Full opinion text
AMENDED ORDER GRANTING MOTION FOR SUMMARY ADJUDICATION AGUILAR, District Judge. I. INTRODUCTION. This case presents an important question of law the resolution of which may have’ implications beyond the scope of this lawsuit. In this motion, plaintiff Intel Corporation (“Intel”) seeks an order requiring defendant The Hartford Accident and Indemnity Company (“Hartford”) to reimburse it for expenses incurred under a “comprehensive general liability” insurance policy. The question presented is whether the policy involved covers costs incurred in connection with the cleanup of hazardous waste located on and beneath the insured’s property. Although addressed in other jurisdictions, it is a question of first impression for appellate courts in the State of California. For reasons explained below, the Court concludes that the policy encompasses the claim made by Intel and that Hartford is liable under the terms of the policy for some, though perhaps not all, of the cleanup costs incurred by Intel. II. FACTUAL BACKGROUND. Plaintiff Intel is an international manufacturer of semiconductors with its corporate headquarters in Santa Clara, California. During the late 1970s and early 1980s, Intel’s production facilities were located in Santa Clara, Mountain View, and Liver-more, California, as well as Aloha, Oregon and Chandler, Arizona. This motion relates to the Mountain View, California production facility located on Middlefield Road on property leased by Intel from an entity called Renault and Handley. From 1968 through 1980, Intel conducted manufacturing operations on the Middle-field Road property leased from Renault and Handley. As part of its manufacturing processes at the Middlefield Road facility, Intel employed certain chemical solvents. These solvents contained one or more of the following elements or compounds: 1,1,1,-trichloroethane (“TCA”), trichloroethylene (“TCE”), trichlorobenzene (“TCB”), dichloroethylene (“DCE”), phenol, and xylene. Each of these solvents had been classified by the federal government as a “hazardous substance” within the meaning of the Comprehensive Environmental Response, Compensation, and Liability Act of 1980 (“CERCLA”) and remain so classified under the 1986 amendments to CERCLA. 42 U.S.C. § 9601(14). See also 40 C.F.R. §§ 116.4A, 302.4 (July 1987). Despite the extremely hazardous nature of these chemicals, Intel stored them in an unsecured underground storage tank. This tank apparently was intended only for temporary use, with the chemicals later transported to off-site locations for permanent disposal. Sometime in 1980, Intel ceased production at the Middlefield Road manufacturing plant and moved its operations to a larger facility. Due to the existence of its lease with Renault and Handley which ran through 1984, Intel undertook to sublet the Mountain View property. Supposedly at the request of a prospective sublessee, Intel commissioned soil sampling and testing of the Mountain View property in September 1981. The test results showed that the site was contaminated by hazardous waste solvents both in the soil and in ground water percolating beneath the soil. Upon learning of the contamination, Intel took several steps. The order and manner in which these steps were taken is not relevant here. It suffices to note that Intel contacted three environmental consulting firms to investigate the nature and extent of the problem, and notified several relevant government agencies with jurisdiction encompassing such problems. After the investigation revealed that the preliminary report was correct, i.e., that the soil adjacent to the solvent tank was contaminated and the ground water beneath the property had been despoiled, Intel began cleanup efforts. The tank was excavated, the adjoining soils removed, and the contamination of the ground water was addressed by the use of a pump-absorption filtration system. Ultimately, in August 1985, Intel entered into a consent decree with the United States Environmental Protection Agency (“EPA”) for purposes of the remediation of the site. In The Matter of: Middlefield-Ellis-Whisman Area Mountain View, California (Fairchild Camera and Instrument Corp.; Intel Corp.; and Raytheon Corp.) (hereafter, simply the “Consent Decree”). The Consent Decree is typical of the settlements achieved by EPA with “potentially responsible parties” (“PRPs”) at CERCLA cleanup sites. EPA is more interested in achieving cleanup quickly than resolving the question of liability. Thus, Intel, as well as Fairchild and Raytheon (collectively, “respondents”) accepted responsibility without admitting liability. See Consent Decree at 2. Prior to achievement of the Consent Decree, on January 10, 1985, all three sites were listed on the California State Priority List by the California Department of Health Services (“DOHS”) pursuant to California Health and Safety Code § 25356. On April 30, 1985, the California Regional Water Quality Control Board (“RWQCB”) issued separate waste discharge requirements to each of the respondents and certain other parties requiring them to undertake further investigation of ground water quality and to prepare and implement plans for interim containment and cleanup at their respective facilities. At that time, the California RWQCB referred the sites to EPA for action under CERCLA. On May 15,1985, respondents and others were notified by EPA of the agency’s intent to conduct a remedial investigation and feasibility study (“RI/FS”) at the site. Intel and its co-dumpers were invited to participate in formulating a plan whereby the private parties rather than the EPA would undertake such an investigation. Subsequently, the EPA asked for and received written work plan proposals from each of the PRPs. Based on the submissions by the PRPs and the agreement of the California authorities, EPA and the respondents entered into the aforementioned Consent Decree. EPA made several findings in the Consent Decree which are relevant to this litigation. EPA stated that soil and ground water has been polluted with organic solvents including, but not limited to: TCE, TCA, TCB, DCE, xylene, and phenol. The sites containing the facilities of each of the respondents already had been proposed for inclusion on the National Priorities List. See 40 C.F.R. Part 300, Appendix B, 49 Fed. Reg. 40320 (October 15, 1984). The hazards these sites pose to the public are underscored by the fact that the City of Mountain View has one public water supply well within one-half mile of the contaminated area and additional public water supply wells within one mile of the area. Consequently, EPA concluded that the “actual and threatened release of hazardous substances from facilities at the Site may present an imminent and substantial endangerment to the public health or welfare or the environment.” Consent Decree at 6(G). Acting pursuant to its authority under CERCLA, EPA certified that the work plan submitted by Intel, Fairchild, and Raytheon is consistent with the National Contingency Plan (“[W]ork to be performed by Respondents under this Consent Order ... is consistent with all applicable requirements of the National Contingency Plan”). See 42 U.S.C. § 9601(25). EPA further stated that “all costs reasonably incurred for such work are necessary costs of response.” Consent Decree at 6(1). Intel and its corespondents were required to remit to the EPA $50,000 to cover EPA’s costs for oversight and response costs in connection with the cleanup. Finally, the EPA stated: The actions and work required by this Consent Order are necessary to enable EPA or any other party to remove or remedy the existing or threatened conditions at the Site, and to further evaluate, prevent or minimize the release or migration or the threatened release or migration of hazardous substances to the environment and to protect the public health and welfare and the environment. From April 1,1976 through April 1, 1988, Intel owned a continuing insurance policy with defendant Hartford covering the Mountain View production facility. The “comprehensive general liability” insurance policy (hereafter, the “Policy”) contained within its terms the following passage: The company will pay on behalf of the insured all sums which the insured shall become legally obligated to pay as damages because of Coverage A — bodily injury or Coverage B — property damage to which this insurance applies, caused by an occurrence, and the company shall have the right and duty to defend any suit against the insured seeking damages on account of such bodily injury or property damage, even if any of the allegations of the suit are groundless, false or fraudulent, and may make such investigation and settlement of any claim or suit as it deems expedient, but the company shall not be obligated to pay any claim or judgment or to defend any suit after the applicable limit of the company’s liability has been exhausted by payment of judgments or settlements. Upon learning of its predicament, Intel notified Hartford in October 1981 of a claim for reimbursement for reasonable and necessary investigation and cleanup costs incurred by Intel in connection with the Mountain View facility. The basis for its claims was the above-mentioned portion of its policy and the portion of the policy which defined an “occurrence.” As discussed below, there is some dispute as to when the claim actually was lodged, but it is clear that Hartford denied the claim on May 19, 1982, because of “Exclusion K” which excluded coverage for damage to “property owned or occupied by or rented to the insured.” After Hartford’s denial of coverage, the parties engaged in a protracted squabble regarding their respective positions. Unable to reason with the insurer and facing potentially serious liability, Intel filed suit in California Superior Court in the County of Santa Clara. On November 20, 1986, Hartford removed the case to federal court. The matter was assigned to Judge William A. Ingram of this Court. After entertaining a motion to remand, Judge Ingram dismissed the action without prejudice because of lack of derivative jurisdiction under the applicable removal statute and lack of diversity. See Intel Corp. v. The Hartford Accident and Indemnity Co., 662 F.Supp. 1507 (N.D.Cal.1987). On June 2, 1987, Intel refiled in California Superior Court in Santa Clara County. Once again, Hartford removed to this Court. This time, however, the case was assigned to the RPA docket. Intel did not make a motion to remand and for unknown reasons the case was not related to Judge Ingram’s prior filing. Thus, the case remains as docketed and this Court must now address Intel’s motion for summary adjudication. III. DISCUSSION. (A) Introduction: The complaint contains a host of claims against Hartford including, but not limited to, fraud, deceit, intentional and negligent misrepresentation, breach of contract, breach of fiduciary duty, tortious breach of the implied covenant of good faith and fair dealing, breach of the insurer’s statutory duties, civil conspiracy, and, of course, violation of the Racketeering Influenced and Corrupt Organizations Act (“RICO”), 18 U.S.C. §§ 1961-68. In this motion, Intel seeks summary adjudication of the issue of the scope of the insurance policy. As stated by Intel, “[t]he sole issue presented in this Motion is whether insurance coverage exists under the policy for Intel’s reasonable and necessary costs incurred in the investigation and clean-up (sic) of hazardous chemicals at its Mountain View facility in order to prevent or mitigate claims or possible claims of injury or damage to third parties.” (B) Is There A Case or Controversy? Before taking up the contract issues presented, the Court must address defendant’s arguments surrounding the case or controversy requirement of Article III of the United States Constitution. The bulk of Hartford’s opposition to this motion focuses on the technical issue of the proper filing of an insurance claim. Hartford insists that Intel has not established that it actually has filed a claim with Hartford or that it has incurred costs for which Hartford is responsible. The argument is concisely stated as follows: Intel has not demonstrated that it ever made a specific claim for anything beyond some shapeless amalgam of “costs for investigation and clean-up [sic],” and has certainly not shown that it has spent even a dollar which might be arguably covered under the pertinent comprehensive general liability insurance policy. It is true that Intel has not filed any receipts or proof of payment with this Court. Nor has Intel filed any proof of claim. But no such proof is necessary. Intel has made these accusations in its complaint, and Hartford has admitted them in its answer. To wit, in paragraph 22 of its answer, Hartford responds to Intel’s assertion of having filed a claim and incurred costs in the following language: [Hartford] admits that on or about February 16, 1982, plaintiff notified Hartford of contamination at one of plaintiff’s facilities on Middlefield Road in Mountain View, California, and admits that plaintiff has since made claims for reimbursement for investigation and cleanup costs at that site. This admission satisfies the ease or controversy requirement of Article III. Hartford’s arguments now are akin to the proverbial sealing of the barn door after the bovines have gone to graze in the pasture. Moreover, Hartford’s argument here, in light of its previous admission, borders on sanctionable conduct. To admit that an insured has made a claim and then later to deny that any claim exists smacks of bad faith. There is a real dispute between the parties. Intel has filed a claim and Hartford has denied it. The matter properly has been presented to the Court as a case for judicial resolution. (C) By Its Terms Does the Insurance Policy Cover This Claim? The parties agree that the Policy covers claims made for damages to third parties. As quoted above, coverage includes claims made against Intel for bodily injury or property damage. Intel has made a claim against the Policy for coverage of damage presently and prospectively to third-party property, and prospectively for the threat of injury to third persons. In the face of apparent coverage, Hartford raises two provisions of the Policy as defenses to avoid coverage — Exclusion F and Exclusion K. (D) Analysis of Exclusion F — The Pollution Exclusion: Hartford denies coverage on the basis of two exclusions. The first of these exclusions is Exclusion F which provides: This insurance policy does not apply ... (f) to bodily injury or property 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 the land, the atmosphere or any water course or body of water; but this exclusion does not apply if such discharge, dispersal, release or escape is sudden and accidental. The construction of this exclusion and its application is an issue over which courts have differed. See New Castle County v. Hartford Accident and Indemnity Co., 673 F.Supp. 1359, 1363 (D.Del.1987) (listing cases and stating that “there is clearly a split of authority”); Broadwell Realty Services, Inc. v. Fidelity & Casualty Co. of New York, 218 N.J.Super. 516, 528 A.2d 76 (N.J.Super.App.Div.1987) (listing cases and noting “some degree of disarray”). However, this Court need not address the merits of this claimed exclusion because the Court concludes that Hartford has waived its right to invoke the exclusion. The factual basis for the Court’s conclusion is found in the history of the dispute over this claim by Intel against the Hartford Policy. Declaration testimony of Rossmore C. Robert, Division Supervisor in the Claims Department of Hartford Insurance, establishes that Hartford received notice of the claim no later than February 16, 1982. On March 31, 1982, a Hartford claims representative responded to the claim by referring Intel to certain provisions of the Policy. Specifically, Hartford asked Intel to “[pjlease review [the Policy]/’ while directing Intel’s attention to “exclusions F, Kl, and L, Broad Form Property Damage Endorsement form L3252-2[,] exclusion Y1 and Contractual Liability Insurance Coverage Part 3523-0 exclusions G and H.” Hartford stated that “[i]t would appear [that] these exclusions may apply,” but the claims representative went on to emphasize that “we will continue our investigation” and “we will contact you as soon as a decision regarding coverage is made.” Thus, it is unequivocally clear that as of March 31, 1982, Hartford had made no decision on the applicability of the aforementioned exclusions; none of the at least six potential exclusions had been invoked or ruled out from Hartford’s point-of-view. On May 19, 1982, three months after the claim was filed and seven weeks after reporting that it continued to investigate the matter, Hartford informed Intel that it was denying the claim. Aside from the greeting and closing salutation and an introductory and concluding sentence on either side of the greeting and salutation, the entirety of the text of the letter is as follows: After a careful review of the facts of this loss and the terms of your policy # 57 CM X0230, policy term, 4/1/81 through 4/1/82.- (sic) It has been determined that coverage will be afforded for bodily injury or property damage claims brought against Intel by the Silvas. There will be no coverage for any claims for damages, loss of rent, etc. to the facility at 365 Middlefield Rd., Mt. View, CA. Exclusion K form L 3503-1 would apply. This letter unmistakably documents the fact that the sole basis for denial of Intel’s claim was Exclusion K — which excludes coverage for damage to the insured’s property as opposed to that of third parties. The pollution exclusion, Exclusion F, is not stated as a basis for denial of coverage, nor is the exclusion even mentioned. Similarly, in a letter from the supervisor of the Hartford claims division in San Jose, California, Nate Swainson, Hartford reiterated that the basis for its denial of coverage was Exclusion K. Again, no mention was made by Swainson of Exclusion F. Another letter in March 1986 also referenced Exclusion K but omitted any mention of Exclusion F. In connection with this motion, Hartford submitted the declaration of Rossmore C. Robert, Hartford’s Division Supervisor in the Claims Department. Mr. Robert stated that he was the custodian of the original claims file on Intel’s policy. Mr. Robert stated that “[a]ll materials which are generated in connection with a ‘claim’ ... are kept in the ‘claims file’.” Declaration of Rossmore C. Robert at 2:20-21. This file contains not only records of contact by the insured to the insurer, but also “copies of all internal memoranda or correspondence from Hartford to its insured or other parties.” Id. at 2:18-19. Mr. Robert’s declaration constituted a synopsis of the claims file of Hartford with respect to Intel’s claim with the actual documents appended as exhibits. Robert’s declaration and the claims file contain no mention of Exclusion F after the preliminary letter of March 31, 1982. Notwithstanding the Robert declaration, the claims file, and Intel’s assertion that “Hartford for the very first time in this case, has raised exclusion (f) as a policy defense, the ‘sudden and accidental’ provision [sic],” Intel Reply Brief at 9:22-24 (emphasis added), the parties have filed papers which contradict their own and each others’ positions. In its complaint, Intel asserted the following: On or about November 19, 1982, and upon several separate occasions thereafter, both orally and in writing, Hartford notified Intel that it was denying coverage for Intel’s claim arising out of the leakage at its Santa Clara III and Santa Clara Magnetics plants, as well as at other plants, based on Exclusion (f) in the policy, in that the loss was not sudden and accidental, the same exclusion relied upon by Hartford unsuccessfully in denying its insured’s claim in the Coastal Chemical case, in addition to Exclusion (k) previously referred to herein. Thus, although not explicitly stating so, Intel implies in paragraph XXIV of the complaint that Hartford had invoked Exclusion F as a basis for denial of coverage of November 19, 1982, and thereafter. In its answer to the complaint, Hartford explicitly stated that it had told Intel that coverage was excluded under Exclusion F. At paragraph 24 in its answer, Hartford stated: 24. Hartford denies the allegations of paragraph 24, except admits that on or about November 19, 1982, Hartford advised plaintiff that there was no coverage available for damage to plaintiff’s own property at its site on Middlefield Road in Mountain View, California, because of exclusion “f” in the policy. Rather than inducing the parties to square each’s own apparently contradictory statements, the Court will simply view the matter in the light most favorable to Hartford. In that light, Hartford may be supposed to have invoked the pollution exclusion on or about November 19, 1982. The question presented, then, is whether after citing and relying upon one exclusion as the basis for denial of coverage, Hartford may later invoke an entirely different exclusion as an additional basis for denial of coverage? Judge Patel of this Court confronted precisely this question in the case of McLaughlin v. Connecticut General Life Ins. Co., 565 F.Supp. 434 (N.D.Cal.1983). After concluding that the California Supreme Court had not spoken on the subject, Judge Patel ruled: [Tjhis court concludes that the California Supreme Court would hold that an insurer waives its right to rely on defenses not specified in its denial [of coverage] which a reasonable investigation would have uncovered. 565 F.Supp. at 452. Having retraced the research done by Judge Patel and examined California law since the decision in McLaughlin, this Court finds that it remains true that the California Supreme Court has not spoken on the subject. Furthermore, the Court finds that the holding in McLaughlin is both fair and a faithful extrapolation of the laws of California. As the above discussion points out, and as has been highlighted in the development of the covenant of good faith and fair dealing in California courts within the past few years, California places significant obligations on insurers in handling claims of insureds. For example, the California Supreme Court in Egan v. Mutual of Omaha Insurance Co., 24 Cal.3d 809, 817, 169 Cal.Rptr. 691, 620 P.2d 141 (1979), cert. denied, 445 U.S. 912, 100 S.Ct. 1271, 63 L.Ed.2d 597 (1980), held that an insurer may breach the implied-in-law covenant of good faith and fair dealing if it fails properly to investigate its insured’s claim. See also Hanson v. Prudential Ins. Co. of America, 772 F.2d 580, 584 (9th Cir.1985) (citing Egan), modified on different grounds, 783 F.2d 762. A natural concomitant of this ruling is that an insurer has a duty of candor to explain or at least make reference to the bases for denial of policy coverage. In this case, Hartford informed Intel of its reliance on Exclusion F at least six months after denying the claim because of Exclusion K. The fact that Hartford was aware of the potential availability of Exclusion F at the time it denied the claim is established by the fact that Hartford wrote the contract, and Hartford itself listed the exclusion in the letter of March 31, 1982 among the six exceptions that Hartford thought “may” be available. Furthermore, Hartford was fully aware of the nature of the risk. In the letter of November 1, 1982, from Nate Swainston to Intel, Swainston actually commended Intel for its cleanup operations. Swainston stated that “[w]hat has been established thus far is that there was a leakage at the Mountain View cite (sic) that resulted in some water and soil contamination of that cite (sic).” Possessing undisputed knowledge of the nature of the risk and the availability of Exclusion F to exclude that risk from coverage, Hartford had an obligation to Intel to inform it if Exclusion F was to be invoked. This conclusion is buttressed by the fact that at least twenty other states have adopted laws imposing an obligation on insurers to state all the bases for their denial of coverage which they know or should have known at the time of the first denial. For a list of cases, see 16C Appleman, Insurance Law and Practice § 9260, at 393 (West 1981 & Supp.1987) (hereafter. “Appleman ”). The proposition, as stated by Appleman with abundant citations, is that once a specific ground of forfeiture or exclusion is offered to deny the claim, “all other grounds are waived.” Id.; see, e.g., Luria Bros. & Co. v. Alliance Assur. Co., Ltd., 780 F.2d 1082, 1090 (2d Cir.1986) (quoting Appleman and citing New York law). Hence, for example, the Ninth Circuit has concluded that: Under Oregon law when an insurer denies liability upon a specific ground, other grounds of forfeiture then within its knowledge are waived. “Good faith” requires an insurer to 'apprise the insured “fully of its position, and, failing to do this ... [the insurer] is estopped from asserting any defense other than that brought to the notice of the plaintiff. Dillingham Corp. v. Employers Mutual Liability Ins. Co. of Wisconsin, 503 F.2d 1181, 1185 (9th Cir.1974), quoting Ward v. Queen City Fire Ins. Co., 69 Or. 347, 138 P. 1067 (1914). The Ninth Circuit’s analysis of Oregon law provides a sound basis for inferring this doctrine of waiver to California law, but it also raises one issue respecting construction of the doctrine. Dillingham refers to both waiver and estoppel as if the two were interchangeable, when in fact there is a subtle but significant distinction between the two. Waiver relates to one party’s forfeiture of certain rights. For example, waiver of the right against self-incrimination (in the criminal context) or waiver of notice. Generally, it makes no difference whether the opposing party relied upon the waiver or not. The unilateral act is decisive. In contrast to waiver, estoppel always involves some elements of reliance by the opposing party. The classic definition of equitable estoppel is provided in the Restatement 2d of Torts § 894 which requires some action or forbearance on the part of a third party. In other words, estoppel is a bilateral activity which involves inquiry into the actual relationship of the parties. The question is whether California law would adopt a unilateral or bilateral focus, i.e., would the California Supreme Court require some element of reliance or prejudice on the part of the insured before foreclosing an insurer from raising a ground for denial of liability that was known or reasonably knowable but not raised at an earlier date. This Court concludes that the California Supreme Court’s approach would be to focus on the unilateral action of the insurer in failing to investigate and raise all proper defenses to the claim at the outset. This conclusion is rooted, first, in Egan and other cases emphasizing the insurer’s duty to investigate. Once a claim is presented, this duty exists regardless of the posture of the insured. Second, the California Supreme Court’s analysis of the implied covenant of good faith and fair dealing suggests that a great deal flows from the relationship between the parties. Both parties owe good faith obligations to the other. “Sandbagging,” in this case bringing a defense that should have been known but was not raised earlier, would seem condemnable because it runs contrary to these good faith obligations. The third and final reason for constructing the doctrine in terms of waiver rather than estoppel is that the waiver approach is already an important component of California insurance law. For example, under California law, an insurer may be held to have waived the terms of a policy if it defends an insured without reservation of its right. Miller v. Elite Ins. Co., 100 Cal.App.3d 739, 754, 161 Cal.Rptr. 322 (1980); Phoenix Ins. Co. v. U.S. Fire Ins. Co., 189 Cal.App.3d 1511, 1527-28 n. 15, 235 Cal.Rptr. 185 (1987) (citing Miller), modified on different grounds, 190 Cal.App.3d 825A, rev. denied, June 17, 1987. Miller and the decisions following it are in accord with the application of a theory of waiver to defenses that are known and not raised by an insured. For this reason, and for all the reasons outlined above, the Court concludes that the California Supreme Court would hold “that an insurer waives its right to rely on defenses not specified in its denial which a reasonable investigation would have uncovered.” McLaughlin, 465 F.Supp. at 452. (emphasis added) Applying this law to the facts of this ease yields the inescapable conclusion that Hartford has waived its right to raise Exclusion F as a defense in this lawsuit. Hartford's letters to its insured clearly establish that it was aware of the exclusion. It is equally clear that Exclusion F, although mentioned early on as a potential basis for denial of coverage, was never invoked by Hartford in denying Intel’s claim. Some six months later, Hartford sought to raise the Exclusion F defense. This Court holds that Hartford cannot raise Exclusion F because it has waived the application of the exclusion. (E) Applicability of the Theory of Estoppel: The Court is aware that there are cases stating that estoppel is the proper approach to this issue. For example, in Wright v. Newman, 598 F.Supp. 1178 (W.D.Mo.1984), aff'd, 767 F.2d 460 (8th Cir.1985), the court stated that “under Missouri law the loss of an insurer’s defense is a matter governed by the principles of estoppel, with an attendant requirement of some showing of prejudice to the insured.” 598 F.Supp. at 1201. The court in Wright went on to state that “I believe ... that this position indeed reflects the mainstream of thinking in American law.” Id. Although the court in Wright may be correct, the citations in the Appleman treatise on which the Wright court relied do not resolve the issue. Appleman provides references from approximately a dozen jurisdictions. In contrast, the waiver rule described in Appleman was referenced in some twenty jurisdictions. The best that can be said is that the cases are split. Courts do often discuss the issue in terms of estoppel and, at times, use the term “estoppel” interchangeably with “waiver” as in Dillingham, supra, 503 F.2d 1181. But this is not surprising given the facts of the eases. Where an insurer has “sandbagged” an insured to the obvious detriment of the insured, equity cries out for estoppel. At that point, the question of estoppel versus waiver is an irrelevant doctrinal distinction. In this case, involving a district court’s educated assessment of what the California Supreme Court would do, it is appropriate to attend to the doctrinal distinction. Accordingly, for purposes of clarity, this Court rules that under an estoppel theory, Intel has not put forth evidence of injurious reliance necessary to support invocation of estoppel. The mere passage of six months between the denial of coverage and the invocation of Exclusion F does not establish prejudice. Without some showing of injurious reliance, Intel cannot lay the foundation required to assert estoppel. Therefore, although the Court finds that the waiver theory is appropriate in this case, there is insufficient evidence to sustain the theory of estoppel. Nevertheless, for the reasons described above in the discussion of waiver, as a matter of law, Hartford may not raise Exclusion F as a defense in this lawsuit. (F) Exclusion K: Before addressing the parties’ assertions relating to the interpretation of Exclusion K, it is necessary to review relevant California law concerning the construction of insurance contracts in general, and exclusions in particular. (i) California Law on Insurance Contracts: California has well-developed principles which inform and constrain the interpretation of insurance contracts. To begin with, “an interpretation of [a] policy is a legal rather than a factual determination.” Congleton v. Nat’l Union Fire Ins. Co., 189 Cal.App.3d 51, 59, 234 Cal.Rptr. 218 (1987) (emphasis in original), rev. denied, March 25, 1987. Moreover,. an insurance contract, like any other contract, is construed against the party who prepared the document. Federal Leasing Consultants, Inc. v. Mitchell Lipsett Co., 85 Cal.App.3d Supp. 44, 48 (Super.App.Dept.1978). In other words, where contract terms are ambiguous they will be construed against the author of the document. Id. This is particularly true in the context of insurance contracts where California courts have made it a policy to give the benefit of any doubts to insureds. See Admiralty Fund v. Peerless Ins. Co., 143 Cal.App.3d 379, 385, 191 CahRptr. 753 (1983) (providing citations and discussion of cases and principles favoring insureds over insurers). Courts are admonished to consider each insurance contract with an eye toward achieving “its manifest object of securing indemnity to the insured for the losses to which the insurance relates.” Id., quoting Crane v. State Farm Fire & Casualty Co., 5 Cal.3d 112, 115, 95 Cal.Rptr. 513, 485 P.2d 1129 (1971); Beaumont-Gribin-Von Dyl Management Co. v. California Union Ins. Co., 63 Cal.App.3d 617, 622, 134 Cal.Rptr. 25 (1976). Thus, “[ejxclusions are narrowly construed and uncertainties resolved in favor of the insured.” Nat’l Union Fire Ins. Co. v. Estate of Meyer, 192 Cal.App.3d 866, 871, 237 Cal.Rptr. 632 (1987), rev. denied, August 20, 1987; Reserve Ins. Co. v. Pisciotta, 30 Cal.3d 800, 807-08, 180 Cal.Rptr. 628, 640 P.2d 764 (1982) (“Whereas coverage clauses are interpreted broadly so as to afford the greatest possible protection to the insured ..., exclusionary clauses are interpreted narrowly against the insurer____”); see, e.g., Silberg v. California Life Ins. Co., 11 Cal.3d 452, 464-66, 113 Cal.Rptr. 711, 521 P.2d 1103 (1974) (interpreting ambiguous policy provision against the insurer). With these principles in mind, the Court now turns to an analysis of Exclusion K. (ii) Introduction to Analysis of Exclusion K: The relevant text of Exclusion K is as follows: This insurance does not apply ... (k) to property damage (1) to property owned or occupied by or rented to the insured____ The Court must conclude that given the facts of this case, Exclusion K does not apply. Clearly, there is property damage to the land leased by Intel, but that is not the sole damage. The water beneath the subject property has been polluted and lives and property nearby have been placed in peril. These facts are borne out in the “Administrative Order On Consent” (hereafter “Consent Decree”) entered into by the EPA, Intel and two other polluters on August 15, 1985. (iii) The Consent Decree and Relevant Background: Before discussing the content of the Consent Decree, several background facts are noteworthy. The EPA’s practice in administering Superfund has been to encourage settlement and private cleanup wherever possible. When conducting its initial assessment of sites included on the National Priority List (“NPL”), EPA’s seeks to identify PRPs and to determine whether public monies must be spent to address the problem. If solvent PRPs are identified, EPA almost invariably will attempt to secure their cooperation. In particular, EPA encourages PRPs to perform, or at least participate in, the execution of a remedial investigation (“RI”) and the subsequent (“FS”). In the RI, the extent of the environmental degradation is assessed. The goal is to evaluate the scope and severity of the threat to the public and the necessity of immediate cleanup. In the FS, EPA examines all potential modes of removing the public threat and detoxification of the property and underground water. In view of their joint and several liability for the entirety of an cleanup expenses, PRPs usually will seek to participate in the RI as well as the FS. For those PRPs who are financially able, often the best course is for them actually to conduct the RI/FS. In that way, PRPs are not only fully informed of their potential liability, but also are able to control the FS process of selecting the appropriate cleanup method and technologies. The consequence is that PRPs may be able to keep cleanup costs to a minimum by urging the adoption of the least-cost alternatives. In this case, the EPA investigated the site, identified three PRPs and “invited” them to undertake the RI/FS themselves. See Consent Decree at 5. The Consent Decree embodies the respondents agreement to take remedial action pursuant to the specifications articulated by EPA. Several findings in the Consent Decree are relevant to the applicability of Exclusion K. First, EPA’s initial investigation disclosed soil and ground water contamination. Several biological harmful organic solvents were discovered lurking in and beneath the site including TCE, TCA, xylene, TCB, phenol, and DCE. Second, EPA determined that the actual or threatened release of those substances from the site may present “an imminent and substantial endangerment to the public health or welfare of the environment.” Consent Decree at 6. Third, in support of its finding of endangerment to the public, EPA found that the site was within one-half mile of one public water supply well and within one mile of additional wells. Furthermore, private wells exist “at or near” the site. Consent Decree at 4. Fourth, EPA concluded that the cleanup effort of which Intel plays a part is “necessary” to remove the harm and threat of harm, to “further evaluate, prevent or minimize the release or migration or the threatened release or migration of hazard- and “to protect the public health and welfare and the environment.” Consent Decree at 7 (emphasis added). ous substances' Fifth, EPA also concluded that all activities performed pursuant to the Consent Decree “are necessary costs of response,” Consent Decree at 6, and are consistent with the NCP, the blueprint for Congress’ assault on hazardous waste. All costs incurred are “apportionable expenses of response and implementation.” Sixth, the parties were required to pay $50,000 into the Hazardous Substance Response Trust to cover EPA’s oversight and response costs for the site. Seventh, the Consent Decree contains joint and several penalty provisions whereby the respondents must pay the EPA as much as $5,000 per week for delays or shortcomings in the RI/FS process. The sum and substance of the Consent Decree is that Intel, along with its co-polluters, must bear the expense of investigating and remedying damage and potential damage to the property itself, the underground water, adjoining property, and the public. Under California Water Code § 102 (West 1971 & 1988 Supp.), “[a]ll water within the State is the property of the people of the State.” While individuals may acquire a right to use water, id., no one has a right to pollute or otherwise damage the quality of the State’s water. This is codified in California Water Code § 12922 which declares that the people of the State of California “have a primary interest in the correction and prevention of irreparable damage to, or impaired use of, the ground water basins of this State.” These laws are enforced through penalty provisions under California law including California Water Code §§ 13300-13755. Criminal penalties may be imposed in certain circumstances. See California Water Code § 13387. Thus by polluting the ground water, Intel has damaged the property of all Californians — third-party damage of such a serious nature that it could warrant criminal prosecution in certain circumstances. Although the Court’s conclusion that ground water contamination is damage to third parties is sufficient to eliminate Exclusion K, other damage to third parties is also present. The EPA’s assessment of Intel’s property is tantamount to a governmental finding that the land is a public nuisance. Indeed, by some definitions the property is more than a mere nuisance for it presents “an imminent and substantial endangerment to the public health or welfare or the environment.” Consent Decree at 6. The costs incurred by Intel to abate this public danger must fall outside the ambit of Exclusion K. Moreover, Hartford’s concern that Intel is spending money merely to repair its own former leasehold property is obviated by the Consent Decree’s express finding that all costs incurred in connection with the cleanup “are necessary costs of response,” i.e., they are necessary to evaluate, prevent or minimize the migration or release or threatened migration and release of hazardous substances and are necessary “to protect the public health and welfare and the environment.” Consent Decree at 7. Under such circumstances, it is clear that Intel’s response costs are not encompassed within Exclusion K. It is worth emphasizing in this connection that Intel’s actions are not taken merely to restore the property of its lessor. As a PRP, Intel faces joint and several liability for all of the consequences of its contamination. Although Intel has consented to begin the cleanup and to conduct the RI/FS (along with Raytheon and Fairchild), the earlier discussion highlights the fact that Intel will have to bear the costs of cleanup. By volunteering to participate in the cleanup operation, PRPs are able to mitigate their liability by keeping response costs down. Regardless of consent, the truth is that these PRPs are marked for liability to the public and they are answerable to the EPA — as evinced by the penalty provisions of the Consent Decree. See Consent Decree at 15. Thus, it would be disingenuous and misleading for Hartford to suggest that Intel’s acts involve solely the cleanup of damage to the leased property. The EPA’s involvement serves to underscore the public nature of the damage. (iv) Case Law on the Subject: The parties agree that there is no, or virtually no, California case law addressing the issue presented here: whether Exclusion K can be applied to bar Intel’s claim for coverage under its comprehensive general liability policy. Hartford has supplied one recent case relating to the subject, a superior court decision from Los Angeles County, Protective Nat’l Ins. Co. of Omaha v. Union Oil Co., slip op. No. C514 463 (November 24, 1987). The case involves an attempt by Union Oil to receive reimbursement for investigative, cleanup and response costs incurred in connection with a petroleum processing site in the State of Washington. The policy provided coverage for loss resulting from “property damage.” The Court concluded that reimbursement for response costs — including investigation and cleanup costs — “does not constitute ‘property damage’ within the meaning of the Insurer’s policies.” Union Oil at 3-4. In assessing the holding of Union Oil, this Court must begin from the premise that although the California Superior Courts are “courts of record,” California Constitution art. 6,' § 1, their decisions are not dispositive in reflecting the status of California law. Whatever guidance to be derived from the Union Oil decision must be from its logic. Unfortunately, the opinion offers little reasoning to support its conclusion. Instead, the opinion merely cites the holdings in two opinions from the United States Court of Appeals for the Fourth Circuit (which will be discussed below). Hence, although Union Oil is a decision rendered by a California court, it offers nothing in the way of independent analysis or an evaluation of the competing approaches to the issue. With no California appellate precedent speaking directly to the issue, this Court has the task of determining what course the California Supreme Court would take if presented with this issue. As stated above, the Court concludes that California statutory law mandates a finding that contamination of ground water constitutes injury to third-party property. Furthermore, after examining decisions on this issue handed down by state and federal courts around the country, the Court finds that the best analyses support the Court’s conclusion that Exclusion K literally does not apply in this case. In cases applying state law that, like California, recognizes public ownership of all underground water, several courts have ruled that contamination to ground water is not excluded from coverage by provisions like Exclusion K in this case. In Port of Portland v. Water Quality Ins. Syndicate, 796 F.2d 1188 (9th Cir.1986), the Court addressed the issue whether water is “tangible property” for purposes of an insurance policy’s definition of “property damage.” The court applied Oregon law and noted first that in Oregon “all water in the state belongs to the public.” 796 F.2d at 1193. Second, the court observed that Oregon prohibits the discharge of pollutants into the water and provides a statutory vehicle to collect damages for illegal discharges. Id. However, because the Oregon Supreme Court had not spoken on the subject, the panel was obligated to anticipate the position that the Oregon court would adopt if given the opportunity. After examining cases from other jurisdictions, the court stated that it agreed with the district court “that the ‘reasonable, enlightened view’ that the Oregon Supreme Court would adopt would be that discharge of pollution into water causes damage to tangible property.” Id. at 1194. Another case deciding the issue in a similar vein is United States Aviex Co. v. The Travelers Ins. Co., 125 Mich.App. 579, 336 N.W.2d 838 (1983). The court examined an exclusion like “K” in this case and rejected its application. Under Michigan law, ground water does not belong to the landholder above the water. Hence damage to the ground water is damage to a third party and is not excluded from coverage. See also Fireman’s Fund Ins. Co. v. Ex-Cell-O Corp., 662 F.Supp. 71 (E.D.Mich. 1987) (following Aviex and applying Michigan law). A similar conclusion was reached by the court in Upjohn Co. v. New Hampshire Ins. Co., slip op. No. 85-288651-CK (Mich. C.C. January 5, 1987). In Upjohn, the insured had released 15,000 gallons of distillate fluids into the soil of Puerto Rico. Ground water contamination ensued. Under Puerto Rican law, ground water is a public resource. Following Aviex, the court in Upjohn determined that contamination of ground water could not be excluded from coverage because it is not property of the insured. The court concluded that “the damage caused by the spill was damage to the property of third persons notwithstanding the fact that the spill occurred on and most of the spill was removed from Upjohn’s property.” Id. at 10. The law of California is analogous to the laws of Oregon, Michigan, and Puerto Rico in that underground water is a public resource. Contamination of that resource constitutes a palpable and tangible injury to the welfare of the entire State which for purposes of Intel’s policy can be seen as a monolithic third party. Accordingly, the Court holds that the California Supreme Court would hold that damage to underground water is damage to tangible property of a third party within the terms of a comprehensive general liability policy. Hence, Exclusion K is no bar to coverage in this case. (v) Does Cleanup of Contamination Constitute “Damages” Within the Terms of the Policy? There is an analytical step between injury to the property of a third party (California’s water) and payment of damages. The terms of the policy provide that Hartford "will pay on behalf of the insured all sums which the insured shall become legally obligated to pay as damages____” While Intel indubitably has injured the property of a third party, can cleanup, investigation, and mitigation costs be considered a form of “damages” compensable under the policy? This raises a disputed theoretical problem and brings the discussion back to the two cases cited and relied upon by the California Superior Court in Union Oil: The Maryland Casualty Co. v. Armco, Inc., 822 F.2d 1348 (4th Cir 1987), cert. denied, — U.S. -, 108 S.Ct. 703, 98 L.Ed.2d 654, and Mraz v. Canadian Universal Ins. Co., Ltd., 804 F.2d 1325 (4th Cir 1986). In Mraz, plaintiff insured sought to recover costs owed to the United States for a cleanup of hazardous waste executed pursuant to § 107 of CERCLA. 42 U.S.C. § 9607(a)(4). The court concluded that response costs were not covered within the scope of the policy’s comprehension of “property damage.” The court’s conclusion was grounded on CERCLA’s distinction between property damage and response. The court ruled that response costs are “economic loss” which may be incurred independently of property damage which the policy defined as “injury to or destruction of tangible property.” 804 F.2d at 1329. Thus, the court ruled that response costs were not “damages” within the meaning of the policy. In Maryland Casualty, the Fourth Circuit again examined the issue of the meaning of “damages” in the context of a comprehensive general liability policy. Maryland Casualty sought a declaratory judgment concerning its liability to its insured, Armco, Inc., for reimbursement costs demanded by the U.S. in connection with the cleanup of a hazardous waste site in Missouri. The court began its analysis from the premise that: “ ‘[djamages,’ as distinguished from claims for injunctive or restitutionary relief, includes ‘only payments to third persons when those persons have a legal claim for damages____’” 822 F.2d at 1352, quoting Aetna Casualty and Surety Co. v. Hanna, 224 F.2d 499, 503 (5th Cir 1955). It followed from this, according to the court, that “damages” should be “construed in consonance with its ‘accepted technical meaning in the law.’ ” 822 F.2d at 1352, quoting Hanna, 224 F.2d at 503. Utilizing this approach, the court concluded that expenditures which result from compliance “with the directives of regulatory agencies” are not damages within the technical meaning of the word. 822 F.2d at 1352. Rejecting two appellate decisions finding coverage on similar facts, the court asserted its determination not “to extend the obligations of insurance carriers beyond the well-illumined area of tangible injury and into the murky and boundless realm of injury prevention.” 822 F.2d at 1354. Mraz and Maryland Casualty offer some guidance to the question of whether there are compensable “damages” in this case. Because it believes that response costs are not “damages”, the Fourth Circuit clearly would not rule for Intel here. However, these decisions present several problems. For example, Mraz bases its analysis of the definition of property damage on a federal statute, CERCLA, rather than the laws of the state in question. Mr. Mraz was in federal court on the basis of diversity of citizenship, see 804 F.2d at 1326-27, and it is well established that federal court hearing cases based upon diversity of citizenship apply the laws of the forum state. Erie Railroad v. Tompkins, 304 U.S. 64, 78-79, 58 S.Ct. 817, 822-23, 82 L.Ed. 1188 (1938). Yet if one examines the court’s opinion, there is not a single reference to Maryland law nor an explanation of whether Maryland law is silent on the issues presented. Most important, on the question of whether response costs are distinguishable from property damage, the court relies on the distinctions inhering in CERCLA — a federal statute — rather than examining the relevant Maryland precedent. To describe response costs as “economic loss” does not answer the question whether it is a loss which is compensable as a form of “damages” under the law of the relevant state. Although the Fourth Circuit in Maryland Casualty properly does elect and attempt to apply the law of the forum state, the court’s decision is no less flawed. It is not clear why the court elected a definition of “damages” drawn from a thirty-year old case in another circuit. Apparently, the Fourth Circuit sought the narrowest available definition to constrict the scope of insurers’ liabilities. Whatever the reason, the panel stated that the Hanna decision reflects Maryland law. 822 F.2d at 1352, citing Haines v. St. Paul Fire and Marine Ins. Co., 428 F.Supp. 435 (D.Md.1977). However, an examination of Haines reveals that it, like Mraz, does not rely upon Maryland law. As noted in Maryland Casualty, the Haines Court ruled that a claim for restitution of profits made in violation of the federal securities laws was an action for “traditional equitable relief and cannot be considered damages within the policy coverage.” 428 F.Supp. at 441. The logic of Haines derived from the Supreme Court’s decision in Curtis v. Loether, 415 U.S. 189, 94 S.Ct. 1005, 39 L.Ed.2d 260 (1974), which dealt with the right to jury trials in Title VIII actions. The Court distinguished Title VIII actions from those arising under Title VII and noted that the Title VII provision for “back pay” awards is committed to the discretion of the trial judge and is “equitable” in nature. 415 U.S. at 196-97, 94 S.Ct. at 1009-10. Similarly, the court in Haines concluded that restitutionary character of the SEC’s prosecution was equitable in nature and that the disgorgement of illegal profits is not “damages” in the traditional sense. With all due respect to the court in Haines, this Court concludes that the proper source to determine the definition of damages in a diversity case is the law of the forum state. Haines fails to respect this fact, as does the Fourth Circuit in Mraz. Although the Fourth Circuit in Maryland Casualty does acknowledge the need to consult state (Maryland) law to decide the issue, the court failed to do so except for a citation to Haines. Before examining California law, a brief survey of other states’ treatment of the issue is instructive. For example, in a case cited earlier, a Michigan appellate court determined that “damages” under Michigan law includes monies spent to reimburse the government for costs incurred in investigating and cleaning up chemical contamination of underground water. United States Aviex Company v. Travelers Insurance Company, 125 Mich.App. 579, 336 N.W.2d 838 (1983). The insurance policy provisions in Aviex were virtually identical to those in this case. The court squarely addressed the issue of damages and rejected the argument that investigation and cleanup costs were not “property damages” within the terms of the policy. Noting three decision from other jurisdictions finding no coverage on similar facts, the court in Aviex rejected their reasoning as interpreting “damages” “too narrowly”. The court observed that under Michigan law, the state’s interest in natural resources would allow the State Attorney General to sue the insured for contamination of subterranean water. If such a suit were prosecuted, the insurer undeniably would be obligated to cover any damages awarded — including cleanup costs. The court then stated: “It is merely fortuitous from the standpoint of either plaintiff or defendant that the state has chosen to have plaintiff remedy the contamination problem, rather than choosing to incur the costs of cleanup itself and then suing plaintiff to recover those costs. The damage to the natural resources is simply measured in the cost to restore the water to its original state.” [citations omitted] [no page-Lexis] Aviex is an important case on the topic of damage because it focuses the issue. A recent en banc opinion of the United States Court of Appeals for the Eighth Circuit further crystalizes the issue. In Continental Insurance Co., v. Northeastern Pharmaceutical & Chemical Co., 842 F.2d 977 (1988), the Eighth Circuit en banc addressed the issue whether environmental cleanup costs are “damages” within the scope of a comprehensive general liability (“CGL”) policy. By a 5-8 vote, the court held that cleanup costs are not “damages” within the meaning of CGL policies. This decision reversed a prior panel decision published at 811 F.2d 1180 (8th Cir.1987). Several valuable points can be distilled from the sharply divided opinions of the court. First, both sides agreed that as a diversity case in federal court, the law of the forum state applied, i.e., the law of Missouri. Second, both sides recognize that the “[c]ase law on this issue is sharply divided.” However, the division of cases on this subject is “sharp” only in respect to the contrast between the two positions. The clear bulk of authority has held that cleanup costs are damages within the scope of CGL policy coverage. Yet, as implicitly noted by the dissent, the fact that the law of the state applies is dispositive; each state’s own law may provide its own answer to this question. (vi) Assessment of Relevant California Law: Although no California appellate court has decided the issue, there is California law to answer the question of whether cleanup and investigation costs are “damages” within the meaning of a CGL policy. Indeed, there is a statute speaking to the issue, California Civil Code § 3281, which provides: Every person who suffers detriment from the unlawful act or omission of another, may recover from the person in fault a compensation therefor in money, which is called damages. The Court notes that § 3281 speaks to “damages” which technically is not identical to “damage.” “Damage” denotes “injury to the person or property of the plaintiff as the result of a wrongful act or omission on the part of the defendant,” 23 Cal.Jur.3d § 2 at 12 (1975), whereas “damages” is the compensation awarded to one who suffers “damage.” Section 3281 essentially incorporates the former to arrive at a definition of the latter so that the two are inextricably intertwined. This is consistent with time-tested California case law. See Wainscott v. Occidental Bldg, and Loan Association, 98 Cal. 253, 255, 33 P. 88 (1893). Applying California law, it is clear that damage has occurred for which compensation in the form of damages would be appropriate. The State and People of California have suffered a detriment from the unlawful acts or omissions of Intel and are entitled to damages as a result. The detriment, which is defined as a “loss or harm suffered in person or property,” California Civil Code § 3282, is not yet fully determined, but at this point there is established detriment to California’s ground water through contamination, and a serious health threat to the people of the state. As noted earlier, the contamination of California’s environment, in particular the water, is illegal. See, e.g., California Water Code § 13387. In entering into the Consent Decree along with Intel and EPA, California’s Regional Water Quality Control Board and Department of Health Services already have demanded and received the payment of $50,000 in money compensation from Intel and its co-polluters. Further compensation will be paid in kind, for instead of EPA and the State of California performing the investigation and cleanup and demanding payment of “compensation therefor in money,” California Civil Code § 3281, Intel and its peers will conduct the work under the conditions imposed by the government, which include strict government supervision. Thus, under California law, damages have been and are being paid by Intel. This analysis is consistent with the requirement in California that insurance contracts be read “in the light of the reasonable and normal expectations of the parties as to the extent of the coverage.” Globe Indemnity Co. v. State of California, 43 Cal.App.3d 745, 751, 118 Cal.Rptr. 75 (1974); Hanson v. Prudential Ins. Co. of America, 783 F.2d 762, 765 (9th Cir.1985), citing Reserve Insurance, 30 Cal.3d at 809, 180 Cal.Rptr. 628, 640 P.2d 164, and McLaughlin, 565 F.Supp. at 441 (N.D.Cal.1983). At the time this contract was negotiated, neither party anticipated a claim of this sort. What the parties did contemplate, as expressed b