This matter comes before the court on the Motion in Limine of the Plaintiffs, S. W. Heischman and University Village, Inc., seeking a ruling on the burden of proof applicable under the facts and circumstances of this case.
The Plaintiffs are owners of a 46-unit condominium building, known as University Village, located at 2401 Old Ivy Road in Albemarle County. During the initial construction period, the building project was covered by a builder’s risk insurance policy issued by Aetna Casualty and Surety Company, with a policy period extending from. July 24, 1990, to July 24, 1991. Subsequently, the University Village building was covered by a “condominium package” insurance policy, issued by the Reliance Insurance Company. The Reliance policy was a multiperil policy as defined in Va. Code §§ 38.2-132 and 38.2-1921 and began on July 1, 1991, continuing to July 1, 1992.
In their Amended Motion for Judgment, Plaintiffs allege that the building’s air conditioning system caused water damage in various condominium units and common areas throughout the building, result*236ing in total damages of $1 million. These damages allegedly occurred over both Aetna’s and Reliance’s policy periods, and Plaintiffs claim that they are therefore entitled to recover under both policies.
In this Motion, Plaintiffs do not seek to avoid the burden of proof regarding the amount of their overall damages. Rather, they assert that they should not bear the burden of proving specifically during which policy period the particular elements of the entire loss might have occurred. They claim instead that the burden of proof should rest with the Aetna and Reliance, as insurers, to prove and decide between themselves during which policy period the respective losses occurred.
For its part, Reliance does not appear to claim the benefit of any exclusions in its policy but rather contends that the Plaintiffs have failed to meet their initial burden of proof regarding coverage. In Reliance’s view, the Plaintiffs have overlooked the most important question, which is whether or not their losses are covered at all under the Reliance policy. Reliance contends that the damage suffered by the Plaintiffs manifested itself before coverage under the Reliance policy began and that the Plaintiffs' losses are thus not covered at all. Consequently, Reliance argues that no apportionment of damages between itself and Aetna is necessary.
Discussion
A brief examination of the Plaintiffs’ Motion reveals a basic but questionable assumption. As the court understands it, the Plaintiffs now seek to place the burden on the two successive insurance carriers, Aetna and Reliance, to apportion liability between them for damages resulting from the defective air conditioning system. Yet shifting the burden of proof in this manner could only be appropriate if coverage exists under both policies at the same time for injuries flowing from a single, continuing cause. If simultaneous coverage for such injuries could not exist, then only one carrier could be responsible, and no apportionment would be necessary.
A ruling on Plaintiffs’ Motion thus requires this court to examine the allocation of indemnity between successive first-party property insurers when the loss is continuous and progressive. This is not a question that has yet been addressed by Virginia courts, however, and its resolution is far from obvious. Essentially, the Plaintiffs are asking the court to adopt the “continuous exposure theory” of coverage in the *237context of first-party property insurance.1 This is a difficult issue, raising many policy concerns, and the lack of Virginia precedent is unfortunate. However, other courts have considered the issue of successive first-party coverage for continuing property damage, and the reasoning they applied will prove instructive in this case.
Landmark
The California courts have been particularly active in addressing the issues now raised by the Plaintiffs. In California Union Insurance Co. v. Landmark Insurance Co., 193 Cal. Rptr. 461 (Ct. App. 1983) (Landmark I), a case on which the Plaintiffs rely, the Second District Court of Appeals considered a situation involving damages resulting from the installation of a leaky swimming pool. There were two insurance companies that had issued policies on the property, one of which was in effect at the time the swimming pool began to cause damage, and the other which was in effect at the time the swimming pool was actually determined to be the cause of the damage. As the Plaintiffs note, Landmark I determined that the damage did constitute “one occurrence” but found that both insurers were jointly and severally liable for the damage anyway. In holding a post-manifestation insurer liable for premanifestation injuries, the Landmark I court apparently created quite a bit of controversy in California. The holding was later limited by the Second District to the distinct facts of the case. Harbor Insurance Co. v. Central National Insurance Co., 211 Cal. Rptr. 902 (1985).
The Fourth District subsequently found that, at least in the context of first-party property insurance, it was the date of manifestation of property damage that determined which policy was triggered, and not the (several) dates at which the insured was exposed to damage. Home Insurance Co. v. Landmark Insurance Co., 253 Cal. Rptr. 277 (1988) (Landmark II). There, two insurance companies insured a hotel against *238property damage for successive policy periods. The damage was initially discovered during the first of the two policy periods and continued on through the second. Apparently, it was impossible to determine the extent of damage occurring during each period, and the amount owed by each insurer could not be determined. In a declaratory judgment action, the trial court held that, since the injuries had manifested themselves during Home’s policy period, it was solely at risk, and no apportionment was necessary. The decision was affirmed on appeal, and the Landmark II court concluded that a property insurer whose policy is in effect at the time damage is first discovered is solely liable for the entire claim, including those damages which continue after its policy expires. Id., at 280-91. The court noted, however, that its decision was limited, as it had failed to distinguish between first-party and third-party claims. Id., at 281, n. 3.
Prudential-LMI
The first-party, successive coverage issue was finally settled, in California at least, in Prudential-LMI Commercial Insurance v. Superior Court, 798 P.2d 1230 (1990). There, an apartment building owned by the insured sustained continuing and progressive property damage over a period of years. The insured sued four successive companies, including Prudential, which had insured the building over the period in which the damage was sustained. Prudential’s coverage of the insured had been terminated five years before the damage became manifest, however, and Prudential argued that its policy was therefore not triggered by the insured’s property damage. Deciding in favor of Prudential and against the insured, die California Supreme Court held that only the carrier insuring the property at the time of manifestation of property damage is responsible for indemnification. Id. at 1232. That carrier will, however, be responsible for all continuing property damage flowing from that initial manifestation, even after the policy expires. Id. As a policy matter, the Prudential-LMI court reasoned that a manifestation rule is more appropriate for first-party property insurance claims because those claims are less uncertain and more capable of estimation than those in the third-party liability or health insurance contexts. The insurer knows with reasonable accuracy the value of the property under coverage during the policy period and is therefore better able to assess potential exposure. Accordingly, the court noted that the manifestation rule in the first-party context will promote certainty *239in the insurance industry and allow insurers to gauge premiums with greater accuracy. Id. at 1246. In addition, consumers would presumably benefit because insurers would be better able to set aside reserves for well-defined coverages and avoid increasing reserves to cover potential financial losses caused by uncertainty in the definition of coverage. Id.
Furthermore, the court noted that a failure to apply the manifestation rule in the first-party context would violate the “loss-in-progress” rule, a fundamental principle of insurance law which states that an insurer cannot insure against a loss that is known or apparent to the insured. Id. at 1243-44. See also, Bartholomew v. Appalachian Insurance Co., 655 F.2d 27, 28-29 (1st Cir. 1981). Once a loss is manifested, the court noted that it is no longer contingent, and the policy coverage currently in effect is triggered immediately. Moreover, that coverage will continue, even for resulting damages which occur beyond the initial policy period. Insurers whose policy terms commence after initial manifestation cannot therefore be found responsible for any potential claim resulting from previously discovered and manifested loss. Id. at 1246.
Application under Virginia Law
Although the ruling in Prudential-LMI is not dispositive of the matter now before the court, the reasoning adopted by the California Supreme Court in that case is persuasive. The policy grounds cited by the Prudential-LMI court are just as persuasive in Virginia as it is in California. It is clear that the interests of both consumers and insurers will be furthered if insurers are allowed to accurately estimate their liability, and such estimates are much easier to make under a manifestation rule. If an exposure rule were adopted, an insurer could find itself liable for damages discovered years after its policy with an insured expired. While this might be a necessary evil for third-party liability or health coverage, such is not the case in the first-party property insurance context. There, an insured knows its insurance needs with a greater degree of certainty and can always look to the insurer on risk at the time of manifestation to cover the entire loss. There is no need to include either predecessor or successor insurers to expand coverage.
*240 Fortuity Doctrine
In addition, Virginia has adopted the so-called “fortuity doctrine,” which is entirely consistent with the “loss-in-progress” rule relied upon by the Prudential-LMI court. Under the fortuity doctrine, an insurance policy cannot cover events which are either known to or intended by the insured and are thus not fortuitous. A fortuitous loss has been defined as one which is dependent on chance so far as the parties are aware. Insurance Co. of North America v. United States Gypsum Co., 678 F. Supp. 138, 141 (W.D. Va. 1988).
The fortuity doctrine is actually implicit in the statute authorizing the commercial multi-peril policy issued to University Village by Reliance. Va. Code § 38.2-132. This section authorizes insurers to issue policies “insuring risks incident to a commercial enterprise” (emphasis added). The General Assembly’s use of the term “risks” implies that they did not intend to allow an insurer to issue policies covering losses that were certain or known to either party.
Furthermore, the Supreme Court has recently applied the “fortuity doctrine” directly to an analysis of a commercial multi-peril policy. Fidelity and Guaranty Insurance Underwriters, Inc. v. Allied Realty Co., 238 Va. 458 (1989). There, the insured had suffered damage to its warehouse walls due to earth movement around its foundation. The insurer argued that the loss was not fortuitous but instead was caused by the insured’s own actions in filling in the area behind the warehouse wall. Acknowledging the vitality of the fortuity doctrine in Virginia, the court nevertheless held that the insured’s loss was fortuitous since it did not know with certainty that the fill it placed near the walls would actually cause any damage.
Loss-in-Progress
Although it has clearly adopted the fortuity doctrine, the Virginia Supreme Court has not yet had the opportunity to address the “loss-in-progress” rule. The “loss-in-progress” rule, which the Prudential-LMI court heavily relied upon, is closely related to the fortuity doctrine. Insurance Co. of North America v. U.S. Gypsum Co., 870 F.2d 148, 152 (4th Cir. 1989). Indeed, it would appear dial in adopting the latter, the Virginia Supreme Court implicitly embraced the former, as well. Where the fortuity doctrine precludes recovery for all losses either known to or intended by the insured, the “loss-in-progress” rule bars only those losses which have already begun and are therefore *241known to the insured. Any claim barred by the “loss-in-progress” rule must therefore also be denied as non-fortuitous.
In addition, the sound public policy rule underlying the “loss-in-progress” is clear and undeniable. To hold an insurer liable for a progressive and continuing loss which was discovered before the insurer undertook the risk to the property would be to impose upon the insurer a guaranty of the good quality of the property insured, a liability the insurer never assumed. Greene v. Cheetkam, 293 F.2d 933, 937 (2d Cir. 1961). Accordingly, the court finds that the “loss-in-progress” rule would be followed by the Supreme Court and that it applies to the transaction in question as a matter of law.
Manifestation
In light of the relevant policy concerns and in order to comply with the “loss-in-progress” rule, this court must therefore rule that it is the manifestation of injury, and not the continuous exposure to damage, that is the relevant trigger in the first-party property damage claim made by the Plaintiffs. “Manifestation of the loss,” as used to determine which of several successive insurers is solely responsible for indemnification in the first-party progressive property damage context, is “that point in time when appreciable damage occurs and is or should be known to the insured, such that a reasonable insured would be aware that notification duty under the policy has been triggered.” Prudential-LMI, 798 P.2d at 1247. In most cases, the time of manifestation is an issue of fact for the jury to decide, although summary judgment may be appropriate where the evidence supports only one conclusion. Id.
Conclusion
It is well-established that a plaintiff in any action bears the burden of proof regarding all the elements of her claim. In an action on an insurance policy, this means that the insured bears the burden of proving that the loss occurred while the policy was in force and in effect. Aetna Casualty and Surety Co. v. Goldman, 217 Va. 419, 420 (1976). To recover under a property insurance policy, such as the one now in question, an insured must therefore make out an initial showing that the loss for which she is seeking compenstion actually occurred during the effective policy period. Bituminous Casualty Corp. v. Sheets, 239 Va. 332, 336 (1990).
*242In light of the above discussion, this court concludes that the Plaintiffs must therefore establish when damages resulting from the defective air conditioning system first manifested themselves. Once they have made this showing, the question of coverage will be clear. If manifestation occurred prior to the inception of the Reliance policy, Aetna will be solely liable for all resulting damages, even those continuing into Reliance’s coverage period. If, on the other hand, the evidence shows that manifestation was delayed until after the Reliance policy took effect, then Reliance will be solely accountable. In neither event, however, will both carriers be held liable for the same continuing injury. Accordingly, there is no need to apportion damages between carriers, and the Plaintiffs’ motion is denied.
Motion to Compel
A second matter comes before the court on Plaintiffs’ Motion to Compel Production of Documents which the Defendant asserts are protected by either the attorney-client or the work-product privilege. In Virginia, Rule 4:1(b) governs tiie scope of discoverable information and authorizes discovery of any matter which is not privileged and which is relevant to the subject matter of the pending action.
The Defendant has produced a number of documents for the court’s in camera review and argues that these documents are, in whole or in part, protected by one of the two privileges in question. Granting or denying a request for discovery is a matter within the trial court's discretion. Rakes v. Fulcher, 210 Va. 542 (1970). Before considering the specific issues raised by the parties, the court will briefly review the applicable Virginia law.
Attorney-Client Privilege
Confidential communications between an attorney and his client made because of that relationship and concerning the subject matter of the attorney’s employment are privileged from disclosure, “even for the purpose of administering justice.” Grant v. Harris, 116 Va. 642, 648 (1914). The privilege attaches to communications of the client made to the attorney’s agents, when such agent’s services are indispensable to the effective representation of the client. Commonwealth v. Edwards, 235 Va. 499, 508 (1988). However, the privilege is an exception to the general duty to disclose, is an obstacle to the investiga-, tion of the truth, and should be strictly construed. Id. As in most other *243jurisdictions, the proponent of a claim that certain documents are privileged has the burden of establishing that the privilege exists. Id. at 509.
Work-Product Privilege
The traditional work-product doctrine has been adopted in Virginia under Rule 4:1(b)(3), which reproduces Federal Rule of Civil Procedure 26(b)(3). Thus, while materials prepared “in anticipation of litigation” will be granted a qualified immunity from disclosure, those compiled in the ordinary course of business will not. State Farm Fire and Casualty Co. v. Perrigan, 102 F.R.D. 235, 238 (W.D. Va. 1984). As with the attorney-client privilege, the party asserting the privilege bears the burden of showing that it applies. The general purpose of this rule is to encourage prompt and thorough preparation by the attorneys and to avoid unfairness. Leigh B. Middleditch, Jr., and Kent Sinclair, Virginia Civil Procedure, § 12.4 (1992).
The difficult issue in many work-product cases, including the case at bar, is to determine when litigation was actually anticipated by the parties. Since the work-product doctrine is intended to grant an attorney the freedom to prepare a case without worrying about improper disclosure, there is no need to apply it where litigation is not yet imminent. This question is particularly troublesome in the insurance context, where the nature of the business requires a routine investigation prior to the determination of any claim. The Fourth Circuit has held that materials prepared by an agent of an insurance company during such a routine examination are made in the “ordinary course of business,” and not in anticipation of trial, and are therefore discoverable. McDougall v. Dunn, 468 F.2d 468, 473 (4th Cir. 1972).
Some federal courts have adopted what amounts to a bright-line rule, applying the work-product privilege only to those documents prepared after the insurance company has decided to deny the claim. Westhemeco, Ltd. v. New Hampshire Insurance Co., 84 F.R.D. 702, 708-09 (S.D. N.Y. 1979). Others have adopted a somewhat looser test, which asks:
Whether, in light of the nature of the document and the factual situation in the particular case, the document can fairly be said to have been prepared or obtained because of the prospect of litigation. But the converse of this is that even though litigation is already in prospect, there is no work product im*244munity for documents prepared in the regular course of business, rather than for purposes of litigation.
Perrigan, 102 F.R.D. at 238. This latter approach has been said to recognize that at some point, an insurance company shifts its activity from the ordinary course of business to pending litigation, and that no hard and fast rule governs when this change occurs. Id. The critical moment is the point “where the probability of litigating the claim is substantial and imminent.” Id. Accordingly, the decision to extend or deny the work-product privilege to insurance investigation reports will depend on the facts of each case. Id.
Application
The court has examined the documents provided by Reliance for in camera review. As to the documents included in Attachment A, the court has indicated which it has found to be discoverable based on the case law discussed above. There are several documents in Attachment A for which the court has not been able to make a finding, largely because insufficient information has been provided by Reliance. To make a proper determination regarding these documents, the court must know: (1) who prepared the document; (2) when the documents were prepared; and (3) what the items in the documents refer to. As Reliance bears the burden of persuasion on this issue, the documents will be discoverable unless Reliance can prove sufficient information to allow the court to rule otherwise.
The document numbered 0485 by Reliance, the letter from Barger Insurance dated March 26, 1992, is found not privileged. Although nominally an insurance “agent,” Mr. Meador was clearly not acting as the agent of Reliance at the time the letter was written. Rather, he was writing on behalf of the insured, trying to obtain coverage. Accordingly, the work-product privilege is inapplicable.
Regarding the documents included in Attachment B, the court has indicated which items it considers subject to the work-product privilege. The items which the court has included in the category “Documents Ordered Produced” are to be produced in their entirety.
The “continuous exposure theory” apportions liability among all insurance carriers whose policies were in effect from the time damage first occurred to the date it was first discovered by the insured. This approach was first adopted in Gruol Construction Co. v. Insurance Co. of North America, 524 P.2d 427 (Wash. App. 1974), a case involving a third-party claim for a construction defect. More recently, the continuous exposure theory has found application in numerous asbestos cases involving recovery for bodily injury. See, e.g., Keen Corp. v. Insurance Co. of North America, 667 F.2d 1034 (D.C. Cir. 1981); North America v. Forty-Eight Insulations, 633 F.2d 1212 (6th Cir. 1980).