Judge interprets potentially conflicting conditions precedent and the meaning of “left unattended”


A fire at the insured’s premises destroyed the vast  majority of its stock. The evidence suggested  that the fire was started by an intruder. Although a fire alarm at the premises detected the fire,  the insured’s burglar alarm had not been set at the time and in fact was no longer being monitored  by the provider, SECOM, because its monitoring charge had not been paid by the insured. At the time of the fire, which happened at 01.00 am on a  Saturday, the owner of the insured had been asleep in a separate dwelling area on the premises.

The insured’s property insurer repudiated liability on  the basis that there had been a breach of  two conditions precedent. The policy contained:

  1. A condition precedent (GC7) which provided that “the whole of the protections including any  Burglar Alarm provided for the safety of the premises shall be in use  at all times out of business  hours or when the Insured’s premises are left unattended and such protections shall not be  withdrawn or varied to the detriment of the interests of Underwriters without their prior consent”
  2. A further clause (PW1) which provided that “It is a condition precedent to the liability of the  Underwriters in respect of loss or damage caused by Theft and/ or attempted Theft that the Burglar Alarm shall have been put into full and proper operation  whenever the premises…areleftunattended…”(emphasisadded).  The case therefore turned on the  inter-relationship between GC7 and PW1 (it being accepted, in accordance with English law  principles that, since they were both conditions precedent, a breach did not have to be shown to  have caused the loss). Jay J held as follows:
    1. Wording in a quotation sheet which indicated that PW1 was intended to override GC7 did not form  any part of the contract of insurance
    2. The correct approach to construction of GC7 is to examine its commercial purpose in light of  the other relevant contractual provisions. The judge rejected the approach of some American courts  that “each clause must be seen as an island unto itself”. Here,  it could not be said that the sole  purpose of the burglar alarm was to reduce the risk of theft (and hence a burglar alarm differs  from a fire alarm). A burglar alarm protects against the risk of intrusion which might in turn lead  to damage to property. Accordingly, it could not be said that PW1 entirely predominates over GC7.  Furthermore, the burglar alarm had a “domain of activity” outside of theft and so PW1 was not  surplusage
    3. Did PW1 qualify GC7 with the effect that the duties under GC7 could be no more onerous than  under PW1? The judge agreed that it did. There was no hard and fast rule that PW1 must be “hermetically sealed and ring-fenced off GC7”. The two must be read together and the commercial purpose behind  them considered. Jay J concluded that here GC7  had to be “read down” so that the insured was only  required to set the alarm when the premises were unattended (rather than also just when out of  business hours) and thus the requirements were no more onerous than they would have been had this  been a theft claim (and nor did the clauses have to match only for theft claims)
    4. Had the premises been “left unattended”? The situation here was said to differ from earlier   cases regarding motor insurance, where someone “attending” a vehicle cannot be asleep whilst doing so. The construction of the term was one of  fact, degree and circumstance. Here, there was a large and complex building which could not  possibly be subject to effective, active surveillance by one person and: “In ordinary parlance, houses or premises become unattended when their occupants leave”.  Accordingly, there had been no breach of GC7
    5. Even if GC7 was to be read as a stand-alone provision, it should not be construed in such a way  that there was an obligation on the insured to activate the alarm in circumstances where it would have been triggered by the movement of persons  who were legitimately on the premises (there being no reference in the clause to “part of” the  premises). Nor did it matter what the insured actually used to do in practice: “this cannot be determinative  of the claimant’s contractual obligations…those obligations must be ascertained by the application of ordinary principles of construction of  insurance documentation”
    6. However, the insured’s case failed because the judge held that it had breached the second part  of GC7. Regardless of the use of the burglar alarm on one specific occasion, the insured had failed   to pay SECOM’s monitoring charge and this had given rise to a real risk that SECOM would cease its monitoring service. The insured had known that  the charge was payable in advance, that it had not paid for over six months, and that SECOM would  not allow this situation to continue indefinitely. It made no difference that the insured  considered that SECOM was in breach of contract for causing or permitting too many false alarms. The insured  had been reckless as to the risk of the monitoring service being cut off

So, with a “degree of reluctance” the judge found in favour of insurers.

COMMENT: The judge’s construction of the two conditions precedent conflicts with some textbook  commentary that “each exclusion “is meant to be read with the insuring agreement independently of  every other exclusion. The exclusions should be read seriatim, not cumulatively … There is no  instance in which an exclusion can properly be regarded as inconsistent with another exclusion,  since they bear no relationship with one another.” The meaning must be sought first in context: within the confines of the clause  in question” (see Clarke, The Law of Insurance Contracts,  4th edn, para 15-6). However, the judge was clearly swayed  by the view that a mismatch between the insured’s  obligations in relation to a theft claim as opposed to any  other claim arising out of non-working burglar alarm made  no commercial sense and could not have been what the  parties intended.