Opinion
No. E037627.
December 28, 2006. CERTIFIED FOR PARTIAL PUBLICATION Modified January 12, 2007. Modified January 25, 2007. REVIEW GRANTED April 18, 2007
Pursuant to California Rules of Court, rules 8.1105(b) and 8.1110, this opinion is certified for publication with the exception of parts II.A. through C, II.D.1., II.D.3., and II.D.4.
Appeal from the Superior Court of Riverside County, Nos. CIV239784 and RIC381555, Erik Michael Kaiser, Judge.
Cotkin, Collins Ginsburg, Roger W. Simpson, David W. Johnson, Jr.; Bill Lockyer, Attorney General, Darryl L. Doke and Jill Scally, Deputy Attorneys General; Law Offices of Daniel J. Schultz, Daniel J. Schultz; Anderson Kill Olick, Robert M. Horkovich and Edward J. Stein for Plaintiff and Appellant. Gauntlett Associates, David A. Gauntlett and Eric R. Little as Amicus Curiae on behalf of Plaintiff and Appellant. Berkes Crane Robinson Seal, Steven M. Crane, Barbara S. Hodous; Nixon Peabody, Bruce E. Copeland, Alan S. Feiler; Berman Aiwasian, Alan S. Berman, Steven R Haskell; Riedl, McCloskey Waring and Andrew McCloskey for Defendants and Respondents. Wiley Rein Fielding, Laura A. Foggan; Sinnott, Dito, Moura Puebla, Randolph R Sinnott and John J. Moura as Amicus Curiae on behalf of Defendants and Respondents.
OPINION
This is a coverage dispute between the State of California (the State) and four of its liability insurers. The insurers are Allstate Insurance Company, Century Indemnity Company, Columbia Casualty Company, and Westport Insurance Corporation, to whom we shall refer collectively as Insurers. The dispute concerns whether Insurers are required to indemnify the State against liability for damage caused to third parties by the discharge of pollutants from the State's "Stringfellow Acid Pits" waste disposal site. The trial court granted summary judgment in favor of Insurers, based on exclusions in their policies for liability based on pollution and on the discharge of pollutants into a watercourse. The State contends Insurers are estopped from asserting the pollution exclusion, and at any rate neither that exclusion nor the watercourse exclusion excludes coverage here. We reverse the summary judgment, because we conclude the record raised a triable issue whether the State sustained liability for discharge of pollutants that fell within the "sudden and accidental" exception to the pollution exclusion and did not fall within the watercourse exclusion.
I
FACTUAL AND PROCEDURAL BACKGROUND
A. The Site In 1956, the State opened a class I hazardous waste site (Stringfellow Site) near Glen Avon in Riverside County. The State's geologist, who investigated the site to determine whether it was suitable, did no soil analysis; he assumed the site was underlain by impermeable rock and there was no water in the bedrock or granite. In fact, there were two buried alluvial channels, and water was moving through bedrock that consisted of decomposed granite and broken rock. A class I site had to be impermeable or underlain by unusable water. The State designed the site, which included a concrete barrier dam eight feet high at the downstream (southern) boundary of the site, diversion channels, and ponds. The State admits it negligently investigated, selected, designed, and supervised the construction of the site, failing to ensure adequate diversion channels and other safeguards to prevent or protect against heavy rains. The Stringfellow Site operated for about 16 years. During that time, with the knowledge and consent of the State, more than 30 million gallons of liquid industrial wastes were deposited directly into unlined evaporation ponds at the site. B. Discharges of Pollutants from the Site
This information, and other information in our statement of the facts, comes from the report of a special master appointed to make findings of fact in U.S. v. Stringfellow (C.D.Cal., Nov. 30, 1993, No. CV83-2501 JMI) 1993 WL 565393. The special master was appointed by the court in that case to conduct a hearing to determine the State's liability for pollution caused by the escape of wastes from the Stringfellow Site. His report, dated November 30, 1993, is almost 500 pages, including appendices, and was submitted by Insurers in support of their summary judgment motions. The district court, in an unpublished decision in 1995, adopted the special master's findings, conclusions, and recommendation as modified. ( United States v. Stringfellow (C.D.Cal. Jan. 24, 1995, No. CV-83-2501) 1995 WL 450856 at pp. *1, *6.) The facts we recite from the report appear not to be disputed for purposes of this appeal.
1. Subsurface discharges
According to a report prepared by an expert for the State, by 1960 contaminants exited the subsurface of the site around the east and west ends of the concrete barrier and through the fractured bedrock underneath the barrier. From the moment the contaminants left the site, the soils and groundwater became contaminated progressively farther downstream of the site due to the continuous motion of the groundwater to the southwest. Through at least the late 1980's, the plume of contamination was moving progressively farther from the site. As of the date of the report, July 2004, damage to the soils and groundwater downgradient of the site was ongoing.2. 1969 discharge
For purposes of the summary judgment motion from which this appeal arises, the parties agreed to the following facts. According to rainfall records, it rained heavily in January and February of 1969, with nearly seven inches of rainfall in January and eight inches in February. Not later than March 17, 1969, a once-in-50-year rainstorm of some 20 inches inundated the site, causing the contaminants at the site to overflow into the surrounding environment, including the City of Glen Avon. In November 1972, the State found contamination in the groundwater, and the site was closed. No later than January 1973, signs of leaking were observed at the site. The leakage was worse by 1975. A 1974 report by the State's chief geologist recommended (1) a hydraulic barrier to capture waste flowing out of the site in the subsurface, to protect groundwater; and (2) leveling the site and putting an impervious cap on it, to prevent overflow in case of rain.
3. 1978 discharges
By the beginning of the 1978-1979 rainy season, the recommended measures had not been taken. After heavy rains in early 1978, all of the ponds at the site were full. On March 5, 1978, they began to overflow. The State decided to make a "controlled discharge" of waste from the site. The waste from the controlled discharge went directly into Pyrite Creek and from there across a roadway, down a channel, across a street just below a school, and into the Santa Ana River. Three days after the first controlled discharge, a section of the dam had given way and was moving, and there was a 50-foot crack in the dam as well. To prevent the failure of the dam, the State made another controlled discharge, again discharging waste directly into Pyrite Creek and affecting areas as much as six miles downstream from the site. The two controlled discharges in March 1978 released more than one million gallons of rain-diluted waste into the environment. In addition, during later rains the contaminants in the downstream surface soils repeatedly migrated and further damaged the environment. By December 1979, the contaminant plume had reached a street in the adjacent community. The release of waste in 1978 would not have occurred if the State had installed the hydraulic barrier and cap. C. The Federal Action In 1983, the United States of America and the State brought a civil action in federal district court (the federal action) against companies that had disposed of waste at the Stringfellow Site. ( U.S. v. Stringfellow, supra, No. CV83-2501 JMI.) The companies counterclaimed against the State for damages caused by progressive environmental contamination occurring at and emanating from the site. In September 1998, the court in the federal action held the State 100 percent liable for past and future costs of remediating the contamination. According to the State, the costs exceed $500 million. The State alleges it paid $99.4 million and received from the counterclaimants a waiver of an estimated $100 million in return for dismissing its appeal from the judgment in the federal action. D. The Policies After the site was closed, but before the 1978 discharges, the State purchased comprehensive general liability (CGL) excess insurance policies from Insurers. The terms of the policies varied, but together they provided coverage from September 1976 to May 1978. Although the policies were purchased after the site ceased operations, Insurers have not argued for purposes of summary judgment or this appeal that there is no coverage for that reason. While the language of the policies varied, Allstate's, Century's, and Westport's were in the same form and said essentially the same thing. For convenience, in this opinion we will quote the policy issued by Allstate's predecessor, which is sufficiently representative of the other two policies for our purposes. The coverage clause of the policy obligates the insurer "[t]o pay on behalf of the Insured all sums which the Insured shall become obligated to pay by reason of liability imposed by law, including Chapter 1681 of the State of California Statutes of 1963, or liability assumed by contract, insofar as the State may legally do so, for damages, including consequential damages, because of direct damage to or destruction of tangible property (other than property owned by the Insured), including the loss of use thereof, which results in an Occurrence during the policy period." An "Occurrence" is "an accident, event or happening including continuous or repeated exposure to conditions which results, during the policy period, in . . . Property Damage neither expected nor intended from the standpoint of the Insured." The policy also contains a "pollution exclusion." The exclusion states that the policy does not apply to damage "arising out of the discharge, dispersal, release or escape of smoke, vapors, soot, fumes, acids, alkalis, toxic chemicals, liquids or gases, waste materials or other irritants, contaminants or pollutants into or upon land or the atmosphere, but this exclusion does not apply if such discharge, dispersal[,] release or escape is sudden and accidental." (Italics added.) The Columbia policy also contains this exclusion, with the modification noted below. Finally, the policy contains a "watercourse exclusion," which states: "It is further agreed that the Policy does not apply to Personal Injury or Property Damage arising out of the discharges, dispersal, release or escape of smoke, vapors, soot, fumes, acids, alkalis, toxic chemicals, liquids or gases, wate [ sic] materials or other irritants, contaminants or pollutants into or upon any watercourse or body of water." The watercourse exclusion is set forth in a separate paragraph and does not contain the exception for sudden and accidental discharges. Columbia's policy differs from the standard policy in two ways that should be noted here. First, instead of the coverage clause quoted above, the Columbia policy merely states that the insurer will "indemnify the insured for the amount of loss which is in excess of the applicable limits of liability of the underlying insurance" to which the policy is excess. "Loss" is defined as "the sums paid as damages in settlement of a claim or in satisfaction of a judgment for which the insured is legally liable," minus certain deductions. The policy states it is excess to "Underwriters at Lloyds, London Policy #TBA," but there is no policy number filled in. The parties stipulated that this copy of the Columbia policy was "the most complete evidence" of the policy that was available. The record contains no subsequent stipulation supplying the number of the underlying policy, nor does it contain a copy of the underlying policy. Therefore, we are unable to determine what, if any, effect the terms of the underlying policy might have on the coverage provided by the Columbia policy. The second difference between the standard policy and the Columbia policy is that the Columbia policy does not contain a separate paragraph setting forth the watercourse exclusion. Instead, the pollution exclusion provides that it also applies to discharges into or upon "any watercourse or body of water. . . ." The effect is that Columbia's watercourse exclusion is subject to the same "sudden and accidental" exception as is the rest of its pollution exclusion. E. The Present Action The State filed the present action in September 2002. The operative complaint, the first amended complaint (hereafter the complaint), alleged claims for declaratory relief, breach of contract, and breach of the implied covenant of good faith and fair dealing. Although the complaint named about 30 insurance companies, Insurers are the only defendants involved in this appeal. In the complaint, the State alleged that it notified Insurers of the federal action and the judgment against it and demanded that Insurers indemnify it against any liability arising from the judgment, but Insurers denied coverage.
We note the district court's unpublished 1995 decision states that the court found the State 65 percent liable on claims asserted under federal law and 100 percent liable on claims asserted under state law. ( United States v. Stringfellow, supra, 1995 WL 450856 at p. *6.)
We are concerned here only with the coverage for property damage (coverage "B" in the policy), because there is no claim that the State was held liable for any personal injury (coverage "A" in the policy).
In 1985, the insurance industry modified the standard general liability policy to delete the "sudden and accidental" exception from the pollution exclusion. ( MacKinnon v. Truck Ins. Exchange (2003) 31 Cal.4th 635, 644 [3 Cal.Rptr.3d 228, 73 P.3d 1205].)
Insurers claim the Columbia policy "follows form to the same language as do the other policies," and that "the policy stipulation simply did not physically attach the underlying language." This court, of course, is limited to the record on appeal. Insurers did not move to admit the missing language as evidence on appeal. (Cal. Rules of Court, former rule 22(c) (see now rule 8.252(c)).)
F. The Trial Court's Grant of Summary JudgmentIn October 2004, Insurers moved for summary judgment or summary adjudication of issues based on the pollution exclusion and the watercourse exclusion. The court heard the motions in November 2004. The court accepted Insurers' argument that the 1969 and 1978 releases were excluded from coverage because the event that triggered the coverage was the initial deposit of wastes into the site, which was neither sudden nor accidental. The later escape of pollutants from the site was not a covered event and therefore could not provide a basis for coverage even if it were "sudden and accidental." The court also found that the watercourse exclusion precluded coverage. Finally, the court found that the State was required to allocate damage between covered and uncovered causes and had failed to do so. In December 2004, the court granted summary judgment in favor of all four Insurers. The State erroneously appealed from the order granting summary judgment instead of the judgment itself, but we construe the notice of appeal to be from the judgment. ( H.N. Frances C. Berger Foundation v. City of Escondido (2005) 127 Cal.App.4th 1, 6-7, fn. 5 [ 25 Cal.Rptr.3d 19].)
II DISCUSSION
A. Burden of Proof and Standard of Review A party moving for summary judgment "bears the burden of persuasion that there is no triable issue of material fact and that he is entitled to judgment as a matter of law." ( Aguilar v. Atlantic Richfield Co. (2001) 25 Cal.4th 826, 850, fn. omitted.) A defendant moving for summary judgment "must present evidence that would require a reasonable trier of fact not to find any underlying material fact more likely than not," thus establishing that the plaintiff could not prevail at a trial. ( Id. at pp. 851.) An insurer moving for summary judgment based on the pollution exclusion in a case involving the duty to indemnify "has the additional burden . . . of making an affirmative showing that the insured cannot establish that the claims fall within the sudden and accidental exception to the pollution exclusion." ( Travelers Casualty Surety Co. v. Superior Court (1998) 63 Cal.App.4th 1440, 1456.) Finally, in construing an insurance policy, "exclusionary provisions are narrowly interpreted while exceptions are broadly construed. . . ." ( E.M.M.I. Inc. v. ZurichAmerican Ins. Co. (2004) 32 Cal.4th 465, 476.) Accordingly, the "sudden and accidental" exception to the pollution exclusion "will be construed broadly in favor of the insured. [Citations.]" ( Aydin Corp. v. First State Ins. Co. (1998) 18 Cal.4th 1183, 1192.) B. Contentions on Appeal; Summary of ConclusionsThe State contends the court erred in granting summary judgment for the following reasons:(1) Insurers are estopped from asserting that the pollution exclusion should be interpreted to bar coverage in this case, because the insurance industry represented to state regulators that the exclusion would not be interpreted that way. (2) Even if Insurers are not estopped, the pollution exclusion does not apply because the discharges of pollutants in this case fell under the "sudden and accidental" exception. (a) "Sudden and accidental" only means that the discharge of pollutants must be unexpected and unintended, not that it must be abrupt in a temporal sense. (b) Even if "sudden and accidental" means that the discharge must be abrupt, the 1969 and 1978 discharges satisfied that requirement. (3) The watercourse exclusion does not apply because the 1969 and 1978 discharges were not confined to Pyrite Creek. (4) The State is not required to differentiate the damages caused by covered discharges of pollutants from those caused by uncovered causes in order to obtain coverage of its liability for the damages. We conclude solely for purposes of the summary judgment motion from which this appeal arises: (1) Insurers are not estopped from invoking the pollution exclusion to deny coverage. (2) There is evidence from which a reasonable trier of fact could find the 1969 discharge was "sudden and accidental" but insufficient evidence to support such a finding as to the other discharges. (3) There was evidence from which a reasonable trier of fact could conclude the watercourse exclusion does not apply. (4) The State is not required to allocate its damages to obtain indemnity for any covered liability. C. Estoppel 1. Background One of the principal issues in this case, and a matter of substantial controversy in insurance law generally, is whether, in the "sudden and accidental" exception to the 1970 pollution exclusion, "'the word "sudden" was intended to be given a strictly temporal meaning such that, in order for the exception to apply, the discharge of pollution had to have been "abrupt." [Citation.]'" ( MacKinnon v. Truck Ins. Exchange, supra, 31 Cal.4th at p. 644) The issue is significant here because the State concedes that if "sudden" means "abrupt," it is entitled to coverage only for the 1969 and 1978 discharges and not for the discharge of contaminants that occurred gradually over the years. While the California Supreme Court has observed that the proper interpretation of "sudden and accidental" has "'generated an enormous amount of litigation'" ( ibid.), that court has not addressed the issue itself. In the meantime, the Courts of Appeal have held unanimously that "sudden" is to be given a "temporal" meaning, and therefore the "sudden and accidental" exception only covers liability for "abrupt" discharges of pollutants. ( Shell Oil Co. v. Winterthur Swiss Ins. Co. (1993) 12 Cal.App.4th 715, 754 ["'sudden' necessarily contains a temporal element in addition to its connotation of the unexpected"]; ACL Technologies, Inc. v. Northbrook Property Casualty Ins. Co. (1993) 17 Cal.App.4th 1773, 1779 ["sudden" does not allow for "liability coverage arising from gradual pollution"]; Truck Ins. Exchange v. Pozzuoli (1993) 17 Cal.App.4th 856, 861 ["[t]he trial court's conclusion that the word 'sudden' has a temporal meaning was correct"]; Standun, Inc. v. Fireman's Fund Ins. Co. (1998) 62 Cal.App.4th 882, 889 ["'[s]udden' has a temporal element and does not mean a gradual or continuous discharge"]; Travelers Casualty Surety Co. v. Superior Court, supra, 63 Cal.App.4th at p. 1458 ["California courts have uniformly ruled that the language of the sudden and accidental exception is unambiguous and its plain language meaning is 'abrupt,' 'unintended, and unexpected'"].) The State acknowledges these decisions and the rule they enunciate, that "sudden" means abrupt, and not just unexpected. It argues, however, that even if that rule is a correct statement of California law, Insurers are estopped from invoking the rule because the insurance industry represented to insurance regulators that the exception would not eliminate coverage for gradual pollution. Rather, the State asserts, the industry represented that to be covered, pollution damage need only be "unexpected and unintended," not "sudden" in the sense of abrupt. The State thus seeks to invoke the doctrine of regulatory estoppel. The State further argues that at least Allstate, if not all four Insurers, should be estopped from contending the "sudden and accidental" exception does not cover gradual pollution damage. The State asserts Allstate previously argued in a federal court action in Massachusetts that the pollution exclusion did not exclude coverage for pollution that had occurred from seepage over more than five years, as long as the damage was "unexpected and unintended" by the insured. The State thus seeks to invoke the doctrine of judicial estoppel. In response, Insurers assert the State waived any estoppel claim by not arguing estoppel in the trial court and at any rate failed to provide adequate evidentiary support for an estoppel claim. We consider each form of estoppel in turn. 2. Regulatory estoppel a. Elements As far as we can tell, there is no published California decision discussing regulatory estoppel. The State accordingly relies on four out-of-state decisions. In what appears to be the leading case on the doctrine, Morton Intern., Inc. v. General Acc. Ins. Co. (1993) 134 N.J. 1, [ 629 A.2d 831], the court acknowledged that "an interpretation of 'sudden' that does not acknowledge its temporal quality is unfaithful to its core meaning. . . ." ( Id. at p. 71, italics added.) However, the court noted that before the exclusion was added to the standard policy, the policy generally was understood "'to cover pollution liability that arose from gradual losses[.]'" ( Id. at p. 32.) In fact, the Mutual Insurance Rating Bureau (MIRB), an insurance industry group, had stated in a memorandum to the New Jersey Department of Banking Insurance that the standard policy as drafted in 1966 eliminated "' the connotation of suddenness previously intended as respects coverage on an "accident" basis.'" ( Id. at p. 38.) Moreover, the Morton court noted, when the insurance industry had proposed the pollution exclusion in 1970, the Insurance Rating Board (IRB) and the MIRB "apparently" submitted to "most if not all states in which approval was sought" — including New Jersey — a standard explanatory memorandum stating that under the proposed exclusion, "'[c]overage is continued for pollution or contamination caused injuries when the pollution or contamination results from an accident. . . . '[Citation.]" ( Morton Intern., Inc. v. General Acc. Ins. Co. of America, supra, 134 N.J. at pp. 35-36, italics added.) The memorandum said nothing to the effect that besides being accidental, the pollution had to be "sudden" in order to be covered. In view of this history, the Morton court concluded: "Although we have not heretofore applied the estoppel doctrine in a regulatory context, its application to these circumstances is appropriate and compelling." ( Morton Intern., Inc. v. General Acc. Ins. Co. of America, supra, 134 N.J. at p. 75.) Therefore, the court stated, it would construe the pollution exclusion to preclude liability only where the insured "intentionally discharged, dispersed, released, or caused the escape of a known pollutant." ( Id. at p. 31; see also id. at p. 78.) In the next case the State cites, Claussen v. Aetna Cas. Sur. Co. (1989) 259 Ga. 333 [ 380 S.E.2d 686], the Georgia Supreme Court construed the phrase "sudden and accidental" to mean "unexpected and unintended," so that the pollution exclusion would not preclude coverage for "liability for environmental contamination caused by the discharge of pollutants over an extended period of time." ( Id. at pp. 333, 338.) The court relied, in part, on its conclusion that "[d]ocuments presented by the Insurance Rating Board (which represents the industry and on which Aetna participated) to the Insurance Commissioner when the 'pollution exclusion' was first adopted suggest that the clause was intended to exclude only intentional polluters." ( Id. at p. 337.) In Joy Technologies, Inc. v. Liberty Mut. Ins. Co. (1992) 187 W.Va. 742 [ 421 S.E.2d 493], the West Virginia Supreme Court construed the pollution exclusion to exclude only pollution damage that was expected or intended, "even if it resulted over a period of time and was gradual. . . ." The court relied, in part, on evidence that "the insurance group representing Liberty Mutual unambiguously and officially represented to the West Virginia Insurance Commission that the exclusion in question did not alter coverage under the policies involved, coverage which included the injuries in the present case. . . ." ( Id. at p. 749.) In addition, the West Virginia Insurance Commissioner had submitted an affidavit stating that in approving the exclusion he had relied on representations of the insurers that "'the proposed endorsement forms did not limit or narrow coverage and were not intended to do so.," ( Id. at p. 748.) Finally, in Sunbeam Corp. v. Liberty Mut. Ins. Co. (2001) 566 Pa. 494 [ 781 A.2d 1189], insureds who had been held liable for gradual pollution damage argued in their action for coverage that "sudden and accidental" should be interpreted to mean "unexpected and unintended" rather than "abrupt." ( Id. at p. 499.) The complaint alleged "that in 1970 the insurance industry, including the defendant insurers, submitted to the Pennsylvania insurance department a memorandum which asserted that the disputed language — excluding coverage for pollution unless it was 'sudden and accidental' — would not result in any significant decrease in coverage." ( Id. at p. 499.) The court held these allegations adequately pled a claim for coverage based on "regulatory estoppel, a form of judicial estoppel." ( Ibid.) b. Waiver "As a general rule, theories not raised in the trial court cannot be asserted for the first time on appeal, out of fairness to both the trial court and the opposing parties. [Citation.]" ( Howard S. Wright Construction Co. v. BBIC Investors, LLC (2006) 136 Cal.App.4th 228, 242, fn. 13.) Here, at no time did the State advise the court or Insurers that it was claiming regulatory estoppel. The State asserted in opposition to summary judgment that "when seeking regulatory approval of the pollution exclusion, insurers informed regulators that the language was intended only to exclude coverage for expected or intended pollution. . . ." However, nowhere did the State ever use the term "estoppel," nor did it cite Morton Intern., Inc. v. General Acc. Ins. Co. of America or any of the other regulatory estoppel cases. It was not the court's obligation to divine on its own that the State might be trying to argue regulatory estoppel. A court "'is not required to make an independent, unassisted study'" to try to find grounds for a party's position ( Guthrey v. State of California (1998) 63 Cal.App.4th 1108, 1115), nor will a court "develop the [parties'] arguments for them. . . ." ( Dills v. Redwoods Associates, Ltd. (1994) 28 Cal.App.4th 888, 890, fn. 1.) Moreover, the rule against asserting a new theory on appeal "is especially applicable to the doctrine of estoppel, which includes factual elements that must be established in the trial court. [Citation.]" ( Honig v. San Francisco Planning Dept. (2005) 127 Cal.App.4th 520, 530.) As we have seen, a claim of regulatory estoppel necessarily "includes factual elements," because the doctrine applies where insurance industry representations have induced state regulators to approve the policy provision at issue. Thus, Morton Intern., Inc. v. General Acc. Ins. Co. of America and the other regulatory estoppel cases on which the State relies all have in common the fact that in each case, the party seeking estoppel presented evidence of specific representations to state regulators in the state in which the case was pending. The State presented no comparable evidence here that the insurance industry made any representations to California regulators concerning the effect of the "sudden and accidental" exception. Instead, the State offered only a hearsay declaration and unauthenticated exhibits indicating that the MIRB made representations to regulators in New York. Even those documents are not properly before us, because the court granted Insurers' objections to them, and the State has not challenged those rulings. ( Johnson v. City of Loma Linda (2000) 24 Cal.4th 61, 65-66; Alexander v. Codemasters Group Limited (2002) 104 Cal.App.4th 129, 140.) Consequently, we lack an adequate record to rule on regulatory estoppel even if we were inclined to overlook the State's failure to timely assert it. In addition, Insurers represent to this court that there is substantial evidence they could and would have presented to the trial court to show that California regulators did not rely on any insurance industry representations about the effect of the "sudden and accidental" exception. They did not present this evidence because the State never claimed regulatory estoppel. It is unfair to consider a claim on appeal when the parties had no incentive to fully litigate the theory. ( People v. Storm (2002) 28 Cal.4th 1007, 1028, fn. 10.) The State contends it was not required to show what representations were made to California regulators, or by whom, but may rely simply on evidence that the insurance industry in general understood, and represented publicly, that "sudden and accidental" means "unexpected and unintended," not abrupt in a temporal sense. The State cites Montrose Chemical Corp. v. Admiral Ins. Co. (1995) 10 Cal.4th 645 ( Montrose) and MacKinnon v. Truck Ins. Exchange, supra, 31 Cal.4th 635 ( MacKinnon) as examples of cases in which the Supreme Court relied on drafting history to interpret policy provisions without finding that the specific insurers involved had ever discussed or given a particular meaning to the policy provisions at issue. Montrose and MacKinnon stand only for the proposition that a court may consider drafting history in interpreting an insurance policy. Montrose recognized that interpretative literature could be helpful in "construing standardized insurance policy language." ( Montrose, supra, 10 Cal.4th at p. 671.) MacKinnon, similarly, recited the history of the pollution exclusion in deciding how to interpret the 1985 version of the exclusion. ( MacKinnon, supra, 31 Cal.4th at pp. 643-645.) The State makes clear it is not relying on drafting history to interpret the policies at issue, but rather as a basis for claiming regulatory estoppel. Montrose and MacKinnon do not support the use of drafting history for that purpose, and as discussed, ante, the out-of-state cases that apply regulatory estoppel all involved representations made to regulators in the forum state, not merely generalized evidence of drafting history. At any rate, the general drafting history of the pollution exclusion, as reflected in secondary sources, in inconclusive at best. (Compare Salisbury, Pollution Liability Insurance Coverage, the Standard-Form Pollution Exclusion, and the Insurance Industry: A Case Study in Collective Amnesia (1991) 21 Environmental Law 357, 372, 391-392 ( Salisbury) [industry took position in 1970 that coverage would not be reduced by pollution exclusion, because "sudden" meant only "unexpected and unintended," but later took opposite view when claims for coverage were made] with Zampino, et al., Morton International: The Fiction of Regulatory Estoppel (1993) 24 Seton Hall L. Rev. 847, 861 [quoting New York insurance regulator as stating that the exclusion's temporal restriction on coverage was "clearly communicated" to regulators, who "were not in the slightest bit misled" about the effect of the provision].) Finally, the vitality of the regulatory estoppel doctrine as a general principle of law is dubious. A federal district court writing in 1996 — after all but one of the four regulatory estoppel cases on which the State relies — stated: "The regulatory estoppel argument has been rejected by virtually every other state and federal court to address the issue. [Citations.]" ( SnyderGeneral Corp. v. Great American Ins. Co. (N.D. Tex. 1996) 928 F.Supp. 674, 682.) Two years later, another federal district court similarly stated: "Since Morton, a number of courts have faced similar arguments based on claims of 'regulatory estoppel' or 'regulatory fraud.' Most courts have rejected these claims. [Citations.]" ( Wysong and Miles Co. v. Employers of Wausau (M.D.N.C. 1998) 4 F.Supp.2d 421, 427.) This case, especially given the lack of an adequate factual record, is not an appropriate one in which to decide whether California should adopt the regulatory estoppel doctrine. We therefore decline to consider the State's regulatory estoppel claim. 3. Judicial estoppel a. Elements Judicial estoppel, unlike regulatory estoppel, is well established in California. The doctrine "prohibits a party from taking inconsistent positions in the same or different judicial proceedings. [Citation.]" ( M. Perez Co., Inc. v. Base Camp Condominiums Assn. No. One (2003) 111 Cal.App.4th 456, 463.) Judicial estoppel is intended to prevent a litigant from playing "fast and loose" with the courts by first advocating one position, and later, if it becomes beneficial, advocating the opposite one. ( Ibid.) b. Waiver The State did not cite in the trial court any case discussing judicial estoppel. Moreover, as with its regulatory estoppel claim, the State failed to present an adequate factual record for a claim of judicial estoppel. It offered only a hearsay declaration claiming that in the Massachusetts federal court case, Allstate had argued the "sudden and accidental" exception covered liability for gradual pollution and an unauthenticated purported excerpt from Allstate's brief in that case. The court sustained Insurers' objections to the declaration and exhibit. Because the court sustained Insurers' objections, we have no evidence of what Allstate actually argued in the Massachusetts case concerning the "sudden and accidental" exception. Among the requirements for judicial estoppel are, first, that "the two positions are totally inconsistent," and second, that the party to be estopped "was successful in asserting the first position (i.e., the tribunal adopted the position or accepted it as true). . . ." ( Jackson v. County of Los Angeles (1997) 60 Cal.App.4th 171, 183.) Without any evidence of what Allstate argued in the Massachusetts case, we cannot determine whether its argument was inconsistent with its position in this case, or whether it prevailed in the Massachusetts court. The State cites the published opinion of the district court in the Massachusetts case, in which the court ruled that gasoline contamination from a pipe that leaked intermittently for six years was "sudden and accidental." ( Allstate Ins. Co. v. Quinn Const. Co. (D. Mass. 1989) 713 F.Supp. 35, 37, 41.) However, the opinion does not show what Allstate argued. Even if we could infer that Allstate must have argued in favor of the position the court took, the ruling was based on Massachusetts law and was made before the California decisions holding that "sudden" means abrupt had been decided. We cannot assume Allstate's argument in 1989 that a gasoline leak was "sudden and accidental" under Massachusetts law and its argument in 2004 that the release of contaminants from the Stringfellow Site was not "sudden and accidental" under California law are "totally inconsistent." ( Jackson v. County of Los Angeles, supra, 60 Cal.App.4th at p. 183.) The Massachusetts decision also fails to disclose whether Allstate ultimately prevailed. As Insurers point out, the original district court decision was vacated, and the action was dismissed with prejudice about a year later. ( Allstate Ins. Co. v. Quinn Const. Co. (D. Mass. 1990) 784 F.Supp. 927.) While the State argues there is no "hard and fast" rule that the party to be judicially estopped must have prevailed in the prior case, even the decision it cites for that proposition recognized that the majority of courts do require a showing of prior success. ( Thomas v. Gordon (2000) 85 Cal.App.4th 113, 118-119.) Further, the Thomas court departed from the majority rule only because it concluded the case presented a "rare" situation involving "egregious" dishonesty, in which judicial estoppel should apply even without a showing of prior success. ( Id. at p. 119.) The State has not shown Allstate was guilty of any comparable dishonesty. "It is axiomatic that it is the burden of the appellant to provide an adequate record to permit review of a claimed error, and failure to do so may be deemed a waiver of the issue on appeal. [Citations.]" ( People v. Akins (2005) 128 Cal.App.4th 1376, 1385.) Given the absence of an adequate factual basis for the State's judicial estoppel claim and the State's failure to argue the claim in the trial court, it would not be appropriate to consider the claim for the first time on appeal. We therefore decline to do so.
D. Application of the Pollution Exclusion to This Case1. Interpretation of the policy terms Having determined that Insurers are not estopped from contending that the pollution exclusion bars coverage, we address the coverage issue on the merits. As we have noted, Courts of Appeal hold unanimously that the use of "sudden" in the exception to the pollution exclusion means there is no coverage except for "abrupt" discharges of pollutants. ( Shell Oil Co. v. Winterthur Swiss Ins. Co., supra, 12 Cal.App.4th at p. 754; ACL Technologies, Inc. v. Northbrook Property Casualty Ins. Co., supra, 17 Cal.App.4th at p. 1779; Truck Ins. Exchange v. Pozzuoli, supra, 17 Cal.App.4th at p. 861; Standun, Inc. v. Fireman's Fund Ins. Co., supra, 62 Cal.App.4th at p. 889; Travelers Casualty Surety Co. v. Superior Court, supra, 63 Cal.App.4th at p. 1458.) Although we are not bound to follow these decisions, we agree with their reasoning, for the following reasons. The Supreme Court has repeatedly set forth the rules to be applied in interpreting insurance policies. These rules apply to the interpretation of standard form policies. ( Powerine Oil Co., Inc. v. Superior Court (2005) 37 Cal.4th 377, 391.) The court has stressed that the fundamental goal of contractual interpretation is to give effect to the mutual intention of the parties. That intent is inferred, if possible, solely from the written provisions of the contract. Only if the terms are ambiguous may the court resort to other methods of construction. ( Id. at p. 390; County of San Diego v. Ace Property Casualty Ins. Co. (2005) 37 Cal.4th 406, 415.) The State suggests the rule limiting a court to the policy terms unless they are ambiguous no longer applies, because in Montrose, supra, 10 Cal.4th 645 and MacKinnon, supra, 31 Cal.4th 635 the Supreme Court considered extrinsic evidence of drafting history without expressly finding the policy terms to be ambiguous. However, Powerine and County of San Diego were decided after Montrose and MacKinnon and demonstrate that the rule is alive and well. In fact, Montrose itself cited the rule. ( Montrose, at p. 666.) In addition, the Montrose court followed the rule. It found the policy language unambiguous and interpreted it on that basis alone. ( Montrose, supra, 10 Cal.4th at p. 668.) It merely found the drafting history to "confirm" its interpretation. ( Id. at p. 669). Further, the court said the drafting history was relevant in evaluating the insurer's argument that the insurance industry had never anticipated the interpretation the court adopted. ( Id. at p. 671.) Thus, the court only considered drafting history to determine a matter it obviously could not determine from the policy language alone, i.e., what the industry actually anticipated. We do not read Montrose as adopting a general rule that extrinsic evidence may be considered even if policy language is not ambiguous. Therefore, we confine our analysis to the policy language in the absence of ambiguity. " . . . California courts have uniformly ruled that the language of the sudden and accidental exception is unambiguous and its plain language meaning is 'abrupt,' 'unintended, and unexpected.' [Citation.]" ( Travelers Casualty Surety Co. v. Superior Court, supra, 63 Cal.App.4th at p. 1458; see also ACL Technologies, Inc. v. Northbrook Property Casualty Ins. Co., supra, 17 Cal.App.4th at pp. 1790-1791.) We, too, find no ambiguity. We agree with the Court of Appeal decisions, ante, that if "sudden and accidental" only meant "unexpected and unintended," then the word "sudden" would be surplus, because accidents are always, by definition, "unexpected and unintended." (See, e.g., Shell Oil Co. v. Winterthur Swiss Ins. Co., supra, 12 Cal.App.4th at p. 755.) A court should, of course, "disfavor constructions of contractual provisions that would render other provisions surplusage. [Citation.]" ( Boghos v. Certain Underwriters at Lloyd's of London (2005) 36 Cal.4th 495, 503; see also Civ. Code, § 1641.) The California Supreme Court recognized in 1959, well before the "sudden and accidental" exception was added to the standard policy, that "accident" may or may not mean abrupt. The court noted that while "accident" had been defined as something that happens "suddenly," it also had been defined to include "'"any event which takes place without the foresight or expectation of the person acted upon or affected by the event."' [Citations.]" ( Geddes Smith, Inc. v. St. Paul Mercury Indemnity Co. (1959) 51 Cal.2d 558, 563, italics omitted.) The court concluded that "[n]o all-inclusive definition of the word 'accident' can be given." ( Ibid.) It is reasonable to infer, then, that the word "sudden" was inserted to make clear that the exception only applies if the discharge is not only unexpected but also abrupt. If all that was intended was to require that the discharge be unexpected, it would have been sufficient merely to provide that a discharge is covered by the exception if it is "accidental." In that case, the exception would cover gradual discharges of pollutants, because a gradual discharge may well be unexpected if the insured is unaware it is happening, even though it is not abrupt. We therefore conclude, as the decisions cited, ante, unanimously held, that the "sudden and accidental" exception only covers abrupt, not gradual, pollution. The State does not dispute that if this interpretation of "sudden and accidental" is correct, the only discharges of pollutants that could be considered sudden and accidental were the releases that resulted from the heavy rains in March of 1969 and 1978. We consider whether those discharges were abrupt and therefore "sudden and accidental." First, however, we must determine whether Insurers are correct that the discharge that must be "sudden and accidental" is the initial dumping of wastes into the site and not the later discharge of the wastes from the site. 2. Relevant discharge Insurers point out that the court in Standun, Inc. v. Fireman's Fund Ins. Co., supra, 62 Cal.App.4th 882 held: "Where hazardous waste material is deposited directly into a landfill, the relevant discharge of pollutants for purposes of the pollution exclusion is the initial release of the hazardous waste into the landfill, not the subsequent release of pollutants from the landfill into the water, air and adjoining land. [Citation.]" ( Id. at p. 891.) However, while Standun's holding may have been proper under the circumstances of that case, those circumstances are not present here. In Standun, toxic substances migrated from a landfill into adjoining properties, in part because of the landfill operator's poor waste management practices. The insured had nothing to do with operating the landfill. Instead, the insured was sued under a strict liability theory simply because wastes generated at its place of business were transported to the landfill and deposited there. ( Standun, Inc. v. Fireman's Fund Ins. Co., supra, 62 Cal.App.4th at p. 890.) Therefore, the Standun court concluded: "The relevant discharge as to Standun is the discharge of its wastes into the landfill. That discharge was purposeful and regular. Accordingly, the relevant discharge of pollutants was neither sudden nor accidental and coverage under the policy is barred by the pollution exclusion." ( Standun, Inc. v. Fireman's Fund Ins. Co., supra, 62 Cal.App.4th at p. 892, italics added.) Here, in contrast, the State was not held liable for dumping wastes into the site. It was held liable for negligently selecting, designing, building, and operating the site. Its liability was based not on the release of wastes into the site — that was, after all, the intended purpose of the site — but on the release of wastes from the site when, because of the State's negligence, the site failed to contain them properly. Because the bases for the underlying liability in Standun and this case were different, Standun does not support denying coverage here. Insurers argue that application of the pollution exclusion is not controlled by the basis of the insured's liability, but rather by whether coverage is sought for property damage arising from a discharge to land. If so, the pollution exclusion bars coverage unless the discharge is sudden and accidental. According to Insurers, here there is no dispute that the property damage arose out of discharges to land that were not sudden and accidental. Insurers' argument begs the question by assuming that if wastes intentionally deposited into a disposal site later escape the site and cause injury to adjacent property, the damage "arose out of" only the initial deposit and not the later event that actually caused the wastes to escape. Nothing in the pollution exclusion or elsewhere in the policy compels that interpretation. The exclusion says the policy does not apply to "Property Damage arising out of the discharge, dispersal, release or escape" of pollutants "into or upon land or the atmosphere," but the exclusion "does not apply if such discharge, dispersal[,] release or escape is sudden and accidental." The phrase "arising out of" is not defined. "'California courts have consistently given a broad interpretation to the terms "arising out of" or "arising from" in various kinds of insurance provisions.'" (Medill v. Westport Ins. Corp. (2006) 143 Cal.App.4th 819, 830.) This language "'broadly links a factual situation with the event creating liability,'" and "'"requires [the court] to examine the conduct underlying the . . . lawsuit, instead of the legal theories attached to the conduct."' [Citation.]" (Ibid.) Here, the "'event creating liability'" and the "'"conduct underlying the . . . lawsuit"'" was the escape of the wastes from the site, not the deposit of the wastes into the site. It is not, as Insurers claim, undisputed that the escape of the wastes from the site was not sudden and accidental. Furthermore, even if the phrase "arising out of" reasonably could be construed to refer either to the initial deposit or the later escape, as an exclusionary provision, the pollution exclusion would have to be interpreted against Insurers. (TRB Investments, Inc. v. Fireman's Fund Ins. Co. (2006) 40 Cal.4th 19, 27.) Therefore, even if the property damage "arose out of" both sudden and accidental discharges and discharges that were not sudden and accidental, the policy would have to be construed to afford coverage for the damage. The court in Travelers Casualty Surety Co. v. Superior Court, supra, 63 Cal.App.4th 1440 refined the principle stated in Standun in a way that is instructive here. While acknowledging, and accepting, Standun's holding that ordinarily "the relevant discharge is the initial disposal of toxic waste into the landfill" ( Travelers, at pp. 1458-1459), the court nonetheless stated: "This is not to say that environmental contamination damages connected with industrial dumping are automatically barred from coverage under the sudden and accidental exception to the pollution exclusion. An intervening event may occur between the initial 'disposal of waste on the landfill and the actual damage that eventually resulted,' and that intervening event may have been sudden and accidental." ( Id. at pp. 1459-1460.) Therefore, under Travelers, even though the initial depositing of wastes into the Stringfellow Site was not "sudden and accidental," the State is still entitled to coverage if the 1969 discharge and/or the 1978 discharges were "sudden and accidental." For these reasons, we conclude that the relevant discharge as to the State was the release of the wastes from the site after they had been deposited there by other entities. The later release is the relevant discharge for purposes of determining whether the State's discharge of pollutants was "sudden and accidental." 3. The 1969 discharge In assessing whether the 1969 discharge was "sudden and accidental," we note first that none of the Court of Appeal decisions that denied coverage on the basis that the release of pollutants was not "sudden and accidental" involved a factual situation comparable to the 1969 discharge. In the first decision, Shell Oil Co. v. Winterthur Swiss Ins. Co., supra, 12 Cal.App.4th 715, Shell sought coverage for liability for contamination of groundwater and soil from a manufacturing site. As in this case, the wastes had been discharged into unlined ponds and were expected to evaporate or seep into the ground and stay there. Water flowing underground from the site was the most likely cause of the pollution. ( Id. at p. 734.) There was no overflow from the site as in this case. Thus, the court found that the discharge was gradual rather than abrupt and therefore could not be sudden. ( Id. at p. 754.) ACL Technologies, Inc. v. Northbrook Property Casualty Ins. Co., supra, 17 Cal.App.4th 1773 involved liability for leakage of pollutants from corroded underground storage tanks. The court stated: "Corrosion is, by definition, a gradual process. On the other hand, there was no evidence of any specific trauma to the tanks during the Northbrook policy period." ( Id. at p. 1795; see also Truck Ins. Exchange v. Pozzuoli, supra, 17 Cal.App.4th at pp. 860-861 [leakage from underground storage tank that went on for at least 60 days was not "sudden and accidental," though there would have been coverage if tank had exploded].) Again, there was no overflow from the site as in this case. In Standun, Inc. v. Fireman's Fund Ins. Co., supra, 62 Cal.App.4th 882, the insured deposited wastes at a landfill for about seven years. Later, property adjoining the site became polluted due to the site operator's discharge of waste water into the storm drainage system, as well as storm runoff and "problems with the structural integrity and capacity of underground leachate storage tanks." ( Id. at pp. 885-886.) Thus, again there was no precipitous release of wastes as occurred from the 1969 overflow. Finally, in Travelers Casualty Surety Co. v. Superior Court, supra, 63 Cal.App.4th 1440, a case apparently involving the same landfill as in Standun (see Travelers, at p. 1445), the contamination resulted from leachate seepage, landfill gas, noxious odors, and groundwater contamination. ( Id. at pp. 1445-1446.) All of these processes are gradual in nature, occurring on a continuous basis as wastes accumulate and begin to escape from the landfill. Here, in contrast, Insurers themselves assert that the 1969 discharge occurred because on March 17, 1969, the ponds "'topped out'" after the rains had filled them, causing the contaminants at the site to overflow into the surrounding environment. A reasonable trier of fact could find on this evidence that the overflow was a "discharge, dispersal[,] release or escape" that was "sudden and accidental," making the pollution exclusion by its terms inapplicable. The closest case factually that we have found is Aetna Cas. Sur. Co. v. Dow Chemical Co. (E. D. Mich. 1998) 10 F.Supp.2d 771. There, the insured presented evidence "that there was a sudden, massive levee break, which was different in kind from the leaking and seepage which was also allegedly present at the site, and furthermore, the massive levee break contributed to the extent and nature of contamination at the site." ( Id. at p. 781.) The insured also presented evidence that "the relevant property damage at the [insured's] site — groundwater contamination — may arguably be traced to the 1970 massive levee break." ( Ibid.) The court ruled: "Dow's proferred evidence is sufficient to create a genuine issue of material fact whether the 1970 levee break fits within the 'sudden and accidental' exception to the pollution exclusion in Dow's policies." ( Ibid.) An overflow, like a levee break, is reasonably viewed as a "sudden" event. This view also is consistent with the intent of the insurance industry in drafting the "sudden and accidental" exception. In Truck Ins. Exchange v. Pozzuoli, supra, 17 Cal.App.4th 856, the court noted: "[T]he principal draftsman of the pollution exclusion clause has stated it was intended to wholly eliminate coverage for pollution except in the case of a 'classical accident,' which he defined as a 'sudden, boom-type accident' such as an explosion. [Citation.]" ( Id. at p. 859, fn. 2, quoting Williams, Intent of the Drafters: The Pollution Exclusion And The Insurance Commissioner Documents in Environmental and Toxic Tort Claims Insurance Coverage in 1990 and Beyond (1990) pp. 25-26; see also Staefa Control-System v. St. Paul Fire Marine (N.D. Cal. 1994) 847 F.Supp. 1460, 1468, fn. 5 [quoting same source].) A dike washout, or even an overflow with no washout, reasonably can be seen as a "classical accident." Insurers suggest several reasons why the 1969 discharge should not be considered "sudden and accidental." First, they contend that because the discharge came after several weeks of rain, it was not sudden. This argument confuses the discharge of wastes with the events that led up to the discharge. The "sudden and accidental" exception applies if the "discharge, dispersal[,] release or escape" of pollution is sudden. It does not require that the events causing the discharge be sudden. Here, it is at least inferable that the 1969 discharge occurred suddenly, because according to Insurers, it did not rain on the day the discharge occurred and therefore the discharge could not have occurred as part of a gradual process of water accumulation and release. Second, Insurers argue that Travelers Casualty Surety Co. v. Superior Court, supra, 63 Cal.App.4th 1440 stated that "the expected natural phenomena of rain falling upon a landfill and causing migration of contaminants cannot constitute sudden and accidental discharge of pollution." ( Id. at p. 1463.) It is evident from the context of that statement, however, that the court was referring to the gradual migration of contaminants as rain soaks a landfill and the contents start to seep from the landfill. As noted, ante, the groundwater contamination in Travelers resulted from leachate seepage, not an overflow into the surrounding area. ( Id. at pp. 1445-1446.) Third, Insurers argue, the 1969 discharge was not "accidental" because it was not unexpected. Rather, the State had known years before that heavy rains could cause the site to overflow. However, theoretically it can be said that any party operating an uncovered site "knows" that if it rains enough, the contents of the site will be washed outside of its boundaries. That does not mean the party is bound to have "expected" a discharge that results when an unprecedented period of rain occurs. Here, it is not disputed for purposes of summary judgment and this appeal that the 1969 storm of some 20 inches was a once-in-50-year event. Nothing in the record indicates the site had ever overflowed before, though it had operated for 13 years. In any ordinary, reasonable sense of the word "expected," there is at least a triable issue whether the State should have expected that the site would overflow and cause the 1969 discharge. Finally, Insurers note that under Travelers Casualty Surety Co. v. Superior Court, supra, 63 Cal.App.4th 1440 a loss caused by an intervening event is only "sudden and accidental" if the event did not arise from the ordinary course of business and caused an "appreciable" amount of damage over and above routine dumping into the disposal site. Here, Insurers assert, the 1969 discharge arose out of the routine dumping over the years, because it would not have happened had the dumping not occurred. Further, because the State admitted it could not differentiate damage caused by the 1969 discharge from damage caused by routine (see part II.F, post), it could not show the 1969 discharge caused any "appreciable" amount of damage over and above the routine damage. While Insurers are correct that the 1969 discharge would not have occurred but for the routine dumping, it is simply unrealistic to claim the discharge "arose" only out of the dumping, and not out of the overflow. As we have noted, the phrase "arising out of" is given a "broad interpretation." (Medill v. Westport Ins. Corp., supra, 143 Cal.App.4th at p. 830.) Moreover, there was at least enough evidence to create a triable issue whether the 1969 discharge caused an "appreciable" amount of damage. The State's expert concluded in 2004, "[I]t is reasonable to expect that much of the soil contamination detected [after the 1978 discharges] was from the 1969 event." At his deposition, the expert explained the basis for his conclusion: "[I]n 1965 [ sic; 1969] the site was still active and so the ponds contained mostly waste liquids. In '78, the site had been shut down for some considerable period of time, about six years, actually, and so the fluid that would have overflowed in 1978 would have, at a minimum, been diluted from the time when the site was active." He further explained that by 1978, the fluid would have been diluted by "the influx of variety of rainfalls" between 1969 and 1978. The expert's conclusion that the more concentrated wastes that were released in 1969 were likely responsible for much of the soil contamination detected later was at least sufficient to support an inference "that an appreciable amount of environmental damage was caused by the intervening event," i.e., the 1969 overflow, "over and above that caused by routine dumping into the disposal site." ( Travelers Casualty Surety Co. v. Superior Court, supra, 63 Cal.App.4th at p. 1460.) In sum, construing the "sudden and accidental" exception "broadly in favor of the insured" ( Aydin Corp. v. First State Ins. Co., supra, 18 Cal.4th at p. 1192), we view the overflow of a liquid waste site due to massive rains as an event within the exception. We therefore conclude summary judgment on the basis that the policies did not cover the 1969 discharge was improper. 4. 1978 discharges We reach a different conclusion regarding the 1978 discharges, not because the discharges were not "sudden," but because they were not unexpected and therefore were not "accidental." The State argues it did not expect the 1978 discharges might occur, because after the 1969 discharge, it improved the storm drainage at the site and would not have reopened the site if it had not believed it was safe against overflowing. However, at least five years before the 1978 discharges, the State was on notice that the site was not safe against overflowing. It is undisputed that in 1973 Mr. Franks, the State's chief hydrologist, told Mr. Anderson, the executive officer of the State's regional water quality control board, "about the risk of the site being inundated," which was "exactly what happened in 1978." It is further undisputed that in 1974, Franks "recognized that the site had the potential to fill up and overflow during a heavy storm." It is likewise undisputed that he "suggested leveling the site and putting an impervious cap on it" and that one reason for the cap was "to prevent overflow in the event of rain." It also is undisputed that the State did not install a cap as suggested. The State points out that the rains in early 1978 were so intense that Riverside County was declared a disaster area by the state and federal governments. The State asserts it could not have expected rains of this magnitude. However, the state government's declaration came on February 5, 1978, and the federal declaration came on February 15, 1978, more than two weeks before the 1978 discharges. The discharges did not occur until March 5 and 8, 1978. It is not reasonable to believe that by that time, the possibility of a discharge was still "unexpected." The State asserts that even on March 5, 1978, Anderson believed the sandbags and pumping that had been put in place would keep the ponds from further overflowing. However, in the deposition testimony to which the State cites, Anderson said only that it was his " best hope" that "the sandbags would be sufficient and the rains would abate." (Italics added.) In addition, Anderson testified that he dug a trench "[i]n anticipation that the sandbags would not be sufficient." (Italics added.) Finally, the State notes that, while the pumping was in progress, a flood worker stepped on the discharge pipe of one of the pumps, and it had to be repaired. According to Anderson's notes, however, this happened after one of the ponds had filled to the brim and "overtopping started." Thus, even though the State may not have expected the particular mishap with the pump, at least some discharge would have occurred even without the mishap. In sum, years before the 1978 discharges, the State knew at least that (1) there was a risk of the site becoming inundated by rain and overflowing, as actually happened in 1978; (2) the site had already overflowed and released pollutants before; and (3) it needed to place a cap on the site to prevent this from happening again. Courts applying the "sudden and accidental" exception have recognized that "[w]here a discernable pattern of releases of hazardous waste has occurred at the same site over a period of time, it becomes difficult to maintain that those releases were 'unexpected,' even if they were not deliberate or intentional." ( Quaker State Minit-Lube, Inc. v. Fireman's Fund Ins. Co. (D. Utah 1994) 868 F.Supp. 1278, 1324; accord, Associated Indem. Corp. v. Dow Chemical Co. (E.D. Mich. 2003) 248 F.Supp.2d 629, 637 ["evidence of the pollutant container's design, licensing, and past violations is probative as to whether there was an expectation of containment"].) Here, though there was a theoretical chance that after 1969 it would never again rain heavily enough to cause any discharge, if that were enough to make a discharge "accidental," the term would cease to have any practical meaning. A person who digs a 20-foot pit on his or her property with no warning sign and no fence around it theoretically may not "expect" anyone to fall in, since no one may pass that way. However, it would not be reasonable to suggest that if someone did fall in, the event was "unexpected" by the owner. Considering all of the circumstances from the closure of the site in 1972 up to the 1978 discharges, a reasonable trier of fact could not conclude the discharge was "unexpected." Consequently, the court correctly ruled that the 1978 discharges were not covered. E. Watercourse Exclusion Insurers contend the 1969 discharge was excluded from coverage by the "absolute" watercourse exclusion. As they point out, an expert retained by the State noted in a 2004 report that when the site overflowed in 1969, waste was discharged into Pyrite Creek, a watercourse. Although the State acknowledges that the discharges went into Pyrite Creek, a watercourse, it argues that Insurers did not show the discharges did not also enter the land, and therefore the court should not have granted summary judgment based on the watercourse exclusion. Insurers counter that at least the 1969 discharge was initially confined to Pyrite Creek, and the initial discharge was the relevant event for purposes of determining whether the watercourse exclusion applied. As evidence that the discharge first went into Pyrite Creek and only later contaminated the land, Insurers point to exhibits they submitted in support of their summary judgment motions. However, those exhibits fail to demonstrate that the 1969 overflow discharged waste only into the creek, even initially. Insurers cite an October 1972 letter from the California Regional Water Quality Control Board to the County of Riverside that stated: "In the spring of 1969, the heavy rains exceeded the capacity of the storm water diversion ditches and runoff flowed through the dump site carrying some of the waste out of the dump and down a natural drainage ditch parallel to Pyrite Street crossing Highway 60 and Mission Boulevard." (Italics added.) Insurers also cite an October 1980 report prepared by the State Water Resources Control Board that stated: "The [1969] storm caused the hazardous wastes ponded behind a subsurface concrete dam to overflow into the Pyrite Creek drainage channel." (Italics added.) Contour maps included in the record show that the Stringfellow Site was situated at the head of a canyon that sloped southwest toward Glen Avon. A drainage channel designated "Pyrite Channel" extended from the mouth of the canyon southwest across Highway 60, through Glen Avon, and beyond. Naturally, any liquid waste discharged from the site eventually would enter the channel if there was enough of it. However, neither the maps nor any other evidence of which we are aware showed whether the creek filled the entire channel, or whether it began immediately at the boundary of the site or somewhere farther down the channel. Accordingly, the statements that pollutants went into the "drainage ditch" or "drainage channel" did not necessarily show that the discharges were confined to the creek that flowed through the channel. In fact, there was deposition testimony (presented by Insurers) that waste water from the 1969 discharge contaminated "[t]he soil within the canyon from the site down to the community." (Italics added.) Thus, there was at least a factual issue whether the discharge entered only the creek, or also entered the soil. Insurers argue that a "watercourse" includes not only a creek but also the entire channel through which it flows and the ground where water usually flows, even when there is no water. They cite decisions involving riparian rights. Assuming, without deciding, that those decisions are relevant in applying the watercourse exclusion, we note the California Supreme Court has defined "watercourse" for purposes of riparian rights as follows: ". . . 'A watercourse is defined as a natural stream of water usually flowing in a definite channel, having a bed and sides, or banks, and discharging itself into some other stream or body of water.'" (Chowchilla Farms Inc. v. Martin (1933) 219 Cal. 1, 16.) The record in this case did not show that water usually flowed in the entirety of Pyrite Channel, over all of the land into which the 1969 discharge deposited contaminants, or that the channel discharged into some other stream or body of water. Hence, there was at least a factual issue whether the 1969 discharge was confined to a watercourse. In addition, there was evidence that pollution from the 1969 discharge entered groundwater at sites away from the creek. The October 1980 report stated: "Groundwater contamination downgradient of the site was first observed after the spring 1969 storm in the Stringfellow well 1Q1." According to the maps, well 1Q1 was located to the east of the creek channel rather than directly in its drainage path. Thus, the presence of groundwater contamination at the well site at least supported an inference that waste water could have been discharged onto the land adjacent to the creek, rather than only into the creek itself. Insurers also cite a report prepared by the State's chief hydrogeologist, evidently in 1979, which stated: "Subsequent to a major storm system, the Riverside County Flood Control District reported to the Regional Board on March 17, 1969, that stormwater runoff carried waste out of the disposal site down Pyrite Creek and across Highway 60." However, the next page of the report stated that a technical report in January 1973 had "indicated that the degradation of the groundwater at the monitoring well was the result of surface runoff from the disposal site caused by storms during the spring of 1969." (Italics added.) The only monitoring well mentioned was well 1Q1. Thus, like the 1980 report, the 1973 report supported an inference that discharge of waste water onto the surface of the adjacent land, as opposed to discharge directly into the creek, could have polluted the groundwater. In our view, discharge of pollutants into groundwater does not fall within the watercourse exclusion. There appear to be no California decisions on point, and apparently out-of-state courts are in conflict on the issue. (See Aetna Cas. Sur. Co. v. Dow Chemical Co. (E.D.Mich. 1998) 28 F.Supp.2d 440, 447.) However, we find persuasive the view stated in Dow that "body of water" means "an aggregate of water having defined boundaries. . . ." ( Ibid.; see also Lumbermens Mut. Cas. Co. v. Plantation Pipeline Co. (1994) 214 Ga.App. 23 [ 447 S.E.2d 89, 91-94].) Construing the phrase that way, groundwater contamination would not fall within the exclusion, as groundwater has no "defined boundaries." Therefore, if the 1969 discharge polluted the groundwater by entering the land, the watercourse exclusion would not apply. Again, it appears there is a least a triable issue whether this happened. For all of the above reasons, we conclude the court should not have granted summary judgment based on the watercourse exclusion.
We note that some of the defendants in the federal action claimed "that the heavy rainfall in 1969 and 1979 [ sic] was a natural disaster which constituted an act of God." ( U.S. v. Stringfellow (C.D. Cal. 1987) 661 F.Supp. 1053, 1061.) The defendants sought to invoke the exception to liability under the Comprehensive Environmental Response, Compensation, and Liability Act of 1980 (CERCLA; see 42 U.S.C. § 9601 et seq.) for damage caused solely by "an act of God." ( 42 U.S.C. § 9607(b)(1).) The district court found, however, "that the rains were not the kind of 'exceptional' natural phenomena to which the narrow act of God defense of [title 42 United States Code section 9607(b)(1)] applies. The rains were foreseeable based on normal climatic conditions and any harm caused by the rain could have been prevented through design of proper drainage channels. Furthermore, the rains were not the sole cause of the release." ( Stringfellow, at p. 1061.)
An "act of God" for purposes of CERCLA is defined as "an unanticipated grave natural disaster or other natural phenomenon of an exceptional, inevitable, and irresistible character, the effects of which could not have been prevented or avoided by the exercise of due care or foresight." ( 42 U.S.C. § 9601(1).) The parties have not cited Stringfellow, and we express no opinion on what significance, if any, the finding of the court in that case that the rains were not an "act of God" might have in this case.
As noted, ante, the exclusion was not absolute in the Columbia policy.
Insurers also cite this court's decision in Ortega Rock Quarry v. Golden Eagle Ins. Corp. (2006) 141 Cal.App.4th 969. They provide no internal page citation, so it is unclear what part of the decision they claim is relevant here. In any event, Ortega Rock Quarry did not involve the watercourse exclusion or the definition of "watercourse." The case involved an "absolute" pollution exclusion, and the issues pertaining to the application of the exclusion were whether the language of the exclusion was ambiguous as to the definition of "pollutant" and whether dirt and rocks deposited into a creek were "pollutants." (Id. at pp. 979-980.) We do not see how the decision has any relevance here.
F. Allocation Between Covered and Uncovered LossesThe State, in response to a request for admission, stated: "The State admits it cannot differentiate the work performed to date to remedy the property damage caused by the escape of contaminants [from one alleged release, e.g., through the fractured granite] from the work performed to date to remedy the property damage caused by, [ sic] the escape of contaminants [from another alleged release, e.g., the '1969 overflow' or '1978 overflow']. . . ." (Original brackets and some capitalization omitted.) The State similarly admitted it could not differentiate the expenses it had paid to remedy damage caused by one release from expenses paid to remedy damage from another release. Based on these admissions, Insurers argue that Golden Eagle Refinery Co. v. Associated Internat. Ins. Co. (2001) 85 Cal.App.4th 1300 [ 102 Cal.Rptr.2d 834] ( Golden Eagle) and Lockheed Martin Corp. v. Continental Ins. Co. (2005) 134 Cal.App.4th 187 [ 35 Cal.Rptr.3d 799] ( Lockheed) bar the State from recovering at all, whether or not the 1969 discharge was "sudden and accidental."
1. Golden Eagle and Lockheed In Golden Eagle, the insured, Golden Eagle Refinery Co. (Golden Eagle), operated an oil refinery for 26 years and during that time routinely discharged crude oil and crude oil products on and into the ground, contaminating the soil. Some of the discharges were sudden and accidental, and others were not. ( Golden Eagle, supra, 85 Cal.App.4th at p. 1304.) Golden Eagle admitted it was unable to attribute any particular part of the property damage to any particular discharge. ( Id. at p. 1310.) After operations ceased, the State required Golden Eagle to clean up the site. Golden Eagle then sought coverage from its liability insurers for the cleanup costs. The court stated that because some discharges were "sudden and accidental" and therefore covered by the policies, but other discharges were not, "[i]t would be the insured's burden at trial to prove that all of the damages it seeks to recover were caused by a covered event of discharge, failing which, Golden Eagle will recover nothing." ( Golden Eagle, supra, 85 Cal.App.4th at pp. 1309-1310, italics added.) Since Golden Eagle admitted it could not meet that burden, summary adjudication of the coverage claim was properly granted against it. ( Id. at p. 1317.) In Lockheed, the insured, Lockheed Corp. (Lockheed), operated an aero-space research and manufacturing facility from about 1928 to the early 1990's, causing soil and groundwater contamination at the site. The federal government required Lockheed to clean up the contamination, and Lockheed sought coverage from its liability insurers for the cleanup costs. Lockheed submitted evidence of 14 accidents at the site between 1967 and 1980 in which a pollutant, PCE, was discharged. These discharges were covered by the policies. However, most of the pollution-causing events were routine leaks and spills that were not covered by the policies. ( Lockheed, supra, 134 Cal.App.4th at pp. 212-215.) Lockheed admitted it could not separate the cleanup costs according to the individual pollutants involved and failed to show that the "handful" of accidental discharges "resulted in increased migration or penetration of the PCE contamination." ( Id. at pp. 217-218.) Relying on Golden Eagle, the court held that to prove the causation and damages elements of its case Lockheed had "to prove that these accidents affected the investigation and cleanup costs for which Lockheed was liable." ( Lockheed, supra, 134 Cal.App.4th at p. 217.) That would require"'the identification, allocation, and quantification of the contamination in relation to its source.' [Citation.]" ( Id. at p. 218.) Since Lockheed admitted it could not do that, the trial court properly excluded Lockheed's evidence. ( Id. at pp. 218-219, citing Golden Eagle, supra, 85 Cal.App.4th at p. 1316.) 2. Partridge The State argues Golden Eagle and Lockheed are contrary to the decision of the California Supreme Court in State Farm Mut. Auto. Ins. Co. v. Partridge (1973) 10 Cal.3d 94 [ 109 Cal.Rptr. 811, 514 P.2d 123] ( Partridge). In Partridge, the insured negligently filed the trigger mechanism of his pistol to lighten the trigger pull. Later, he used the gun to shoot jackrabbits out of the windows of his truck as he drove through the countryside with two friends. The truck hit a bump and the gun fired, wounding one of the passengers. The question was whether the insured's liability for the damage was covered by the general liability provision of his homeowner's policy, which excluded liability arising from the use of a vehicle. ( Id. at pp. 97-99.) That policy provided: "'This Company agrees to pay on behalf of the Insured all sums which the Insured shall become legally obligated to pay as damages because of bodily injury or property damage, to which this insurance applies, caused by an occurrence.'" ( Id. at p. 99, fn. 5.) An "occurrence" was "'an accident, including injurious exposure to conditions, which results, during the policy term, in bodily injury or property damage.'" ( Ibid.) The court held that although the damage arose in part out of an excluded activity, the use of the truck, the homeowner's policy covered the damage: "Here the 'use' of Partridge's car was not the sole cause of Vanida's injuries but was only one of two joint causes of the accident. . . . Defendants correctly contend that when two such risks constitute concurrent proximate causes of an accident, the insurer is liable so long as one of the causes is covered by the policy." ( Partridge, supra, 10 Cal.3d at p. 102.) The court further explained: "In the instant case, . . . although the accident occurred in a vehicle, the insured's negligent modification of the gun suffices, in itself, to render him fully liable for the resulting injuries. Under these facts the damages to Vanida are, under the language of the homeowner's coverage clause, 'sums which the Insured . . . (became) legally obligated to pay' because of the negligent filing of the trigger mechanism. . . ." ( Partridge, supra, 10 Cal.3d at p. 103.) Thus, the Partridge court interpreted the phrase "all sums which the Insured shall become legally obligated to pay as damages" to mean all amounts for which the insured would be held liable in a tort action. As the court further explained: "If, after negligently modifying the gun, Partridge had lent it to a friend who had then driven his own insured car negligently, resulting in the firing of the gun and injuring of a passenger, both Partridge and his friend under traditional joint tortfeasor principles would be liable for the injury. In such circumstances, Partridge's personal liability would surely be covered by his homeowner's policy. . . ." ( Partridge, supra, 10 Cal.3d at p. 103.) The result should be no different, the court said, "simply because, in the instant case, both negligent acts happened to have been committed by a single tortfeasor." ( Ibid.) 3. Garvey Since Partridge, the California Supreme Court has not addressed the issue of coverage in a third party liability case involving concurrent causes. However, in Garvey v. State Farm Fire Casualty Co. (1989) 48 Cal.3d 395 [ 257 Cal.Rptr. 292, 770 P.2d 704] ( Garvey), the court in addressing coverage in a first party case involving multiple contributing causes made several observations concerning Partridge that are instructive in this case. In Garvey, the insureds' home was damaged as a result of earth movement (an excluded cause) and negligent construction (a covered cause). The trial court ruled the damage was covered by the insureds' homeowner's policy, based on the concurrent causation analysis in Partridge. (Garvey, supra, 48 Cal.3d at pp. 399-44-01.) The Supreme Court reversed, holding that the "efficient proximate cause" analysis of Sabella v. Wisler (1963) 59 Cal.2d 21 [ 27 Cal.Rptr. 689, 377 P.2d 889], and not the "concurrent cause" analysis of Partridge, should control first party coverage cases. (See Garvey, at pp. 412-413.) Under the Sabella analysis, "'where there is a concurrence of different causes, the efficient cause — the one that sets others in motion — is the cause to which the loss is to be attributed, though the other causes may follow it, and operate more immediately in producing the disaster.'" ( Sabella, at pp. 31-32.) In explaining why the Partridge analysis did not govern first party cases, the court first stated that in contrast to coverage in first party cases, "the right to coverage in the third party liability insurance context draws on traditional tort concepts of fault, proximate cause and duty." ( Garvey, supra, 48 Cal.3d at p. 407, italics added.) Thus, in third party cases "the focus is, at least initially, on the insured's legal obligation to pay for injury or damage arising out of an 'occurrence.'" ( Id. at p. 408, italics added.) Further, the court said, application of tort law analysis in third party cases means that in such cases "the insurer agrees to cover the insured for a broader spectrum of risks" than in first party cases. ( Garvey, supra, 48 Cal.3d at p. 407.) In fact, since in most first party cases "the insured can point to some arguably covered contributing factor," the court observed that "if the rule in Partridge [citation] were extended to first party cases, the presence of such a cause, no matter how minor, would give rise to coverage." ( Id. at p. 408.) The Garvey court also held the Court of Appeal in that case had misinterpreted Partridge in determining "that in order for coverage to be found under Partridge, the concurrent event alone must have been a 'sufficient condition' of the loss — i.e., capable of producing damage itself." ( Garvey, supra, 48 Cal.3d at p. 409, fn. omitted.) The court explained that "[t]he term 'sufficient condition' as used by the Court of Appeal missapplied [ sic] the Partridge holding because it implied that negligent driving alone could have caused plaintiff's injury in that case." ( Id. at p. 409, fn. 8.) As Garvey was a first party case, the Garvey court expressed no opinion on the proper analysis to be applied in third party cases involving multiple causes. Instead, it simply clarified that Partridge "should be limited to the third party tort liability context." ( Garvey, supra, 48 Cal.3d at p. 410, fn. 9.) The court did say, "[W]e leave the application of Partridge in the liability context to a future liability case that raises the concurrent causation issue," suggesting that it might revisit the issue if a suitable opportunity presented itself. ( Id. at p. 409, fn. 8.) Justice Kaufman concurred in the result in Garvey but not in the court's reasoning. He believed Sabella, not Partridge, should govern not only first party cases but also third party cases. In Justice Kaufman's view, under the Sabella analysis there would be no coverage in Partridge because the filing of the trigger — the covered cause — "was activated by the other cause," the driving of the truck. "Thus, under Sabella principles it was the negligent driving of the vehicle over the rough terrain with the gun pointed at the passenger that was the 'efficient' or 'predominant' cause of the injury and coverage for such accidental injury arising out of the use of a vehicle was expressly excluded by the homeowner's policy." ( Garvey, supra, 48 Cal.3d at p. 415 (cone. opn. of Kaufman, J.).) As this case involves third party liability, Garvey has no direct application here. Garvey does, however, make clear that under Partridge: (1) "the right to coverage in the third party liability insurance context draws on traditional tort concepts of fault, proximate cause and duty." ( Garvey, supra, 48 Cal.3d at p. 407); (2) in a third party liability policy "the insurer agrees to cover the insured for a broader spectrum of risks" than in first party cases. ( Garvey, supra, 48 Cal.3d at p. 407); (3) in a third party case, the presence of a covered cause, "no matter how minor, would give rise to coverage" ( Garvey, supra, 48 Cal.3d at p. 408); and (4) the insured in a third party case need not show that the covered cause was "a 'sufficient condition' of the loss — i.e., capable of producing damage itself." ( Garvey, supra, 48 Cal.3d at p. 409, fn. omitted.)
Partridge was decided before the Supreme Court, in Li v. Yellow Cab Co. (1975) 13 Cal.3d 804, 828-829 [ 119 Cal.Rptr. 858, 532 P.2d 1226], adopted the comparative negligence doctrine. However, "adoption of comparative negligence . . . does not warrant the abolition or contraction of the established 'joint and several liability' doctrine; each tortfeasor whose negligence is a proximate cause of an indivisible injury remains individually liable for all compensable damages attributable to that injury." ( American Motorcycle Assn. v. Superior Court (1978) 20 Cal.3d 578, 582 [ 146 Cal.Rptr. 182, 578 P.2d 899].) Joint and several liability does not, however, apply to liability for noneconomic damages in cases controlled by Proposition 51. (Civ. Code, § 1431.2, subd. (a).)
"[T]he distinction between first and third party claims can be summarized as follows: If the insured is seeking coverage against loss or damage sustained by the insured, the claim is first party in nature. If the insured is seeking coverage against liability of the insured to another, the claim is third party in nature." ( Garvey, supra, 48 Cal.3d at p. 399, fn. 2.)
4. Incompatibility of Golden Eagle and Lockheed with Partridge and GarveyIt is readily apparent from the foregoing discussion that Golden Eagle and Lockheed are incompatible with the views expressed in Partridge and Garvey. Golden Eagle held that "insurers are not required to indemnify for any damages not caused by a covered event." ( Golden Eagle, 85 Cal.App.4th at p. 1314.) By this the court meant that the insured must negate the possibility that any part of the damages for which it seeks indemnity was caused by an uncovered event. Partridge, on the other hand, held that "coverage under a liability insurance policy is equally available to an insured whenever an insured risk constitutes simply a concurrent proximate cause of the injuries." ( Partridge, supra, 10 Cal.3d at pp. 104-105, fn. omitted.) Thus, under Partridge the insured is entitled to coverage even if the damages were partially caused by an uncovered risk. In a similar vein, the Golden Eagle court stated that the insured's damages were not covered because it admitted "that its damages were 'indivisible.'" ( Golden Eagle, supra, 85 Cal.App.4th at p. 1315.) Yet that was precisely the situation in Partridge as well. The damage caused by modifying the trigger was not divisible from that caused by the use of the vehicle. One injury, from one bullet, resulted from both causes. The court nonetheless held the damage was covered. The holding of the Golden Eagle and Lockheed courts that an insured must prove the damages for which it seeks coverage were caused solely by a covered cause and not by a combination of covered and uncovered causes also is contrary to the Supreme Court's interpretation of Partridge in Garvey. The Garvey court stated that the Court of Appeal in that case had misinterpreted Partridge by concluding "that in order for coverage to be found under Partridge, the concurrent event alone must have been a 'sufficient condition' of the loss — i.e., capable of producing damage itself." ( Garvey, supra, 48 Cal.3d at p. 409, fn. omitted.) If the insured need not show that the concurrent cause was capable of producing damage by itself, then it follows that, contrary to Golden Eagle and Lockheed, the insured cannot be required to trace the damages for which it seeks coverage to one cause and one cause only. Finally, as was made clear in both Partridge and Garvey, the Partridge court adopted a tort liability analysis in determining coverage in third party cases. Because the insured's negligent modification of the gun was sufficient by itself "to render him fully liable for the resulting injuries," the damages to the passenger were "'sums which the Insured . . . [became] legally obligated to pay' because of the negligent filing of the trigger mechanism. . . ." ( Partridge, supra, 10 Cal.3d at p. 103.) The criterion for coverage is not, therefore, whether the insured can differentiate between the amount of the damages that are allocable to the covered and uncovered causes, but whether the insured would be held liable in tort for the damages. Where, as in Partridge, the damages are indivisible, the insured is liable for all the damages and hence is covered for the entire amount. Golden Eagle and Lockheed, however, expressly rejected any reliance on tort liability principles. Both Golden Eagle and Lockheed stated: "A claim for indemnity is a contract claim. . . ." ( Golden Eagle, supra, 85 Cal.App.4th at p. 1316; see Lockheed, supra, 134 Cal.App.4th at p. 217 [quoting Golden Eagle].) The Golden Eagle court added: "Golden Eagle's argument that it need only prove that a sudden and accidental event caused an appreciable amount of the contamination is wrong because it is essentially a tort approach. . . . 'Substantial cause' may be sufficient to make a prima facie case in a tort action in order to support a joint and several judgment, but in the context of a coverage dispute relating only to the duty to indemnify, the tort threshold is not sufficient." ( Golden Eagle, at p. 1316, italics added.) As the Partridge court's joint tortfeasor example illustrated, however, in that court's view the damage was covered precisely because the evidence would support a joint and several judgment against the insured. The incompatibility of Golden Eagle's causation analysis with Partridge is best illustrated by the fact that Golden Eagle used exactly the analysis Justice Kaufman used in Garvey in disagreeing with Partridge. Justice Kaufman stated: "The first flaw in Partridge is, as the majority suggests, that it imported into the determination of coverage, concepts and rules of tort law inapplicable to the contractual question of the coverage afforded by an insurance policy, and, based on them, adopted the tort rule of concurrent causation to determine coverage." ( Garvey, supra, 48 Cal.3d at pp. 413-414 (cone. opn. of Kaufman, J.), italics added.) Justice Kaufman, however, like the Golden Eagle court, thought that coverage "is to be determined by contract principles, not tort principles, in both first party and third party cases." ( Garvey, at p. 414.) Golden Eagle did not, of course, acknowledge that it was adopting a position that the Supreme Court had rejected in Partridge and had implicitly rejected again in Garvey when it declined to adopt Justice Kaufman's analysis and left intact the application of Partridge to third party cases. In fact, neither Golden Eagle nor Lockheed acknowledged Partridge at all. The failure of the Courts of Appeal in Golden Eagle and Lockheed even to mention a potentially conflicting Supreme Court decision is puzzling enough. In the case of Golden Eagle, it is even more puzzling, because the insured in Golden Eagle cited Partridge in its petition for rehearing, arguing: "[T]he Court appears to hold that where third party property damage results from both covered and non-covered causes, a policyholder losses [ sic] coverage unless it segregates and quantifies the damage resulting from covered causes versus non-covered causes. [Citation.] This apparent ruling conflicts with California Supreme Court authority on the 'concurrent causation' doctrine that provides that a policyholder is entitled to full liability coverage for losses caused 'concurrently' by independent covered and non-covered events. State Farm Mutual Auto Ins. Co. v. Partridge[, supra,] 10 Cal.3d [at pages] 104-105." Thus, the potential conflict between Partridge and Golden Eagle was squarely before the Golden Eagle court. The court's failure, nonetheless, to address Partridge at least calls into question Golden Eagle's reliability as precedent. We are not the first to make that observation. The authors of a treatise on California insurance litigation have noted that Golden Eagle "may not be reliable precedent" because "the court seems to have overlooked the causation standards set forth by the California Supreme Court in Partridge, above." (2 Croskey et al., Cal. Practice Guide: Insurance Litigation (The Rutter Group 2005) ¶ 7:244:8, pp. 7A-78 to 7A-79.) Moreover, even apart from its inconsistency with Partridge, Golden Eagle's analysis is not convincing. In support of its conclusion, the court cited Civil Code section 3300, which provides: "For the breach of an obligation arising from contract, the measure of damages, except where otherwise expressly provided by this code, is the amount which will compensate the party aggrieve for all the detriment proximately caused thereby, or which, in the ordinary course of things, would be likely to result therefrom." Golden Eagle read the statute to mean that unless the insured can show what proportion of its total liability was the direct result of a covered cause, its damages for breach of the duty to indemnify are zero. (See Golden Eagle, supra, 85 Cal.App.4th at p. 1316.) However, a liability insurance policy is a contract not for a good or a personal service, but for indemnification against legal liability in tort. {Stein-Brief Group, Inc. v. Home Indemnity Co. (1998) 65 Cal.App.4th 364, 369-370 [ 76 Cal.Rptr.2d 3].) Thus, although an action for failure to indemnify is an action for breach of contract, the insurer's liability for breach of the policy is measured by the tort liability incurred by the insured that the insurer failed to cover. It follows that the "detriment proximately caused" by the breach of the duty to indemnify is the amount the insured is liable in tort to pay because of a covered risk. Under the principle of joint and several liability, that amount includes all of the damages of which the covered risk is a proximate cause, even if it is not the only cause. Thus, to say as the Golden Eagle court did that a coverage action is an action for breach of contract and therefore the insured is limited to contract damages ignores the fact that the measure of the insured's liability, and hence the measure of its damages, is governed by tort and not by contract law. For this same reason, we must reject Insurers' contention that the State's position conflates the concurrent cause doctrine and the requirement that in a breach of contract case the insured must prove what damages flowed from a covered event. In the liability insurance context, the two are the same because liability resulting even in part from a concurrent cause is wholly, not partially, covered. Golden Eagle also relied on FMC Corp. v. Plaisted Companies (1998) 61 Cal.App.4th 1132 [ 72 Cal.Rptr.2d 467] ( FMC Corp.) According to the Golden Eagle court, FMC Corp. "stands for the proposition that where both covered and not covered events cause damages a failure to differentiate and allocate is fatal to a claim for indemnity." ( Golden Eagle, supra, 85 Cal.App.4th at p. 1314.) In fact, FMC Corp. held exactly the opposite — the insured wanted to allocate, and the court held it could not do so. FMC Corp. involved pollution at multiple sites operated by the insured. ( FMC Corp., supra, 61 Cal.App.4th at pp. 1143-1144.) Pollution was only covered if it unexpectedly and unintentionally caused damage. ( Id. at p. 1149.) At certain sites, there were several pollution sources, some unexpected and some not. The insured wanted to argue at trial that the unexpected sources could be treated as separate "occurrences" that would be covered even though the expected sources were not covered. ( Id. at p. 1161.) However, the insurance policies covering the sites provided: "'All such exposure to substantially the same general conditions existing at or emanating from one premises location shall be deemed one occurrence.'" ( Id. at p. 1149.) At the beginning of its discussion, the FMC Corp. court described the insured's proposed argument: "A site might (for example) have had two sources of pollutants, each of which had contributed to groundwater contamination, and while FMC could hope to prove that it had not expected one of the sources to cause the damage it could not reasonably expect to prove that the damage caused by the other source was unexpected. If the totality of damage attributable to the two sources were considered a single occurrence, then FMC's inability to establish the 'unexpectedly' element as to one of the sources would jeopardize its prospects for coverage as to either source." ( FMC Corp., supra, 61 Cal.App.4th at p. 1161.) The Golden Eagle court quoted this statement ( Golden Eagle, supra, 85 Cal.App.4th at p. 1311), andInsurers rely on it as additional support for their allocation argument. However, both Golden Eagle and Insurers overlook the statement's context and the FMC Corp. court's subsequent discussion. The court only reached the merits of the insured's proposed argument with respect to one site. At that site, there were four groundwater pollution sources, only three of which were unexpected. The insured argued there were multiple "occurrences," because the different pollution sources were separate "premises locations" within the site as a whole. (See FMC Corp., supra, 61 Cal.App.4th at pp. 1166, 1169.) The court disagreed, holding that "premises location" meant the entire site and not a particular location within the site. ( FMC Corp., supra, 61 Cal.App.4th at p. 1169.) Thus, FMC Corp.'s holding turned on the issue of the proper interpretation of the "one occurrence" and "one premises location" clauses in a multiple-source pollution case. Golden Eagle not only did not discuss that issue, the Golden Eagle court did not even indicate whether the policies in that case contained "one occurrence" or "one premises location" clauses. Instead, Golden Eagle relied on a wholly different analysis, founded on general principles of contract law and proof of damages. Accordingly, despite the Golden Eagle court's claim that FMC Corp. was "squarely on point" ( Golden Eagle, supra, 85 Cal.App.4th at p. 1314), FMC Corp. does not support Golden Eagle's "allocation" analysis. In fact, FMC Corp. actually is inconsistent with Golden Eagle's allocation analysis, since FMC Corp. did not allow allocation of pollution at one site among various sources of pollution at that site. The policies in this case contained "one occurrence" clauses, but Insurers do not rely on those clauses, and the parties have not briefed or argued the effect of the clauses. Therefore, it is not appropriate for this court to address the matter, and FMC Corp. is no more relevant here than it was in Golden Eagle. At any rate, FMC Corp., like Golden Eagle and Lockheed, failed to consider Partridge's concurrent cause analysis and is unreliable for that reason as well. For the foregoing reasons, we do not consider Golden Eagle to be persuasive precedent for applying the "allocation" analysis to deny recovery in this case. Since Lockheed relied uncritically on Golden Eagle to reach the same holding, that case is no more persuasive. Even if the two cases were persuasive in themselves, as stated they are incompatible with Partridge and Garvey which, as Supreme Court precedent, must control.
It is interesting to note that Justice Kaufman was one of the original coauthors of the cited treatise. Evidently, though he did not agree with Partridge, he did agree that Golden Eagle was not consistent with Partridge.
5. Application of Partridge to this caseThe property damage coverage clauses of the policies in this case obligated Insurers "[t]o pay on behalf of the Insured all sums which the Insured shall become obligated to pay by reason of liability imposed by law, including Chapter 1681 of the State of California Statutes of 1963, or liability assumed by contract, insofar as the State may legally do so, for damages, including consequential damages, because of direct damage to or destruction of tangible property (other than property owned by the Insured), including the loss of use thereof, which results in an Occurrence during the policy period." The policies defined an "Occurrence" as "an accident, event or happening including continuous or repeated exposure to conditions which results, during the policy period, in . . . Property Damage neither expected nor intended from the standpoint of the Insured." The insuring language in Partridge was substantially the same, providing: "'This Company agrees to pay on behalf of the Insured all sums which the Insured shall become legally obligated to pay as damages because of bodily injury or property damage, to which this insurance applies, caused by an occurrence.'" ( Partridge, supra, 10 Cal.3d at p. 99, fn. 5.) "Occurrence" was defined as "'an accident, including injurious exposure to conditions, which results, during the policy term, in bodily injury or property damage.'" ( Ibid.) Based on that language, Partridge found coverage because "although the accident occurred in a vehicle, the insured's negligent modification of the gun suffices, in itself, to render him fully liable for the resulting injuries." ( Partridge, supra, 10 Cal.3d at p. 103.) Therefore, the resulting damages were "under the language of the homeowner's coverage clause, 'sums which the Insured . . . [became] legally obligated to pay' because of the negligent filing of the trigger mechanism. . . ." ( Ibid.) The insured's liability for modifying the gun existed "independently of any 'use' of his car," and therefore was covered. ( Ibid.) Applying that same reasoning to the substantially similar policy language in this case would support the conclusion that the damages from the escape of contamination from the site likewise were covered. There is evidence from which a reasonable trier of fact could conclude the 1969 discharge was "neither expected nor intended. . . ." If it were proven that the 1969 discharge was a concurrent cause of indivisible damages for which the State was held liable, then under tort law principles the State's liability for that discharge would suffice, in itself, to render the State liable for all of the damages. In that event, all of the damages would be "sums which the Insured shall become obligated to pay by reason of liability imposed by law" under the insuring clauses of the policies. To borrow Partridge's joint tortfeasor analogy, suppose that in this case the State's negligence caused the 1969 discharge, but another party's negligence caused the other discharges from the site and that all of the discharges contributed to an indivisible injury. Under Partridge, because the State would be jointly and severally liable for all of the resulting damage and not just the amount directly traceable to its own negligence, if the 1969 discharge were covered Insurers would be liable to indemnify the State against all of its joint and several liability. Partridge makes clear that the result is no different merely because all of the negligent conduct is committed by one tortfeasor instead of two. Insurers argue that a court should not apply the concurrent cause analysis until the insured satisfies its initial burden to "prove what damages resulted from the particular occurrence for which it seeks coverage, not simply 'all sums.'" Therefore, Insurers assert, the State must establish "how much of the damages it seeks coverage for was caused by the 1969 event." However, Partridge holds that to determine whether an injury was "caused" by a particular event for purposes of deciding whether the injury is covered, a court must ask whether the event was "a" cause of the injury. It is not necessary that the event be the only cause. Therefore, the concurrent cause analysis affects the coverage question, not just the question of what damages the insured has suffered as a proximate result of the breach of the insurance policy. Moreover, the policies in this case do not say that liability for damages is covered only if the damages solely "result from" or are solely "caused by" a particular event. Instead, they say that liability is covered if it is imposed "because of property damage that "results in" an occurrence. That the liability also may have been imposed partially "because of property damage that does not result in an occurrence does not preclude coverage. At least, that is one reasonable interpretation of the policy language. If there is a reasonable interpretation of an insurance policy that supports coverage, this court is, of course, required to adopt that interpretation. ( TRB Investments, Inc. v. Fireman's Fund Ins. Co., supra, 40 Cal.4th at p. 27.) Insurers in the joint tortfeasor situation would, of course, have a right of subrogation against the other tortfeasor. In the insurance context, the doctrine of equitable subrogation "'permits the paying insurer to be placed in the shoes, of the insured and to pursue recovery from third parties responsible to the insured for the loss for which the insurer was liable and paid.' [Citation.]" ( United Services Automobile Assn. v. Alaska Ins. Co. (2001) 94 Cal.App.4th 638, 645 [ 114 Cal.Rptr.2d 449].) Presumably, to exercise that right Insurers would have to show what proportion of the State's joint and several liability was caused by the other tortfeasor. Under Partridge, though, Insurers would still be liable to indemnify the State against its total joint and several liability and not just the part the State could prove was caused by its own negligence. As Insurers point out, Partridge differs from this case that in Partridge the damage was instantaneous, while in this case the damage from the discharge of contamination accrued gradually. However, Partridge only required for coverage that the covered cause be a "concurrent" cause of the damage. There was evidence from which a reasonable trier of fact could conclude the 1969 discharge was a concurrent cause of the damage for which the State was held liable. The State's expert report from 2004 concluded that by 1960 contaminants exited the subsurface of the site around the east and west ends of the concrete barrier dam and through the fractured bedrock underneath the dam. The report stated that each of these three "flow pathways" leading to contaminant exit from the site — east, west, and under the dam — was separate from the others and resulted from different causes. However, the report also identified a "fourth path flow" that existed downstream of the site in the alluvium, weathered bedrock, and fractured bedrock. The report stated that once the contaminants exited the site through the three pathways already described, the contamination plume continued to migrate downgradient due to the continuous motion of the groundwater to the southwest. The plume contaminated both soils and groundwater farther downstream and extended several thousand feet downgradient of the site. Through at least the late 1980's, the plume was moving progressively farther from the site, and damage to the soils and groundwater downgradient of the site was still ongoing in 2004. The report further stated that a program of soil sampling conducted downstream of the site after the 1978 discharges indicated waste contaminated soils downgradient of the site and that "it is reasonable to expect that much of the soil contamination detected was from the 1969 event." The report went on to state: "Because the soils in the canyon below the Site are permeable, it is probable that the 1969 flood contributed to downstream groundwater contamination in addition to soil contamination." The author of the report, V. Stephen Reed, testified at a deposition that the soil contamination from the 1969 discharge was "part" of the contamination that was measured in the sampling program after the 1978 discharges, because "some" of the contaminated soil remained in "some places" during the time between the 1969 and 1978 discharges. Similarly, a report of January 1977 prepared by consulting engineers concluded that the abrupt appearance of higher contamination levels at a monitoring well in 1972 was most likely caused by continued leaching of contaminants deposited during the 1969 floods. A reasonable trier of fact could conclude from the above evidence that contamination from the 1969 discharge entered and remained in the soil and groundwater at the same time that the subsurface leakage around and under the barrier dam and the 1978 discharges were occurring. Reed's conclusions that "some" of the contamination found in the sampling program after the 1978 discharges was caused by the 1969 discharge and that the 1969 discharge "contributed" to the downstream soil and groundwater contamination meant that the remainder of the contamination must have been caused by one of the other sources of contamination. Therefore, the 1969 discharge and the other discharges must have combined to cause the entirety of the contamination, or so, at least, a reasonable trier of fact could infer. For this reason, Insurers' claim that the injury in this case was not indivisible is unfounded. In fact, Insurers effectively premised their allocation argument on the assumption that the injury was indivisible, because the thrust of the argument is that the State is required to, but cannot, divide the injury among the multiple causes that produced it. Similarly, Insurers supported their argument with Golden Eagle, in which the court found no coverage because the insured admitted its damages were "'indivisible.'" ( Golden Eagle, supra, 85 Cal.App.4th at p. 1315.) If, as Insurers now argue, the injury was not indivisible, then their reliance on Golden Eagle was incorrect. Thus, if the 1969 discharge was covered, the situation in this case would be analogous to that in Partridge, with covered and uncovered causes acting concurrently to cause an injury that could not be allocated among them. In that event, a reasonable trier of fact could conclude that the amount for which the State was held liable represented the damage attributable to an indivisible injury resulting from several causes. The fact that not all of the damage occurred instantaneously as in Partridge should not affect the applicability of Partridge here. That point is illustrated by State Farm Fire Cas. Co. v. Kohl (1982) 131 Cal.App.3d 1031 [ 182 Cal.Rptr. 720], another case not mentioned in either Golden Eagle or Lockheed. In Kohl, the insured hit a motorcyclist while driving his truck, injuring her. After the collision, he dragged her out of the street, which she alleged caused her additional serious injury. ( Id. at p. 1034.) Even though not all of the damage occurred instantaneously, or even concurrently in a strict temporal sense, the court held, citing Partridge, that liability for the damage was covered by the insured's homeowner's policy. Since one of the acts that caused the injury, dragging the victim from the street, did not arise from the use of the insured's vehicle, the vehicle exclusion did not apply. ( Id. at p. 1039.) Coverage in this case also is supported by the definition of "occurrence" in the policies. As noted, "occurrence" was defined to mean not only a discrete event, but also "continuous or repeated exposure to conditions" if the exposure resulted in property damage. An occurrence therefore need not happen instantaneously in order to be covered. Insurers contend, however, that Partridge should not apply because in Partridge there was no doubt that the same injury and the totality of that injury resulted from the concurrent causes, while here, the injury began before and continued after the covered event, and the State only claims the 1969 discharge caused part of the total injury. First, the Supreme Court has made clear that a concurrent tortfeasor is fully liable for an indivisible injury regardless whether that tortfeasor's conduct by itself would have caused the same injury or the totality of that injury. In American Motorcycle Assn., the court explained that in some cases, "it is simply impossible to determine whether or not a particular concurrent tortfeasor's negligence, acting alone, would have caused the same injury. Under such circumstances, a defendant has no equitable claim vis-á-vis an injured plaintiff to be relieved of liability for damage which he has proximately caused simply because some other tortfeasor's negligence may also have caused the same harm." ( American Motorcycle Assn. v. Superior Court, supra, 20 Cal.3d at pp. 588-589.) Since under Partridge coverage is measured by the extent of the insured's liability to the injured party, the quoted language from American Motorcycle Assn. means that there is coverage for the entire injury even if it is not possible to prove the insured's covered act caused the same injury or the total injury. Second, the record does not show that the injury for which the State seeks coverage was not caused or contributed to by the 1969 discharge. All that is clear is that the discharge of contamination began before 1969. There is no indication when the resulting injury to the downstream property occurred. Under the policies' insuring clauses, coverage turns on the timing of the injury, not the act that causes the injury. An act is covered if it "results in an Occurrence during the policy period," which means that it "results, during the policy period, in . . . Property Damage neither expected nor intended from the standpoint of the Insured." Thus, the injury, though not the act, must occur during the policy period for the act to be covered. The policies in this case covered September 1976 to May 1978. Accordingly, any covered injury necessarily occurred after the 1969 discharge and therefore could have been contributed to by that discharge. The State's expert, in fact, concluded it was reasonable to expect that much of the contamination after the 1978 discharge was from the 1969 discharge. Insurers finally contend that because they are excess insurers, they have no liability to the State unless the State proves that the damage caused by a covered event exceeded the underlying coverage. This argument is founded on the same premise — that an insurer's duty to indemnify only extends to liability that the insured can prove is traceable to a covered cause — that we have already concluded Partridge requires us to reject. Therefore, we need not address the issue of excess insurance separately. For these reasons, we conclude Partridge's analysis should govern this case. Under that analysis, the State is not required to allocate its liability based on the cause of the underlying damage, as long as a covered cause is a concurrent contributing cause. Since there is at least evidence raising a reasonable inference that the 1969 discharge contributed to the damage for which the State was held liable, under Partridge a reasonable trier of fact could conclude the State's liability was covered.
Insurers assert that in the unpublished part of this opinion this court described the 1969 rain event as an "intervening cause," and that an intervening cause cannot be a concurrent cause. What we actually said was that the 1969 discharge could qualify as a sudden and accidental "intervening event" as the term was used in Travelers Casualty Surety Co. v. Superior Court, supra, 63 Cal.App.4th at p. 1460, to mean an event after the initial discharge of waste into a site that causes appreciable damage over and above the routine dumping. ( Ibid.) In contrast, an "intervening cause" is "a later cause of independent origin" that relieves the original tortfeasor of liability if the intervening cause and the resulting damage were not foreseeable. ( Ballard v. Uribe (1986) 41 Cal.3d 564, 587 [ 224 Cal.Rptr. 664, 715 P.2d 624].) The 1969 discharge was not an "intervening cause" with respect to the subsurface discharges around and under the dam, because it was not wholly "later" (the discharges overlapped in time) and was not "of independent origin" (the State's negligence caused all of the discharges). Therefore, the earlier discharges and the 1969 discharge could be "concurrent causes" with respect to one another even though the 1969 discharge was an "intervening event" with respect to the original dumping.
At trial, the State would have to prove its damages were indivisible to claim coverage under Partridge. Insurers would not be foreclosed from offering proof that in fact the damages were not indivisible. Our point is only that, for purposes of summary judgment, Insurers effectively assumed in their allocation argument that the damages could not be divided.
Insurers note that in answers to interrogatories the State said that the subsurface escape of contaminants to the east, west, and under the dam, and the 1969 and 1978 discharges, were five separate "occurrences." However, the State only said that the damage from the five sources of contamination "began" at different times and locations. The State also said that the damage from all five sources continued through the policy periods and that each of the five occur rences had "contributed to the plume of contaminants." Thus, the State did not foreclose the possibility that the five sources combined to produce an indivisible injury. Moreover, the answers were verified on July 15, 2004, two weeks before the date of the expert report, July 30, 2004, and presumably were made without the benefit of the report. At any rate, Insurers have not argued that the answers have preclusive effect here, and the effect, if any, of the answers should be determined on remand, as it depends on factual as well as legal determinations.
As noted, ante, the insuring clause of the Columbia policy as it appears in our record does use the term "occurrence" to describe the event that triggers coverage. However, Columbia's insuring clause provides it will indemnify the State for "loss," with "loss" defined to mean "the sums paid as damages in settlement of a claim or satisfaction of a judgment for which the insured is legally liable. . . ." There is no requirement in the insuring clause that all of the damage from which the loss arises occur instantaneously.
III DISPOSITION
The summary judgment in favor of Insurers is reversed. The matter is remanded with directions to grant Insurers' alternative motions for summary adjudication establishing that liability for the gradual escape of pollutants from the site over the years and the 1978 discharges was excluded by the pollution exclusion. The parties shall bear their own costs on appeal. Ramirez, P. J., and Hollenhorst, J., concurred.