Conservation Law Foundation, Inc. v. Shell Oil Company

Opinion

UNITED STATES DISTRICT COURT DISTRICT OF CONNECTICUT

Conservation Law Foundation, Inc.,

Plaintiff, Civil No. 3:21-cv-00933 (JAM)

v.

Shell Oil Co., et al., March 29, 2024

Defendants.

RULING AND ORDER ON MOTION TO COMPEL (ECF No. 326-1)

This is a citizen suit under the Clean Water Act (“CWA”), 33 U.S.C. §§ 1251 et seq., and the Resource Conservation and Recovery Act (“RCRA”), 42 U.S.C. §§ 6901 et seq. The plaintiff, Conservation Law Foundation, Inc. (“CLF”), is a nonprofit corporation “dedicated to the conservation and protection of New England’s public health, environment, and natural resources.” (Am. Compl., ECF No. 47, ¶ 9.) It claims that the bulk oil storage terminal next to the New Haven Harbor (the “Terminal”) has not been adequately prepared for “the reasonably foreseeable risks . . . posed by severe weather, including risks associated with climate change.” (ECF No. 149, at 2.) In this suit, it seeks to hold five companies accountable under the CWA and RCRA for those alleged failures – Shell Oil Co. n/k/a Shell USA, Inc. (“Shell Oil”), Equilon Enterprises LLC (“Equilon”), Shell Petroleum Inc. (“Shell Petroleum”), Triton Terminaling LLC (“Triton”), and Motiva Enterprises LLC (“Motiva”) (together, “Defendants”). The case has been plagued by discovery disputes. By the Court’s count, the parties have filed fifty-five discovery or discovery-related motions. In the most hard-fought motion to date, CLF sought an order compelling the Defendants to comply with eleven interrogatories and sixty- four requests for production to which they had objected on a variety of grounds. (ECF No. 149) (“First Motion”). The Court resolved that motion in a fifty-one-page Ruling and Order, generally agreeing with CLF on many of the parties’ relevance disputes, but nevertheless denying relief with respect to ten interrogatories and fifty-three requests for production, principally on grounds of undue burden, overbreadth, and lack of proportionality. See generally Conservation L. Found. v.

Shell Oil Co., No. 3:21-cv-933 (JAM) (TOF), 2023 WL 5434760 (D. Conn. Aug. 22, 2023) (hereinafter “Prior Ruling”). The Court added, however, that the denial was “without prejudice to more narrowly tailored and targeted discovery into the subject matter of that interrogatory or request for production.” (Id. at *24.) CLF then served revised requests for production on the Defendants on September 5, 2023. (ECF No. 309-2.) Upon reviewing them, the Defendants contended that they were still irrelevant in many instances, and were also unduly burdensome, overbroad, and disproportionate to the needs of the case. A round of negotiations ensued. (See discussions, ECF No. 326-1, at 3–6; ECF No. 325-1, at 9–11.) CLF revised its requests yet again on October 3, 2023 (ECF No. 309-5), after which negotiations continued, but the parties ultimately reached impasse. (ECF No. 326-1, at 6;

ECF No. 325-1, at 11.) CLF has now filed another motion to compel. (ECF No. 326-1.)1 It seeks an order compelling the Defendants to comply with thirty-nine of the forty-one requests for production served on September 5, 2023, as revised on October 3, 2023. (Id. at 1; see also ECF No. 309-5 (the “Final Revised Requests”).) The parties filed over 500 pages of briefs and exhibits, and the Court heard nearly an hour and a half of oral argument. (ECF No. 342.) For the reasons that

1 The motion was originally filed in partially redacted form at ECF No. 309. In a rare instance of cooperation between these parties, they were evidently able to resolve any disputes over how much of the motion and supporting exhibits should be sealed in response to the Defendants’ confidentiality designations. (See ECF No. 326.) CLF then re-submitted the motion in unredacted form at ECF No. 326-1. follow below, CLF’s motion will be granted in part and denied in part, as set forth more fully in Section III. I. APPLICABLE LEGAL STANDARDS In its Prior Ruling, the Court discussed the legal standards applicable to motions to compel

at considerable length. Prior Ruling, at *10–12. There is therefore no need for a similarly lengthy recital here; a summary will suffice. “Subject to the proportionality requirement and other limitations set forth in Rule 26, a party may discover relevant, nonprivileged information in the other party’s possession.” Prior Ruling, at *10. “While the Federal Rules of Civil Procedure do not define ‘relevant,’ the operative definition can be found in Rule 401 of the Federal Rules of Evidence.” Gaynor v. City of Meriden, No. 3:17-cv-1103 (CSH), 2019 WL 2537669, at *2 (D. Conn. June 20, 2019). Under that rule, “[i]nformation is ‘relevant’ if it ‘(a) has any tendency to make a fact more or less probable than it would be without the evidence, and (b) the fact is of consequence in determining the action.’” Prior Ruling, at *10. At the discovery phase of the case, the concept of relevance is accorded an

“extremely broad” construction. Martino v. Nationstar Mortg. LLC, No. 3:17-cv-1326 (KAD), 2019 WL 2238030, at *1 (D. Conn. May 23, 2019). “When a discovery-seeking party moves to compel responses to its requests, it bears the burden to demonstrate that the requests are within the scope of Rule 26(b)(1).” Prior Ruling, at *11. “Once the requesting party has made a prima facie showing of relevance, however, it is up to the responding party to justify curtailing discovery.” Doe v. Wesleyan Univ., No. 3:19-cv-1519 (JBA) (TOF), 2021 WL 4704852, at *3 (D. Conn. Oct. 8, 2021) (citation and quotation marks omitted). “In particular, ‘[a] party may object to a relevant discovery request . . . if it is ‘overly broad’ or ‘unduly burdensome.’” Prior Ruling, at *11 (quoting Sullivan v. StratMar Sys., Inc., 276 F.R.D. 17, 19 (D. Conn. 2011)). Moreover, “[d]iscovery requests must be proportional to the needs of the case . . . and where the requests lack proportionality, the responding party may validly object.” Id. at *12 (citing Conservation L. Found. v. All-Star Transp., LLC, No. 3:21-cv-201 (JBA) (TOF), 2022 WL 16901999, at *3 (D. Conn. Nov. 11, 2022) (“All-Star”). The proportionality

factors are set forth in Fed. R. Civ. P. 26(b)(1), and they “focus[] on the marginal utility of the requested discovery.” Conn. Mun. Elec. Energy Coop. v. Nat’l Union Fire Ins. Co., No. 3:19-cv- 839 (JCH) (RAR), 2020 WL 6074204, at *6 (D. Conn. Oct. 15, 2020). “[E]ven relevant information must be reasonably proportional to the value of the requested information, the needs of the case, and the parties’ resources.” Parimal v. Manitex Int’l, Inc., No. 3:19-cv-1910 (MPS) (SALM), 2021 WL 1978347, at *5 (D. Conn. May 18, 2021). “Each party bears a burden with respect to proportionality.” Prior Ruling, at *12. “For example, ‘a party claiming that a request is important to resolve the issues should be able to explain the ways in which the underlying information bears on the issues as that party understands them.’” All-Star, 2022 WL 16901999, at *2 (quoting Fed. R. Civ. P. 26 advisory committee notes to 2015

amendments) (brackets omitted). “Conversely, ‘the responding party bears the burden on the “burden and expense” element of the proportionality analysis.’” Id. (quoting Huseby, LLC v. Bailey, No. 3:20-cv-167 (JBA) (TOF), 2021 WL 3206776, at *8 n.3 (D. Conn. July 29, 2021)). This burden is not met “simply by making a boilerplate objection that [the requested discovery] is not proportional.” Endurance Am. Spec. Ins. Co. v. Wm. Kramer & Assocs., LLC, No. 3:18-cv- 192 (MPS) (RAR), 2020 WL 6940761, at *7 (D. Conn. Mar. 30, 2020). “To successfully resist discovery on grounds of undue burden or expense, the responding party ordinarily must submit affidavits or offer evidence revealing the nature of the burden.” All-Star, 2022 WL 16901999, at *2 (citation and quotation marks omitted). “District courts may, however, overlook the lack of an affidavit or other evidence of burden when the requests are overly broad or unduly burdensome on their face.” Id. II. DISCUSSION A. Issues That the Parties Have Addressed on a Blanket Basis

The parties’ briefs do not discuss the relevance, burden, and proportionality of the Final Revised Requests on an individual, request-by-request basis. (See, e.g., ECF No. 326-1, at 8 n.2) (contending that “[a]ddressing each objected-to Revised RFP individually is not possible within the page limitations because of Defendants’ capacious objections”); see also D. Conn. L. Civ. R. 37(b) (stating that “[w]here several different items of discovery are in dispute” in a motion to compel, “counsel shall, to the extent possible, group the items into categories in lieu of an individual listing of each item”). Instead, the parties invite the Court to address several issues on a blanket basis, as they did with the First Motion. Prior Ruling, at *13. First, the parties continue to dispute the relevance of those Final Revised Requests that implicate the Defendants’ petroleum storage facilities other than the New Haven Terminal. Many

of the requests at issue in the First Motion likewise encompassed other facilities – and indeed, several ostensibly encompassed all Shell facilities anywhere in the world. See, e.g., Prior Ruling, at *22 (discussing Request No. 18, which sought “[a]ll Documents related to the work of Shell’s global Soil & Groundwater Team”). Yet since the Court addressed those requests principally on burden and proportionality grounds, see id., the relevance of documents specifically concerning, say, Shell’s facilities in Port Fourchon, Louisiana, or Sewaren, New Jersey, was not an area of focus. In its Final Revised Requests, however, CLF narrowed its inquiry. CLF has learned through other discovery that the Defendants maintain a “Risk Assessment Matrix” that “scores and classifies risks at facilities” with respect to several categories of hazard, including “weather conditions” such as “[f]loods, [w]indstorm, . . . [s]ea conditions,” and so forth. (ECF No. 326-1, at 12–13.) Facilities are given an alphanumeric rating, with the number indicating the severity of the risk being run and the letter indicating the likelihood of the risk coming to fruition. (Id. at 13

n.4.) “Each alphanumeric rating combination is [then] assigned a . . . color – from the lowest severity/likelihood of light blue . . . to yellow . . . to the highest severity/likelihood of red[.]” (Id.) For the most part, CLF’s Final Revised Requests no longer ask for documents concerning all Shell facilities worldwide; rather, many requests are now limited to “the [New Haven] Terminal or coastal facilities that store or handle petroleum in bulk and that have a Risk Assessment Matrix (RAM) rating of yellow or red for category H-09.01 (weather conditions).” (E.g., Final Rev. Req. 10, ECF No. 309-5, at 2.) CLF says that the Defendants gave the New Haven Terminal a score of “4C” in the dimension of “weather conditions,” which corresponds to a “yellow” color code. (See ECF No. 326-1, at 13.) Thus, in asking for documents concerning coastal facilities that merited a “yellow” or “red” code in that same dimension, CLF is no longer asking about all facilities,

anywhere in the world; instead, it is essentially asking for documents concerning coastal bulk petroleum storage facilities that are as risky or riskier than the New Haven Terminal in the Defendants’ own scoring system. And CLF does not seek documents related to all such facilities, but rather only fifteen of them. (ECF No. 326-1, at 38.) The Defendants argue that other facilities are irrelevant to the claims and defenses in this case – or, at least, not relevant enough to support the burdens that the requests would impose. (ECF No. 325-1, at 16–19 (contending that the scope of the requests “sweeps in irrelevant information”) and 23 (“Even if . . . there were some marginal relevance, the scope of what CLF is requesting is far out of proportion” to the needs of the case).) The Defendants acknowledge, as they must, that the question of whether “best industry practice” required consideration of climate change risks in the design and operation of the New Haven Terminal is a key issue in the case. (See ECF No. 325-1, at 20–21; see also Tr. of Hrg. on Mot. for Partial Summ. J., ECF No. 305, at 83.) But they contend that this should lead, at most, to discovery about “facilities ‘like’ the New

Haven Terminal.” (ECF No. 325-1, at 22) (citing Tr. of Hrg. on Mot. for Partial Summ. J., ECF No. 305, at 85:13-15). And they argue that there are no other facilities sufficiently “like” New Haven to support discovery. (ECF No. 325-1, at 22–23.) They claim, for example, that the much- discussed Port Fourchon facility “couldn’t be less ‘like’ the New Haven terminal” because it has “unique climate risk considerations (e.g., extreme subsidence) that do not apply to the Port of New Haven.” (Id. at 23–24.) They argue that facilities like Port Fourchon are “demonstrably poor fit[s] for determining best industry practices applicable to [Stormwater Pollution Prevention Plans, or] SWPPPs[,] and consideration of climate change risks in Connecticut.” (Id. at 24.) The Court disagrees. The key inquiry in this case is about best industry practice, and because the Defendants are of course members of the industry, what they do at their other facilities

may inform the meaning of that term. (Cf. Tr. of Hrg. on Mot. to Compel, ECF No. 344, at 29:7– 12) (acknowledgment by Defendants’ counsel that “other Shell companies” are “members of the industry” whose activities “inform industry practice,” if only “narrowly”). As Judge Meyer observed in ruling upon the Defendants’ motion for partial summary judgment, even other oil companies’ facilities and procedures can be relevant to a best industry practices analysis, let alone the Defendants’ own practices at their other facilities. (See Tr. of Hrg. on Mot. for Partial Summ. J., ECF No. 305, at 83:11–15) (rejecting suggestion that “only Shell’s practices rather than industry-wide practices establish what constitutes best industry practice”). And while Judge Meyer did indeed suggest that another facility must be “like” the New Haven Terminal to be relevant to that analysis, the rest of his sentence is important: he framed the key question as “whether best industry practice requires consideration of climate change for a facility like the Shell facility that sits adjacent to the New Haven Harbor.” (Id. at 84:19–22) (emphasis added). Thus, in limiting its requests to certain “coastal” facilities, CLF has reasonably tailored its inquiry to

facilities that are “like” the New Haven Terminal in Judge Meyer’s contemplation. Moreover, CLF is not obligated to accept the Defendants’ unilateral determination of what constitutes a “like” facility at the discovery stage. See Prior Ruling, at *18 (quoting Xchange Telecom Corp. v. Sprint Spectrum, L.P., No. 14-cv-54 (GLS) (CFH), 2015 WL 773752, at *3 (N.D.N.Y. Feb. 24, 2015) for the proposition that “[n]o party possess[es] the unilateral ability to dictate the scope of discovery based on their own views of the parties’ respective theories of the case”). In asking only about coastal facilities that store petroleum in bulk and that have “yellow” or “red” Risk Assessment Matrix scores, CLF has sufficiently tailored its requests to “like” facilities for discovery purposes and, by extension, has made a relevant and reasonably limited inquiry. To be sure, it may ultimately prove true that “best industry practices” are as site-specific

as the Defendants contend. Put another way, it may ultimately prove true that “best industry practices” at Port Fourchon or Sewaren have no bearing on “best industry practices” in New Haven. But this is an unresolved factual issue (see Tr. of Hrg. on Mot. for Partial Summ. J., ECF No. 305, at 85:13–16), and CLF is entitled to explore it with proportional discovery. Indeed, at oral argument the Defendants acknowledged that events at another facility can be relevant to this case, so long as they “inform[] what goes on in New Haven.” (Tr. of Hrg. on Mot. to Compel, ECF No. 344, at 59:11–21.) The Defendants raise other objections under the heading of relevance, but the Court does not find them persuasive. They first argue that each Final Revised Request encompassing other facilities is worded so broadly as to “sweep[] in irrelevant information.” (ECF No. 325-1, at 16.) They cite Request No. 9 as one example, noting that it sought production of documents containing the word “flood,” and arguing that it would pull in documents in which someone used that word in a sense not relevant to this case – for instance, an e-mail in which someone complained about

receiving a “flood of information.” (Id. at 18.) Issues of this sort can, however, be addressed in the production process, and they present no argument against the relevance of the request in a case that is very much about the potential for flooding. (See Am. Compl., ECF No. 47, ¶¶ 340–52, 393–94.) The Defendants also complain that CLF has “conflated Defendants with other Shell group entities” and has wrongfully directed some of its requests to “Shell” (ECF No. 325-1, at 19), but CLF has clarified that it is only seeking documents and information in the Defendants’ “possession, custody, or control, as required under Rule 34(a)(1).” (ECF No. 330, at 3); see also Prior Ruling, at *15 n.5. The Defendants also note that the Prior Ruling “ruled that CLF’s discovery into facilities other than the Terminal for purposes of its operator liability theory is not permitted” (ECF No. 325-1, at 19 n.5), and this is true as far as it goes. Prior Ruling, at *15. But

while the Court did indeed observe that policies and procedures “beyond those that govern the Terminal” were “not supported by CLF’s operator liability claims,” Prior Ruling, at *16, those claims are not the only claims in the case. The Court did not say that documents concerning other facilities are per se irrelevant to the other claims, including the claims that turn on the “best industry practice” issue. While document requests directed to other facilities can of course raise issues of undue burden and lack of proportionality (see Section II.B infra), the Court disagrees that they categorically lack logical relevance under Fed. R. Civ. P. 26 and Fed. R. Evid. 401. The second issue that the parties address on a blanket basis is whether CLF genuinely narrowed its requests in response to the Prior Ruling. (See, e.g., ECF No. 325-1, at 17.) In that ruling, the Court denied relief to CLF on ten interrogatories and fifty-three requests for production on grounds of undue burden, overbreadth, and lack of proportionality. Prior Ruling at *23–24. Among other reasons, the Court noted that many of CLF’s requests for production “ask[ed] for ‘[a]ll documents’ ‘related to’ broad subject areas,” and it observed that “[a]ny production request

that begins this way is immediately suspect, because starting a request ‘with “All Documents” or “All Communications” . . . often is a red flag for overbreadth and undue burden.” Id. at *21 (quoting Hedgeye Risk Mgmt., LLC v. Dale, No. 21-cv-3687 (ALC) (RWL), 2023 WL 4353076, at *2 (S.D.N.Y. July 5, 2023)). The Court added, however, that its denial of relief was “without prejudice to more narrowly tailored and targeted discovery into the subject matter of that interrogatory or request for production.” Id. at *24. In its new motion, CLF contends that it “narrowed its requests to address the concerns raised in the” Prior Ruling. (ECF No. 309, at 4.) The Defendants disagree, asserting that “CLF had the opportunity to meaningfully narrow and tailor these requests (and follow the Courts’ [sic] ruling based on its own arguments), but . . . it has not.” (ECF No. 325-1, at 19.)

The Court has carefully reviewed each Final Revised Request and compared it with its predecessor, and it concludes that CLF is right much more often than the Defendants. Take, for example, Request No. 6. In its original formulation, that request sought production of “[a]ll Documents related to corporate control of any activities at the Terminal, including policies governing Defendants’ corporate interactions with each other and any parents and subsidiaries.” Prior Ruling, at *21. The Court previously observed that, “[w]hen considered alongside the temporal scope of the request and CLF’s broad definition of ‘relate,’ [the original] RFP 6 essentially asks for all documents constituting interactions between the Defendants, or between them and their parents, over the past twelve years,” including “[e]very e-mail between Shell Oil and Shell Petroleum on the one hand, and Shell plc on the other.” Id. The analog in the Final Revised Requests, by contrast, seeks production only of “policies” establishing, acknowledging, or limiting each Defendant’s authority to conduct activity at the Terminal – certainly a much narrower request, and one that is relevant to CLF’s operator liability theory. (ECF No. 309-5, at

1.) Similarly, the original Request No. 12 sought production of “[a]ll Documents related to Shell’s participation in” certain research initiatives, but Final Revised Request No. 12 is limited to “[r]eports that Shell authored.” (Id. at 3.) And original Requests 17 and 18 – which had demanded production of “[a]ll documents related to the work of” two entire business units anywhere in the world – has now been narrowed down to certain documents encompassing or referencing the Terminal. (Id. at 4.) There are, however, some requests that CLF did not meaningfully narrow. For example, the original Request No. 11 had sought “[a]ll Documents concerning risks to the Terminal’s physical infrastructure from climate change, flooding, sea level rise, and/or severe weather.” (Id. at 2.) Its Final Revised counterpart is no narrower; it seeks production of “[a]ll Documents

concerning risks to the Terminal’s physical infrastructure . . . posed by hurricanes, flooding, sea level rise, and/or severe weather, whether affected by Climate Change or not.” (Id.) To take another example, the original Request No. 40 sought “[a]ll Communications related to stormwater monitoring policies and practices at the Terminal,” but the Final Revised version seeks “[a]ll Communications concerning stormwater monitoring policies at the Terminal.” (Id. at 8.) CLF contends that the replacement of “related to” with “concerning” constituted a material narrowing of the request (ECF No. 309, at 4), but at oral argument, it struggled to explain how. (Tr. of Hrg. on Mot. to Compel, ECF No. 344, at 9:7) (remarks by counsel, indicating that she was “having a hard time coming up with a hypothetical” example of a document that would be responsive to the “related to” formulation, but not responsive to the “concerning” formulation). The Court concludes that this issue cannot be addressed on a blanket basis. There are a few instances in which CLF has not narrowed its request in any meaningful way, and in those

instances, it has provided no reason for the Court to depart from its Prior Ruling. But in most cases, the Final Revised Request is materially different from its original counterpart. The Court will address each request in Section II.B. Third, the parties dispute whether the Defendants have adequately supported their undue burden, overbreadth, and lack-of-proportionality objections to the Final Revised Requests. (See ECF No. 326-1, at 15–23, 30–34; ECF No. 325-1, at 19–31.) This dispute has two principal aspects, one of which concerns those requests that inquire after other “coastal facilities that store or handle petroleum products in bulk” and have “yellow” or “red” Risk Assessment Matrix codes for weather. (E.g., ECF No. 309-5, at 1.) In purported support of their undue burden and proportionality objections, the Defendants have offered a declaration from their “Health, Safety,

Security and Environment . . . Team Lead for the United States in Distribution Operations,” Ms. Kalisha Bennett. (ECF No. 325-1, at 26; ECF No. 324-5.) Ms. Bennett says, in substance, that it would be unduly burdensome for the Defendants to even identify which facilities are “coastal,” because they keep “no one list of facilities that includes [this] information.” (ECF No. 324-5, ¶ 7.) She likewise says that it would be burdensome to identify which facilities “store or handle petroleum products in bulk,” because again, “[t]here is no one place” in which that information resides. (Id. ¶ 9.) Finally, she says that “[g]enerating a list of facilities with a RAM rating of yellow or red for weather would be extremely difficult, if it is in fact even possible,” because “there is no one list or database that captures what all the individual RAM ratings have been or are currently for all the Americas terminals.” (Id. ¶¶ 10-11.) These claims do not adequately support the Defendants’ undue burden and lack-of- proportionality objections, for two independently sufficient reasons. First, it is simply implausible

that the Defendants – member companies of a large, multinational petrochemical conglomerate – cannot identify which facilities are “coastal” without undue burden, particularly when one considers that CLF expressly incorporated Shell’s own definition of “coastal” in its requests. (ECF No. 309-2, at 5 n.2.) It similarly strains credulity to suggest that they cannot easily identify which facilities are bulk petroleum storage facilities, or which have “yellow” or “red” RAM codes for weather risks. When asked at oral argument why those tasks “would take more than a couple of hours,” the Defendants could not say; instead, they pivoted to the separate question of the burdens of producing documents related to those facilities once they are identified. (Tr. of Hrg. on Mot. to Compel, ECF No. 344, at 38:24–39:4) (“I think the issue is not how much time to identify the facilities, but rather the process that would be entailed once they’re identified what the information

is that needs to be collected in that process.”). Second, even if the Defendants’ information systems are as deficient as Ms. Bennett asserts, it is well established that a party may not premise an undue burden or lack-of-proportionality objection on the obsolescence of its own systems. See, e.g., Malzberg v. New York Univ., No. 19-cv-10048 (LJL), 2020 WL 3618962, at *4 (S.D.N.Y. July 2, 2020) (“[T]o the extent that Defendant’s burden is a function of its own record-keeping, that factor should not prejudice Plaintiff’s ability to obtain clearly relevant information.”). The second aspect of this dispute concerns the burdens of collecting and producing responsive documents once other responsive facilities are identified. On this issue, the Defendants have come forward with a declaration from their e-discovery analyst, Mr. Frank Contreras (ECF No. 324-3), who also provided a declaration in opposition to CLF’s First Motion. (ECF No. 177- 7.) The new declaration explains that, much earlier in the case, the Defendants ran ninety-five search terms across twenty-five custodians’ electronic document repositories in response to CLF’s initial requests. (ECF No. 324-3, ¶ 17.) That search returned 1,747,411 documents, and the

Defendants used a combination of “software-based artificial intelligence” tools, “targeted searches,” contract attorney review, and litigation counsel review to winnow that cohort down to the 41,135 documents that they ultimately produced. (Id. ¶¶ 21–23; see also ECF No. 330, at 6 n.5.) Mr. Contreras’s declaration then goes on to disclose that the Defendants did not search any additional repositories or custodians in response to the Final Revised Requests. Instead, they ran three searches across a fourteen-custodian subset of the existing 1,741,411-document cohort, using search terms suggested by three of those requests. For example, in response to Final Revised Request No. 23 – which inquires about “berms,” “secondary containment areas,” and so forth – the Defendants ran a search using those terms, and the search returned 109,611 documents. (Id. ¶ 27.) And in response to Final Revised Request No. 48, which inquires after “wet weather,”

“hurricanes,” “storms,” and “floods,” a similar process returned 15,038 documents. (Id. ¶ 28.) These are figures for the New Haven Terminal, and the Defendants concede that they have no similar figures for any other facility because, again, they have not run any new searches across any other bodies of documents. But reasoning that each other facility may return a similar number of documents, Mr. Contreras asserts that it would cost over $5,000,000 and take “up to 250 days” to collect and conduct a first-level review of documents responsive to the eleven Final Revised Requests that implicate other facilities, leaving aside the remaining requests and assuming that there will be only twenty such facilities. (Id. ¶¶ 34–35.) Yet these claims do not account for the ways in which CLF has narrowed its requests. For example, Request No. 48 no longer seeks production of “[a]ll documents related to Terminal inspections . . . after storms, floods, inundation, and/or weather events,” but rather only “[p]olicies or presentations sufficient to show” these things. (ECF No. 309-5, at 10.) Similarly, Request No.

23 no longer seeks “[a]ll documents related to changes in physical infrastructure at the Terminal,” but instead seeks only “[d]ocuments sufficient to show the design, construction, maintenance, or modification of” berms, secondary containment areas, and so forth. (Id. at 5.) The Defendants have provided no reason to suppose that, after these limitations, the Final Revised Requests will require the same degree of effort or return the same numbers of documents as the initial requests. Mr. Contreras’s treatment of Request Nos. 25 and 26 is particularly telling on this point. As originally phrased, those two requests sought “[a]ll Documents related to Shell’s consideration of climate change in investment and spending decisions” and “[a]ll Documents related to the study of climate change and/or weather-related risks created, used, or relied on by Defendants[.]” (ECF No. 309-5, at 5–6.) In his initial declaration, Mr. Contreras used document counts from a similar

search in the parties’ Rhode Island case to predict that these two requests alone might implicate 233 bankers’ boxes of documents. (ECF No. 177-7, ¶ 23.a.) CLF then revised these requests in response to the Prior Ruling; in the Final Revised Requests, Number 25 is now limited to “policies concerning the consideration of Climate Change” at a limited range of facilities, and Number 26 is now limited to those facilities as well. (ECF No. 309-5, at 5–6.) Yet despite these new limitations, Mr. Contreras’s second declaration reports the same 233-box figure as the first. (ECF No. 324-3, ¶ 26.) That cannot be the case. In the declaration that he filed in connection with the First Motion, Mr. Contreras clearly described undue burdens that plainly outweighed the other Rule 26(b)(1) proportionality factors. For example, he explained that just one of CLF’s requests – Request No. 7, which inquired after documents related to environmental monitoring at the New Haven Terminal – might require review of 1,135 bankers’ boxes of documents, as the request was then phrased. Prior Ruling, at *20. But CLF has narrowed almost all its requests, and because Mr. Contreras does not account for that

narrowing in his new declaration, it cannot be said with the same assurance that the new requests are disproportionate to the needs of the case. As CLF explained at oral argument, the amount in controversy is substantial, the Defendants have greater relative access to the key information, and the issues raised by the case are important ones. (Tr. of Hrg. on Mot. to Compel, ECF No. 344, at 19–21, 26); see also Fed. R. Civ. P. 26(b)(1) (stating that proportionality requires consideration of, inter alia, “the importance of the issues at stake in the action, the amount in controversy, [and] the parties’ relative access to relevant information”).2 Fourth, the parties dispute the proper temporal scope of CLF’s requests. (See ECF No. 326-1, at 24–29; ECF No. 325-1, at 31–33.) Most of CLF’s Final Revised Requests seek documents from “January 1, 2011 to present.” (ECF No. 309-2, at 4.) The Defendants objected

to that timeframe as “not relevant, overbroad, unduly burdensome, and not proportional to the reasonable needs” of the case in many instances. (ECF No. 309-7, at 5.) In most cases they did not produce responsive documents “that pre-date 2017.” (See, e.g., id. at 7, 10.)

2 As this Ruling and Order was undergoing final proofreading and editing, CLF filed a motion for a discovery status conference. (ECF No. 369.) In that filing, it said that it has learned of reasons to disbelieve the statements in the Contreras declaration, and it asserted that it “has grounds to file a motion to strike Defendants’ statements in affidavits and briefs regarding the scope of their searches and associated undue burden objections[.]” (Id. at 2.) CLF suggested that the Court should withhold its decision on this motion until these concerns can be aired in more detail at a discovery conference. (Id. at 4–5.) Having held, however, that the Contreras declaration does not adequately support the Defendants’ undue burden or lack-of-proportionality objections even if it is true, the Court sees no reason to withhold decision on this motion pending resolution of any disputes over the declaration’s accuracy. The parties contended over this issue in the First Motion as well. When it filed that motion, CLF understood the Defendants’ objection to be rooted in the CWA’s five-year statute of limitation. (ECF No. 149, at 24.) It argued that a “statute of limitations is not a rigid barrier separating discoverable information from information outside the scope of discovery.” (Id.)

(quoting Hall v. Marriott Int’l, Inc., No. 3:19-cv-1715 (JLS) (AHG), 2021 WL 1906464, at *8 (S.D. Cal. May 12, 2021)). Beyond that legal point, its discussion of the relevance of pre-2017 documents was brief. CLF acknowledged that Motiva had been the permittee of the New Haven Terminal before 2017, and it further acknowledged that the CWA claims against Motiva had been dismissed. But it argued that, in part because the 2017 permit was merely a renewal of Motiva’s 2011 permit, the Defendants’ “knowledge and actions throughout the course of the Permit, and their statements to CT DEEP during the Permit term, are relevant for evaluating the state of Defendants’ knowledge and experience at the start of the current permit term and during their application for registration under the Permit.” (Id. at 25.) CLF also argued for a 2011 start date because it would encompass Tropical Storm Irene and Superstorm Sandy, and it regarded as

relevant any 2011–2016 documents showing how the Defendants adapted the New Haven Terminal in response to those events. (Id. at 26.) In other words, in the First Motion CLF’s brief argument for the relevance of pre-2017 documents focused on its so-called “disclosure” and “adaptation” claims. In their opposition to the First Motion, the Defendants agreed to the 2011 start date with respect to those requests directed to CLF’s RCRA claims, because they acknowledged that “historic waste can ‘currently’ be the cause of an alleged endangerment.” (ECF No. 177, at 32.) But they continued to object with respect to the many requests directed to the CWA claim, because “CLF has made no argument explaining why information from 12 years ago . . . will inform whether there are current permit violations.” (Id.) And as to Storms Irene and Sandy, the Defendants argued that “[d]uring that time period, Motiva was the operator of the Terminal[,]”, and “[a]ll Permit-based claims against Motiva have been dismissed.” (Id. at 33.) In the Prior Ruling, the Court “agree[d] with CLF on the general proposition that

information pre-dating a limitation period can sometimes be discoverable if relevant to claims within the limitation period.” Prior Ruling, at *13. But it added that “this general proposition only takes CLF so far,” because as the party bearing the burden to demonstrate the relevance of its requests, it still needed “to show that [the] pre-2017 portion of each request seeks relevant information.” Id. (citing Doe, 2021 WL 4704852, at *3). Because CLF had cited its so-called “disclosure” and “adaptation” claims in its argument for the relevance of pre-2017 materials, the Court held that it had not met its burden with respect to several requests premised on its “operator liability” theories. Id. at *19 (declining to order production of pre-2017 materials responsive to Requests 2–5 because CLF “has not shown the relevance of pre-2017 documents”). Yet because it decided most of the other CWA-related requests on grounds of undue burden, overbreadth, and

lack of proportionality, the Court did not expressly address the relevance of pre-2017 documents to CLF’s other claims. See generally id. at *19–22. Indeed, after it issued the Prior Ruling, the Court encouraged the Defendants to continue negotiating on this point. (Tr. of Hrg. on Mot. to Compel, ECF No. 284, at 30:17–22) (“I wouldn’t suggest that you guys dig your heels in too much in the negotiations about January 1st, 2017[,] as a hard stop on any production obligations.”). The parties were nonetheless unable to resolve this dispute, and they have now renewed it in the context of CLF’s Final Revised Requests. In its new motion, CLF discusses the relevance of pre-2017 documents to those of its requests premised on its “operator liability” theory, asserting that each Defendant can be liable in that period “due to their ownership of Motiva . . . and their employees’ participation in the operations of the Terminal during the Motiva ownership[.].” (ECF No. 326-1, at 25.) It cites Final Revised Request No. 45 as an example to which pre-2017 documents would be relevant to operator liability; that request inquired after “[p]olicies or presentations sufficient to show employee training for stormwater discharge monitoring at the

Terminal,” and CLF argues that “[i]f any of the non-Motiva Defendants have such trainings or policies, it would support CLF’s contention that those Defendants were directly involved in CWA operations prior to the transfer.” (Id. at 26.) CLF also argues that under the “continuing violation doctrine,” “if Defendants began violating the permit as soon as it went into effect in 2011, then they may be liable for civil penalties for the entirety of the period of ongoing violations.” (Id. at 27.) Finally, CLF notes that the Defendants have conceded the relevance of the 2011 start date with respect to its RCRA claims, but have “incorrectly assert[ed], without any justification, that information sought in the Revised RFP is relevant only to the CWA claims.” (Id. at 28.) It cites Final Revised Request 46 as an example of a request that explores claims under both statutes. (Id.) The Defendants respond by arguing that pre-2017 materials are irrelevant to the extant

claims in CLF’s Amended Complaint. (ECF No. 325-1, at 31–32.) They say that “CLF has not alleged pre-2017 violations of the General Permit against the non-Motiva Defendants,” and “these claims have already been dismissed as to Motiva.” (Id. at 31.) They also note that the Court did not accept CLF’s operator liability claims as a basis for pre-2017 discovery in the context of Initial Requests 2–5 and 62, and although CLF did not renew those requests in the Final Revised set, the Defendants urge the same result here. (Id.) While the Court will not call the dispute over the relevance of pre-2017 documents to CLF’s operator liability claims “much ado about nothing,” it is much ado about very little. At the time of the First Motion, only six of CLF’s sixty-five requests were exclusively premised on those claims – requests 2, 3, 4, 5, 6, and 62. (ECF No. 149-6.) CLF’s Final Revised Requests did not include analogs to requests 2, 3, 4, 5, or 62, nor does its current motion seek any relief with respect to them. (ECF Nos. 309-5, 326-1 at 1.) Thus, only one Final Revised Request is exclusively based on CLF’s operator liability claims – Request No. 6. That request sought production of “[a]ll

policies that establish, acknowledge, or limit Your authority to participate in, exercise control over, manage, direct, or conduct activities at the Terminal, including the Shell Manual of Authorities.” (ECF No. 309-5, at 1.) The Court concludes that pre-2017 documents responsive to Final Revised Request No. 6 are relevant to CLF’s operator liability claims. The Court disagrees with the Defendants’ contention that CLF pled no such claims for the 2011–2017 period; as CLF points out in its reply brief, it clearly alleged in paragraphs 260–63 of its Amended Complaint that “‘Shell,’ defined as all Defendants, reported violations of the Permit in 2011.” (ECF No. 330, at 7.) Moreover, there is no credible claim of undue burden, overbreadth, or lack of proportionality with respect to pre- 2017 documents responsive to Final Revised Request No. 6. The request is limited to the New

Haven Terminal, and if indeed the Defendants did not “manage, direct, or conduct activities” there until the Permit transferred in 2017 as they contend, there should be little if any burden in responding to the pre-2017 portion of the request. This leaves the relevance of pre-2017 documents to those requests based on CLF’s “best industry practice,” “disclosure,” “adaptation,” and other claims. The Court did not resolve the parties’ disputes over these issues in the Prior Ruling, because it declined to compel responses to the then-current versions of those requests on grounds of undue burden, overbreadth, and lack of proportionality. See Prior Ruling, at *20–22. Now that CLF has narrowed most of its requests to an acceptable breadth, these disputes need to be resolved. The Court concludes that the pre-2017 portions of these requests seek relevant discovery. Take, for example, Final Revised Request No. 23, which CLF premises on its “best industry practices” claim. (ECF No. 309-5, at 5; ECF No. 149-6, at 2.) That request seeks production of “[d]ocuments sufficient to show the design, construction, maintenance, or modification of . . .

berms, secondary containment areas, outfalls, stormwater drainage or pipes, storage tanks, and oil/water separator” at the New Haven Terminal or at other “yellow” or “red” coastal facilities. (ECF No. 309-5, at 5.) If the Defendants made changes to, say, berms and containment areas to protect facilities against rising sea levels prompted by climate change, documents showing those changes would be no less relevant to CLF’s “best industry practices” claim if they were prepared in 2014 than if they were prepared in 2018. Or take Final Revised Request No. 29, which asked for “[a]ll Clean Water Act permits, registration documents, permit applications, and technical filings for the Terminal,” and which CLF premised on its “disclosure” claims. (ECF No. 309-5, at 6; ECF No. 149-6, at 2.) If in 2011 the Defendants informed state regulators of some risk to the New Haven Terminal from severe weather and climate change, that would be relevant to CLF’s

disclosure claims respecting the 2017 permit. Similarly, pre-2017 documents responsive to requests like Final Revised Request No. 64 are relevant to CLF’s “adaptation” claims. That request sought documents sufficient to show how, if at all, the Defendants adapted the New Haven Terminal or other coastal “yellow” or “red” facilities “to address the cause(s) of the loss of containment at the Sewaren facility during Superstorm Sandy.” (ECF No. 309-5, at 13.) A 2016 document showing such an adaptation would be just as relevant as a 2018 document. In short, CLF has carried its burden to show how the pre-2017 component of its Final Revised Requests is relevant to at least one of its claims. Fifth, the parties dispute whether, in responding to the Final Revised Requests, the Defendants are obligated to review the information behind the redactions in the documents they already produced. (See ECF No. 326-1, at 21–23; ECF No. 325-1, at 36.) When they assembled their initial productions, the Defendants encountered documents that contained both (a)

information about the New Haven Terminal, the relevance of which they conceded; and (b) other information that they contended was irrelevant or non-responsive, including information about other facilities. They redacted the latter, and CLF challenged these redactions in its First Motion. See Prior Ruling, at *6. The Court agreed that “most authorities do not support” the practice of redacting allegedly irrelevant information from documents that also contain admittedly relevant material. Id. (citing, inter alia, Berney v. Apple, Inc., No. 3:20-cv-1379 (JAM) (RMS), 2021 WL 6334985, at *2 (D. Conn. May 27, 2021)). Yet because CLF had not objected until well after the Defendants “invested time and effort into redacting the later production installments,” the Court did not order them to un-redact their entire production. Id. To balance CLF’s information needs with the Defendant’s interest in avoiding an expense that could have been averted if CLF has

objected sooner, the Court allowed CLF to choose “up to fifty documents in which the redactions have rendered the unredacted information unintelligible or lacking in key context,” and it ordered the Defendants to produce those documents.3 Id. at *6 (citing In re State Street Bank & Tr. Co. Fixed Income Funds Inv. Litig., No. 08-cv-0333 (RJH) (DFE), 2009 WL 1026013, at *2 (S.D.N.Y. Apr. 8, 2009) as a precedent for this technique). In short, the Court chose a practical solution to a practical problem.

3 CLF says that the Defendants have produced only thirty-six of the fifty requested documents and, therefore, have not complied with this order. (ECF No. 326-1, at 22 n.7.) The Court will address that issue separately. The Defendants now seek to parlay this practical solution into a substantive rule that the information behind the redactions is off-limits for all discovery purposes (see ECF No. 325-1, at 36), but nothing in the Prior Ruling supports this. The Court denied relief to CLF on many of the requests at issue in the First Motion because they were “facially overbroad and disproportionate

to the needs of the case,” but it expressly authorized CLF to serve “more narrowly tailored and targeted discovery into the subject matter of that . . . request.” Prior Ruling, at *21, 24. Nothing in the Prior Ruling suggests that, if information behind a redaction proved responsive to one of these new, “narrowly tailored and targeted” requests, that information would not be discoverable. A hypothetical example will explain this point and guide the Defendants as they comply with the orders set forth in Section III below. In the original Request No. 24, CLF sought “[a]ll Documents related to Shell’s consideration of climate change in its design and engineering of terminals, refineries, and other infrastructure” – in other words, all infrastructure, anywhere in the world. (ECF No. 149-4, at 31.) The Defendants objected on grounds of overbreadth, undue burden, and lack of proportionality, but “[s]ubject to” those objections, they committed to produce

that subset of responsive documents “related to ‘Shell’s consideration of climate change in the design and engineering’ of the New Haven Terminal that are relevant to Plaintiff’s claims[.]” (Id. at 32.) The Court sustained the objections, and in response, CLF served a revised request that now inquires after only fifteen “yellow” or “red” facilities. (ECF No. 309-5, at 5; ECF No. 326-1, at 38.) Let us suppose that, in assembling their initial productions, the Defendants came across a document that contained information about fifty facilities including the New Haven Terminal. Under their view of what they could permissibly redact, they would have cloaked information about the other forty-nine facilities, leaving only the New Haven Terminal visible. Now let us further suppose that among those forty-nine facilities are some “yellow” and “red” facilities that CLF is inquiring about in its now-narrowed Final Revised Request. Having acceptably limited the scope of its request, CLF would be entitled to that information, and should not be barred from it merely because the Defendants redacted it in response to an earlier, impermissibly broad request. Nothing in the Prior Ruling suggested otherwise.

B. The Court’s Rulings on the Individual Requests for Production and CLF’s Other Requests for Relief With those issues addressed, the Court now turns to the individual, contested, Final Revised Requests for Production. The Court has carefully considered all the arguments raised in the parties’ papers and at oral argument, and it has carefully considered all the proportionality factors listed in Rule 26(b)(1). Having done so, and applying the conclusions reached in Section II.A above, the Court rules as follows: The following Final Revised Requests (a) seek relevant information; (b) have been materially narrowed since their prior iteration and, therefore, are not controlled by the Prior Ruling; (c) have not been shown to be unduly burdensome, overbroad, disproportionate to the needs of the case, or otherwise objectionable, except to the extent that they inquire after documents protected by an applicable privilege or by the work product doctrine; and (d) have a proper temporal scope: 6, 7, 12, 18, 20, 29, 31, 32, 33, 34, 36, 38, 39, 41, 42, 43, 44, 45, 46, 47, 49, 51, 53, 54, 63. The Defendants will be ordered to comply with these requests. Final Revised Request Nos. 9, 10, 23, 24, 25, 26, 27, 28, 48, and 64 are not limited to the New Haven Terminal. They also inquire about “coastal facilities that store or handle petroleum

products in bulk and that have a Risk Assessment Matrix (RAM) rating of yellow or red for category H-09.01 (weather conditions).” (E.g., ECF No. 309-5, at 1.) In their October 3, 2023 form, these requests evidently inquired about all such facilities, anywhere in the world. (See id.) By the time it filed its motion, however, CLF was willing to limit these requests to fifteen facilities. (ECF No. 328, at 1) (“CLF seeks discovery regarding 15 facilities”). Without this limitation, the Court would regard these requests as facially overbroad, unduly burdensome, and disproportionate to the needs of the case. See Prior Ruling, at *22 (declining to compel compliance with “requests for ‘all documents’ ‘related’ to the work of two entire business units, anywhere in the world . . .

[as] facially overbroad and disproportionate to the needs of the case”). With this limitation, however, these requests are not facially objectionable on those grounds, nor have the Bennett and Contreras Declarations shown them to impose undue burdens. The Court will therefore order the Defendants to comply with these requests, not as stated in their October 3, 2023 form, but rather as limited in CLF’s brief. Final Revised Request Nos. 11 and 40 have not been materially narrowed from the version at issue in the First Motion, and accordingly the Court adheres to its Prior Ruling that they were facially overbroad, unduly burdensome, and disproportionate to the needs of the case. CLF’s motion will be denied with respect to these requests. Final Revised Request No. 17 seeks “[a]ll Documents concerning” three categories of

information. The first is “Shell’s assessments of Climate Change Factors and associated future climate conditions, including the Global Sea Level Rise Review (2030-2050) . . . and assessments that include or encompass the Terminal within its geographic scope.” (ECF No. 309-5, at 4.) The second is “Shell’s design standards applicable to the Terminal” or other “yellow” and “red” facilities “that take into account Climate Change or are being revised to account for Climate Change.” (Id.) The third is “[a]ny adaptation plans . . . that have been integrated into Shell’s existing procedures and processes . . . that are applicable to the Terminal” or “yellow” and “red” facilities. (Id.) The Court concludes that the request is facially overbroad and unduly burdensome with respect to the first category of information, but not the second or third. CLF’s motion will therefore be granted in part and denied in part with respect to Final Revised Request No. 17. Finally, CLF asks the Court to order the Defendants to comply with its requests within fourteen days. (ECF No. 326-1, at 38-39.) Rule 37(d) of the Local Rules of Civil Procedure for

the District of Connecticut provides that “[u]nless a different time is set by the Court, compliance with discovery ordered by the Court shall be made within fourteen (14) days of the filing of the Court’s order.” Although the Court has rejected many of the Defendants’ undue burden arguments, it acknowledges that the orders set forth in Section III below will require the collection, review, and production of more than a few documents. Accordingly, the Court will exercise its discretion to set a different time for compliance. D. Conn. L. Civ. R. 37(d). III. CONCLUSION AND ORDER For the foregoing reasons, the Court orders as follows: A. The Defendants’ objections, other than privilege and work product, to Final Revised Requests for Production Nos. 6, 7, 12, 18, 20, 29, 31, 32, 33, 34, 36, 38, 39, 41, 42, 43,

44, 45, 46, 47, 49, 51, 53, 54, and 63 are overruled. The Defendants are ordered to comply with these Final Revised Requests by April 29, 2024. B. With respect to Final Revised Requests for Production Nos. 9, 10, 23, 24, 25, 26, 27, 28, 48, and 64, and sections (b) and (c) of Final Revised Request No. 17: a. The Defendants shall provide to CLF, by April 5, 2024, a list of all coastal facilities owned or operated by any Defendant, that store or handle petroleum in bulk and that have a Risk Assessment Matrix rating of yellow or red for category H-09.01 (weather conditions); b. CLF shall, by April 12, 2024, identify no more than fifteen such facilities; and c. The Defendants’ objections, other than privilege and work product, as to documents that are responsive to Final Revised Requests for Production Nos. 9, 10, 23, 24, 25, 26, 27, 28, 48, and 64, and sections (b) and (c) of Final Revised Request No. 17, as limited by the foregoing identification process, are

overruled. The Defendants are ordered to produce non-privileged, non-work- product documents by April 29, 2024. C. In complying with the orders set forth in Sections III.A and III.B above, the Defendants must review the information behind the redactions in their previous document productions. D. If the Defendants withhold any documents under a claim of privilege or work product protection, they must log each document on a privilege log that meets the requirements of D. Conn. L. Civ. R. 26(e). E. CLF’s motion is denied with respect to Final Revised Requests for Production Nos. 11 and 40, and with respect to section (a) of Final Revised Request No. 17.

CLF seeks an award of reasonable expenses incurred in connection with the motion pursuant to Fed. R. Civ. P. 26(g) and 37(a)(5)(A). (ECF No. 326-1, at 39.) CLF also points out one respect in which the Defendants did not comply with the Prior Ruling, and although it does not directly ask the Court for any specific relief on this point, it notes that “the Court at all times has authority to issue sanctions sua sponte over violations of its own discovery orders.” (Id. at 22 n.7) (citing In re 60 East 80th St. Equities, Inc., 218 F.3d 109, 117-18 (2d Cir. 2000)). The Court reserves decision on these two issues. The Court has carefully considered all the arguments and claims for relief raised in the parties’ briefs and at oral argument. Other than the two issues referenced in the preceding paragraph, upon which decision is reserved, any claims for relief not specifically addressed above are denied on the current record. This is not a recommended ruling. It is a ruling by a Magistrate Judge on “nondispositive motions . . . relating to discovery,” D. Conn. L. Civ. R. 72.1(C)(2), and as such it is reviewable

pursuant to the “clearly erroneous” statutory standard of review. See 28 U.S.C. § 636(b)(1)(A); Fed. R. Civ. P. 72(a); D. Conn. L. Civ. R. 72.2(b). It is an order of the Court unless reversed or modified upon timely objection under Local Rule 72.2(a).

/s/ Thomas O. Farrish Hon. Thomas O. Farrish United States Magistrate Judge