Royal Devon & Exeter NHS Foundation Trust v Atos IT Services UK Ltd

A negotiated commercial limitation clause should be construed by its natural meaning, viewed through the reasonable person with the parties’ available background knowledge. Commercial common sense may assist where competing constructions are available, but it does not justify departing from clear contractual language.

Where a clause fixes different aggregate limits for defaults occurring in separate, non-overlapping periods, the word or may operate conjunctively. The clause may therefore create separate caps for each period. A reference to a claim in such a clause may, in context, mean the underlying default rather than the later communication of a claim.

The Trust engaged Atos under a five-year contract to supply and support an electronic patient-records system. Schedule G contained a negotiated limitation clause. It set a cap equal to the total contract price for claims arising in the first 12 months, and a lower cap based on charges paid in the preceding 12 months for claims arising later.

Following project difficulties and termination, the Trust claimed damages for alleged contractual breaches. On preliminary issues, the Technology and Construction Court held that Schedule G imposed one aggregate cap, selected by the date of the first default. The Trust appealed. Atos cross-appealed, contending that the later cap should be calculated by reference to notification of the claim rather than occurrence of the default.

The central issue was whether the limitation clause imposed one cap or two separate caps, and how the later cap was to be calculated.

1. Appeal allowed; cross-appeal dismissed. The judge’s declaration that the clause imposed a single aggregate cap was set aside. The proper construction was that paragraph 9.2 imposed two separate aggregate caps: one for defaults occurring in the first contractual year and a separate, lower cap for defaults occurring thereafter.

2. The parties were commercially sophisticated and had legal advice. Although paragraph 9.2 was poorly drafted, it was valid and enforceable. Its natural meaning, assessed from the perspective of a reasonable person with the parties’ background knowledge, accorded with business common sense.

3. The words aggregate liability did not require one overall cap. They were capable of referring to the aggregate liability within each relevant period. Nor did or make the alternatives mutually exclusive in operation. The subparagraphs addressed discrete, non-overlapping periods and could operate conjunctively.

4. The clause did not provide multiple caps for individual defaults. Rather, defaults in the first year were subject collectively to the total contract-price cap, while defaults in years 2 to 5 were subject collectively to the separate cap based on charges paid during the preceding 12 months. This reflected the higher-value implementation work in year 1 and prevented a first-year default from exhausting all liability for later defaults.

5. For paragraph 9.2.2, the references to claims arising meant defaults occurring. The later reference to that claim referred back to those defaults, not to notification of a claim. The singular caused no difficulty because clause 1.2 provided that the singular included the plural. The later cap was therefore calculated by reference to the 12 months preceding the relevant default.

• Court of Appeal (Civil Division): Allowed the Trust’s appeal and dismissed Atos’s cross-appeal: [2017] EWCA Civ 2196.
• Technology and Construction Court: O’Farrell J, on preliminary issues, held that the clause imposed one aggregate cap selected by the date of the first default. That construction was set aside on appeal.