Gerald Lebovits, J.

O'Melveny & Myers LLP, New York, NY (Jonathan Rosenberg, Anton Metlitsky, and Leah Godesky of counsel), and Los Angeles, CA (Daniel Petrocelli and Craig P. Bloom of counsel), for plaintiff.

Covington & Burling LLP, Washington, D.C. (Benjamin Razi, Michael Lechliter, and Brian E. Foster of counsel), and San Francisco, CA (Gretchen Hoff Varner and Ryan Buschell of counsel), for defendant Brooklyn Union Gas Company.

Gerald Lebovits, J.

This decision arises from the first trial in two-decades-old litigation between Brooklyn Union Gas Company and Century Indemnity Company. The underlying dispute in the litigation concerns whether and to what extent Century, as an excess insurer to Brooklyn Union from 1941 to 1969, must cover the vast costs that Brooklyn Union is incurring through its contributions to the environmental remediation of the sites of a number of former manufactured-gas plants (MGPs) in New York City and of the Gowanus Canal in Brooklyn.

This trial, potentially the first of several in this litigation, concerns remediation of three MGP sites in Brooklyn and the associated remediation of the nearby Gowanus Canal. After four weeks of trial, a jury rendered a special verdict in late-April 2022. During the course of trial, the court made a number of rulings about the form and content of the jury charge and verdict sheet that the court then lacked time to explain fully. The court informed the parties that it would give them more detailed written explanations of these rulings once time permitted. This decision is intended to provide those fuller explanations.

DISCUSSION

I. Whether the Environmental Damage at Issue was "Accidental" or "Expected or Intended"

It is undisputed that Brooklyn Union can be covered under the Century policies only if Brooklyn Union proves that the investigation and remediation costs for which it seeks coverage were incurred to address environmental damage that was accidental, rather than expected or intended by Brooklyn Union. The parties had numerous disagreements at trial, however, about what parameters the court should give the jury for answering the accidental versus expected-or-intended question.

A. The Point in Time for Assessing Whether Damage was "Accidental" or "Expected or Intended"

The six Century policies underlying this litigation, taken together, ran from 1941 to 1969. The parties agree that the Century policies are triggered as long as property damage happened during their various policy periods. It is not material whether the acts causing that damage had also happened during the policy periods or had instead happened before the policies were purchased or incepted. But the parties dispute whether the accidental versus expected-or-intended inquiry turns on what Brooklyn Union knew or did not know at the time of the acts causing the damage (Brooklyn Union), or instead what it knew or did not know at the start date of each policy (Century). This court agrees with Brooklyn Union.

It is undisputed that for property damage to be expected or intended, it is not enough for the party whose acts caused the damage to have intended to commit the damage-causing acts. The party must also have intended to cause the resulting damage. The expected-or-intended inquiry necessarily focuses on the time of the act itself—the point in time when the party chooses to act or to refrain from acting, based on its knowledge of the consequences of its acts.

As this court explained in deciding a pretrial motion in limine, New York precedent in this context holds that where a party takes a "calculated risk" that damage might (but might not) result from its deliberate acts, any damage that does eventuate was not expected or intended. (See Continental Cas. Co. v Rapid-Am. Corp. , 80 NY2d 640, 649 [1993], citing Allstate Ins. Co. v Zuk , 78 NY2d 41, 46 [1991] ; Union Carbide Corp. v Affiliated FM Ins. Co. , 101 AD3d 434, 435 [1st Dept 2012].) Similarly, if it is merely foreseeable that an act could , through several intervening causal steps, lead to harms down the line, that damage is not expected or intended. ( Salimbene v Merchants Mut. Ins. Co. , 217 AD2d 991, 994 [4th Dept 1995].) Where, on the other hand, the party acted knowing that damage would result directly from the party's acts (see id. ), or where the damage is inherent in the nature of the act itself (see Pistolesi v Nationwide Mut. Fire Ins. Co. , 223 AD2d 94, 97 [3d Dept 1996] ), the act is expected or intended under New York law. Each of these different scenarios necessarily considers whether damage was accidental or expected-or-intended from the actor's standpoint when it acted.

Thus, "a disclaimer of coverage based on ‘expected or intended’ injury requires inquiry into whether, at the time of the acts causing the injury , the insured expected or intended the injury, an inquiry that generally asks merely whether the injury was accidental." ( Union Carbide Corp. v Affiliated FM Ins. Co. , 2010 NY Slip Op 51656[U], at *3 [Sup Ct, NY County, Sept. 9, 2010] [emphasis added; internal quotation marks omitted].)

Century's contrary argument relies on the potential for inequity should a party gain greater knowledge of the harmful consequences of its acts between the time it acted and the time its insurance coverage begins. But that problem is most naturally dealt with through a known-loss-type limit on coverage, not through distorting the expected-or-intended inquiry.

Century relies on a suggestion by the U.S. Court of Appeals for the Second Circuit, applying New York law, that the expected-or-intended inquiry is an appropriate way to protect insurers against parties who purchase coverage despite knowing that future harm is likely. (See National Union Fire Ins. of Pittsburgh, PA. v Stroh Cos., Inc. , 265 F3d 97, 108 [2d Cir. 2001].) The Second Circuit's decision in Stroh Companies does not change this court's conclusion.

As an initial matter, it is not clear from the passage in Stroh Companies on which Century relies that the Second Circuit intended there to opine on the point at which the policyholder's knowledge and expectations should be assessed. Regardless, the passage cited by Century appears to rest on a misreading by the Second Circuit of the "expected or intended" doctrine as it exists in New York.

The part of Stroh Companies on which Century relies addressed the insurer's argument that it could deny coverage because the policyholder knew of a risk of future loss at the time it purchased the policy. (See id. at 107-108.) The Second Circuit rejected this argument, adhering to the court's prior holding in City of Johnstown, NY v Bankers Standard Ins. Co. (877 F2d 1146, 1152-1153 [2d Cir. 1989] ) that New York law does not recognize a "known risk" defense to coverage. The Stroh Companies court then asserted that the lack of a known-risk defense "does not leave insurers unprotected under New York law when insuring such risks." ( 265 F3d at 108 ). Instead, under City of Johnstown "such protection" lies elsewhere, in "narrower and better-established doctrines, such as rules barring recover for damages that are ‘expected’ or ‘intended’ by the insured." (Id. [internal quotation marks omitted].)

As City of Johnstown itself makes clear, though, future harm will qualify as "expected or intended" only if it is certain or substantially certain to result at the time of the harm-producing act. (See 877 F2d at 1150-1151.) Indeed, if a risk of harm is sufficiently likely to be realized as of the start of the policy period that it would be "expected or intended," then that harm it is not a "known risk"; rather, it is a "known-but-not-yet-realized loss," which Stroh Companies addresses separately from "known risks." On the flip side, if sufficient uncertainty remains about future damage that it is a "known risk," rather than a "known loss" or "inevitable loss," that damage could not be "expected or intended." (See id. at 1150 [noting principle that damages are not expected or intended if the insured, upon being warned of the possibility of damages, "decided to take a calculated risk and proceed as before"].)

Given these shortcomings in Stroh Companies ’s analysis of the expected-or-intended limit on the scope of coverage, this court finds it unpersuasive on the point for which Century seeks to use it.

Century also relies on the terms of a verdict sheet in a 2013 jury trial held in the decades-long federal litigation over coverage for environmental harm between the Olin Company and its insurers. That verdict sheet, which was used by the U.S. District Court for the Southern District of New York and approved on appeal by the Second Circuit, asked the jury whether, "[b]y the time of the policy period of 1970 ... Olin expect[ed] or intend[ed] the property damage that it was obligated to remediate." ( Olin Corp. v OneBeacon Am. Ins. Co. , 864 F3d 130, 146 [2d Cir. 2017].) Century contends that the Second Circuit's approval of this question on the Olin verdict sheet evinces the court's recognition that the proper point in time for the expected-or-intended inquiry is the time of the policy period and not the time of the damage-producing act. This court disagrees.

The aspect of the Olin litigation dealt with in the decisions on which Century relies concerned sites in which industrial operations began before, and continued after, the 1970 policy period. As a result, the insurer's challenge to this jury question was not, as here, that the jury's expected-or-intended inquiry was focused on too late a point in time. Instead, the insurer argued that because damage continued after the policy period, the question for the jury should have encompassed an inquiry into whether the post-period damage was expected or intended. That is, the insurer was arguing that the question on the verdict sheet, as phrased, focused on too early a point in time. (See Olin , 864 F3d at 146 [describing this argument].) The Second Circuit's rejection of that challenge under the particular facts of that litigation sheds little light on whether it would accept Century's converse challenge to the verdict sheet here—let alone on whether accepting that challenge would be the correct course.

B. Whether the Presence of a "Substantial Probability" of Property Damage Suffices to Make that Damage Expected or Intended

One of the topics on which the parties disagreed repeatedly during the court's preparation of the jury charge for this trial was how to instruct the jury on whether Brooklyn Union had proven that the environmental damage at issue was accidental (rather than expected-or-intended). Century proposed that the last sentence of the accidental-damage instruction should state that "if the operator of the plant was aware of a substantial probability of damage as a result of the manner in which the plant was operated, you may find that the damage was intentionally caused and was therefore not accidental." (See NYSCEF No. 1169 at 14.) This court does not agree that the accidental-damage instruction should include a "substantial probability of damage" sentence.

This court concluded prior to trial that "substantial probability of damage" does not provide the correct test under New York law for determining whether environmental damage was accidental, or instead expected or intended. (See Century Indem. Co. v Brooklyn Union Gas Co. , 2022 NY Slip Op 50192[U], at *13-*15 [Sup Ct, NY County Mar. 14, 2022].) And this court's conclusion on this point was further confirmed by the decision of the Appellate Division, First Department, in Certain Underwriters at Lloyd's London v NL Indus., Inc. (2022 NY Slip Op 02056 [1st Dept Mar. 24, 2022] ), issued during jury selection in this trial. In the course of argument over the wording of the accidental-damage instruction, Century contended that this court should not give weight to the NL Industries decision because it addressed a legally distinct issue. This court disagrees.

NL Industries is a coverage action arising from a lawsuit in California against NL for formerly manufacturing and promoting a brand of lead paint. In the California action, the California courts found that NL had caused a public health hazard/nuisance, and ordered NL and other defendants to contribute to a fund for lead-paint abatement. ( See 2022 NY Slip Op 02056, at *1.) NL's insurers then brought a coverage action in Supreme Court, New York County. The insurers contended among other things that the costs of contributing to the abatement fund were not covered because the damage being abated was expected or intended rather than accidental. (See id. ) The motion court (Andrea Masley, J.) denied the insurers’ motion for summary judgment. (See Certain Underwriters at Lloyd's, London v NL Indus., Inc. , 2020 NY Slip Op 34331[U], at *10-*14 [Sup Ct, NY County Dec. 29, 2020].) On the insurers’ appeal, the First Department affirmed. (See 2022 NY Slip Op 02056.) The Court held that the trial and appellate decisions in the California action determined that NL's acts posed a serious or significant risk of harm that formed a public health hazard, and that this determination fell short of a "clear finding that NL either expected or intended to harm any person or property." (Id. at *1.)

On first considering the First Department's NL Industries decision when it was issued, this court took it to be consistent with this court's prior "substantial probability" conclusion in its motion in limine ruling. (See Century Indem. , 2022 NY Slip Op 50192[U], at *13-*15.) During briefing and argument on the jury charge, however, Century advanced a different interpretation of the holding in the case. On Century's reading, the First Department did not hold that "conduct would be accidental even if it were undertaken with knowledge of a serious risk of harm to a particular person or particular property "—only that "knowledge of a serious risk of harm to the ‘public health’ generally does not represent a clear finding of knowledge of a serious risk of harm to any particular ‘person or property.’ " (NYSCEF No. 1172 at 3 n 1 [emphases in original].) Century's argument is creative; but this court is not persuaded it is correct.

The motion court's analysis in NL Industries of the expected-or-intended issue did not give weight to the difference between a finding that NL Industries's conduct posed a risk of harm to public health and a finding that the conduct posed a risk of harm to a particular person or property. Rather, the motion court applied New York law's distinction between actions that pose a risk of harm and actions that will lead directly to harm (or are taken with the purpose of causing harm). ( See 2020 NY Slip Op 34331[U], at *12, *13.) On Century's reading, this holding was wrong: According to Century, New York law provides that when a serious risk of harm to property comes to pass, the resulting property damage is expected or intended. Yet the First Department did not conclude in NL Industries that the motion court had erred in finding insufficient the presence of a serious risk of harm. Nor, for that matter, did the Court suggest that the motion court should have focused instead on the fact that the California courts found that NL Industries's actions risked harm to the public health. Instead, the First Department held that the motion "court correctly rejected the insurers’ argument" that the rulings in the California action required a conclusion in the New York coverage action that the damage was expected or intended. ( 2022 NY Slip Op 02056, at *1.)

Additionally, the First Department supported its conclusion on this point with citations to three decisions assessing whether property damage was expected or intended. (See id. ) Each of those decisions considered whether a given policyholder had acted knowing that harm would result from the policyholder's actions, or instead had known only of a risk of resulting harm. None involved the distinction that Century now seeks to draw between risk of harm to persons or property and risk of harm to the public health at large. The content and context of the First Department's NL Industries ruling foreclose Century's narrower reading of that decision.

C. Whether Brooklyn Union May Argue that the Property Damage was Concurrently Caused by both Intentional and Accidental Conduct

Brooklyn Union's proposed accidental-damage instruction sought to carve out space for the possibility that the jury might find some of the damage at issue to be expected or intended and some accidental. In that circumstance, Brooklyn Union proposed, the jury should be instructed that it should "still provide a ‘Yes’ answer to the question of whether at least some of the property damage was caused by accident" and therefore covered under the Century policies. (NYSCEF No. 1134 at 5.) This court's initial communications with the parties, both responding to the parties’ proposed sets of instructions and proposing the court's own draft language, reflected skepticism of Brooklyn Union's "concurrent-causation" argument. The court, relying on the First Department's decision in Throgs Neck Bagels v GA Ins. Co. of NY (241 AD2d 66, 69-70 [1st Dept 1998] ), concluded that Brooklyn Union's proposed language would require the jury to determine the "dominant and efficient cause" of the property damage (and whether that cause was expected/intended or accidental)—and that the jury lacked the evidence necessary to make that determination.

In letter briefing responding to the court's proposed draft instructions (and in ensuing oral argument), Brooklyn Union contended that this court's skepticism was ill-founded. Brooklyn Union took the position that the line of authority on which the court had relied applied only in the context of first -party insurance, not the sort of third-party liability insurance at issue in this action. (See NYSCEF No. 1202 at 3-4; NYSCEF No. 1205 at 3.) This court does not agree that New York law draws Brooklyn Union's proffered distinction. Rather, as Century contended at argument, New York courts have employed a version of the dominant/efficient-cause test in actions concerning third-party liability insurance, as well as in first-party insurance cases.

The First Department's decision in U.S. Fire Insurance Co. v New York Marine & General Insurance Co. (268 AD2d 19, 22-23 [1st Dept 2000] ) provides a good example. There, two insurers were disputing whether coverage existed for claims in an underlying personal-injury action brought against the policyholder by third parties. The question before the court was whether a relevant exclusion in the defendant insurer's policy foreclosed the defendant's duty to indemnify, when some of the theories of liability in the underlying actions were within the scope of the exclusion and some were not. The First Department held that answering this question required identifying "the operative act giving rise to this accident." ( Id. at 24.) Because all the theories of liability in the underlying action rested, causally speaking, on an operative act within the scope of the policy exclusion, no coverage existed under the defendant insurer's policy. (See id. at 23-24.) Similarly, a recent First Department decision described a prior decision construing a third-party policy as having held that "an exclusion bar[red] coverage because the excluded activity was the ‘immediate and efficient cause of the injury.’ " ( New Hampshire Ins. Co. v MF Global Fin. USA Inc. (––– AD3d ––––, 2022 NY Slip Op 01880, at *6 [1st Dept Mar. 17, 2022] [emphasis omitted], quoting New Hampshire Ins. Co. v Jefferson Ins. Co. of NY , 213 AD2d 325, 327 [1st Dept 1995].)

The First Department's decisions in U.S. Fire Insurance , MF Global , and New Hampshire Insurance demonstrate that the first-party/third-party distinction advocated for by Brooklyn Union is not present in precedent. Brooklyn Union thus would have been entitled to a concurrent-causation instruction only if the trial evidence would permit a reasonable jury not only to distinguish between accidental and intentional property damage where both forms of damage were present, but also to determine that the accidental damage was the efficient cause of the investigation/cleanup costs for which Brooklyn Union now seeks coverage. The evidence at trial would not support those findings.

II. How the Policies’ Per-Occurrence Limits Should be Applied to Brooklyn Union's Covered Losses

The Century policies at issue in this action contained per-occurrence limits on Century's coverage obligations. The parties have repeatedly contested how those limits should be applied, and in so doing laid the groundwork for the court's determinations discussed below.

In particular, several of the Century policies remained in place for more than one year. Brooklyn Union has contended that these policies were renewed annually, thereby also resetting the policies’ per-occurrence limits annually. Century has contended instead that the policies each had one multi-year term (rather than a one-year term renewed annually several times), such that the policies’ per-occurrence limit should be applied once over the full term of the policy.

At summary judgment, this court held that the multiyear policies comprised one multiyear term rather than several one-year terms and that the language of the policies unambiguously provided that the per-occurrence limits would be applied on a term, rather than annual basis. (See Century Indem. Co. v Brooklyn Union Gas Co. , 2018 NY Slip Op 52018[U], at *9 [Sup Ct, NY County July 18, 2018].) On appeal, the First Department modified that aspect of this court's summary-judgment ruling. The court held that "the policies that had multi-year terms or contained a multi-year renewal are ambiguous as to whether the per-occurrence limits were limits for the respective policies’ entire terms or were annual per-occurrence limits." ( Century Indem. Co. v Brooklyn Union Gas Co. , 170 AD3d 632, 633 [1st Dept 2019].)

This court then held, in deciding a pretrial motion in limine, that the first step in seeking to resolve this ambiguity would be to consider extrinsic evidence submitted by the parties, if any. (See Century Indem. Co. v Brooklyn Union Gas Co. , 2022 NY Slip Op 50083[U], at *2-*3 [Sup Ct, NY County Feb. 11, 2022].) This court also held that Century would be able to introduce testimony from an expert witness, Robert M. Anderson, about industry custom-and-practice with respect to per-occurrence limits. (See id. at *3-*5.) The court held that although Anderson's custom-and-practice opinion, as previewed in his report, was potentially open to criticism that the expert lacked sufficient expertise and that the opinion was insufficiently supported, these criticisms were best addressed at trial (including through cross-examination). (See id. at *5.)

At trial, Brooklyn Union did not introduce any extrinsic evidence on the per-occurrence limits issue. Century offered the custom-and-practice testimony of its expert witness, Anderson. He opined that (i) Brooklyn Union was a sophisticated policyholder, given its separate internal insurance department, maintenance of a large self-insured retention and multiple layers of excess policies, and use of a well-known insurance broker and outside insurance counsel; (ii) as a sophisticated policyholder, Brooklyn Union would have been aware of industry custom and practice about when per-occurrence limits would (and would not) be applied on an annual basis; (iii) the industry custom and practice during the period of the Century policies was that per-occurrence limits applied annually only if the policy expressly provided so.

At the close of Century's case, Brooklyn Union moved for a directed verdict that would preclude the jury from being asked to resolve the issue of industry custom and practice with respect to per-occurrence limits. Brooklyn Union contended that Anderson's opinion was insufficiently probative to raise a jury question about the presence or absence of industry custom and practice on this issue. This court granted Brooklyn Union's motion. The court explained that the concerns it had expressed in its motion in limine ruling about the basis and support for Anderson's custom-and-practice opinion were not assuaged—indeed, were exacerbated—by his trial testimony. This testimony was insufficient to put a custom-and-practice question before the jury.

In other words, Brooklyn Union provided no extrinsic evidence addressing the per-occurrence-limits ambiguity, and Century's proffered extrinsic evidence lacked sufficient probative value to raise a jury question that might determine the issue. The resolution of the ambiguity in the Century policies—i.e. , the choice between the parties’ two reasonable interpretations of the per-occurrence limits in those policies—therefore reverted to this court, to be decided as a matter of law.

In reaching its decision, this court must consider two interpretive issues: First , whether the ambiguity in the multiyear Century policies must be construed in Brooklyn Union's favor under the interpretive doctrine of contra proferentem. This court concludes, for the reasons discussed in Section II.A, that contra proferentem does not apply here. Second , whether, absent contra proferentem, the best reading of the multiyear policies’ per-occurrence limits is that they apply on an annual or term basis. As discussed in Section II.B, the court concludes that the limits apply on a term basis.

Given these holdings, a third question arises: Whether the property damage that happened during the multiyear policies’ periods stemmed from one occurrence per period or several occurrences per period. This court concludes, as set forth in Section II.C, that the property damage stemmed from a single occurrence in each multiyear policy period.

A. Whether the Contra Proferentem Doctrine Requires the Per-Occurrence Limits to be Applied on a Term Basis

Brooklyn Union argues that under Court of Appeals precedent, the contra proferentem doctrine requires this court to resolve the per-occurrence-limits ambiguity in Brooklyn Union's favor. Century contends in response that the issue of the applicability of contra proferentem is controlled by decisions of the Appellate Division; and that these decisions establish that contra proferentem does not apply here, because Brooklyn Union is a sophisticated policyholder that maintained a large self-insured retention.

No New York court has previously addressed whether Court of Appeals decisions mentioning contra proferentem control that doctrine's applicability when the policyholder is sophisticated, or whether the governing decisions have instead been issued by the Appellate Division. Considering the question as a matter of first impression, this Court agrees with Century that the Appellate Division supplies the controlling authority—and that the Appellate Division has held that contra proferentem does not apply in the circumstances of this action.

1. Whether contra proferentem applies when the policyholder maintains a large self-insured retention

The Appellate Division, Fourth Department, has held that a policyholder in Brooklyn Union's position is not entitled to the benefit of contra proferentem. (See Loblaw, Inc. v Employers’ Liab. Assur. Corp. , 85 AD2d 880, 881 [4th Dept 1981], affd on other grounds 57 NY2d 872 [1982].) There, plaintiff Loblaw maintained a $25,000 self-insured retention for workers’ compensation claims and purchased excess insurance to cover claims above that amount. (See id. at 880.) After Loblaw's insurer disclaimed coverage of an excess claim on the ground of untimely notice, the parties contested the excess policy's notice requirements. (See id. at 881.) The Fourth Department rejected Loblaw's argument that contra proferentem required resolving this dispute in Loblaw's favor. The Court noted the general principle that contra proferentem "is not applicable in a contest between two insurance companies." (Id. ) Although Loblaw was not itself an insurance company, that Loblaw was "act[ing] as a self-insurer with respect to Workers’ Compensation claims through its agent, a licensed insurance adjusting company," meant that "its status is more akin to that of an insurance company than to that of an individual who is inexperienced in matters of insurance coverage for whose benefit the rule was promulgated." (Id. ; accord Cummins v Atlantic Mut. Ins. Co. , 56 AD3d 288, 290 [1st Dept 2008] [holding that the policyholder's maintaining of a self-insured retention was a factor weighing against application of contra proferentem].)

Here, Brooklyn Union, like Loblaw, was acting as a self-insurer up to a $100,000 limit and had extensive experience in insurance-coverage-related issues—as reflected in the fact that it had an entire department for insurance-related matters and worked with a prominent insurance brokerage, Johnson & Higgins, to obtain coverage. Under the Fourth Department's decision in Loblaw , Brooklyn Union is not entitled to the benefit of the contra proferentem doctrine. (Cf. Morgan Stanley Grp. Inc. v New England Ins. Co. , 225 F3d 270, 279 [2d Cir. 2000], quoting Loblaw , 85 AD2d at 881 [acknowledging that contra proferentem would not apply in a scenario in which "the policy holder is ‘akin to ... an insurance company’ because of its sophistication in matters of insurance coverage"].)

Brooklyn Union contends that Loblaw is inapposite here because that decision referred in its first sentence to the policy at issue being a "Workmen's Compensation Excess Reinsurance Contract." ( 85 AD2d at 880.) Brooklyn Union asserts that this reference shows the decision held only that contra proferentem "does not apply to reinsurance agreements, which are contracts between insurers and not at issue in this case." (NYSCEF No. 1202 at 11 n 4; see also NYSCEF No. 1208 at 4 [same].) This contention is without merit. Nothing in Loblaw suggests it turns on whether the policy at issue was a reinsurance contract, in particular. As discussed above, the Court specifically cautioned that it was treating Loblaw as akin to an insurance company, not as an insurer itself. And later decisions relying on this part of the Fourth Department's decision have not described its holding as being about (or limited to) reinsurance.

Brooklyn Union also argues that this court should not treat the presence of a large self-insured retention as dispositive because "[c]ontra proferentem is regularly applied in coverage disputes even where policyholders maintain sizable self-insured retentions." (NYSCEF No. 1202 at 11.) But Brooklyn Union provides no decision applying New York law in which a court has held that the presence of a large self-insured retention does not foreclose application of contra proferentem. And the cases Brooklyn Union does cite does not support its argument.

Brooklyn Union relies on a passage of the Court of Appeals's decision in Continental Casualty Co. v Rapid-American Corp. (80 NY2d 640 [1993] ), a case in which the policyholder maintained a self-insured retention. The aspect of the decision on which Brooklyn Union relies concerns whether a pollution exclusion in the policy barred coverage so as to negate the insurer's "exceedingly broad" duty to defend. ( Id. at 648.) To negate coverage, an exclusion must be "be stated in clear and unmistakable language" and be "subject to no other reasonable interpretation"—i.e. , must not be ambiguous. ( Id. at 652.) And the Court held that the pollution exclusion at issue was ambiguous in the circumstances of the case (see id. at 653-654 ), and as a result did not oust the insurer's duty to defend (see id. at 655 ).

To be sure, the Continental Casualty Court stated its conclusion in terms that might suggest application of contra proferentem: "[S]ince the pollution exclusion is ambiguous in these circumstances and could be construed against the insurer-drafter and in favor of potential coverage for the policyholder, [the insurer] cannot escape its duty to defend." (Id. ) But this statement is brief, also tracks closely the basic interpretive principle that an ambiguous exclusion does not oust coverage, and was made in the context of the duty to defend rather than the more narrow duty to indemnify. This court is skeptical that this passage from Continental Casualty should be treated as controlling on the scope-of-coverage question that the parties are disputing here with respect to Century's duty to indemnify.

Brooklyn Union also cites Oneida Ltd. v Utica Mut. Ins. Co. (178 Misc 2d 572 [Sup Ct, Madison County 1998], affd 263 AD2d 825 [3d Dept. 1999] ). (See NYSCEF No. 1202 at 11.) But in Oneida , the motion court's description of the relevant interpretive principles relied on Loblaw as authority for the contra proferentem rule, without addressing Loblaw ’s discussion of the significance of self-insured retentions in this context. (See 178 Misc 2d at 593.) This is likely because the motion court's application of contra proferentem in the case before it was limited to concluding that the doctrine would furnish a second alternative basis for rejecting an unpersuasive argument by the insurer. (See id. at 594 ). And the Third Department did not reach that argument in affirming the motion court's ruling.

This court is thus unpersuaded by Brooklyn Union's suggestions that the Fourth Department's decision in Loblaw is not on point here or that Loblaw ’s holding has been undermined by other, more recent precedents.

2. Whether contra proferentem applies when the policyholder is sophisticated

Even if the Fourth Department's decision in Loblaw did not control this case, this court would reach the same conclusion: The First Department has held that contra proferentem does not govern interpretation of insurance policies when the policyholder is sophisticated. (See Westchester Fire Ins. Co. v MCI Communications Corp. , 74 AD3d 551, 551 [1st Dept 2010].)

In Westchester Fire , the parties contested on appeal whether contra proferentem should apply to the construction of an endorsement to the underlying insurance policy. Defendant MCI, citing the Court of Appeals's decision in Thomas J. Lipton, Inc. v Liberty Mutual Insurance Co. (34 NY2d 356 [1974] ), argued that contra proferentem applied whether or not MCI was a sophisticated policyholder. Westchester Fire Insurance, citing the Appellate Division decisions in Loblaw and Cummins , argued the doctrine did not apply, on the ground that MCI was a large company that had a separate risk-management department and obtained insurance through the Johnson & Higgins brokerage. The First Department agreed with Westchester Fire Insurance. The Court's decision held in categorical terms that contra proferentem "would be inapplicable to this sophisticated policyholder." ( Westchester Fire , 74 AD3d at 551, citing Cummins , 56 AD3d at 290 ; accord Certain Underwriters at Lloyd's, London v Essex Global Trading, Inc. , 147 AD3d 595, 596 [1st Dept 2017], citing Westchester Fire , 74 AD3d at 551 [same].)

Brooklyn Union contends that Westchester Fire is "inapposite" here because the First Department also held that the provision at issue in that case was unambiguous. (NYSCEF No. 1202 at 11 n 4; NYSCEF No. 1208 at 4.) This court takes this statement as contending that the Court's contra proferentem ruling in that case was merely nonbinding dicta. This court does not agree that the ruling was dicta. In context, the disposition of Westchester Fire rests on at least two alternative holdings, each rejecting a separate argument by MCI.

The Westchester Fire Court held (i) the policy endorsement at issue, a coverage provision rather than an exclusion, unambiguously meant what Westchester Fire said it did; and (ii) the endorsement should not be read instead in MCI's favor under contra proferentem because MCI was a sophisticated policyholder. (See 74 AD3d at 551.) In these circumstances, each of these holdings has binding precedential effect: "[W]hen two or more points arise and are argued and the appellate court passes upon them all, no branch of the decision is merely incidental, but all the grounds thereof must be taken to be equal in force and together constitute the judgment." ( Broderick v City of NY , 295 NY 363, 368-369 [1946] ; accord Clough v Board of Educ. of Spencerport Cent. School Dist. , 56 AD2d 233, 236 [4th Dept 1977].) Indeed, the First Department has held that it is "erroneous[ ]" to "refer[ ] to as ‘dicta’ " several alternative holdings that each support the same disposition. (Matter of Nonhuman Rights Project, Inc. v Breheny , 189 AD3d 583, 583 [1st Dept 2020].)

Brooklyn Union asserts that its contra-proferentem position is supported by consistent Court of Appeals authority. (See NYSCEF No. 1202 at 9-11; NYSCEF No. 1208 at 2.) Were Brooklyn Union correct, the court would be put in the difficult position of having to choose between conflicting precedents from the two courts that sit in review of this court's decisions. But Brooklyn Union is not correct. On close examination, the Court of Appeals decisions on which it relies do not say nearly as much about contra proferentem as Brooklyn Union would have it.

For example, Brooklyn Union has relied heavily (both in its papers and in oral argument at trial) on the Court of Appeals's recent decision in J.P. Morgan Securities Inc. v Vigilant Insurance Co. (37 NY3d 552 [2021] ). (See NYSCEF No. 1202 at 11; NYSCEF No. 1208 at 1-2). But J.P. Morgan Securities does not apply contra proferentem. The decision's sole reference to that doctrine comes in an early passage that serves only to lay the groundwork for the Court's analysis by articulating the general principles that govern interpretation of insurance policies. (See 37 NY3d at 561-562.) The Court's actual stated bases for its interpretive conclusion in J.P. Morgan Securities are, instead, the principles of (i) giving "exclusion[s] ... a ‘strict and narrow construction’ " that is (ii) "consistent with the expectations of a reasonable insured at the time of contracting." ( Id. at 569, quoting Seaboard Sur. Co. v Gillette Co. , 64 NY2d 304, 311 [1984].) Those principles are not contra proferentem.

The only Court of Appeals decision cited by Brooklyn Union that applied contra proferentem, standing alone, to resolve a policy ambiguity in an action involving a sophisticated policyholder is State v Home Indem. Co. (66 NY2d 669, 672 [1985] ). And neither Home Indemnity nor any of the other New York decisions relied upon by Brooklyn Union indicate that the appellate courts issuing those decisions even considered whether the doctrine applies when the policyholder in the case is sophisticated. Absent a showing that the sophistication issue was both considered and decided in any of those decisions, the decisions lack precedential effect on this issue now. (See People v Bourne , 139 AD2d 210, 216 [1st Dept 1988] [explaining that "[a] case ... is precedent only as to those questions presented, considered and squarely decided"].)

Brooklyn Union contends that this court should reject Century's position for lack of "guideposts" to use in "assessing whether, when and how a corporate policyholder supposedly crosses the threshold" of sophistication. (NYSCEF No. 1208 at 3.) But the potential for difficulty in applying on-point appellate precedent is not a basis to disregard that precedent. Regardless, it might well be that future cases involving this issue might present close and challenging line-drawing problems. That is different, though, from saying that this court lacks a sound basis to conclude that Brooklyn Union was a sophisticated policyholder, in the relevant sense, when it purchased the excess policies at issue in this case.

Here, Century's custom-and-practice expert, Anderson, gave unrebutted testimony that Brooklyn Union was sophisticated because, in addition to its large self-insured retention, it—like the policyholder in Westchester Fire —maintained a separate insurance department for handling claims, had several layers of excess coverage, and obtained insurance through the large corporate brokerage of Johnson & Higgins. Given this evidence, the court has little difficulty concluding that Brooklyn Union was sophisticated.

Thus, only two appellate precedents squarely consider whether a policyholder with Brooklyn Union's characteristics is entitled to the benefit of the contra proferentem interpretive presumption—the First Department's decision in Westchester Fire , and the Fourth Department's decision in Loblaw . Each of those precedents holds that contra proferentem does not apply in those circumstances. And no New York authority holds to the contrary.

B. Whether the Per-Occurrence Limits Apply on an Annual or Term Basis

Absent either extrinsic evidence or the contra-proferentem presumption, the issue before this court is, once again, the question the court faced in 2018: Whether the per-occurrence limits in each multiyear Century policy applied (and reset) on an annual basis, or whether one per-occurrence limit applied for the full multiyear policy period.

With respect to the Century policies in place for multiple years, one policy, XPL-3661, was issued for an initial term of one year and then repeatedly renewed for more than a decade—once for a 25-month period, otherwise for 12 months at a time. (See generally NYSCEF Nos. 615, 616.) The remaining Century policies were each issued for an open-ended period, under which they would remain in force until cancelled by either party with appropriate advance written notice. (See id. )

Century does not dispute that each of the one-year renewals of XPL-3661 gave rise to a new annual per-occurrence limit. The question is whether the 25-month renewal period of XPL-3661, and the multiyear periods of the other policies, were subject to annual or full-period per-occurrence limits. Brooklyn Union bears the burden of proof on this question. It is the insured's burden initially to establish the presence of coverage. ( Consolidated Edison Co. of NY v Allstate Ins. Co. , 98 NY2d 208, 218 [2002].) And the Appellate Division has held that the amount of a policy's limit or sublimit goes to what coverage obligations the insurer owes in the first place, as distinct from whether an exclusion ousts a coverage obligation that would otherwise exist.

Brooklyn Union has advanced various arguments in this litigation to meet its burden of proof. None of these arguments is persuasive.

1. At summary judgment, Brooklyn Union contended that each party's right to cancel the policies unilaterally on advance written notice meant that the open-ended policies are renewed each year, giving rise to a new policy with a new per-occurrence limit. (See NYSCEF No. 386 at 23-24, citing Moore v Metro. Life Ins. Co. , 33 NY2d 304, 311-312 [1973].) But the motion-court and appellate decisions in Moore are inapposite here.

In Moore , the policy expressly provided that its term was " ‘for one year commencing on the date of issue hereof and shall be renewed from year to year thereafter" unless terminated as of the policy's anniversary date through a written notice served at least 90 days in advance. ( 33 NY2d at 311.) The question in that case was not whether the policy was renewed each year. It undisputedly was. The issue was instead whether the nature of that renewal—i.e. , one that occurred automatically by virtue of the passing of the policy's anniversary date without termination—gave rise each year (i) to a new policy that would be subject to an intervening, prospective statutory amendment governing its terms, or (ii) a renewed version of the original , pre-amendment policy that would not be subject to the new statutorily imposed terms. (See id. at 311-312 [holding that the renewal gave rise to a new policy, and distinguishing Mulligan v Travelers Ins. Co. (280 AD 764 [1st Dep 1952], affd 306 NY 805 [1954] ), in which the renewal of the policy had merely carried forward the terms of the original policy].)

That holding in Moore has nothing to do with the issue presented here—i.e. , whether the multiyear policy periods comprised multiple one-year policies, renewed annually, or a single unrenewed multiyear policy.

2. Both at summary judgment and in pretrial motions in limine, Brooklyn Union suggested that its annual payment of premiums to Century (pursuant to the terms of annually issued premium endorsements) necessarily establishes that the multiyear policies were renewed annually, as well. (See NYSCEF No. 386 at 22, 24; NYSCEF No. 679 at 2-3, 8.) But Brooklyn Union cites no authority for this proposition. And appellate authority demonstrates that Brooklyn Union's reliance on annual premium payments as proof of annual renewals is misplaced.

In Shared-Interest Management Inc. v CNA Financial Ins. Group (283 AD2d 136, 137-138 [3d Dept 2001] ), the Appellate Division, Third Department, construed a policy issued for a period extending from the date of issuance "Until Cancelled." The policy had a single coverage limit for the entire policy period. On the third and sixth anniversaries of the issue date, the insured received notices from the insurer "indicat[ing] that the policy had been renewed for [a] ‘Premium Term’ " lasting an additional three years. ( Id. at 138-139.) The insured contended that each of these notices constituted a renewal of the original policy giving rise to a new three-year policy with a new policy limit.

The Third Department squarely rejected this contention. The Court instead agreed with the insurer that "what [the insured] would characterize as separate policy periods are simply three-year premium terms, i.e., the time period during which the policy premium is established." ( Id. at 139.) As a result, the original policy—and the original whole-period policy limit—remained in force throughout.

Under Shared-Interest Management ’s holding, a policy term's providing for the annual determination and annual payment of premiums does not, without more, give rise to the annual renewal of the policy. And Brooklyn Union does not provide more. To the contrary, the language of the premium endorsements from one of the multiyear policies itself undercuts Brooklyn Union's position.

As discussed above, policy XPL-3661 was renewed repeatedly pursuant to the terms of various "renewal endorsements." (See NYSCEF No. 615 at 22-30, 32-34, 38-40, 44, 48.) Each of these endorsements expressly provided that in consideration of a further premium, the policy's coverage would "continue in force for a further period"—12 months each time, except for the 25-month period from November 1959 to the beginning of 1962. The annual premium endorsements for policy XPL-5910, in place from 1962 to 1965, on the other hand, did not describe themselves as "renewal endorsements" or contain equivalent "continue in force for a further period" language. Instead, those endorsements merely stated the premium for the next year and provided that this premium is a "fixed premium not subject to audit" unless the policyholder's revenues fluctuated by a sufficiently large amount or unless "the policy is cancelled " pursuant to its terms. (NYSCEF No. 615 at 65, 67, 68 [emphasis added].)

Similarly, policy SRL-2221, though it was cancelled after one year, was initially purchased for a three-year term. (See NYSCEF No. 615 at 84.) SRL-2221, as purchased, included a "premium computation endorsement" setting forth the process for determining the premium to be paid by Brooklyn Union "upon the expiration of each twelve (12) months of the policy period , or if sooner terminated." (Id. at 94 [emphasis added].) The endorsement contains no language indicating that payment of this annual premium constituted consideration for the policy to "continue in force for a further period" beyond the initial three-year period the policy would be in effect.

Brooklyn Union also suggests that it would be illogical and inequitable for an insured to gain no additional coverage protection (in terms of a reset per-occurrence limit) when paying premiums for additional years beyond the first. (See NYSCEF No. 679 at 8, citing National Union Fire Ins. Co. of Pittsburgh, Pa. v. Farmington Cas. Co. , 1 Misc 3d 671, 674 [Sup Ct, NY County 2003].) But Farmington Casualty is different in a crucial respect from the Century policies at issue here. In that case, it was undisputed that the insured had renewed the policy for a second year, along with paying a second year's premium. (See 1 Misc 3d at 674.) The dispute in that case was instead over a question uncontested here—whether the insured's renewal of the policy reset the policy's liability limit. (See id. at 673-674.) The motion court's skepticism there of the insurer's position (i.e. , that renewing the policy and paying a fresh premium should have no effect on the policy's aggregate limit) has no bearing on whether payment of a fresh premium, standing alone, constitutes a renewal of the policy.

3. At trial, Brooklyn Union emphasized that the multiyear "policies are simply silent on [the] issue" whether the per-occurrence limits apply annually or for a multiyear term. (NYSCEF No. 1208 at 5.) True enough; but that silence does not get Brooklyn Union to where it wants to go. The most reasonable way to interpret a policy limit that does not specify the period over which it applies is that the limit applies for the length of time the policy is in effect—whether that be a year, two years, or five. If, as occurred here, Brooklyn Union chose for its own business reasons to purchase, renew, and maintain policies for varying lengths of time during the overall 1941-1969 duration of coverage, that choice necessarily carried with it a choice by Brooklyn Union to make itself subject to policy limits that would be applied over those same varying policy periods.

Brooklyn Union does not explain why, notwithstanding the numerous variations in the length of the policy periods (and renewal periods) that it purchased, the per-occurrence limits of those policies should be computed annually every year from 1941 to 1969. Brooklyn Union could have obtained coverage that expressly provided that it would be renewed annually, or that the per-occurrence limits would apply annually regardless of the length of the policy's period of coverage. But it did not. This court declines to read these renewal or policy-limit-duration provisions into the terms of the policies that Brooklyn Union did purchase.

C. Whether Brooklyn Union's Covered Losses in the Multiyear Policy Periods Stemmed from a Single Occurrence or Multiple Occurrences

This court's ruling late in the course of trial that the multiyear policies should be interpreted only to have term-based per-occurrence limits did not finally resolve all the policy-limit-related questions in the trial. In particular, after the court issued that ruling, Brooklyn Union contended in letter briefing that the investigation/cleanup costs for which Brooklyn Union is seeking coverage should be understood as the product of "a separate occurrence (at each of the three former MGP sites) for each annual period (or portion thereof) encompassed by the 29 years in which Century provided coverage." (NYSCEF No. 1209 at 5.) Because each of these asserted occurrences would be subject to its own separate per-occurrence limit, the result—were this court to agree—would have been to substantially increase Century's coverage liability during the multiyear policy periods. But this court does not agree.

As an initial matter, Brooklyn Union is incorrect in asserting that the relevant standard for determining whether "separate incidents should be aggregated into one occurrence or treated as separate occurrences" is necessarily the "unfortunate event" test. (Id. ) The Court of Appeals has made clear in this context that the initial analytical question must instead be whether the language of the policy itself articulates a specific approach to defining "accident" or "occurrence" that would apply to the coverage issue at hand. (See Appalachian Ins. Co. v General Elec. Co. , 8 NY3d 162, 172-173 & n 3 [2007].) It is only when the policy language does not itself govern how different incidents are to be grouped into occurrences, or when the policy's "grouping" language does not cover the incidents at issue, that a court is to apply the unfortunate-event standard. (See Appalachian Ins. , 8 NY3d at 173 & n 3 ; Mt. McKinley Ins. Co. v Corning Inc. , 96 AD3d 451, 452 [1st Dept 2012] ; ExxonMobil v Certain Underwriters at Lloyd's, London , 2007 NY Slip Op 51138[U], at *8 [Sup Ct, NY County June 5, 2007], affd 50 AD3d 434, 434-435 [1st Dept 2008] [holding that policy language grouping together claims arising from exposure to the same general injury-producing conditions at a given location did not apply to claims arising from the same injury-producing conduct committed at many locations].)

Here, the three Century policies in effect between 1963 and 1969 each provide that "[a]ll damages arising out of ... exposure to substantially the same general conditions shall be considered as arising out of one occurrence." ( Century Indem. , 2018 NY Slip Op 52018[U], at *4; see also NYSCEF No. 615 at 59, 78, 89.) Under this grouping language, the cumulative environmental damage that resulted from MGP operations during the terms of the three policies—which is "spatially identical and temporally close enough that there were no intervening changes in the injury-causing conditions"—constitutes one occurrence for each of the three MGP sites at issue. ( Nesmith v Allstate Ins. Co. , 103 AD3d 190, 194 [4th Dept 2013], affd 24 NY3d 520, 525-526 [2014] ; accord Hiraldo v Allstate Ins. Co. , 5 NY3d 508, 511-512 [2005] [holding that under this same grouping language, physical injury from continuous exposure to lead paint in one apartment over a three-year period constitutes one occurrence]; see also Mt. McKinley Ins. , 96 AD3d at 452 [explaining that under an "identical or similar grouping provision[ ]," exposures "emanating from the same location at a substantially similar time" constitute a single occurrence].)

The other three Century policies, in effect between 1941 and 1963, do not have this grouping language. For these policies, the governing standard is, as Brooklyn Union contends, the unfortunate-event test. Applying that test, though, requires this court first to determine which instances of environmental damage are even at issue in this context.

That is, Brooklyn Union is contending that under the unfortunate-event test, the investigation/cleanup costs for which Brooklyn Union is being held liable under Superfund are the product of a separate occurrence for each annual period in which Century provided coverage. As discussed above, though, policy XPL-3661, in effect between 1952 and 1963, was for nearly all of that time a one-year policy renewed annually. It is undisputed that for each of these one-year policy periods, the costs at issue derive from a single occurrence and are subject to annual per-occurrence limits.

The relevant disagreement between the parties over the results of the unfortunate-event test thus applies only to policies that (i) lack grouping language; and (ii) under this court's per-occurrence-limit ruling discussed above, contain multiyear, term-based per-occurrence limits, rather than annual limits. Those policies are P-84739, in effect from 1941 to 1946; XPL-572, in effect from 1946 to 1951; and the 25-month renewal of XPL-3661 from November 1959 until the beginning of 1962. This court concludes that under the unfortunate-event test, the environmental damage happening during these three policy periods for which Brooklyn Union has been held liable constitutes one occurrence per policy per MGP site, for a total of nine occurrences. This court is unpersuaded by Brooklyn Union's argument that the correct number of occurrences under these policies is instead one occurrence per year per MGP site (a total of 36 occurrences).

Under the unfortunate-event test, courts are "to focus on ‘the nature of the incidents giving rise to damages.’ " ( Roman Catholic Diocese of Brooklyn v National Union Fire Ins. Co. of Pittsburgh, Pa. , 21 NY3d 139, 150 [2013], quoting Appalachian Ins. , 8 NY3d at 171 [alteration omitted].) In this analysis, "several factors emerge as relevant": In particular, "whether there is a close temporal and spatial relationship between the incidents giving rise to injury or loss, and whether the incidents can be viewed as part of the same causal continuum, without intervening agents or factors." ( Appalachian Ins. , 8 NY3d at 171-172.)

Here, the incidents giving rise to damage are the release of contaminants from the three MGPs and/or the post-release migration of those contaminants into groundwater under those MGP sites and into the sediments under the Gowanus Canal. These incidents, by their very nature, compose gradual, continuous, and cumulative processes at each MGP site, which flow (literally) from common causal starting points and continue without intervening causal factors, occur in substantially the same locations, and cannot meaningfully be separated into discrete temporal segments. These processes are most naturally thought of as comprising a single unfortunate event at each MGP site. Moreover, treating these processes as a single occurrence tracks the Century policies’ definition of an occurrence as a "continuous or repeated exposure to conditions." (E.g. NYSCEF No. 615 at 59; See Roman Catholic Diocese of Brooklyn , 21 NY3d at 151 [discussing how "injuries caused by exposure to environmental ... hazards ... comported with" this definition of occurrence].)

To support its contrary argument that the release and migration of environmental contaminants should instead be understood "discrete incidents of property damage" forming separate occurrences, Brooklyn Union provides only the bare assertion that these incidents "have no close temporal and spatial relationship" to one another. (NYSCEF No. 1209 at 3.) Brooklyn Union does not explain why this court should think that so. Nor is that explanation self-evident.

Additionally, the only method of separating this ongoing process of environmental contamination that Brooklyn Union does offer—breaking the process into many 12-month-sized pieces, each piece a different occurrence—is not only unrelated to the nature of that contamination but also inherently arbitrary. On Brooklyn Union's position, an eleven-month period of release or migration of contaminants would constitute one occurrence, if that period fell within a given 12-month period; but a two-month (or two-week) period of contaminant release or migration would constitute two occurrences, should the period fall at the end of one 12-month period and the beginning of another. Breaking up incidents into separate occurrences in this manner makes little sense.

For the reasons discussed above, this court concludes that (i) the ambiguity in the Century policies’ per-occurrence limits must be resolved without applying the interpretive presumption of contra proferentem; (ii) the best resolution of this ambiguity is that the per-occurrence limits in the multiyear Century policies should be understood to apply on a term, rather than annual basis; and (iii) the investigation/cleanup costs for which Brooklyn Union has coverage under the multiyear policies stem from one occurrence under each policy, not several.

III. How to Allocate Brooklyn Union's Covered Losses on a Pro Rata Basis

Two further disagreements between the parties relate to the issue of pro rata allocation. Under this court's 2018 summary-judgment ruling—as affirmed by the First Department—Century's coverage liability must be determined based on allocating pro rata Brooklyn Union's total covered losses. (See Century Indem. , 2018 NY Slip Op 52018[U], at *7-*9, affd 170 AD3d at 633.)

As this court discussed in a February 2022 motion-in-limine ruling, it follows from this principle that the total amount of covered losses suffered by Brooklyn Union must substantially exceed the Century policies’ face-value attachment point of $100,000 (excess of a self-insured retention) before liability will attach, because the total loss must be large enough that it still exceeds the policies’ applicable attachment points after the loss has been allocated pro rata. (See Century Indem. Co. v Brooklyn Union Gas Co. , 74 Misc 3d 733, 736 [Sup Ct, NY County 2022].) Additionally, under that same February 2022 ruling, the pro rata allocation analysis must also be undertaken to determine when Brooklyn Union would first have been required to provide notice of occurrence to Century—and whether Brooklyn Union's February 1993 notice was timely. (See id. at 736-737.)

The parties disagreed at trial about two aspects of how to conduct the pro rata allocation process under the prior rulings in the action: (i) How to determine the post-allocation attachment points of the Century policies for notice and liability purposes ; and (ii) how to allocate evidentiary burdens in determining the total number of years over which to allocate Brooklyn Union's covered losses.

Neither party provided New York authority that would determine these issues. This court's resolution of the parties’ disagreements, considering the issues essentially as questions of first impression, appears below.

A. How to Determine the Century Policies’ Post-Allocation Attachment Points

Brooklyn Union's position at trial was that determining the Century policies’ post-allocation attachment points would require the jury simply to multiply the total number of years of injury that it had found for each site by the policies’ face-value attachment point of $100,000. Century contended, on the other hand, that determining the policies’ post-allocation attachment points entailed a policy-by-policy analysis, under which the jury would need to multiply the total number of years of injury by dollar figures that varied based on the length of a given policy's term.

Century's approach has obvious practical drawbacks—it entails a more complicated inquiry and more complicated arithmetic by the jury. (See NYSCEF No. 1216 at 23-25, 33-35, 43-44 [verdict sheet].) This court concludes, though, that Century is correct that these complexities are required under the language of the policies.

The language of the Century policies provide that they go into effect for covered losses "in excess of" $100,000 incurred "as a result of any one occurrence or accident," whether that initial $100,000 was covered by a primary policy or by a self-insured retention. (NYSCEF No. 615 at 56 [policy XPL-5910].) It is undisputed that Brooklyn Union maintained a $100,000 self-insured retention rather than purchasing first-dollar primary coverage in that amount. Thus, the post-allocation point attachment for a given Century policy is that amount of money in covered losses from any one occurrence which, after having been spread across the total period of injury, would still be a large enough covered loss in each year of the policy to exhaust the $100,000 self-insured retention.

The Century policies do not, however, provide that the full amount of the self-insured retention (or any primary coverage) must be exhausted in a given year for the policies’ coverage to attach. (See id. ) Where a policy is in effect for multiple years, Brooklyn Union could exhaust its retention from covered losses assigned to each year that are only a fraction of the retention's total amount (as long as all of the losses stem from one occurrence). With respect to multiyear policies with single self-insured retentions, therefore, ascertaining the policies’ attachment point requires not only prorating the covered losses over the length of the period of injury, but also prorating the self-insured retention over the length of the policy.

Two simple contrasting examples illustrate the point.

First: A policyholder purchased five one-year policies, each with a $100,000 self-insured retention, and suffered 100 years of covered losses. The attachment point of those policies is $10 million. That amount, when allocated evenly across the 100-year period of injury, will lead to $100,000 of losses in each year, exhausting the $100,000 retention for each of the annual policies.

Second: A policyholder purchased one five-year policy with a $100,000 self-insured retention, and suffered 100 years of covered losses. The attachment point of the policy is not $10 million, but $2 million. The $2 million figure, allocated evenly across the period of injury, will lead to the policyholder experiencing $20,000 of covered losses in each year. If the policyholder experiences $20,000 of covered losses in each of the five years of the policy, the $100,000 self-insured retention will be exhausted by the end of the policy's term.

These examples show why Brooklyn Union errs in asserting that the post-allocation-attachment-point inquiry can be satisfied simply by "compar[ing] the attachment point of the policies ($100,000) with the extended period over which the damage likely occurred (e.g. , 100 years)." (NYSCEF No. 1211 at 2.) Instead, whether the loss allocated to a particular year must exceed $100,000 for liability to attach, or must only exceed some lesser amount, depends on the details of the policies at issue—and, as a result, may vary depending on the policy.

Brooklyn Union contends that "no objectively reasonable policyholder would engage in th[is] type of in-the-weeds computation of attachment points on a policy-by-policy basis in deciding whether notice is due." (NYSCEF No. 1211 at 2 [italics in original; bolding omitted].) But determining for notice purposes the post-allocation amount at which liability attaches will frequently entail this kind of retrospective inquiry. Even on Brooklyn Union's position, for example, a factfinder in 2022 would, in effect, have to assess how a reasonable policyholder in 1992 would have approximated the length of the period of environmental damage to the groundwater and the Gowanus Canal sediments caused by Brooklyn Union's MGP operations. The ultimate question, though, remains the same: How a reasonable policyholder would view the likelihood that covered losses would reach the excess policies, considered in light of the factfinder's determination of the amount in losses at which the excess policies would be reached.

Further, this court concluded pretrial that as a matter of both precedent and the law of the case, Brooklyn Union correctly argued that "[i]f pro rata allocation must be taken into account for purposes of determining when Century can be held liable, it must also be taken into account for purposes of determining when notice was due." (NYSCEF No. 618 at 2; see Century Indem. , 74 Misc 3d at 737-744.) That conclusion cuts both ways. As discussed above, this court has concluded that the attachment-point liability inquiry is more involved than Brooklyn Union suggests, because the multiyear Century policies’ self-insured retentions and per-occurrence limits must be applied over the full term of those policies. By the same token, the attachment-point notice inquiry, running in parallel to the liability inquiry, becomes more complex as well. Brooklyn Union may (and does) disagree with this court's conclusion about how to apply the retention/per-occurrence limit language of the policies. But, taking that conclusion as given, Brooklyn Union does not provide a basis under which it may be properly applied only to the liability inquiry, and not to both liability and notice.

B. How to Determine the Total Period of Injury for Allocation Purposes

The parties have also disputed how to determine the total period of injury across which Brooklyn Union's covered losses must be allocated. At trial, Brooklyn Union and Century each proposed endpoints for the periods of environmental damage at the three MGP sites—i.e. , when the damage began, and when it ended (or will end). The parties agreed that the choice among those different endpoints, and thus the choice among different-length periods of damage, was for the jury. They disagreed, though, about how to guide the jury in making that decision. In particular, Brooklyn Union contended that the jury should be instructed that Century bears the burden of proof in establishing its proposed period of injury. Century argued that neither side bears a burden of proof, and that the jury should be instructed simply to choose appropriate endpoints based on the evidence presented at trial. This court agrees with Century.

Here, it is undisputed that Brooklyn Union has proven at trial that property damage triggering the Century policies happened throughout the overall Century coverage period of 1941 to 1969. It is also undisputed that Brooklyn Union does not have coverage under the Century policies for any property damage that happened before 1941 or after 1969. And under time-on-the-risk allocation, Century is responsible for that fraction of Brooklyn Union's total losses of which the period of coverage is the numerator and the overall period of injury is the denominator. (See Century Indem. , 74 Misc 3d at 736 n 5.)

Ascertaining the length of that overall period of property damage determines the proportion of the policyholder's total losses that is potentially covered due to having happened during the policy period. In other words, the question is what losses are—and are not—potentially covered under the policy. And it is ordinarily the policyholder , not the insurer, that bears the burden to prove that losses are covered to begin with. (See Consolidated Edison , 98 NY2d at 218.)

Brooklyn Union contends that Century nonetheless bears the burden of proof with respect to the total duration of injury, because as that duration grows, the share of Century's potential liability for Brooklyn Union's losses shrinks. (See NYSCEF No. 1211 at 2-3.) But Brooklyn Union does not provide authority for this contention. To the contrary, precedent teaches that an insurer does not, as Brooklyn Union suggests, bear the burden of proof on an issue merely because the resolution of that issue may decrease the insurer's liability.

For example, resolving a question about the size of a policy's coverage limit may substantially reduce the amount for which the insurer can be held liable. Yet as discussed above, the existence and scope of these limits goes to the initial presence or absence of coverage, making it an issue on which the policyholder, not the insurer, bears the burden of proof. (See Presbyterian Hospital , 220 AD2d 733, 733 [2d Dept 1995] ; XL Ins. Am., Inc. v Howard Hughes Corp. , 154 AD3d 505, 506-507 [1st Dept 2017].) Similarly, Insurance Law § 3420 requires a prompt notice of disclaimer in bodily injury cases for disclaimer grounds that would "bar coverage or implicate policy exclusions." ( Pav-Lak Indus., Inc. v Arch Ins. Co. , 56 AD3d 287, 288 [1st Dept 2008].) The Court of Appeals has held that this prompt-disclaimer requirement does not apply to an argument by the insurer for applying pro rata allocation and thereby reducing the insurer's personal-injury coverage liability: That argument "do[es] not relate to an argument of exclusion or disclaimer, but rather, focus[es] on the extent of alleged liability under the various policies." ( Roman Catholic Diocese of Brooklyn , 21 NY3d at 147.)

Given this analysis, this court disagrees with Brooklyn Union that Century should bear a burden of proof with respect to the length of the overall period of injury. This court does not, however, conclude that Brooklyn Union should bear a burden of proof. The policyholder does ordinarily bear the burden of proof on questions going to the extent of coverage. But this is not an ordinary scenario.

The trial record contains sufficient evidence to support a wide range of starting and ending dates for the underlying property damage at issue here. The question for the jury was simply to choose what it found to be the proper endpoints within that wide range. As Century argued at trial, it does not make sense to think that one side or the other bore the burden of proof to establish that a particular choice of endpoints was warranted. The jury inquiry here differs from the more typical binary, yes/no jury question for which it might make sense to speak in terms of burdens of proof.

Brooklyn Union provides no New York caselaw to support its argument that a burden of proof should be imposed with respect to the particular jury question at issue here—much less that the burden should be imposed on the insurer, in particular. Instead, Brooklyn Union relies on six decisions from other jurisdictions, applying the law of other states. (See NYSCEF No. 1211 at 3 [collecting cases].) None of those cases is apposite.

Harlor v Amica Mutual Insurance Co. (150 A3d 793, 801-802 [Me 2016] [applying Maine law] ) and Peabody Essex Museum, Inc. v U.S. Fire Insurance Co. (2010 WL 3895172, at *3, *6-*7 [D Mass Sept. 30, 2010] [applying Massachusetts law] ) each involved applications of the rule that an insurer that breaches its duty to defend then bears the burden to prove that damages for which the policyholder was held liable in the underlying action were outside the scope of the policy. That rule derives in both Maine and Massachusetts from the holding of the Massachusetts Supreme Judicial Court in Polaroid Corp. v Travelers Indemnity Corp. (414 Mass 747 [1993] ). The Polaroid Court emphasized, in reaching that holding, that "delay in determining an insurer's defense obligation and delay in the settlement of the claim could make it more difficult for an insured to prove that the underlying claim that it settled fell within policy coverage." ( Id. at 764.) This "information asymmetry rationale" is not implicated by the circumstances here. ( Peabody Essex , 2010 WL 3895172, at *6.)

Aerojet-General Corp. v Transport Indemnity Co. (17 Cal 4th 38, 69-71 [Cal 1997] ) holds that an insurer may obtain reimbursement from its insured for defense costs incurred by the insurer, if the insurer establishes that those costs were incurred with respect to injuries that necessarily were sustained outside the policy period. The California Supreme Court's ruling that the insurer bears the burden of proof flows from the Court's view of the right to reimbursement as a claim sounding in quasi-contract: I.e. , because it is the insurer seeking monetary relief from the insured in quasi-contract, the insurer, like any other party seeking relief, bears the burden of proof. (See id. at 69 ; see also Buss v Superior Ct. of Los Angeles County , 16 Cal 4th 35, 50-51, 53 [Cal 1997].) That principle has no application to this case.

Insurance Co. of North America v Kayser-Roth Corp. (770 A2d 403, 413-414 [RI 2001] ) dealt with an insurer's right to obtain a setoff of losses that it had been held liable to pay, in order to account for prior settlements by other insurers that were concurrently responsible for the same losses. In other words, Kayser-Roth concerned the apportionment among insurers of responsibility for losses after they have been found to be covered—not, as here, an apportionment as between insurer and insured of losses as covered or not covered.

Thomson Inc. v Insurance Co. of N. Am. (11 NE3d 982 [Ind Ct App 2014] ), and the other Indiana cases it cites on this point, hold that the insurer "bears the burden of proof as to any basis for avoiding coverage " for a loss—not that the insurer bears the burden of proof as to any basis for limiting the extent of the insurer's liability for a covered loss. ( Id. at 994-995 [emphasis added]; cf. Roman Catholic Diocese of Brooklyn , 21 NY3d at 147 [distinguishing between defenses based on "an argument of exclusion or disclaimer" of coverage and defenses that "focus on the extent of alleged liability" under the policies at issue].)

Finally, Domtar, Inc. v Niagara Fire Ins. Co. (563 NW2d 724 (Minn 1997]) addresses a different pro rata allocation scenario from that presented by this action. In Domtar , the Minnesota Supreme Court explained that the policyholder bears the initial burden of proving that the beginning and ending dates of the underlying property damage encompass the period of the policy or policies for which the policyholder is seeking coverage, and that if the policyholder meets that burden, the trial court should presume that property damage was continuous from the beginning date to the ending date. ( Id. at 732.) If the insurer then contends that "no appreciable damage occurred during a particular time period" within the resulting range of time, the insurer bears the burden of proof on that point. (See id. ; see also Northern States Power Co. v Fidelity & Cas. Co. of NY , 523 NW2d 657, 663-664 [Minn 1994] [NSP ] [articulating this three-step framework].)

Domtar and NSP put the burden of proof on the insurer to disprove coverage in a particular time period (on the ground that no injury happened that would trigger the policy) only once the court has presumed, based on the policyholder's factual showing, that coverage exists during that period—just as an insurer bears the burden of proving an exclusion exists that defeats otherwise-applicable coverage. Here, on the other hand, Century is not seeking to disprove the existence of coverage on the ground that the policies at issue were not triggered to begin with. Relatedly, the disagreement between the parties is not about whether property damage did or did not occur for a particular subperiod within the established temporal endpoints of that damage. The parties are instead disputing what those damage-period endpoints are , because that will affect the extent of Century's liability for the covered losses that occurred. Domtar and NSP do not provide any guidance on that inquiry.

In short, Brooklyn Union's out-of-state cases do not persuade this court to impose a burden of proof in the unusual context presented here—nor to impose that burden on Century, in particular.