'5.7 Details of the sterilisation process became an important aspect of the investigation into the cause of the leakage. But . . . the Defendant's request for documentation relating to the sterilisation process were [sic] met with the answer that the Claimant had no relevant documentation and that it simply did not know whether sterilisation had been undertaken or not.

5.8 At the Hearing . . ., the only factual evidence about the sterilisation process was given by [a witness] on behalf of the Defendant. [The witness] indicated that he had received reports that the Claimant's sub-contractor . . . did chlorinate the pipework system. He also confirmed that certain forms had been submitted by the Claimant which indicated that tests had been carried out by [the subcontractor] on the pipework installation. However, he explained, these forms in themselves did not constitute any approval or certification that the testing in question had been carried out in compliance with the Specification or in accordance with the Contract. [The witness]'s evidence in this connection was not challenged. It is notable that some three months after the . . . hearing, the Claimant disclosed a series of records that, it claimed, had "just come to light" which also showed that sterilisation had been performed in Blocks 1 and 2 of the development.

5.9 At the . . . hearing, the Defendant produced technical papers on the subject of chlorination . . . The Defendant submitted that these papers supported its position that poor sterilisation practice is a well known and recognised problem likely to cause pitting corrosion in copper pipework such has [sic] occurred at this case. These papers were, however, criticised by [an expert] at the Hearing as giving no boundaries and excess levels to the effect of chlorine. Subsequently, in its Response Case, the Claimant's experts refuted the conclusions drawn by the Defendant from these technical papers and expressed the view that on the contrary these papers contradict the Defendant's position.

5.10 Turning to MIC, the Claimant responded to the Defendant's contentions by claiming that all the evidence pointed to the corrosion being microbially induced. MIC is now a well-documented phenomenon wholly unconnected with sterilisation, but still requiring a chloride content in the incoming water. Although the particular mix of water supplied to the building varies (being a mixture of desalinated and borehole water), tests indicated that the chloride content may vary from around 20 to 80 parts per million (ppm). These figures were not considered unusual and were indeed typical of much of the domestic supply in the UK. MIC requires the existence of microbes or bacteria in the water and the formation of a biofilm over the inner circumference of the pipe. The biofilm protects the copper from chloride in the water. However, if the biofilm is locally damaged, the anions, such as chloride and sulphate, will migrate to and concentrate upon small areas of exposed copper, which will act as anodes. Should such circumstances persist, a stable corrosion pit, protected by a nodule formed by the product of the corrosion process, will be formed which will eventually perforate the pipes. The biofilm, whilst initiating the pitting, is not required once the pit is formed.

5.11 MIC is usually associated with poorly buffered supplies and it was not suggested that MIC could not have developed from the [local] water supply. The Defendant's response to the Claimant's experts concentrated on demonstrating that the particular conditions necessary for MIC had not or were unlikely to have occurred.

5.12 The mechanism leading to MIC requires that the biofilm should first be formed and then disturbed and punctured. The suggested cause of disturbance is claimed to be turbulence within the water. The circumstances postulated by the Claimant, therefore, consist of a period of stagnation, (although the Claimant's experts emphasised that this need not be complete stagnation) lasting for a number of weeks, followed by turbulence sufficient to break the biofilm. Chloride salts in the water coming into contact with the copper create a small anode, which leads to progressive seeding and growth of a chloride nodule, leading eventually to perforation. It is the initiation of the process that is important. The subsequent growth of the nodule is progressive and less dependent on the initial conditions.

5.13 In its Closing Submissions . . ., the Claimant made two points of general principle. The first related to the burden of proof. The Claimant submitted that the Defendant had to prove its allegations that the Claimant had not only carried out the sterilisation process but also that it had done so improperly. Furthermore, if the Defendant maintained its claim that all the pipework had to be replaced, the Defendant had to prove that the sterilisation process had been improperly carried out in each of the 125 apartments.

5.14 The second point of general principle related to causation in that, whilst the Claimant genuinely believed that all the evidence pointed to the corrosion being microbially-induced, it is not for the Claimant to demonstrate that case on the balance of probabilities. It was enough that the Claimant could demonstrate, on the balance of probabilities, that the Defendant's case as to causation is incorrect. In this connection, the Tribunal was referred to The Popi M [1985] 1 WLR 948, where the House of Lords made clear that:


(a) the burden of proof remains at all times on the Plaintiff;


(b) the Court can conclude, even after extensive evidence, that the proximate cause of loss, even on the balance of probabilities, remains in doubt, with the result that the Plaintiff fails;


(c) although it is open to a Defendant to suggest and seek to prove some cause other than that asserted by the Plaintiff, there is no obligation on them to do so, neither is there any obligation on them to prove, even on the balance of probabilities, the whole of their alternative case.

5.15 The Claimant also submitted that the Defendant had failed to provide a plausible explanation of the mechanism of how the sterilisation process had caused the damage to the pipework. The Claimant contended that each of the mechanisms which the Defendant identified as capable of causing the corrosion, namely:


- too much chlorine left for too long;


- incomplete draining down of the water pipework; and


- failure to flush the water system after draining down,


were mere possibilities only, none of which could be proven as the actual cause of corrosion.

5.16 Quoting from a paper . . ., the Claimant submitted that the main problem in the type of corrosion caused by chlorination is the substantiation that excessive levels of chlorine had in fact been used or that the period of exposure had exceeded the British Standards recommendations. The Claimant submitted that the Defendant could not explain how much chlorine was too much, or how long might be too long for the chlorine to remain in the system.

5.17 In support of its position, the Claimant argued that at the . . . Hearing, Mr [A], the Defendant's Expert, confirmed that without the records being available, he could not say whether excessive chlorine had been introduced in any apartment, or whether it remained for an excessive period of time. Mr [A] also confirmed that in apartments where there were no samples, he was unable to say whether or not excessive levels had been used. Mr [A] also accepted in cross-examination that, although there was a wealth of technical and academic literature, nowhere is there a single case history on a project where the corrosion of copper pipework was due to the failure to carry out sterilisation properly. Similarly Mr [B], the Defendant's second Expert, accepted that whilst there was evidence in technical papers of the dangers of chlorination, there was no case history or case study that would show that corrosion resulted from improper sterilisation.

5.18 The Claimant also argued that the technical evidence was such that the cause of corrosion is more properly attributable to MIC. The Defendant conversely argued that the conditions required for MIC to develop in copper pipework were absent at the locations where corrosion had occurred in the water system.

5.19 In assessing which of the mechanisms contended for by the Parties was the real cause of the corrosion, one of the difficulties was the Claimant's refusal to call evidence from those involved in the sterilisation process, particularly the Claimant's sub-contractor . . . who undertook to carry out the process. The Claimant's statement that it had no documents relating to sterilisation, other than the tender and sub-contract made with [the sub-contractor] added to the difficulties. The Claimant's stance was criticised by the Defendant, who asked the Tribunal, in the circumstances, to draw adverse inferences against the Claimant. The Tribunal, while recognising that it had no power under English procedural law to call a witness, shared the Defendant's concern and found it impossible to accept that the Claimant could not have obtained documents which would have thrown more light on the process of sterilisation had it so chosen. The Tribunal considers it appropriate to draw adverse inferences in these circumstances, to the extent such inferences can properly be drawn. However, such an approach does not necessarily assist the Tribunal in coming to a conclusion.

5.20 Such documents as are available show that, before the sterilisation process was carried out, [a building consultant] gave approval to the Claimant's proposal to sterilise at a lower concentration than that provided under the contract. The [consultant's] Specification provided for chlorination to be undertaken with free residual chlorine levels being monitored. For "Water tanks and Distribution Pipelines", the Claimant was obliged to repeat the chlorination process "if the concentration of free residual chlorine is less than 30mg/l(ppm)". . . . [A]pproval was sought and obtained for the use of a weaker solution of chlorine for sterilisation than that specified, to be left in the system for two days, which was a longer period than that specified in the . . . Specification. Although the Tribunal has no hesitation in accepting the possibility of excessive sterilisation leading to corrosion damage, there is little assistance in the published literature as to what may constitute excessive contact time or concentration. The approved change from the contract levels gave a reduced concentration for a more prolonged period of time. Bad practice could have resulted in both of these values being exceeded-and if an adverse conclusion is to be drawn this could be an explanation.

5.21 The Defendant's case concentrated on the likelihood of improper flushing after sterilisation. This should have been carried out by flushing the charged system through with water. The Defendant contended that if, in fact, the sterilising fluid were simply drained it would have left droplets, which would have tended to concentrate in horizontal runs where the great majority of both perforations and nodules had been found. Again, there is no direct evidence as to whether or how this operation was carried out. The only reason suggested as to why the contractor should choose this course is that it would involve use of less water, which was to be paid for by the contractor. The specification is clear as to the method that was to be adopted. The system was to be flushed and measurements of the chlorine content were to be taken until levels at the inlet and outlet points were equal. The operation was to be supervised and should have been recorded.

5.22 The Claimant on the other hand argued that corrosion was likely to have been induced by MIC and put forward reasons why corrosion could not be due to over-chlorination. These included the following:


- Time of attack: if the sterilization . . . was the cause of the corrosion, then the damage would have occurred much more quickly and would have been much more widespread;


- Uniformity: if sterilisation were the cause of corrosion the outcome would neither be sporadic nor localised, as described by the Defendant;


- Pipes of all diameters & positions would have been affected: if sterilisation were the cause of corrosion, then the failures would not have occurred mainly in the 15mm horizontal sections of the pipework;


- Damage to the internal surface was confined to the top 270 degrees, which would be contrary to the sterilisation theory.

5.23 The scientific evidence postulated by the Defendant in favour of sterilisation being the cause of corrosion as against MIC was that the corrosion pits were broad, rather than narrow and deep as would be expected if the source of chlorine was the supply water. Also, the singular penetrations observed were claimed to be consistent with sterilisation as the cause, rather than the "pepper pot" formation normally expected with MIC. The Claimant's experts argued against reaching either of these conclusions and the Tribunal does not regard either factor as strongly indicative of one mechanism rather than the other. Of somewhat more significance is the fact, already noted, that virtually no literature exists on the dangers of, and conditions under which, corrosion may result from sterilisation. There are no published manufacturers' warnings. The Tribunal is prepared to draw adverse conclusions, as already stated, and would be prepared to conclude that the Claimant, or [its sub-contractor ] may have acted irresponsibly. There nevertheless appears to be no guidance in the literature, beyond the association of a particular concentration of sterilising fluid with a particular period of contact, on actions or inactions which may amount to abuse or irresponsibility. The Tribunal is left with the conclusion that the type of carelessness, which might have led to sterilisation-induced corrosion, was something not hitherto anticipated or thought worthy of being recorded in the literature.

5.24 In addition to the above, while improper draining down could be reflected in penetration occurring predominantly (if not exclusively) in horizontal pipes, no mechanism was suggested which would favour corrosion of 15mm pipes as against 22 or 28mm. The collated figures for perforations and nodules show both perforations and nodules to be predominantly in 15mm piping and conspicuously absent from 28mm, with a small number occurring in 22mm pipe. In favour of sterilisation as against MIC was the existence of perforations and nodules in the hot water system, although both are heavily concentrated in the cold water systems. The evidence on corrosion in the hot water system was less than conclusive, since the possibility existed that the particular length of pipe could have been outside the hot circulation loop, and also that the heating or the circulation pump could have been switched off at the time that the corrosion was initiated.

5.25 The Defendant . . . alleged extensive bad workmanship by the Claimant and its sub-contractor . . . It was contended that on a balance of probabilities the only credible explanation for the corrosion problem was incorrect sterilisation practice and that there were fundamental deficiencies in the Claimant's arguments that MIC and its associated pitting were present at the Development.

5.26 . . . [T]he Claimant contended that the Defendant had got nowhere near proving its claims on the balance of probabilities. The Claimant contended that on the contrary, there was a huge weight of evidence supporting the Claimant's contentions. The Claimant submitted that if the Defendant had carried out regular flushing of water from the pipework systems of unoccupied or little used apartments in accordance with good practice and the requirements of BS 6700 the chronic stagnation of the water in the pipework systems of such apartments would not have occurred with the result that continuous biofilms would not have formed in those systems and that MIC would therefore have been prevented.

5.27 Furthermore, as recommended by [a water quality expert], to cure the corrosion problem, the Claimant proposed that the pipework should be cleaned using Citric Acid wash in order to remove corrosion nodules, scale and other deleterious deposits to the internal surfaces of the pipework and scour the corrosion pits. [An expert ] also suggested that should it be found necessary to replace the copper pipework, an alternative to using copper should be considered in the form of installing semi-rigid tubing threaded through the existing service ducts in the apartments, in single lengths, with no joints.

The Tribunal's conclusions

5.28 The Tribunal is satisfied that the evidence presented has established that microbes did exist within the pipework and that, where tested for, a "biofilm" was found in all of the samples in which perforations or nodules were found. These tests were carried out by or under the direction of [an expert]. The Defendant's experts accepted the test results.

5.29 MIC is consistent with corrosion being concentrated in the 15mm pipe runs, where the water velocity and consequently the turbulence would be higher. The Claimant also demonstrated that a rough correlation existed between the height of affected apartments and the number of perforations: hydrostatic pressure will be lower at the top of the building where a smaller number of penetrations have occurred.

5.30 Conversely, although sterilisation-induced corrosion is a possible explanation, there is no firm evidence on which to base such a conclusion. Chlorisation would not be expected to be concentrated in 15mm pipe runs.

5.31 Weighty arguments have been presented for and against both mechanisms. It has been established that the elements necessary for the development of MIC were present, whereas the facts necessary for sterilisation-induced corrosion remain a matter of speculation. MIC is a now well-known (although by no means widespread) phenomenon, whereas sterilisation-induced corrosion appears not to be a recognized or documented risk. The Tribunal accepts that the Defendant has the onus of proof. Taking into account all the evidence and arguments, the Tribunal concludes that the Defendant has failed to establish its case that corrosion was caused by improperly performed sterilisation.

5.32 To the extent it is necessary to make any finding on the cause of corrosion, the Tribunal is satisfied on a balance of probability that the most likely explanation of the observed facts is the existence of MIC within the pipework system, caused or induced by the nature of the water supply, coupled with intermittent use of parts of the system. The Tribunal is unable to accept the Respondent's contentions as to the cause of corrosion.'