Thames Water Utilities Ltd. v London Regional Transport & Anor

In negligence, nuisance and statutory breach claims arising from the same facts, causation ordinarily requires a single approach where the statutory duty does not materially differ from the common-law duty. The but for test is necessary but not determinative. It operates as an exclusionary filter, after which the court must assess whether the defendant’s conduct was a material, real or effective cause, using common sense and the scope of the duty. A merely nominal or de minimis contribution is insufficient. On the evidence, tunnelling, shaft sinking and associated works caused differential settlement, which materially contributed to the failure of a cast-iron water main.

Thames Water claimed damages after a 36-inch cast-iron water main burst in St Thomas Street, Southwark, in October 1999. The defendants had undertaken Jubilee Line extension works, including tunnelling, shaft sinking and compensation grouting, pursuant to statutory powers.

The claim was advanced under London Transport Act 1963, in negligence and in nuisance. The central issues were whether the works materially caused the pipe failure and whether any contribution was merely de minimis because the pipe already contained historic locked-in strains.

1. Liability and causation. Judgment was given for Thames Water on liability and causation.

2. There was a single causation test applicable to the pleaded statutory, negligence and nuisance claims because the statutory wording did not impose a materially different obligation. The claimant had to prove causation on the balance of probabilities.

3. The but for test was necessary but not determinative. It could identify viable causes or exclude irrelevant ones, but the court had also to evaluate materiality by a common-sense assessment informed by the scope of the defendant’s duty. The law did not require exceptional policy grounds to relax the ordinary approach.

4. The court accepted the evidence that the works caused both predicted and consolidation-related movements in complex made-up ground. Changes in groundwater conditions and localised alluvial deposits could produce differential settlement and a metastable event. Reliable road-pin and building-pin data established significant differential settlement near the failed pipe.

5. Laboratory testing established the relationship between joint rotation and socket strain. The rotation-induced strains, added to operational loads and residual strains, were sufficient to produce prising and shearing failure. The competing theory of an abnormal, rapidly applied external point load was unsupported by the fracture pattern, the absence of an external witness mark and the evidential analysis.

6. The historic condition of the pipe did not make the defendants’ contribution de minimis. The works were the prime and effective cause of the differential settlement and resulting loadings. The suggestion that they were merely the final straw was rejected.