[2006]JRC059
royal court
(Samedi Division)
20th April, 2006
Before : |
M. C. St. J. Birt, Esq., Deputy Bailiff, and Jurats Bullen and Le Cornu. |
|||
Between |
Jersey Electricity Company |
Plainitff |
|
|
|
|
|
|
|
And |
Brocken & Fitzpatrick Limited |
Defendant |
|
|
|
|
|
|
|
And |
Dewplan Limited |
Third Party |
|
|
Advocate D. J. Benest for the Plaintiff.
Advocate S. J. Young for the Defendant.
Advocate M. L. Preston for the Third Party.
judgment
deputy bailiff:
INTRODUCTION
1. The Jersey Electricity Company ("JEC") is the owner and occupier of La Collette Power Station in St Helier. In November 1999 there was on the site a water demineralisation plant ("demin plant") which included two tanks containing concentrated hydrochloric acid. In September 1999 the defendants (B&F), a firm of scaffolders, had erected scaffolding around the tanks so that certain work might be undertaken. On 19th November 1999 they were in the process of dismantling ("striking") the scaffolding when a clip (used to secure scaffold poles together) was dropped and struck the drainage valve at the base of one of the tanks. The valve snapped off with the result that a substantial quantity of hydrochloric acid spilt into the protective bund which surrounded the tanks. This in turn gave rise to extensive acid fuming which caused considerable damage to the demin plant and to other equipment in the power station. The plaintiff's claim was for a total of approximately £1.3 million.
2. JEC brought proceedings in tort and contract against B&F. B&F in turn alleged contributory negligence on the part of JEC. Thus questions were raised as to who, as between JEC and B&F should have been responsible for taking precautions to protect the valve whilst B&F were carrying out operations above the bund. B&F also raised issues as to whether JEC had mitigated its loss after the spillage by taking all reasonable steps to remove or dilute the acid as soon as possible. B&F also joined Dewplan Limited ("Dewplan"), who designed and installed the demin plant, as third party. B&F alleged that Dewplan owed a duty of care towards JEC to exercise reasonable skill and care in the supply and installation of the demin plant and that Dewplan would, if sued, have been liable to JEC in tort so that B&F could therefore seek a contribution or indemnity from Dewplan in accordance with Article 5(1)(c) of the Law Reform (Miscellaneous Provisions) (Jersey) Law 1960 in the event of B&F being held liable to JEC in tort. In a preliminary judgment the Court ruled that the contract between Dewplan and JEC had not excluded liability in tort on the part of Dewplan (2004 JLR 289).
3. The case came on for hearing in June 2005 during which, over some eight days, the Court heard evidence from the witnesses of fact and experts on scaffolding. Unfortunately the time estimate turned out to be insufficient and it was not possible to resume the hearing until November 2005. By that time JEC had settled its claim against B&F on the basis that B&F should pay a proportion of the sum claimed. However the Court was still required to resolve B&F's third party claim against Dewplan in that B&F argued that Dewplan should indemnify it or contribute towards the amount which it had agreed to pay JEC in settlement. Under the terms of the contract, Dewplan's liability was capped at the amount paid for the demin plant by JEC, namely £326,670. The Court heard evidence from expert witnesses and received submissions over a further six days in November, at which time it reserved its judgment.
4. Given the settlement of JEC's claim against B&F, much of the evidence (which was directed to the issues of the respective responsibilities of JEC and B&F and mitigation of loss) is no longer relevant and we do not need to refer to it. Accordingly we propose to concentrate only on the expert evidence which we heard in November and such of the factual evidence as is relevant to the third party claim against Dewplan.
THE FACTUAL BACKGROUND
(i) The installation of the demin plant
5. On this aspect the main witnesses were Mr David Killip, the production manager for JEC, Mr Kenneth Piddington, a consultant to Dewplan who was responsible for the tendering process and the initial plans, Mr David Clark, who was contract manager for Dewplan and had overall responsibility for the supply and installation of the plant once the tender had been accepted and Mr Andrew Hilditch, who was the commissioning manager for Dewplan responsible for commissioning the plant after its installation.
6. As part of the process of generating electricity JEC needs very pure water. This requires mains water to be treated using a demineralisation process. In 1997 it became necessary to replace the existing demineralisation plant which used sulphuric acid. It was decided to replace it with a new plant which used concentrated hydrochloric acid and sodium hydroxide (caustic soda).
7. Tenders were sought and one of those tendering was Dewplan, one of the UK's leading companies in this field. As already mentioned Mr Piddington was in charge of the tendering process on behalf of Dewplan and he visited JEC to discuss the position. Dewplan's tender was accepted in December 1997.
8. At that stage Mr Clark took over primary responsibility on behalf of Dewplan. Technical drawings were prepared and installation started towards the middle of July 1998 using the components and materials provided by Dewplan although they were actually installed by a sub-contractor; but nothing turns on that. JEC undertook and was responsible for the civil works i.e. matters such as the construction of the plinth and the bund. After a commissioning period the demin plant was handed over to JEC on 6th October 1998.
9. After commissioning there were various teething problems during the one-year defects liability period. These related substantially to minor fuming leaks whereby acid fumes escaped from joints in the pipe work and also to the need for a ventilation system. Mr Young made much of these with a view to suggesting that the quality of workmanship had not been very good. But we do not see that anything turns on these matters which are not relevant to the incident at hand. The defects liability period expired on 6th October 1999.
10. It is not necessary to describe the demin plant in detail. It was situated in a building which we shall refer to as the demin room. For our purposes the only relevant aspect of the demin plant is that it contained two 18 cubic metre plastic tanks of 36% concentrated hydrochloric acid. The two tanks were built on a concrete plinth and were surrounded by a concrete bund with a wall just over 2 metres high. A bund is a form of moat. It is there to protect against a catastrophic failure of the storage tanks so that, if acid escapes from the tanks, it will be contained within the bund and not flood the surrounding area where it would clearly do great damage. In accordance with industry practice the bund could take the entire contents of both acid storage tanks with 10% spare capacity.
11. We were told by Mr Piddington and Mr Clark that storage tanks for demin plants are usually situated outside. Indeed Mr Piddington said that he had designed over 500 such plants and that in only one of them could he recall the storage tanks being situated inside. This point was raised with JEC by both Mr Biddington and Mr Clark during the design stage but it proved impossible to find an outside area of the power station which was free to be used in this way and accordingly JEC instructed that the storage tanks had to be situated in the demin room.
12. The relevance of this is that if there is leakage of acid into the bund, it will clearly evaporate comparatively safely into the atmosphere if the bund is situated outside whereas, if it is in a building, the resulting acid fumes may well cause damage in the room in question.
13. Each tank had a drainage valve at its base. The drainage valves were not intended for regular use and were not intended for the purpose of removing acid from the tanks. The idea would be that the acid should be removed with specialist advice (probably into some holding tank) leaving a very low level remaining which could then be drained off through the drainage valve in order for the tank to be cleaned. In other words, the valve was intended merely to drain out the 'dregs' at the bottom of the tank. In order to prevent any release of acid in case of inadvertent release of the valve, it is standard practice to bolt a blank flange to the outlet of the valve and this was done in this case.
14. In cross-examination, Mr Young extracted two concessions from the Dewplan witnesses which are worth mentioning at this stage:-
(i) Dewplan was a company which was accredited under the ISO 9001 Quality Standard. As we understand it, this means that the company must have procedures in place to ensure quality which meet the standard required by ISO 9001. According to Mr Piddington's answers when questioned by Mr Young, the accreditation requires there to be control points during the design and installation process where the quality of design and installation is checked by a different department; there must be a quality plan for each project; and, most importantly, there must be an audit trail for all design processes through to construction. The witnesses for Dewplan accepted that there was no evidence that the company had fulfilled these requirements in connection with the design of the demin plant for JEC. Indeed Mr Clark stated that it was largely a standard package with standard materials and equipment and therefore he did not think that there was any need for these procedures to be in place on this occasion.
(ii) Dewplan did not at any stage carry out a HAZOP analysis (hazard and operability study). A HAZOP is an exercise to try and anticipate potential problems in order to see if the design addresses them satisfactorily. Again, Mr Clark said that he did not think it necessary to carry out a HAZOP analysis in this case bearing in mind that it was a standard plant with standard design. He said that no thought had been given as to whether any extra measures over and above the standard package should be introduced because the storage tanks were inside a building rather than out in the open.
(ii) The incident
15. In September 1999 B&F had erected scaffolding around the two storage tanks in the demin plant in relation to certain work which had to be carried out as a result of the fuming leaks. On 19th November they were asked to return to strike the scaffolding. It was a matter of some urgency and three employees of B&F, who were working on another job in Grouville, were instructed in the latter part of the morning to attend at the power station that afternoon. They arrived around lunchtime and began striking the scaffolding. There was then an interruption for a major emergency exercise which had been long planned. The scaffolders recommenced work at about 2.30 p.m. The three men were Mr Stephen Young, Mr Brian Norfolk and Mr Michael Fitzpatrick. At about 2.50 p.m. Mr Young dropped a scaffolding clip which struck the drainage valve to one of the tanks causing it to break. Acid poured out into the bund. The scaffolders raised the alarm and the demin room was evacuated. The fire brigade were called and they used chemical protection suits and breathing apparatus to try and address the situation. However the available pumps were all soon corroded by acid. In due course the Hampshire Fire Service were called upon and ultimately a satisfactory method of pumping the acid out of the bund was commenced at about 3.00 p.m. on Monday 22nd November and completed about midnight that day, some 3½ days after the incident. The two storage tanks were estimated to be about 60% full at the time of the accident which would mean a spillage of approximately 21.6 cubic metres. The acid was at all times safely contained in the bund but considerable damage was caused not only to the demin room but also to other parts of the power station by the fumes given off by the acid in the bund.
16. Two very contrasting versions of what the scaffolders were doing immediately before the incident were presented to the Court. Mr Young and Mr Norfolk were not available to give evidence. Their written statements were presented as hearsay evidence but the weight to be attached to them clearly suffers from the fact that they have not been tested by cross-examination. Mr Fitzpatrick gave evidence before us as did Mr Jones, an employee of Dewplan. Mr Jones happened to be at the power station on the day in question whilst working on another demin plant which JEC was planning to commission. He happened to go into the demin room and saw the three scaffolders. One, (who was clearly Mr Fitzpatrick), was standing on the floor outside the bund whilst the other two were either on the bund wall or on the scaffolding boards at the level of the first lift; he was not very clear as to which. These two scaffolders were tossing clips etc down to Mr Fitzpatrick who was catching them and then stowing the equipment. Mr Jones was somewhat concerned about this. He was also concerned about the fact that the scaffolders' equipment was leaning against his equipment and might cause damage. He decided to go and see Mr Gordon Mahrer, the duty manager of JEC. Shortly after he had left the demin room in order to do this, one of the scaffolders rushed out from the demin room saying that something inside was leaking. Mr Jones put his head into the demin room. He could hear liquid running and smelt hydrochloric acid fumes. He shouted to the scaffolders to get out.
17. The scaffolders' version was very different. Mr Fitzpatrick said that Mr Young was in the bund undoing the first lift of the scaffolding (i.e. the first floor of the scaffolding) from below and then passing equipment up to Mr Norfolk who was on the bund wall. Mr Norfolk was then passing the equipment down to Mr Fitzpatrick who was on the outside of the bund as stated by Mr Jones. He could not see Mr Young inside the bund but he heard something fall. He then heard a hissing sound and Mr Norfolk shouting to Mr Young to get out of the bund and that he could see liquid coming out. Mr Fitzpatrick could not see into the bund and therefore could not give direct evidence of how the valve came to break. Initially he was reluctant to accept that there had been any type of tossing of equipment from the other two down to him but he appeared to accept at times in his evidence that there might have been some 'controlled dropping' of equipment down to him.
18. In his written statement Mr Young said that he was in the bund dismantling the scaffolding. He said that he accidentally dropped a clip from a height of about six foot, just above his head. As it hit the floor he looked to see where it had gone. It was on the floor and did not seem to have done any damage. He turned his back and carried on dismantling. He then heard Mr Norfolk shouting at him and as he turned he saw liquid spurting out with a cloud of vapour. He ran to the back of the tank and got out of the bund. He was coughing and spluttering and his eyes were red.
19. In his statement made on 30th November 1999 to the Health and Safety Department in connection with the prosecution of JEC and B&F for breach of the relevant Health and Safety legislation, he said broadly the same thing although he suggested that he had been standing on the plinth when he dropped the clip. He again stated that he saw the clip on the ground and that it did not appear to have damaged anything. It was only when Mr Norfolk shouted that he turned and saw fluid and fumes coming from where he had dropped the clip.
20. In his written statement Mr Norfolk said that he was on the bund wall. He received the equipment from Mr Young and then passed it down to Mr Fitzpatrick on the outside of the bund. He heard a rushing noise like water from a fast hosepipe. He looked down and saw liquid coming from the bottom of the tank and a mist rising. He shouted to Mr Young to get out. He jumped off the wall and Mr Young followed him out of the building.
21. Thus no one actually saw the clip strike the valve. There is however general acceptance that this is what must have happened because of the fact that acid poured out of the valve very shortly after the clip was dropped. We find therefore that the cause of the incident was that Mr Young dropped a clip on the valve causing it to break.
22. Had we still been considering the claim by JEC against B&F the resolution of the conflicting evidence as to what the scaffolders were doing at the time of the incident would have been very necessary. However we are now concerned only with whether Dewplan was in breach of its duty of care towards JEC in respect of the design and installation of the demin plant (including the valve). On B&F's version of events, the clip was dropped by Mr Young while he was standing in the bund and undoing the scaffolding from below. Dewplan, on the other hand, assert that this is highly improbable and that it is much more likely that the clip was dropped when being tossed from one scaffolder to another at a time when they were on the scaffold itself or at any rate on the bund wall.
23. The evidence concerning the height of the scaffold is not very satisfactory. No one who had measured it gave evidence. It is clear that, although a first lift (level) of scaffolding is normally at 6' 6", it was higher on this occasion so that access could be gained from the first (and only) lift to all the areas where work had to be undertaken. Mr Fitzpatrick estimated it as being 9' high although it was not clear whether he was taking this from the main floor of the building or from the floor of the bund, which was some 250 mm higher. Mr Fitzpatrick thought that the scaffold was some 2' - 3' above the top of the bund wall. That would seem to be broadly correct having regard to the photographs which were before the Court, although the angle from which the photographs were taken is such that one cannot be precise. It would also appear to be broadly consistent with the relationship of the first lift of the scaffold to two signs (one yellow and one white) painted on the side of the tank as shown in the photographs. What is agreed is that the height of the bund wall from the bund floor (as opposed to the external floor) was 1.9 mtrs (6' 3"). Thus the first lift was between 8' 3" and 9' 3" above the bund floor.
24. The only other evidence as to height emerged during the evidence of Mr Kendall, the scaffolding expert called at the June sitting by JEC. During his evidence, reference was made to a statement made in the criminal proceedings by a Mr Andrew East, a health and safety expert from England. He inspected the site on 25th November 1999 and measured the height of the scaffold at 4.3 mtrs (sic) or 9'. The metre figure is obviously an error but the figure in feet would appear to be broadly consistent with the other evidence. However it is not clear whether he was taking this measurement from the bund floor or the external floor. We were also referred to a statement from Mr David Martin, who accompanied Mr East. He said that the scaffolding was 'some 2.5m' above floor level. However his use of the word 'some' suggests that it was not an exact measurement. When read in context it would appear that his measurement is from the bund floor. It equates to 8'2".
25. Doing the best we can with this rather unsatisfactory evidence, we find that the scaffolding was in the region of 8' 3" above the bund floor. Accordingly, if B&F's evidence is accepted, the clip fell from this height when being undone from below by Mr Young. If, on the other hand, the scaffolders were on the scaffold when the clip was dropped, we find that it was dropped from an additional 3' or so (being half the height of a man and taken on the basis of a man holding the clip in his hand whilst standing on the scaffold). In each case we must take into account that the valve itself was a short distance (we have not been given any measurements) above the plinth which was in turn 1' 2" above the bund floor. We propose therefore to deduct 1' 3" in order to find the height which the clip actually fell. Accordingly we find that it fell a minimum of 7' and a maximum of 10'. In the context of whether Dewplan is in breach of its duty of care towards JEC in its design and installation of the plant, including the valve, we do not consider this to be a significant difference and accordingly it is not strictly necessary for us to find which version of events we prefer.
26. However, in case we are found to be wrong and it should be thought significant, we find that Mr Young was not in the bund when he dropped the clip. Firstly, it is clear that he could not have reached the first lift from the bund floor. Mr Fitzpatrick said that he was 5' 11" and Mr Young was slightly taller than him. During the evidence of Mr Kendall an experiment was carried out in the Court and Advocate Benest, who is 6' 1", could only just reach to a height of 7' 7" with one arm stretched to its limit. He clearly could not have put two hands to that height and carried out any task at that height such as undoing a clip. That accords with common sense even without the benefit of the demonstration. Mr Young could have reached the scaffolding by standing on the plinth or on the valve itself but this would have been an extremely hazardous stance given that the plinth was only 6" wide and the scaffolding was not immediately above him. As the plinth was 1' 2" high, he would still have been reaching up 7' 1". That is more than a conventional first lift (which is undone from below) which is 6' 6". It is not therefore very probable. Secondly, Mr Young gave evidence (in both his witness statement for these proceedings and his statement in the criminal proceedings) of a gap between his dropping the clip and the escape of the acid. This does not ring true. It is clear from the evidence of Dr Graham, which we accept, that, this being a brittle fracture, the valve would have broken immediately upon impact. All the witnesses were agreed that the acid would have poured out at a considerable rate ("like a fast garden hose"). It is simply not credible that there was a gap as suggested by Mr Young before he became aware of acid pouring out. Thirdly we found Mr Jones to be a credible witness and we accept his evidence as to what the scaffolders were doing when he saw them. We accept that it is theoretically possible that, after he left, they moved positions, with Mr Young climbing down into the bund. We have no precise evidence as to how long after Mr Jones left the incident occurred but he does not appear to have got very far and it seems to have been fairly shortly afterwards. Furthermore, Mr Fitzpatrick was reluctant to accept that the scaffolders had at any stage been behaving as Mr Jones asserted, which adversely affects his credibility. For these reasons we find on balance that the clip was dropped from the greater height of approximately 10'.
Why did the valve break?
27. In order to decide whether Dewplan was negligent we must first consider why the valve broke. The Court heard evidence from four expert witnesses: for B&F Dr Jan Graham and Mr Peter Lumley; for Dewplan Dr Paul Shipton and Mr Rolf Clayton. In relation to the issue of why the valve broke, the key experts were Dr Graham and Dr Shipton.
28. Dr Graham is a partner of Geoffrey Hunt & Partners, consulting engineers and scientists. He has a PhD in mechanical engineering (strength of materials). He is a consulting engineer specialising in all aspects of materials performance, including elastic failure, plastic failure, creep, brittle fracture and other relevant areas. Dr Shipton has a PhD in the compounding of fibre re-enforced thermo plastics. At the material time he was head of plastics at Rapra Technology Limited and was responsible for providing a professional plastics and polymer consultancy. He has also had substantial experience at an international plastics company in the development of new products and plastics materials technologies. The Court is satisfied that both are highly qualified experts in the field, inter alia, of plastics. The Court is grateful for their reports and found them both to be impressive and helpful witnesses.
29. The valve and associated pipework was made of a plastic called unplasticised polyvinyl chloride (known as uPVC and also referred to as PVC-u). The inlet pipe of the valve was 1" bore and was bonded into a 2" x 1" reducer. The reducer was in turn bonded to a 2" flange attached to the tank. On the other side of the valve the 1" outlet pipe was attached to a further 2" x 1" reducer with a further 2" flange with a blank plate. The inlet pipe extended into the reducer and was bonded to it by solvent adhesive. It is important to note that what fractured was the inlet pipe of the valve at the point where it entered the reducer.
30. As one might expect much was agreed between Dr Graham and Dr Shipton. Both were very exact in their use of scientific terms. We hope we will be forgiven if we describe things in laymen's terms. We appreciate that in this respect that we may not be scientifically exact but we hope to give a more easily understood impression of the evidence. Both agreed that, uPVC being a plastic, it normally behaves in a ductile (i.e. plastic) manner in response to applied loads. Thus it will tend to 'stretch' as the load increases until it eventually fractures. In very simple and inexact terms one can compare it with a substance such as toffee which as one pulls or bends it will stretch until it breaks. This is what is referred to as a ductile fracture. If there is a ductile fracture there will be visible evidence of 'stretching' in the plastic at the scene of the fracture. However, in certain circumstances, the plasticity of uPVC may be reduced, with the result that a comparatively small applied load may lead to a brittle fracture. Essentially in that case there is no 'stretching' of the material. A very low temperature can lead to a brittle fracture as can other features, which are not relevant here. At ordinary temperatures uPVC will only be vulnerable to a brittle fracture if there is a 'notch' in the material. A notch is essentially any physical feature which results in a state of stress at a particular location which reduces the ability of the material to respond plastically to an applied load so that a brittle fracture may be initiated and result in a fast running crack which requires very little energy to continue and thereby lead to a fracture. Typically, notches are micro cracks or some other abrupt change in the material. Both witnesses were agreed that the fracture in this case was a brittle fracture. The question therefore is why there was such a fracture when the clip fell on the valve.
31. It was agreed by both experts that temperature affects uPVC. Although Dr Graham was inclined to take a figure of 0ºC as being the relevant temperature, he did not seriously differ from Dr Shipton's opinion that, below 5ºC uPVC was liable to suffer a brittle fracture under impact. The Court heard no evidence as to what the temperature was in the demin room on 19th November 1999. However, Mr Killip gave evidence that, because of earlier difficulties when the caustic soda in the demin room had got too cold, JEC had been advised that it must keep the demin room at about 15ºC. The ventilation system had been modified to achieve this. In the absence of any evidence to the contrary, we therefore assume that the temperature in the demin room on the day in question was not less than 15ºC. On that basis, neither party contended that temperature alone could have caused the brittle fracture.
32. Dr Graham placed considerable reliance upon an experiment carried out by Mr Lumley at the premises of MIRA Limited ("the MIRA test"). Mr Lumley obtained an identical valve assembly to that which failed. He arranged for a weight of 2.1 kilos to fall from a height of 2.75 metres on to the valve. We should explain briefly how he reached these figures. Although we have referred (and will continue to refer) in this judgment to a 'clip' falling on the valve, that is not strictly correct. A clip of the sort used by B&F in the present case consists of two parts, known as a band and a plate. These two parts are joined together when the scaffolding is secured but have to be separated when striking the scaffolding. In both his statement to this Court and his witness statement for the health and safety prosecution, Mr Young said that it was the band that he had dropped as he undid it from the plate. A clip weighs a total of approximately 2.1 kilos of which the band represents 0.7 kilos and the plate about 1.4 kilos. It follows that, if Mr Young is right, the weight of the object which fell upon the valve was only 0.7 kilos. However, in order to try and replicate the worst case scenario in relation to the facts of the present case, Mr Lumley took a weight of 2.1 kilos (being the weight of a clip as a whole) and arranged for it to fall from a height of 2.75 metres (which he considered at the time to be the maximum height from which it had fallen). Thus, in all likelihood, the object which struck the incident valve on 19th November 1999 was a band which weighed only 0.7 kilos rather than 2.1 kilos as taken in the experiment. But against that, upon the findings made in para 26, it fell from a slightly greater height than the 2.75 metres assumed by Mr Lumley. The experiment was carried out at 23-25oc whereas, as stated in the preceding paragraph, the temperature in the demin room on the day of the incident was probably not much more than 150c. However we accept the evidence of Dr Graham that a difference in temperature of this nature is irrelevant and would not materially affect the performance of the uPVC. The Court saw a video recording of the experiment. Whilst the valve assembly flexed under the impact and showed 2mm of permanent deformation, it did not fracture. Subsequently a man weighing 82 kilos stood on the valve assembly. It did not fail. The man then bounced on the assembly six or seven times and this eventually caused the assembly to fail. Dr Graham examined the valve used in the MIRA test, from which it was clear that what had occurred was a ductile fracture, not a brittle one. In other words, the uPVC pipe had stretched before eventually breaking under the load of the man jumping upon it. From this experiment, Dr Graham concluded that the valve assembly should not have failed under the impact of the clip in the present case. The only explanation for the brittle fracture was the presence of a notch and one had to look for an explanation for the presence of that notch.
33. Dr Graham concluded in his report that the notch was probably caused by the wrong adhesive being used to join the inlet pipe to the reducer. The background to that conclusion is as follows. The type of adhesive (sometimes referred to as "cement") used to bond pipes of this nature is solvent adhesive. This acts by partially dissolving the plastic on each side so that, in effect, the two plastic surfaces are joined to each other chemically rather than just sticking to the glue which remains in the middle, as is the case with conventional adhesives. It is a very strong form of bonding.
34. A number of the documents prepared by Dewplan suggested that "TANGIT" or Durapipe uPVC cement should be used to bond the pipework in the demin plant. It is not known which of these two was in fact used but they have very similar qualities.
35. TANGIT is made by a company called George Fischer. They produce a plastics technical manual entitled "A guide to the design and installation of plastic pipelines systems". In chapters 4 and 7, there are passages which comment upon the appropriate solvent adhesive to use with concentrated hydrochloric acid. The relevant passages are as follows -
"Chapter 4
"2.3 pipe joints
2.3.1 Solvent cement joints (PVC-u)
Solvent cement joints made with TANGIT Cement are generally as resistant as the PVC-u material itself. The following chemicals are, however, an exception:
Sulphuric acid H2SO4 in concentrations above 70%
Hydrochloric acid HCL in concentrations above 25%
Nitric acid HNO3 in concentrations above 20%
Hydrofluoric acid in any concentration
In conjunction with the above media the solvent cement jointing is described as "conditionally resistant". In order to obtain "resistant" joints it is necessary to use "DYTEX" solvent cement supplied by Henkel, in accordance with the manufacturer's instructions."
Later in chapter 7 there is a passage which reads -
"Users of PVC-u pipe and fittings will be familiar with TANGIT, a solvent cement based on the George Fischer formulation and containing THF as the principal solvent. TANGIT makes a sound joint which remains unaffected by almost all fluids for which PVC-u pipe is suitable. However, prolonged exposure of the inside joint edge to any of the acid concentrations below, may eventually cause some deterioration of the cement bonding...........
Hydrochloric acid over 25% .............
In order, therefore, to ensure maximum service life from pipelines conveying these fluids, we recommend the use of DYTEX which offers a high degree of resistance......."
36. After the occurrence of the incident, all the companies involved carried out their own investigations and there is an internal Dewplan memo dated 14th December 1999 which reads as follows -
"My investigations into the Jersey incident are, shall we say, interesting. I am indebted to Andy Hilditch for the vital lead to the following information -
"Standard uPVC solvent cement, including TANGIT, is not suitable for use on systems carrying hydrochloric acid at concentrations above 25%. The only product recommended for such applications is DYTEX which is exclusively made by George Fischer. If there is another product on the market, I am not aware of it.
I have spoken to Terry Pope of GF who advises that the problem arises "after a period of time" when joints begin to weep and therefore drip. He quoted an instance involving concentrated sulphuric acid where this occurred after five years. There is, however, no known instance of a catastrophic failure or anything worse than weeping.
He pointed out that whereas uPVC cement has gap filling capabilities (he quoted up to 1mm), DYTEX is not a gap filling cement, and on slack joints requires building up in layers. Presumably, adequate instructions are supplied with this product of which I had not heard before Andy mentioned it".
37. Dr Graham concluded in his report that the most likely explanation for the brittle fracture in this case was the presence of a notch in the form of micro cracks through the bond line between the inlet pipe and the reducer attributable to exposure of the adhesive joints to 36% hydrochloric acid.
38. In the light of the report from Dr Graham Dewplan consulted Dr Shipton. He carried out certain tests on the incident valve. He was asked to advise whether the hydrochloric acid could have significantly weakened the bond between the inlet pipe and the reducer and whether the acid could have penetrated to any extent along the bond line (i.e. where the inlet pipe had been cemented to the reducer by use of solvent adhesive) so as to compromise the position at or near the point of failure (which was where the inlet pipe entered the outer end of the reducer). The length of the bond line (i.e. the extent to which the inlet pipe was inserted in the reducer and joined to it by the adhesive) was 24mm.
39. Dr Shipton noted that the excessive adhesive at the inner end of the pipe (where it had therefore been in contact with the acid in the tank for about a year) had turned a milky brown colour and had become brittle in nature. When he took samples from that adhesive, it was quite hard to do as it fragmented on cutting due to its hard nature. Conversely, the excess adhesive at the other end of the inlet pipe by the fracture site was transparent and soft in nature. Slivers of this material were easily cut. It was pliable and showed no physical signs of degradation. Dr Shipton had hoped to take samples of the adhesive at various points along the bond line in order to see whether there was any evidence of acid penetration at various points along the line. However, the joint proved too narrow to enable this to be done. He therefore had to confine himself to the two samples to which we have referred, namely one at the fracture site and the other at the inner end of the pipe where it entered the tank.
40. Dr Shipton carried out an analysis of the samples of adhesive by infrared spectroscopy in order to establish whether there had been any polymer degradation in the adhesive. He said in evidence that he fully expected to find that the adhesive at the inner end (which had been in contact with the acid and showed signs of physical change) had degraded. However, somewhat to his surprise, he found no evidence of chemical degradation in that sample of adhesive or in the samples taken from the adhesive near the fracture site.
41. Dr Shipton's conclusion was that there was no evidence that the acid had permeated along the bond line so as to reach the adhesive near the point of fracture; there was no evidence of any chemical degradation of the adhesive at that point; nor was there any physical evidence of degradation near the point of fracture, in that the adhesive remained in the condition which he would expect, namely pliable and not discoloured. As to the adhesive at the inner end, where it had been in contact with the acid, there were physical signs of change in the adhesive which was almost certainly attributable to contact with the acid, but there was no evidence (on the tests which he had been able to carry out) of any chemical degradation.
42. When giving evidence, Dr Graham emphasised that he felt that Dr Shipton had not been asked the right questions. He (Dr Graham) did not suggest that the adhesive itself had failed because of acid permeation. This was a brittle fracture of the inlet pipe, not a failure of adhesive. His view was that some change in the adhesive caused by the acid had led to the formation of a notch which therefore rendered the uPVC of the inlet pipe liable to a brittle fracture. When pressed as to why he considered this had occurred, he was very frank and made it quite clear that it was simply a process of elimination. He accepted that there was no direct evidence connecting the adhesive used with the fracture. His reasoning was as follows. The MIRA test showed that the pipe should not have fractured under this load. However it had done so in a brittle way and there must therefore have been a notch. The only difference between the valve used in the MIRA test and the incident valve was that the latter (or more strictly the inner end of the inlet pipe of the latter) had been in contact with concentrated hydrochloric acid for a year with an adhesive which the manufacturer said was not suitable for use in conjunction with such acid. This led him to conclude that the failure must be connected with the incorrect adhesive.
43. The difficulty with this theory is that there is no evidence that the acid ever came into contact with or could possibly have had any effect upon the adhesive in the area of the fracture site. On the contrary, the evidence from Dr Shipton satisfies us that in all probability there had been no permeation of acid along the bond line to the fracture point. The adhesive in that area remained in exactly the condition one would expect it to be, unlike the adhesive at the other end which had been in contact with the acid. If there had been no contact between the acid and the adhesive at the site of the fracture, we do not see how the fact that the 'wrong' adhesive (in the sense that it was more vulnerable to acid) was used could have had any effect and could have led to the formation of a notch.
44. Dr Graham pointed, without great enthusiasm, to certain voids in the excess adhesive near the fracture site which appeared from the micrographs in his report and Dr Shipton's report. There were also indications of micro-cracks in the excess adhesive. Although these were in the excess adhesive rather than in the inlet pipe itself, he thought that these might possibly be stress points which could be connected with the creation of a notch which itself had disappeared in the fracture. Dr Shipton stated - and we accept - that excess adhesive often has voids. The difficulty with Dr Graham's theory is that, even if one accepts it, there is no evidence to connect the existence of the voids or micro-cracks in the excess adhesive with acid penetration and therefore with the fact that the 'wrong' adhesive was used. It seems to us much more likely that the existence of the voids and micro-cracks in the excess adhesive was simply a common occurrence which had nothing to do with the fact that TANGIT was used rather than DYTEX. One would expect to find voids or micro-cracks in either case.
45. Mr Young at one stage referred to the fact that there had been some acid fuming leaks on several occasions during the previous year which fumes would have been in contact with the outer end of the bond line near the fracture site and could therefore have affected the adhesive and contributed to the failure. Dr Shipton thought this unlikely. Furthermore, the available material from George Fisher suggests only that prolonged exposure to concentrated acid may eventually cause some deterioration of the cement bonding if TANGIT is used. Not only is there no question of any prolonged exposure (there was only occasional exposure to fumes) but we return again to the point that there is no evidence (whether visual or chemical) of any degradation of the adhesive bond in the area of the fracture. We therefore discount this possibility.
46. Given our difficulty with the theory that the choice of adhesive led to the formation of a notch, we explored with both experts in evidence whether there could be any other explanation for the notch. One theory explored was that the scaffolders may have caused a notch by standing on the valve in the course of erecting or striking the scaffolding. Dr Graham accepted that this could in theory occur but it would inevitably lead to some stretching of the pipe caused by the scaffolder's weight. If such stretching had occurred, there would be evidence of it even after the fracture and there was none in this case. He therefore discounted the possibility of a man standing on the valve having caused the notch. Dr Shipton did not dispute Dr Graham's evidence on this point and we accept it.
47. Dr Shipton said that, whilst uPVC was an extremely commonly used material with good chemical resistance, it did occasionally suffer from brittle fracture. It was not therefore a complete surprise if uPVC failed in a brittle manner. Over the lifetime of a uPVC pipe a notch could be caused by a number of things. For example, it could arise because of some imperfection of the manufacturing process or the pipe could be stored at low temperature and sustain a notch as a result of a minor impact at such temperature. In other words, there were a number of ways in which a notch in this particular inlet pipe could have arisen, thereby leading it to fracture in a brittle manner when struck by the clip. Dr Graham was not very enthusiastic about these alternative theories and thought them rather unlikely but he did eventually accept that something in the life of this particular pipe, from manufacture to its installation (particularly if it happened at low temperature) could possibly have caused a notch.
48. Our conclusion is that it is impossible to know what caused the notch in this particular pipe. However, for the reasons which we have given, we are quite satisfied that it was not caused by the fact that TANGIT or Durapipe PVC cement was used instead of DYTEX. On the balance of probabilities, we find that some event in the past history of this particular plastic pipe led to the formation of the notch, which in turn rendered it liable to fracture in a brittle manner under a comparatively small load.
49. With that finding in mind, we turn to consider the particular allegations of negligence which B&F have made against Dewplan.
The allegations of negligence
50. On this aspect, as well as the evidence of Dr Graham and Dr Shipton, the Court heard from Mr Lumley and Mr Clayton. Mr Lumley gave expert evidence on behalf of B&F. He is a mechanical engineer who specialises in mechanical engineering design in the power and process industries. He has many years experience of mechanical process design and engineering in the power and petrochemical industries with the Manderstam Group of consulting engineers. His work has included water demineralisation plants. Mr Clayton, called by Dewplan, is a Fellow of the Institute of Chemical Engineers. He has his own consultancy firm which advises on the environmental and water industry market. Before that he managed the Water Research Centre, a research laboratory consisting of some 150 staff. He has experience of designing water treatment plants.
51. The specific criticisms of Dewplan made by B&F varied a little as between those pleaded and the evidence of Mr Lumley, which included some matters which had not been raised in the pleadings. In his closing submissions Mr Young helpfully summarised those matters upon which he relied and we propose to concentrate on them.
(i) The wrong adhesive
52. We have already dealt with this allegation. For the reasons given earlier we find that the fact that Dytex was not used played no part in the failure of the valve. It follows that such omission cannot give rise to liability in negligence on the part of Dewplan.
(ii) 1" instead of 2" valve
53. The plans drawn up by Dewplan specified a 2" valve. However, the valve actually fitted was a 1" valve i.e. the inlet and outlet pipes from the valve had a diameter of 1" instead of 2". As the flange at the exit from the tank was a 2" flange, this meant that a reducer had to be applied between the flange and the inlet pipe of the valve.
54. Mr Lumley considered that a 1" valve was not fit for its purpose. The tank had a 2" flange and there was therefore no obvious reason for a reduction in size to 1". Importantly, a 2" valve would be more robust. He also pointed to the booklet produced by Hays Chemicals Limited entitled 'Hydrocholoric Acid' ("the Hays booklet") which stated at page 10:-
"The diameter of the outlet line will of course depend upon the rate of usage by the process but it is recommended that it should not be less than 50mm (2") nominal bore."
55. Because Dewplan failed to provide an audit trail of the design process in accordance with the requirements of its ISO 9001 status, it is not possible to ascertain from the records when and why the decision was taken to change from a 2" to a 1" valve. However, Mr Clark stated in evidence that Dewplan always used a 1" drainage valve. He said that this was because a drainage valve was not used very often but, when it was, it was important that the flow should not be too great. A 1" valve had a much slower rate of flow than a 2" valve and this was why it was chosen. He accepted that a 2" valve would be stronger but said that the incident valve was inherently strong and suitable for its application. Such valves were used very widely in the water treatment industry and had been so used over many years. He accepted that a diaphragm valve gave one the ability to control the rate of flow but was of the view that one would still get a much greater rate of flow with a 2" valve whereas one could control the flow very carefully with a 1" diaphragm valve. Mr Clayton was also of the view that a 1" valve was perfectly acceptable and fit for the purpose. He pointed out that the comment in the Hays booklet was concerned with an outlet line (i.e. the pipe used to extract the acid from the tank during ordinary usage) whereas what was involved in this case was a drainage valve, which would hardly ever be used and then only to extract the 'dregs' at the bottom of the tank. It was not unusual to find a 1" drainage valve. He also supported Mr Clark's evidence that the smaller size valve (even allowing for the degree of control which could be exercised by a diaphragm valve) would help ensure that the rate of flow was not excessive.
56. In our judgment it was not negligent of Dewplan to use a 1" valve. Our reasons are as follows:-
(i) Mr Young is quite correct to point out that Dewplan failed completely to comply with the ISO 9001 standards so that there was no paper trail to show why a 1" valve had been substituted for a 2" valve. However, we accept Mr Clark's evidence that Dewplan always used a 1" drainage valve, that this was commonly done in the industry and that it had caused no problems. We also accept that the reasons given by Mr Clark for preferring a 1" valve (i.e. to control the rate of flow) were reasonable.
(ii) We also accept the evidence of Mr Clayton that the use of 1" drainage valves is not unusual in the industry. Whilst not determinative, these first two factors support the contention that Dewplan did not fall below the standard of a competent and prudent supplier in electing for a 1" valve. Indeed the MIRA test carried out by B&F's own expert, Mr Lumley, showed that a 1" valve ought to have been sufficient to withstand the impact in this case.
(iii) The only reason that the valve failed was because it had become subject to brittle failure because of the presence of a notch. Although all parties accepted that a 2" pipe would be more robust than a 1" pipe, B&F has produced no evidence as to whether a notch brittle 2" pipe would have been able to withstand the impact of the clip in this case. We simply have no evidence on that point. Accordingly, even if we had found that Dewplan had fallen below the standard of the careful and prudent supplier in using a 1" valve, we would not, on the evidence, have been able to find that such failure was causative of the fracture of the valve in this case.
(iii) Use of valve limited to 20oc
57. Mr Lumley obtained a quotation from GEMU Valves Limited (the manufacturer of the valve in question) for a valve of the identical type to the incident valve for use with concentrated hydrochloric acid. The quotation referred to both a uPVC valve and a polypropylene valve. The former was stated in the quotation to have a maximum working temperature of 20oc whereas the latter was stated as having a maximum working temperature of 60oc. From this, Mr Lumley drew the conclusion that Dewplan had supplied a valve which was not fit for the purpose because the temperature in Jersey would undoubtedly exceed 20oc during the course of the year. Accordingly he said that propylene valves should have been used.
58. We are not satisfied from this slender basis that uPVC is not suitable for use at temperatures over 20oc. Thus:-
(i) Mr Lumley himself did not rely on anything other than the quotation. No explanation was given to support the assertion in the quotation and Mr Lumley did not proffer any opinion based on his own experience that uPVC was unsuitable for use at temperatures over 20oc.
(ii) Dr Graham, B&F's scientific expert, stated that uPVC was commonly used for hot water pipes, which suggests that it is clearly thought to be a suitable material for temperatures over 20oc.
(iii) The quotation from GEMU was accompanied by another document which was described by GEMU as setting out the properties of uPVC. That document gives 60oc as the maximum operating temperature of uPVC and 0-60oc as the normal working temperature range for the material.
(iv) Mr Clayton supported the view of Dr Graham and said that it was his experience that uPVC was commonly used at higher temperatures than 20oc. For example, he referred to the George Fischer catalogue, which shows uPVC as a satisfactory material for use with 36% hydrochloric acid up to a temperature of 40oc.
59. Even if we had thought that uPVC should not be used at temperatures over 20oc, such a finding would not have assisted B&F. This valve failed because of the existence of a notch which caused it to be liable to brittle failure. The Court explored with Dr Graham and Dr Shipton whether excessive temperature followed by a cooling down could lead to the formation of a notch but both experts were agreed that it could not. Accordingly we see no grounds for concluding on the evidence that the use of the uPVC valve at temperatures over 20oc could in any way have contributed to this failure.
(iv) Rubber lined mild steel instead of uPVC
60. Mr Lumley stated that, in the absence of any other piping specification, it would be industry practice to select the ASME B31.3(1999) Edition standard. ASME is the American Society of Mechanical Engineers. According to Mr Lumley this standard provides that, for 36% hydrochloric acid, non-metallic valves should not be used. He would therefore have recommended rubber lined mild steel pipes rather than plastic ones. In evidence he did however accept that plastic was very widely used in the water industry in the United Kingdom.
61. Mr Clayton accepted that rubber lined mild steel would be stronger than uPVC but said that the use of plastic pipework and valves in water treatment plants in the United Kingdom was widespread and routine, with uPVC being one of the most commonly used plastics. The ASME guidelines were American and were applied more in petro-chemical plants. It was very unusual to find them used for water treatment plants in the United Kingdom. He also pointed out that, following this incident, Dewplan had started to use rubber lined mild steel but that this had placed them at a cost disadvantage as against their competitors who continued to use plastic.
62. Given that it is clearly standard industry practice in the United Kingdom for uPVC or other plastics to be used in connection with drainage valves in water treatment plants, we would require much stronger evidence before concluding that the use of such material by Dewplan in this case had caused it to fall below the standard of the reasonable and prudent supplier/designer of water treatment plants.
(v) The use of a diaphragm valve
63. Mr Lumley referred to the Hays booklet at page 10:-
"The valves on both the outlet and drain connections of hydrochloric acid tanks should preferably be plug or ball valves. Diaphragm valves must not be used for the primary isolation of outlet or drain connections on hydrochloric acid tanks because of the possible consequences should the diaphragm fail. As a safety measure, independent drain valves should be securely blanked off when not in use."
He stated that it was not good design to use a diaphragm valve for primary isolation.
64. Mr Clayton, on the other hand, said that it was common place in the water treatment industry for a diaphragm valve to be used in such circumstances. Furthermore he referred to a guidance note issued by ICI on the bulk storage of hydrochloric acid in which a diaphragm valve is recommended. However, it is not entirely clear whether, when read in context, the recommendation of ICI relates to the main outlet valve rather than to a drain valve.
65. On the limited evidence available, we do not find that the use of a diaphragm valve as a drainage valve situated in a bund falls below the standards to be expected of a reasonable and prudent supplier/designer of water treatment plants. However, there is a more straightforward answer to the allegation, namely that the use of a diaphragm valve in this case had no connection whatsoever with the leakage of acid. The diaphragm did not fail. The cause of the acid spill was that the inlet pipe of the valve suffered a brittle fracture when struck from above by a clip. The nature of the valve (i.e. diaphragm, plug or ball) was completely irrelevant.
(vi) Failure to provide a backup cut-off valve
66. Mr Lumley referred again to page 10 of the Hays booklet where it states:-
"Every bottom outlet and drain connection should be provided with an isolation valve which should be located in the line as close to the tank as practicable to reduce the risk of leakage. For added protection, essential with all large tanks, two valves in series should be fitted."
He said that it was negligent not to have a second valve in this case.
67. As in relation to the previous submission, the answer to this is that the failure to have a second valve, even if negligent, had nothing to do with the cause of the incident. There was of course a blank flange secured at the outer end of the valve which operated as a form of second valve as long as it remained in place. However, the acid did not leak because of a failure of the valve and the absence of a second valve. The acid leaked because the inlet pipe of the valve fractured where it entered the reducer to connect it to the storage tank. This was the closest point of the valve to the tank. Even if there had been a second valve outside the incident valve, it too would have fallen off when the pipe fractured and its existence would not therefore have prevented the acid leak. The absence of a second valve did not contribute in any way to the occurrence of this failure or to the extent of the acid leak.
(vii) No remote cut-off valve between the tanks
68. Although there was a valve between the two tanks, this had to be operated manually. Once the acid leak had occurred, it was impossible to enter the bund and disconnect the two tanks. Accordingly, the contents of both tanks emptied into the bund. Mr Lumley argued that there should have been an ability remotely to close the valve between the two tanks. If this had been possible, the amount of acid leaking into the bund would have been reduced by approximately 50% because only one tank would have emptied.
69. It is of note that the standard version of the water treatment plants supplied by Dewplan includes only one storage tank. In this particular case, because of the height constraints of the demin room, it was necessary to have two smaller tanks of 18 cu. metres each rather than one large tank of 30 cu. metres. It follows that, if the standard package had been used, there would have been no facility for preventing the whole tank from emptying following a failure of this nature. We do not consider that that would have been negligent. As it happens, Dewplan did provide for a connecting valve between the two tanks but, in the circumstances, it does not seem to us to have fallen below the standard of a reasonable designer not to have gone further and to have provided for a remote switch for the connecting valve.
70. Furthermore, the existence of such a valve would not have prevented the fracture of the drainage valve, nor would it have prevented one tank from emptying into the bund. The only effect therefore would have been to have confined the leak to one tank. The evidence before us was that the extent of acid fuming depends upon the surface area of acid which is exposed, not on its volume. Thus, once sufficient acid had leaked to fill the full area of the bund, the extent of the fuming would not have been affected by the depth of acid in the bund unless or until it had all evaporated. There is no evidence before us to suggest that half the depth of acid which leaked had already evaporated by the time the remainder was pumped out on the Monday after the incident. On that basis, there is nothing to suggest that, if only one tank had leaked into the bund, there would have been any less acid fuming. In the circumstances, B&F has failed to show that the existence of a remote valve between the two tanks (even if it should have been provided) would have made any difference to the extent of the damage in this case.
(viii) Covering or supporting the valve
71. The valve protruded from the bottom of the tank and was not independently supported in any way. B&F contend that the design of the valve should have included either some form of support or some form of covering in order to protect it. The experts essentially agree that, had either of these steps been taken, the falling clip would not have caused the valve to fracture. There is therefore no problem over causation.
72. Although Mr Young listed these as two separate matters, we propose to take them together because essentially they raise the same issues. Furthermore we propose to concentrate mainly on the question of support rather than covering because the expert witnesses for B&F also concentrated on this aspect. Thus Mr Lumley said that he would not have recommended any permanent covering of the valve because he would have included a support from underneath. Dr Graham also concentrated on the issue of support.
73. Dr Graham stated that the valve should have been supported rather than hanging as a cantilever from its mounting flange and that temporary protection should have been installed above it during the course of the scaffolding work. However he did not go into any detail as to the reasons for this opinion. That was left to Mr Lumley.
74. Mr Lumley referred again to the Hays booklet which states at page 10:-
"All valves connected to stock tanks must be adequately and independently supported to prevent strain on the pipework if excessive force is applied to the valve key. This is of vital importance with main isolation valves."
He also referred to the ASME standards which specify the requirements for pipe support. He accepted that one could not have a rigid support because the plastic tanks are prone to flex, but he said that it would be perfectly possible to provide flexible support for the valve from below. He noted that the valve had an integral mounting face designed to provide location for support. It was particularly important that there should be support because a 1" rather than a 2" valve had been used. Both in his report and in evidence he agreed that it was not normal practice to provide a permanent protective cover or shield for drainage valves but contended that there should be proper support which would enable such a valve to withstand items being dropped from above. He considered it to be bad engineering practice not to have so provided and did not consider it normal for such a valve to be unsupported.
75. Mr Young also relied upon the evidence of Dr Shipton to the effect that it was foreseeable that a uPVC pipe could fail because uPVC could become brittle when a notch was present. He also relied upon the ICI guidance referred to earlier which states at para 5.6 under the heading 'Pipelines' as follows:-
"uPVC is not recommended, however, due to its brittle nature and tendency to fail under the influence of pressure surges and vibration. For this reason it is totally unsuited for use in inlet pipework."
76. Mr Clayton took a very different view. He said that, in all his time in the business, he had never seen a drainage valve supported. What was done in this case was standard practice in the water treatment industry. This was because such valves were situated in a bund. He agreed that vulnerable pipework and valves in any place where persons regularly worked should be protected; but a bund was an area which was entered only in exceptional circumstances. It was at that time that precautions should be taken. Thus, on his view, once it was decided to put scaffolding in the bund in this case, a risk analysis should have been undertaken and steps should have been taken to protect the valve properly by covering it. Ultimately, he felt that Mr Lumley was being wise after the event and speaking with the benefit of hindsight. Dewplan should not be criticised for acting in accordance with standard industry practice. He agreed that most of the water treatment plants which he had dealt with were cases where the storage tanks were situated outside rather than inside a building but considered that this did not make any difference.
77. Mr Clark said that there was no need for the valve to be supported and he was not aware of anyone else in the water treatment industry who supported drainage valves from underneath or arranged for permanent covering. This was because such valves were invariably situated in a bund which would only be entered as a last resort and under controlled circumstances. This was shown by the fact that there was no permanent ladder to enable access to be gained to the bund. If work was to be carried out in the bund, that was the time to take precautions in respect of any pipework or valves in the bund and that was the responsibility of the customer or the contractor carrying out the work in question.
78. He said that what was supplied in this case was a standard package which had been supplied many times over and was wholly in accordance with industry standards. Furthermore Dewplan sent the plans for the JEC proposal to Hays Chemicals for their approval and they wrote back on 15th May 1998 confirming that the plans fulfilled their (Hays) requirements (apart from a few minor points which are not relevant). He agreed that it would not be impossible to construct a support for the valve although he emphasised that it would have to build in some flexibility because of the tendency of the tanks to flex when being filled. A rigid support of the valve would have been potentially disastrous.
79. He accepted that no HAZOP had been carried out but questioned its value in relation to a bunded area where no one was allowed to go in normal circumstances and where strict controls should be exercised when access was allowed. He personally had been involved in the installation of some twenty to thirty water treatment plants which had the same characteristics as the present plant in the sense of the use of uPVC, the use of a 1" valve, the non-use of Dytex and the fact that there was no support for the drainage valve. There had been no problems with any of the others. What Dewplan had done was the industry standard. He accepted that the difference in this case was that the storage tanks were inside rather than outside and that Dewplan had not given any specific thought to whether any extra measures were necessary as a result; but he did not consider that this would have made any difference as the conclusion would still have been reached that the bund remained the essential protection against failure and that pipework and valves in a bund did not need protection or support.
80. Finally on this point, we would refer to the evidence of Mr Killip, the production manager of JEC who had the lead responsibility in connection with the purchase and installation of the demin plant. He said that JEC carried out a risk assessment following installation and decided to cage in the external pipework in the demin room outside the bund. This was to protect the pipework and valves from accidental damage in an area where people worked. However JEC did not consider it necessary to take any such precautions in the bund, where no operational activities took place.
81. He also confirmed that, as stated by some of the witnesses for B&F, JEC used to employ a Mr Arthur Dow as Buildings Maintenance Manager. It was Mr Dow's practice, whenever a contractor (such as a scaffolding firm) came on site to carry out particular work, to review the location in the power station where the work was to be undertaken and arrange for JEC carpenters to box in or otherwise protect equipment (such as pipes, valves, cables etc) which he thought might be vulnerable to damage during the course of the contractor's work. However Mr Dow had left the employment of JEC several years before the incident and this practice appeared not to have been continued since then. Mr Killip also pointed out that it would not have been viable to box in the valve on a permanent basis as this would have rendered it difficult to access and operate satisfactorily.
82. We have carefully considered the submissions on this point. Whereas a number of the other criticisms made of Dewplan by Mr Young were either clearly not causative of the accident or could fairly easily be rejected, the Court has not found this point easy to resolve and it represents the high point of B&F's case. However, we have come to the conclusion that it was not negligent of Dewplan to provide a drainage valve without permanent support or covering. Our reasons are as follows:-
(i) We accept that Hays recommends that valves be supported. However it seems to us that this advice is given particularly in the context of outlet valves because of the reference to excessive force being applied to the valve key. An outlet valve may of course be used regularly and manual force is often applied to open and close it. A drainage valve will be used only rarely. We also accept that, as stated by Dr Shipton and the ICI guidance note, it is not wholly unforeseeable that uPVC may be liable to brittle fracture. Having said that, the MIRA test carried out by Mr Lumley suggests that, in ordinary circumstances, an unsupported 1" uPVC valve was strong enough to withstand the sort of impact which occurred in this case and it was only because, unknown to anyone, the uPVC in this case had become notch brittle that it failed.
(ii) However, we also accept the evidence of Mr Clayton and Mr Clark that neither of them has ever come across any example in the water industry of a drainage valve in a bund being supported or permanently covered. Whilst that is, of course, not decisive, it is certainly powerful evidence of what the competent and careful supplier of water treatment plants regards as appropriate.
(iii) In our judgment the practice of the industry accords with the realities of the situation. Drainage valves of this nature are situated in a bund. A bund is the primary protection so that, if there is a catastrophic failure, the acid will be contained in the bund. For this reason, access to a bund is invariably very restricted. It is only allowed in special circumstances. A bund is not an area where workmen are likely to operate. Furthermore, a drainage valve (unlike an outlet valve) is not a valve which is routinely used. Thus the chances of someone dropping something on such a drainage valve are extremely remote.
(iv) Given our acceptance of the invariable practice in the UK water treatment industry, we agree with Mr Clayton that the ASME standards are not often applied in the United Kingdom and accordingly, although they may well be applied in the petrochemical industry, the weight to be attached to them is not as great as Mr Lumley suggested. It is of note that, in this connection, Mr Lumley did not point to a single specific example of a water treatment plant in the United Kingdom where a drainage valve in a bund has been supported or permanently covered.
(v) We note that JEC carried out its own risk assessment and decided that prudence required that pipes, valves etc in the demin room outside the bund should be protected on the basis that this was an area where workmen were present and there was a risk of damage to such equipment. However JEC too, having considered the position in the course of a risk assessment, decided that the risk of damage to the pipes and valves in the bund was insufficient to require them to be protected.
(vi) It is of course true that the vast majority of drainage valves in bunds are situated outside and that accordingly the consequences of an acid leak into the bund are likely to be less serious than where the bund is inside. However, we have come to the conclusion that this factor does not mean that Dewplan fell below the standard of a prudent and competent supplier in this case. In our judgment, given that this was a bund, Dewplan was entitled to reach the view that the chances of an object falling on the valve with sufficient force to break it was extremely remote. The risk would only become material if work was to be undertaken in the bund. In those circumstances we agree that it was the responsibility of JEC and/or B&F to consider the nature of the work being undertaken in the bund and to provide temporary protection (by covering or otherwise) for the pipework and valves in the bund, having regard to the risk of serious damage from fuming should acid leak into the bund. In this respect we think that the practice followed by Arthur Dow was the proper and prudent course to be followed by the occupier of a power station such as JEC.
For these reasons we find that Dewplan was not in breach of its duty of care towards JEC in respect of the lack of support or covering for the valve.
(ix) General
83. Mr Young made certain other criticisms of Dewplan such as a failure to show that it had used properly trained and skilled persons to apply the adhesive in this case. We do not think that these points take him any further. We agree however that we must stand back and look at the matter in the round in order to decide whether Dewplan fell below the standard of care to be expected of a competent and prudent designer/supplier of a demin plant and, if so, whether such failure was causative of the fracture of the valve in this case.
84. We fully accept that, in certain respects, Dewplan did not exactly cover itself with glory in this matter. Thus it failed to comply with the ISO 9001 standard despite its accreditation in that it failed to keep a proper paper trail of its decisions such as that to substitute a 1" for a 2" valve; it never supplied any 'as built' plans; it did not apply any defined piping standard; it did not use the recommended adhesive; it did not carry out a HAZOP at any stage; it did not give any consideration to the fact that the storage tanks in this case were to be inside the demin room rather than outside as normal; and it did not follow the Hays recommendations in certain respects.
85. However, for the reasons we have given, these failures were either not causative of the fracture of the valve or were not negligent. Ultimately, this was a standard piece of equipment built to standard specifications which has been supplied routinely by Dewplan and which complies with the standards of its competitors. Thus, uPVC is widely used and drainage valves in bunds are not supported or covered. Had these tanks been situated outside, there would have been no question of any negligence in our view. Does it make any difference to the outcome that the tanks were inside, on the basis that it was foreseeable that greater damage would be caused following an acid leak if the tanks were inside rather than outside? In our judgment, it does not for the reasons which we have given; essentially that Dewplan was entitled to expect that activities in a bund would be absolutely minimal and that if there were to be any such activity, that would be the time to consider whether steps should be taken to provide temporary protection for the pipes and valves situated in the bund. That responsibility lay with those undertaking the activity namely the customer and any contractor. The responsibility for such protection does not lie with the supplier who in normal circumstances would be unlikely even to know that such activities were being undertaken.
Conclusion
86. We therefore reject the claim of B&F to a contribution or indemnity on the part of Dewplan in respect of the amount which B&F has agreed to pay to JEC. If we had come to the conclusion that Dewplan could be liable in tort to JEC, we would have had to give consideration to the respective allocation of responsibility as between Dewplan and B&F. We would have been of the view that a substantial degree of responsibility lay with the JEC because of its failure, as occupier and operator of the power station, to ensure that any temporary protection measures were taken in the way that Mr Arthur Dow used to do. However, any responsibility of JEC must be taken to have been reflected in the amount of damages which they have accepted from B&F, which is far less than the full claim. When considering the position as between B&F and Dewplan, it could be said on the one hand that the fracture would not have occurred if Dewplan had made the valve of a different material or had provided support or covering. On the other hand it could also be said that the fracture would not have occurred if B&F had not dropped the clip or if they had taken steps to provide temporary protection for the valve in conjunction with the JEC. We would have found that the greater responsibility lay with B&F and accordingly, we would have ordered Dewplan to contribute 25% of the amount paid by B&F to JEC in settlement of its claim, subject of course to a maximum contribution of £326,670.87.
87. However, for the reasons given, we dismiss B&F's third party claim.
Authorities
Law Reform (Miscellaneous Provisions) (Jersey) Law 1960.