England and Wales Court of Appeal (Civil Division) Decisions

IN THE SUPREME COURT OF JUDICATURE
COURT OF APPEAL (CIVIL DIVISION)
ON APPEAL FROM THE HIGH COURT OF JUSTICE
QUEEN'S BENCH DIVISION (COMMERCIAL COURT)
Mr. Justice Burton
[2008] EWHC 722 (Comm)

		Royal Courts of Justice Strand, London, WC2A 2LL
		19 December 2008

Lord Justice Moore-Bick :

1. This is an appeal against the order of Burton J. made on the trial of a preliminary issue in proceedings between the appellant, John Reilly, and the respondent, National Insurance and Guarantee Corporation Ltd ("the insurer"). Mr. Reilly carries on the business of supplying and installing fire protection and detection systems under the name J & J Services. The present proceedings relate to a policy of insurance described as a "Tradesmen Insurance Policy" under which he was insured against employer's liability and public liability for the period 1^st March 2004 to 28^th February 2005.
2. The policy documents consisted of the standard form of policy and a schedule identifying those sections applicable in the present case together with certain endorsements and warranties. Section 2 of the policy provided cover in respect of public liability, subject to certain specific exclusions. Exclusion 8 provided that the insurer should not be liable in respect of liability arising out of products supplied by the insured, save in two cases which are immaterial for present purposes. However, the Section was subject to a number of extensions, including the following:

"K Products Liability

Despite Exclusion 8 of this Section the Company will indemnify the Insured against liability in respect of

1 . . .

2 accidental Damage to Property

. . . caused by any Products Supplied . . . "

3. The schedule contained a number of endorsements and warranties applicable to the cover in respect of public liability, of which only the following is material:

"TP34S

This section does not does not indemnify the Insured in respect of any claim arising out of:-

(i) . . . ;

(ii) the failure of any fire or intruder alarm switchgear control panel or machinery to perform its intended function."

4. In January 2001 Mr. Reilly entered into a contract to supply and install fire suppression equipment to three printing presses owned by one of his customers, Print Design & Graphics Ltd. ("PDG"). The equipment consisted of cylinders of carbon dioxide gas which could be discharged through fixed pipes to the likely seats of fire in each press. On 12^th May 2004 a fire occurred in one of the presses and steps were taken to activate the system, but it failed to operate as intended and as a result the fire spread, causing considerable damage to the press itself and surrounding property. PDG made a claim against Mr. Reilly and on 21^st November 2006 obtained judgment against him by consent in the sum of £2 million.
5. Mr. Reilly made a claim under his policy of insurance, but the insurer declined to indemnify him on a number of grounds and so on 12^th April 2007 he started proceedings against it in the Commercial Court. At a case management conference held on 11^th September 2007 Flaux J. ordered that the following question be determined as a preliminary issue:

Is the claimant's claim excluded from cover under tradesmen's policy numbered GLA/003910953 issued by the defendant by operation of clause TP34S?

and for that purpose he directed that the parties produce a statement of those facts that were agreed and those that were to be assumed for the purposes of the trial.

6. The facts agreed for the purposes of the trial included the following:

"2. Each such CO₂ system consisted of a main and a reserve supply. Each supply comprised a bank of 3 gas cylinders (one master and two slave cylinders) which were connected to distribution pipework by a manifold.

3. The systems were to be operated either:

(a) by the breaking of a glass call point which should in turn have activated an explosive activator fitted to a valve actuator mounted on to the discharge valve in the head of the master cylinder of the relevant system; [or]

(b) manually using a mechanical release lever fitted to the master cylinder.

4. The principle of operation of the explosive actuator was to force a piston in the valve actuator on to a pin that protruded from the top of the master cylinder discharge valve, which should in turn have caused the cylinder valve to open and allow the cylinder contents to be discharged into the pipe manifold via a siphon tube inside the cylinder. A "detent" pin in the valve actuator housing was intended to latch the piston, and thus the cylinder valve, in the fully open position.

5. In each supply, the discharge valves of the two "slave" cylinders were connected to the pipe manifold by flexible hosing. The pressure generated in the manifold by the release of liquefied CO₂ from the master cylinder was supposed to force pneumatic valve actuators to open the discharge valves of the slave cylinders, which would discharge their contents into the manifold.

6. The contents of the cylinders were supplied to the presses via rigid branched pipework. Nozzles were located at the ends of pipe branches along the length of the presses. The purpose of the nozzles was to distribute CO₂ effectively into the semi-enclosed volume of the presses."

7. The parties also agreed that the following facts were to be assumed for the purposes of the trial:

"A. Following the discovery of the fire . . . members of PDG's staff activated the fixed CO₂ system supplied by [Mr. Reilly] by the breaking of a glass call point.

C The cause of the failure of the fixed CO₂ system for press K2 was either:

(a) insufficient pressure in the master cylinder in the main CO₂ system for press K2, resulting either from the master cylinder having been incorrectly filled or pressurised or leakage from the discharge valve of the master cylinder; or

(b) failure of the actuator piston on the cylinder valve of the master cylinder in the main CO₂ system for press K2 to latch. As a consequence of this, CO₂ escaped through the pilot orifice in the cylinder valve and not the main orifice. This in turn resulted in there being pilot pressure in the manifold but not full pressure and prevented the slave cylinder valves from opening."

8. On 11^th April 2008 Burton J. delivered judgment following the trial of the preliminary issue. He held that the system, including the actuator, valves and cylinders, was "machinery" within the meaning of clause TP34S and that the clause was to be construed as referring to fire and intruder alarms, switch gear, control panels and machinery as separate items of equipment. He therefore held that the insurer was not liable to indemnify Mr. Reilly because the failure of the fire suppression equipment arose out of the failure of a piece of machinery to perform its intended function. It follows that in his view the question posed by the preliminary issue was to be answered 'Yes'.
9. On behalf of Mr. Reilly Mr. Tozzi Q.C. submitted that the judge had reached the wrong conclusion and that the question should be answered 'No'. In essence his submission was that the words "fire or intruder alarm" govern everything that follows, that is, switchgear, control panel, and machinery, so that the clause as a whole is limited to the failure of fire and intruder alarm systems to perform their intended functions. If that is right, the clause would not exclude Mr. Reilly's claim because the equipment he supplied was fire suppression equipment. Mr. Eklund Q.C. for the insurer submitted that the clause is to be read as containing a list of various items of equipment in respect of whose failure to perform as intended cover is excluded. He argued that the failure of the system as a whole, or at any rate of the actuator and valve of the master cylinder, constituted a failure of machinery to perform its intended function with the result that the insurer is not liable under the policy.
10. The principles to be applied in the construction of a clause of this kind were not in dispute. In paragraphs 13 and 14 of his judgment Burton J. summarised them in uncontroversial terms, referring to the judgment of Longmore L.J. in Absalom v TCRU Ltd [2005] EWCA Civ 1586, [2006] 2 Lloyds Rep 129, the speech of Lord Diplock in Antaios Compania Naviera v Salen Rederierna AB [1985] A.C. 191 and MacGillivray on Insurance Law (10^th ed.). It is unnecessary to repeat them here, but it is worth noting that they include the following: a presumption that the words in question should be construed in their ordinary and popular sense; that a commercial document, such as an insurance policy, should be construed in accordance with sound commercial principles and good business sense; that the commercial object of the contract as a whole, or the particular clause in question, will be relevant in resolving any ambiguity in the wording; and that in a case of true ambiguity, the construction which produces the more reasonable result is to be preferred. I would only add by way of comment that difficulty of construction is not the same thing as ambiguity.
11. The starting point must be the language of clause TP34S itself. Paragraph (ii) contains a list of items of equipment not obviously related to each other and not separated by commas. The absence of any commas makes it possible to argue that the words "fire or intruder alarm" governs everything that follows. The use of such a phrase in an adjectival sense, although perhaps inelegant, would not be particularly surprising in a document of this kind, but a comparison with a preceding endorsement, which refers among other things to "towers steeples chimney shafts" and other constructions, also not separated by commas, shows that no reliance can safely be placed on the absence of punctuation in clause TP34S.
12. Mr. Tozzi relied on the fact that the words "fire or intruder alarm" come at the beginning of the list, and as such are words to which the reader's attention is most naturally and immediately drawn, whereas the word "machinery" comes at the end, as tending to support his construction. In my view, however, like the absence of commas, it does no more than make that construction possible. (Whether, as he submitted, the inclusion of the word "machinery" at the end of the list creates a trap for the unwary is a point to which I shall return.) One way of testing his construction, however, might be to ask whether the other items of equipment mentioned in the clause are of a kind that might be expected to form part of a fire or intruder alarm system. If any of them could not, that would be a strong argument for rejecting it.
13. A control panel almost invariably would form part of an alarm system, but there is greater uncertainty when it comes to switchgear and machinery. In my view there is a distinction between a simple switch and what would ordinarily be described as switchgear. The latter suggests to me heavy duty switching apparatus of a kind that one would expect to find in a high voltage supply, a view reflected in the Oxford English Dictionary which gives as the meaning of switchgear "the assembly of switching devices and associated equipment used in the generation and transmission of electric power." A similar point can be made in relation to the word "machinery", which has no precise definition but in ordinary usage generally connotes a mechanical device of a certain level of size and complexity that operates by means of moving parts. Again, machinery is not something that one would naturally expect to find as a component of a fire or intruder alarm. Mr. Tozzi submitted that in the absence of evidence of the construction of fire and intruder alarms, particularly in large industrial premises, the court should be cautious about relying on arguments of that kind as a basis for rejecting his construction. In all the circumstances I accept that it would be unwise to treat them as determinative of the question, but in my view they can properly be taken into account as a pointer in favour of the construction for which Mr. Eklund contended.
14. Mr. Tozzi's next point was that his preferred construction was more consistent with the prospectus in which the cover provided by this form of policy was described as "wide-ranging" in nature and as providing "protection against the common risks faced by most contractors". He submitted that the construction adopted by the judge diminished the scope of cover to the point at which it could not properly be described in those terms.
15. In my view there is a limit to the amount of weight that can be attached to a statement of that kind in a document which by its own terms sets out to do no more than provide a summary of cover. Although the general description given in the prospectus cannot be dismissed as mere advertising puff, I do not think that it will bear the weight that Mr. Tozzi sought to place on it, particularly because the following paragraph directs the reader to the broker for an explanation of the details of cover and a specimen policy. Moreover, anyone who takes out a policy of this kind can be taken to understand that the scope of cover will be defined by the detailed terms, including any endorsements. Quite apart from that, however, it is questionable whether, taken as a whole, the policy would not provide wide-ranging protection against the common risks faced by most contractors, even if the insurer's construction of clause TP34S were to be preferred. The standard policy document provides cover against public liability and also, as additional options, cover against employer's liability and the loss of or damage to tools and equipment. Cover against product liability, that is, against liability in respect of accidental bodily injury and damage to property caused by any equipment supplied by the insured is provided by Extension K, but it does not extend to a failure of equipment which has no effect on persons or property other than itself.
16. The effect of the insurer's construction would be to exclude from cover liability for injury, loss and damage caused by the failure of certain kinds of equipment supplied by the insured to perform their intended functions. One obvious example might be the failure of a pump in a sprinkler system to operate resulting in the destruction of the premises by fire. However, the exclusion would not be as wide as might appear, both because it would extend only to equipment of the kind to which the clause refers and because it would not exclude liability, for example, in a case where the equipment disintegrated causing damage to its immediate surroundings. I do not think that the insurer's construction would result in the scope of the cover being at odds with the description given in the prospectus.
17. Nor do I think that the insurer's construction creates a trap for the unwary, despite the fact that the last word in the clause is "machinery". The endorsement is clearly intended to limit the scope of the cover; that much at least would have been plain to any insured reading it. It is certainly not easy to interpret, as the arguments before the judge and on this appeal have shown, but I do not think it can be said that an insured would readily assume that its effect was limited to fire and intruder alarms. One of the curious aspects of the clause, if Mr. Tozzi is right, is that it is directed not simply to fire and intruder alarms, but to particular items of equipment forming part of such systems: switchgear, control panels and machinery. I can see no sensible reason why other parts of the system, such as sensors and wiring, should have been omitted. Indeed, if it had been the intention to exclude damage caused by the failure of a fire or intruder alarm to perform its intended function, the words "switchgear control panel or machinery" could have been omitted, thereby shortening the clause and improving its clarity. That, in my view, is a powerful argument against his position.
18. On the other hand, it can be said that the choice of equipment covered by the endorsement is curious. I can understand why insurers should wish to exclude liability resulting from the failure of alarm systems, since the loss may be very substantial by comparison with the value of the equipment, but it is by no means clear why in that case they should not wish to exclude the failure of protection equipment, such as a sprinkler system or the gas injection system supplied by Mr. Reilly in this case. Nor is it clear why the insurer should wish to exclude liability caused by a failure of switchgear, control panels and machinery in particular to perform their intended functions. Given the wide meaning capable of being given to the word "machinery", it would have been more understandable if the insurer had applied the exclusion to any equipment supplied by the insured. Again, the clause would have gained both in brevity and clarity. Considerations of this kind suggest that the insurer's construction does not make complete commercial sense either.
19. There was a certain amount of discussion about other material that could be brought to bear on the question of construction, in particular the description given by Mr. Reilly in his proposal of the work he expected to undertake and certain information about the application of the endorsement to policies issued to different trades that is to be found in the underwriting guide issued by the insurer to its agents. In his proposal Mr. Reilly described his business as the installation of fire protection and detection systems, a description which eventually found its way into the schedule to the policy. In my view the nature of the insured's business, as described in the schedule, is a matter that can and should be taken into consideration when construing the policy, but I do not think that it sheds much light on the construction of endorsement TP34S.
20. The underwriting guide appears to have been provided to brokers who acted as agents of the insurer to negotiate, and in some cases issue, policies. There is no reason to think that it was disclosed to Mr. Reilly or that any reference was made to its contents in the course of his discussions with the brokers. It did not form part of the background to the contract common to both parties and is not, therefore, relevant to the construction of the policy, any more than other private information which an underwriter might draw on when deciding whether to accept a risk and if so how to rate it.
21. Finally Mr. Tozzi submitted that since the clause is ambiguous it should be construed in a manner favourable to Mr. Reilly. I agree that if a provision in a policy of this kind, particularly one which imposes a restriction on cover, is equally capable of bearing two quite different meanings, it should be construed in the manner more favourable to the insured. However, that does not mean that the court can simply choose out of a number of possible constructions the one that is most favourable to the insured without first trying to identify the true meaning of the clause. Sometimes that can be difficult, but usually it is possible to prefer one construction to another on rational grounds. In my view there are rational grounds in the present case for preferring the construction put forward by Mr. Eklund. There is no commercial sense in construing the clause in a way that requires a distinction to be drawn between different components of an alarm system when damage has been caused following its failure and, if liability is to be excluded at all under those circumstances, there is good commercial sense in not doing so. What matters to the insurer is whether the alarm system has worked as intended. If it has not, it does not matter which part has failed to perform correctly. I think that some additional support for that conclusion can be derived from the incongruity of treating switchgear and machinery as components of an alarm system. Why the insurer should have wanted the exclusion to encompass switchgear, control panels and machinery is unclear, but I do not think that matters. For whatever reason it chose to do so.
22. For these reasons I agree with the judge that the endorsement is not limited to fire and intruder alarms and any switchgear, control panels and machinery associated with them.
23. This makes it necessary to consider whether there was in this case a failure of machinery to perform its intended function. It is to be assumed that the system failed either because, for one reason or another, there was insufficient pressure in the master cylinder to generate the pressure in the manifold required to operate the pneumatic actuators of the slave cylinders, or because the actuator piston in the valve of the master cylinder failed to latch with the consequence that insufficient pressure was generated in the manifold to ensure the discharge of the slave cylinders.
24. The judge's attention was drawn to three cases in which the courts have considered the meaning of the word "machinery", The 'Arsa' (1925) 23 Ll. L. Rep. 273, (1926) 24 Ll. L. Rep. 219 (an outlet valve in the hull of a vessel), Lovelidge v Anselm Odling & Sons Ltd [1967] 2 Q.B. 351 (a flexible revolving drive shaft) and Aktion Maritime Corporation of Liberia v S Kasmas & Brothers Ltd [1987] 1 Lloyds Rep 283 (an anti-corrosion system forming part of a vessel's hull), but since the context in which the word fell to be construed in each of those cases differed significantly from that of the present case, both parties accepted that none of them provided a great deal of assistance. Perhaps the most relevant to our case is Lovelidge v Anselm Odling & Sons Ltd insofar as the court appears to have thought that the drive shaft connecting an electric motor to a cutting wheel could itself constitute machinery. It is not necessary for present purposes to attempt an exhaustive definition of the word "machinery", which is the kind of everyday word whose precise meaning may differ depending both on the user and the context in which it is used. In my view it does not have the same meaning as "machine"; a computer is a good example of something that might aptly be described as an (electronic) machine, but is unlikely to be described as a piece of machinery. Considerations of this kind led me to express the view earlier that the word "machinery" is capable of encompassing a wide range of devices which operate by means of physical movement to perform a particular function.
25. The gas injection equipment in the present case consisted of gas cylinders connected to fixed distribution pipework by valves, each of which was operated by an actuator. The actuators fitted to the valves of the master cylinders were themselves operated by explosive or mechanical means, but in either case they depended on the operation of moving parts, in particular a piston and valve collet and a detent pin whose function was to lock the valve open. The judge took the view that the scope of the word "machinery" in endorsement TP34S was not limited by reference to the equipment mentioned earlier in the clause and I think he was clearly right about that. He also held that it would not be right to consider individual components of the system in isolation and came to the conclusion that the system as a whole, including the cylinders, actuators and valves, constituted "machinery" within the meaning of the endorsement.
26. Although I agree with a large part of the judge's conclusions, I have difficulty in accepting them all. I agree with him that it would not be right to direct one's attention solely to the individual part that failed to operate correctly, for example, the detent pin, because that would not give a fair meaning to the endorsement. An individual part of that kind will rarely, if ever, constitute machinery, if viewed in isolation, although it may represent an important part of a (larger) piece of machinery. However, I am unable to accept that the equipment as a whole in this case can properly be described as "machinery". Rather, it is equipment which incorporates various components, some of which could be described as machinery and some of which could not. In my view the master cylinder valves and actuators, which were designed to operate as a single mechanical unit to allow the release of gas, can properly be described as "machinery" because of their complexity and reliance on moving parts. However, I would not extend that description to the cylinders themselves or the pipework, despite the fact that they are all physically connected, because they are separate components which, as a matter of the ordinary use of language, do not qualify for that description.
27. It follows, in my view, that if the failure of the system to operate correctly was caused by the failure of the piston in the actuator mechanism to latch properly, there was a failure of machinery to perform its intended function and the claim is excluded from cover.
28. If, however, the failure was caused by insufficient pressure in the master cylinder, the position is different. The master cylinder itself was not, in my view, "machinery". It was simply a container whose function was to hold gas and it did not fail to perform that function just because it had been incorrectly filled or pressurised. In any event, if the system as a whole failed for either of those reasons, it was not because the cylinder failed to perform its function. On the contrary, it contained the gas and released it when the valve was opened. If that was the cause of the system's failing to operate as intended, the claim would not be excluded by the endorsement.
29. The only other possibility is that there was a loss of pressure as a result of leakage from the discharge valve of the master cylinder, presumably (though this is not stated in the agreed assumptions) because of a partial failure of the sealing mechanism between the valve and the body of the cylinder. The sealing mechanism itself is not "machinery", but it forms an integral part of a larger mechanism made up of the valve and actuator which, for reasons given earlier, can in my view properly be described as "machinery". Attention has naturally been concentrated on the way in which the actuator and valve are designed to work when the system is put into operation, but it might be said that in fact it has two functions: to retain the gas in the cylinder until it is required and then to release it. However, whereas the actuator and valve depend on the movement of internal parts to release the gas, the seal between the valve and the cylinder is entirely static. Endorsement TP34S is intended to exclude liability where loss is caused as a result of the failure of machinery to perform its intended function. The intended function in that sense of the machinery represented by the valve and its actuator was simply to release the gas. I do not think that the leakage of gas from a faulty seal between the valve and the cylinder can be regarded as a failure of the machinery to perform its intended function in the sense meant by the endorsement. If that was the cause of the system's failure, therefore, I do not think that the insurer's liability is excluded.
30. I would therefore allow the appeal in part and answer the question posed by the preliminary issue in the following way:

(a) If the cause of the failure of the fixed CO₂ system was insufficient pressure in the master cylinder resulting from the master cylinder having been incorrectly filled or pressured, No;

(b) If the cause of the failure of the of the fixed CO₂ system was insufficient pressure in the master cylinder resulting from leakage from the discharge valve of the master cylinder, No;

(c) If the cause of the failure of the fixed CO₂ system was the failure of the actuator piston on the cylinder valve of the master cylinder to latch, Yes.

Lord Justice Thomas:

31. I agree.

Sir Anthony May:

32. I also agree that this appeal should be allowed in part to the extent and for the reasons given by Moore-Bick LJ.