South MacOmb Disposal Authority v. American Insurance

572 N.W.2d 686, 225 Mich. App. 635

• Court: Michigan Court of Appeals
• Decided: January 8, 1998
• Docket: Docket Nos. 186277-186280 and 189612
• Status: Published

Opinions

Corrigan, PJ.

These consolidated cases are before us on remand from our Supreme Court for consideration as on leave granted. South Macomb Disposal Authority v American Ins Co, 448 Mich 947 (1995); South Macomb Disposal Authority v Westchester Fire Ins Co, 448 Mich 947 (1995); South Macomb Disposal Authority v Westchester Fire Ins Co, 450 Mich 873 (1995). Plaintiff, South Macomb Disposal Authority (SMDA), which operated several landfills, instituted these declaratory judgment actions to compel defendant insurers to defend claims brought against it. The claims arose from the leaking of some of plaintiffs landfills. The trial court concluded that certain of the defendants owed plaintiff a duty to defend with respect to the underlying claims based on personal injury, but denied in part defendants’ motions for summary disposition insofar as they were based on the absence of an “occurrence” within the meaning of the policies or the applicability of the pollution exclusion clauses contained in insurance policies issued to plaintiff because of genuine questions of material fact concerning those issues. Together, these cases comprise seventy-three lower court files and many additional pleadings, exhibits, deposition transcripts, and videotapes. We affirm in part, reverse in part, and remand.

Among the other issues involved, these cases raise three issues of first impression: (1) whether an identifiable discharge properly may be separated from a larger pattern of discharge or leakage; (2) whether a personal injury liability endorsement provides coverage for the pollution claims of private plaintiffs where those claims otherwise are barred by a pollution exclusion clause in a comprehensive general liability policy; and (3) whether letters from the Michigan Department of Natural Resources (dnr) constitute a “suit” triggering insurance coverage.

To answer those questions, this opinion is structured as follows.

First, we set forth the underlying facts and precedural history. Next, we consider whether defendant Westchester Fire Insurance Company should have been granted summary disposition on the basis that there was no “occurrence” within the meaning of its policy and conclude that summary disposition should not have been granted because Westchester did not demonstrate sufficiently, under a subjective standard, that there was no occurrence. Additionally, we hold that certain letters from the DNR constituted a “suit,” which triggered coverage under the defendants’ policies.

We next examine the pollution exclusions and hold that the proper focus for determining whether a discharge was “sudden and accidental” is the discharge of pollutants from the landfill into the surrounding soils, groundwaters, and surface waters. We then discuss and reject the contention that plaintiff presented a genuine issue of material fact regarding other sources of contamination and hold that the distinct, identifiable discharges of leachate from the sides of one of the sites at issue may be considered separately from the overall leaking at the site.

Next, we reject plaintiffs argument that the personal injury liability endorsements afford plaintiff coverage outside the pollution exclusion. We thereafter address costs, holding that defense costs may include the costs of designing a plan for remediation of pollution. Also, we note that because questions of fact arose under the loss-in-progress doctrine, the circuit court properly denied summary disposition on this point. Finally, we conclude that plaintiffs waiver and estoppel argument against defendant Westchester should fail because Westchester did not misrepresent the policy terms to plaintiff.

I. UNDERLYING FACTS

These suits involve four municipal landfill sites, which were developed and operated at various times from the late 1960s to the mid-1980s. Sites 9 and 9A are located in Macomb Township. Sites 7 and 11 are located in Washington Township. During their operation and after their closure, leachate¹ leaked from these landfills, polluting the surrounding soil and groundwater.

Plaintiffs declaratory judgment action in lower court number 84-002686CZ resulted from a lawsuit against plaintiff, captioned: Bielat v SMDA, Macomb Circuit Court No. 84-612AA. In Bielat, residential property owners in the immediate vicinity of landfill sites 9 and 9A sued plaintiff, alleging personal injuries and property damage on theories of nuisance, trespass, and negligence. The dnr and the Michigan Department of Public Health (dph), defendants in the Bielat action, cross-complained against plaintiff. Plaintiff filed the second declaratory judgment action, South Macomb Disposal Authority v American Ins Co, lower court number 90-001995CK, against the same insurer-defendants in response to letters plaintiff received from the dnr regarding contamination of groundwater at sites 7 and 11, which the dnr identified as having been caused by plaintiffs landfills.

A. SITES 9 AND 9A

The circuit court set forth in great detail the complex history of sites 9 and 9A; we include only a summary of pertinent facts. The site 9 property, acquired in 1967, had been operated previously as a sand and gravel mine. Dr. Andrew Mozola, a geologist, investigated the suitability of the site for use as a landfill. Plaintiff hired Warren Anderson, a civil engineering consultant, to design and maintain the landfill.

Soil boring tests indicated that the water table at the site ranged in depth from eleven feet to the surface. Mozola told plaintiffs director in 1967 that he was not enthusiastic about the site’s use as a landfill. Mozola determined that the water table was high and that a leachate problem could develop if plaintiff used the property as a landfill. Mozola outlined several design considerations should the property be used as plaintiff proposed. State law required plaintiff to keep landfill operations at least two feet above the water table and to place six inches of compact suitable cover material over all exposed refuse at the end of each day. Plaintiff knew that, because of the previous sand mining, the site consisted of nearly thirty acres of low land with water pockets.

In April 1968, the dph issued a license to plaintiff to operate the site as a landfill. The license stipulated that drainage effectiveness was to be reviewed after construction and that modifications to operations might be necessary, depending on watertable control. Within six months, plaintiff received unsatisfactory comments about its operation of the site. A Macomb County Health Department (MCHD) evaluation in September 1968 noted that the site was not approved because it was a burning, open dump. The MCHD ordered plaintiff to cover the material and stop dumping in the water. Groundwater problems occurred at site 9, which plaintiff worked to overcome. The site was licensed until it was closed in 1975.

Plaintiff sought to expand its landfill operations by including site 9A in 1971. Again, the dnr and the dph specifically informed plaintiff about the foreseeable watertable problems at the site. Although the dph licensed site 9A in February 1971, the license included stipulations regarding the replacement of sand, underdrain modifications, the construction of a perimeter clay dike and the placement of fill material not less than four feet above the subsurface drain. Site 9A was closed in 1979.

While site 9A was operating, specific complaints regarding leachate outbreaks arose. The first occurred in 1971, shortly after the site opened, and the second in June 1976. A third outbreak occurred in 1980. These outbreaks will be discussed later. Throughout the time when site 9A was operating, numerous problems occurred, including water in the dumping trenches, lack of cover material, inadequate drainage, and severe leaching of leachate. The under-drain system at the site also had problems. Residents repeatedly complained, alleging that leachate had leaked into a basement, leachate had flowed and bubbled up onto property, cattle and pets had died and suffered deformities, and human health problems had occurred. The McBride Drain, which ran near the site, had shown evidence of degradation. On several occasions, leachate from site 9A discharged into the McBride Drain.

B. SITES 7 AND 11

Site 7 is located on twenty acres formerly owned by the Macomb County Road Commission. In the 1950s and the 1960s, the road commission operated an asphalt plant on the property and excavated gravel from portions of it. During those years, residents of Washington Township were permitted to dump household refuse in the areas from which gravel had been excavated. The dump area was not lined or otherwise prepared for landfilling by the road commission. Instead, the refuse apparently was dumped directly on the land, sometimes into water. The site was essentially an open dump. When the refuse material reached a height of approximately six to ten feet, it was covered with soil. Sometime in the 1960s, a substance that had been dumped in the area leached into a nearby lake, killing the fish that had been stocked in it.

In 1967, plaintiff purchased the twenty acres comprising site 7 from the road commission, intending to obtain a license and operate the site as a sanitary landfill. The site, licensed in 1968, operated as a sanitary landfill until 1974. When plaintiff began its operations at site 7, it covered the existing refuse and continued dumping on top of it. Although residential dumping continued, plaintiff eventually limited residents’ access to the site. Plaintiff filled the bottom of the fill area to raise it two feet above the elevation of nearby lakes. Evidence shows, however, that groundwater fluctuations could raise the water table to within two feet of the bottom elevation of the site.

In 1974, plaintiff leased another twenty acres from a private company that had operated the Predmore Sanitary Landfill on the property. Plaintiff developed this newly acquired parcel and the twenty acres acquired from the road commission as phase I of site 11. During the development of phase I of site 11, a water sample taken from the north face of the Predmore Landfill by the mchd indicated that leachate was being generated. Eventually, plaintiff leased another parcel of property, the development of which became phase n of site 11. Plaintiff deposited clay on the bottoms of the areas that had not been used previously as a dump or landfill. These sites were licensed and operated until approximately 1986.

While the sites operated, regular evaluation inspections occurred. The dnr and mchd repeatedly cited plaintiff for using inadequate daily cover on the refuse material and failing to properly control erosion and to seed completed areas. Several leachate outbreak problems occurred over the years, and leachate leaked into a nearby pond or lake. The refuse was often exposed. At one point, the mchd cited plaintiff for improperly using incinerator ash, a waste material, as cover. Anderson was aware that the sand and gravel mixture used as daily cover was not impervious to water and that other more impervious substances were available, but plaintiff initially did not use them. In later years, clay was brought in as cover material.

In April 1990, the dnr informed plaintiff that groundwater sampling from downgradient monitor wells reflected contamination with volatile organic compounds (vocs), as well as inorganic and organic chemicals, and ordered plaintiff to respond through a remedial investigation and feasibility study. The dnr furnished written notice of contamination again in January 1992 and in January 1993. Defendant insurers Citizens Insurance Company, American Insurance Company, and National Surety Corporation thereafter refused plaintiff’s request to defend it regarding the remedial action oirdered by the dnr.

n. THE INSURANCE POLICIES

Defendant insurers .provided plaintiff with comprehensive general liability policies. Defendants American Insurance Company and National Surety Corporation are insurers through Fireman’s Fund Insurance Company and collectively are referred to as FFIC. Citizens issued policies from April 1971 through August 1975, and FFic’s policies covered September 1975 through April 1985.² These Citizens policies, as well as the FFIC policies covering September 1975 through February 1984, contained a pollution exclusion with a “sudden and accidental” exception. From February 1984 through April 1985, the FFIC policy contained an “absolute pollution” exclusion. Defendant Westchester Fire Insurance Company provided umbrella policies from September 1975 through January 1984. The September 1975 through August 1976 and December 1983 through January 1984 policies contained the sudden and accidental pollution exclusion. The remaining policies contained the absolute pollution exclusion. The Citizens and FFIC policies also contained a personal injury liability provision.

HI. THE PROCEDURAL HISTORY

The trial court initially granted in part defendants’ motions for summary disposition regarding sites 9 and 9A. Defendants sought summary disposition in part because no “occurrence” triggered coverage and because the policies’ pollution exclusion provisions barred plaintiff’s claims. Because the court found a factual question regarding whether an “occurrence” triggered coverage, it denied defendants’ motions. It also denied defendants’ motions regarding the pollution exclusion clauses, except those claims based on the long-term leaking, because that long-term leaking was not sudden and accidental. The court also partially granted plaintiffs motion for summary disposition with respect to claims that fell within the personal injury liability coverage, finding that Citizens and ffic had a duty to defend plaintiff with respect to those claims under the personal injury provisions of the policies issued by those defendants. The court denied defendants’ motions for rehearing.

In 1994, this Court denied defendants’ applications for leave to appeal. Defendants applied for leave to appeal to our Supreme Court, which vacated the circuit court’s opinions and remanded the matter to that court for reconsideration in light of Auto-Owners Ins Co v City of Clare, 446 Mich 1; 521 NW2d 480 (1994). SMDA v National Surety Corp No 1, 447 Mich 951 (1994); SMDA v National Surety Corp No 2, 447 Mich 951 (1994); SMDA v Westchester Fire Ins Co, 447 Mich 951 (1994). The circuit court again granted in part defendants’ motions for summary disposition, on the same bases, but also concluded that coverage was precluded under the FFIC policy that contained the absolute pollution exclusion. The court granted in part plaintiff’s motion on personal injury liability coverage, finding that Citizens and ffic had a duty to defend under the personal injury liability provisions. Finally, the circuit court found that Citizens’ and ffic’s duties to defend included payment of the costs of a hydrogeological study and reasonable attorney fees in defense of the underlying Bielat action.

This Court again denied defendants’ applications for leave to appeal in 1995. Our Supreme Court, in response to defendants’ applications for leave to appeal, remanded the matter as on leave granted to consider the issues defendants had raised.

Regarding sites 7 and 11, an April 1990 letter from the DNR informed plaintiff that groundwater testing at site 11 had revealed that the groundwater contained vocs, which indicated the landfill was leaking. After Citizens and FFIC initially refused to defend, plaintiff filed a declaratory judgment action in 1990 to compel defendants’ defense. The circuit court originally granted partial summary disposition for plaintiff, finding that defendants had a duty to defend. The court denied Citizens’ and ffic’s motions for reconsideration. After an unsuccessful application for leave to appeal to this Court in 1994, defendants sought leave to appeal to our Supreme Court, which, in lieu of granting leave to appeal, vacated the prior opinions and remanded the matter to the circuit court for consideration in light of City of Clare, supra, and Michigan Millers Mut Ins Co v Bronson Plating Co, 445 Mich 558; 519 NW2d 864 (1994). SMDA v American Ins Co No 1, 447 Mich 952 (1994); SMDA v American Ins Co No 2, 447 Mich 952 (1994).

The circuit court again granted plaintiff’s motion for partial summary disposition. First, it found, contrary to defendants’ argument, that the letters sent to plaintiff from the dnr constituted the functional equivalent of a suit and triggered coverage. The circuit court also found that, because questions existed regarding the source of the contamination of the groundwater and plaintiff’s expectation of the contamination, defendants had a duty to defend plaintiff. The court also granted plaintiff’s motion for partial summary disposition regarding coverage for its claims under the personal injury liability endorsements. After this Court again denied defendants’ application for leave to appeal in 1995, our Supreme Court remanded the matter to this Court for consideration as on leave granted. Our Court has consolidated these cases for appellate review.

IV. THE STANDARD OF REVIEW

In ruling on a motion brought under MCR 2.116(C)(10), a court must consider the pleadings, affidavits, depositions, admissions, and documentary evidence filed or submitted by the parties. MCR 2.116(G)(5). The court should grant summary disposition when, “[e]xcept as to the amount of damages, there is no genuine issue as to any material fact, and the moving party is entitled to judgment or partial judgment as a matter of law.” MCR 2.116(C)(10). The court should give the nonmoving party the benefit of all reasonable doubt and consider the evidence in its favor. It must then determine whether a record could be developed upon which reasonable minds could differ. McGuirk Sand & Gravel, Inc v Meridian Mut Ins Co, 220 Mich App 347, 352; 559 NW2d 93 (1996); Kent Co v Home Ins Co, 217 Mich App 250, 261; 551 NW2d 424 (1996), vacated in part and remanded to the circuit court in part 456 Mich 858 (1997).³ The circuit court’s decisions concerning motions for summary disposition are reviewed de novo. Kent Co, supra at 261.

We apply general rules of construction in inteipreting insurance policies. Interpretation of a contract with clear language is a question of law, which is reviewed de novo. Auto Club Ins Ass’n v Lozanis, 215 Mich App 415, 419; 546 NW2d 648 (1996). A court determines whether the policy is clear and unambiguous on its face. Upjohn Co v New Hampshire Ins Co, 438 Mich 197, 206; 476 NW2d 392 (1991). Courts may not create ambiguities where none exist, but must construe ambiguous policy language in the insured’s favor. Fire Ins Exchange v Diehl, 450 Mich 678, 687; 545 NW2d 602 (1996). Clear and unambiguous language may not be rewritten under the guise of interpretation; contract terms must be enforced as written, and unambiguous terms must be construed according to their plain and commonly understood meaning. Upjohn, supra at 207; Lozanis, supra. Additionally, an insurance contract should be viewed as a whole and read to give meaning to all its terms. Fresard v Michigan Millers Mut Ins Co, 414 Mich 686, 694; 327 NW2d 286 (1982). Conflicts between clauses should be harmonized, and a contract should not be interpreted so as to render it unreasonable. Id.

Exclusionary clauses are strictly construed in the insured’s favor. Auto-Owners Ins Co v Churchman, 440 Mich 560, 567; 489 NW2d 431 (1992). A clear and specific exclusion must be given effect; an insurance company may not be held liable for unassumed risks. Id. If any exclusion in an insurance policy applies to a claimant’s particular claims, coverage is lost. Id.

V. WHETHER AN “OCCURRENCE” TRIGGERED COVERAGE

Defendant insurer Westchester argues that the trial court erred in failing to grant summary disposition on the basis that no “occurrence” triggered coverage. We disagree.

The Westchester policies provided coverage for losses “arising out of an occurrence.” To decide whether an insurance policy covers a specific act, a court must decide the policy terms upon which the parties agreed. Diehl, supra at 683. To do so, the court first determines if the occurrence section of the policy includes the action at issue. Id. In this analysis, the court examines whether the policy language, specifically the definition of an “occurrence,” excludes coverage for injuries that were either (1) objectively intended or (2) subjectively intended by the insured. Id. at 684. In other words, a court must decide whether the policy language requires use of an objective standard or of a subjective standard. See id. at 685-686.

Westchester argues that an objective standard should be used to determine whether plaintiff expected or intended the damage. Under an objective standard, Westchester asserts, the evidence demonstrates that no “occurrence” transpired to trigger coverage.

Here, Westchester failed to include in its policy language that signals which standard applies to its policy. The Westchester policy defines an “occurrence” as “either an accident or happening or event or a continuous or repeated exposure to conditions which unexpectedly and unintentionally causes injury to persons or tangible property during the policy period.” This definition contains neither the word “reasonably” (an objective term) nor the phrase “from the standpoint of the insured” (subjective terminology). The policy is susceptible to more than one meaning and therefore is ambiguous. Vanguard Ins Co v Racine, 224 Mich App 229, 233; 568 NW2d 156 (1997).

Recently, this Court ruled that the subjective test applies where an insurance policy uses the term “accident” but is otherwise silent with respect to from whose perspective the event is to be deemed an accident. Frankenmuth Mut Ins Co v Masters, 225 Mich App 51, 57-59; 570 NW2d 134 (1997). As indicated above, however, the Westchester policy did not limit its definition of occurrence to include only accidents; under the policy, an occurrence may be something other than an accident.

Moreover, the Westchester policy does not define the term “accident.” Accordingly, we rely on the common meaning of the word “accident.” Arco Industries Corp v American Motorists Ins Co, 448 Mich 395, 404; 531 NW2d 168 (1995).

The Court in Masters, supra at 57, relied on the expanded definition of accident contained in Group Ins Co of Michigan v Czopek, 440 Mich 590, 597; 489 NW2d 444 (1992), quoting Guerden Industries, Inc v Fidelity & Casualty Co of New York, 371 Mich 12, 18-19; 123 NW2d 143 (1963):

“ ‘[Ajnything that begins to be, that happens, or that is a result which is not anticipated and is unforeseen and unexpected by the person injured or affected thereby — that is, takes place without the insured’s foresight or expectation and without design or intentional causation on his part. In other words, an accident is an undesigned contingency, a casualty, a happening by chance, something out of the usual course of things, unusual, fortuitous, not anticipated, and not naturally to be expected.’ ”

Because that definition references both the insured’s and the injured person’s perspectives, it is ambiguous concerning the applicable standard.

Under Masters, we should apply the subjective standard because the Westchester policy is silent regarding from whose perspective the question of intent should be viewed, but does speak in terms of an accident. The analysis does not stop there, however. Because the Westchester policy did not define occurrences only as being accidents and because the definition of “accident” is ambiguous concerning the applicable standard, we include additional analysis regarding the objective and subjective standards.

In Allstate Ins Co v Freeman, 432 Mich 656, 685; 443 NW2d 734 (1989) (Riley, J.), the policy excluded “bodily injury or property damage which may reasonably be expected to result from the intentional or criminal acts of an insured person or which is in fact intended by an insured person.” (Emphasis added.) The Court ruled that the use of “reasonably be expected” in the provision unambiguously directed use of the objective standard of expectation. Id. at 688.

Conversely, in Metropolitan Property & Liability Ins Co v DiCicco, which was submitted and considered with Freeman, supra, the policy excluded “bodily injury or property damage which is either expected or intended from the standpoint of the insured.” Id. at 672 (lower case characters substituted, emphasis added). Although the dissenting justices opined that the use of the word “expected” in the exclusion denotes that the actor “knew or should have known that there was a substantial probability that certain consequences will result from his actions,” thus connoting an objective standard, id. at 675 (Riley, J.), the majority found the policy language unambiguously directed application of the subjective standard. Id. at 708-710 (Boyle, J.).

Myriad cases decided after Freeman/DiCicco have addressed this distinction between objective and subjective language. For example, in Auto-Owners Ins Co v Harrington, 212 Mich App 682, 685-686; 538 NW2d 106 (1995), aff'd 455 Mich 377; 565 NW2d 839 (1997), this Court applied a subjective standard because the policy language provided that damage was not covered when “expected or intended by an insured person.” (Emphasis added.) In contrast, this Court used an objective standard when a policy stated that coverage was not provided for damage “which may reasonably be expected to result from the intentional . . . acts of an insured person or which is in fact intended by an insured person.” Allstate Ins Co v Keillor (On Remand), 203 Mich App 36, 39-40; 511 NW2d 702 (1993), modified on other grounds 450 Mich 412; 537 NW2d 589 (1995) (emphasis added).

General rules regarding the interpretation of ambiguous insurance policy provisions provide guidance. Where policy language is clear, this Court is bound by the specific language set forth in the policy. Heniser v Frankenmuth Mut Ins Co, 449 Mich 155, 160; 534 NW2d 502 (1995); Racine, supra. If a policy does not define a term, however, this Court interprets the terms of the policy in accordance with their commonly used meanings. Arco Industries Corp v American Motorists Ins Co, 448 Mich 395, 404; 531 NW2d 168 (1995); Racine, supra. Ambiguity exists when a policy provision is reasonably susceptible to more than one meaning. Racine, supra. This Court construes ambiguous insurance contract provisions in favor of the insured. Group Ins Co of Michigan v Czopek, supra at 595; Michigan Basic Property Ins Ass’n v Wasarovich, 214 Mich App 319, 321; 542 NW2d 367 (1995).

Similarly, when this Court interprets exclusionary provisions, we strictly construe the policy language against the insurer. Czopek, supra at 597; Wasarovich, supra at 323. If the exclusion language is clear and unambiguous, this Court will apply it as written. Auto Club Group Ins Co v Marzonie, 447 Mich 624, 631; 527 NW2d 760 (1994); Wasarovich, supra at 323.

While no Michigan court has addressed the exact language that appears in this policy, other states have resolved policies with similar language in favor of the insured. In a pollution coverage case, the Supreme Court of Washington addressed an insurance policy defining an occurrence as “an accident or happening or event or a continuous or repeated exposure to conditions which unexpectedly and unintentionally results in personal injury, property damage . . . .” Queen City Farms, Inc v Central Nat’l Ins Co of Omaha, 126 Wash 2d 50, 64; 891 P2d 718 (1995), replacing 124 Wash 2d 536; 882 P2d 703 (1994). The court noted that the policy failed to define whose state of mind the policy intended to use to define an occurrence. Id. at 66. That court concluded that the policy language is “at least ambiguous as to whether an objective or a subjective standard applies.” Id. at 68. As in Michigan, that court construed the ambiguity against the insurer and adopted the subjective standard. Id.

Because the Westchester policy is susceptible to more than one meaning, it is ambiguous. Applying the above principles, we construe the policy in favor of the insured. In doing so, we adopt the subjective, insured-based perspective for the analysis of the occurrence. The absence of objective terminology such as “reasonable” and “should have known” in the Westchester policy also supports the application of a subjective standard.

Under the policy-blended subjective standard, the court examines the insured’s conduct from the insured’s perspective and evaluates either the insured’s intent to cause injury or the insured’s awareness that harm was likely to follow from the performance of the act. Frankenmuth Mut Ins Co v Piccard, 440 Mich 539, 549-550; 489 NW2d 422 (1992). Our Supreme Court applied this two-pronged test to the environmental issue in Arco, supra at 410, and asked (1) whether the conduct, from the perspective of Arco, the insured, evidenced an intent to cause the contamination to the environment and (2) whether Arco had the awareness that harm was likely to follow from the performance of its actions.

Thus, we ask whether the conduct of -plaintiff, as the insured, evidenced an intent to cause the contamination and whether plaintiff was aware that harm was likely to follow from its actions. Westchester has failed to demonstrate as a matter of fact, as required under MCR 2.116(C)(10), that plaintiff actually knew that the groundwater and soil contamination would occur. Westchester did not show that plaintiff evidenced an intent to cause the contamination or that plaintiff knew that harm was likely to follow from the performance of its actions. Accordingly, because no facts demonstrate that the contamination was expected or intended by plaintiff, the contamination meets the definition of “occurrence” in the Westchester policy, and coverage thus was triggered. The trial court properly denied Westchester’s motion for summary disposition and those of the remaining defendants on this issue.

VI. WHETHER THE DNR LETTERS CONSTITUTED A “SUIT”

Defendants argue that the circuit court erred in concluding that the DNR letters constituted a “suit” that gave rise to coverage under the policies. We disagree.

Although our Supreme Court has addressed this issue in Bronson Plating, supra, in the context of demand letters from the federal Environmental Protection Agency (epa), the Court has not addressed it in the context' of similar letters from the DNR to alleged polluters. See Arco Industries, supra at 417, n 15.⁴ Whether demand letters from the DNR give rise to coverage in the same manner as demand letters from the epa raises a question of law, which we review de novo. Cardinal Mooney High School v Michigan High School Athletic Ass’n, 437 Mich 75, 80; 467 NW2d 21 (1991).

The first dnr letter in April 1990 stated that annual groundwater sampling at site 11 demonstrated that the groundwater resources at the site had been affected. Water samples taken in August 1989 and January 1990, from monitor wells at the facility, revealed elevated levels of chlorinated halocarbons and hydrocarbons, vocs, and inorganic and organic chemicals. The dnr letter stated: “The presence of volatile organic compounds and the elevated levels of the inorganic parameters downgradient of the site indicates [sic] that SMDA # 11 is impacting the groundwater and contaminating a usable aquifer.” The dnr letter informed plaintiff that the data indicated a leak from the landfill. The dnr advised plaintiff that it was violating the water resources commission act, MCL 323.1 et seq.\ MSA 3.521 et seq., and its license under 40 CFR 257.3-4(a) to 257.3-4(c).

The letter requested that plaintiff submit a work plan detailing a hydrogeological investigation of the site to define the extent of the groundwater contamination and provide an explanation of the direct and immediate measures that plaintiff would implement to prevent further degradation of the area’s potable aquifer. Additionally, following the remedial investigation, plaintiff would be required to conduct a feasibility study to investigate possible remediation techniques.

In ■ January 1992, the dnr notified plaintiff about widespread groundwater contamination and instructed it to address the situation. Finally, in January 1993, the DNR sent another letter, reporting that annual groundwater samples taken from 1989 through 1992 showed the landfill contaminated the groundwater resources at the site. The dnr informed plaintiff that such discharges violated the water resources commission act. The letter stated:

Therefore, by submittal of this letter, the Department is requesting that smda submit a work plan detailing a hydrogeological investigation of the smda #11 Landfill. The hydrogeological investigation must define the lateral and vertical extent of groundwater contamination and assess the hydrogeological characteristics of the site. The work plan should also include direct and immediate measures that smda will implement to prevent further degradation of the groundwater. After the hydrogeological investigation, a feasibility study will be required to investigate possible remediation techniques that could be used at the site.

... Please be advised that violation of Act 245 is punishable by fines of no less than $2500.00 nor more than $25,000.00 for each violation that has occurred. Additionally, a fine of $25,000.00 for each day during which the unlawful discharge occurred may also be imposed. Any such unauthorized discharges may be considered a violation of other environmental regulations, which are not in the jurisdiction of this Division. Please note that other mdnr Divisions may pursue additional enforcement actions, civil and/or criminal, and/or penalties under their own delegated authority.

The DNR warned plaintiff that its failure to submit the requested work plan by a specific date might “result in escalated enforcement action taken against smda.”

The issue is whether these letters constituted a suit, triggering defendants’ duty to defend. In Bronson Plating, our Supreme Court examined the issue “whether the epa letter notifying Bronson [the insured] of its potential liability for the alleged environmental contamination constitutes a ‘suit’ that gives rise to the insurers’ duty to defend under the terms of the applicable insurance contracts.” Bronson Plating, supra at 566. In that case, Bronson, an electroplating company, used a variety of chemicals and compounds in its processes. The epa and the dnr identified waste-water that Bronson had released as a possible source of environmental contamination. Id. at 562-563. The epa, under the Comprehensive Environmental Response, Compensation and Liability Act (cercla), 42 USC 9601 et seq., notified Bronson of its potential liability for contamination of the site and advised it that the epa was planning to spend public money to control and investigate the releases of contaminants. The epa requested that Bronson participate voluntarily in certain studies, cautioning Bronson that its refusal could lead to joint and several liability for funds expended in carrying out the request. It also demanded that Bronson supply relevant information, encouraged good-faith negotiations between Bronson and the epa with other potentially responsible parties, and explained that a failure to comply could result in a civil enforcement action by the epa against Bronson. Bronson Plating, supra at 564. The insurers contended that this EPA letter did not rise to the level of a “suit,” which would trigger their duty to defend.

Our Supreme Court concluded that the term “suit” as used in the insurance policies was ambiguous. Id. at 571. It decided, however, that in considering the general definition of the term as well as alternative definitions, “a typical layperson might reasonably expect the term to apply to legal proceedings other than a court action initiated by a complaint.” Id. at 568-569.

Next, the Court considered whether the epa’s letter constituted the functional equivalent of a suit. It ultimately concluded that in Bronson Plating, the letter from the epa triggered the insurers’ duty to defend. To determine whether the letter initiated a “suit,” the Court examined the letter’s contents, which informed Bronson of its status as a potentially responsible party (prp). It also stated that if none of the prps voluntarily undertook to conduct the remedial investigation and feasibility study (Rl/FS), the EPA would do so, that if the epa conducted the Rl/FS, Bronson could be held jointly and severally liable for all costs, and that in the event Bronson did not answer the epa’s request for documentation of waste released onto the site, a civil action and fines could result. Id. at 572-573. The Court concluded: “Taking into account the various components of this PRP letter and its ramifications, we find that the legal proceeding initiated by the réceipt of that notice is the functional equivalent of a suit brought in a court of law.” Id. at 573. The Court considered several characteristics of cercla actions:

Of critical importance is the creation of the administrative record and the role it may play in future litigation. Documentation sought by the epa, and which Bronson must produce under the force of law, will determine the amount and type of waste generated by Bronson and discharged onto the site. Given the strict liability stance of cercla, this information is all that is needed to establish both the fact and proportional share of Bronson’s liability at the site.

Moreover, because the epa may implement any investigatory and remedial action it deems necessary at the site, subject only to an abuse of discretion review, the total cost of the project will also be determined before litigation is brought. The significant authority given to the epa in such matters allows it essentially to usurp the traditional role of a court of law in determining and apportioning liability. Such matters are concluded by the epa before the action is ever brought to court.

The epa’s powers may also be viewed as coercing the “voluntary” participation of prps. The entire CERCLA scheme revolves around “encouraging” prps to engage in voluntary cleanups. Only in so doing may a PRP have a voice in developing the record that will be used against it and in determining the amount of its liability through selection of investigatory and remedial methods and procedures. The significance of these incentives is underscored by the fact that EPA-conducted cercla actions have historically been considerably and, some would suggest, needlessly more expensive than those actions conducted by PRP groups. [Id. at 573-575.]

Defendants highlight our Supreme Court’s language that not every request for relief should be considered as the initiation of a suit, triggering an insurer’s duty to defend. Id. at 573, n 13. The Bronson Plating Court stated that its decision that the epa letter constituted a “suit” was “made primarily based on the unique aspects of cercla actions and the authority given to epa under the statute.” Id. It noted that the PRP notice, unlike general demand letters, carried immediate and severe implications. In cercla actions, the. substantive rights and ultimate liability of a prp are affected from the beginning of the administrative process. Id.

This Court previously has ruled that the intent of the Environmental Response Act, MCL 299.612(1); MSA 13.32(12)(1), is similar to that of its federal counterpart, the cercla. Accordingly, it is appropriate to examine federal law in interpreting similar issues. Farm Bureau Mut Ins Co v Porter & Heckman, Inc, 220 Mich App 627, 637; 560 NW2d 367 (1996); Haworth, Inc v Wickes Mfg Co, 210 Mich App 222, 228; 532 NW2d 903 (1995). By analogy, it is appropriate to look to the epa’s'powers in determining those of the DNR. Consequently, we may employ the Bronson Plating analysis of the epa to the DNR.

Under Bronson Plating, we consider the nature of the dnr’s authority under the water resources commission act to determine whether the dnr’s letters to plaintiff in this case constituted the functional equivalent of a suit. The water resources commission act, MCL 323.1 et seq.) MSA 3.521 et seq., was repealed after the dnr letters in this action were sent, but the pertinent sections were incoiporated into the Natural Resources and Environmental Protection Act, MCL 324.101 et seq.; MSA 13A.101 et seq.⁵

The Natural Resources and Environmental Protection Act prohibits direct or indirect discharge of injurious substances into the state’s waters without a permit. MCL 324.3109; MSA 13A.3109. Any discharge made without the necessray permit is prima facie evidence of the existence of a public nuisance, and, in addition to remedies provided in the act, the Attorney General may bring a court action for abatement of the nuisance. MCL 324.3109(4); MSA 13A.3109(4). The DNR may enter into or upon private or public property to inspect and investigate conditions relating to the pollution of state waters. MCL 324.3105; MSA 13A.3105. When the dnr is of the opinion that a person is causing or is about to cause unlawful pollution of Michigan waters, it may notify the alleged offender of its determination and order the person to abate the pollution or may refer the matter to the Attorney General for legal action. MCL 324.3112(2); MSA 13A.3112(2). A person receiving, or otherwise aggrieved by, such an order may seek a contested case hearing under the Administrative Procedures Act, MCL 24.201 et seq.) MSA 3.560(101) et seq. MCL 324.3112(3); MSA 31A.3112(3). Thus, a polluter who is required by a dnr order to take action to abate pollution may follow the order or contest it through administrative hearings.

Moreover, the dnr itself may enforce the act through lodging a criminal complaint, MCL 324.3114; MSA 13A.3114, or requesting the Attorney General to commence a civil action. MCL 324.3115(1); MSA 13A.3115(1). In addition to other relief granted, the court hearing the civil action may impose a civil fine of at least $2,500 but no more than $25,000 for each day of violation and award the prevailing party reasonable attorney fees. Id. If it is determined that a person knew or should have known that a discharge was made without a permit or that a person intentionally made a false statement, representation, or certification in a permit application, the offender is guilty of a felony and faces a fine of between $2,500 and $25,000 for each day of violation, with higher penalties for successive convictions. A two-year term of imprisonment is possible. MCL 324.3115(2); MSA 13A.3115(2). The Attorney General may commence a civil suit to recover the full value of the injuries to the natural resources that result from the violation and the costs of surveillance and enforcement of the statute. Id. The act provides for additional civil and criminal penalties upon a finding that the conduct posed a substantial endangerment to the public health, safety, or welfare. MCL 324.3115(3), (4); MSA 13A.3115(3), (4).

The dnr plainly could take substantial action against plaintiff. In its final letter to plaintiff, it indicated that “other MDNR Divisions may pursue additional enforcement actions, civil and/or criminal, and/or penalties” and that “[f]ailure to comply with this request may result in escalated enforcement action taken against SMDA.” Clearly, the DNR considered that its actions amounted to enforcement of the statutory prohibition against unpermitted discharges. Applying controlling legal authority to these facts, we hold that the DNR letters constituted a “suit” for purposes of triggering the insurers’ duty to defend.

Moreover, the three DNR letters, sent over the course of three years, exceeded the scope of general demand letters. The letters were not “garden variety” demand letters, which expose one only to a potential threat of litigation, Bronson Plating, supra at 573, n 13. The letters did not merely potentially threaten litigation, they instructed plaintiff to remedy the groundwater contamination. Further, the DNR is not an ordinary party to litigation; rather, it is the administrative agency responsible for highly specialized environmental concerns that arise in this state.

Our holding should not be construed as bestowing additional powers upon the DNR. Rather, in the context of this case, we hold only that the DNR letters constituted the initiation of a suit.

VII. THE POLLUTION EXCLUSIONS

Next, defendants argue that the pollution exclusion clauses contained in their policies bar the claims for which'plaintiff seeks coverage. Defendants assert that the trial court erred in failing to grant them summary disposition because of the pollution exclusion clauses.

A. SUDDEN AND ACCIDENTAL

Thé comprehensive general liability policy provisions at' issue contain the “sudden and accidental” exception to the pollution exclusion clause. The provisions stated that comprehensive general liability insurance did not apply

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 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.

Our Supreme Court has concluded that the standard terms used in pollution exclusion provisions are unambiguous. Upjohn, supra, 438 Mich 207 (citing United States Fidelity & Guaranty Co v Star Fire Coals, Inc, 856 F2d 31 [CA 6, 1988]; United States Fidelity & Guaranty Co v Murray Ohio Mfg Co, 875 F2d 868 [CA 6, 1989]; FL Aerospace v Aetna Casualty & Surety Co, 897 F2d 214 [CA 6, 1990]). The Upjohn Court stated:

We conclude that when considered in its plain and easily understood sense, “sudden” is defined with a “temporal element that joins together conceptually the immediate and the unexpected.” Star Fire Coals, supra at 34. The common, everyday understanding of the term “sudden” is “ ‘happening, coming, made or done quickly, without warning or unexpectedly; abrupt.’ ” FL Aerospace, supra at 219. “Accidental” means “[o]ccurring unexpectedly and unintentionally; by chance.” The American Heritage Dictionary: Second College Edition, p 72. [Upjohn, supra at 207-208.]

The term “accident” has been defined more specifically as “ ‘ “an undesigned contingency, a casualty, a happening by chance, something out of the usual course of things, unusual, fortuitous, not anticipated, and not naturally to be expected.” ’ ” Arco Industries, supra, 448 Mich 404-405. (Citations omitted.) Thus, unless the discharge is immediate and unexpected (sudden) and by chance and “out of the usual course of things” (accidental), coverage for bodily injury or property damage arising out of that discharge is not triggered.

B. DISCHARGE OF WASTE VERSUS LEAKAGE OF LEACHATE

Defendants argue that, in deciding whether the “discharge” was sudden and accidental, the trial court erred in failing to focus on the discharge of waste into the landfill rather than the leakage of leachate from the landfill. They assert that under Protective Nat’l Ins Co of Omaha v Woodhaven, 438 Mich 154; 476 NW2d 374 (1991), and Traverse City Light & Power Bd v Home Ins Co, 209 Mich App 112; 530 NW2d 150 (1995),⁶ the trial court should have concluded that because the placement, or discharge, of waste into the landfill was intentional and occurred over a number of years, the discharge was not sudden and accidental. Plaintiff counters that the trial court properly ruled that the discharge of leachate from the landfill into the soils and groundwater is the relevant discharge.

This Court previously decided this issue in Kent Co, supra, which involved a licensed landfill. When the Sparta landfill began operating, it was considered “state of the art,” but it was not artificially lined; a seven-foot barrier of soil separated the bottom of the landfill from the water table. Id., 217 Mich App 257. It was not until 1979, two years after the landfill had been closed, that the plaintiff received complaints from surrounding property owners regarding their well water. Id. at 258. After the dnr ordered the plaintiff to take remedial action, the plaintiff brought a declaratory judgment action to compel its insurers to defend and indemnify it in any potential dnr action regarding clean-up and contamination. Id. at 254. Some of the insurance policies, which excluded pollution-based claims, contained the “sudden and accidental” exception. Others contained an exception for discharges that were neither expected nor intended by the insured.⁷ Id. at 255-256.

In Kent Co, regarding the policies containing the “sudden and accidental” exclusions, this Court decided that it need not characterize the relevant discharge as either the initial placement of waste into the landfill or the subsequent discharge of leachate into the groundwater because the evidence did not show that either discharge had the requisite temporal element of suddenness. Id. at 262-264. However, recognizing that those policies contained the “neither expected nor intended by the insured” language, this Court concluded that the trial court erroneously applied Woodhaven. Id. at 264-271.

In analyzing the relevant discharge for purposes of the pollution exclusions, this Court emphasized that the landfill was constructed and licensed with the intent to contain the waste materials placed in it. Id. at 269-271. This Court concluded:

[Application of the pollution exclusion requires that the court focus on the initial discharge into the environment. If that discharge is intended, there is no coverage, notwithstanding that the damage may have been unintentional. If waste materials are placed in a contained area or structure and later escape into the environment, the latter discharge is the relevant discharge. [Id. at 288.]

In this case, the landfills were designed consistent with then-contemporary standards. Experts testified that plaintiff responded sufficiently to the outbreaks under those standards. From the outset of their operations, these landfills, although approved by the dnr, were fraught with problems arising from their operation and maintenance. These continuing problems, however, do not render it improper to focus on the discharge of contaminants from the landfills. The refuse was placed into the landfills that had been constructed to contain the waste. Consistent with the holding in Kent Co, we agree that the proper focus is the leaking of leachate and contaminants from the landfill into the surrounding soil and groundwater because the refuse had been placed into a landfill designed and licensed to contain waste.

C. OTHER SOURCES OF CONTAMINATION

Because the trial court correctly focused on the discharge of leachate and contaminants from the landfills, we next determine whether the “sudden and accidental” exception to the pollution exclusions applies. Plaintiff, however, raises a preliminaiy matter, asserting that it presented evidence that created a genuine issue of material fact regarding other sources of contamination.

In Polkow v Citizens Ins Co of America, 438 Mich 174; 476 NW2d 382 (1991), our Supreme Court addressed whether coverage existed under a sudden and accidental exception to a pollution exclusion. The plaintiff testified about alleged accidental spills over the years in the transfer of oil from the tanker trucks to underground storage tanks. On that basis, the Court ruled that reasonable jurors could determine that the plaintiff had expected the release of contaminants. The evidence also reflected, however, that the contaminants were not from oil and may have been unrelated entirely to the plaintiffs operations. The contamination possibly was caused by a discharge from the underground tanks. Id. at 178-179. The Supreme Court concluded:

Therefore, resolution of whether the sudden and accidental exception applies is impossible, given the current state of factual development. The resolution of this question requires an examination of whether the discharge of pollutants was sudden and accidental. On a record where it is unclear even what the discharge was, how can we, absent some form of augury, possibly declare that this unknown form of discharge was not sudden and accidental? ... It was the “duty of [the insurer] to undertake the defense until it could confine the claim to a recovery that the policy did not cover.” Jonesville Products, Inc v Transamerica Ins Group, 156 Mich App 508, 513; 402 NW2d 46 (1986), cited with approval by the majority in Protective Nat’l Ins Co of Omaha v City of Woodhaven, 438 Mich 154; 476 NW2d 374 (1991), a companion to this case. [Polkow at 179.]

In this case, plaintiff argues that other possible sources contaminated each landfill area, and that, until the nature and origin of the contamination are determined, the court cannot decide whether the releases were sudden and accidental.

The trial court found that plaintiffs evidence raised a question of fact regarding the source of the contamination at sites 9 and 9A. The court relied on the preliminary findings of a September 6, 1994, hydrogeologic investigation that addressed other sources of chemicals, elevated organics, and metals. We reject the trial court’s holding as clearly erroneous, because the report at best merely identifies possible sources of contamination. Plaintiff did not meet its burden.

The report identified one landfill within a one-mile radius of sites 9 and 9A and one residential well listed under the Michigan Environmental Response Act, MCL 299.601 et seq.; MSA 13.32(1) et seq.8 The report does not reflect any problems at the landfill and draws no conclusions about whether the landfill affected water or soil quality.

Although tetrachloroethylene (pce) appeared in samples, the report concluded that the source of the pce was unknown.⁹ Also, samples revealed the presence of benzene; however, the report characterized the source as unknown.¹⁰ Moreover, storage of automobile bodies, parts, and other debris on property north of the sites may have affected adversely soil and groundwater.¹¹ The report also investigated potential origins of diethyl ether (de) in groundwater at the site, but concluded that the DE could have naturally occurred.¹² Further, it posits that a possible source of chemicals was the drilling muds.¹³ Finally, the report discusses metals that are present in naturally occurring brines and chemicals utilized in septic tanks.

Plaintiff suggests the foregoing as other sources of contamination. Plaintiff has presented nothing beyond mere speculation that these sources contributed to the contamination associated with sites 9 and 9A. Mere speculation is insufficient: “A party opposing a motion for summary disposition must present more than conjecture and speculation to meet its burden of providing evidentiary proof establishing a genuine issue of material fact.” Cloverleaf Car Co v Phillips Petroleum Co, 213 Mich App 186, 192-193; 540 NW2d 297 (1995). Plaintiff has failed to provide sufficient proof to establish a genuine issue of material fact regarding other sources of contamination.

With respect to sites 7 and 11, plaintiff has attempted to prove that the contamination was caused by landfilling activities: (1) on the property while the Macomb County Road Commission owned and operated the site, (2) at the Predmore Landfill, and (3) at a landfill located across the road from sites 7 and 11. Regarding plaintiff’s first and second proposed causes, the Washington Township dump and the Predmore Landfill were located on property that became sites 7 and 11. Plaintiff has provided no authority demonstrating that it cannot be held responsible for contamination caused by these landfills, upon which it constructed sites 7 and 11.

Regarding the adjacent Walker Land Reclamation site, plaintiff asserts that it has produced evidence of direct dumping of chemicals and wastes. The record does not support plaintiff’s claim. The alleged evidence either involves self-serving statements and unsubstantiated allegations or fails to implicate the Walker site as a contributor to the contamination problem. Plaintiff properly presented evidence only that the groundwater flows from the Walker site to sites 7 and 11; however, no evidence suggests any leakage from the Walker site. Plaintiff has failed to “set forth specific facts showing that there is a genuine issue for trial” as MCR 2.116(G)(4) requires.

Plaintiff also asserts that the dnr letters instructed plaintiff to perform a hydrogeological investigation to determine other sources of contaminants. After explaining that it found vocs present at site 11, the dnr in its April 17, 1990, letter states:

Therefore, [the dnr] is requesting that smda submit a work plan detailing a hydrogeological investigation of the SMDA #11 site. The vertical and lateral extent of the groundwater contamination must be fully defined, and the investigation must assess the hydrogeological characteristics of the site. Included in the work plan should be direct and immediate measures that smda will implement to prevent further degradation of the area’s potable aquifer. After the remedial investigation, a feasibility study is required to investigate possible remediation techniques that could be used at the site.

In this letter, the dnr unequivocally identified plaintiffs landfill as the cause of the groundwater contamination. It did not suggest other possible contamination sources. Contrary to plaintiff’s assertion, the DNR ordered the hydrogeological investigation to determine the extent of contamination, not other possible sources of the contamination. Plaintiff has not created a genuine issue of material fact regarding other sources of contamination at sites 7 and 11.

Thus, the trial court erred in concluding that plaintiff presented a genuine issue of material fact regarding other sources of contamination associated with either sites 9 and 9A or sites 7 and 11. Therefore, we must determine whether the court erred in refusing to grant defendants’ motions for summary disposition on the pollution exclusions.

The trial court found that with respect to sites 9 and 9A, the long-term leaking through the porous bottom of the sites was not sudden and accidental. Regarding the three allegedly distinct outbreaks from the walls of site 9A, however, it concluded a genuine issue of material fact existed whether these outbreaks were expected or instead were sudden and accidental. Regarding sites 7 and 11, the trial court found that a determination whether the leaking at the landfill was sudden and accidental would be premature because plaintiff had presented a genuine issue of material fact regarding other sources of contamination. It therefore denied defendants’ motions for summary disposition on the “sudden and accidental” issue as to both landfill areas. We will first discuss sites 9 and 9A.

D. IDENTIFIABLE DISCHARGES VERSUS LARGER PATTERN OF DISCHARGE/LEAKING

Whether it is proper to separate identifiable discharges from a larger pattern of discharge or leakage presents a question of first impression. The policies covered bodily injury or property damage arising from a discharge, dispersal, release, or escape that is sudden and accidental. The policy language indicates that, if a specific discharge that causes bodily injury or property damage can be deemed sudden and accidental, then coverage is provided. Defendants contend that courts have rejected a “microanalytical” approach to characterizing discharges as sudden and accidental. Defendants cite cases that primarily involve unsubstantiated claims of sudden and accidental discharges in the face of repeated, continuous discharges in the course of business. See Ray Industries, Inc v Liberty Mut Ins Co, 974 F2d 754, 768 (CA 6, 1992) (court rejected argument that, although barrels containing waste were routinely and regularly crushed, each discrete release was sudden: “under this theory, all releases would be sudden”); A Johnson & Co, Inc v Aetna Casualty & Surety Co, 933 F2d 66, 75 (CA 1, 1991) (allegation in prp letter that leakage from cracked tanks was not sufficient to support a finding of a sudden and accidental release where a variety of disposal methods allegedly contributed to contamination in a long-term course of conduct: “Mere speculation . . . that any individual instance of disposal . . . occurred ‘suddenly’ cannot contradict a reasonable reading of the allegations that the entire pattern of conduct was not a ‘sudden and accidental’ occurrence”); Smith v Hughes Aircraft Co, 10 F3d 1448, 1453 (CA 9, 1993) (contamination resulted from practice of dumping chemicals into unlined ponds; court refused to “break down [Hughes’] long-term waste practices into temporal components in order to find coverage where the evidence unequivocally demonstrates that the pollution was gradual” and rejected Hughes’ argument that “sudden” polluting events, which caused dumping directly into the ponds, partially caused the contamination).

In Lumbermens Mut Casualty Co v Belleville Industries, Inc, 938 F2d 1423, 1430 (CA 1, 1991), the court concluded that the application of a sudden and accidental clause depended on “the nature of an insured’s enterprise and its historical operations.” Thus, where the insured’s manufacturing operations involved continual pollution-releasing activity, the insured’s attempt to characterize discharges following a torrential rainstorm and fire as sudden and accidental was futile.

In the cases cited, the courts refused to focus on individual discharges. The present case is distinguishable, however, because (1) the outbreaks purportedly were identifiable and isolated, (2) they arguably were different in kind from the overall leakage through the porous bottoms of the landfills, (3) specific damage arguably may be traced to these outbreaks, and (4) plaintiff has provided evidence of their occurrence.

The Sixth Circuit Court of Appeals has held that if a company whose ongoing activities result in contamination of soil presents a scintilla of evidence that some of the spills were sudden and accidental, the insurer’s motion for summary judgment is properly denied. Employers Ins of Wausau v Petroleum Specialties, Inc, 69 F3d 98 (CA 6, 1995). Petroleum Specialties, Inc. (psi), owned and operated an oil refinery in Flat Rock, Michigan, from the early 1930s until about 1964. After 1964, the property was used as a petroleum storage facility. In 1989, the dnr conducted a site evaluation and discovered an open lagoon of petroleum; soil contaminated with polychlorinated biphenyls (PCBs), lead, arsenic, cyanide, and benzene; asbestos construction material; leaking electrical transformers; and leaking drums of xylene, toluene, and lead. Id. at 100. The dnr notified psi that it was considered a potentially responsible party to the contamination at the facility. One of psi’s insurers, Zurich Insurance Company, covered PSI for property damage caused by an “accident” that occurred during the policy period, 1951 to 1956. The policy contained no pollution exclusion. Id. at 101.

One of the dnr’s concerns regarding the property was the allegedly severe long-term damage. The Sixth Circuit Court of Appeals concluded that, to the extent PSI could present evidence that the damage stemmed from accidental spills during the period the Zurich policy was in effect, the long-term damage might be covered. Id. at 103. In answer to Zurich’s motion for summary judgment, PSI presented sworn discovery answers that described a series of discrete releases from a condenser box and leaking valves and pumps that occurred between 1951 and 1956. The court found that because this evidence presented a factual issue, summary judgment for Zurich was inappropriate. Id.

We find Petroleum Specialties persuasive. Plaintiff has presented facts regarding the 1971, 1976, and 1980 outbreaks of leachate from the sides of site 9A. The evidence supports plaintiffs position that some of the contamination and damage was attributable directly to these three leachate outbreaks. The reasoning of Petroleum Specialties governs this case. The trial court did not err in separating these claimed isolated “sudden and accidental” discharges from the overall continuous leaking of the landfills.

E. WERE THE LEACHATE OUTBREAKS SUDDEN AND ACCIDENTAL?

Defendants next contend that the trial court erred in denying summary disposition because the leachate outbreaks were not sudden and accidental. In policies like those at issue,

[t]he actual language of this exclusion clause “unambiguously links the terms ‘sudden and accidental’ with the release as opposed to the knowledge, intent, or expectation of the insured.” Matakas v Citizens Mut Ins Co, 202 Mich App 642, 653; 509 NW2d 899 (1993). Accordingly, under this policy the insured’s actions must be analyzed under an objective rather than subjective standard. [Traverse City Light & Power, supra, 209 Mich App 119.]

Defendants assert that because plaintiff expected the discharge of leachate and contaminants from the landfills, the discharge was not “accidental.” This argument is premised on plaintiffs failure to design the landfills properly consistent with Mozola’s recommendation and its repeated violations of operating and maintenance requirements.

In City of Clare, supra, the Court followed Upjohn in determining what constituted a “sudden and accidental” discharge. Where the insured should have expected the discharge, the discharge cannot be considered “sudden and accidental” because the discharge was not “unexpected.” City of Clare, supra, 446 Mich 13-14 (citing Upjohn, supra at 215, n 3, 216-217). City of Clare involved a landfill operated by the City of Clare. Four years after the city began operating the landfill, the Solid Waste Management Act, 1978 PA 641, repealed the act under which the landfill was licensed. Nonetheless, the landfill remained open. City of Clare, supra at 3.

Two years later, the dnr informed the city to upgrade or to close the landfill and reminded the city that the landfill was unlicensed and that the city should complete a hydrogeologic survey. The landfill remained unlicensed. Id. at 4. Five years later, the dnr received the city’s consent to perform the hydrogeologic study itself. The DNR suspected that the landfill was contaminating groundwater and notified the city to close and cap the site. Id. at 4-5. The next year, the dnr informed the city that the study indicated that contamination was emanating from the landfill and advised the city to terminate operations immediately and to close and cap the site. Id. at 5-7.

Hatton Township thereafter sued the city, alleging violations of the Solid Waste Management Act and the Michigan Environmental Protection Act, MCL 691.1201 et seq.; MSA 14.528(201) et seq. Two residential wells near the landfill were contaminated, and contaminants had been detected in surface water seeping from the landfill. The dnr then ordered the city to cease and desist its operation of the landfill and to prepare for closure and cleanup. City of Clare, supra at 6-7.

The Court applied the Upjohn reasoning and found that the “City of Clare must have expected the release of pollutants into nearby lands and waters.” Id. at 14. The DNR repeatedly had informed the city that the landfill did not comply with the law and that it would be best to close and seal the site. Tests taken over a seven-year period confirmed ground and surface water contamination. Thus, the Court found that the release was expected and not “sudden and accidental.” Id.

In contrast to City of Clare, the sites in this case satisfied the minimum requirements for licensure and remained licensed while operating. Also, in City of Clare, the DNR ordered the city to cease operating the landfill, while in this case the DNR continued to allow operation of sites 9 and 9A. In City of Clare, the city knew the landfill was not in compliance with pertinent statutes. Here, plaintiffs sites met the threshold requirements. On these facts only, it cannot be said that plaintiff expected the discharges within the meaning of the holding of City of Clare.

In light of City of Clare, however, the violations noted in the numerous inspection reports bear on whether plaintiff expected the leaking of leachate from the landfill. We are persuaded by Harleysville Mut Ins Co, Inc v Sussex Co Delaware, 831 F Supp 1111 (D Del, 1993), aff'd 46 F3d 1116 (CA 3, 1994). Maryland Casualty, one of the insurers in Harleysville, argued that claims against a landfill did not fall within the sudden and accidental exception because Sussex County, the operator of the landfill, expected and intended to discharge leachate. The insurer alleged that the county designed and operated the landfill in violation of the state solid waste disposal code and that its employees expected the discharge of leachate. 831 F Supp 1125. The court found that the evidence raised a genuine issue of material fact whether the county expected that the landfill would discharge leachate. Id. Maryland Casualty’s evidence raised a question regarding whether the county could have expected the discharges, based on the noncomplying design and construction of the landfill. The evidence indicated that waste may have come into contact with the groundwater and that the county failed to perform tests to determine the high water table underlying the landfill. The evidence also raised a question whether, when it repeatedly failed to comply with daily cover requirements, the county knew that a primary cause of leachate was the percolation of precipitation through the waste material and that the daily cover was intended to minimize the amount of percolation of precipitation. Id. at 1126-1127.

In this case, each of the three alleged distinct leachate outbreaks from the walls of site 9A can be considered separately from the overall leakage problem for purposes of the “sudden and accidental” exception to the pollution exclusion in the policies. We therefore consider each outbreak individually to determine whether the trial court erred in finding that genuine issues of material fact existed about whether the outbreaks were sudden and accidental.

The first of the outbreaks occurred in 1971. Warren Anderson testified that problems with the under-drains caused this outbreak. An expert who reviewed plaintiff’s documentation on this outbreak explained that the outbreak appeared to have resulted from the plugging of the system, although he did not specifically so state. On this evidence and the evidence regarding the design and operation of the site, a genuine issue of material fact exists whether this outbreak was sudden and accidental. See Harleysville, supra. Therefore, on remand, the circumstances of this outbreak must be further developed so that the court may determine whether it was expected or sudden and accidental.

The second leachate outbreak occurred in June 1976. Anderson investigated this outbreak and found that heavy rains had caused erosion gullies in the cover material, which exposed the waste. Rainwater then contacted the rubbish, generating leachate. This leachate ran into the perimeter ditch and into the McBride Drain. Anderson also testified that heavy rains typically cause erosion gullies in landfills and that even after remedying this particular situation they reoccurred. Such erosion gullies normally developed at the point where the surface of the landfill would have a one percent slope and then would change to a steeper slope. The erosion would occur at that point because the velocity of the water flowing over the surface would increase at the point where the slope changed.

On the basis of Anderson’s testimony, plaintiff must have expected the development of erosion gullies and the subsequent generation of leachate. Anderson testified that this was a typical occurrence. The trial court erred in finding a question of fact regarding this outbreak. The 1976 outbreak could not be termed sudden and accidental.

Finally, the third outbreak occurred in 1980. Anderson investigated this outbreak by digging trenches into the waste material. The trenches filled with water that appeared to migrate from the upstream area of the slope. The water penetrated the cover material and ran along the bottom of the cover material but on top of the waste. The water ran down to the dike, creating hydrostatic pressure on the dike and eventually breaking through. Anderson determined that the surface water penetration was caused by small settled areas that held water and erosion gullies. As noted above, plaintiff must have expected that erosion gullies would develop. Because plaintiff should have expected the gullies, the resulting outbreak could not be classified as sudden and accidental.

Also, an expert indicated that this outbreak was caused by water ponding on the top of the refuse under the surface and migrating southward to the south wall, where it broke from the landfill. Although Anderson testified that this was an unusual phenomenon, the evidence shows that the top of the landfills had a low degree of slope, approximately one percent or less. Further, an expert stated that the cover material on the site was not effective in preventing or minimizing infiltration and did not promote surface runoff to a desirable extent because the cover material was not uniformly low in permeability and was of fairly low slope. Another expert testified that the one percent slope found on the sites is minimal for landfill cover. Water is shed slowly, so there is more of an opportunity to infiltrate. Moreover, landfills settle, and when a one percent slope settles, it may become a slope of zero or less than zero percent, and rainwater will not be shed. Additionally, plaintiff frequently had been cited for failing to provide adequate daily cover on the site and to properly maintain the site.

Under Harleysville, we consider the violations from plaintiff’s landfills as well as the design of the landfills, including the degree of slope. The evidence shows that plaintiff had been cited for violations for inadequate cover and improper maintenance — plaintiff did not rectify the situation. We also consider the minimal cover slope, the expected phenomenon of erosion gullies, and the known characteristics of the cover material. That evidence does not support a finding that the 1980 outbreak was sudden and accidental.

Regarding sites 7 and 11, the trial court concluded that summary disposition on the application of the pollution exclusion was improper because the facts established several possible sources of contamination. The court also found a genuine issue of material fact regarding whether plaintiff expected the contamination to occur. As discussed above, plaintiff’s evidence supporting its claim of other sources of contamination is insufficient to survive summary disposition. Plaintiff has presented no facts to show that the leakage from sites 7 and 11 was “sudden.” Therefore, the trial court erred in denying defendants’ motions that the claims regarding sites 7 and 11 were not precluded by the pollution exclusions because the discharges were not “sudden and accidental.”

vm. THE PERSONAL INJURY LIABILITY PROVISIONS

Plaintiff next contends that the personal injury liability endorsements in the Citizens and ffic policies afford it coverage outside the pollution exclusion. We reject this contention.

Ffic’s policies provided that “[s]uch insurance as is afforded by the policy for Comprehensive General Liability Insurance ... is amended to include the following additions and extension of coverage,” which set forth the provisions of personal injury liability. The Citizens policies, on the other hand, provided coverage for personal injuries with no statement that the coverage added to or extended the comprehensive general liability coverage. The personal injury liability provisions of the policies provided coverage for damages arising out of,,among other things, “wrongful entry or eviction, or other invasion of the right of private occupancy.” Plaintiff asserts that the claims for which it seeks coverage fall within either “wrongful entry” or “other invasion of the right to private occupancy.”

This Court previously addressed this argument in Kent Co, supra. Kent County similarly sought coverage for the dnr claims under these personal injury liability offenses. The dnr’s action against Kent County arose from the contamination of groundwater, as in the instant case. Kent Co, supra, 217 Mich App 289. This Court determined that the offenses “must affect a right of private occupancy,” but that the action involved a public right — assuming that the people of the state owned the groundwater. Id. at 290 (citing Harrow Products, Inc v Liberty Mut Ins Co, 833 F Supp 1239 [WD Mich, 1993], aff'd in part, rev’d in part, and remanded 64 F3d 1015 [CA 6, 1995]). Thus, the claims for which Kent County sought coverage did not fall within the personal injury liability provision. Applying this reasoning, we hold that the dnr claims do not fall within the personal injury liability endorsements of plaintiffs policies with Citizens and FFIC.

Whether a personal injury liability endorsement provides coverage for the pollution claims of private plaintiffs where those claims otherwise are barred by a pollution exclusion clause in a comprehensive general liability policy is a question of first impression. We adopt the reasoning of the Sixth Circuit Court of Appeals in Harrow for application to the claims of the private plaintiffs in the Bielat action. Whether pollution-related claims excluded by pollution exclusion clauses may be resurrected through a personal injury liability endorsement is an issue upon which the authorities are split.¹⁴ The Sixth Circuit Court of Appeals examined several cases on both sides of the issue¹⁵ and decided that our Supreme Court would conclude that where a comprehensive general liability pollution exclusion bars claims, a personal injury liability endorsement will not provide coverage for damage to property by pollution. Harrow, supra, 64 F3d 1024. To do so, explained the court, would “render the pollution exclusion clause a nullity.” Id. at 1025. The Sixth Circuit Court of Appeals cited favorably the following language in Columbia Co v Continental Ins Co, 83 NY2d 618, 627-628; 612 NYS2d 345; 634 NE2d 946 (1994):

“[T]he coverage under the personal injury endorsement provision in question was intended to reach only purposeful acts undertaken by the insured or its agents. Evidence that only purposeful acts were to fall within the purview of the personal injury endorsement is provided, in part, by examining the types of torts enumerated in the endorsement in addition to wrongful entry, eviction and invasion: false arrest, detention, imprisonment, malicious prosecution, defamation and invasion of privacy by publication. Read in the context of these other enumerated torts, the provision here could not have been intended to cover the kind of indirect and incremental harm that results to property interests from pollution.” [Harrow, supra, 64 F3d 1022.]

It is illogical to conclude that an insurer specifically excluded claims under the comprehensive general liability provision, only to include them within the personal injury liability endorsement. Such an interpretation would neither harmonize conflicts between clauses nor avoid an interpretation that renders the provision unreasonable. Fresard, supra, 414 Mich 697-698. Thus, to the extent the pollution exclusion clauses in the various policies bar coverage of a claim, plaintiff may not seek coverage for those claims under the personal injury liability endorsements.¹⁶ The trial court therefore erred in granting plaintiffs motion for summary disposition regarding the personal injury liability coverage issue.

ix. defense costs

Ffic contends that the trial court erred in ordering it to pay plaintiff’s alleged defense costs. It asserts that the trial court erroneously characterized certain costs as defense costs rather than indemnification costs because they were related to implementing the remediation of pollution. Furthermore, FFIC asserts that because it had no duty to defend plaintiff, it is entitled to reimbursement of monies paid on plaintiff’s behalf. We disagree.

This Court recently has stated:

The duty to defend is essentially tied to the availability of coverage. . . . [T]he duty to defend arises in instances in which coverage is even arguable, though the claim may be groundless or frivolous. . . . Consistent with this premise, any analysis of an insurer’s duty to defend must begin with an examination whether coverage is possible. ... If coverage is not possible, then the insurer is not obliged to offer a defense. [Arco Industries Corp v American Motorists Ins Co (On Remand), 215 Mich App 633, 636; 546 NW2d 709 (1996), lv gtd 454 Mich 873 (1997).]

Moreover, an insurer has a duty to defend, even where only some of the theories of liability are covered by the policy. Woodhaven, supra, 438 Mich 159 (quoting Detroit Edison Co v Michigan Mut Ins Co, 102 Mich App 136, 141-142; 301 NW2d 832 [1980]). In Polkow, supra, 438 Mich 180, our Supreme Court further explained:

Fairness requires that there be a duty to defend at least until there is sufficient factual development to determine what caused the pollution so that a determination can be made regarding whether the discharge was sudden and accidental. Until that time, the allegations must be seen as “arguably” within the comprehensive liability policy, resulting in a duty to defend.

Thus, until it can be demonstrated that none of the claims fall within the policies, defendants have a duty to defend and are responsible for defense costs.

The duty to defend, however, is broader than the duty to indemnify. American Bumper & Mfg Co v Hartford Fire Ins Co, 452 Mich 440, 450-451; 550 NW2d 475 (1996). It follows that an insurer may be responsible for defense costs but not indemnification costs. Indeed, indemnification costs are damages. An insurer “is not obligated to indemnify or pay damages to another until one’s liability for the injury has been established.” Gelman Sciences, Inc v Fireman’s Fund Ins Cos, 183 Mich App 445, 448-449; 455 NW2d 328 (1990). Defense costs are those expenses related to developing and setting forth a theory establishing that the defendant is not liable or otherwise challenging liability. Id. at 448. Therefore, it is necessary to determine whether the costs ordered by the trial court are defense or indemnification costs.

Our Supreme Court in American Bumper agreed with the analysis of Gelman and explained that “site investigation costs incurred during the RI/FS are defense costs, rather than indemnification costs, if they were expended in order to disprove or limit the scope of liability for cleanup . . . and if they do not represent an ordinary cost of doing business.” American Bumper, supra at 463. However, costs expended for remediation or making a potentially injured party whole are indemnification costs. Id. at 461. The conclusions of the federal district courts are persuasive. For example, in Fireman’s Fund Ins Co v Ex-Cell-O Corp, 790 F Supp 1318, 1321 (ED Mich, 1992), the court included as defense costs those “that are reasonable and necessary to limiting the scope and/or costs of remediation.” Also, in Hi-Mill Mfg Co v Aetna Casualty & Surety Co, 884 F Supp 1109, 1117 (ED Mich, 1995), aff’d 98 F3d 1341 (CA 6, 1996), the court included as defense costs those costs associated with a ri/fs that identified, evaluated, and compared alternative remediation plans.

The costs at issue are associated with designing the most cost-effective remediation plan. These were costs plainly associated with limiting the damages that plaintiff would be required to pay. Considered in this light, defense costs should include the costs related to proving that plan for remediation of the landfills may be more cost-effective than those proposed by the plaintiffs in the underlying action. The trial court did not err in ruling that defense costs included costs relating to the studies conducted to determine the extent of remediation and costs to develop a remediation plan.

X. THE loss-in-progress doctrine

Westchester argues that the trial court erred in failing to grant its motion for summary disposition under the “loss-in-progress” doctrine. We disagree.

As noted in American Bumper, supra at 459, Michigan courts have not adopted the loss-in-progress doctrine. Nonetheless, the Sixth Circuit Court of Appeals in Inland Waters Pollution Control, Inc v Nat’l Union Fire Ins Co, 997 F2d 172, 178-179 (CA 6, 1993), reasoned that our Supreme Court would adopt the doctrine because courts throughout the country apply the doctrine, a fundamental principle of insurance law.

The “loss-in-progress” doctrine defeats an insured’s claim against an insurer under a policy “only where the insured is aware of a threat of loss so immediate that it might fairly be said that the loss was in progress and that the insured knew it at the time the policy was issued or applied for.” Id. at 178. Thus, application of the doctrine requires “foreknowledge of loss or an awareness of an immediate threat of loss on the part of the insured.” Id.

Plaintiff first argues that the loss-in-progress doctrine does not apply to third-party liability insurance. In Inland Waters, however, the doctrine applied to a third-party liability situation: Inland Waters, the insured, was sued by Strieker Paint Products, Inc., for damages done to Strieker’s property through Inland Waters’ crushing of drums of waste paint materials on Strieker’s property. Id. at 173. Plaintiff’s argument is without merit.

Plaintiff also challenges the doctrine on the bases of the allegedly absurd consequences that would result from the application of the doctrine and the fact that this case involves more than one occurrence. Plaintiff asserts that because insureds always are seeking insurance for known risks, any application of the doctrine would negate coverage for losses. The doctrine, however, has limited force: the loss must be so immediate that it can be considered in progress when a policy is applied for or issued. Thus, its application would not result in the absurd consequences posed by plaintiff.

Westchester’s policies covered losses for which plaintiff may be held liable arising from an “occurrence,” which is defined as having “cause [d] injury to persons or tangible property during the policy period.” By definition, therefore, an injury to persons or property must occur to trigger a loss. Even where a loss has occurred, however, the loss may be insured against — provided the occurrence was not known when the insured sought the insurance. Inland Waters, supra at 176. As Inland Waters recognized, plaintiff must have been aware of, at a minimum, an immediate threat of injury to persons or property for the loss-in-progress doctrine to apply.

Thus, application of the doctrine requires that, when the specific policies were issued: (1) the losses had occurred and (2) plaintiff had the requisite foreknowledge. Because questions of fact exist regarding plaintiffs foreknowledge of loss or an awareness of an immediate threat of loss, summary disposition would be inappropriate on this issue. The trial court correctly denied the motion on this basis. Therefore, it is unnecessary to determine whether Michigan courts would adopt the doctrine.

XI. WAIVER AND ESTOPPEL

Westchester finally argues that the trial court erred in applying the doctrine of waiver and estoppel against it because plaintiff raised the issue only against the other defendants. Westchester asserts that it never misled plaintiff about the terms of its policy and did not delay in asserting its defenses. We agree.

“Waiver ... is bottomed on the doctrine of estoppel.” Ruddock v Detroit Life Ins Co, 209 Mich 638, 653; 177 NW 242 (1920). This Court has stated that, generally, “once an insurance company has denied coverage to an insured and stated its defenses, the company has waived or is estopped from raising new defenses.” Lee v Evergreen Regency Cooperative, 151 ; 390 NW2d 183 (1986). The Lee Court indicated that the doctrine may not be used to broaden policy coverage to protect an insured against risks not included in the policy or expressly excluded from the policy. This restriction is based on the rule that the insurer should not be required, through waiver and estoppel, to cover a loss for which no premium was charged. Id.

Exceptions to the general rule have been made in cases within two broad classes:

The first class involves companies which have rejected claims of coverage and declined to defend their insured in the underlying litigation. In these instances, the Court has held that the insurance company cannot later raise issues that were or should have been raised in the underlying action. These cases are closely akin to the principle of collateral estoppel....

The second class of cases . . . involves instances where the inequity of forcing the insurer to pay on a risk for which it never collected premiums is outweighed by the inequity suffered by the insured because of the insurance company’s actions. The insurance company has either misrepresented the terms of the policy to the insured or defended the insured without reserving the right to deny coverage. [Id. at 286-287 (citations omitted).]

The trial court, in finding that the pollution exclusion clause may be perceived as deceptive, apparently concluded that the instant case falls within the second class. The evidence, however, does not reflect that Westchester misrepresented the terms of the policy to plaintiff. Westchester denied coverage to plaintiff on the basis of the pollution exclusion; thus, the doctrine of waiver and estoppel should not be used to require Westchester to cover any loss that falls within the pollution exclusion. The trial court erred in finding that factual questions arose regarding waiver and estoppel because: (1) the pollution exclusion is unambiguous, (2) no evidence demonstrated that Westchester misled plaintiff as to the policies’ terms, and (3) Westchester specifically denied coverage to plaintiff based on the pollution exclusions contained in its policies.

Affirmed in part, reversed in part, and remanded. We do not retain jurisdiction.

M. Warshawsky, J., concurred.

1 Leachate results when water mixes with the waste contained in a landfill.

2 Plaintiff asserted that Citizens provided it with comprehensive general liability policies from April 1968 through August 1975. These policies apparently have been lost. The trial court found that plaintiff had presented an issue of fact about the existence of these policies, and it refused to grant Citizens’ motion for summary disposition regarding plaintiffs claims under these policies. That ruling is not at issue in this appeal.

3 Our Supreme Court vacated a portion of this Court’s judgment in Kent Co while this opinion was circulating among the panel members. However, the Court’s action does not affect our analysis of the issues. The Court vacated the portion of this Court’s judgment holding that the “sudden and accidental” exception to the pollution exclusion clauses in that case was not triggered and remanded the matter to the trial court for consideration because the trial court had not reached that issue. The Court denied leave in all other respects. Therefore, with the exception of this Court’s analysis of the “sudden and accidental” exception, the precedential effect of Kent Co remains intact.

4 The Arco Court declined to address formally whether a letter from the dnr constituted the initiation of a suit. Our Supreme Court stated, however, that it expected that Michigan Millers would provide guidance to this Court on remand. Arco, supra at 417, n 15.

5 MCL 324.107; MSA 13A.107 states: “It is the intention of the legislature that editorial changes in the language of statutes codified as parts within this act not be construed as changes to the meanings of the statutes.” Thus, our reasoning with respect to this issue applies to the dnr letters submitted under the Natural Resources and Environmental Protection Act.

6 In Woodhaven, supra at 162-163, the Court identified the “discharge” as the release of pesticide, the chemical that caused physical injury, into the atmosphere. In Traverse City Light & Power, supra at 117-118, the proper focus was on the discharge of fly ash into an abandoned gravel pit that was not a licensed landfill, not on the discharge of the leachate into the underground aquifer.

7 These policies excluded coverage for

“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 neither expected nor intended by the Insured.” [Kent Co, supra at 255.]

8 This act is now part of the Natural Resources and Environmental Protection Act, MCL 324.20102 et seq.) MSA 13A.20102 et seq.

9 In 1985, water samples collected from the residential well contained low levels of pce, but five samples taken in 1986 and 1987 revealed no vocs or pce.

10 In 1986, water from another residential well in the area was tested. It contained low levels of benzene. Samples collected later in 1986 and in 1987 revealed no vocs or benzene. The report concluded that the source of benzene was not known. Further, it was unknown whether agricultural chemicals or those used for maintenance of a nearby golf course may have affected the sites.

11 Significantly, however, the report stated: “The hydrogeologic conditions and the chemical gradients suggest that the source of chemicals on the properties to the west of Site 9 and north of Site 9A (Mustaffa and Bielat) are from the landfills.”

12 The report indicated that de had been detected in groundwater samples in several wells at the sites and that although it had not been reported to be a naturally occurring organic chemical in soil and groundwater systems, “this does not preclude the fact that de may be a naturally occurring chemical.”

13 Drilling muds are placed inside augers to maintain pressure on the formation during drilling and sampling.

14 Although we disagree with its criticism of the Sixth Circuit Court of Appeals’ decision in Harrow, the federal court in Great Northern Nekoosa Corp v Aetna Casualty & Surety Co, 921 F Supp 401 (ND Miss, 1996), discusses this split of authority and sets forth the positions of various jurisdictions.

15 Columbia Co v Continental Ins Co, 83 NY2d 618; 612 NYS2d 345; 634 NE2d 946 (1994); O’Brien Energy Systems, Inc v American Employers’ Ins Co, 427 Pa Super 456; 629 A2d 957 (1993); Titan Corp v Aetna Casualty & Surety Co, 22 Cal App 4th 457; 27 Cal Rptr 2d 476 (1994); American Universal Ins Co v Whitewood Custom Treaters, Inc, 707 F Supp 1140, 1144 (D SD, 1989); Titan Holdings Syndicate, Inc v City of Keene, New Hampshire, 898 F2d 265 (CA 1, 1990); Pipefitters Welfare Educational Fund v Westchester Fire Ins Co, 976 F2d 1037, 1042 (CA 7, 1992).

16 Moreover, the personal injury liability provision of ffic’s policies amended the definition of “bodily injury” to include “personal injury.” The pollution exclusion, as has been discussed, applies to bodily injury. Because the policy specifically amends the definition of “bodily injury” to include “personal injury,” a plain reading of the policy reveals that the enumerated personal injury offenses are excluded under the pollution exclusion.