Sinclair Wyoming v. Infrassure

FILED United States Court of Appeals Tenth Circuit

PUBLISH August 17, 2020 Christopher M. Wolpert UNITED STATES COURT OF APPEALS Clerk of Court

TENTH CIRCUIT

SINCLAIR WYOMING REFINING COMPANY,

Plaintiff-Appellant/ Cross-Appellee, v. No. 19-8018 INFRASSURE, LTD,

Defendant-Appellee/ Cross-Appellant.

APPEAL FROM THE UNITED STATES DISTRICT COURT FOR THE DISTRICT OF WYOMING (D.C. NO. 2:15-CV-00194-NDF)

Marc J. Ayers and Alicia K. Margolis, Bradley Arant Boult Cummings, LLP (Marcy G. Glenn and JoAnna S. DeWald, Holland & Hart LLP, Denver, Colorado and Cheyenne, Wyoming, with them on the briefs), Birmingham, Alabama, and Jackson, Mississippi, for Plaintiff-Appellant/Cross-Appellee.

Guyon H. Knight, Quinn Emanuel Urquhart & Sullivan (Jane M. Byrne and David M. Cooper, Quinn Emanuel Urquhart & Sullivan, New York, New York, and Randall B. Reed, Long Reimer Winegar Beppler, Cheyenne, Wyoming, with him on the briefs), New York, New York, for Defendant-Appellee/Cross-Appellant.

Before TYMKOVICH, Chief Judge, BACHARACH, and CARSON, Circuit Judges.

TYMKOVICH, Chief Judge.

After a 2013 fire in its Wyoming refinery caused the Sinclair Wyoming Refining Company to restrict operations for several months, it filed a claim with its eighteen insurers, including Infrassure, Ltd., which collectively provided Sinclair coverage for business interruption losses under an all-risk insurance policy. In 2015, after twenty months of claim adjustment, Sinclair and the other seventeen insurers settled the claim. But Infrassure did not agree with the settlement value and eventually exercised its right under the policy to have Sinclair’s covered loss calculated by a panel of three appraisers. The panel valued the loss at $60,365,508, with Infrassure liable for $4,527,413.

Infrassure, still unsatisfied, sought to invalidate the award in district court, arguing that the appraisers relied improperly on the settlement amount rather than independently valuing the loss. The district court rejected this theory and confirmed the award, holding Infrassure failed to show any actionable misconduct on behalf of the appraisers. We agree the record reveals nothing warranting setting aside the appraisal award and therefore AFFIRM.

-2-

I. Background

This appeal arises out of an insurance dispute regarding a September 27, 2013 fire in Sinclair’s petroleum refinery. 1 The fire originated near a heater located within a #4HDS Unit—a piece of machinery that removes sulfur and nitrogen from gasoil, one step in the production of gasoline. The fire damaged the #4HDS Unit and other portions of Sinclair’s facility. As a result, the plant was forced to operate on a limited basis, resulting in lost business for Sinclair.

A. The Policy

At the time of the fire, Sinclair’s refinery was covered under an all-risk insurance policy. Sinclair’s parent companies solicited the Policy on the London insurance market, and eighteen insurers (collectively the Market) separately subscribed to varying portions of the $250 million limit. Infrassure, as one of the participating insurers, subscribed to cover 7.5%—meaning that it assumed several liability for 7.5% of any covered loss up to the Policy limit.

1 Originally this appeal was joined with case number 19-8017 and represented the cross-appeal in the combined matter. We decide here only 19- 8018—an appeal brought by Infrassure challenging the district court’s confirmation of the appraisal award determining the value of Sinclair’s covered business income and expense losses under its insurance policy. The associated appeal—19-8017—involves a question of law that was certified to the Wyoming Supreme Court. We neither decide nor address that appeal in this opinion.

-3-

The Policy covers property damage as well as business interruption losses. The parties do not dispute the amount of Sinclair’s covered property damage. Nor do they dispute that the fire caused Sinclair to lose business from its inability to operate as fully as it had before, generating some covered business interruption loss. The sole point of contention is the amount of that loss.

The Policy defines business interruption loss as “loss resulting from necessary interruption of business . . . caused by loss, damage, or destruction covered” by the Policy. Aple. App. at 127. 2 This constitutes the actual loss sustained by Sinclair, “consisting of the net profit which is . . . prevented from being earned and of all charges and expenses only to the extent that these must necessarily continue during the interruption of business.” Id.

This coverage is cabined in two ways. First, Sinclair undertook a duty to mitigate its business interruption loss to the extent “reasonable means” were available, including by resuming operations on a “complete or partial” basis. Id. Second, the Policy limits the time period—the “Period of Recovery”—during which business interruption losses were recoverable. It states the “length of time for which [a business interruption loss] may be claimed” runs from 75 days after

2 The parties have submitted three appendices. Sinclair, as the cross- appellee, has submitted two. We refer to these as appellee’s appendix and appellee’s supplemental appendix. Infrassure, as the cross-appellant, has submitted one. We refer to this as appellant’s supplemental appendix.

-4-the event giving rise to the loss and “shall not exceed such length of time as would be required with the exercise of due diligence and dispatch to rebuild, repair, or replace the property that has been destroyed or damaged.” Id. at 130.

In the event of a dispute over coverage, the Policy provides any party with the right to have the covered loss determined by a panel of three appraisers. Once a party demands an appraisal, the parties must each name a “competent and disinterested appraiser,” and those two in turn “select a competent and disinterested umpire.” Id. at 148. Under the Policy, the two party-appointed appraisers will meet and attempt to value the loss, with the umpire deciding any disputes.

B. The Claim-Adjustment Period

Shortly after the fire, Sinclair filed a claim under the Policy, assisted by experts from the firm AON. The Market retained independent loss adjusters to review the claim. The Market’s team included Dan McLain, the lead loss adjuster, Ennio Mastracci, a refinery expert, and Mike Clarke, an accounting expert. For approximately the next 20 months—from October 2013 through May 2015—the Market’s adjusters worked with Sinclair and its experts in an attempt to value the claim.

During this claim-adjustment period, Sinclair and the Market, along with their respective experts, disagreed over how best to value Sinclair’s business

-5-interruption loss. Although this analysis depended on numerous factors, two issues became particularly contentious: (1) the amount of time it should have taken Sinclair to repair the #4HDS Unit 3 and (2) the appropriate model to use to estimate Sinclair’s lost profits.

With respect to the repair timeline, the experts disagreed over what constituted a reasonable hypothetical restart date assuming the exercise of due diligence and dispatch in repairing the #4HDS Unit. On May 11, 2015, in one of the final reports on the issue, Mr. Mastracci stated that the appropriate theoretical restart date may have been as early as March 1, 2014—almost two months before the actual restart date of April 26, 2014. See Aplt. Supp. App. at 364–65. Mr. Mastracci arrived at this conclusion by starting with the actual restart date of April 26, 2014 and then subtracting time for numerous categories of work—including fireproofing, welding, repairing errors resulting from poor workmanship, and ordering new parts—that he reasoned were either not attributable to the fire or were not performed with sufficient alacrity or efficiency. Id. In this report, however, Mr. Mastracci noted certain issues required additional discussion with Sinclair. Id. at 365.

3 As Mr. McLain reiterated throughout the claim-adjustment period in “reports, meetings and conference calls” with the Market, “reaching an agreement with [Sinclair] on a reasonable timeline for the period of restoration w[as] the key factor” in resolving Sinclair’s claim. Aplt. Supp. App. at 397 (emphasis added).

-6-

On June 7, 2015—less than a month after his earlier report—Mr. Mastracci issued an updated report listing April 1, 2014 as an appropriate theoretical start date. Id. at 412. In justification for his adjusted timeline, Mr. Mastracci stated, “my comments have been revised to reflect the new information and clarifications provided by Sinclair during the meeting in Salt Lake City on June 2, 2015.” Id. According to Mr. Mastracci, in that meeting, “Sinclair explained that work on the #4HDS Unit undertaken and performed during the month of March 2014,”—work that Mr. Mastracci had earlier opined was unrelated to the fire—“was the result of physical damage sustained during the September 27, 2013 incident.” Id. As such, he concluded, that work “should be considered in preparing a theoretical restoration timeline.” 4 Id.

The Market adjusters and Sinclair also disagreed over what linear program model (LP Model) to use to calculate Sinclair’s lost profits. In this context, an LP Model is a computer program that describes the operations of a refinery, estimating what gasoline, diesel, and other products the facility would have produced in a but-for world where no fire had occurred. Comparing this

4 In his June 9, 2015 final report to the Market, Mr. McLain explained Sinclair’s acceptance of this theoretical start date by saying “[i]n brief, the Insured agreed that repair timeline issues associated with faulty workmanship involving heat exchanger 25EX-2877 (.9.5 days) and fireproofing activities (.16 days) needed to be removed from the loss-related period of restoration. The elimination of these activities ultimately changed the restoration timeline to reflect a startup date on or about 1 April 2015.” Aplt. Supp. App. at 398.

-7-theoretical output against the actual output of the refinery during the repair period, one can calculate Sinclair’s lost business.

In the wake of the fire, a Sinclair employee, Jon Christensen, prepared the November 2014 LP Model. But after testing the model and additional analysis by the Market’s expert, Mr. Christensen agreed to make certain modifications, resulting in the March 2015 LP model. As a result of this agreement, both AON, working on behalf of Sinclair, and the Market’s experts used the March 2015 LP model during the claim-adjustment period. 5

C. The Settlement

On June 2, 2015, approximately 20 months after the fire, four members of the Market—comprising 40 percent of the coverage—met with representatives from Sinclair in an attempt to settle Sinclair’s claim. During the meeting, both sides “presented their claim analysis and answered questions of the other [regarding] methodology of the claim submission, assumptions regarding performance of the business, hypothetical period of restoration vs. actual period of restoration, etc.” Aplt. Supp. App. at 131. At the request of the insurers, forensic accountants completed a revised claim assessment, using a March 31, 2014 theoretical restart date, that resulted in a calculation of $59,995,997. Id.

5 In the subsequent litigation and appraisal process, however, Sinclair’s experts and appointed appraiser advocated for using the November 2014 LP Model, which was more favorable to Sinclair.

-8-Peter Johnson, Sinclair’s president, then met privately with representatives from two insurers and agreed on a settlement figure of $60 million. 6

In his deposition, Mr. Johnson stated that he did not believe Sinclair had provided any “new information concerning timeline issues” in the settlement meeting. But, in response to questions regarding whether consensus had been reached on the appropriate repair timeline in the meeting, Mr. Johnson stated that although there was not “an explicit agreement reached on the repair time” his observation was that “[Mr. Mastracci] seemed satisfied that [Sinclair] had addressed his questions about the repair timeline.” Aplt. Supp. App. at 538. He elaborated that, in his view, “[Mr. Mastracci] conceded that [Sinclair] had answered his questions adequately on the timeline.” Id. at 539.

In the wake of the meeting, seventeen members of the Market agreed to the $60 million settlement amount. Infrassure was the sole holdout. In an attempt to gain further insight into how the settlement figure was reached, Infrassure reached out to the Market’s claim adjusters inquiring into, among other things, the repair timeline. Id. at 126 (requesting “the information provided by Sinclair to justify” the new theoretical restart date of April 1, 2014). Mr. Mastracci

6 In describing his efforts in the breakout session, Johnson relayed that he was attempting to reach a commercial settlement. He stated that he tried “to see if there was a middle ground where the claim could be settled talking purely about dollar amounts and not about principles of claim valuation.” Aplt. Supp. App. at 540.

-9-responded to these inquiries summarizing additional “information provided” by Sinclair employees and discussions that occurred during the settlement meeting. Id. at 124. But Infrassure remained unswayed, and continued to reject the settlement amount.

D. The Litigation and Appraisal

Following Infrassure’s refusal to settle, the matter went to court. 7 As relevant here, Sinclair filed its second amended complaint on June 9, 2017, alleging that its total business interruption loss constituted $99,467,307. 8 Infrassure, meanwhile, had at this point paid Sinclair only what it believed to be due under the Policy: $2,136,197.48—representing Infrassure’s share of an alleged $28,482,633 business interruption loss.

On June 29, 2017, Infrassure exercised its right to appraisal under the Policy. The parties notified the district court of their intention to follow the appraisal process, and the matter was stayed. Sinclair named Karl Killian to serve as its appraiser, and Infrassure selected Brad Ebel. Mr. Killian and Mr. Ebel then

7 The parties’ dispute involves a complicated procedural history. We relay here only the portion relevant to this appeal. 8 Sinclair took the position in litigation that because Infrassure failed to pay its portion of the settlement, it was unable to rely on the $60 million figure, and had to pay Sinclair’s actual business interruption loss, which Sinclair maintains exceeds that amount.

-10-collectively chose R. Dean Graves to serve as the umpire on the panel. Sinclair and Infrassure agreed this panel was acceptable.

The Policy describes the appraisal panel’s substantive task in a single paragraph, stating in pertinent part:

[A]t a reasonable time and place, the appraisers shall appraise the loss, stating separately the value at the time of loss and the amount of loss. If the appraisers fail to agree, they shall submit their differences to the umpire. An award in writing by any two shall determine the amount of loss and shall be paid . . . within 30 days thereafter.

Aple. App. at 148.

To supplement the Policy, Sinclair and Infrassure entered into a negotiated Stipulation Concerning Procedures for Appraisal (Stipulated Procedures). The Stipulated Procedures require an organizational meeting, address ex parte communications, and provide for additional document discovery, expert analysis, and depositions. The Stipulated Procedures also establish a limited record for the panel to consider, including evidence gathered during the 20-month claim-adjustment period—specifically, the reports and deposition testimony of the experts involved. The appraisal record also includes documents and testimony related to the settlement and the Market’s adjusters’ final report issued after the settlement.

With respect to the award itself, the Stipulated Procedures require “a written award setting out [the panel’s] key findings and, where necessary, the

-11-reasons for those findings.” Aplt. Supp. App. at 268. The Stipulated Procedures list six issues which, at a minimum, the panel was required to resolve:

a. The date on which the #4HDS Unit should have been rebuilt, repaired, or replaced following the September 27, 2013 explosion and fire, as defined in the Policy as Period of Recovery;

b. The covered business interruption loss incurred during the Period of Recovery proximately caused by the September 27, 2013 explosion and fire;

c. Whether the buildup of gasoil was caused by the September 27, 2013 fire and explosion, and the amount of covered loss for the Extended Period of Indemnity (as defined in the Policy), if any;

d. The covered business interruption loss incurred, if any, during the Extended Period of Indemnity (as defined in the Policy) proximately caused by the September 27, 2013 explosion and fire;

e. Any reductions to the amount of loss Sinclair could have achieved, if any, by selling or otherwise mitigating the buildup of gasoil; and

f. The total amount of covered loss. Id. at 268–69.

Consistent with the Policy, the party-appointed appraisers worked to narrow their differences with respect to these issues and jointly identified three questions for the umpire to assist in resolving: (1) “the appropriate time period for the time element loss;” (2) which LP Model should be applied and “whether further adjustments were needed;” and (3) how to quantify “the increase in coker gas oil

-12-(‘CGO’) inventory and sales of CGO during the loss period.” 9 Id. at 252–53. The answers to these questions would resolve the six issues identified in the Stipulated Procedures.

The panel met in person, communicated by telephone, and exchanged numerous emails regarding these questions. Each party-appointed appraiser also submitted a report to Mr. Graves presenting their positions. Perhaps unsurprisingly, the party-appointed appraisers remained far apart on the key issues. Mr. Ebel, Infrassure’s appointed appraiser, argued that the appropriate restart date was February 28, 2014 and that additional modifications were required to the March 2015 LP Model. Under his calculations, Sinclair suffered a business interruption loss of only $5,098,601. Meanwhile Mr. Killian, Sinclair’s appointed appraiser, used a restart date of April 26, 2014 and the November 2014 LP Model to arrive at a total business interruption loss of $78,724,351.

On September 14, 2018, over a year after Infrassure sought appraisal, the panel issued its decision over Mr. Ebel’s dissent. The award determined April 16, 2014 was an appropriate theoretical restart date for the #4HDS Unit and that the

9 During the outage of the #4HDS Unit, CGO accumulated that could not be processed as it would in the normal course of operations. The parties in the appraisal process contested whether this excess CGO inventory caused a loss to Sinclair and, if so, the amount of that loss. The appraisal award determined Sinclair suffered a $10,648,885 loss with respect to CGO. On appeal, Infrassure does not make any arguments specifically with respect to this portion of the appraisal award. Accordingly, we do not address it further.

-13-March 2015 LP model, without additional adjustments, should be used to calculate Sinclair’s lost profits. These decisions led to calculating Sinclair’s total business interruption loss as $60,365,508—of which Infrassure’s share constituted $4,527,413.

Mr. Ebel dissented from the panel’s decision. Although he recognized that “the umpire . . . conducted the panel interactions in a very professional and open manner,” he stated “in my opinion, it has become apparent that the umpire was not disinterested and actually had a strong bias in favour of the agreed settlement.” Aplt. Supp. App. at 264. According to Mr. Ebel, “[o]n several issues, Mr. Graves expressed his opinion that, given that 22 qualified and experienced individuals had dealt with an issue, how could their agreement not be correct.” Id. 10

Back in district court, Sinclair moved to confirm the appraisal award, while Infrassure cross-moved to vacate the award. The district court, applying New York law, held that the appraisers complied with the Policy and Stipulated

10 In both his dissent from the appraisal award and his subsequent declaration, Mr. Ebel expands partially on only one alleged example of this. In response to Mr. Ebel’s position that the March 2015 LP Model “had serious flaws which needed to be corrected,” Mr. Graves allegedly advised the panel that Mr. Ebel had “made a compelling argument that an adjustment was warranted,” yet, “[i]n spite of there being no substantive response to refute this ‘compelling argument,’ Mr. Graves refused to give any consideration to it.” Aplt. Supp. App. at 544–45.

-14-Procedures and that the record failed to show any lack of good faith, fraud or bias, failure to consider the evidence before the panel, gross mistake, or lack of substantial thoroughness. Accordingly, the district court confirmed the appraisal award.

II. Analysis

In considering the confirmation of an appraisal award, we review the district court’s legal conclusions de novo and its findings of fact for clear error. Bowen v. Amoco Pipeline Co., 254 F.3d 925, 931 (10th Cir. 2001); see also Salt Lake Tribune Publ’g Co. v. Mgmt. Planning, Inc., 454 F.3d 1128, 1133 (10th Cir. 2006).

A. Legal Standard

The parties agree that under the Policy’s choice-of-law provision, New York law governs the appraisal process. Under New York law, an appraisal “shall proceed pursuant to the terms of the applicable appraisal clause of the insurance policy.” N.Y. Ins. Law § 3408(c). 11 Appraisal awards are presumed valid where the panel “substantially complied with the terms of the submission,” Gansevoort Holding Corp. v. Palatine Ins. Co., 11 Misc.2d 518, 522 (N.Y. Sup. Ct. 1957), and there is no showing of “fraud, bias, or bad faith.” See Zarour v. Pac. Indem.

11 Here the applicable contractual terms include both the Policy’s appraisal clause and the Stipulated Procedures.

-15-Co., No. 15-CV-2663 (JSR), 2017 WL 946332, at *1–2 (S.D.N.Y. Feb. 22, 2017). Unless the parties contract for a different arrangement, “[a]ppraisers are not limited to a single method of valuation.” Olympia & York 2 Broadway Co. v. Produce Exch. Realty Tr., 93 A.D.2d 465, 468 (N.Y. App. Div. 1983). Instead, in the general course, appraisers have “wide discretion as to methods of procedure and sources of information.” Grosz v. Serge Sabarsky, Inc., 24 A.D.3d 264, 266 (N.Y. App. Div. 2005).

New York courts recognize “appraisal is not an exact science.” Id. (“All that we may be concerned with is whether the appraisers acted honestly and in good faith in the exercise of their wide discretion.” (quoting Rice v. Ritz Assoc., 88 A.D.2d 513, 514 (N.Y. App. Div. 1982))). “[T]he determination of an appraiser is to be upheld as long as the appraiser proceeds in good faith and without bias or fraud.” Olympia, 93 A.D.2d at 468.

Appraisals, especially those dealing with business losses, often “involve[] the exercise of judgment and discretion in weighing competing arguments regarding causation and loss.” Amerex Grp., Inc. v. Lexington Ins. Co., 678 F.3d 193, 206 (2d Cir. 2012). These “factually laborious” tasks may generate controversial outcomes, but the fact that “an appraisal panel exercises judgment or produces a controversial result . . . does not turn factual disputes regarding damages into legal disputes” meriting court intervention. Id.

-16-

Absent a showing of fraud, bias, bad faith, misconduct, or a failure on behalf of the appraisers to abide by the governing contract, courts will not set aside an appraisal award. Olympia, 93 A.D.2d at 468. As we have said,

The submission of a disagreement as to the amount of a loss to disinterested appraisers is a lawful and commendable method of determining such a controversy; the awards of such appraisers are presumptively correct, and should not be set aside except upon clear and convincing evidence of fraud, gross mistake, misconduct of the appraisers, or their failure to perform the duties committed to them by the agreement of submission.

Aetna Ins. Co. v. Murray, 66 F.2d 289 (10th Cir. 1933).

To do otherwise would rob the parties of the opportunity to have their dispute resolved according to their agreement, including the ability to have highly specialized calculations made by technical experts of the parties’ choosing as opposed to unfamiliar and less-experienced members of the judiciary. See Gansevoort, 170 N.Y.S.2d at 174 (holding appraisal award “should not be denied effect in the absence of clear and strong proof amounting to fraud, bad faith or misconduct”).

B. The Appraisal Award

Infrassure seeks to invalidate the appraisal award on two bases. First, it argues the panel failed to follow the Policy’s mandate to “appraise the loss.” 2d Cx-App. Br. at 47. Second, Infrassure contends the appraisal award “committed a

-17-gross mistake and evidenced a lack of substantial thoroughness by relying on alleged facts that were unquestionably false.” Id. at 55.

1. Failure to Follow the Policy

Infrassure’s primary allegation of misconduct is that the panel “reverseengineered” its conclusions to reflect the settlement Sinclair reached with the other members of the Market rather than independently appraising Sinclair’s loss as the Policy required. 2d Cx-App. Br. at 48–49. Underlying this argument is the general suggestion that any consideration of the settlement is improper. Accordingly, we address this first before turning to Infrassure’s specific grievances with respect to the panel’s conduct.

We agree that had the panel “started and finished with the commercial settlement,” this would have run afoul of its obligation to appraise the loss. But, at least in the context of this case, mention of the settlement, or, critically, consideration of the expert analysis underpinning that agreement, is not inconsistent with the appraisers’ mandate. Grosz, 24 A.D.3d at 266 (recognizing that appraisers have “wide discretion as to methods of procedure and sources of information”); see also Amerex Grp., 678 F.3d at 205–06 (noting appraisers may be aided by expert analysis in the resolution of factual issues regarding lost business income).

-18-

Not only was the expert analysis underpinning the settlement useful in valuing the claim, 12 but this analysis, including the adjusters’ final report containing information regarding the settlement, was explicitly included in the appraisal record for the panel to consider in reaching its conclusions. Thus, far from representing an infringement of the panel’s duty, consideration of the settlement-related analysis was explicitly envisioned by the parties. Indeed, it may have been a dereliction of the panel’s duty to ignore such evidence. See Gervant, 118 N.E. at 575–76 (upholding lower court’s decision vacating an appraisal award where the umpire refused to consider relevant evidence).

Infrassure relies on St. Paul Fire & Marine Insurance Co. v. Eldracher, 33 F.2d 675 (8th Cir. 1929), and Aetna, 66 F.2d 289, but neither are on point. In Eldracher, the Eighth Circuit sustained the invalidation of an appraisal award where the umpire and one appraiser agreed to abandon their duty to assess the value of the property, instead opting to put forward an award that reflected the expected settlement value of the case. 33 F.2d at 675. In Aetna, we upheld the invalidation of an award that failed to assess the loss, and instead ascertained the defendant’s settlement offer and raised the amount $2,000. 66 F.2d at 291.

12 As Mr. Killian explained, the “professionals who participated in the adjustment process” were a “‘who’s who’ in the industry.” Aple. Supp. App. at 3.

-19-

These decisions confirm the principle that appraisers must abide by their contractual duty to appraise a loss and may not substitute wholesale a separate figure. But neither extends as far as Infrassure suggests. That is, neither case precludes consideration of a settlement figure or settlement-related analysis when such information may assist in appraising a loss and was explicitly included in the appraisal record.

Nonetheless, Infrassure points to certain areas of the record as demonstrating that the panel improperly relied on the settlement. Specifically, it references the face of the appraisal award, emails amongst the panel, and Mr. Ebel’s dissenting opinion. But none of these sources demonstrate the type of exclusive reliance on the settlement that would warrant vacating the award. To the contrary, the record taken as a whole shows the appraisers followed the Policy and Stipulated Procedures in assessing Sinclair’s covered loss.

With respect to the appraisal award, Infrassure contends that a failure to discuss “numerous subsidiary factual issues that go to the proper theoretical” restart date for the #4HDS Unit invalidates the award. 2d Cx-App. Br. at 50. These subsidiary issues include “how many welders should have been used on the repair project, whether Plaintiff unreasonably delayed ordering custom piping, and whether the repairs were delayed by Plaintiff’s choice to perform other upgrades” during the repair process. Id. But the Stipulated Procedures only

-20-required the panel to include its “key findings and, where necessary, the reasons for those findings” in the award. Aplt. Supp. App. at 268. This permissive standard gave the panel discretion to exclude explicit findings on the discrete issues Infrassure raises.

Moreover, as the district court found and Infrassure concedes, the appraisers “thoroughly discussed” these issues in their deliberations. Aplt. Supp. App. at 606; see also 4th Cx-App. Br. at 7 (conceding that the panel “certainly debated the issues”). 13 Infrassure protests their omission from the award on the hypothesis that because they were not mentioned explicitly, the panel must have failed to consider them. See 4th Cx-App. Br. at 7. But we will not overturn a district court’s confirmation of an appraisal award on the basis of speculation that issues that were admittedly “thoroughly discussed” amongst the appraisers played no role in the panel’s final decision. To do so would rewrite the parties’ contractual agreement, which mandates nothing comparable to the detailed explanation Infrassure now seeks to read into it. Such a decision would also run counter to the presumption of validity that appraisal awards possess under New York law. See Gansevoort, 170 N.Y.S.2d at 174.

13 The record amply supports the district court’s factual finding in this regard. See Aplt. Supp. App. at 570, 580, 590, 592.

-21-

Nor does the face of the appraisal award demonstrate an exclusive reliance on the settlement. With respect to the repair timeline, the award analyzes various hypothetical restart dates that were proposed by experts during the claim-adjustment period and appraisal process. These spanned dates from February 8, 2014 to July 25, 2014—both well before and after the April 16, 2014 date ultimately selected by the panel. In its analysis, the award provides conclusions as to why particular dates either were or were not appropriate based on numerous references to the expert reports and testimony in the appraisal record.

Ignoring this expanded analysis, Infrassure focuses on a single sentence in the award as evidence of exclusive reliance on the settlement figure. In discussing the April 1 and April 26 dates, the award states in part,

Consistent with the facts i) that April 1st was a negotiated settlement date, ii) that April 26th is the undisputed restart date, iii) the insurers noted that a date later than April 1st might be appropriate for subrogation purposes, and iv) Sinclair agreed to shorten the actual restart date by ten days, the Appraisal Panel concluded that April 16, 2014 is a reasonable restart date assuming due diligence and dispatch.

Aplt. Supp. App. at 257.

Infrassure argues this shows the panel “relied on just four facts” in establishing the repair timeline, all of which it contends are irrelevant or fatally dependent on the commercial settlement figure. 2d Cx-App. Br. at 50. But this alleged reliance is belied by the award’s language—stating that its decision was merely “consistent” with the four facts discussed—and the panel’s detailed

-22-discussion of other potential theoretical restart dates. See Aplt. Supp. App. at 255–57.

Even considering the four facts Infrassure raises, the record reveals that they are not so divorced from principles of claim valuation as Infrassure suggests. For example, the first fact—that April 1st was a negotiated settlement date—was arguably arrived at using principles of claim valuation, and not the mere product of commercial horse-trading. Indeed, as the award states, use of the April 1 date as the theoretical restoration date was “consistent with Mr. Mastracci’s June 7, 2015 report”—part of the expert analysis retained by the Market, including Infrassure, to value Sinclair’s claim. Id. at 256–57. 14

Nor do the statements of Infrassure’s party-appointed appraiser demonstrate exclusive reliance on the settlement. The record contains only a single specific grievance regarding the umpire’s alleged commitment to the settlement figure. Mr. Ebel states that in response to his position that the March 2015 LP Model “had serious flaws which needed to be corrected,” Mr. Graves advised the panel that Mr. Ebel had “made a compelling argument that an adjustment was warranted,” yet, “[i]n spite of there being no substantive response to refute this

14 Infrassure contends that Mr. Mastracci’s June 7, 2015 report is unreliable as it is itself a product of commercial negotiations and not genuine claim valuation analysis. As discussed in more depth infra, we find this allegation unsupported by the record.

-23-‘compelling argument,’ Mr. Graves refused to give any consideration to it.” Aplt. Supp. App. at 544–45.

This allegation falls short of the type of “clear and strong proof amounting to fraud, bad faith or misconduct” required to set aside an appraisal award. Gansevoort, 170 N.Y.S.2d at 174. Not every compelling argument necessarily carries the day, and the fact that Mr. Ebel’s argument failed to persuade the umpire is evidence only that the panel had to exercise judgment “in weighing competing arguments regarding causation and loss.” Amerex Grp., 678 F.3d at 206. It shows no legally cognizable misconduct. See id. The umpire’s decision to adhere to the March 2015 LP Model was supported by the award’s statement that the panel had “not seen any engineering evidence that would support further adjustment” to that model and the fact that the only two LP experts involved in adjusting the claim agreed on the model as a compromise. Aplt. Supp. App. at 258. Infrassure fails to show that this was an unreasonable exercise of judgment and we will accordingly not second-guess the work of the appraisers.

The email communications amongst the appraisers similarly fail to reveal any misconduct. As the district court found, each appraiser had “every opportunity to argue their positions,” had “significant” communication with the umpire, and discussed the substantive issues underlying both the repair timeline and the appropriate LP Model. Aplt. Supp. App. at 606. These observations are

-24-amply supported by the record, which shows the panel engaged in a thorough discussion of how best to value the claim. The fact that certain email communications challenged each party-appointed appraiser’s arguments by referencing the settlement does nothing to change this in light of the other evidence in the record and the presumption of validity the award is entitled to. See Gansevoort, 170 N.Y.S.2d at 174.

In sum, we agree with the district court that the “record undermines Infrassure’s allegation that the umpire simply accepted the commercial settlement.” Aplt. Supp. App. at 606. Taken as a whole, the appraisal award and communications amongst the panel reveal no lack of good faith, fraud or bias, or failure to comply with the terms of the Policy and Stipulated Procedures. Accordingly, the award cannot be set aside for the panel’s alleged failure to appraise the loss.

2. Gross Mistake and Lack of Substantial Thoroughness

Infrassure also argues the appraisal award should be vacated because it was based on “gross mistake or lack of substantial thoroughness.” 2d Cx-App. Br. at 47. As a preliminary matter, we note that Infrassure cites no authority articulating “gross mistake” or “lack of substantial thoroughness” as a basis for vacating an appraisal award under New York law. Infrassure points to Salt Lake Tribune, but there we held only that, under New Jersey law, a court may “review

-25-the appraisal to determine whether [the appraisers] committed a mistake of law, failed to consider relevant evidence, or exceeded its authority.” 454 F.3d at 1136.

Infrassure also cites Gervant, which applies New York law. But again Infrassure’s articulation of the standard is nowhere to be found. In Gervant, the “principal issue” was whether an appraisal award could be set aside where the umpire and a party-appointed appraiser considered only certain evidence and “flatly refused to consider other pertinent evidence.” 306 N.Y. at 396. The court held that such an arbitrary refusal to consider evidence that was directly presented to the appraisers justified vacating the award. Id. at 399 (“The right of a party to have appraisers receive all pertinent evidence offered is a fundamental procedural right to which plaintiff was entitled, and its denial . . . is sufficient . . . to set aside the award.”). Although one may characterize the refusal to consider evidence a “gross mistake” or evidence of a “lack of substantial thoroughness,” Infrassure has not cited any authority that has done so in applying New York law. In the absence of such authority, we decline the invitation to expand the law by elevating these terms to standards mandating the overturning of an appraisal award. This is the province of New York lawmakers into which we will not intrude.

But even to the extent such standards did apply, they would not merit vacating the award in this instance. Infrassure’s primary allegation is that the

-26-panel erred by relying on the updated timeline in Mr. Mastracci’s June 7, 2015 report, which states that he revised his analysis of the timeline based on “new information” presented at the June 2, 2015 settlement meeting. See 2d Cx-App. Br. at 55–56. Infrassure claims no new information was presented at this meeting, and Mr. Mastracci’s statement to the contrary is “unquestionably false.” Id. at 55. The truth, Infrassure claims, is revealed by Johnson’s deposition and emails between Mr. Mastracci and Mr. McLain. These allegedly show that the parties simply reached a commercial settlement and Mr. Mastracci’s “vague reference to ‘new information’” was inserted as a “fig leaf” to cast the agreement as a result of genuine value-based claim adjustment. Id. at 56.

Like the district court, we find this argument amounts to an oversimplification of the record and lacks merit. See Aplt. Supp. App. at 608. Taken as a whole, the record contains ample evidence that Mr. Mastracci’s updated timeline was the product of legitimate analysis and not, as Infrassure alleges, concocted solely to back-fill a commercial settlement. Mr. Mastracci’s June 7 report states that his conclusions were adjusted to reflect “new information and clarifications provided by Sinclair during” the June 2 settlement meeting. Id. at 412. He elaborates that during the meeting “Sinclair explained that work on the #4HDS Unit undertaken and performed during the month of March 2014, was the result of physical damage sustained during the September 27, 2013 incident

-27-and, as such, the effects of which should be considered in preparing a theoretical restoration timeline.” Id. This explanation is in part what Mastracci references in later saying that he revised his theoretical restart date based on “new information.” Id.

As evidenced by the disparate expert opinions in the record, valuing Sinclair’s business interruption loss is not an exact science. Amerex Grp., 678 F.3d at 206 (noting that appraising business losses “involve[s] the exercise of judgment and discretion in weighing competing arguments regarding causation and loss”). Considering new explanations and revising calculations based on continued discussions is not improper for an expert in Mr. Mastracci’s position. Cf. Olympia, 93 A.D.2d at 472 (rejecting the notion that appraisers must “dogmatically . . . stick to their initial calculations” in a manner that would routinely result in deadlock). And the record suggests that is what occurred here. See Aplt. Supp. App. at 131 (noting that both the insurers and the insured “presented their claim analysis and answered questions of the other [regarding] methodology of the claim submission, assumptions regarding performance of the business, hypothetical period of restoration vs. actual period of restoration” during the settlement meeting); id. at 538 (stating that “[Mr. Mastracci] seemed satisfied that [Sinclair] had addressed his questions about the repair timeline”

-28-during the settlement meeting). Accordingly, the panel’s reliance on Mr. Mastracci’s updated timeline fails to warrant invalidation of the award. 15

III. Conclusion

For the reasons stated above, we AFFIRM the district court’s confirmation of the appraisal award.

15 Infrassure, without citing to any legal authority for support, also points to other alleged factual errors in the award’s assessment of the appropriate theoretical restart date as constituting gross mistake. None of these alleged errors approaches the type of “clear and strong proof” that the appraisers committed fraud or engaged in bad faith or misconduct. Gansevoort, 170 N.Y.S.2d at 174. Accordingly, we decline to vacate the award on these bases. As discussed supra, the record reveals the panel engaged thoroughly in assessing the evidence with respect to the appropriate theoretical timeline. It is not this court’s task to re- weigh that evidence. See Gervant, 118 N.E. at 575–76.

-29-