HORTON, J.,

Retired, specially assigned under RSA 490:3. The plaintiff, EnergyNorth Natural Gas, Inc. (hereinafter referred to, collectively with its predecessors in interest, as ENGI), appeals an order of the Superior Court (Lynn, J.) granting summary judgment to the defendant, The Continental Insurance Company (hereinafter referred to, collectively with its predecessor in interest, as Continental), and denying ENGI's cross-motion for summary judgment in this insurance coverage case. We affirm.

The trial court's order recited the following facts, none of which are disputed by the parties on appeal. ENGI operated a manufactured gas plant (MGP) in Concord from approximately 1852 to 1952. The production of gas at the Concord MGP generated byproducts such as tar, phenol, anthracene, benzene, napthalene, toluene and xylene, some of which meet the definitions of "hazardous waste" or "hazardous substance" under State and federal law. See RSA 147-B:2, VII (1996 Supp. 2000); 42 U.S.C. § 9601(14) (1995). While some byproducts were recovered for resale, the remainder were either stored on-site in a "relief holder" or released into a sewer line that emptied into an area near the Merrimack River known as the "Tar Pond."

In September 1992, the New Hampshire Department of Environmental Services (NHDES) directed ENGI to conduct a site investigation to determine whether waste from the Concord MGP had contaminated the Tar Pond. In October 1996, NHDES ordered ENGI to undertake remedial measures at the Tar Pond site (the site). ENGI has spent over $3.5 million on investigation and remediation at the site and anticipates expending more in the future.

Continental insured ENGI from at least 1965 through 1985 under a number of comprehensive general liability insurance policies. The policies issued prior to 1967 were "accident-based" policies, which required Continental "[t]o pay on behalf of the insured all sums which the insured shall become legally obligated to pay as damages because of injury to or destruction of property, including the loss of use thereof, caused by accident."

The policies issued after 1967 were "occurrence-based" policies. They provided that "[t]he company will pay on behalf of the insured all sums which the insured shall become legally obligated to pay as damages because of . . . property damage to which this insurance applies, caused by an occurrence." Prior to 1973, the occurrence-based policies defined "occurrence" to mean "an accident, including injurious exposure to conditions, which results, during the policy period, in bodily injury or property damage neither expected nor intended from the standpoint of the insured." (Emphasis omitted.) In 1973, the definition was changed slightly to provide that "`occurrence' means an accident, including continuous or repeated exposure to conditions, which results in bodily injury or property damage neither expected nor intended from the standpoint of the insured." (Emphasis omitted.) The trial court noted, and the parties do not dispute on appeal, that "since an `occurrence' is defined in terms of an `accident' . . ., it is apparent that to obtain coverage under either the accident-based or the occurrence-based policies injury must have been caused by an accident."

ENGI brought this declaratory judgment action to obtain coverage for the costs of cleaning up the site. Ruling on cross-motions for summary judgment, the trial court ruled that "none of the policies in question was triggered because the contamination was not accidental," and accordingly granted summary judgment in favor of Continental.

On appeal, ENGI argues that the trial court erred in: (1) failing to find that the term "accident" is ambiguous and should therefore be construed in ENGI's favor; (2) impermissibly expanding the inherently injurious analysis developed in Vermont Mutual Insurance Co. v. Malcolm, 128 N.H. 521 (1986); (3) failing to consider, in conducting the inherently injurious analysis, what was in the mind of the insured; and (4) granting summary judgment when there were material facts in dispute.

[1-3] We review the trial court's interpretation of the policies, including its conclusion as to whether a policy term is or is not ambiguous, de novo. See Hudson v. Farm Family Mut. Ins. Co., 142 N.H. 144, 147 (1997). We generally construe an insurance policy as we do any other contract. See id. at 146. Where an insurance policy's language is ambiguous, however, and one reasonable interpretation favors coverage, we construe the policy in the insured's favor, and against the insurer. See id. "We will not, however, force an ambiguity simply to resolve it against an insurer." Brouillard v. Prudential Prop. Cas. Ins. Co., 141 N.H. 710, 712 (1997) (quotation omitted).

ENGI first argues that the trial court erred in not finding the term "accident" ambiguous and construing it in favor of coverage. Had the trial court done so, ENGI argues, it would have been unnecessary to conduct an inherently injurious analysis because coverage would already have been found. ENGI contends that High Country Assocs. v. New Hampshire Insurance Co., 139 N.H. 39 (1994), established that the term "accident" is ambiguous under the circumstances presented here. In High Country, we recognized a reasonable disagreement as to whether the term "accident" as used in the policy at issue required a sudden event. Id. at 43-44. ENGI argues that the term is similarly ambiguous in this case because the damage occurred gradually and "the temporal meaning of `accident' is unclear." ENGI also contends that because no inherently injurious analysis was performed in High Country, that analysis is inappropriate whenever the term "accident" is found, as a threshold matter, to be ambiguous.

ENGI's argument misconstrues our rules on construction of ambiguous policy terms. Coverage is not mandated whenever a term in the abstract admits of more than one meaning. Policy language is ambiguous where "the language of the policy reasonably may be interpreted more than one way and one interpretation favors coverage." Id. at 41. Once such an ambiguity is established, the court will apply the reasonable interpretation that favors coverage. See Trombly v. Blue Cross/Blue Shield, 120 N.H. 764, 771 (1980).

Thus, even if ENGI were correct that the "temporal meaning" of the term "accident" is ambiguous, we would do no more than give the term an interpretation that, in the temporal context, favors ENGI. We have already given the term such an interpretation in High Country, where we construed "accident" "to mean circumstances, not necessarily a sudden and identifiable event, that were unexpected or unintended from the standpoint of the insured." High Country, 139 N.H. at 44. That construction does not, however, require us to ignore other requirements implicit in the term and find coverage in any event. We note that the trial court did interpret the term "accident" to cover gradual damage, a holding not challenged on appeal.

We also point out that an inherently injurious analysis was not at issue in High Country, presumably because there was no intentional act that could reasonably have been construed to be inherently injurious. High Country involved damage caused by the insured's negligent construction of condominium units. See High Country, 139 N.H. at 41. High Country does not stand for the proposition that ambiguity in the term "accident" automatically renders the inherently injurious analysis unnecessary.

[5,6] We also note that "a term in a contract already clearly defined by judicial decision cannot be considered ambiguous." Coakley v. Maine Bonding Cas. Co., 136 N.H. 402, 410 (1992). "[W]here judicial precedent clearly defines a term at issue, we need look no further than that definition." Id. at 409-10. Thus, the trial court did not err in applying the definition of accident we adopted in Vermont Mutual: "An accident is an undesigned contingency, a happening by chance, something out of the usual course of things, unusual, fortuitous, not anticipated, and not naturally to be expected." Vermont Mutual, 128 N.H. at 523 (quotation, brackets and ellipses omitted).

ENGI next challenges the trial court's application of the inherently injurious analysis developed in Vermont Mutual. First, ENGI contends that public policy in New Hampshire favors coverage for even intentional torts, see American Home Assurance Co. v. Fish, 122 N.H. 711, 715 (1982), and that the inherently injurious analysis carves out only a narrow exception to that general rule for "behavior that is so egregious it should not be entitled to insurance coverage," such as the child sexual abuse at issue in Vermont Mutual. ENGI argues that the trial court impermissibly broadened the inherently injurious analysis by applying it here to behavior that, at the time it took place, was both lawful and conformed to widely accepted industry standards and practices.

We reject ENGI's categorization of our inherently injurious cases, which have dealt with claims such as breach of a residential purchase and sale agreement, see Fisher v. Fitchburg Mut. Ins. Co., 131 N.H. 769 (1989), and encroachment of a newly constructed home onto neighboring property, see A.B.C. Builders v. American Mut. Ins. Co., 139 N.H. 745 (1995). Nor do we agree with ENGI that a meaningful distinction can be drawn on the basis that our inherently injurious cases have involved some allegation of wrongful conduct by the insured, while ENGI's underlying exposure here is based on strict liability. Our concern in conducting the inherently injurious analysis is not with the nature of the insured's underlying liability but with the intentional character of the insured's conduct and its injurious natural consequences. Cf. Jesperson v. United States Fidelity Guaranty Co., 131 N.H. 257, 261 (1988) (implying that even the ultimate determination of the insureds' liability is irrelevant to the inherently injurious analysis). Thus, the trial court did not err in applying the inherently injurious analysis to the conduct at issue here.

ENGI next contends that the trial court erred in conducting the inherently injurious analysis from the standpoint of a reasonable person in the position of the insured rather than from the standpoint of the insured itself. ENGI argues that the trial court's approach conflicts with the both the language of the occurrence-based policies and our caselaw. ENGI notes that the occurrence-based policies require that the damage be "neither expected nor intended from the standpoint of the insured." The inherently injurious analysis, however, is based on the interpretation of the policy term "accident," not the portion of the definition of "occurrence" quoted by ENGI. While the language quoted by ENGI focuses on the damage, the judicially supplied definition of accident focuses on the "cause of injury, as distinct from the injury itself." Vermont Mutual, 128 N.H. at 523. We therefore reject ENGI's policy language argument.

ENGI also argues that our caselaw requires the inherently injurious analysis to be conducted from the standpoint of the insured. In Vermont Mutual, we held that in examining whether an injury-causing act was accidental, "the fortuity should be determined by reference to the insured." Id. at 523. We implicitly incorporated this holding into our definition of "accident" in High Country: "`[A]ccident' in the definition of `occurrence' . . . mean[s] circumstances, not necessarily a sudden and identifiable event, that were unexpected or unintended from the standpoint of the insured." High Country, 139 N.H. at 44.

We reject ENGI's contention, however, that this perspective is based solely on the insured's subjective state of mind. Rather, we have developed two inquiries to determine whether an insured's act was an accidental cause of injury, one subjective, the other objective. These inquiries are based on the two types of acts we recognized as non-accidental in Vermont Mutual and its progeny. First, we held that "an insured's act is not an accidental contributing cause of injury when the insured actually intended to cause the injury that results." Vermont Mutual, 128 N.H. at 523 (emphasis added). Thus, the first inquiry is subjective, looking to the actual intent of the insured. The trial court found it unnecessary to conduct this inquiry, as it found Continental had not claimed that ENGI intended to cause environmental injury.

Second, we held that an insured's intentional act cannot be accidental when it is so inherently injurious that "it cannot be performed without a certainty that some injury will result." Providence Mut. Fire Ins. Co. v. Scanlon, 138 N.H. 301, 306 (1994). In this inquiry, the insured's intent is irrelevant. Id. Thus, the "inherently injurious" test is objective, and should be conducted from the standpoint of a reasonable person in the position of the insured. See Fisher, 131 N.H. at 773; see also N.H. Ball Bearings v. Aetna Cas. Sur. Co., 43 F.3d 749, 753-54 (1st Cir. 1995).

ENGI argues that the reasonable person test erroneously employs a tort standard. We disagree. The reasonable person standard is simply the hallmark of an objective inquiry and is used in contexts ranging from the establishment of probable cause, see, e.g., State v. Ball, 124 N.H. 226, 235 (1983), the determination of whether a person is seized for constitutional purposes, see, e.g., State v. Cote, 129 N.H. 358, 364-65 (1987), contract interpretation, see, e.g., Smith v. Liberty Mut. Ins. Co., 130 N.H. 117, 124 (1987), and recusal of judges, see, e.g., Taylor-Boren v. Isaac, 143 N.H. 261, 268 (1998). Thus, we hold that the trial court did not err in applying the following inherently injurious test: "[W]ould a reasonable company in . . . [ENGI's] position know that its intentional dumping of tar and other byproducts contained in its waste stream was certain to result in some injury to property, although not necessarily the particular injury to the groundwater, surface water and sediment." (Quotation and brackets omitted.)

ENGI further argues that the trial court's inherently injurious analysis constituted a premature evaluation of the "expected or intended" defense, which had not been briefed or argued by the parties. ENGI concedes, however, that its defense turns on its subjective intent to cause injury. The inquiry is therefore the same as the first in the test for accidental causation. See Scanlon, 138 N.H. at 305. As noted previously, the trial court properly found it unnecessary to address this issue.

ENGI contends that summary judgment was still not warranted because there were genuine issues of material fact in dispute. Most importantly, ENGI contends that the state of the gas industry's knowledge at the relevant time was genuinely disputed, particularly by the parties' experts. We will affirm a trial court's grant of summary judgment if, considering the evidence and all inferences properly drawn therefrom in the light most favorable to the non-movant, "our review of that evidence discloses no genuine issue of material fact, and . . . the moving party is entitled to judgment as a matter of law." Iannelli v. Burger King Corp., 145 N.H. 190, 193 (2000). Thus, we begin our review by examining the evidence presented.

A paper read at the 1899 annual meeting of the New England Association of Gas Engineers, entitled "The Nuisance Question in Gas Works," listed the following as some of the incidents of gas making that had been held to be nuisances: "The pollution of private wells adjacent; the pollution of State streams under the protection of the Fish or Game Commissioners; [and] the pollution of rivers used for the water supply of cities . . . ." The paper's author had previously made clear that a ruling of nuisance required proof of "real injury" to the plaintiff. A committee report on "Disposal of Waste from Gas Plants" presented at the 1920 annual convention of the American Gas Association discussed the gas industry's receipt, "sometimes . . . just[ly] and sometimes unjust[ly]," of part of the blame for the pollution of streams by industrial waste discharge. The report stated:

Although the waste from gas plants may be perfectly clear, it will often contain in solution many substances which are poisonous, and some of the most poisonous are those which are commonly regarded as insoluble in water such as napthalene, benzene, toluene, xylene, etc. Table I shows the relative toxicity of the various products associated with the manufacture of gas. This table is taken from a Bulletin of the Illinois State Laboratory of Natural History on An Experimental Study of the Effects of Gas Waste upon fishes, with especial reference to stream pollution, March, 1917, by Victor E. Shelford, Ph.D. More recent work on this same subject has shown that the concentrations of some of the substances in the table are low. In general, the greater the amount of tar emulsions, the more toxic or poisonous the waste, especially in the case of coal tar.

Also before the trial court were the report and deposition testimony of ENGI's expert, Gonzalo J. Mon. Mon opined that "Gas manufacturers did not know that their operating and disposal practices were causing actual or potential long-term damage to the environment in the sense of how the word environment is used today. In today's usage, damage to the environment is a concept that evolved in the 1970s and 1980s." Mon reiterated that theory in his deposition, stating at one point that MGP operators "were not even aware what the word `pollution' means."

Mon's testimony, however, did not controvert Continental's evidence of industry knowledge. Even if the gas industry was not cognizant of the concept of "environmental damage" as we now understand it, Continental's evidence showed that the industry was aware that contamination of water with MGP waste could cause sufficient "real injury" to support a nuisance action. Continental's evidence also showed the industry's knowledge that the harmful nature of MGP waste was due in part to the toxicity of its chemical constituents. "[A]n act is inherently injurious if it is certain to result in some injury, although not necessarily the particular alleged injury." Scanlon, 138 N.H. at 306. The industry's awareness that its waste could contaminate water to the detriment of its neighbors is sufficient under this test, even if the industry did not understand the term "environment" as it is used today.

ENGI nevertheless argues that Continental did not show that some injury was certain to result from ENGI's actions. At best, ENGI argues, Continental's expert, Dr. Allen Hatheway, could only opine that MGP waste had a propensity to cause harm. As ENGI did not provide us with the cited portion of Dr. Hatheway's report or testimony, we do not know what kind of harm he meant. We note, however, that the contamination of water with chemicals having a potential to cause harm is itself an injury regardless of whether the chemicals are certain to cause the ultimate harm of which they are capable. We have no difficulty concluding that a reasonable company in the position of ENGI at the time in question would have known that its intentional discharge of toxic waste into a body of water was certain to contaminate that body of water, particularly where, as here, that waste was known to be "insoluble in water and tends to deposit on the shores of a stream in the form of a shiny coating." L.J. Willien, Chairman, Disposal of Waste from Gas Plants, Report of the 1920 Committee of the American Gas Association 420 (1920). Thus, we hold that the trial court did not err in finding ENGI's acts to be inherently injurious.

In light of that conclusion, ENGI's other asserted facts in dispute are immaterial. For instance, ENGI argues that the specific knowledge of ENGI's MGP operators was genuinely disputed. Such facts are not material, however, where it has been determined that ENGI's conduct meets the objective inherently injurious test. ENGI also argues that there was a significant dispute over how or whether evidence related to a 1938 hurricane should be used in the inherently injurious analysis. Such evidence, however, which went to the issue of knowledge available to ENGI's MGP operators in Concord, was not necessary to a determination under the objective inherently injurious test, and thus is also immaterial.

Nevertheless, ENGI cites the trial court's emphasis on the 1938 hurricane and argues that there must be a reasonable nexus between the insured's act and the damage for which it seeks coverage. ENGI contends that the trial court found that because ENGI was aware that its waste had actually caused damage during the 1938 hurricane, it should have known that its waste was certain to cause some damage to property. ENGI argues that the damage from the 1938 hurricane was of a different type than that for which it is now being held liable.

We need not decide whether the trial court improperly inferred knowledge from the events of 1938 because we have already held that such knowledge is immaterial to a finding of inherently injurious action. "We will not overrule the trial court's grant of summary judgment where it reached the right result, albeit on mistaken grounds." Laro v. Leisure Acres Mobile Home Park Assocs., 139 N.H. 545, 548 (1995).

Finally, ENGI argues that the trial court's decision ignored "certain bedrock principles" of insurance such as that "insurance is bought and sold as a guard against both known and unknown risk" and that "[i]nsurance protects policyholders against unforeseen liability arising out of even their negligent acts." Thus, ENGI argues that "Continental insured ENGI against the precise risk that there would be a change in public attitude (and government regulation) toward the environment and as a result of that change, legal liability would be imposed on ENGI" for activity that, at the time it was conducted was "lawful and in conformity with accepted business standards." We disagree.

"We emphasize that our answer to each question [presented in this appeal] seeks to give effect to the plain meaning of an insurance contract. These answers do not imply that the contract could not have been different." MacKinnon, 124 N.H. at 462. The policies at issue here did not insure against undefined risks, but against the consequences of accidents. Our caselaw has long defined the term "accident" to exclude inherently injurious acts, notwithstanding what the insured may have subjectively intended or known about the consequences of those acts. Thus, although ENGI's acts may well have been lawful and socially acceptable at the time they were taken, they were not accidents as our cases have defined that term, and that term is the one on which coverage hinges.

Affirmed.

BROCK, C.J., and BRODERICK, NADEAU and DALIANIS, JJ., concurred; NADEAU and DALIANIS, JJ., took part in the final vote by consent of the parties.