Lee Ann Sommerville v. Union Carbide Corporation, et al.

IN THE UNITED STATES DISTRICT COURT FOR THE SOUTHERN DISTRICT OF WEST VIRGINIA CHARLESTON DIVISION LEE ANN SOMMERVILLE, Plaintiff, v. CIVIL ACTION NO. 2:19-cv-00878 UNION CARBIDE CORPORATION, et al., Defendants. MEMORANDUM OPINION AND ORDER Pending is Plaintiff Lee Ann Sommerville’s Motion to Certify Class. [ECF No. 313]. Defendants Covestro LLC (“Covestro”) and Union Carbide Corporation (“Union Carbide”) each responded in opposition, [ECF Nos. 358; 359], to which Plaintiff replied, [ECF Nos. 375; 376]. In light of the intervening appellate proceedings and the reopening of this matter on remand, the court permitted supplemental briefing to be filed regarding Plaintiff’s Motion to Certify Class. On July 17, 2026, Plaintiff filed a Supplemental Memorandum of Law. [ECF No. 432]. On August 20, 2026, Defendants each responded in opposition, [ECF Nos. 471; 470], to which Plaintiff replied on September 4, 2026, [ECF Nos. 502; 503]. Also pending is Plaintiff’s Motion to Redefine the Class. [ECF No. 493]. Defendants each responded in opposition, [ECF Nos. 510; 509], to which Plaintiff replied, [ECF No. 511]. Defendants own and operate a manufacturing facility in South Charleston, West Virginia (the “South Charleston Plant”).1 Plaintiff, on behalf of herself and others similarly situated, brings this action against Defendants for alleged exposure to ethylene oxide (“EtO”), an odorless, colorless, carcinogenic gas. [ECF No. 85, ¶ 33 (“Second Amended Complaint”)]. Plaintiff and the proposed class of approximately 37,000 people2 reside in neighborhoods surrounding the South Charleston Plant. Plaintiff seeks a medical monitoring fund for the alleged exposure as a result of the Defendants’ alleged EtO emissions, which Plaintiff alleges has significantly increased hers and the putative class’s risk of developing cancer. Id. ¶ 1; [ECF No. 314, at 2].3 For the reasons 1 The Second Amended Class Action Complaint, which is the operative Complaint in this case, refers solely and exclusively to the South Charleston Plant. Inexplicably, Plaintiff Sommerville’s briefing in support of the Motion to Certify seeks certification of a class affected by emissions from both the South Charleston Plant and Institute Plant. Compare e.g., [ECF No. 85, ¶ 1 (“Plaintiff brings this class action against Defendants, owners and operators of a manufacturing facility in South Charleston, West Virginia (‘the South Charleston Plant’), for damages resulting from Defendants’ dangerous and reckless emission of ethylene oxide (‘EtO’).”), ¶ 4 (“Plaintiff and Class Members have lived in the vicinity of Defendants’ South Charleston Plant . . . ”), ¶ 7 (referring to “Defendants’ South Charleston Plant”); ¶ 8 (“As a result of Defendants’ operations at Defendants’ South Charleston Plant . . . ”), ¶¶ 9, 12–13, 15–16, 18–19, 21, 23, 28–29, 34, 40–46, 49–52, 54, 57–58, 75, 87–88, 94, 98, 101, 105]; with [ECF No. 314, at 2 (“On behalf of herself and all others similarly situated who were harmed by . . . Defendants’[] dangerous and negligent emission of . . . ethylene oxide (‘EtO’), over the course of over four decades in and around their manufacturing facilities in Institute, West Virginia (the ‘Institute Plant’) and South Charleston, West Virginia (the ‘South Charleston Plant’), Plaintiff seeks a medical monitoring fund.”)]. The expert reports also combine calculations based upon emissions from both the South Charleston and Institute plants. E.g., [ECF 313-1, at 11 (Dr. Sahu explaining his methodology and conclusions to determine cumulative exposure are based upon “activities relevant to air pollution controls and waste management practices employed by Defendants at the Institute and South Charleston facilities”); ECF No. 313-3 (Dr. Martin T. Wells’s Report, which calculates relative risk based upon Dr. Sahu’s cumulative exposure figures); id., at 6 (“According to Dr. Sahu’s analysis, . . . [Plaintiff Sommerville’s] cumulative exposure to defendants’ ethylene oxide emissions in the years 2001–2003, 2005–2019 . . . equate[s] to 340.558 in-a-million increase in relative risk.”); ECF No. 313-4 (Dr. Charles Cowan’s report, which calculates the number of class members “exposed to high levels of EtO gas from Defendants’ manufacturing operations”); ECF No. 315-1 (Dr. Cowan’s rebuttal report, which represents “[c]ases of Hodgkin Lymphoma, Non-Hodgkin’s Lymphoma and Leukemia in Kanawha County, WV are . . . highly correlated with emissions of ethylene oxide from the Institute and South Charleston facilities . . . ”)]. Because the allegations concerning the Institute Plant appear for the first time in the class-certification briefing and are not included in the operative Complaint, they present a pleading issue addressed later in this Memorandum Opinion and Order. 2 See [ECF No. 314, at 8 (citing ECF No. 313-4, ¶ 54) (acknowledging Plaintiffs’ expert, “Dr. [Charles] Cowan[,] estimates that there are approximately 37,000 people in the Class”)]. Notably, Plaintiff cites Dr. Cowan’s rebuttal report, which is located at ECF No. 313-4, but the rebuttal report does not contain the information Plaintiff cites. Dr. Cowan also provided Plaintiffs with an expert report dated February 17, 2023, which Plaintiffs did not include with their certification briefing, but which Defendant Covestro included in its Motion to Exclude Dr. Charles Cowan’s Original Report, [ECF Nos. 315; 315-1]. Regardless, I refer to the proposed class as totaling approximately 37,000. 3 All page number references reflect CM/ECF page numbers. discussed below, I DENY the Motion to Redefine the Class, [ECF No. 493], and DENY the Motion to Certify Class. [ECF No. 313]. Plaintiff Sommerville says Defendants have released EtO into the air for over 40 years, that she and proposed class members breathed it, and that they now face a higher chance of getting cancer than they otherwise would. Plaintiff is not currently sick, but she wants Defendants to pay for regular medical tests—for her and putative class members—so that if cancer comes, it will be caught early. West Virginia allows medical monitoring claims. Bower v. Westinghouse Elec. Corp., 206 W. Va. 133, 522 S.E.2d 424 (1999). And the United States Court of Appeals for the Fourth Circuit held Plaintiff has standing to bring it. Sommerville v. Union Carbide Corp., 149 F.4th 408, 421 (4th Cir. 2025), cert. denied, 146 S. Ct. 2677, 224 L. Ed. 2d 857 (2026). The court of appeals also held Plaintiff’s air-modeling expert, Dr. Ranajit Sahu, may testify. Id. at 421, 426–27. I do not now revisit standing or the admissibility of Dr. Sahu’s testimony. Now, the question is, assuming the admissibility of the testimony of the parties’ experts,4 whether any of Plaintiffs’ proposed class definition is certifiable such that the proposed class of 37,000 people can try their cases together. A plaintiff seeking medical monitoring must show the defendant’s tortious conduct significantly exposed her—relative to the general population—to a hazardous substance, that the exposure proximately caused an increased risk of serious disease, that the raised risk makes the 4 “Merits questions may be considered to the extent—but only to the extent—that they are relevant to determining whether the Rule 23 prerequisites for class certification are satisfied.” Amgen Inc. v. Connecticut Retirement Plans & Trust Funds, 568 U.S. 455, 466, 133 S. Ct. 1184 (2013) (citing Wal-Mart Stores, Inc. v. Dukes, 564 U.S. 338, 376 n.6, 131 S. Ct. 2541, 2566 n.6 (2011)); Spurlock v. Wexford Health Sources, Inc., 175 F.4th 232, 252 (4th Cir. 2026) (“While class certification analysis may entail some overlap with the merits of the plaintiff's underlying claim, . . . it is not appropriate to treat class certification as a merits-determination or to delve into the merits further than necessary to satisfy the requirements of Rule 23.”) (citation modified). The court does not now rule on the merits of the Daubert motions, but assumes the admissibility under Daubert of the expert testimony for the purposes of class certification. testing medically necessary, and that monitoring procedures exist. Id. at 416 (citing Bower, 522 S.E.2d at 430–31). Plaintiff seeks medical monitoring not just for herself—but for roughly 37,000 other people too. A plaintiff seeking class certification must show, among other things, (1) the class members can be determined by considering objective criteria without being administratively infeasible for the court, (2) the plaintiff’s claim is typical of the class as a whole, (3) enough similarities exist among the proposed class such that the court can arrive at common answers able to resolve the litigation, and (4) whether common questions among the class as a whole predominate over individual questions such that it makes sense to warrant adjudication by representation. Under these standards, granting the motion for class certification would require finding both that Plaintiffs have satisfied the class certification requirements and that their evidence makes a prima facie showing of each Bower element outlined above. Critically, in order to certify, I must determine that each Bower element is susceptible to generalized, class-wide proof. Plaintiff Sommerville has not carried her burden for class certification. Six independent reasons preclude me from granting the motion. First, the proposed class definitions within the Motion to Certify, [ECF Nos. 313; 314], and the Motion to Redefine the Class, [ECF Nos. 493; 494], rely unpled evidence regarding the Institute Plant. Absent a formal amendment under Federal Rule of Civil Procedure 15 expanding the scope of the claims and the allegations upon which they rest in the operative Second Amended Complaint, this case remains based solely on emissions from the South Charleston facility. Second, the proposed class is a fail-safe class because membership depends on whether a putative class member has a valid claim on the merits. Third, ascertainability is lacking because determining class members would be administratively infeasible since the court will be unable to determine class members until after a legally cognizable injury has been established. Fourth, Plaintiff Sommerville’s claims are not typical of the proposed class because it includes individuals who lived in the Class Zone before 2000 and minor class members. Fifth, predominance is lacking because individualized inquiries regarding causation, fault- allocation, resultant damages, “significant exposure,” and “increased risk” predominate over individual questions. Sixth, and finally, superiority is lacking because Plaintiff Sommerville has not shown a class action is, in fact, the best and most efficient manner of adjudication—particularly in light of the aforementioned defects. Most critically, the proposed class definition is overbroad and would be too difficult to manage. I. BACKGROUND This case concerns Plaintiff and a putative class of 37,000 people in communities surrounding the South Charleston Plant. Union Carbide has owned and operated the South Charleston Plant since at least 1978. [Second Amended Complaint, ¶¶ 41, 44]. In 2000,5 Covestro began “operat[ing] polyols facilities” at the South Charleston Plant. Id. ¶¶ 41, 43–44. Plaintiff alleges Defendants’ operations at the South Charleston Plant released toxic EtO into the surrounding environment, that proposed class members inhaled these emissions, and the resulting 5 Covestro’s predecessors in interest—Bayer Corporation, Bayer Polymers LLC, and Bayer MaterialScience LLC (“Bayer”)—purchased the polyols operations at the South Charleston Plant from the Lyondell Chemical Company and other entities, with the transfer of ownership effective April 1, 2000. [ECF No. 358-1, ¶¶ 4–5 n.1]. Covestro spun off from Bayer in 2015. [Second Amended Complaint, ¶ 43–44]. While Covestro itself did not exist as a separate entity until 2015, and did not own the South Charleston Plant until 2015, the polyols operations that Covestro ultimately acquired have been operated at the South Charleston Plant since April 1, 2000. [ECF Nos. 358-8, at 11; 359-4 at 11 (“Machado Expert Report”)]. exposure has significantly increased their risk of developing cancer. Id. ¶¶ 48, 59, 62. Based on this alleged increased risk, Plaintiff asserts that she and the proposed class have a present need for medical monitoring and seeks to require Defendants to bear the costs of a medical monitoring program. Id. ¶¶ 62, 86. A. Union Carbide and Covestro’s South Charleston Plant Operations The South Charleston Plant is a “sprawling, complex chemical manufacturing operation” in West Virginia’s Kanawha Valley comprising numerous interconnected manufacturing steps and process areas. [ECF Nos. 313-1, at 14, 17; 358-4, at 14, 17; 359-7, at 14, 17 (“Sahu Expert Report”)]. Both Union Carbide and Covestro operate process areas within the South Charleston Plant. Id. at 17–18. Although Defendants both operate within the same overall facility, their respective process units and operations are distinct. [Sahu Expert Report, at 17–18; Sahu Deposition, at 234–36; ECF No. 375-2, at 2–3 (“Sahu Supplemental Expert Report”)]. Union Carbide operates multiple process areas at the South Charleston Plant, including the Chemical Mixing, Oxide Adducts, and TRITON areas, whereas Covestro has operated the South Charleston Plant’s polyols operations since April 1, 2000. [Machado Expert Report, at 17; Second Amended Complaint, ¶¶ 41–44; ECF Nos. 358-1, ¶¶ 4–5 (“Unsworn Declaration of Robert Lenahan”); 358- 2, at 234–36 (“Sahu Deposition”); Sahu Supplemental Expert Report (noting Covestro’s operations include the following “sources of EtO emissions[:] . . . Reactor #8, . . . Building 103, and . . . “Phase IV”)]. Although different before 1988, bulk shipments of EtO currently arrive at Union Carbide’s neighboring Institute Plant6 via rail tank cars and are stored in two underground tanks. [Sahu 6 [Machado Expert Report, at 9 (“the Institute facility . . . is located approximately five miles west of the South Charleston facility”)]. Supplemental Expert Report, at 14, 20]. From the Institute Plant, Union Carbide transports the chemical six miles through an underground pipeline to the South Charleston Plant. Id. at 14, 21. The EtO is then distributed to different batch reactors across the South Charleston facility, where it is reacted with other chemicals to produce various products.7 [Machado Expert Report, at 18]. Because EtO is readily ignitable, its use during reactions to create the products is subject to safety procedures and controls, including limits on the amount of unreacted EtO present in a reactor. Id. During the South Charleston Plant’s operations, EtO may be released into the air through both point sources and fugitive sources. [ECF No. 359-9, at 53–54]. Point-source emissions are releases from designated, identifiable discharge points, such as stacks, ducts, pipes, or vents.8 This occurs at the conclusion of reactions intended to create various products as described above; residual unreacted EtO may be vented to the atmosphere or routed through additional control devices. Id. at 53; see also [Machado Expert Report, at 18]. Fugitive emissions, by contrast, are releases that do not pass through a designated discharge point and instead occur from equipment such as valves, pumps, flanges, and other connectors. Id. at 54.9 Plaintiffs dispute the efficacy of Defendants’ measures to control EtO emissions. [Second Amended Complaint, ¶ 55; see generally Sahu Expert Report; Sahu Supplemental Expert Report]. Plaintiffs allege in the operative Second Amended Complaint that Defendants have significantly underestimated their EtO emissions and failed to adequately address them. [Second Amended Complaint, ¶¶ 53–57]. Plaintiffs allege putative class members in the surrounding communities 7 These products have applications as cleaners, de-greasers, deicers, paints and paint primers, brake fluid, and personal care products, [Machado Expert Report, at 12], as well as chemical additives for gasoline, jet fuels, water-based paints, cheese, baked goods, and other foods, [Sahu Expert Report, at 13]. 8 See Sommerville, 149 F.4th at 423 n.4 (quoting ECF No. 410, at 22 n.5) (“‘Point source’ or ‘stack emissions’ are ‘air releases that occur through identifiable confined air streams, such as stacks, ducts or pipes.’”) (citation modified). 9 See Sommerville, 149 F.4th at 423 n.4 (quoting ECF No. 410, at 22 n.5) (“‘[F]ugitive emissions’ are ‘releases to air that don’t occur through a confined air stream’ such as via ‘equipment leaks.’”) were exposed to elevated ambient concentrations of EtO for decades, increasing their risk of developing cancer, which Plaintiffs allege requires medical monitoring. Id. at 61; see generally [ECF Nos. 313-2 (“Stayner Expert Report”); 313-3 (“Wells Expert Report”)]. B. Plaintiffs’ Medical-Monitoring Theory “A claim for medical monitoring seeks to recover the anticipated costs of long-term diagnostic testing necessary to detect latent diseases that may develop as a result of tortious exposure to toxic substances.” Bower, 522 S.E.2d at 429. West Virginia recognizes a cause of action for medical monitoring expenses “where it can be proven that such expenses are necessary and reasonably certain to be incurred as a proximate result of a defendant’s tortious conduct.” Id. at 431. In order to prevail on a medical monitoring claim, a plaintiff must prove that: (1) he or she has, relative to the general population, been significantly exposed; (2) to a proven hazardous substance; (3) through the tortious conduct of the defendant; (4) as a proximate result of the exposure, plaintiff has suffered an increased risk of contracting a serious latent disease; (5) the increased risk of disease makes it reasonably necessary for the plaintiff to undergo periodic diagnostic medical examinations different from what would be prescribed in the absence of exposure; and (6) monitoring procedures exist that make the early detection of a disease possible. Id. at 432–33; see also Sommerville, 149 F.4th at 416 (quoting id.). Critically, Bower does not require present physical injury—the exposure itself and the concomitant need for medical testing constitute the injury. Id. at 430–431; see also Sommerville, 149 F.4th at 419. Nor does Bower require a particular quantification of the increased risk. Bower, 522 S.E.2d at 433. But the plaintiff must show that the exposure resulted in a significantly increased risk of a serious latent disease compared to the risk that would exist absent the exposure. Id. Finally, the defendant’s conduct must be a proximate cause of the significantly increased risk. Bower, 522 S.E.2d at 433. Thus, where a plaintiff’s prior risk or exposure was already sufficient to warrant the same monitoring sought as a result of the defendant’s subsequent conduct, plaintiff’s subsequent exposure does not itself create a compensable monitoring expense. Id. (quoting Hansen v. Mountain Fuel Supply, 858 P.2d 970, 980 (Utah 1993)). Accordingly, the plaintiff must establish that the defendant’s tortious conduct contributed to a significantly increased risk that would not otherwise have existed. Bower, 522 S.E.2d at 433. Plaintiffs assert Defendants owed a duty of reasonable care to neighboring communities to prevent off-site toxic contamination, and contends that Defendants breached that standard by operating the South Charleston Plant without adequate emission controls. [Second Amended Complaint, ¶¶ 48, 53]. Plaintiffs’ Complaint attempts to distinguish distinguishes between Defendants based on their respective operational timelines and roles. With respect to Union Carbide, Plaintiffs’ theory of liability as pleaded in the operative Second Amended Complaint encompasses over four decades of chemical production, which Plaintiffs allege was marked by unverified flare destruction efficiencies, unmonitored fugitive equipment leaks, failed compliance testing, and historical facility-wide releases across three process units. [Second Amended Complaint, ¶ 41; Sahu Expert Report, at 13, 47–49, 65]. Plaintiffs allege that these practices resulted in the tortious release of carcinogenic EtO into the surrounding community. [Second Amended Complaint, ¶ 41]. As to Covestro, Plaintiffs pled that duty and breach arise from its active operation of polyols facilities at the South Charleston Plant since 2000. [Second Amended Complaint, ¶ 44; ECF No. 358-1, at 4–5 n.1; Sahu Supplemental Expert Report]. Because polyols production relies directly on EtO as a chemical feedstock, Plaintiffs contend that operating these units inherently generated point-source process vents and fugitive equipment leaks, and that Covestro breached its standard of care by conducting these operations without using leak-free components to minimize or eliminate its EtO emissions. [Sahu Deposition, at 296–97]. Plaintiffs likewise allege both Defendants failed to assess the impacts to the densely populated residential community when deciding to emit EtO from the South Charleston facility. [Sahu Expert Report, at 29, 67; Sahu Deposition, at 295–97 (Dr. Sahu testifying that his standard of care opinions apply to all Defendants, including Covestro)]. Plaintiffs assert they will establish tortious conduct through a combination of evidence concerning Defendants’ operations, emissions practice, and compliance with applicable regulations and controls, together with expert opinions regarding the sources and quantities of EtO released from those operations. [Sahu Expert Report]. Whether that evidence establishes a breach of a duty of reasonable care, as opposed to merely demonstrating that EtO was emitted, is a distinct question from whether Plaintiffs can establish the resulting significant exposure and increased risk required under Bower. Plaintiffs must establish that Defendants’ tortious conduct caused the significant exposure and increased risk requiring periodic monitoring as required by Bower. Plaintiffs rely exclusively on expert testimony to make that connection. Plaintiffs’ experts will rely on quantitative air dispersion modeling to estimate ambient EtO concentrations attributable to Defendants’ emissions. [Sahu Expert Report, at 68–69, 78; Sahu Deposition; Sahu Supplemental Expert Report]. The resulting modeled concentrations are then used to calculate the cumulative EtO exposure experienced by Plaintiff Sommerville and members of the proposed class. See [Sahu Expert Report, at 68–69, 78]. Plaintiffs’ experts use those estimated exposures to quantify the resulting relative risk of cancer. [Wells Expert Report; Stayner Expert Report]. Plaintiffs contend the calculated risk for proposed class members exceeds 1 in 10,000—a level that the medical experts opine makes periodic diagnostic medical monitoring reasonably necessary to detect latent cancers. [ECF No. 313-9 (“Knopf Expert Report”)]. Plaintiffs’ theory thus relies on a series of evidentiary links: Defendants’ operations allegedly generated EtO emissions; those emissions allegedly resulted in significant exposure to Plaintiff Sommerville and proposed class members; and the resulting exposure allegedly increased their risk of serious latent disease sufficiently to make medical monitoring reasonably necessary. Different experts address different links in that chain. Their respective opinions, and the extent to which those opinions reliably establish each required Bower element, provide important context for the parties’ class-certification arguments. C. Plaintiffs’ Experts Plaintiffs’ proof runs through five experts, each handing off to the next. Plaintiffs’ experts address distinct aspects of the proposed proof, including emissions, exposure, causation, and increased risk to meet Bower’s requirements. Plaintiffs’ experts include Dr. Ranajit Sahu, [Sahu Expert Report; Sahu Supplemental Expert Report], Dr. Martin Wells [Wells Expert Report], Dr. Leslie T. Stayner, [Stayner Expert Report], Dr. Kevin Knopf, [Knopf Expert Report], and Dr. Charles Cowan [ECF Nos. 359-22 (“Cowan Expert Report”); 313-4 (“Cowan Rebuttal Report”). Dr. Sahu takes the plant’s emissions records, weather data, and other records and uses them to calculate how much EtO was in the air at fourteen census tracts around the South Charleston and Institute plants, year by year to determine Plaintiff’s and putative class members’ cumulative exposure. Dr. Wells takes Dr. Sahu’s cumulative exposure calculations and uses them in his own model to determine a putative class member’s estimated increased lifetime risk of lymphoid cancer as a result of their exposure, which he calls their “relative risk.” [Wells Expert Report, at 3, 5–6]. Dr. Wells opines an increased risk of one in 10,000 is a significantly increased risk. Id. Dr. Stayner opines regarding general causation and identifies which cancers are caused by EtO exposure. Dr. Knopf testifies that anyone whose risk went up by at least one in 10,000, the number identified by Dr. Wells, ought to be enrolled in a monitoring program. Dr. Cowan testifies to the exposed populations, including the number of people who were allegedly exposed to EtO. 1. Dr. Ranajit Sahu (Process Engineering and Emissions)10 Dr. Ranajit Sahu, a mechanical engineer with over thirty years of experience, serves as Plaintiffs’ process engineering and emissions expert. [Sahu Expert Report, at 6–7]. In this expert report, Dr. Sahu provides an expert opinion on EtO “emissions . . . from Union Carbide’s Institute and South Charleston facilities . . . between 1984 and 2019.” Id. at 6.11 Dr. Sahu identifies five process areas at the South Charleston Plant that he associates with EtO use or emissions during the period from 1984 through 2019. Id. at 17–18, 38–39. Within those areas, he identifies both point- 10 I note a potential issue concerning Dr. Sahu’s use of “UCC” and “Union Carbide” in his report. At his deposition, Dr. Sahu testified that he used those terms generically to refer to the operations at the South Charleston and Institute Plants as a whole, without distinguishing among the entities that owned or operated particular facilities or process units during the relevant period. He explained: “I was using UCC generically as a site, both Charleston and West Virginia without getting into ownership.” [Sahu Deposition, at 240–41, 291 (“When I wrote that report, I was referring to the entire plant of Union Carbide without distinguishing between different owners and different operators at different time period[s].”), 292–97]. Dr. Sahu further testified his opinions concerning leak detection programming (“the LDAR”), “fugitive emission estimates,” “the citing of official operations,” “extended cookout,” “reactors” and “standard of care” were applicable to Covestro as well. Id. at 292. 11 I note the following foundational issue with the dates in the proposed class definitions in both the Motion to Certify, [ECF No. 313], and Motion to Redefine Class, [ECF No. 493], in light of Dr. Sahu’s report. The class period as pled in the Complaint begins in 1978. [Second Amended Complaint, ¶ 69]. The Motion to Certify identifies the class period beginning in 1981. [ECF Nos. 313; 314]. On September 4, 2026, Plaintiff filed a Motion to Redefine the Class. [ECF No. 493]. Plaintiff seeks to redefine the class to, among other things, change the relevant class period back to beginning in 1978. See [ECF Nos. 493; 494]. The issue for both dates, which is discussed further below, is that Dr. Sahu’s analysis begins in 1984. See [Sahu Expert Report, at 6 (“The time period of this analysis is between 1984 and 2019.”). Accordingly, the class period could not be defined to include pre-1984 conduct because Plaintiff has no pre-1984 evidence since that’s when Dr. Sahu’s analysis begins. source and fugitive sources of EtO emissions, including process vents, storage tanks, railcar- loading operations, and potential releases from valves, pumps, and pipe fittings. Id. at 38–40. Dr. Sahu also evaluates the measures used to control EtO emissions and the operational practices at the two Plants. He notes that Union Carbide used industrial flares at the Institute Plant to combust excess EtO but that “[n]o flares are known to exist at the South Charleston plant for EtO destruction.” Id. at 29, 65. Union Carbide reported that its Institute Plant flares achieved 99.8- percent EtO destruction efficiency. Id. at 41–45. Dr. Sahu opines, however, that this figure was based on a 1987 internal memorandum rather than testing conducted under actual operating conditions. Id. He further notes that the underlying small-scale testing did not account for wind conditions. Id. at 41–42. When the flares were subsequently tested in May 2002, they failed to satisfy the applicable heat requirements. Id. at 47–49. Dr. Sahu further opines that Union Carbide did not report those failed test results to regulators and did not conduct subsequent testing of the flares. Id. at 49. Dr. Sahu likewise addresses Defendants’ controls for fugitive emissions and their operational practices. He opines that a substantial portion of the EtO emissions from the two Plants consisted of “uncontrolled fugitive emissions,” arising “from leaking equipment and other sources.” [Sahu Expert Report, at 65]. With respect to the facilities’ leak detection programs, Dr. Sahu identifies a 500-parts-per-million threshold for triggering repairs. Id. at 65–66. He opines that the operators did not evaluate the effectiveness of the leak detection programs or quantify the extent to which EtO emissions from fugitive sources reached the surrounding community. Id. Dr. Sahu likewise addresses the use of extended cookout procedures, opining the operators lacked data demonstrating that extended cookout reduced EtO emissions under normal operating conditions. Id. at 65. Dr. Sahu further opines that the operation of the facilities did not adequately account for the proximity of surrounding residential communities. Id. at 67. He found there was no evidence that Defendants considered the proximity of nearby residential communities when locating or expanding their EtO-related operations. Id. Because comprehensive, facility-specific emissions records were unavailable for portions of the relevant historical period (1985 through 1995), Dr. Sahu reconstructed a year-by-year inventory of EtO emissions from the two Plants from 1984 through 2019. [Sahu Expert Report, at 32–39, 55–64]. To fill gaps in the historical records, he relied on a combination of Defendant’s internal records from 1996 through 2019, state environmental filings, historical West Virginia Air Pollution Control Commission records, and federal Toxics Release Inventory (“TRI”) data. Id. Using this reconstructed emissions inventory, Dr. Sahu directed an air-dispersion modeling analysis using the United States Environmental Protection Agency’s (“EPA”) AERMOD model, which Dr. Andrew Gray performed under Dr. Sahu’s supervision. Id. at 68. The analysis was designed to estimate how EtO released from the two Plants would have dispersed through the surrounding air and the concentrations that would have been present at various locations from 1984 through 2019. Id. Dr. Sahu divided the emission sources into four groups: South Charleston source units other than tanks, South Charleston tanks, Institute source units other than flares, and the two elevated Institute flares. Id. at 69, 77. The model accounted for the location and physical characteristics of each emission source so that it could estimate how EtO released from that source would disperse through the air. Id. at 69. The source locations and other parameters used in the model were drawn primarily from West Virginia Department of Environmental Protection (“WVDEP”) modeling files and, for certain additional sources, documents produced in discovery. Id. at 69, 77–78. The model then estimated EtO concentrations at locations throughout a grid surrounding the two Plants, as well as at two specific residential addresses used as examples. Id. at 70–77. To account for the role of weather in determining where emissions travel, the model also incorporated historical meteorological data, including site-specific data for most sources and year- specific data for the two elevated Institute flares based on their height and plume characteristics. Id. Dr. Sahu explained at his deposition that the AERMOD analysis predicts ambient EtO concentrations at outdoor receptor locations—that is, the concentration of EtO in the outdoor air at a particular location. [Sahu Deposition, at 118–19]. Dr. Sahu testified the model can predict the concentration at any location designated as a receptor, depending on how the model is configured. Id. at 119. Dr. Sahu acknowledged that ambient concentrations may differ from concentrations in particular indoor environments, such as homes, workplaces, schools, and vehicles. Id. at 118–19. Still, he opines that, on an annual-average basis, differences between indoor and outdoor EtO concentrations are not material because of the substantial exchange of air between indoor and outdoor environments. [ECF No. 442-3, at 13 (“Sahu Rebuttal Report”). He further notes that EtO can persist in the air for extended periods, such that outdoor air containing EtO that enters a building does not simply dissipate rapidly. Id. Based on these considerations, Dr. Sahu opines that the ambient concentrations estimated by his model provide an appropriate basis for evaluating an individual’s exposure, including while that individual is inside the home. Id. Dr. Sahu next used those modeled ambient concentrations to evaluate the extent of EtO exposure at various locations. He compared the ambient EtO concentrations estimated at the two example residences and throughout the broader modeling area with EPA’s national average background EtO concentration and a cumulative exposure-based cancer-risk threshold. Id. at 78–83. After his May 2023 deposition, Dr. Sahu identified a discrepancy in the emission rates used for certain Covestro-associated sources. Specifically, the model had used a constant 2017 emission rate for those sources rather than the year-by-year emission rates he had provided. [Sahu Deposition; Sahu Supplemental Expert Report, at 1]. Dr. Sahu subsequently submitted a supplemental report presenting revised modeling results using the year-by-year emission rates. Sahu Supplemental Expert Report, at 1. He concluded that the revised inputs did not materially change the estimated EtO concentrations or alter his opinions. 12 Sahu Supplemental Expert Report. Dr. Sahu opines that his modeling can separately account for EtO emissions attributable to Covestro’s operations. Covestro’s South Charleston operations include identified emission points associated with its polyols production, and Dr. Sahu treats those sources as a discrete source group within the South Charleston Plant. [Sahu Expert Report, at 68–69, 78]. He testified that his model can isolate the emissions attributable to that source group and distinguish them from emissions associated with other sources at the Plant. [Sahu Deposition, at 297–98]. In identifying the information specific to Covestro that he considered, Dr. Sahu pointed to Covestro’s reported emissions, the emission sources associated with its polyols operations, and the information used as inputs to his dispersion model. Id. Thus, although Covestro’s sources are incorporated into his broader South Charleston analysis, Dr. Sahu testified his modeling permits him to separately model Union Carbide’s and Covestro’s identified sources to the ambient EtO concentrations. Further, although his model combines emissions from the Institute and South Charleston plants, he testified 12 I do not now revisit the admissibility of Dr. Sahu’s testimony. But I previously noted “[t]he accuracy of the model depends upon the rigor applied in the input gathering process” and, based on a number of issues, I concluded Dr. Sahu based his data on his own assertions rather than critically examining the operation history. [ECF No. 410, at 23–25 (citation modified)]. The Fourth Circuit concluded I abused my discretion in so deciding because I “ignored Dr. Sahu’s testimony” from his supplemental report. Sommerville, 149 F.4th at 426. his model is capable of separately modeling only emissions from the South Charleston Plant. [Sahu Rebuttal Report, at 20]. Dr. Sahu also offers opinions concerning the standard of care applicable to the operation of the relevant process units. He testified that operators of those units could have exercised greater care in handling EtO, including by using leak-free components to reduce or eliminate fugitive emissions. [ECF No. 358-2, at 296–98]. He likewise addresses the use of leak detection programs, fugitive-emission controls, extended-cookout procedures, and reactor operations, and testified that those opinions apply to both Union Carbide and Covestro. Id. at 292, 296–98. With respect to Covestro specifically, Dr. Sahu noted that Covestro continued to report fugitive EtO emissions from its polyols operations in 2017 through 2019 and that he saw no indi