Seele Austria GmbH & Co KG v Tokio Marine Europe Insurance Ltd

This is an appeal from the decision of Field J handed down on 21^st June 2006 whereby he answered certain questions in relation to the proper construction of a Combined Contract Works and Third Party Liability policy (the policy). That policy had been taken out in relation to a building project at St Martins Court, 3 Paternoster Square. Under the policy there were three categories of insured. The third category included the trade subcontractors amongst whom was the appellant. Only Section 1 of the policy, insuring the contract works, applied to that third category (Insurer (3)). The appellant had installed “punched windows” which had been found to be defective. Costs were incurred in gaining access to the windows which had to be taken out and reinstalled, and the main issue (decided by the judge in favour of the insurers) was and is whether those costs can be recovered by the appellant under the policy.

The facts are uncontested and I can take them from the judge’s judgment.

I can also take the description of the policy and the identification of the relevant terms from the judge’s judgment:-

The key issue before us was whether, even though no accidental damage had occurred as a result of the installation of the defective windows, Memorandum 18(3) was a free standing term allowing the appellant, once a defect had been discovered, to recover access costs to enable the defective windows to be replaced. Despite the judge’s ultimate view (when considering whether to grant permission to appeal) that once properly understood the position was plain, argument before us demonstrated that the terms relevant to that key issue are not easy to interpret. It is no criticism to say that the difficulty was demonstrated by some shifting in the submissions on construction by both Mr Reed for the insurers and Mr Williamson QC for the appellant during the argument.

By the end of the arguments before us it was, I believe, reasonably clear where the differences between the parties lay. Mr Williamson’s final position was simple and straightforward. His submission was that the situation in which a defect had been discovered but no accidental damage had taken place took one directly, via the insuring clause of section 1, to Memorandum 18(3). This (he submitted) was a policy which covered accidental damage to property but also covered non-accidental damage where the circumstances fell within Memorandum 18(3) i.e. intentional access damage where a defect had been discovered. On his submission insured property was the whole of the insured property and not limited to the part of the contract works for which the appellant was responsible. His submission was that this was a perfectly sensible contractual arrangement for the insurers to enter into, since they might well wish to give cover for such access costs so as to prevent far worse incidents occurring as a result of a defect. Of course, if his submission was right, it not only enabled a subcontractor to gain indemnity against access costs when the defect posed a risk of a serious accident but to gain indemnity (subject to the deductible) against access costs even if there was no risk of that kind at all.

Mr Reed’s submission for the insurers was to emphasise the nature of section 1 of the policy. That section provided property damage cover. He submitted that for the appellants to have any claim under the policy there must be loss or damage to the insured property and he submitted that that must be accidental damage not deliberate damage. All risk property damage policies were in normal circumstances insuring against accidents, not wilful damage by the insured. [See Professor Clarke, The Law of Insurance Contracts 4^th Edition para 17.3] As a matter of law, and in any event as expressly provided by this policy, a defect in the works did not constitute damage. So, he submitted, since there was in this case no accidental damage, there was no trigger to which the policy had to respond at all. That was, he submitted, the end of the matter.

It was however important to his submissions to demonstrate how the policy would have worked if there had been accidental damage, it being part of his submission that it was sensible for the insurers to provide, as they did by Memorandum 18(3), for indemnity against intentional damage flowing from the existence of a defect once there had been an accident caused by the defect. That made sense, he submitted, by virtue of the fact that the danger of the defect producing other damage would have been demonstrated. That was to be contrasted with an unlikely undertaking on the part of insurers to enable a subcontractor to gain access to every piece of defective work even if the likelihood of it causing further damage was remote or non-existent.

He had alternative submissions as to the proper construction of the cover being offered to trade subcontractors. His primary submission was that Memorandum 15 was a complete definition of the cover given to insured (3). It rested very much on the words at the end of the description of insured (3) “as to the extent stated in Memorandum 15”, and on the words in Memorandum 15 which referred to “the indemnity provided … to insured (3) in respect of any works … shall be restricted…” In other words, he submitted, cover only existed (1) where damage had been caused to a subcontractor’s works, (2) by a defect in the subcontractors' work and, he submitted, (3) the cover was restricted to an insured (3)’s works. Thus on this construction there was no cover if the subcontractors works had been destroyed by fire not due to any defect. There was further no cover for subcontractor A if the defect in subcontractor A’s works destroyed other subcontractors works.

In the alternative he argued that insured (3) had cover by virtue of section 1, for events other than through defects in the subcontractor’s works, and Memorandum 15 simply restricted cover where the circumstances described within it existed, i.e. (1) there was damage to insured (3)’s works and (2) that damage was caused by a defect in design or workmanship. Memorandum 15 then restricted the indemnity provided by section 1 to insured (3) to the extent provided by Memorandum 18(2) and (3). Mr Reed would still submit that by its language Memorandum 15 restricted the indemnity to the works of the subcontractor. He would further submit, the loss or damage must be accidental damage caused by the defect and once again the defect would not itself be damage.

What then if damage was caused to insured property other than the works for which the subcontractor was responsible, through a defect in the work of the subcontractor? Mr Reed would submit that such damage would not be covered. He would suggest the language of Memorandum 15 provided an indemnity only in respect of any works carried out by insured (3). He would further suggest that although the subcontractor or main contractor whose works had been damaged might be covered, the policy would leave the insurers free to sue the subcontractor on a subrogated claim. In the alternative he submitted there would be cover simply by virtue of section 1 of the policy and Memorandum 18(1) would apply in such a situation.

Mr Reed’s submission on Memorandum 18 was that the opening words “This Section includes loss or damage arising out of a defect . . .” still contemplated accidental damage. But, he submitted, recovery for accidental damage caused by a defect in design would be limited in accordance with the terms of Memorandum 18. Memorandum 18(1), he submitted, was dealing with accidental damage which is covered directly under section 1, i.e. where the circumstances in Memorandum 15 did not apply. In that instance the insured must apply 18(1) save where, under the final paragraph of 18, he has an option to apply (2) and (3). In applying 18(1) he will not recover the cost of repairing or replacing “any insured property which is defective” or “damage to the insured property caused to enable replacement repair etc” (access costs) unless damage to the insured property (other than access damage) results from the defect. In that event what cannot be recovered are costs of improving the design or workmanship to eliminate any defect. Thus, submitted Mr Reed, in this instance where the defective property is damaged accidentally but as the result of a defect, the insured recovers the costs of replacement (without betterment) and access costs.

He makes two points flowing from that construction, first the breadth of the cover compared to applying 18(2) and (3), and second that access costs are recovered in this instance and thus (3) is not necessary as an addition to this cover.

Coming then to 18(2) and 18(3) Mr Reed submitted that these apply only to a situation in which Memorandum 15 has brought them into play unless the option under the last paragraph of 18 applies “their wording” to a claim to which Memorandum 15 does not apply. Where Memorandum 15 applies there has been (1) damage to the subcontractor’s works and (2) a defect in design. In this situation, Mr Reed submitted, that (a) the subcontractor is not covered for costs of replacing the property in defective condition or (b) for access costs unless some part of the “insured property free of the defective condition . . . is unintentionally damaged”, in which event he gets (a) the costs of replacing that damaged property and “additionally” (as he would emphasise) under Memorandum 18(3) “intentional damage”, i.e. the access costs to enable the defective work to be repaired.

In Mr Reed’s submission the trigger to recovery of replacement costs and access costs is damage to non-defective property and he submits that will only apply if the damage is to the subcontractors’ works and not more generally.

I am not sure that I follow all of the judge’s reasoning for ruling in favour of the insurers. For example he uses the phrase ‘unintentional damage’ at times when I would prefer to think in terms of ‘accidental damage’ so as to draw a distinction between two quite separate aspects of Mr Reed’s submissions. Mr Reed’s primary submission is that this policy only responds if there has been accidental damage to property; absent that accidental damage, he submits there simply is no claim. He has a separate point which is that, even if there was accidental damage, 18(3) is additional to and in effect part of 18(2) and thus not free-standing.

However I do agree with the judge that the insurers were entitled to succeed, but will seek to explain how I reach that result in my own words. I cannot accept all Mr Reed’s submissions and in particular I do not accept his primary way of putting his case, on the basis that Memorandum 15 defined the cover available to insured (3). Nor can I accept that in Memorandum 15 the insurers were imposing a restriction on the indemnity by reference to the insured (3) works. This is an important point in construing the policy. It is true that one could read the language as language of restriction in that it states that “the indemnity provided by this section of the policy to insured (3) in respect of any works …” But the restriction intended is, in my view, identified by the final words “restricted to the extent provided by Memorandum 18(2) and (3)” Section 1 of the policy would seem to me naturally to cover insured (3) for any damage to his works not caused by a defect and for damage to other works even if caused by a defect. It is unnecessary and makes little commercial sense to read the language of Memorandum 15 as confining recovery to damage to insured (3)’s works or as denying cover to insured (3) where there is no defect. What is more when one comes to consider 18(2) and 18(3) where insured property is used as if it means the whole works, it makes more sense that Memorandum 15 is not read so as to confine any right of indemnity to damage to the subcontract works.

But whether I am right or Mr Reed is right on the points where I have identified disagreement with his submissions, I do accept that Memorandum 18(3) was not intended as the free-standing bargain for which Mr Williamson contended.

Mr Williamson’s contention involves arguing first that there needs to be no accidental damage before a subcontractor can open up works to discover defects and charge insurers. He has to say Memorandum 15 and thus Memorandum 18(2) have no application to such a situation. He did not, I think, explain quite how Memorandum 15 worked if he were right. Presumably, on his argument, it did not apply where there was intentional damage “due to a defect in design”. I do not think it can have been intended that Memorandum 15 would be by-passed in this way. In my view the proper construction of this policy was to cover against accidental damage to the insured property including subcontractors’ works. It included cover for accidents that resulted from defects in design or workmanship, subject to the limitations imposed by Memorandum 18(1), or where the circumstances stated in Memorandum 15 applied the limitations in Memorandum 18(2) and (3).

Where Memorandum 15 did not apply, and thus 18(1) was applicable, if the accident had damaged any insured property, including the defective property, the non-recovery was limited to the costs of improving the design or workmanship. If thus there was damage to defective property, under Memorandum 18(1) the insured did recover both replacement costs of the defective property and the access costs.

Memorandum 15 was to apply when damage has been done to a subcontractor’s own works through a defect in workmanship or design. A restriction is then placed on recovery by reference to Memorandum 18(2) and (3). In my view they provide that, if the damage is simply to the property in defective condition, the subcontractor does not get the costs of replacement or repair or access costs. If however there is damage to a non-defective part of the insured property, then the insured gets the costs of replacing the non-defective property and, additionally, access costs to enable the defective property to be replaced.

My reasons for taking the above view are primarily these. (a) It makes sense for insurers to be prepared to pay access costs once there has been an accident going beyond the property which is in defective condition – the risk of further damage is clear; it makes little sense for insurers to be prepared to pay all trade subcontractors for opening up works to repair defects, whether those defects were likely to cause any further damage or not; (b) if 18(3) was to be construed as free standing, it would have to apply to all insured whether insured (3) or insured (1). But insured (1) can recover access costs under Memorandum 18(1) and in certain circumstances have an option to apply 18(2) and (3); it is not, thus, “free-standing” as regards insureds (1) and (2) and it would be odd if it were free-standing for insured (3) only; (c) if 18(3) is in some way free-standing, even in situations in which Memorandum 15 has brought 18(2) and (3) into play, 18(3) would simply mean that 18(2)(b) never applied. 18(2)(b) is intended to prevent recovery of access costs in a circumstance where there is no damage to non-defective property but, if the appellant is right, that can be ignored and use made of 18(3) even where there is no damage to non-defective property.

I should also explain a little further why I reject Mr Reed’s submission that insured (3)’s right of indemnity is limited to damage to insured (3)’s property. I have already indicated my view as to whether Memorandum 15 would restrict indemnity to cover loss or damage to insured (3) works. It does not. It is damage to that property which brings Memorandum 15 into play but the restriction on coverage is found in Memorandum 18(2) and (3). If it were suggested that by defining some insureds as category (3) insureds it should be implied that the meaning of insured property becomes limited to the property of insured (3), I would reject that suggestion. That might be so if insured (3) had no insurable interest beyond his own works, but that is not so. It would not have been contemplated that this form of insurance should be construed in a way which allowed insurers to bring subrogated claims as between co-insureds arising out of damage to “the works”. Mr Reed submitted that placing the insureds in three categories pointed to an opposite conclusion but in my view it does not, because disputes as to whose defective work had caused what damage could arise as between any or all of the trade subcontractors falling within the insured (3) category. I would thus construe insured property (a) or the insured property as applying to “The whole works etc..” as set out in (a) under the heading Insured Property, and not simply to the subcontractor’s works wherever the words insured property or insured property(a) appear, and in particular in Memorandum 18(3). When applying Memorandum 18(2) insured property (a) will also have its full width although it may well be that in situations in which (2) (a) and (b) prevent recovery it will be because damage has only been suffered by the subcontractor’s property.

In the result I take the view that, because there was no accidental damage in this case, the appellant has no right to indemnity under Memorandum 18(3).

I would thus dismiss the appeal.

I have read the judgments of Moore-Bick and Richards LJJ in draft. The point of construction is a difficult one, but I am not persuaded to change my view. As regards the questions posed and the answers dealt with by Moore-Bick LJ at paragraph 58, it may help if I indicate my view on those issues. As regards Issue 1, I would answer the question ‘No’ and thus not answer Issue 2. I would answer Issues 3 to 7 as he does. I would answer Issue 8 ‘Yes’ and would not on that basis answer Issues 9 to 13.

The circumstances giving rise to this appeal are not in dispute. They are described in the passages of the judgment of Field J. to which Waller L.J. has referred and I need not repeat them.

The appellant, Seele Austria G.m.b.H. & Co. K.G. (“Seele”), was employed as a sub-contractor by the main contractor, BLS St Martin’s Ltd (“BLS”), to design, manufacture and install punched windows forming part of the external curtain walling of St. Martin’s Court, Paternoster Square. The contract required Seele to maintain insurance to cover its liability in respect of death or personal injury and in respect of damage to property, except for damage to the sub-contract works themselves. BLS undertook to insure the sub-contract works and all works in the course of execution. However, the contract did not oblige BLS to effect insurance on the whole of the works on behalf of all those interested in them as co-insureds.

The policy of insurance in this case is described as a ‘Combined Contract Works and Third Party Liability Insurance’. Although reference has already been made to its material terms, it is convenient to set them out at this point. The Operative Clause provided that

The Schedule to the policy identified the insured as

The Schedule did not identify separate sections of cover other than by reference to the sums insured and limits of indemnity and excesses. It is clear from those provisions, however, that Section 1 provided cover in respect of the contract works as a whole. The Specification to Section 1 contained the following provision:

There followed a number of Memoranda to Section 1, only two of which, Memorandum 15 and Memorandum 18, are relevant to the present case. Memorandum 15 provided as follows:

Memorandum 18 provided as follows:

It became clear in the course of argument that the central issue on this appeal is whether in the circumstances of this case clause (3) of Memorandum 18 operates as a free-standing agreement, providing additional cover to the sub-contractor in accordance with its terms, or is conditional upon the operation of clause (2), providing cover only if a defect in the sub-contract works has caused physical damage either to the sub-contract works themselves or to some other part of the works as a whole. The judge held that clause (3) was subject to clause (2) and therefore only provided cover if a defect in the sub-contract works had caused unintentional damage to other work that was free of the defective condition.

Although the parties’ submissions were detailed and sophisticated (as was inevitable having regard to the intricacies of the policy), I think that the kernel of their respective arguments can be stated quite shortly. Mr. Williamson Q.C. for the appellant submitted that the language of clause (3) was clear. It states that the insurers will additionally indemnify the insured against the damage to which it refers and is not expressed to be conditional on the operation of clause (2). Although the Memorandum is structured in a way that suggests that clause (3) is subordinate to the opening lines, it is clear from the way it is worded that it is in fact an independent provision of equal status and not subordinate to either clause (1) or clause (2). Mr. Reed for the respondent placed substantial reliance on the fact that the policy provides insurance against damage to property and does not therefore respond in the absence of physical damage caused by a fortuitous event. He also relied on the fact the sub-contract does not require the employer to maintain insurance for the benefit of all those involved in the project and submitted that sub-contractors were intended to obtain liability insurance to protect themselves against the cost of making good any damage they might cause to other parts of the work.

Mr. Reed was clearly right in saying that Section 1 of the policy provides insurance against damage to the works. That seems to me clear from the operative clause, the Schedule and the Specification to Section 1 itself. One starts from the point, therefore, that the insurers agreed to indemnify the insured against fortuitous damage to the works. It was common ground that Seele fall into the category of insured who are covered only in respect of Section 1 and who therefore come within the designation “Insured (3)”.

The Schedule provides that Insured (3) are insured in respect of the works “as to the extent stated in Memorandum 15”, which, Mr. Reed submitted, had the effect of limiting cover in the case of any particular sub-contractor to the part of the works which it was itself engaged to carry out. The words in the Schedule itself tend to suggest that it may have been intended to have that effect, but in my view the language of Memorandum 15 shows that it is directed to restricting the scope of the indemnity provided in certain cases, not to limiting the scope of the property in respect of which Insured (3) are covered.

Section 1 provides cover in respect of the works as a whole. That much is clear from the Specification which defines the Insured Property in terms of the whole of the works. If it had been the parties’ intention to limit the scope of the property in respect of which an individual sub-contractor was insured to that part of the works on which it was itself engaged, that could easily have been achieved by including appropriate wording in Memorandum 15. Instead, Memorandum 15 states that the indemnity provided in respect of any works carried out by the sub-contractor which are damaged due to a defect in design shall be restricted in accordance with Memorandum 18(2) and (3). Mr. Reed’s argument at the same time proves too little and too much. If Memorandum 15 was intended to restrict cover to cases in which damage was caused to a sub-contractor’s works by a defect in those works, it would leave it uninsured in respect of damage caused by some wholly extraneous event. That can hardly have been intended. On the other hand, unless one can read the Memorandum in that way, it does not limit the scope of cover at all.

In effect, Mr. Reed wanted to read Memorandum 15 as if it said

That brings me to Memorandum 18, which I have found a difficult provision to construe. It begins with the words “This section includes loss or damage arising out of a defect in design etc. . . . . . ”. That would ordinarily refer to physical damage, but the sentence immediately qualifies that with two exclusions which are defined in part by reference to the cost of making good the defect in question. The effect of clause (1) is to exclude the whole of the cost of making good any original defect in the works as a whole and any damage caused to enable that defect to be made good (what for convenience I shall call “access damage”). However, it then goes on to provide that if the defect has caused damage to the works (other than the need for access damage), the exclusion shall be limited to the cost of any additional work needed to improve the original design. In other words, if the defect has caused damage to another part of the structure, the insured can recover both the cost of making good the original defect (but not the cost of any improvement) and any necessary access damage as well as the cost of making good the consequential damage. The insured’s retained liability in this case is substantial: £100,000 each and every occurrence or series of occurrences arising out of one event. That may reflect the fact that the insurer is bearing the risk of making good the original defect, as well as access damage, whenever it causes consequential damage of any kind.

Clause (2), which is expressed as an alternative, also excludes the cost of making good the defect and access damage, but allows the insured to recover in respect of accidental damage “in consequence of” the defect to other parts of the works which are themselves free of the defective condition. In this case, therefore, the cost of repairing the original defect is never recoverable. Intentional access damage is not recoverable either, but it is debatable whether unintentional access damage is recoverable as damage caused “in consequence of” the original defect. Fortunately, it is unnecessary to decide that point on this appeal. In this case the insured’s retained liability is much smaller: £2,500 each and every occurrence or series of occurrences arising out of one event.

It is in that context that clause (3) falls to be construed. On its face it provides additional cover in respect of access damage necessarily caused to the works to enable a defect to be made good. In this case the insured’s retained liability is £10,000 each and every occurrence or series of occurrences arising out of one event and the insurer’s liability is limited to £2,500,000. This clause is not expressly linked either to clause (1) or clause (2) and in any event its language is not couched in terms of a derogation from the exclusions contained in those clauses. In this respect I am unable to agree with the judge that clause (3) is a carve-out from the carve-out contained in each of those clauses. It is not worded in that way, unlike clause (1)(ii) and the rider to clause (2). In my view this points to the conclusion that clause (3), although structured as if it were another exclusion to which the introductory lines of Memorandum 18 refers, is in fact intended to stand as a separate clause in its own right.

It is then necessary to consider the two following clauses, both of which apply to the Memorandum as a whole. The first might be thought to present the greater difficulty since it provides that insured property shall not be regarded as lost or damaged solely by virtue of the existence of a defect. That tends to suggest that the mere existence of a defect will never be sufficient to trigger the policy, but I think a clue to its true meaning lies in the word “necessarily” in clause (3). Neither clause (1) nor clause (2) speaks of damage in any sense other than physical damage, not even clause (1)(ii) which provides extended cover in cases where the defect has caused physical damage. In my view this paragraph was intended to emphasise that the mere existence of a defect does not constitute damage and that the insurers will not indemnify the insured against any of the cost of making good a defect unless rectification is a practical necessity because it poses a risk of physical damage to the works at some later time. Read in that way it is consistent with clause (3), which provides an indemnity against damage necessarily caused to the works in order to make good a defect. That is also consistent with clause (2), which excludes the cost of making good access damage caused merely to rectify a defect that was unlikely to cause any physical damage. An example might be staining to exterior cladding, discoloured or incorrectly shaped pointing or poorly finished internal plasterwork.

Any construction of Memorandum 18 also has to accommodate its final clause. Mr. Reed submitted that the fact that it treats clause (1)(ii) as an alternative to clauses (2) and (3) is a pointer, albeit a small one, to the conclusion that clause (3) is an adjunct to clause (2). For my own part, however, I do not think that much assistance is to be derived from this clause. Clause (1)(ii) provides cover against consequential damage to the works caused by a defect, access damage and the cost of making good the defect itself (but not betterment). The deductible is high and the clause under consideration gives the insured the option to elect an alternative basis of cover when the deductible would otherwise preclude any recovery. Under those circumstances clauses (2) and (3) necessarily both come into play. I do not think that the wording of this clause enables one to draw any inference about the application of these two clauses in circumstances where the defect has not caused consequential damage of any kind.

The judge rejected the argument that clause (3) provides an independent addition to cover partly because he read it as being an appendage to clause (2) (a point which I have already addressed) and partly because of what he perceived to be the disparity that that would produce between the cover afforded to trade contractors and that afforded to the primary insureds. However, I do not think this latter point is sound. If clause (3) provides an independent addition to cover, it is equally available to an insured whose claim would otherwise be excluded by clause (1)(i), although it is true to say that he will only need to take advantage of it in a case where the defect, although requiring rectification, has not caused any damage to any other part of the structure.

The inconsistency which on this view of the matter the judge detected between clauses (2) and (3) is not in my view, a matter that should cause surprise. As the judge observed, the whole of Memorandum 18 is drafted in the form of grants of cover subject to exceptions which are themselves subject to exceptions and qualifications. It bears the hallmarks of a provision which has been fashioned from various different clauses rather than having been drafted as a single integral whole. In that context I do not think that the fact that one paragraph gives back what an earlier one has taken away is a good reason for rejecting the plain meaning of the words used. The judge’s reliance on the effect of Memorandum 15 is a matter to which I shall return. At the end of the day he appears to have been heavily influenced by the fact that the policy provides cover against damage to property. That, of course, is true, but that does not prevent the insurer from providing additional cover of a different kind as an adjunct to the policy, if he is willing to do so. Mr. Reed submitted that to construe clause (3) in the way I have suggested would amount to including a sue and labour clause in the policy, but for the reasons I have given I do not think that is a strong objection.

Mr. Williamson and Mr. Reed both sought to buttress their arguments by broad appeals to what each of them submitted was the general purpose of the policy and what would or would not make business sense for the different categories of insured or the insurer. However, having regard to the intricate way in which Memorandum 18 is drafted, I do not think that it is safe to construe it by reference to considerations of that kind, save to the extent that they emerge with sufficient clarity from the language the parties have chosen to use. Insurers may be perfectly willing to undertake all sorts of risks as an adjunct to the primary cover if the premium is sufficiently attractive. In the end the natural meaning of the words used is usually the most reliable guide to whether they have done so, unless it produces a commercially absurd result. In my view in this case it does not.

The judge was influenced in part by his conclusion that Memorandum 15 limits the scope of the property in respect of which the sub-contractor is insured to the work that he has been engaged to perform. Since Memorandum 15 refers specifically to work carried out by the sub-contractor which is lost or damaged due to a defect in it of one kind or another, that view points strongly to the conclusion that in the absence of some consequential damage, no cover is provided by the policy. However, I have already explained why I do not think that is the correct construction of Memorandum 15 and why each sub-contractor is insured in respect of the whole of the works, so that a sub-contractor who negligently damages another part of the works is insured in respect of the loss. If that is correct, it goes some way towards negating the assumption, which underpins the judge’s decision, that there can be no indemnity of any kind in the absence of damage to the sub-contractor’s own work.

I think it is clear that the defects in the sealing of the windows had to be rectified because they allowed water to enter the building. Sooner or later that would have caused physical damage to internal finishing and perhaps to other parts of the building, such as electrical installations. Mr. Williamson submitted that some damage had been caused to the windows by defective workmanship and it is true to say that the judge found that pin holes had been made in the sealing membrane by welding carried out by Seele’s workmen on site, but I think that is properly to be regarded as part and parcel of inherently faulty workmanship rather than as consequential damage. The other defects identified by the judge are all in the nature of defects in construction or the consequence of slipshod work. This was not a case, therefore, in which work carried out by Seele was damaged due to a defect in workmanship; poor workmanship simply caused each of the windows to be defective. In those circumstances Memorandum 15 as such does not come into play, but Memorandum 18 applies because it contains provisions which apply to a situation of that kind.

The judge rejected a submission by Mr. Sears Q.C., then appearing for Seele, that the plasterboard ceilings damaged when water entered the building during testing were “other Insured Property which [was] free of the defective condition but which [had been] unintentionally damaged in consequence of the defects in the windows” within the meaning of the rider to Memorandum 18(2). He considered that, because they were in close proximity to the defective weather seal and were therefore vulnerable to the ingress of water by reason of the defective design and installation of the windows, they were not free of the defective condition but on the contrary were themselves in a defective condition, as indeed was the whole facade.

I am unable to agree with that conclusion. The precise point at which a line is to be drawn between “Insured Property(a)” which is in a defective condition and “other Insured Property” which is free of the defective condition may be difficult to identify in some cases, particularly where the work being carried out by a single sub-contractor is of a complex nature. However, I think the intention behind the rider was to provide cover in respect of damage accidentally caused in consequence of the defect to parts of the work which in commercial terms are to be regarded as separate and distinct from that part in which the defect exists. For this reason it is not right, in my view, to regard the whole facade as a single item of property for this purpose. In commercial terms the plasterboard ceilings and the external cladding are each to be regarded as separate items of property. Since the testing of the windows was part of the sub-contract works, any damage accidentally caused to adjacent property in the course of testing as a result of defects in the seals would fall within the rider to Memorandum 18(2), but since the judge found that there was in fact no evidence that the ingress of water had caused physical damage to the ceilings, the point does not arise.

External cladding and internal plasterboard ceilings had to be removed in order to enable the sealing of the windows to be replaced. In my view that constituted damage to the works other than those affected by the defect, but it was deliberate damage carried out for the purpose of gaining access to the sealing membranes. The cost of making good the defects themselves and the access damage is not covered under clause (1) of Memorandum 18 because the defects were not themselves damage and did not cause damage to other parts of the works. Nor, for the same reason, is it covered by clause (2). However, because it was necessary to carry out remedial work to preserve the physical integrity of the building, the insured (including Seele) are in my view entitled to be indemnified under clause (3) in respect of the cost of making good the necessary access damage.

At this point it is necessary to consider the effect of the deductible in clause (3), as a result of which Seele must bear the first £10,000 of the cost of each and every occurrence or series of occurrences arising out of any one event. The judge did not deal with this issue in his judgment because he understood Seele to have conceded that the deficiencies in each window constituted a separate occurrence. Mr. Williamson, who did not appear below, was unable to explain quite how the judge had obtained that impression, since as far as he was aware the issue had been contested. Certainly it was not in Seele’s interest to concede it. The most likely explanation seems to be that Seele suggested that Issue 13, which asked whether the defects in each window were to be treated as one occurrence or one event for the purposes of Memorandum 18(3), should be answered ‘Yes’ in the mistaken belief that it was asking whether they were to be aggregated. At all events, the issue was argued fully before us and in my view we can and should decide it.

A deductible or excess charge leaves a certain level of risk with the insured. Its purpose, depending on the level of retained liability, may be to encourage responsible behaviour on the part of the insured or to relieve the insurer from having to respond to numerous low value claims. However, such provisions can work to the serious disadvantage of the insured when several separate incidents of damage all flow from one underlying cause. The purpose of the aggregation provision is to avoid that consequence. Mr. Williamson submitted that the best test of the existence of a single event is to ask whether there is unity of cause, intention, location and time: see Kuwait Airways Corporation v Kuwait Insurance Co. S.A.K.[1996] 1 Lloyd’s Rep. 664, 686 per Rix J. He identified three possible unifying events in this case: the implementation of the programme to make good the defects in the windows; the installation of defective windows; and the water-testing.

One must start by identifying the occurrences of damage in respect of which the insured is entitled to be indemnified, since it is to these that the aggregation provisions apply. Since all the windows were of similar design and construction and since the repair work was essentially the same in each case, I think one can infer that roughly the same amount of disturbance of plasterboard ceilings and cladding was required in each case and that both types of damage were necessary in order to make good the sealing of any single window. Moreover, the work was executed in a single continuous phase to implement a single solution to the problem that had been discovered. To that extent it can be said that the damage had certain unifying characteristics. However, the aggregation provisions of clause (3) require that the separate instances of damage should arise out of (in the sense of being caused by) one event. If there is to be aggregation, therefore, it is necessary to find one event which can properly be regarded as a cause of all the damage.

It is not altogether easy to say precisely what constitutes an ‘event’, but in AXA Reinsurance (U.K.) Plc v Field[1996] 1 W.L.R. 1026 Lord Mustill (with whom the other members of the Appellate Committee agreed) suggested that in ordinary speech, an event is something which happens at a particular time, at a particular place, in a particular way (page 1035). He contrasted it with a ‘cause’, which he considered could be something altogether less constricted and which might include a continuing state of affairs or the absence of something happening. The use of the word “event” is not altogether apposite in a clause dealing with damage which will inevitably have been caused deliberately in response to a need to obtain access to another part of the works, but the expression used in this case represents a well-recognised form of aggregation provision and must, I think, have been adopted with that in mind, the purpose being to direct the reader to the need to identify something in the nature of an event which can sensibly be regarded as the underlying cause of all the damage. In those circumstances I see no reason to construe the word “event” in anything other than its ordinary sense, but if one does that, giving it the meaning suggested by Lord Mustill, it is difficult to identify any unifying event in this case.

I can start with the water-testing. That was not cause of the damage; it merely demonstrated the there were deficiencies in the sealing of the windows, which simply throws one back to the installation of defective windows. I do not think that the installation of defective windows can be regarded as an event for these purposes either, however. If they had all suffered from a common defect in design or manufacture which lay at the root of the problem, it might have been possible to argue, despite the number of separate units involved, that the installation of windows with a common defect was an event for these purposes, but as I understand the judge’s findings, that is not really the case. It is true that there were defects in the design or manufacture of the termination bars, but it is not at all clear that that of itself was sufficient to cause each of the seals to fail. Rather, the impression one obtains from the findings in paragraph 5 of the judgment below is that poor workmanship was really to blame. It seems fairly clear that similar shortcomings in workmanship affected all the windows and I am prepared to assume for present purposes that in each case the same mistakes were made. However, there is no evidence that those mistakes were attributable to a single event, such as giving the workmen wrong instructions which they then conscientiously followed so as to produce a series of similar defects. Again, had that been the case, it might have been possible to argue that giving faulty instructions was the unifying event, but the judge’s findings point to the conclusion that the defects were simply the result of poor workmanship repeated over and over again.

That leaves the implementation of the programme of work to make good the defects in the windows. The fact that all the access damage formed part of a single programme of remedial works no doubt represents a unifying factor of a kind, but I do not think that either the decision to carry out a programme of remedial work or the implementation of that programme amounts to an event of the kind contemplated by the clause. The remedial work provided the context in which the damage was caused, but was not itself the underlying cause of it. That lay in the defects which gave rise to the need for it. In these circumstances I do not think that it is possible to identify a single event that can be regarded as the underlying cause of all the access damage required to enable the defective sealing membranes to be renewed.

For these reasons I would allow the appeal to the extent indicated and answer the questions posed by the preliminary issues as follows:

A number of other issues decided by the judge in the course of answering the questions raised by the preliminary issues were the subject of appeal. However, since they are primarily relevant to the determination of damages, it was agreed between the parties that they should make submissions on the status of those decisions at the trial of damages and that, insofar as they might be held to be binding, they should be challenged, if necessary, on appeal from the later judgment, subject to obtaining permission to appeal in the ordinary way. It is unnecessary, therefore, to consider them on the present appeal.

I approach this case without the extensive experience that both Waller LJ and Moore-Bick LJ have in the field of insurance, and therefore without their knowledge of what contracts of the present kind would ordinarily be expected to contain and be intended to achieve. Fortunately, however, the difference of view between them relates not to any major issue of principle but to a relatively narrow, albeit difficult, point of construction of the particular contract.

At the close of argument I was leaning in favour of Mr Williamson’s submission that the indemnity provided by clause (3) of Memorandum 18 in respect of intentional damage necessarily caused to the works to enable a defect to be made good does not depend on the existence of accidental physical damage to the insured property but is an additional and free-standing indemnity capable of applying even where no such accidental damage has occurred. Moore-Bick LJ’s judgment has brought me down more firmly in favour of that view. I agree with him, subject to one point, and it is only because Waller LJ has reached a different conclusion that I propose to express matters briefly in my own way.

It is not in dispute that the cover provided by the policy is in general against accidental damage to the works. That is how the operative clause to the policy and the insuring clause in the Specification to Section 1 are to be understood. But there is no reason in principle why the policy should not also include specific separate cover against intentional damage occurring in circumstances which do not involve accidental damage; and the question is whether, on the proper construction of this policy, that is what has been done.

Memorandum 15 featured large in the argument but in the end, as I read it, drops out of the picture. I agree that it simply restricts a trade contractor’s indemnity in the circumstances described in it, and those circumstances do not apply in this case. It does not restrict the property in respect of which a trade contractor is covered. Thus it remains open to a trade contractor to rely on the generality of Section 1 and in particular on the provisions of Memorandum 18 in circumstances other than those described in Memorandum 15. The inclusion of the words “(as to the extent stated in Memorandum 15)” in the definition of Insured (3) is an oddity, suggesting a greater role for Memorandum 15 in limiting the cover to be given to Insured (3), but it is the wording of Memorandum 15 itself to which effect must be given.

In so far as Memorandum 18 has a natural reading (Moore-Bick LJ is generous in describing the way it is drafted as “intricate”), clause (3) is in my view to be read as a free-standing provision rather than as being governed by, or conditional on, clause (2). The opening words of Memorandum 18 apply to clauses (1) and (2), not to clause (3). On its own wording, clause (3) provides an additional indemnity in respect of intentional damage which is separately defined and has its own deductible. The fact that, by the last clause of Memorandum 18, clauses (2) and (3) may sometimes apply together as an alternative to clause (1) does not mean that they must always apply together or that clause (3) is conditioned in some way by clause (2). Thus, neither structurally nor linguistically is there a necessary linkage between the two.

The penultimate clause of Memorandum 18, that insured property shall not be regarded as lost or damaged solely by reason of the existence of a defect in design etc., is considered by Moore-Bick LJ to pose a possible difficulty for this construction of clause (3), and he has suggested at para 42 a way of reading the clause so as to reconcile the two. That is the one part of his judgment that I do not find fully persuasive. For my part, however, I do not see why the penultimate clause should be seen as posing a difficulty at all. The fact that a defect in design etc. is not by itself to count as loss or damage to insured property does not impinge on clause (3), in which the relevant damage to insured property is not the defect in design etc. but intentional damage necessarily caused to another part of the insured property to enable the defect to be made good.

To read clause (3) as providing a free-standing indemnity in respect of access damage even where there has been no accidental damage to the works does not seem to me to produce a result so obviously contrary to the commercial purpose of the contract as to require a different approach. Under clause (1), access damage caused in making good a defect is covered if the defect has given rise to accidental damage to the works, however small. It is a relatively small step to cover access damage caused in making good a defect that has not yet given rise to such accidental damage.

On the other matters in respect of which an answer is required from this court, I have nothing to add to Moore-Bick LJ’s judgment, with which I agree. For the avoidance of doubt, I should make clear that I would answer all the Issues in the same way as he proposes.

For the reasons given, I would allow this appeal to the extent indicated by Moore-Bick LJ.