost water weight).

14.   In their brief, Plaintiffs state:Further, [Sanmei] introduced two chemical reaction formulae to explain why certain by-products out-weighted the main product. During the production process, the actual weight ratio of by-product to the main product for fluorine gypsum was 3.7931, a deviation of 0.3931 or 11% from the theoretical ratio and the actual weight ratio of by-product to the main product for hydrochloric acid was 4.0667, a deviation of 0.0067 or 0.2%. The fact that two of the by-products had weights greater than the main product does not detract from the accuracy of the provided by-product offset information.Pls.' Br. at 20.

15.   Commerce did not change its method for valuing inland freight charges between the Preliminary Determination and the Final Determination. As stated in the surrogate value memorandum:We valued foreign inland freight charges using the World Bank's Doing Business 2020: Russia. This report gathers information concerning the distances and costs to transport products in a container for export to the border crossing at the St. Petersburg Port and import from the border crossings at Krasnaya Gorka, Smolenskaya Oblast, and St. Petersburg Port. We calculated a per-MT, per-km inland freight rate for export of 1.069 USD, using the 15,000 kg weight provided in Doing Business 2020: Russia. We calculated a per-MT, per-km inland freight rate for import of 1.068 USD, using the 15,000 kg weight provided in Doing Business 2020: Russia. Because the data being used pre-dates the [period of investigation] (i.e., it is recent as of May 2019), we adjusted the values using the Russian [producer price index], which resulted in a surrogate inland freight rate for export of 1.039 USD per-MT per-km and a surrogate inland freight rate for import of 1.038 USD per-MT per-km.Preliminary Surrogate Value Mem. at 10.

16.   In Plaintiffs' words:To properly calculate an average [U.S. dollars per metric ton per kilometer (USD/MT/KM)] freight rate from two transactions that incorporate three distinct variables, it is necessary to first calculate the average cost of the two transactions, the average weight of the two transactions, and the average distance of the two transactions. Then the averaged dollar cost of the two transactions should be divided by the averaged weight of the two transactions and then divided by the averaged distance of the two transactions.Pls.' Br. at 12. The “two transactions” that Plaintiffs refer to are the Moscow and St. Petersburg freight rates, used for calculating the average export freight rate and the average import freight rate.

17.   Plaintiffs additionally argue that Commerce acted arbitrarily by using in its initiation notice the weighted average method set forth in the petition, but then using a simple average for its Final Determination. See Pls.' Br. at 13 (arguing that “[i]t was arbitrary for Commerce to use different methodologies to calculate the same surrogate freight rates at different points in the same investigation.”). When Commerce issues an initiation notice, however, it is simply announcing its decision to initiate an investigation. See 19 C.F.R. § 351.203(c)(1). Commerce issues this notice when it determines that “the petition alleges the elements necessary for the imposition of a duty under [19 U.S.C. § 1673] and contains information reasonably available to the petitioner supporting the allegations.” 19 U.S.C. § 1673a(c)(1)(A)(i). To make this determination, Commerce examines the “accuracy and adequacy of the evidence provided in the petition,” based on “sources readily available to [Commerce].” Id.; see also 19 C.F.R. § 351.202(b)(7)(i)(B) (requiring a petition requesting the imposition of antidumping duties to contain “[a]ll factual information (particularly documentary evidence) relevant to the calculation of the export price and the constructed export price of the subject merchandise and the normal value of the foreign like product ․”). Thus, Plaintiffs' argument cannot be credited because all Commerce did in its initiation notice was determine, based on the information available to it in the petition, that a formal investigation was warranted. 19 U.S.C. § 1673a(a)(1). The Final Determination, on the other hand, was based on the record developed after the investigation had been initiated.

Richard K. Eaton Judge