CHAPTER SEVEN - CAUSATION 3
7.1. Mr Vernon’s Knowledge of PSV 504
A matter that was extensively argued and certainly is an important question in these cases is the state of Mr Vernon’s knowledge about the condition of PSV 504 at the time of the accident. Assuming that Mr Vernon contributed to the cause of the accident by seeking to restart Condensation Injection Pump A at a time when the relief line from the pump was not available because PSV 504 had been removed from the line for service did Mr Vernon know that this was the position. If he knew then this would go towards meeting the pursuers’ need to establish that the accident was caused by Mr Vernon’s negligence. Of course the pursuers’ case is not that Vernon was negligent by attempting to start the pump when the valve was missing but rather that he had no knowledge of the state of the pump and that this was due to his failure to carry out his duty to inspect the worksite. If he had carried out such an inspection, they say, it would have brought home to him that the valve was missing. The pursuers have a delicate balance to achieve on this matter and this is due to the position taken by the defenders. The defenders aver that if the accident happened because Mr Vernon was attempting to start Pump A then he knew that what he was doing was contrary to the Operators’ safety procedures so that his decision to start the pump was a deliberate failure to follow good and prudent work practices and consequentially was "wilful misconduct" as defined in the Indemnities. This would free the defenders from any liability they may otherwise have had under these. Thus this matter while by no means the only point in the defenders’ defence is undoubtedly a pivotal point. Moreover it should be noted that it is not critical to the defenders’ position that Vernon knew that the PSV was missing for if he did not know this they say that his ignorance points to a deliberate failure on the part of the Operators to institute and maintain good and prudent handover and permit to work practices which could also point to "wilful misconduct". Of course the state of Vernon’s knowledge could have a bearing on the question of whether or not he was attempting to start Pump A about the time the accident happened. The pursuers’ position is not that Mr Vernon never knew about the state of the valve during the evening of the accident but that if he had at some point known then the information had not at the time registered in his mind or that he inadvertently had forgotten about it at the time he made his decision about Pump A. It was said that the thought uppermost in Mr Vernon’s mind towards 10pm on the evening of the accident was that the pump had been handed over to maintenance for a 24-month planned maintenance and his concern was to discover if any maintenance work had been performed on the pump which might preclude taking the pump back into production. In any event since as Lead Production Operator he ought to have known that the valve was absent he was negligent if he began to start the pump and the pursuers have a prima facie right to recover under the indemnities whether or not Vernon had actual knowledge that the valve was absent. Accordingly the pursuers submit that provided that Mr Vernon pressurised the pump when the valve was missing, if the defenders want to claim that his conduct was wilful misconduct, they have the onus to aver and prove this.
7.1.2. The Relevant Evidence
One fact that was agreed by all the evidence of experienced operators was that it was contrary to prudent and good practice to run a Condensate Injection Pump with the safety PSV missing. Indeed it was accepted that no operator, including Mr Vernon, would ever do such a thing. I think this evidence is soundly based. The safety element in the situation is not difficult to understand. The purpose of the PSV is to open to relieve pressure should there be a surge in pressure above acceptable levels when the pump is running. Such surges occur from time to time and if the PSV is not in place the increased pressure has nowhere to go. The Safety Procedures include a proposed procedure for running a pump in an emergency without its PSV. This would involve a person being posted at the blind flange. None of the witnesses favoured nor indeed could explain the advantage of this procedure and indeed the witness Mr Henderson said that if it was applicable at all it was only applicable to apparatus with two PSVs (of which Pump A was not one).
The witness Mr Bollands took over his duties as Control Room Operator at about 5.15pm on 6 July 1988 and he replaced Mr Price. It should be noted that all Production Operators rotate so that Mr Bollands was perfectly experienced in the duties of a Production Operator. About the same time Mr Richard replaced Mr Grant as the Phase 1 Operator. Mr Bollands saw Mr Flook and Mr Vernon conducting their handover which took place at the Lead Production Operator’s desk. He expected that matters of significance would at that time be recorded in the Log kept on the Lead Operator’s desk. This was an A4 notepad. Either the incoming Lead Operator was handed notes on this pad or he took his own notes on the pad. The contents of these notes is not clear and it may well be that each Lead Operator decided what he considered to be important and worth noting. Generally just before the shift turnover on the Day Shift was a busy time in the Control Room because performing authorities would be returning their permits to be cancelled, continued, suspended or whatever. The busy period usually began about 5.30pm. This means that if Mr Rankin followed the usual practice with regard to the valve permit which would have needed to be suspended overnight then the probability is that Mr Vernon would have been asked to suspend it. If on the other hand Mr Rankin had finished what he could do that day to the valve at an early time and had taken his permit say at 5pm to the Control Room then it may have been Mr Flook who suspended his permit.
When just before 9.45pm that evening Mr Vernon was informed that an alarm had gone off which indicated that Pump B had tripped he went to assist Mr Richard after having a short conversation with Mr Bollands. Then the JCP Panel Alarm went off and shortly after this Mr Vernon returned to the Control Room. The JCP alarm would be consistent with the tripping of the Condensate Injection Pump and would indicate that the JCP level was increasing unduly. Mr Vernon was informed that the JCP alarm had gone off and he also informed Mr Bollands that he had unloaded and recycled the Reciprocating Pumps. This procedure would relieve the pressure on the JCP Drum. Mr Vernon informed Mr Bollands that he had been unable to get Pump B to restart and that he expected that there was trouble with the lube oil system on the pump. Mr Vernon then indicated that he would try and see if he could get Pump A back from Maintenance and get it running. At this point he said nothing about PSV 504. Indeed Mr Bollands asserts quite categorically that he personally did not at the time of the accident know that PSV 504 was not in place. Mr Bollands said that if Mr Vernon was intending to run Pump A without its safety valve he would have expected him to discuss this. As he put it "it was such a drastic step". He thought that Mr Vernon would have valued his opinion. Mr Bollands also thought that if such a decision had to be taken Mr Vernon would not have taken it without recourse to his superiors. If Mr Vernon had radioed to his superiors other operators with their radios switched on, like Mr Grieve, would have heard. When Mr Vernon saw Mr Bollands within minutes of the explosion the first thing he did was to ask Mr Bollands what had happened. This may indicate that Mr Vernon himself did not associate the explosion with anything he had been doing to the injection pump. Of course later Mr Vernon was killed. Mr Clark was also adamant that Mr Vernon had not raised with him the matter of PSV 504. Mr Clark was an experienced Lead Maintenance Hand and if Mr Vernon was about to do something that was irregular and possibly risky one might have expected him to take the opportunity of discussing it with Mr Clark. Equally it can be said that if Mr Clark knew that the PSV was absent it might have been expected that he would have mentioned this to Mr Vernon. Mr Clark also made other significant observations. He declared that if he had known that the PSV was off he and Vernon would never have started the pump and also that if the valves were missing the electricity would have been immobilised from a second source so that he would not have signed the red tags as he had.
The witness Mr Grieve who was down at the 68-foot level at the time of the accident also says that he had no idea that the PSV was missing.
With regard to Mr Vernon’s character the witness Henderson said that he had worked with Vernon for 8 years and found him to be responsible. He accepts that it was very bad practice to run a pump with the safety valve missing and says that if necessary the plant would have been shutdown to prevent that happening. Moreover Mr Vernon was hoping to get promotion and in that situation I would wonder if he would be inclined to take a highly irregular step that almost inevitably would be found out.
The Phase I Operator would normally know the work going on in his Modules during his shift and certainly Mr Grant, the Phase 1 Operator during the day shift on 6 July, would have known since he at one point helped in removing the valve. Of course when he went off duty he may not necessarily have known that the valve had not been replaced since until the last moment the valve fitters were trying to secure a crane to replace the valve. The question, of course, is whether or not he would have communicated the status of the pipe to the incoming Operator, Mr Richard, at handover particularly if no work was to proceed at the valve over the night shift and the valve calibration was being done in conjunction with the planned maintenance. On the other hand the fact that the planned maintenance was proceeding was something that he certainly would have been expected to mention. This does not however mean that he would have thought it necessary to communicate to Mr Richards what was happening to the PSV since the pump had passed out of the control of operations and was with maintenance. Moreover as I have said he may not have known the precise position regarding the PSV.
Mr Rankin gave evidence to the effect that after 6pm he met Mr Smith in the accommodation module. He describes Mr Smith as his Maintenance Supervisor which is incorrect in that he was the Maintenance Lead Hand. Mr Rankin was right in thinking that as the employee of a specialist contractor he should report to the Maintenance Supervisor but of course he got the wrong man. Mr White was his proper supervisor and from Mr White’s lack of response when Mr Clark got his telephone call shortly before 10pm, it can perhaps be assumed that Mr White did not know the state of the PSV. Mr Rankin was possibly confused in that Mr Smith had assisted him to get started with the valve in the morning and as he had only recently begun to act as a Score Supervisor he may not have been as familiar as some Score employees with the personnel on the platform. Mr Rankin claims to have returned his permit to work to the Control Room (although he is vague about the detail of this) and then to have visited the Score container before proceeding to the Accommodation Module where he met Mr Smith. I have no reason to doubt that the encounter with Mr Smith took place and since the circumstances are such that it can be assumed to have been after 6pm, when Mr Smith was off-duty, we have one time reference that can perhaps be relied on. Indeed another witness Mr Mochan confirms, at least to a degree, the time of the meeting. It would appear that Mr Smith had gone off-duty about 5.45pm and before meeting Mr Rankin in the recreation room would have required time to make a telephone call and to have changed out of his work clothes. Mr Smith asked Mr Rankin if blind flanges had been fitted and the latter confirmed that this was so. It is perhaps unfortunate that Mr Smith did not also ask Rankin if he had reported to Mr White but it would perhaps be understandable if he just assumed that this normal procedure would have been, or would be, carried out. This confusion may also reinforce Mr Clark’s claim that he had no opportunity to find out about the state of the PSV. Also if Mr Smith did not know what had happened to the valve before he went off-duty (he may well not have known that the work had not been completed if indeed he knew that the permit had ever been issued) then it is unlikely he would have mentioned the situation during his handover with Mr Clark. Thus whatever Mr Vernon’s paths to knowledge about the position regarding the PSV it is unlikely he gleaned any information about it from Mr Clark.
The evidence of Mr Rankin may be of crucial importance in relation to Mr Vernon’s knowledge of the valve work. Mr Rankin was uncertain in some of his evidence and this is perhaps not surprising. He had survived a dreadful accident which had killed his workmate Sutton. Then he finds himself associated with the critical events said to be responsible for the catastrophe. Moreover the accident happened shortly after he had become supervisor which of course must have added to his concern. Prior to the accident Mr Rankin had been employed by Score for about two years as a valve technician and had previously been employed by Otis Pressure Control. Thus in a general sense he had reasonable experience of valve maintenance operations. Before the accident he had worked on valve related work on Piper Alpha and other offshore platforms. He had worked on Piper Alpha for about 6 weeks during March 1988. However during that period he had not been acting as supervisor. He was appointed to that position on the day before returning to the platform for the tour of duty that led to the accident. He had not acted previously as supervisor. Mr Sutton and he began their tour on Piper Alpha about 27 June 1988 and their tour was expected to last about 4 weeks. Before the accident he had on that tour maintained about 10 valves. He claimed, and I can accept it, that it was a routine job to fit blind flanges. Yet at one point in his evidence he indicates that he cannot ever remember having been involved in the fitting of valves to open-ended pipework. Before he came to Piper Alpha as supervisor he had no experience of operating the permit to work system on the platform. He realised quite clearly that his function as supervisor included being responsible for following the permit to work procedure, to refurbish and certify the valves, and to oversee the work. In dealing with the permit to work situation he was helped by the fact that in the Score Container a list was posted indicating the various steps which had to be taken to follow the permit to work procedure. There is no doubt that he appreciated that the permit to work procedures were important for the regulation of safety. However it might appear that the specific instructions which his employers gave him about the detail of the permit to work system were fairly skimpy. He was given no specific instructions about the fitting of blind flanges. However he knew that blind flanges always required to be tightened with a flogging hammer or combination spanners. His view of the function of a blind flange was that it was attached to prevent accidental spillage from the system or to prevent dirt getting into the pipes. Since he must largely have acquired this knowledge while working as an ordinary valve technician this is an indication of what such a person might have been expected to appreciate about the fitting and the function of blind flanges.
When Mr Rankin began his tour as supervisor he was introduced to Mr Todd who asked him about his knowledge of the permit to work system. Mr Todd was presumably satisfied that Mr Rankin knew sufficient about the system. Mr Rankin had informed him that he had no problems with this and that he had run through it with his own supervisor.
Mr Rankin claimed possibly to have had experience in the suspension of permits to work but his evidence in this respect was distinctly vague. However the documentary evidence shows that at least on occasions during the tour Mr Rankin had suspended permits or had them extended.
Mr Rankin’s shift began about 6 am and was due to run until 6pm. Normally a valve would be completed during the one shift. Once a permit had been issued for Mr Rankin’s work he would be given a copy of the permit. He claimed that he would have kept it on his person which is not exactly how he should have dealt with it.
The defenders argue that the evidence shows that matters involving the extension or suspension of permits would normally be handled by the incoming Lead Production Operator and they are right about this. Of course the practice would assume that there would be an incoming Lead Production Operator recently come on duty when the suspension or extension was required. Thus were it the case that Mr Rankin had brought the permit back to the Control Room at 5pm it would be consistent with the evidence that Mr Flook would have dealt with any suspension required at that time.
There was evidence which was not disputed that it was not unknown for a pump which had been withdrawn for planned maintenance to require to be brought back to service and the defenders found on this as indicating that the Lead Maintenance Operator should have had this in mind when apprising himself of the working conditions on the platform. It was certainly part of his duties to be familiar with the running of the plant during his shift. He controlled it on the floor of the platform. As we were told the tripping of a pump is not unusual and that may have been why Mr Seddon when he telephoned the platform from onshore was rather uneasy about doing the planned maintenance.
In relation to Mr Vernon’s state of mind when he went to restart the pump it is certainly most unfortunate in every sense that he did not live to give his own account of the facts of the night of the accident for which he is being blamed. However Counsel for the defenders maintained that the test of a person’s state of mind would in any event have to be decided on the objective data. He referred me to Walker and Walker "0n Evidence" where at par. 24 the authors say that a party’s state of mind can be proved only by inferences drawn from other facts so that the circumstances associated with these facts are relevant and admissible evidence. I did not understand that this expression of the law was disputed. Of course there may be many cases unlike this one where direct evidence of a state of mind was available but even in such a case the reliability of the witness as to that would have to be assessed by reference to evidence of the other facts and circumstances. Nor do I have any difficulty with Mr Keen’s argument that if the proved facts would have conveyed the necessary knowledge to any reasonable person then the required knowledge will be imputed to the relevant person. I was also referred to some of my own dicta in the wilful fireraising case of John Scott Blane v HMA 1991 S.C.C.R. 576 where I observed:
"however, since we cannot see into the fireraiser’s mind his intention must be derived from the objective circumstances surrounding the incident. A man will be presumed to intend the natural and foreseeable consequences of his acting ... a reckless indifference to the consequences of a deliberate act of fire raising can be taken as equivalent to these consequences being intentional".
That quotation deals of course with intention to be attributed to actings and not with state of knowledge. However the problem in this case is not the law on this matter but the difficult "Jury" question of ascertaining what the surrounding facts were and what inferences can be taken from them. Defenders’ counsel’s point was of course that since Mr Vernon must be taken to have been a reasonable man then from the facts known he must be presumed to have been aware that Pump A was lacking its PSV. However the problem is compounded by the fact that the issue is not what Mr Vernon may have known at some time but whether or not he retained that knowledge at a particular time.
The defenders also made a submission that far from being pressed for time Mr Vernon had plenty of time to take alternative measures. He had over an hour before the Suction Condensate Vessel would overflow to flare. The de-isolation of the electricity would involve some time because this involved the identification of the tags and the relevant keys, the retrieval of the keys from the key safe, the removal of padlocks and the return of the racks. This would have to be done not only for the motor but for the lube oil system as well. These would have required about 15 minutes. However the argument was continued by the defenders to the effect that there was not such an emergency that an experienced Operator would have thrown caution to the wind. If the word "deliberately" is added I doubt that the pursuers would disagree with that sentiment. The defenders placed great emphasis on the fact that for Mr Vernon not to be aware of the PSV’s absence he would have to have ignored or not absorbed not one factor but a great many. The Defenders summarised these as being, firstly, that the PSV permit would have been suspended at the end of the day shift and it is clear that it must have been Mr Vernon himself who suspended it. Secondly Mr MacGregor had a permit suspended at about 9.15pm on the evening of the accident. It would have been Mr Vernon’s duty to inspect this worksite at the point of suspension and this site was close to where PSV 504 should have been located. In fact Mr MacGregor’s work had been to fix a plate to the Suction Scrubber near the scaffolding which had been erected for the PSV removal. Thirdly there were many people directly in contact with Vernon who would have known about the PSV maintenance. Fourthly it was said that even if Mr Vernon did not himself suspend the PSV permit at the shift handover he would have known about it assuming that he had followed the other safety procedures. Fifthly the events in the Control Room just before the explosion suggest that Mr Vernon must have known about the position regarding the PSV. I do not think that there is any doubt that it was Mr Vernon’s responsibility to keep himself informed of features of the process which may have affected safety. If anyone in management should have known about the status of the PSV it was he.
The pursuers themselves make averments in their pleadings which the defenders said are important to the resolution of the problem being considered. They aver:
"It was the duty of Robert Vernon to have the worksite checked in terms of the Permit to Work procedures then operating on the installation. Robert Vernon had not checked said work site prior to the initial explosion. Had he inspected the work site he would have become aware that PSV 504 had been removed and had not been replaced. In that respect Robert Vernon was negligent and so contributed to the accident".
The pursuers argue that the defenders originally denied these averments and only admitted them by amendment at a late stage of the proof. Thus at the relevant time the pursuers had simply not proved these averments which at the time they led relevant evidence were in dispute. It is of course obvious enough why the pursuers made these averments since they were anxious to show that they had little choice but to make compensation payments to the victims. The question remains if they have at the same time established one of the defences to the application of the indemnities by themselves apparently accepting that the accident was caused by an intentional departure from good and prudent practice. This is certainly the defenders’ contention and indeed they say that since they have admitted the averments in question these must be taken as among the established facts in the case. In terms of the permit to work procedures the Lead Production Operator in his capacity as Designated Authority has an obligation to satisfy himself with the job site in the event that he is responsible for the suspension of the permit (paragraph 3.6(a) of the permit to work procedures in the General Safety Procedures Manual number 12/405 of process). However it must be noted that the responsibility on the Designated Authority is not specifically to inspect the site but to "satisfy himself". Whatever the pursuers take to be the import of the rule about the procedure to be followed upon suspension there is little doubt that in practice not all the operatives considered that it was necessary to carry out an actual inspection to satisfy themselves about the worksite. Thus Mr Henderson thought that it was sufficient to satisfy oneself by indirect inquiry and that an inspections would only be necessary if there had been serious work carried out such as breaking into the condensate system. On that view what would be a reasonable procedure to satisfy himself may vary with the particular circumstances. Thus in the case of an absent PSV it may be a sufficient safety check to inquire if blind flanges have been fitted particularly if the pump has already been isolated. In any event it is by no means clear that a Lead Hand who did not carry out a physical inspection would consider that he was violating the safety procedures. Mr Snape thought that upon suspension it was the duty of the Lead Production Operator to inspect the site but he may not have been as familiar with the Rules on this point as the production operators who were following the system from day to day. The matter of the blind flanges seems to have been Mr Smith’s only concern when he spoke to Mr Rankin in the recreation area. In any event whatever the wording of the Safety Procedures the procedure followed on the platform seems to have been that granted that it was often impracticable to inspect a site it sufficed in many cases simply to be satisfied that the site had been left safe. Applications for suspension of permits were likely to be coming in thick and fast at or about the changeover period. In any event there is a fundamental illogicality in the pursuers’ averments on this matter. If Mr Vernon did have a duty to inspect the worksite this could only be because he was aware that the valve work had been proceeding. Since no-one suggests that he had been asked to cancel the permit it is intrinsic to the situation that he knew that the work had not been completed. If he knew this then it is not likely that on inspection he would have discovered anything he did not know before the inspection. In any event in relation to the Court’s attitude towards implied admissions the pursuers referred me to Charles Lee v The National Coal Board 155 S.C. 151 and Wilson v Clyde Rigging and Boiler Sealing Co. &c 1959 S.C. 328. I do not find it necessary to examine the implications of these cases. In any event no attempt was made by the defenders to have excluded evidence that showed that the alleged admission was totally inapplicable to the facts. Even if an admission on Record is to be regarded in the same light as evidence (as Lord Sorn suggests in Lee at page 160) if further evidence is without objection admitted to the proof which contradicts it then the Court must assess the whole evidence. Cancellation of a permit (as distinct from suspension) involves a duty to inspect the worksite but the permit to work relating to PSV 504 would only have been cancelled if the work on the valve had been completed and it is one the uncontested facts in this case that the work had not been completed.
It should perhaps be noted that when Mr Todd the Maintenance Superintendent was asked if it was good practice to run the Condensate Injection Pump with the PSV missing he had no hesitation in saying that it would not be good practice. Mr Clark was somewhat ambivalent about his own knowledge of what would be bad practice on this matter but given Mr Todd’s position I think it can be assumed that Mr Clark as a Lead Maintenance Hand familiar with the equipment would well know that it was not satisfactory to run the pump in the absence of an important safety feature such as the pressure relief valve.
On the matter of signing-off electrical de-isolation tags although Mr Clark was not entirely consistent about this I think it is clear that as Maintenance Lead Hand he had authority to sign these tags albeit that he was not the Designated Authority on the permit. The production of documents relating to such de-isolations establishes this.
One matter that may affect the evidence of Mr Bollands to the effect that on the day of the accident Instrument Technicians had been working on Pump A under the planned maintenance permit is that it seems undisputed that on the day preceding the accident Maintenance Technicians had been working on Pump B. In his memory Mr Bollands could have got the two pumps confused. On the other hand if such technicians had been working at Pump A during the preceding shift Mr Clark would not have had direct knowledge of this since he would not have been on duty.
It should also be noted that extending a permit does not require any inspection. Upon a handover any suspended permits are handed over in a separate pile.
If two permits relating to the same equipment were issued then there was evidence that they would be folded together but this did not necessarily happen if the permits related to work at different platform levels and given the system for grouping the permits this would make sense. Moreover it was not made clear that say a live permit or a permit application would be folded with a suspended permit. Certainly Mr Bollands never referred to seeing two permits in Mr Vernon’s hand at any time. In any event the defenders argued that if there was only one set of electrical isolation tags then this must have related to the PSV permit. This is based on evidence that the planned maintenance permit had not been issued and evidence that electrical isolations would not be carried out until the work is about to begin. The pursuers made the point that it had not been proved that the replacement of the PSV actually required electrical isolation. Certainly the evidence was that only some designated authorities required such isolations for PSV work. Indeed were the position with regard to electrical isolations critical to the defenders’ attempted reliance on the exceptions in the Indemnities it has not been proved that the calibration of the Pressure Safety Valves invariably led to such isolation although I would accept that it often did.
There was evidence that when the air lines to the pump were disconnected a tag saying "do not operate" is attached. It was suggested that if Mr Vernon had attempted to restart the pump he would have seen these tags. However since his objective was to take the pump out of maintenance I am not quite clear how it can be assumed that he would be deflected by the tags.
It was argued by the defenders that the pursuers cannot challenge Mr Clark on the matters contradicted by Mr Bollands because they did not challenge either witness. However the defenders did not object to any evidence from Mr Clark on the basis that it contradicted evidence which the pursuers had led and not challenged. They were prompt to make such challenges on other occasions and if a party has been misled by a pursuer’s approach to a witness this may be a good ground for objecting to an attempt to lead contradictory evidence. This arises on the basis of failure to give fair notice of the case being developed. However in the case of Mr Clark’s evidence far from claiming prejudice the defenders came to adopt and rely upon much of his evidence. In any event where witnesses are giving evidence on relatively small matters of detail where from the circumstances recollections might be expected to be unreliable about such detail I am inclined to think that there is nothing irregular in not challenging all such material. In such a case the integrity of the witness is not being challenged and the point is often not whether a concession might be rung from the witness but whether at the end of the day one version of events more closely fits other evidence than another. Thus for example Mr Clark indicated that the electrical isolations were taken out not for the PSV but for the planned maintenance and the defenders themselves did not appear eager to challenge Mr Clark on this and other aspects of his evidence. Mr Clark could hardly have supposed that the electrical isolation tags had been referable to the PSV contract since he claims that he did not know about this and yet at handover he was obviously told about the electrical isolation. When Mr Vernon telephoned him he did not say "I do not know what you are talking about. The permit application is in the Safety Office and in accordance with practice there would be no red tags attached to it until the work is about to begin". His attitude would only be consistent with a belief that electrical isolation tags were already attached to the permit application since he seems not to have known about any PSV tags. The point made may rise more sharply if one is left only with contradictory evidence from the witnesses whose evidence is under review. Then if neither witness has been cross-examined it may well be inappropriate to prefer one witness to the other if there is no other evidence on which a decision can be based. However I need reach no conclusive view on the technical point made for in my opinion it is misconceived. Mr Bollands not only gave his understanding of the status of the Planned Maintenance permit but actually claimed that he had read it. This is at the core of his evidence and it is true that the pursuers did not challenge him about this. Mr Clark did not for his part say that he had seen an unissued planned maintenance permit in the Safety Office or indeed anywhere else. What he founded his opinion on was the evidence that Mr Smith had told him that the permit was not issued. He came after Mr Bollands (who was not cross-examined by the defenders) and the pursuers at least to a degree challenged Mr Clark’s assertion that the maintenance permit was never live. Unfortunately Mr Smith did not survive the accident so that the account Mr Clark gave of his conversation with Mr Smith could not be challenged nor could the basis for Mr Smith’s information be tested. In these circumstances the pursuers may well have decided that there was little point in challenging Mr Clark’s evidence as to what he had been told. There is of course a division in the evidence as to whether Mr Vernon saw Mr Clark in the Control Room and one or other of the two persons who spoke about this must be mistaken. As I have already indicated it would certainly be difficult to decide between them on this point so that it is fortunate that it does not matter. It may also matter little whether the red tags remained on a permit at the time when Mr Clark signed them or whether they had earlier been detached by Mr Vernon. These incidental details are exactly what one might expect a witness to be mistaken about.
The defenders submit that Mr Bollands must be wrong in suggesting that instrument technicians were working on the pump during his shift because evidence was given by the night technicians that contradicts this. However it would not be at all difficult for Mr Bollands to confuse technicians working towards the end of the day shift with technicians working in the evening. Whatever else is clear it is that Instrument Technicians were not working on planned maintenance of Pump A when Mr Vernon came down to investigate pump B. Certainly Mr Clark said that no work had been done on the pump that night and this therefore could be true in relation to the night shift. On the other hand Mr Bollands is very certain that Mr Vernon took the pink permit with tags on it from the box of active permits for the 68-foot level. Two details are particularly important here. Firstly if the permit to work had related to the PSV it would have been a blue permit. Secondly if the permit had related to the valve maintenance then it would have been kept at the 84-foot level slot and not that for the 68-foot level. It is perhaps unlikely that Mr Bollands would have been in error in respect of both details. Of course he might be mistaken about the detail but if he is right it is difficult to see why a revalidated permit relevant to the situation was in the live permits box unless some work had indeed been proceeding under it. As the timings may help to clarify whether or not Mr Clark saw Mr Vernon in the Control Room I think they would permit the view that an encounter there was perfectly possible. On the other hand as I have said I do not find it critical to reach a concluded view about that fact.
The pursuers referred me to authorities aimed at assisting me to arrive at a solution to the problem of witnesses who may give conflicting evidence about factual questions. The first case they referred to was Keenan v Scottish Co-Operative Society Ltd 1914 S.C. 959. In that case (which was a jury trial) three witnesses had given important evidence for the pursuer but the defenders led conflicting evidence which was not cross-examined. It was held that the pursuer must be held to have conceded the veracity of the account given by the defenders’ witnesses. The pursuers distinguished that case on the basis that on the vital matters they had cross-examined Mr Clark. Moreover, of course, there is the fact that both witnesses were led by the pursuer and in circumstances where it might be conceded that each was trying his best to remember detailed events of some complexity in circumstances where genuine lapses of memory might be expected. The other case cited by the pursuers was Margaret Robertson or Stewart v The Corporation of Glasgow 1958 S.C. 28. This was a reparation case where one expert had examined a pole which had caused the accident and was not cross-examined on his views. The second expert had not examined the pole but on the technical questions agreed with the first expert on the basis that he accepted the results of the latter’s examination of the pole. A question arose as to the corroboration of the condition of the pole and the pursuer sought to corroborate this point from the fact that the first expert had not been challenged. It was held that failure to cross-examine a witness did not supersede the need for corroborative evidence on the point under discussion. It should be noted that the case was concerned with credibility in relation to a critical matter. If one were applying the above cases strictly then the fact is that it was the defenders in the first place who accepted the evidence of Mr Bollands without challenge. The defenders in other words did not challenge Mr Bollands credibility. They accepted that he was trying his best to remember detail of the important events. In cases of this sort one can understand why Counsel may hesitate to suggest to a witness that he is not credible. Counsel for the parties were prepared in general to accept that each witness was doing his best to tell the whole truth and to base any criticism of the accuracy of a witness on other evidence of detail. In any event what the pursuers argued is that the evidence of Mr Bollands and of Mr Clark can largely be reconciled. This is true and even if there is divergence on matters of detail these do not destroy the pursuers’ position in relation to Mr Vernon. On the really essential matters the evidence is sufficiently consistent particularly upon matters that were not challenged with Mr Clark.
One fact beyond doubt is that shortly before the accident occurred Mr Young the instrument technician had been summoned to the 68-foot level. The defenders say this shows that by that stage Mr Vernon’s intention was to restart Pump B. Certainly it can be argued that a Instrument Technician might well be of more service in starting the pump which had tripped rather than the pump to be retrieved from the maintenance programme. On the other hand the fact that Mr Vernon was trying to get Pump A started would not preclude continuing efforts to trace the trouble in pump B and get it in working order again. Further Mr Vernon had expressed the view that the trouble with Pump B was in the Lube Oil system which would not require an Instrument Technician to cure. The fact is that in my view Mr Young’s summons does not advance matters significantly.
There is certainly no evidence that there was any talk of suspending any planned maintenance permit before doing any work restarting Pump A. If Mr Clark had been presented with an issued planned maintenance permit one would have expected that he and Mr Vernon would have suspended this permit when they signed the tags. Neither Mr Bollands nor Mr Clark speak to that happening. Thus either Clark is right when he says he was only given tags or at least he is right when he suggests that the permit had never been issued.
In relation to the question of timings I have already in Chapter 5 discussed these at some length and I have also there elaborated on my views as to whether or not a permit to work for the pump had been issued. The question is difficult and although the evidence that the permit had in fact been issued is marginally stronger as I have already stated I am not sure that the matter is one which impinges critically on Mr Vernon’s state of knowledge.
7.1.3. Mr Rankin’s Part
In relation to the work he had done earlier in the tour Mr Rankin thought that generally he had been working on plant with more than one PSV available to it. The valve fitters could not work on a valve on a line which was in production and accordingly they had to wait until a particular piece of equipment became available. This is what happened with PSV 504. In the morning of 6 July Mr Rankin was told by Mr Smith that PSV 504 would be made available to the valve technicians during the day. It was the last valve they had to attend to before their tour of duty ended. Mr Smith mentioned that there would be further ongoing work on the pump which would allow the removal of the valve for maintenance. The conversation with Mr Smith took place in the Maintenance Lead Hand’s Office which was close to the Maintenance Superintendent’s Office. The timing of the conversation with Mr Smith is not exactly clear but we do know from Mr Lynch that by about 7.15pm Mr Smith knew that he would be getting Pump A for maintenance and that he had applied for the requisite permit. This timing is vouched by the fact that Mr Curtis signed the permit about 7.40 am. Mr Rankin came to the Control Room just after 8 am to inquire if he could get the valve without a separate permit. This fact itself may reflect his relative inexperience. Mr Rankin stated that after he got his intimation from Mr Smith he obtained a permit form for the PSV from his Container Room which he thought had been filled in by Mr Whalley the preceding Score supervisor. He is again vague about the details but seems to remember that some information had already been filled in.
After procuring the permit application Mr Rankin indicated that he took it to Mr Curtis for approval. It is perhaps also a measure of his lack of familiarity with the procedure that he had to consult his check list before taking this step. Mr Rankin was able to confirm that the application was for a cold work permit. Mr Rankin does not remember at any stage going to Mr White to have the permit requested. Of course Mr Smith may have taken this step for him. Mr Curtis is unlikely to have signed the application until Mr White signed it and although the copy permit application which was recovered and produced (number 12/233 of process) does not show any decipherable indication of Mr White’s signature this may be because it had been damaged in the catastrophe. In the case of permit applications the Approval Authority keeps a copy and this is the copy that was produced. It was recovered from the accommodation area of the platform wreck. As Mr Rankin proceeded with his evidence the impression that he had little recollection of certain details was reinforced. Mr Lynch speaks to the fact that Mr Rankin came to see him between 8 and 9 am on the morning of the accident and mentioned that the Condensate Injection Pump would be available that morning so that he wanted permission to do the valve work. Mr Lynch sent him to Mr Smith to get the necessary paper work attended to. Mr Rankin does not remember any visit to the Control Room that morning but I have little doubt that Mr Lynch has a more accurate recollection. He also had Mr Flook there at the time of Mr Rankin’s intervention. One problem however with Mr Lynch’s account is that the copy permit shows a signature of the Approval Authority which purports to have been appended at 7.40 am so that Mr Lynch may not be accurate as to his timings, or as he suggests, Mr Smith may have taken the application for a permit to Mr Curtis at an earlier time. Mr Lynch does not remember seeing the completed application before he withdrew about 9.45 am to leave the platform so that if anyone signed the permit as Designated Authority it must have been Mr Flook who unfortunately did not survive the accident. Even if the permit application had been presented to Mr Lynch he would not have signed and issued it until the isolations were complete. In fact it is unlikely that any permit would have been issued until the pump was ready for handover in the late morning or early afternoon.
Mr Rankin does remember that after he had obtained a signature from Mr Curtis he experienced further delay because he needed to arrange for a scaffold and this may be the sort of detail which stood out sufficiently in his memory to be retained. He did not remember going to the Control Room to have the permit issued but agrees that he must have done this. On the other hand he is not entirely consistent about this because at certain stages of his evidence he claims that he took the permit to the Control Room in the morning. Moreover, however restricted Mr Rankin’s experience may have been he was certainly aware of the need to obtain a permit to work and I think it unlikely that he would have embarked upon the work without such a permit. This view is enhanced by the fact that an application for a permit was signed by Mr Curtis. If he got a permit then a copy of this as a live permit would have been retained in the Control Room and the Lead Operators would have expected some procedure relating to this to be followed at the end of the shift. Thus Mr Vernon would have been expected to know something about the permit when he came on duty supposing of course that the permit was actually issued. Even if Mr Rankin had not been available as a witness I should have been reluctant to believe that he so flagrantly breached his obligations as to carry out work without any permit unless there was at least some evidence to suggest this possibility.
Assuming then that he had his permit Mr Rankin then had to address himself to the removal of the valve. Since the scaffolding permit could be issued before the permit for the valves I think it is reasonable to suppose that Mr Rankin had arranged this at some time during the morning as his vague recollection supports. Mr Grant gave him some help and he remembers Mr Grant closing a valve in Module C which was close to the deck level. This does not make too much sense because the only valve in Module C which Mr Grant may have closed was the Manual Isolation Valve for PSV 504 which was close to the PSV and some height above the deck. Mr Rankin thought that it was about 2pm when he began to remove the valve. He said that it was after lunch, that Peter Grant was available, and the valve was removed. The procedure followed for this is that the nuts would first be open at either side to see if gas is escaping. Once the nuts are loosened the rigging is put on the valve, the bolts are extracted, the valve suspended and then it is lowered to the deck. The blind flanges are then put on and until this is done the pipework is open. The practice seems to have been to release the nuts slowly to ensure that no gas is escaping and this is done in the presence of an OPCAL representative to ensure that there is no sudden escape of gas. Just what the OPCAL representative was supposed to do in the event of an escape was never made clear. Mr Rankin stated that his job in the operation in question was to remove the nuts and bolts and then he could depart. He remembered that a rigger was present and the valve was taken from the module on a small trolley to be carried further by crane. As another illustration of his bad memory Mr Rankin could not remember if he had participated in the removal of the valve outside the module. He remembers the crane coming and lifting the valve to the Score Container. He cannot remember how the services of the crane was secured and indeed his recollection of the valve removal is seriously flawed. However Mr Rankin seems to be certain that he personally had loosened the bolts holding the valve. Then he had gone off leaving Mr Sutton to fit the blind flanges.
Some evidence about the valve removal was given by Mr James McDonald. He was aged 56 years and was a rigger employed by Wood Group. This witness speaks to a fellow rigger Rutherford assisting Mr Sutton with the valve. Unfortunately because he was medically unfit to give evidence Mr Rutherford did not appear at the proof although neither party made any attempt to introduce his recollection by way of hearsay except in one respect. The witness McDonald said that he had discussed with Rutherford the timing of when the work on PSV 504 was carried out and Mr Rutherford’s recollection was that the work had not begun until the afternoon. It was Mr Rutherford and not McDonald who was directly involved in the work. Mr McDonald had himself assisted Mr Sutton on other occasions. Mr McDonald claims that about 8am Mr Sutton came to him in the Gas Module and asked for help in the removal of a valve. This timing must be wrong since the pumps had to be run and Pump A de-pressurised before any work on the valve could proceed. Mr McDonald was not free to assist Mr Sutton at that time but he sent another colleague. When Mr McDonald visited Module C later in the day the valve was already down on the deck and the only persons at the scene were Mr Rutherford and Mr Sutton. That is to say Mr McDonald confirms that at that stage Mr Rankin was not present. Again the timing is doubtful for Mr McDonald claims that his visit to Module C was at "dinner time" which he said was about 12 o’clock. The meal lasts for about an hour and cannot be taken late. The valve was put in a barrow and taken outside the module to be moved further by crane. If Mr McDonald’s recollection is accurate it is perhaps rather surprising that Mr Rankin was not present supervising when the valve was being transported outside but there may of course be some explanation. Mr McDonald was initially adamant that he had only gone for his "dinner" after he had attended to the transport of the valve and he thought that this operation would have taken about 20 minutes. Later however he was prepared to admit the possibility that his timings could be wrong. It would have taken about three hours to calibrate and quality test the valve after it had been taken to the Score Container. The whole job of removing the valve from its initial location on the deck and transporting it to the Container may have taken about half an hour. Mr Rankin states that the work on the valve did not begin until after lunch but he is not a reliable witness on matters of detail. Like Mr McDonald he accepted that he could have been wrong on matters of timing. A starting time after lunch would certainly be more consistent with the time needed to test Pump B and isolate Pump A before work on the valve could commence.
It should be noted that Mr Rankin never returned to the valve site to check the blind flanges after they had been fitted. I do not say he was remiss in this respect for the matter is not raised but of course it does mean that if the flange had been fitted improperly no-one would have noticed. He remembered the OPCAL Quality Assessor coming to the Score Container to check the valve about 5.30pm. If Mr Rankin is more or less right in relation to timings this would mean that the valve was ready for replacement about 6pm whereas with Mr McDonald’s timings it could have been ready by 5 o’clock or earlier.
Mr Rankin claims that when the valve was finished it was quite close to the end of the shift and that may be the sort of detail he would remember. He claims that he remembers that he returned the permit to the Control Room and thought that this happened about 6pm. He had gone there not only to see about the permit but to see if he could get a crane to move the valve that evening. His evidence is that there was only one person in the Control Room but if his timings are accurate it is unlikely there would only be one person there at that particular time. Indeed Mr Clark was in the Control Room for his changeover about 6pm and states that it was busy at the time as was always the case between 6pm and 7pm. Mr Rankin cannot remember whom he saw in the Control Room nor if he had dealt with that person earlier in the day. Of course if he only saw one person it is always possible that the person he saw was not the Lead Production Operator but the Control Room Operator. On the other hand he claims that he was told in specific terms that there would be no crane available that night and it may be unlikely that the Control Room Operator would give him such information. Mr Rankin also says that his not very certain recollection is that the permit was thereafter suspended. The Control Room Operator would not have done that. Accordingly although the evidence is not entirely convincing I think there is at least a fair possibility that Mr Rankin did return a permit to work to the Control Room and there saw Mr Vernon or Mr Flook depending on the time of his visit. After all because Mr Rankin had little experience of being a supervisor he was working off a checklist in the work container and it is unlikely, even were it necessary to remind him, that this would not clearly relate the need to return permits to the Control Room at the end of shifts. If others were present he may simply have forgotten about this. Mr Rankin’s conversation in the Control Room was said just to have lasted a few minutes and he says that thereafter he left the permit on the Control Room desk. Whichever of the two Lead Production Operators Mr Rankin saw would have put a suspended permit among the pile of such permits rather than retaining it as a live permit. Mr Rankin’s evidence is that when he left the Container to visit the Control Room he had left Mr Sutton there talking to two others. After being in the Control Room Mr Rankin returned to the Container and Mr Sutton was still there. I mention this evidence because otherwise it may have been a possibility that when Mr Rankin went to the Control Room hoping to get a crane Mr Sutton had begun to loosen the blind flanges in preparation for the refit. However in relation to this evidence Mr McDonald raises another point about timing. He claims that he had visited the container and been told by Sutton and Rutherford that a crane would not be available. Apart from other considerations Mr McDonald is likely to have finished his shift just before 6pm. Mr McDonald places his visit to the Score Container at about 5pm (although he later modified this to bring his visit nearer to 5.30pm) and he has a point of reference in relation to time in respect of his tea-break which was about 4pm.
Mr Rankin stated that at the time when the valve came to being replaced he and Mr Sutton would have taken off the blind flanges together.
Mr Bollands when asked about his recollection of events in the Control Room at the changeover of shifts on 6 July stated that as from about 5.30pm the Control Room was rather busy as would normally be the case. Mr Bollands said that after he came on duty about 5.15pm he would acquaint himself with what was going on and then retreat to a coffee table in the Control Room to have a cup of coffee. This could put him out of sight of persons coming into the Control Room. Mr Bollands also indicated that the Lead Production Operator would normally carry out his inspection before 6pm. This would mean that if Mr Rankin had come to the Control Room just after 6pm the normal inspection may have finished. Of course if Mr Vernon went to perform his inspection between say 5.45pm and 6pm it is just possible that Mr Flook had delayed to finish his shift and would have dealt with Mr Rankin.
The defenders made the point that it was only in respect of his timings that the pursuers challenged Mr Rankin’s evidence and this was not disputed.
Mr Rankin claimed not to remember the person he had taken the permit to at the end of the shift but the defenders say, as they can, that he had had some previous contact with Mr Flook whereas he may never have met Mr Vernon who had only come on to the platform that day. Another point made by the defenders is that when Rankin went along to the Control Room he did not know if his permit would be suspended or extended. His hope was it would prove possible to extend it and finish the work that evening. Permits were only extended at the end of the shift so that he may not have seen any point in visiting the Control Room before the end of the shift. Thus for example in number 15/4 of process we find a permit to work for a PSV refurbishment that was extended by Mr Rankin on 3 July 1988. This permit was extended at 6pm until 9pm. There is another permit in that bundle of productions which shows another extension of a permit by Mr Rankin. This however is not a conclusive point for if on the day of the accident he had done all he needed to do on the valve rather than waste time he may have tried to get a crane as soon as possible so that he could get on with the work.
Unfortunately Mr Rankin’s evidence is rather indecisive in relation to his return of the permit to the Control Room. He does make it clear that he still had some hope of obtaining a crane so that the work could be finished that night. It is just possible that when he went to the Control Room, if it was a little early, that Mr Bollands was there on his own. He indeed claims that there was only one person in the Control Room. If he had merely seen Mr Bollands he could not have had much success in regard to commandeering a crane. If the person alone in the Control Room was Mr Bollands it is unlikely that he would have given Mr Rankin the negative information about the crane that Rankin reports. Indeed Bawdens were the agency which made cranes available. I suppose it is possible that in the absence of a Lead Operator Mr Rankin took the permit away until he found out what was going to happen in respect of the crane. If Bawden’s had told him no crane was available he may have gone off to find Mr Smith to see if anything more could have been done. Mr Smith may have been one of the few Lead Operators Mr Rankin as an inexperienced supervisor knew. He certainly seems to have sought out Mr Smith. It is plain that Mr Smith was only interested in suspending the permit (if it was not already suspended) since he inquired about the blind flanges. In that kind of situation Mr Rankin may have neglected to return once more to the Control Room with the permit. The possibility I have been discussing is extremely conjectural and is not consistent with Mr Rankin’s recollection that he left the permit in the Control Room. However it was clear from the content of his evidence that whatever the reality he does not have a precise recollection of how he acted at the end of the shift in question. His rather confused version about what took place in the Control Room is not consistent with what might have been expected. The peculiarity of this case is that however possible it seems that there is merit in the defenders’ contention that Mr Vernon should have known of the absence of the PSV there are in my view strong reasons to believe that he did not have this knowledge at least at the time there was pump failure. It follows that the precise cause of his ignorance at the critical time must in fact be at best somewhat conjectural. The high point of Mr Rankin’s position in relation to the suspension of his Permit may be contained in his observation "my understanding is that we suspended the permit". Of course if Mr Rankin did not return the permit to the Control Room then someone should have noticed that the live permit had not been cancelled, extended or suspended. If this situation in fact arose it is difficult to know if there was a deliberate departure from practice or simply an act of inadvertence.
Given the procedure that had been laid down for Mr Rankin and with which he was at least to some degree experienced it remains eminently possible that Mr Rankin returned the permit to work to the Control room for suspension. However doubts about the matter remain. Just what happened to the suspended permit if it was returned to the Control Room is difficult to discover. Unfortunately Mr Rankin (no doubt because of the shock of the accident) was unable to give any convincing account of what he had done with the permit at the end of his shift. If the permit had been left in the Control Room then Mr Vernon must I think be the person who at that stage dealt with it. However some hours later he was not aware of the existence of the permit and although there could be a number of reasons for this, one might be that he had never seen the permit.
7.1.4. The Manual Isolation Valve
It was the procedure that if a Condensate Injection Pump was closed down the Manual Isolation Valve on the relief line was closed as an additional isolating factor. Once closed they would normally be locked and chained (in this connection it is perhaps worth noting that there was no evidence as to who kept the key or as to how long it would have taken to procure this if the valve had to be re-opened). As I have already observed this valve was approximately at the same height as the PSV and about 18 inches away from it. Obviously the isolation valve, if it had been closed, would have to be opened before the pump could be re-started. Failing this the PSV could not function and there would be no pressure relief on the pump. There was evidence of a practice to open the manual valve before jagging to re-pressurise the pump but it was not clear that everyone would have followed, or was expected to follow, this sequence in respect of the valve. If Mr Vernon himself had opened the valve before carrying out the alleged jagging of the pump then he would have seen that the PSV was missing. I would accept that a person manually operating the Manual Valve would be unlikely not to notice the absence of the PSV. Equally if Mr Richard had gone to open the Manual Valve then he to would have seen that the PSV was absent and might have been expected to tell Mr Vernon. However it should be noted that there was no direct evidence to suggest that either Mr Vernon or Mr Richard had in fact opened the Manual Valve at the time of any jagging. There was no evidence of any express direction to open the valve before jagging and there was no evidence that it would have been dangerous to keep the valve closed until after jagging. Indeed the witness Mr Murray an experienced Operator saw no danger in re-pressurising the pump before opening the Manual Valve. The PSV would be set at a much higher pressure than the system pressure which would be achieved at re-pressurisation (the system pressure would be about 650 psi and the PSV would be set at about 1750 psi). Mr Seddon the Operations Superintendent agreed with this. It is only when the pump starts running that pressure could be built up. Because of the circumstances there would have been little time between the tripping of the pump and the alleged jagging. During this time Mr Vernon was also engaged in trying to restart Pump B. The precise order of procedures may have depended on the number of persons available for the re-pressurising process. As it happens if Mr Vernon jagged the pump Mr Richard had already been called away before the vital second jagging procedure. Mr Henderson had said that normally three men were required for a re-pressurising operation and his reference to the practice of opening manual valves possibly related only to an operation with that number of men. Mr Murray said that he would only send someone to open the Manual Valve at an early stage of re-pressurising the pump if he had the men available. Moreover we were not told very clearly how an operator was expected to gain access to the Manual Valve which was about 20 feet above the deck as was the PSV. Perhaps the PSV scaffolding could have been used as an access but this is by no means clear. Possibly a ladder or some other equipment would have to have been arranged. In any event if other factors point to Mr Vernon not knowing or remembering about the PSV I do not think the requirements relating to the Manual Valve can be taken as excluding the other grounds of inference.
The witness Mr Henderson spoke to the fact that there was a practice whereby in re-pressurising the pumps the Manual Valves would be opened as a first step. The defenders objected to any attempt to lead evidence from Mr Murray and Mr Seddon that there was no such practice on the grounds that Mr Henderson had not been challenged on the point. In the interests of fairness I sustained this objection when first made with the result that the pursuers would have been precluded from leading any evidence to the effect that there was any other established practice. Other objections in a similar vein were reserved. There was some discussion as to whether my ruling related to the sequence of de-pressurisation or re-pressurisation or indeed to sequence at all. The pursuers argued that in fact the objections covered the circumstances in which the Manual Isolation Valve would be used rather than precise sequence. However even if the objections can be taken as relating to the sequence of re-pressurisation as I have indicated there was no evidence as to the source or scope of the practice. It may be that the sequence of events spoken to was generally more convenient. As I have said no reason was advanced as to why individual operators could not have deviated from the practice if it were more convenient to do so.
It should be noted that Mr Henderson, a reasonably impressive witness, opined that in the kind of situation where one pump had tripped and there was an emergency attempt to de-isolate a second pump the Lead Production Operator (whom he supposes has two other men available) would be "getting that pump pressured up to run whenever we had electrical power available". This of course is what the pursuers claim that Mr Vernon was doing.
Mr Murray informed us that the Manual Valve is operated by a straight bar handle so that once access was gained the actual opening should take no time at all. However Mr Murray could not say anything very useful about the Manual Isolation Valve since he had never operated it. A fact confirmed by Mr Murray was that Mr Vernon would have about one hour before the Suction Vessel would be liable to overflow if the flow of condensate were not stopped. He also said that the situation confronting Mr Vernon and his response to it was similar to what had happened on a previous occasion. However I do not think in this observation he was including an absent PSV.
Mr Seddon, an Operations Superintendent on Piper Alpha, had considerable hands-on experience of the Production process. He had in particular considerable experience of operating Manual Isolation Valves. In fact from his evidence it appears that the de-pressurisation and re-pressurisation of equipment such as pumps is a very common procedure indeed. He said that invariably if a pump trips it is a "very, very minor fault". An important observation of Mr Seddon is that if he had required to start a pump which had been with maintenance he would ascertain the status of the pump from the Lead Maintenance Technician. He does not suggest for example that he would look through the permits to work. One point he makes is that before starting the work the permit for the maintenance work should be cancelled for it had not been completed. However in relation to the Manual Valve the point he makes is that there is a considerable difference between jagging with the Manual Valve still closed (the pressure in these circumstances being comfortably within the tolerance of the system) and with the pump running where pressure surges beyond the capacity of the Manual valve might be expected.
The Defenders contended that Mr Vernon might well not have arranged to open the Manual valve which of course would be consistent with a position where he had realised that there would be no point in opening the manual valve with the PSV missing. On the other hand if Mr Vernon had discovered through an attempt to open the Manual Valve that PSV 504 was missing (assuming that he had for some reason forgotten this) the knowledge may explain why when Mr Grieve came on the scene he saw that Mr Vernon had apparently switched his attention to Pump B. But as other evidence shows the main effort could always be expected to be to try to re-start the pump which had tripped.
7.1.4. Handover from Flook to Vernon
The defenders are undoubtedly correct to claim that it was the incoming Lead Production Operator’s responsibility to familiarise himself with the state of the plant and of anything in the preceding shift that could affect the safety of the ongoing production process. Thus Mr Vernon should have had a comprehensive handover from Mr Flook and should have also had an opportunity to discuss the state of the platform with the Operations Superintendent. Since neither of the parties who would have been involved in the latter process have survived it is speculative to consider what, if anything passed between Mr Vernon and Mr Curtis. In relation to PSV 504 Mr Curtis may not even have been aware of the up-to-date situation. Thus I doubt if it is fruitful to bring Mr Curtis into the matter. In any event if Mr Vernon was likely to hear about the PSV from anybody one might expect that it would be from Mr Flook. Mr Keen contended that it was the duty of the incoming Lead Production Operator to familiarise himself with permits that had been suspended during the shift and in a general sense this too must be correct. Moreover the defenders pointed out that Vernon had not only come onto a fresh shift but was actually commencing a new tour of duty so that it would have been particularly necessary that he should thoroughly familiarise himself with the state of the operations he was responsible for. It was also to be his first encounter with the switch to Phase 1 and this may have dominated his thoughts. It was said that the late Mr Flook was a conscientious and meticulous workman and I do not doubt that this is true but all the victims including Mr Vernon were given first class "references" from their surviving colleagues and one could hardly expect anything else. Because of the nature of the situation I should not like to make any vital decision on the basis of any such evidence alone. Moreover I can accept that the likelihood is that Mr Flook knew that the PSV work had been going on and at the point he came off shift he may have had no reason to believe that the work had not been completed. To find out about this he may have been awaiting the return to the Control Room of the PSV permit. In any event this may not have worried him if he thought that the pump was going to be immobilised for a number of days. The pursuers in their pleadings "believe and aver" that Mr Flook knew that the PSV had been removed. However knowing about the valve does no necessarily mean the information was thought to be important. Moreover if Mr Flook had been keeping his log with reasonable accuracy the fact that the PSV had been worked on that day should have appeared in his log. Thus Mr Flook is likely to have given Mr Vernon material from which at least it was possible to derive the information that the PSV had been worked on that day. Whether or not Mr Flook specifically discussed the valve work with Mr Vernon is a matter of conjecture. As I have already indicated Mr Flook may well have thought that the work on the valve would not be of great interest to Mr Vernon since it was expected that work on the motor would commence that night and the possibility of the pump being recalled may well have seemed rather remote given its condition. Since there was a live permit for the PSV work Mr Flook may have considered that given that he personally may not have been sure if the work had been completed Mr Vernon would find out about the pump more informatively when the permit was presented to Mr Vernon for extension or whatever. However the pursuers accept that Mr Flook could be expected to have given Mr Vernon some information to the effect that work on the PSV had been proceeding that day. Mr Grant undoubtedly knew about the PSV work since he had participated in it and the defenders founded on the fact that he would have recorded such work in his own log. There was uncontested evidence that he would have done so and it may have been surprising had he not. The pursuers asked for a specific finding in fact that in accordance with normal and proper practice Mr Flook would have handed over to Mr Vernon information that PSV 504 had been worked on that day. The defenders further argued that the removal of the PSV was not simply a matter of a calibration exercise that might have been of little interest to the Lead Production Operator but represented the removal of a major piece of equipment that would have become very significant should the need arise to bring the pump precipitately into commission which was in fact what happened. Indeed the objective of the permit to work system was to make sure that the Lead Production Operator was aware of everything going on at the plant which might affect production. There seems little doubt that if a PSV was removed during a shift and was likely still to be absent at the end of the shift the outgoing Lead Operator would normally tell the incoming Operator or hand him a log containing the information. On the other had as I have said the outgoing Lead Operator might form the view that the up-to-date status of the valve would inevitably be conveyed to the incoming Lead Operator when the Permit was returned to him at the end of the shift (assuming as he would that the permit would be returned). There was certainly uncontradicted evidence that the work on PSV 504 would, if proper practice had been followed, be recorded in the outgoing Lead Production Operator’s log. The defenders also wanted a finding that the removal of the PSV would be recorded in the Phase 1 Operator’s log which would be handed over to the incoming Phase 1 Operator, in this case Mr Richard. Mr Grant had certainly spent time assisting in the removal of the valve so that some reference to this might have been expected although just what would be said is more conjectural. Defenders’ counsel pointed out that the relief pipe beyond the Manual Relief valve was still connected to the Condensate Suction Vessel which remained a significant part of the ongoing production process so that it was important for a Lead Production Operator to be aware if he could safely open the Manual Valve. It was said that one would have expected Mr Vernon to take particular care to ascertain the state of the plant because he was beginning a new tour of duty. Of course this could be a double edged point since to begin a new tour of duty could mean that there was more than normal to remain mindful of. The defenders argued that the visit by Mr Vernon or Mr Richard to unload and recycle the Reciprocating Compressor Pumps would have taken them to Module C and the scaffolding would have been noticed. I do not think that there is much doubt that these pumps had been unloaded and recycled. Mr Vernon said so to Mr Bollands when he last returned to the Control Room and the matter is consistent with Captain Clegg who noticed a sudden increase in flaring minutes before the accident.
There was of course evidence that because the Pressure Control Valves PCV 1000 A and B did not always set effectively it was practice to adjust the larger valve B if recycling. This valve had a tendency to "hunt" which meant that the valve struggled to find its proper setting. The valves allowed the gas released by the unloading to pass through them to flare. The object was to drop the set point of the valve. Such resetting would have taken the person doing it quite close to the absent PSV and on the same gangway. The problem is that although it was obviously quite normal to adjust PCV 1000 B this was only superficially explored. For example it was not discussed in the evidence whether this was a critical requirement or merely a preferred practice. Mr Bollands seemed to suggest that it was a convenience rather than a necessity. The question of course is whether the PCVs would normally be readjusted if the recycling and unloading had to be accomplished in circumstances of some urgency. In addition it is not at all clear how much of the operator’s attention the valve adjustment might have been expected to take. It is not even clear that it was necessary to go to the valve itself to effect any necessary adjustment. Moreover the valve which had to be reset was located much lower than PSV 504. The evidence raised the possibility that in Phase 1 the valve was not operated at all in which case it would not require adjustment. I do not think too much help can be extracted from the demands of the valves in question. For the recycling operation itself 14 switches had to be thrown. For Phase 2 operation it would be necessary to go to both the PCVs and reset them but with Phase1 because some gas is already passing through PCV 1,000 A it would only be necessary to reset B. It should also be noted that Mr Henderson was somewhat tentative in suggesting that in Phase 1 PCV 1000 A could cope with the normal escape to flare alone. He personally had no direct experience of resetting these valves during Phase 1. It is accordingly also not clear that Mr Vernon or Mr Richard would have had experience of what, if anything, was required of these valves during Phase 1 and the same could be said of Mr Bollands.
The defenders strongly maintain that Mr Clark did not know of the valve maintenance work. However Mr Smith had to a degree been involved in the original instructions to the valve technicians to commence the PSV work and in accordance with the normal procedures spoken to by witnesses the valve work was an incident of the day’s work that Mr Smith might have been expected to record in his log. Indeed Mr Rankin thought that Mr Smith was the person to whom he had to report. If the material was in the log then this should have been handed over to Mr Clark at handover. Thus either Mr Smith’s log was not entirely complete, or Mr Clark had not registered all the material in it or had forgotten it. The only significance of this is that it cannot be assumed that the procedural system will always work effectively in every instance.
There was a point made by the defenders that I have to weigh in my mind in deciding this aspect and indeed all aspects of the case. That is that if the pursuers’ case at any point depends on hypothesis and I make one finding in fact that is inconsistent with that hypothesis then the hypothesis fails. In general as a matter of logic as well as law that must be true. However although the rule would normally be applicable in a circumstantial case it is difficult to apply in the abstract. A factual finding which would otherwise be attractive may be displaced as an available possibility because the other evidence in support of a particular hypothesis is so strong that it can only be supposed that the otherwise contradictory fact cannot be justified.
7.1.5. Conclusions on Mr Vernon’s Knowledge
If the position is that the accident was caused by the fact that Mr Vernon was attempting to put Pump A into Production then the defenders would have a good defence to claims under the Indemnities if they could show that he was acting deliberately in the knowledge that at the time PSV 504 was not in place. There are difficult questions as to what he may have known at certain stages of his shift but I think since it is his state of mind at the time he would have been jagging the pump that is critical, it assists to begin at that point. The question of onus of proof could arise but since I see a reasonable way through the difficult tangle that surrounds this matter I do not require to found my determination on onus. Thus onus is a matter I need not at this point decide although my initial impression is that the onus rests on the defenders to prove wilful misconduct. The alternative would be that the pursuers had to prove a negative.
My starting point must be that it would have been surprising if Mr Vernon had been prepared to ignore safety procedures to the extent that he was prepared to run the pump without a relief system to cope with any build up of pressure. Claims by his former workmates that he was exceptionally conscientious must of course be viewed cautiously but there is certainly no indication in the evidence that he might be capable of downright irresponsibility. It is clear from the evidence that all the Lead Production Operators and the ordinary Operators were well aware that it was bad practice to run a pump without a PSV in place. Indeed there was no suggestion from the considerable number of qualified workmen who gave evidence that such a thing has ever been done previously. There is of course always a first time but one would have to ask why Mr Vernon would have taken a risk. Personally he had nothing to gain. He had indeed a lot to lose because the course he is alleged to have followed was very exposed to discovery by his superiors and could have attracted a severe reprimand or worse. If Mr Vernon did consider that the exigencies of the situation he was faced with might justify a departure from reasonably safe procedures then one would at least have expected that he would discuss the problem with his colleagues. He said nothing either to Mr Bollands nor to Mr Clark both of whom he could readily have consulted. Moreover supposing that in fact Mr Clark was unaware that the PSV had been removed equally it would not have been surprising if he had known. He might have noticed the missing PSV while walking around the platform or Mr Smith might have told him at the handover. Thus Mr Vernon would not have known what Mr Clark would have known about the PSVs and yet allegedly he blithely ignored both Mr Clark and Mr Bollands on the matter. Furthermore Mr Richard would have found out that the PSV was missing when he went to release the Manual Valve if he did not know already. The prospect of losing Production was serious but would probably have been confined to the loss of condensate production since condensate would have had to be released to flare if the pumps had continued to remain disabled. However if Mr Vernon was concerned that he would lose production it is remarkable that he did not consult the various superiors in the management line who were on the platform. Mr Lynch indicated that it was acceptable, if necessary, to wake them out of their beds if there were a breakdown. This would have relieved Mr Vernon of the responsibility were indeed anyone likely to take a decision to run the pump without the PSV. Of course it is unlikely that any of his superiors in management would have authorised the use of Pump A so that there was little risk of his being criticised for not putting that pump into production. There were many routes by which senior management could have found out about Mr Vernon’s irregularity. Moreover there is a further consideration. The defenders’ contention is that Mr Vernon would have known that the valve maintenance was incomplete and suspended. If Mr Vernon had restarted the pump A in these circumstances then there was always the prospect that next shift the valve fitters would seek to resume what they were doing. This would mean removing the blind flange with condensate running through the system. Of course if the permit to work system had worked properly the valve permit would not have been renewed but there was clearly an undesirable risk of a mistake or misunderstanding which simply highlights the risk of running the pump with the PSV missing.
The defenders contend that although it was bad practice to run the pump without a PCV it may not have seemed such a great risk to Mr Vernon. The valve is only needed if the pressure builds up and there are additional safety devices including safety trip devices that could have relieved the flow pressure if the pressure began to build up excessively. However the immediate risk of pressurising the pump was not only that when the pump was in use the pressure valve would not be there as a relief. The relief line had been opened up and was only protected by a blind flange. This may have sufficed if the flange had been pressure tested. This was the procedure with flanges in use on an active line. There was no suggestion that Mr Vernon had made any arrangement to have the flange pressure tested and indeed we are not even certain how quickly this could have been done if desired. Of course if the line had been pressure tested any deficiency in the fitting of the flange would have been discovered. As soon as the pump begins to be pressurised one is dependent on an untested flange to contain the condensate.
One question is the origin of the red de-isolation tags which indisputably Mr Vernon released to Mr Clark to sign off. As between Mr Bollands and Mr Clark I formed the view that Mr Bollands had in general been the more accurate observer but it would be remarkable if any witness could be taken as totally reliable in all matters of detail and this must include Mr Bollands. Certainly a point was made that if Mr Clark knew that the PSVs were missing he would have had a lot to answer for because he released the electrical isolations. I have little doubt that he would have known that it was bad procedure to run the pumps without PSVs. It was also argued that Mr Bollands would have had an interest to cover his tracks by denying all knowledge of the PSV maintenance but this is less obviously so since in the presence of Mr Vernon and Mr Clark he could have had little direct responsibility. However although Mr Clark’s evidence has to be read with a certain caution and is not in all respects convincing I am by no means in a position to suggest that he was consciously misleading the Court. Both Mr Clark and Mr Bollands suffered frightening experiences when the accident occurred. Mr Clark got blown across the Control Room and was injured. With such a shock it would perhaps be expecting too much for Mr Clark in particular to have a reliable recollection of detail some years after the accident.
According to Mr Bollands a permit for the planned maintenance work had been issued and it was this pink permit that Mr Vernon took out of the live permit slot to present to Mr Clark for signature. Mr Clark on the other hand states that the permit application and tags were in the Safety Office and the permit had never been completed and issued. I doubt if either witness is entirely accurate. Since the maintenance permit was Mr Clark’s responsibility it would indeed be surprising if he did not know about its issue. The matter would have inevitably been mentioned to him at handover and one would expect that he would have been responsible for having it extended if it were to be a permit in the live permit container. Moreover it would not have been extended unless work was proceeding during the evening shift and it would be surprising if Mr Clark did not know about this since he would have been responsible for it. Mr Clark at least purports to remember that Mr Smith had told him specifically that the Permit had not been issued and this account of what he had been told may well be true. However, whatever Mr Smith said about the whereabouts of the permit application, I doubt if this was in the Safety Office even if the permit was at the stage of only being an application. There are parts of Mr Bollands’ evidence where it is possible his memory is not accurate. However there are certain details which are more likely to be accurate than others. As Control Room Operator Mr Bollands would have required to know the work proceeding during his shift. Thus he may well not have looked at the suspended permits which would have included the one relating to PSV 504. On the other hand he remembers looking at a permit which certainly gave him the impression that the Pump was undergoing planned maintenance and this he remembers was a pink permit (as it would be). I think when he came on duty he must have looked at something which related to planned maintenance. He claims that this was a live permit but if he is wrong about the live permit coming out of the live permit container the reality may be that what he saw was the permit application. I think that despite Mr Clark’s account of what he had been told the permit application (if there were such) was left lying in the Control Room ready for Mr Clark’s immediate attention when he came to proceed with the Voith Coupling work which he hoped to perform during the nightshift. There was evidence that sometimes after the necessary isolations had been carried out a permit was left lying in the Control Room for issue. The situation may be quite different when it is not known precisely when work will proceed but this was work where urgent completion was desired particularly after isolations had been completed. It is possible that Mr Clark had retrieved the permit himself from the Safety Office. He was challenged as to whether or not he had been in the Safety Office after coming on duty and his evidence was rather vague on this. However it got to the Control Room I think it was there when Mr Bollands looked through what he conceived to be permits relevant to his duties and also when Mr Vernon looked for the maintenance electrical tags. It is of course suggested that what Mr Vernon must have done was to take his tags from the PSV permit but I think this unlikely for reasons I shall expand upon shortly. If the permit had not in fact been issued then I would conclude that the permit form with its tags was available at the relevant time in the Control Room. It is strange that Mr Bollands had the impression that instrument work had been carried on at the pump but since no other evidence supports this I cannot exclude the possibility that this is simply some quirk of his memory. That I think must be so since it would be an odd tale to make up.
One circumstance that Mr Clark accepts without hesitation is that the original electrical isolations had been effected for the planned maintenance proposal and had been completed. This would make sense for once the pump had been isolated by de-pressurisation one would not want someone accidentally switching on the pump. The defenders suggested that what Mr Clark was saying is that the permits had been prepared ready for signing but the isolations had not taken place. However this does not at all accord with Mr Clark’s evidence. He claims not to have known about the PSV work or he would not have signed off the tags. Indeed when he signed the red tags he would not according to his evidence have known that they could possibly relate to valve work (since he claims not to have known about this) and must have had grounds for believing that the maintenance isolations had been completed. Otherwise what tags did he think that he was signing. The defenders’ Counsel attempted to draw some particular conclusions from general evidence that the electrical isolations would not be effected until the work was about to start. However in relation to the Maintenance programme all the evidence was to the effect that the electrical isolations necessary for this programme had been completed. The work was not simply a single job but was expected to be extensive and it was anticipated that some might begin during the night shift. In particular it was expected that the Voith Coupling work would be carried out during the night shift and although this on its own may not have required a hot work permit the work was to be carried out under the maintenance programme permit which would have required electrical isolation. Mr Clark was an experienced Maintenance Lead Hand. He claimed that the Maintenance permit had not been issued. Nevertheless he signed off tags which he assumed related to the maintenance permit. This I think shows that any practice of electrically isolating within a very short time of the work beginning was not in all circumstances followed particularly perhaps if it was part of a more substantial isolation process. The position was therefore that under the maintenance application the electrical switches would have been immobilised and sections of the tags attached thereto so that the power could not be restored unless the electricians were presented with discharged isolation tags. This is very important since it means that the electricity could not be restored unless the tags attached to the maintenance permit or application had been signed off by Mr Vernon and Mr Clark. Mr Clark accepts this and goes beyond it for as he claims he would not have signed off tags (which he assumed were Maintenance permit tags relating to the pump) if he had known that PSV tags were in force. Mr Clark would have had no obvious authority to sign off any valve tags since the performing authority for that work was Mr Rankin. There was a preponderance of evidence to the fact that a PSV Maintenance permit would be a blue cold permit and many (though not all of the relevant authorities) would require electrical isolations. The witness Mr Lloyd thought that the normal practice was to have electrical isolations for PSV calibrations. The Safety Procedure Rules specifically provide that each permit will have the tags relevant to it attached to it. This alone makes sense and there was no evidence that the operators ever departed from their prescribed safety rules. For example if only one set of tags were provided for two distinct jobs requiring isolation at the same time then if these were attached to one permit the risk would exist that when work relating to one permit was completed and the permit cancelled the electricity would be restored while the other work is proceeding. This approach can be illustrated in relation to the Planned Maintenance Permit. Although the various items of work were likely to be proceeding at the same time separate tags were required for the pump motor and the lube oil motor. Thus Mr Clark obviously knew what he was talking about when he said that if he had known that a PSV permit with electrical tags were out, there would have been no point in signing off tags for the Planned Maintenance Permit. The same would apply if the PSV permits alone were signed off. The point is that there would be attached to the switch lock two sets of tags, the tags relating to the two pump motors being one set and the tags relating to the PSV being the other. Then electricians could not restore electricity unless they had authority to discharge all the tags attached. I do not think it was suggested that Mr Vernon was proposing to override the whole tags system. The trouble he was taking to get Mr Clark’s signature indicates the opposite.
Everything about Mr Vernon’s communications in the Control Room suggests that he was concentrating on the planned maintenance situation. There is no suggestion that he visited the Safety Office at any time. Even if we suppose that Mr Vernon must have visited the Safety Office then on the basis of the defenders’ submissions he would only have found the maintenance tags there and these alone would not have been effective if he knew that there were two separate sets of isolation tags. When Mr Vernon spoke to Mr Clark on the telephone his first enquiry was about the position of the planned maintenance. He was apparently keen to get the pump back from that. Moreover his question to Mr Clark was somewhat curious if the position were that the maintenance permit had not been issued. He first asked what the status of the pump was and if work had been done on it. This would be an odd approach for a Lead Production Operator if there was no permit issued for he would know that no work at all could proceed without a permit being issued. Thus either there was a live permit or he had forgotten that the permit had not actually been issued. He did not say to Mr Clark that there was a PSV permit outstanding and that he would like these isolations signed-off (assuming indeed that the PSV permit had provided for such isolation). This is what one would have expected if he knew there was no planned maintenance permit and was standing with the PSV tags in his hand. Instead the whole thrust of his short conversation with Mr Clark seemed to be directed at planned maintenance. He seems certainly to have known about the proposed planned maintenance. He would not have asked if work had been done if he had not known about the isolations and this is what one would expect since that must have been an important element in Mr Flook’s handover. The various steps to isolate the pump must have figured prominently in Mr Flook’s log. However if Mr Vernon had it in his mind that the PSV permit was alive (with its isolations in force) he must have realised, just as Mr Clark did, that there was no way the electricity could be restored without releasing both sets of electrical isolations.
The defenders placed great store in the effectiveness of the procedures in insuring that the Lead Production Operator was at all times well aware of the state of the production situation. However at best for the defenders Mr Vernon must have inadvertently forgotten at least one important fact. Either it had escaped his mind that there were planned maintenance isolations or it had slipped his mind that there was a PSV permit. He could hardly have supposed that he could electrically de-isolate both permits unless he had access to all the tags. Moreover as I have said when he spoke to Mr Clark the emphasis in his conversation was on Maintenance and one would have supposed that the conversation about the red tags related to that. This clearly is what Mr Clark thought. There was certainly no evidence that Mr Vernon got hold of all the tags. If he had wanted the PSV tags (assuming there were some) he would have required to retrieve the permit from the suspended permits relating to the 84-foot level. Of course this would not have bothered him if he was not at the time aware that a PSV permit and possibly tags relating to it were in existence. If Mr Clark is right and the maintenance permit application was in the Safety Office he would also have required to visit there to get hold of the Maintenance Tags. Mr Clark would have been presented with more tags than he speaks to. Mr Bollands remembers Mr Vernon having in his hand a permit with tags attached. This seemed a vivid enough recollection to be substantially true. He may be mistaken in thinking that this was a live permit or even that it came from the 68-foot permit container and not simply from the desk nearby but his belief that it was a maintenance permit could only have arisen from it being red (or pink as the colour is sometimes described) and everything points to Mr Vernon being interested in discharging the planned Maintenance and not the PSVs alone as would be the alternative hypothesis. If it is clear that at the relevant time the PSV permit must have slipped Mr Vernon’s mind (assuming that is that he had ever been aware of it) it may not matter what he had known at the earlier stage of the evening. The defenders of course approached the problem from the opposite end and contended (at considerable length) that he must have known in the early evening and possibly later but in any event it must therefore be assumed that he knew when allegedly he went to restart the pump. One problem of course is that at least in one respect the pursuers seem to accept that Mr Vernon suspended the permit. This was originally denied by the defenders and as I have discussed the averment is based on material which is contradicted by the evidence that has been allowed to emerge. However the matter is not critical. First of all if the averments govern the position this means that the pursuers are accepting that there had been a live PSV permit. I do not think that there can be much doubt about this. Mr Rankin may have had his problems as a witness but he was undoubtedly sufficiently experienced to know that he could not calibrate Valves without a permit. If at one stage he may have been confused about this because of the intervention of the planned maintenance isolations he took the trouble to ask about the position and was put right. There was considerable evidence about the timing of his progress but on balance as I have said I prefer the view that he did not finish all he was able to do on the day of the accident until about 6pm and then he took his permit to the Control Room at least for some purpose. If as the defenders contend the pursuers are bound to a position that the permit was properly suspended Mr Vernon was in all probability the person who suspended the permit and this is what the pursuers aver. As I have said the pursuers’ averments on this element of their case seem to me to involve a massive non sequitur. If the pursuers’ averments are to be taken literally the only occasion when the Lead Production Operator could under the alleged safety rules possibly have a duty to inspect a site relating to an uncompleted job would be when a permit is suspended. However this would involve the Lead Operator knowing about the state of the work in the first place for the application to suspend the permit would tell him that. It is therefore difficult to see what the inspection would add to his knowledge. Moreover certainly as far as an operator was concerned he would on the evidence not necessarily be aware that it was a prudent procedure to inspect a site if he could properly be satisfied that the work had been left in safe condition. As I have said in reality there appears to be a doubt as to whether in the early evening of the accident Mr Vernon had the opportunity to discover from a permit to work that the valve was out of position. However it is not necessary for me to rely on my view of the technical arguments relating to the pursuers’ own pleadings. I shall assume that as the pursuers seemed to assert at least at one point of the case that Mr Vernon did suspend the permit. However he derived any information in the permit against a background when he had just begun his tour and may have had many different situations to worry about. We know for example that the process had just switched over to Phase 1 from Phase 2. This was a considerable change in the process and Mr Vernon was about to experience it for the first time at least for some years. He may therefore have paid limited attention to a valve that was missing from a Pump that was in any event to be unavailable for some time. It also seems that the permit if presented to him for suspension was probably presented rather later than normal - perhaps just after six o’clock. When Mr Flook made his handover he may not have realised that the valve work would spill over until the next day. In any event if that were to happen Mr Vernon would be able to ask for the details and he may have thought that little needed to be said about the PSV particularly with the pump in planned maintenance. The same points can be made in relation to Mr Grant’s handover to Mr Richard. Mr Grant may not have appreciated that the PSV job had not finished particularly as there was no indication that Mr Rankin would report to him nor did he have any responsibility. If Mr Grant was concentrating on matters that the incoming Operator would have to involve himself with then there may have seemed little point in making reference to the missing PSV. Indeed Mr Grieve in his evidence said that the Phase I operator would go through with the incoming operator anything which started during the shift which would relate to the work of the operators. This of course would not include work handed into the control of maintenance. Indeed Mr Grieve indicated that the entry in the log would simply record that the machine had been handed over to maintenance. Mr Bollands took a rather different view of what would have been recorded in the log but even if he was right the evidence would simply be recording that the detailed implementation of the practice could vary from operator to operator. In any event we do not know just how studiously Mr Richard required to look at the log if he had an apparently full oral handover. I am not convinced that the manual valve had been opened before any jagging took place and with the switchover to Phase 1 the proper procedure for resetting of the relief valves to flare may have been ignored by say Mr Richard who may not have been familiar with how to cope with that situation. This would also be likely if Mr Vernon was working under pressurised conditions with inadequate manpower.
It was suggested that Mr Vernon may have learned about the missing PSVs after leaving the Control Room - say from attempting to open the Manual Valve or from Mr Richard. The suggestion goes wider than the question of knowledge for it is that he may have decided not to proceed with the re-introduction of Pump A. If that had been the position I should have expected him to radio Mr Clark to cancel the electricians. Mr Vernon would hardly have wanted to de-isolate the pump and then proceed to isolate it again for completion of the valve work. Regarding inspection of the pump it may well be the case that Mr Vernon had completed his normal walkabout before Mr Rankin reported to him that he required the pump suspended. We were not really told what his practice would have been if subsequent inspections were required. This may have depended on the other tasks at hand and the importance of urgency in the inspection. Thus the inspection of a valve site and the nameplate work carried out by Mr MacGregor were hardly likely to cause much danger and may have justified low priority. When Mr Smith heard that the valve work had not been fitted his only concern was that blind flanges had been fitted. In general there was a limit to what a walkabout could reveal. Mr Clark said that he would not have noticed scaffolding. There was a great deal of it about and much was concerned with matters such as painting and other projects that do not involve production. Mr Clark also revealed that PSV 504 is situated in an area where there is a mass of complex piping. There are about 800 valves on the platform and he could not even identify PSV 504 without reference to a drawing. It is perhaps interesting that none of the operators giving evidence who might have been expected to know that PSV 504 was missing seem to have known so. Mr Bollands, Mr Clark, and Mr Grieve claim not to have known. It is unlikely that Mr White, the Maintenance Superintendent knew since he was present when Mr Clark agreed to return the pump to the operating department. If Mr Vernon hoped to restart the pump in the knowledge that the PSV was missing he might have expected some resistance from Mr Richard if he had known but there was no evidence to suggest this. The problem might well have been that no-one was focusing on the pump because it was presumed to be out of commission for maintenance. Mr Vernon for his part, if he had recollected the absence of the PSV valve could not have known that all these persons did not know the position but there is no suggestion that he attempted to conspire with them to breach safety practices.
The situation on the platform could admit of many circumstances which might explain how Mr Vernon came to forget about the PSV or not to know that it was not in position. The problem for the defenders is that there is in my view fairly clear eyewitness evidence from which it can be inferred that in the final period before the accident Mr Vernon was not aware that the PSV was absent. On the other hand the evidence of what might have taken place to produce that situation is vague and leads to no convincing conclusion. Many of the possibilities remain totally speculative and we shall never know what happened at and following on the handover. Among the various possibilities the likeliest is that at some stage in the evening Mr Vernon had suspended or at least seen the PSV permit but that either in the whole circumstances surrounding the planned work on the pump it had failed to register in his mind that the state of the valve had any significance or if it did so register it had slipped his mind later.
In arriving at a view of Mr Vernon’s behaviour I am in no way suggesting that he was not seriously negligent in allowing the PSV permit to escape his mind, or in any event as the Lead Production Operator on duty not having discovered that it was absent.
7.2. Mr Sutton and the Blind Flange
7.2.1. His Involvement
There was no dispute that in terms of OPCAL’s contract with Score the latter was required to fit blind flanges to all open ended pipework. The fitting of blind flanges was also a condition of the PSV permit to work. Nor was it disputed that such blind flanges ought to have been secured tightly by use of a flogging hammer or by combination spanners. In the case of a PSV being removed the purpose of the blind flange was not only to prevent dirt or other impurities entering the pump but also to prevent any hydrocarbon remaining in the pipe from escaping. These purposes are fairly obvious and should be known to any reasonably experienced valve fitter.
The tests carried out by Mr Stanton showed that if the flange is finger tightened alone it would not withstand the pressure which would be introduced to the valve upon re-pressurisation of the pump. On the other hand if it is tightened by flogging or by use of combination spanners then it should be able to withstand that pressure. A lesser degree of tightening would permit a leak to develop.
It was accepted by the pursuers that their allegation that Mr Sutton had caused the leak by not flogging or using combination spanners on the blind flange depended entirely on inference. No-one observed him when he was actually putting the bolts on the relevant flange. Moreover most unfortunately Mr Sutton did not survive the accident so that we do not have his own testimony.
It is I think obvious that given that a properly fitted flange should not leak then if the flange did leak the probability must be that the flange was not fitted properly. The defenders’ case is that there was no reason why Mr Sutton should not have followed his normal practice and tightened the flange adequately. The pursuers say that there are three reasons why Mr Sutton may not have fitted the flange properly. One is that he expected that the PSV would be replaced that evening and therefore did not expect that tightening the flange would be important. Another possibility that the pursuers advance is that he was working in a restricted space and therefore found it difficult to use tools. The third possibility is that Mr Sutton or Mr Rankin had selected the wrong size of flange for the pipe (a 1500 standard flange for a 900 pipe). It was conceded that this alone would not cause a leak but it was suggested that Mr Sutton had found that he had the wrong size of flange then thinking that the valve would be replaced that day he may not have bothered to tighten the flange properly. In other words it is suggested that the problem with size may have deflected him from tightening the flange as carefully as he might otherwise have done.
The pursuers averred quite plainly that Mr Sutton failed to secure the blind flange securely. The defenders submitted that these were not derivative or inferential averments and that the pursuers had therefore obliged themselves to prove by direct evidence that Mr Sutton had not secured the flange properly. They are correct in stating that there is no direct evidence that Mr Sutton failed to tighten the flange fully. They argue that if the pursuers sought to rely on an inferential case they should have introduced their averment with words such as "believed and averred" to show that their case was based on inference. They should also have set out clearly the circumstances which might permit them to draw any inference they proposed to found on. The only foundation which the pursuers plead in support of their general averment is that the flange leaked so that it must have lacked proper tightening. I think at this stage of the case where the whole factual background has been extensively explored it is a narrow technical point to complain about the pleadings in the manner I have indicated. If the evidence now before me allows for an inference that the blind flange was not properly secured by Mr Sutton I see no pleading difficulty about a finding in these terms. The defenders further argue that the pursuers can only establish their case of incorrect tightening of the flange if the Court first makes a finding that the flange leaked. This point I think has merit. There is no other material from which it could be inferred that the flange was not fixed properly. However it was contended there was no direct evidence that condensate leaked from the flange. Nobody actually saw this happen. I would agree with the defenders to the extent that the position is that there is no evidence about the fitting of the flange that would enable the pursuers to claim that such evidence in itself supports the view that the accident was caused by leakage from the flange. The negligence of Sutton could only arise if it is clear that the tragedy was caused by a leak of condensate from the blind flange.
The defenders develop their argument to the effect that if Mr Sutton did not follow the established practice for the fitting of the blind flange this must have been a deliberate departure from normal and prudent practice. This could of course have implications for the application of the indemnities. I think it would be rather strained to regard a failure by Mr Sutton to tighten the bolts as "wilful misconduct" even as defined by the Contracts. There is always the possibility that any failure to tighten the bolts was simply inadvertence promoted by careless application to the job. Even if he decided not to flog the bolts this may have been due to a misunderstanding as to what was required when the blind flange was thought only to be in place for a short time. He did know the established practice for tightening bolts but may not have realised that it was unsafe to depart from the practice in any circumstances whatsoever. Against the background of the proposed maintenance of the pump it may have been beyond Mr Sutton’s comprehension that circumstances could arise which would necessitate the hurried re-introduction of the pump. He should not have presumed that in the situation he was faced with anything less than a fully tight flange would be safe but his error may have been due to a misunderstanding of the dangers rather than a deliberate ignoring of them. If he was finding it too difficult to flog the bolts or manage the heavy flange in the cramped situation in which he found himself then he had a duty not simply to leave matters but to seek assistance. I think this would be an extension of the case that the pursuers have pleaded. However his conduct in such circumstances would not be deliberate departure from practice. As Mr Sutton was tightening the bolts someone may have arrived and distracted him so that he inadvertently forgot to complete the tightening sequence. In the whole circumstances if Mr Sutton did fail to bolt the flange insecurely I could not proceed to conclude that he did this as an act of wilful misconduct. The defenders seek to extend this argument by saying that if Mr Sutton knew that he had not fitted the flange properly he was also guilty of wilful misconduct by not informing Mr Smith of the position when he met him at the end of the shift. In fact it may never have occurred to him that he was misleading his superiors in any material way. It may be possible to infer that the bolts on the flange were not properly tightened by Mr Sutton. However if this is what happened it is not possible to infer what led him not to do his work properly.
The defenders again referred me to the case of McWilliams v Sir William Arroll 1962 S.C.(H.L.) 70 where it was observed that an inference unfavourable to a deceased should not be drawn except upon a strong balance of probabilities. This of course only creates a presumption and it must yield to the evidence before a Court. The defenders submitted that if it is assumed initially that Mr Sutton was a reasonable man (the appropriate assumption to begin with) then unless it is positively proved that he acted unreasonably it would not be possible to infer that anything happened due to his misconduct. However in my view if other evidence shows that the perceived state of facts could only have arisen if Mr Sutton had not acted unreasonably then I find no difficulty in arriving at that conclusion even in the absence of immediate evidence pointing to the fact that he had acted irregularly. Generally the valve calibration work was completed on the day it began.
Understandably the defenders also submit that Mr Sutton was an experienced valve fitter, he was familiar with the platform and there was no evidence that he or indeed other valve fitters had ever failed to tighten flanges properly. Of course one answer to this may be that it is perhaps unlikely that the very peculiar circumstances which preceded the accident had ever before occurred so that even if a blind flange had in the past not been secured properly the chances are that this would have been noticed.
When Mr Rankin gave his evidence he accepted that the blind flanges had to be fitted securely. He said that "There is, I think, only one way to fit a flange and that is with the correct flange and you tighten up the studs properly." Mr Rankin thought that "the reason for applying these blind flanges to the open pipework was primarily to protect the faces of the flange, to stop any accidental spillage from the system, or to prevent any dirt from entering from outside". These illustrate the experience of a person who until a few weeks earlier had been working as an ordinary valve fitter. Mr Rankin confirmed that he and Sutton had been able to do all the other valves they had dealt with on their tour in one shift (sometimes extended with overtime). Certainly Mr Rankin said that PSV 504 was the last valve he and Sutton had to work on during their tour and that he expected to finish the job on the day he had started it. He confirmed that some flanges for the work were kept in the Score Container and that other flanges could be obtained from a store at a lower level. Mr Rankin did not assist Mr Sutton in the fitting of the blind flanges and it is clear that Mr Sutton fitted them on his own. After the flanges were fitted by Sutton, Mr Rankin, who had left the site did not return to the site to check Mr Sutton’s work. There was no attempt to develop a case that Mr Rankin failed in a duty to inspect the work but his failure to check would ensure that any failure to tighten the bolts would pass unnoticed.
Mr McDonald the rigger who gave evidence had sometimes helped Sutton to remove valves on other jobs. He confirmed that Sutton seemed to be a reasonably competent workman. Mr McDonald said that he had seen Mr Sutton tighten up flange bolts by hand but the matter was not pursued in evidence so that it is not entirely clear if the witness meant that he had seen Sutton tighten bolts with his fingers alone or if he meant tightened by hand with combinations spanners. He affirmed that the blind flanges were heavy and that was obvious from the ones produced in Court. Mr McDonald was able to tell me that Mr Sutton’s worksite had been a very congested area. In fact he thought that given the valves were heavy his view would be that in that constricted site it would require two men to fit the valve properly. On the other hand the witness Bagnall, who seemed to have wider experience of the flanges thought that one man could fit the flange but with difficulty. He thought for example that there might have been difficulty in keeping the mating flanges parallel. He thought that one flange would take from half an hour to three-quarters of an hour with one man but perhaps fifteen to twenty minutes with two men. In fact Mr Bagnall would have put two men on the job. He also expressed the opinion that it could have been difficult to use the combination spanner in the confined space that would have been available to Mr Sutton. If two men had been available then one on either side of the pipe could have wielded a combination spanner. Thus if Mr Sutton had only got combination spanners available to him he may have been faced with the problem of going off to procure a flogging hammer or persuading Mr Rankin to return and help him. It should perhaps be noted that a combination spanner is a spanner which has a round headed spanner at one end and an open-ended spanner at the other end. There is no case against Score for not having enough men on the job but in any event it was Mr Sutton’s duty not to leave the valve in a half tightened state without reporting any difficulty to Mr Rankin. What is significant however is that he is likely to have had some trouble in fitting the valve properly and if he thought that the work was only for a short time he may have had a temptation to do the tightening half-heartedly.
The evidence showed a degree of uncertainty as to the size of the valve unit being serviced. Mr Wylie who was a Quality Assessor who certified re-calibrated valves was shown the certificate which related to the earlier re-calibration of PSV 505 (the valve on Pump B equivalent to PSV 504 and agreed to be the same) and he indicated that on the certificate the valve was shown as having a rating of 1500 lbws on the downstream side and of 600 lbws on the upstream side. However the information on the certificate had been filled in by the valve engineer who was Mr Bruce. There was therefore a suggestion that possibly PSV 504 had a similar rating. However Mr Wottge’s clear view which I have no reason to doubt was that PSV 504 was a 900 lbws rated valve although the defenders challenged him on this point (this may be because the pursuers themselves had raised in their pleadings the possibility that the wrong blind flange size had been chosen). The pressure that the valve was planned to withstand did not require a higher rating than 900 lbws. Moreover a drawing which was part of OPCAL’s records and which purported to be from the manufacturers of the pump (12/170 of process) showed that the flange for the PSVs had a rating of 900 lbws. Other documentation relating to the valve in the pursuers’ possession after the accident were to the effect that originally the valve had a 900 lbws rating. One drawing (12/146) perhaps suggests that the valve may have had a different rating but Mr Wottge did not think this was accurate particularly in the light of the manufacturers’ own drawing. Whether or not the certificate for PSV 505 was accurate need not concern me. Confusion could have arisen because the valve itself was rated at 1500 lbws although the outlet was rated 900 lbws. Mr Standen said that even if a valve had a mismatched size of flange this would not matter provided the flange was effectively tightened. However there was evidence from a Mr Pirie that one could not put a 900 lbws flange on a 1500 lbws pipe because the bolts would not marry up. Mr Bagnall on the other hand indicated that in the opposite case one could put a 1500 lbws flange on a 900 lbws pipe. In any event it was Mr Sutton’s job to see that he was fitting the correct size of blind flange but there was really no evidence that he did not do so. If he was in fact trying to fit a flange which was a higher rating than needed (which I doubt) not only had the fitters selected the wrong flange for the job but he had made the task of fitting the flanges harder since the 1500 lbws flange was much heavier. According to Mr Bagnall a fitter at the valve site would have no difficulty in recognising that he had the wrong size of flange. At the end of the day I did not understand the pursuers to press their case that the wrong size of flange may have been fitted.
Number 12/230 of process was a valve calibration certificate dated 22nd June 1984 relating to PSV 504 and it shows the rating at 4 inch 900 lbws. In the certificate 12/228 the valve size of PSV is shown as 4 inch 1500 lbws. The date of the certificate is October 1985. A certificate of the same date relating to PSV 505 also shows that valve with the higher rating. However Mr Wottge confirmed that there was no change to the relevant line between 1984 and 1985. A question may arise as to whether the fitters may have been confused by looking at earlier certificates and seeing perhaps a certificate that declared the valve to have a 1500 rating. However if the wrong size of flange had been taken to the site as Mr Bagnall said this should immediately become obvious to the fitter and all that was required was to fetch the proper size. There was a store of flanges kept on the platform. Mr Rankin did not suggest that such a problem had arisen. In any event Mr Sutton had worked on PSV 504 not long before so that he should have known the flanges that were fitted to these PSVs notwithstanding that the certificate (which he would not have himself prepared) had been inaccurate. A Mr Reid gave evidence. He was a Production Manager with the Wood Group. In his earlier post as Quality Assessor he had completed the certificate relating to October 1984. He explained that the outlet from the valve was not large and he had rated it at 900 lbws. He thought he may have got the rating from Information Cards with all the valve details on them although he could not be sure of this. It was Mr Pirie who had carried out the calibration of the two PSVs in 1985 and he indicated that it had been a rushed job. It was not at all clear that the person who fitted the flange was the person who had filled details on the certificates.
The witness Mr Barclay was a Maintenance Engineer who had formerly worked as a Quality Assessor with the Wood Group. He had also filled in one of the certificates with an incorrect rating. He said that he may have obtained his information from a tag on the valve but this as I have already observed could have misled him since the rating of the valve was higher than that at the outlet.
There clearly had been occasional confusion as to the correct size of the flange that ought to have been fitted to the pipework leading to PSV 504. This whole area of investigation however may have largely been a red herring. It was the fitters’ job to secure a blind flange and nothing emerged to indicate that they would have had any difficulty in doing this had they applied themselves to the work with adequate care. If there was any problem about the size of the blind flange this would have become clear immediately an attempt was made to fit the flange at the site. There was no suggestion that OPCAL’s store of blind flanges lacked the necessary range. If this had been the problem one would have expected Mr Rankin to have mentioned it. In any event if there was any problem with the blind flange it was that the flange itself was too big and if such a flange is fitted tightly it should not leak. Thus one of the valve fitters doing the fitting job properly (and Mr Sutton in particular as the person effecting the fitting) should have been able to achieve a securely fitting blind flange. There was no evidence that there was any defect in the ring seal and presumably if there had been this would be noted by a competent valve fitter and the seal or flange replaced. There was no evidence to suggest that the seal was faulty. The real issue appears to be how the blind flanges which obviously were fitted were tightened.
7.2.2. Implications in Law - Foreseeability
The defenders submitted that even were it the case that Mr Sutton had not tightened the blind flange on the PSV effectively and that this had caused the leak which led to the explosion then Mr Sutton (and correspondingly those who employed him) were not legally liable for any consequences. I was referred to Gloag and Henderson - 10th Edition p 509 under the heading "Duty of Care". The proposition which the defenders drew from this citation was that a duty of care is not owed to the world at large but to those to whom injury may .reasonably and probably be anticipated if the duty is not observed. As a statement of general principle that does not cause me too much difficulty. The extension of the foregoing formulation of the principle which was urged upon me was that legal liability is limited to those consequences of our acts which a reasonable man of ordinary intelligence and experience so acting would have in contemplation. The kind of injury sustained and the manner in which it was sustained should be reasonably foreseeable but the precise chain of events leading up to the particular accident need not be foreseeable. Negligence is not to be equated with carelessness for one can be careless although it is not anticipated that any harm will result from the careless act. Again in general I can see nothing exceptionable about those assertions.
The defenders sought to analyse the case of Hughes v The Lord Advocate 1963 S.C.(H.L.) 31. In that well-known case small children intermeddling with an allurement which had been left accessible to them resulted in one of the boys being injured by fire following upon an explosion. The fire arose from an unusual and intrinsically unlikely set of circumstances. However it was found that burning (the kind of injury actually sustained) and the manner in which it was sustained were reasonably foreseeable. However the House of Lords held that whereas the events leading to the injury might have been unforeseeable the danger was not different in kind from a danger arising from fire in some form. A manhole, lamps and equipment (all involved in the accident) were obvious allurements to children. At the Proof it had been held that it was reasonably foreseeable that if the manhole was left unattended (as had happened) children might enter the shelter. The defenders in this case contended that this means that the manner in which the accident came about was reasonably foreseeable. However this is rather simplifying the situation. All the finding in fact indicates is that the first step without which the accident could not have occurred was foreseeable. The defenders contended I think with some justification that there was never any doubt that the accident had been caused by a lamp left in the manhole and it is only the particular way in which the lamp came to be implicated in the disaster which was unforeseeable. It should be noted that the Lord Ordinary held that it was reasonable to anticipate that danger would be likely to result from interference by children with the lamps that had been left unattended. The anticipated danger was formulated in the Inner House by Lord Guthrie as being "The Lord Ordinary had held that it should have been anticipated that a boy might in the circumstances fall into the manhole and sustain injuries by burning from the paraffin lamp". Thus the danger to be anticipated had become a rather more comprehensive narration of what actually happened. What the defenders argued was that the essence of the House of Lord’s decision was that the human activity which converted the potential danger into actual injury was entirely foreseeable. Lord Guest in his opinion in the House of Lords observed "the explosion was an immaterial event in the chain of causation. It was simply one way in which burning might be caused by the potentially dangerous paraffin lamp". However the defenders insisted that the decision in Hughes was dependent on the finding that the nature of the injury, that is to say burning, was foreseeable.
An observation of Lord Reid in Hughes is also interesting. He said "no doubt it was not to be expected that the injuries would be as serious as those which the appellant in fact sustained. But a defender is liable, although the damage may be a good deal greater in extent than was foreseeable." I mention this because the defenders seemed to be striving to suggest at points in their submissions that Mr Sutton could never have anticipated that any slackness in carrying out his particular job could ever have resulted in the massive catastrophe which in fact occurred. If an act which could potentially and foreseeably cause injury is committed the person at fault cannot limit his liability by seeking to estimate and restrict the degree of injury which he supposes might result. Lord Reid did concede that "the intrusion of a new and unexpected factor could be regarded as the cause of the accident rather than the fault of the defender". The defenders said that in Hughes the cause of the accident was not a new and unexpected factor. It was the lamp. However it behaved in an unexpected way. The present pursuers argued that from Hughes what could be derived is that the possibility of personal injury from the act or omission complained of must be foreseeable. If you can foresee that people are so proximate to the danger that personal injury to them can be foreseen then you have a duty not to expose them to that risk.. That proposition in my view seems to be entirely consistent with the law.
Lord Jenkins in Hughes gives expression to a useful test of foreseeability. He asks "Is this the sort of thing that is going to happen if I don’t do my job properly?" Lord Sorn in the Inner House gives a test which although its application was not followed in the House of Lords may well give a further indication of the sort of question that can be asked. He asks if the event which caused the explosion was "an essential event outside the kind of events which might have been foreseen". One element in Lord Carmont’s opinion in the Inner House which I have no hesitation in accepting and which must always be borne in mind is that each case must depend on it own particular facts. In Lord Morris’ opinion in Hughes he refers to the fact that there does not seem to be anything "fantastic or highly improbable" in the chain of events which caused the accident. The test he employs is taken from Lord Keith in Miller v The South of Scotland Electricity Board.
The defenders referred to the pursuers’ pleadings and to the fact that the pursuers aver that under the contracts the contractors, Score, had warranted that the work would be performed competently and there were specific stipulations that the contractors would ensure that all employees should conform and abide by Company procedures and regulations pertaining to safety. The pursuers also aver that it was a requirement of the pursuers in terms of their contract that blind flanges should be fitted securely to open ended pipework after the removal of an pressure safety valves. The defenders did not challenge the averments as narrating the relevant contractual conditions but disputed that these were safety requirements. I find that a difficult proposition. Apart from other considerations the fitters knew that the work was regulated by the permit to work system which is one of the principal safety requirements on the platform. The permit for the maintenance of PCVs provided that a blind flange was to be fitted to open pipework (see number 12/233 of process). The pursuers also aver that one of the reasons for the said requirement was to prevent a leakage from pipework in the event of the inadvertent admission to it of hydrocarbons.
The defenders make the point that a fitter was entitled to suppose that if a valve was ever to be exposed to other than incidental passage of gas - that is exposed to full system pressure - it would first be pressure tested. However the implication of that is that a valve fitter is entitled to suppose that if he fits the valve incorrectly his inadequacy will be discovered by pressure testing before the flange is exposed to serious pressure. This assumes a great deal of reliance and certainty in any system of pressure testing. Indeed the very existence of a system of pressure testing may emphasise how important it is to check that through inadvertence or otherwise flanges have not been fitted insecurely.
The defenders argued that the possibility of the flange being exposed to system pressure was highly improbable. Witnesses gave evidence that operators would not deliberately re-pressurise a pump with the PSV missing. It was said that it would be equally improbable given the elaborate safety systems that a lead operator would not know that a PSV was absent.
It should be noted that the evidence showed that there is always thought to be a danger when pipework is breached. That is no doubt why there is an elaborate system to deal with such situations by flanges, valves, spading and pressure testing. The point that was being made was that it is always necessary to be careful when breaking into hydrocarbon.
The defenders claimed that the pursuers had nowhere averred that there was a foreseeable risk of fire or explosion. However if there was one plain fact in the case it was that the pursuers were attributing explosion and the consequent fire to an incorrectly fitted blind flange. It is implicit in the pursuers’ pleadings that Mr Sutton knew or ought to have known he was creating the danger which caused the disaster if he did not tighten the bolts of the blind flange properly. Whether or not the matter can be proved is another question but after Proof lasting for years I doubt if the defenders could have been in any doubt as to the pursuers’ case on this matter. Another approach taken by the defenders was that it is essential to the pursuers’ case as pleaded that Mr Sutton must have known that if he did not tighten the blind flange there was a risk that someone would inadvertently re-pressurise the pump. It is of course contended that the pursuers have not proved that it was foreseeable that someone would attempt to re-pressurise the pump when the valve was missing. However if it were shown that it was foreseeable that there was a risk that a dangerous quantity of gas was likely to escape if the flange was not properly tightened I cannot see that the further refinement that the fitter should have foreseen precisely how this came about is needed. The defenders also submitted that it was only by re-pressurisation that one could get an flammable mass. I think this is not a very useful point. The flammable mass that is related to the particular explosion that the pursuers say occurred is of course related at least in the pursuers’ view on the kind of flammable mass that would be created upon re-pressurisation. It is also related to the fact that in traversing the module the gas cloud would have to pass gas detectors. Because it may not be easy to create flammable clouds at particular detectors does not mean that there could not be a flammable mass locally. In any event it cannot be asserted that even a small escape of hydrocarbon does not pose some risk of fire of explosion. The whole thrust of the evidence is that hydrocarbon is highly unstable and is, if uncontrolled, at least to some degree a danger. I would be surprised if an experienced valve fitter like Mr Sutton working on oil platforms did not realise that even a small escape of hydrocarbon was undesirable because of its volatile nature. Whether that takes the pursuers far enough of course is the critical question. There is force in the view that the re-pressurisation of the pump with the PSV missing is an abnormal procedure and a valve fitter might not even consider that particular possibility. None of the witnesses from the production operating staff on the platform claim to have ever seen a pump re-pressurised with a PSV missing. To do this would be bad practice. Indeed Mr Henderson thought that it would be hazardous. On the other hand Mr Murray thought that it might be done in dire circumstances with the approval of the OIM. Mr Bollands also thought that if such an irregular practice were ever to be followed it would require the approval of a higher authority than Mr Vernon. As to a valve fitter it might be questionable if he would have any experience at all of the detailed procedures for the re-pressurisation of a condensate injection pump.
The defenders cited to me the case of Maloco v Littlewoods Organisation Limited 1987 S.C.(H.L.) 37. While contractors were doing work connected with the conversion of a cinema to a supermarket it became clear that young persons were overcoming the security of the building. The building had been secured at night but the miscreants were breaking in. Eventually the intruders set fire to the building causing serious damage to adjoining buildings. Their owners sued the owners of the cinema. The House of Lords on appeal held that there was no general duty to prevent a third parties’ deliberate wrongdoing even although there was a high degree of foreseeability that it might occur. This was modified in relation to the special circumstances of an occupier who negligently allowed a source of danger to be created on his land. The matter was one for the facts of each case although cases where liability would arise were likely to be rare. However it was held that the defenders were not liable in the case under consideration since if there was a fire risk they had no means of knowing that it existed. Lord Mackay of Clashfern in his judgment made the important observation:
"It is plain from the authorities that the fact that the damage, upon which a claim was founded, was caused by a human agent quite independent of the person against whom a claim in negligence is made does not, of itself, preclude success of the claim, since breach of duty on the part of the person against whom the claim is made may also have played a part in causing the damage".
Another passage in Lord Mackay’s judgment of possible importance is:
"human conduct is particularly unpredictable and every society will have a sprinkling of people who behave most abnormally. The result of this consideration, in my opinion, is that where the only possible source of the type of damage or injury which is in question is the agency of a human being for whom the person against whom the claim is made has no responsibility, it may be easy to find that as a reasonable person he was bound to anticipate that type of damage as a consequence of his act or omission."
His Lordship then goes on to say that:
"The more unpredictable the conduct in question, the less easy to affirm that any particular result from it is probable and in many circumstances the only way in which a judge could properly be persuaded to come to the conclusion that the result was not only possible but reasonably foreseeable as probable would be to convince him that, in the circumstances, it was highly likely".
The defenders sought to derive comfort from the observations of Lord Mackay that the anticipated action of the third party must be highly likely. If as the defenders suggest the only critical danger that Mr Sutton had to anticipate was that some employee of the operators might attempt to re-pressurise the pump then Lord Mackay’s comments might indeed be highly relevant. The defenders also argued that if it was objectively improbable that that someone would re-pressurise the pump Mr Sutton could not be in a worse position if he did not possess that knowledge than if he did.
Reverting to Maloco what Lord Goff of Chieveley states at page 77 also has some interest. He is discussing circumstances in which a defender may be held liable in negligence to the pursuer although the immediate cause of the damage suffered by the pursuer is the deliberate wrongdoing of another. His Lordship observes:
"This may occur where the Defender negligently causes or permits to be created a source of danger, and it is reasonably foreseeable that third parties may interfere with it and, sparking off the danger, thereby cause damage to persons in the position of the Pursuer".
I was referred by the defenders to the case of London Passenger Transport Board v Upson &c 1949 A.C. 155. This case was cited for the dicta of Lord du Parcq at page 176 of the report. His Lordship said:
"The correct principle was stated by Lord Dunedin when he said ‘If the possibility of the danger emerging is reasonably apparent, then to take no precautions is negligence; if the possibility of danger emerging is only a mere possibility which would never occur to the mind of a reasonable man, then there is no negligence in not having taken extraordinary precautions’. I regard this statement and that of Lord MacMillan in the same case, which was to the like effect as applying generally to actions in which the negligence alleged is an omission to take due care for the safety of others".
The case involved the running down of a pedestrian by a taxi driver and Lord du Parcq added "A driver is never entitled to assume that people will not do what his experience and common-sense teach him that they are in fact likely to do so". The analysis I have just quoted is in general terms but is supported by judges of considerable authority.
It should again be noted that Regulation 32(3) of the offshore regulations say "It shall be the duty if every person while on or near an offshore installation not to do anything likely to endanger safety or health of himself or other persons on or near the installation, or to render unsafe any equipment used on or near it". This duty is hedged about with criminal sanctions. The defenders did not attempt to dispute that the statutory duty was absolute and thus they argued that it does not assist much to decide what is a foreseeable danger at common law. However what the Regulation does do in my opinion is to bring home to those working on offshore platforms the need in the interests of safety to take great care.
A further case the defenders referred me to was Woods v Duncan 1946 A.C. 401. This case which is mentioned elsewhere in this opinion in another context involved a submarine undergoing diving trials. When the submarine was submerged a torpedo officer opened a rear door of one of the torpedo tubes not realising the bow-cap was open. Accordingly the submarine sank. The mistake had occurred at least in part because a test-cock which had been painted had stuck. It was held that those responsible for painting although negligent in the painting operation had no reason to suspect that the blocking of a hole with paint would be dangerous to the personnel of the submarine. In the House of Lords, Lord Russell observed that he agreed with the view of the Court of Appeal that although the contractors were negligent their negligence was not the cause of the accident. He accepted that if the test-cock had worked the Torpedo Officer would not have opened the relevant door. Thus to that extent the condition of the instrument lead to the accident but notwithstanding this if the bow-cap had been shut the accident would not have happened. What caused the accident was the double event. However the core of Lord Russell’s opinion is that the contractors could not have foreseen that the choking of the test cock could have endangered the lives of those aboard the submarine. Thus the case is, like so many in this area of the law, very much fact driven. It has to be noted that in Woods the contractors were painting a ship which task would not on the face of it suggest serious safety considerations. It was held in effect that they had no reason to know or anticipate the safety function of the test-cock. The defenders contended that Score had no reason to anticipate that the function of the blind flange could include holding fast against the system pressure.
The defenders placed great emphasis on the fact that many witnesses qualified to discuss the matter were not asked what they considered to be the function of the blind flange. However that is not to say that there was not a mass of evidence making the general function of a blind flange fairly clear. If the defenders thought that the flange had a limited specific function rather than a general function it was of course open to them to develop a challenge on this point by cross-examination of any witnesses they considered could support their view. It has to be noted that before the witness Todd ventured an opinion on what he thought might be the function of the PSV 504 he was asked what purpose there might be in fitting a blind flange to open pipework and he replied "that is a precaution which we took on every piece of pipework". Mr MacDonald testified that even small bore water pipes would be fitted with blind flanges. Mr Todd when asked specifically why blind flanges would be fitted to the PSV piping answered that it was to prevent any leakage of fluids past the isolation valve escaping to the atmosphere. Thus he gave an illustration of why they might be fitted. Nobody asked him if that was the only purpose of having a blind flange. He no doubt thought of the most likely contingency. Also it must be noted that it was thought that even seepage from the isolation valve ought to be prevented. Mr Todd did say that the fitting of blind flanges to PSV was an invariable requirement and that fitted means "fitted properly". Indeed he goes on to declare "It should be put in place and secured as it was meant to be secured, tightly, to prevent gas escaping or any fluids escaping from the open pipework". It would perhaps be odd that if a fitter was aware that the flange was there to prevent escape of gas or fluids he should think that the only such fumes or fluids it might be required to trap were such as might escape from seepage. Moreover if this was in fact Mr Sutton’s expectation he may well have required to surmise it since there was no evidence that he had ever been told that there was this limited function. Even if he knew that an escape of gas can be dangerous and that one reason the blind flange was there was to prevent seepage it would be a curious definition of his duty to take care if this depended on a fine calculation of how much gas could seep past the GOV and the potential of that particular amount for causing fire or explosion. The duty to take reasonable care for one’s neighbour cannot be founded on such narrow considerations.
Counsel for the defenders suggested that one reason why a blind flange was fitted to prevent seepage escaping was to avoid the risk of false gas and fire alarms. He did not cite any evidence in support of this supposition. Moreover if a seepage was sufficient to set an alarm off it would be safer to avoid such a contingency being possible by fitting the blind flange properly. If an alarm goes off it is considered sufficiently dangerous to merit immediate investigation.
When the valve is removed the pipework will be open for a short time while the blind flange is being fitted and certainly it was never suggested that this creates any particular danger. This may indicate that the kind of escape that is to be prevented is such as could occur over a material period and therefore would be more likely to generate a dangerous quantity of gas. The question of valve security, potential seepage, and consequent explosive mass was not explored in any depth in the evidence and if it had been discussed with Mr Sutton during his lifetime it is difficult to suppose it would have meant much to him.
The defenders suggested that there was no evidence that a blind flange was designed to withstand system pressure but this is not quite so for there was evidence about the Nowsco tests which suggested that a properly fitted blind flange would be leak-tight when subjected to the equivalent of such pressure. On the other hand the same tests established that all combinations of flange and seal types would give what are described as gross leaks if the bolts were only finger tight (Mr Standen’s Report which is number 13/79 of process). It should perhaps also be noted that there was no evidence that the flange was likely to warp and looking at the weight and solidity of the flanges I should expect that any kind of warping would be most exceptional.
The witness Mr Reid had a more expansive idea of the purpose of a blind flange. He stated that the purpose was "the same as any line, you don’t want an escape of the medium into the module or whatever". I have no reason to suppose that the awareness of Mr Sutton was any less than that of Mr Reid. The defenders suggested that Mr Reid may have had water lines in mind but I think that is a bit far fetched. In any event Mr Sutton could hardly have expected an escape of water from the relief line.
The defenders’ witness Mr Pirie was a valve technician with the Wood Group. It is worth mentioning that he confirmed that the normal practice for the fitting of blind flanges was to tighten them by flogging or the use of combination spanners. He agreed that one of the reasons for fitting a blind flange was to prevent leakage from the pipework. It should be noted that this answer was prompted by a leading question and of course he was only referring to one of the reasons.
Mr Rankin also had comments to make about blind flanges. He said that he saw no difference in principle to fitting a flange to a valve removed for testing and fitting one to the pipework. On the other hand this was in the context of the valve being tested under pressure. As he stated "a blind flange must be used to contain the pressure within the body of a valve". He makes what is claimed by the defenders was a distinction between a situation where a foreign substance is to be introduced to a valve for testing purposes. He says in relation to the open pipework the purpose of the flange is "to stop any accidental spillage from the system, or to prevent any dirt entering from outside". He considers that if a valve is supposed to contain pressure it ought to be pressure tested. On the other hand he seems to be acknowledging that one purpose of a blind flange is to contain such material as might be expected to arrive at the flange. Any hydrocarbon that escapes from the pipework into the atmosphere is after all an accidental spillage and the flange was there to prevent that happening. The defenders’ Senior Counsel expressed surprise that Mr Rankin was not examined more closely about what was to be expected from a blind flange particularly as in the pleadings he shares the imputed blame for the accident.
The purpose of the submissions of the defenders I have been discussing was to convince me that the pursuers have failed to prove that the kind of accident that may have occurred from an escape of hydrocarbon at the blind flange fitted by Mr Sutton was not such as would have been anticipated by him as a potential risk which could arise from the manner in which he fitted the blind flange. I have no doubt that it probably did not cross Mr Sutton’s mind that any action or omission on his part could expose the platform and its personnel to the kind of catastrophe which actually occurred. However because the consequences of an act are unexpected in scale would not be enough to exonerate a person who otherwise had been negligent. In this area of the law a good starting point is the observation of Lord du Parcq in London Passenger Board (which I have already referred to above) where his Lordship said "if the possibility of the danger emerging is reasonably apparent then to take no precautions is negligent". It must be noted that Mr Sutton was working in what was intrinsically a very dangerous situation if normal safety precautions were ignored. The situation is quite different to the one that arose in Woods. There the painters were carrying out work which was essentially innocuous and they could not have conceived that any careless painting might have caused the submarine to sink. Even if they had been aware of the specific function of the test-cock it was not to warn if the bow was open but to check the water level in the tube when torpedoes are fired. It is well known to those working on a platform that the most critical safety concern is to contain the massive quantities of hydrocarbon that are circulating around the platform. Moreover that there is a risk of fire or explosion if any of this escapes is very evident. On the production modules there is an array of fire and gas alarms. If a workman is faced with a situation where a quantity of hydrocarbon is at risk of escaping I do not think it behoves him to say that he was not aware of the precise amount of the substance it needed to form an explosive mass at any point on the platform. He well knows that it is not his function to calculate precise risks arising from a particular operation. His responsibility is to take the precautions laid down by the platform management. There are rigorous statutory safety regulations which bring home to him how important it is for him to carry out his work safely. There is a complex system laid out by the defenders to control safety. Every person working on the platform knows the important function of the permit to work system. There are hot work permits and cold work permits and it is not a difficult inference to understand that the special requirements of hot work permits are connected with the prevention of fire and explosion. In the case of PSV calibration the Permit which regulated the work carried out by Mr Sutton actually specifies that blind flanges were to be fitted as a precaution. There was no doubt that the proper way to fit blind flanges is to flog the bolts or to tighten them with a combination spanner. There is no point in fitting them if they are not properly fitted. Apart from any other consideration it would seem to me to be careless in the circumstances of work on an oil platform to ignore a safety instruction stipulated by the employers. The valve fitter can hardly be expected to be in a position to assess what dangers may or may not arise from the system of working on the platform. I do not think it is for him to consider what the purpose of a blind flange is or what danger there is of hydrocarbon accidentally being introduced to the relief line. This surely would be for management. It is not for the fitter to consider "if I do not do this job as I am supposed to there is no obvious risk". How is he supposed to be sure about this. The appropriate approach in a situation where generally potential dangers abound is to think "the Management have asked me to take a particular precaution with safety implication so I had better do what they have asked me to do and not try and decide the matter for myself". I do not think there is any doubt that the managers expected a blind flange to be properly fitted so as to exclude any hydrocarbon that may have got into the line accidentally. It may be that the possibility of an operator deliberately introducing pressure into the valve when the PSV was missing would be relatively remote. However I do not require to consider what would have been the result if this had occurred since my view is that Mr Vernon, if he re-pressurised the pump, did so as an act of inadvertence. Such acts are always conceivable and they can present themselves in most unexpected ways. The platform management (and consequently Mr Sutton) could not anticipate precisely what accidental events might put more pressure into the pump than was intended. I do not see this case, as some other cases which have been mentioned, as being one where only the agency of a third party could have caused a danger. There could have been a variety of circumstances causing a small or large but at least dangerous amount of hydrocarbon to enter the pump and this could have been caused not by human intervention but by mechanical problems such as a GOV allowing more seepage than was normal or perhaps not closing fully. The important point was that the pump was still connected with the system and it was clear from the evidence that any breach of the flow system creates a situation requiring special care. Moreover the evidence also showed that it was part of the general system to have checks and counterbalances. Thus if a valve closed this was generally not relied upon to seal the system but flanges or a second valve would be in place as an extra precaution. In my view the blind flange sealing the relief line was at the end of the day part of the layer of precautions. Emphasis has been placed on pressure testing. This seems to me to be an additional precaution to ensure in situations where pressure is to be reintroduced to a line that the flanges have been fitted properly. Pressure testing is not meant as a substitute for correct fitting of a flange but an additional safety measure. Of course the provisions for sealing a blind flange in a valve calibration operation did not require pressure testing. We do not know why. It may be that because the valve is only there as a short term measure or because the likelihood of the valve being subjected to extreme pressure is remote that it was not considered a practical necessity. It should be again noted that the Nowsco tests showed that a properly fitted flange should be able to contain such pressure as would develop on re-pressurisation. Of course once the pump began to pump normally larger pressures could develop and it may need pressure testing to be sure that the flange could handle these. However I think OPCAL were entitled to assume that any precaution aimed at assisting in the sealing off the process line would at least function to anticipated capacity in that direction. The fact is that assuming a leak from an ill-fitting blind flange caused the accident then if it had been fitted as the Lead Production Operator intended when he issued the permit to work the accident would not have happened. I think OPCAL were entitled to expect that the safety systems they laid down would operate to best effect and that workmen would not decide to depart from them. Indeed the defenders make the same point in relation to the completion of the permit to work forms when they argue that production staff should not off their own bat seek to adjust the terms of the prescribed forms.
7.2.3. Novus Actus Interveniens
The defenders argued that since what is alleged to have occurred at the accident was not foreseeable to Mr Sutton he had no duty to take care in respect of what is said against him. However as a quite separate point the defenders argued that there was a break in the chain of causation to the effect that any slackness on the part of the valve fitter was not a cause of the accident. The defenders argued that an act which is a sine qua non in the chain of causation is not necessarily to be regarded as rendering the actor liable if the chain of causation is interfered with and completed by the act or omission of a third party. Senior Counsel for the defenders suggested that the prerequisite for liability is that the act complained off should be a direct contributor to the harm even if the harm would not have occurred without the earlier event. I do not think I can disagree with these generalities.
I was referred to the case of McKew v Holland & Hannen and Cubitts 1970 S.C.(H.L.) 20 on appeal to the House of Lords. A workman had been injured by an accident at work for which his employers were admittedly responsible. As a result of that accident his leg was left weakened. Three weeks later he sustained a further accident while he was descending some stairs and this was attributable to his weakened leg. The House of Lords affirmed the judgment of the Second Division in holding that the pursuer ought to have realised that certain precautionary steps were needed to descend the stair safely, that he acted unreasonably in omitting to take these steps, and thus his unreasonable conduct was a novus actus interveniens breaking the chain of causation. It was noted by Lord Reid that foreseeability did not come into the situation. Lord Reid said (at page 25) that if a man is injured in such a way that his leg may give way at any moment he must act reasonably and carefully. However it has to be observed that in the case being considered the pursuer was aware of all the circumstances dictating that he required to take care. Thus the reasonableness of his actions can be assessed without reference to the facts surrounding the earlier accident. I consider that the test applied by the Court was that the accident was caused exclusively by the pursuer’s unreasonable conduct. The defenders attempted to take from this case that if the chain of causation is broken by unreasonable conduct independent of the first act of negligence then it is immaterial that the earlier wrongdoer may have been able to foresee that the second event could happen. Lord Guest suggests that what is required to break the chain of events for which the second wrongdoer is accountable is "a new cause which disturbs the sequence of events, something which can be described as either unreasonable or extraneous or extrinsic". It also has to be noted that the conduct in McKew which was held to be unreasonable was the conduct of the victim himself.
The defenders also contended that if an act is so extraordinary as to bebeyond possible identification as being in the natural course of events then that would bring the doctrine of novus actus interveniens into play.
The next case cited to me by the defenders was Wright v Lodge &c (1993) 4 All.E.R. 299. A motor car had broken down on the motorway and while stopped for repair on the near-side lane a lorry drove into the back of it seriously injuring a passenger in the back seat of the stationary car. Unfortunately the lorry veered out of control on to the opposite carriageway and a collision took place which caused a fatality and injuries. The root cause of this accident was that the lorry had been travelling at excessive speed. The lorry driver was sued and while admitting liability claimed contribution from the original car driver. The judge of first instance found that the lorry driver had been driving recklessly and that this was the sole cause of the accident to dead and injured drivers. On the other hand in respect of the injured passenger the judge found that the driver of the car had been partly to blame in that she had left her vehicle unlit and stationary on the carriageway. The question in the Appeal was whether or not the negligence of the original driver had also been a factor causing the accident to the other drivers. It was held that the fact that a driver was negligent in leaving his car on the carriageway and causing an accident partly by his own negligence did not necessarily render him liable for subsequent events as a result of another driver’s reckless driving. In the case under consideration the lorry driver, driving an articulated lorry, had been driving at 60 mph in thick fog. The answer would have been different had only careless driving on the part of the lorry driver been involved. However what took the lorry onto the opposite carriageway was not so much the collision but the reckless speed of the lorry. Thus the presence of the lorry on the opposite carriageway was held to be wholly attributable to the reckless driving of the lorry driver. Given that determination of the facts it is difficult to see how any other result could have been arrived at than what the trial judge decided. Lord Parker in his judgment observed that although there has been much discussion in the authorities as to the tests to be applied in a case such as he was considering such tests must be related to the facts of each case. There is in his view no single test. He considered that the questions to be decided must be determined by applying common-sense to the facts of each particular case. Although it is not always as easy as it might appear at first sight to apply common-sense norms to such complicated matters I agree entirely with Lord Parker that the common-sense approach is the only safe way forward. However the matter which determined this appeal was the distinction that was drawn between what might normally be expected by way of careless conduct and the totally extraordinary implications when someone drives recklessly. The trial judge, Mr Justice Hobhouse, had held that the lorry driver had shut his eyes to the obvious risks that existed and that, Mr Currie contended, was exactly what had happened in the present cases. "Recklessness" in his submission was gross negligence in the face of obvious danger. In his judgment in Wright Lord Justice Woolfe comments that in Clerk and Lindsell on Torts the editors suggest the explanation for reckless driving being treated differently in causation is that deliberate or reckless conduct is so obviously more blameworthy than carelessness that it can be singled out as being the substantive cause. However even that approach in my view is not so easy to justify because there are many cases where the relative contributions of joint wrongdoers to the accident are very disproportionate. The justification of course would be that a distinction has to be made between quality and quantity but this is not always easy to discern.
The defenders of course as I have already stated were urging me to find that Mr Vernon was acting recklessly in that he knew that the PSV was missing but nevertheless violated safe procedure by jagging the pump without at least pressure testing (assuming of course that he did jag the pump) or if the fact was that he did not have it in mind that the PSV was not in place then that represented a gross failure to take account of his responsibilities and to keep himself informed as to what was happening in the production modules.
The next case the defenders referred m to was Marvin Sigurdson v British Columbia Electric Railway Company Ltd 1953 A.C. 291, a decision of the Privy Council. In this case a street-car driver ran into the appellant’s motor car which was stationary on the street-car track in the middle of a junction as the driver waited for an opportunity to complete crossing the road. The judge of first instance found that the street-car driver was solely to blame in not keeping a proper lookout but on appeal the motor car driver was held to be equally to blame. The respondents contended that where one party knows of the dangerous situation created by the negligence of another but fails to take reasonable steps to avoid the danger he is generally solely liable. On the other hand if that party did not actually know of the danger, or by his own negligence or deliberate act has disabled himself from becoming aware of the danger, he can only be held liable for a proportion of the resulting damage. It was held that there is no warrant for such a far reaching proposition. The implication that the defenders sought to take from this case was that if Mr Vernon did not know that the valve was absent this was due to his own negligence and he could not hide behind that negligence to deprive the defenders of a result in respect of Mr Sutton’s part that would otherwise be available to them. I must confess that I find that argument rather difficult to follow. If a dangerous hole is left on the pavement and someone falls into it because by looking at some distraction he has precluded himself from being aware of the danger is he totally excluded from some degree of recovery. The argument may make sense if what is being suggested is that only Mr Vernon’s reckless conduct or what is described as ‘gross negligence’ could have prevented him being aware of the risk represented by the missing PSV. To keep oneself in the dark deliberately would of course be akin to wilful misconduct.
In relation to the question of recklessness I was referred to the Stair Memorial Encyclopaedia of the Laws of Scotland, Volume 7, paragraph 82 on page 69. For criminal liability it is suggested that for there to be recklessness it is required that there be a culpable indifference to the consequences, or a blameful disregard of the results, or a total indifference to and disregard for the safety of the public, or an utter disregard of what the consequences will be, or a culpable disregard of the consequences. There were other references in the text-book. It was argued that the two basic notions behind the authorities were a very high degree of indifference or disregard relative to the potential consequences of the conduct in question or alternatively gross negligence.
The final case I was referred to by the defenders in relation to this part of the case is Knightly v Johns &c (1982) 1 W.L.R. 149. A motor car driven negligently in a tunnel overturned. The Police Inspector who took control of the locus forgot to close the tunnel immediately to traffic as required by police procedures. He ordered two constables to go back for that purpose and to fulfil that task they motor-cycled against the traffic flow. The plaintiff was injured when he thus travelled round a blind bend. It was held that the original motorist was not responsible because the new and direct cause of the plaintiff’s accident (the conduct of the inspector) broke the sequence of events. The remarks of Lord Justice Stephenson in the Court of Appeal are interesting because he says:
"I conclude from these rescue cases that the original tortfeasor, whose negligence created the danger which invites rescuers, will be responsible for injury and damage which are the natural and probable results of the wrongful act and that those results include injury and damage of a kind or class which might normally be foreseen or contemplated, although the particular accident could not be expected".
He seems to accept these same principles as applying to chain of causation cases. After a long analysis his Lordship sees no reason for departing from the established concept of a new cause breaking the chain from another link in the chain. He repeats what has repeatedly been said that human beings, however well trained, when they have to cope with a crisis, mistakes and mischance are to be expected. He concludes by finding that (following Lord Reid in Dorset Yacht Company) in the long run the question remoteness of damage should be answered as has often been stated, not by the logic of philosophers but by the common sense of plain men. The case of Knightly shows that whereas recklessness represents one end of the spectrum in assessing novus actus a chain of events not perhaps involving negligence can be so tortuous and extraordinary to make the accident too remote from what may have originally preceded it.
The pursuers referred additionally to The Oropesa 1943 Probate 42. In that case two steam vessels had come into collision and although one, the M.R., was badly damaged her Master thought she would survive. He had sent 50 of his crew in two boats to assist the Oropesa. Later he decided to go himself to the Oropesa with a number of men. The weather roughened, the boat capsized and nine men drowned including an engineer. The parents of the engineer sued the owners of the Oropesa for the loss of expectation following upon the death of their son. It was held that the Master of the Oropesa had acted reasonably and that the death was not the result of novus actus interveniens but was directly caused by the collision. Lord Wright in his judgment (page 37) states that human action does not per se sever the connected sequence of acts. For the rule to apply there must be a breach of the causal chain (I doubt if that would ever be disputed). If the interveniens acts unreasonably it was accepted that there can be a breach in the causal chain but Counsel for the pursuers contended that one never gets to that point in relation to Mr Vernon because the causal chain was never broken. According to the submission it is necessary first to ask is there a new cause and this is a question of fact. If there is a new cause - a break in the chain of causation - it would be necessary then to ask if the interveniens producing that new cause acted reasonably. It was said that there was no break in the link between Mr Sutton’s negligence and Mr Vernon’s conduct. Mr Sutton could be sued because despite what Mr Vernon did or did not do Mr Sutton’s negligence remained a contributory cause of the accident. I was referred to a further passage from Lord Wright in Oropesa where he approves the statement "Cause and consequence in such a matter do not depend on the question whether the first actions which intervenes is excusable or not but on the quest whether it is new and independent or not". In the light of the authorities I have no difficulty in accepting that authoritative encapsulation of the law.
The pursuers further referred me to Rouse v Squires (1973) 1 Q.B. 889. It was held in that case on appeal that if a driver creates an obstruction on the highway which creates a danger to other road users (including those driving too fast though not those who may deliberately or recklessly have driven into the obstruction) then his negligence contributed to the causation of an accident although the immediate cause was the negligent driving of another driver. In any event the essential finding of Rouse was that if there is an act of negligence and the consequential source of danger still exists to a substantial degree at the time of a second act of negligence, and thereby contributes to the accident then a defence of novus actus interveniens can never apply. In Admiralty Commissioners v S.S. Voluta Lord Birkenhead said essentially the same thing.
The final case to which the pursuers referred me was Haynes v Harwood (1935) 1 K.B. 146. I was referred to this case because of the dicta of Lord Greer who said:
"If what is relied upon as novus actus interveniens is the very kind of thing which is likely to happen if the want of care which is alleged takes place, the principle embodied in the maxim is no defence. The escape of hydrocarbon because the flange was loose is of course the kind of eventuality a valve fitter could have anticipated if a flange is not properly tightened".
Lord Greer continued:
"It is not necessary to show that this particular accident and this particular damage were probable; it is sufficient if the accident is of a class that might well be anticipated as one of the reasonable and probable results of the wrongful act."
The defenders’ submissions about novus actus interveniens undoubtedly raise difficult and interesting questions. The main thrust of their argument of course is derived from their principal submission that Mr Vernon either acted in deliberate contravention of the safety procedures or at least in a manner that was grossly negligent. If I had agreed with this I should have had to decide the important question as to whether I could accept that an act which is wilful or reckless wrongdoing will break the chain of causation. However I am not in that position. As I have set out above Mr Vernon in my view had let it slip from his mind (that is assuming it was ever there in the first place) that the PSV was missing. He had arrived that very evening on a fresh tour of duty. He probably had to familiarise himself with a number of matters. Given OPCAL’s systems he probably had proper opportunity to discover that the valve work was progressing. There is at least a possibility that some time in the evening he would have known this or had the opportunity to discover it. On the other hand given that the pump A had been withdrawn from service for maintenance and isolated the situation in relation to the PSV may not have impacted very strongly in his mind. Later in the evening he was confronted with what on any view could have been described as a minor crisis. The pump which pumped out the condensate production stopped working. The defenders suggested that he had ample time to deal with the situation comfortably. However the matter may have caused him concern and he certainly had to take decisions. When pump B did not restart Mr Vernon decided that he would bring pump A back from maintenance and bring it back to production. He knew that to bring a pump back into production when the PSV was not available was unacceptable practice. Nevertheless he confidently and openly made arrangements to re-introduce pump A to the system. He informed Mr Bollands and Mr Clark of his intentions yet never mentioned the matter of the PSV. He made no attempt to consult senior staff who were available on the platform. If he knew that the PSV was not in place he was openly and unnecessarily exposing himself to reprimand by his superiors and that in a situation when he could have hoped for a promotion. He arranged for Mr Clark to restore the electricity and yet if he had known that the pump was also isolated under a PSV permit his actings would not have made sense because the pump would have remained isolated under the PSV tags. It is clear that his measures took place only under one permit or permit application. The reason why Mr Vernon had allowed the position in relation to the PSV to escape his mind (if this is the case) will never be known. However I cannot conclude that his actings represented gross negligence or recklessness. If he had forgotten that the PSV was not in place he was not showing a reckless disregard for the risks since such risks had never entered his mind. Assuming he was guilty of inadvertence this is a serious matter for someone in his responsible position but a serious degree of inadvertence is not the same as a reckless or even irresponsible state of mind.
Unlike the novus actus interveniens situations set out above, Mr Sutton (assuming he only put the flange on finger tight) had deliberately ignored instructions which he knew or ought to have known had safety implications for the platform. It is difficult to know what he was thinking. The flange was heavy and it is likely that because of cramped space he found it difficult to hold the heavy flange in position and to use flogging or combination spanners. He could have called for extra assistance if this were the case. He may well of course have thought that the flange would only be on for a few hours and that he would be present during that time if any problem developed or was likely to develop at the relief line. Were that so when he discovered that the flange was going to be left in position overnight he should have taken steps even at that stage to tighten it properly. It is very sad and unfortunate that Mr Sutton has not had the opportunity to give his own version of events but I have to proceed on the basis of the evidence available. As I have already said one purpose of the flange being fitted was to exclude foreign substances from escaping from the pump and entering the Module. If the flange had been properly fitted the probability is that it would have fulfilled that purpose and the accident would not have happened. As it happened the loose flange remained a danger even at the time of the accident albeit that Mr Vernon’s actings had intervened. The complex events leading up to the escape of hydrocarbon perhaps could not have been anticipated in all their detail. However that some hydrocarbon could escape was perfectly foreseeable. It was exactly the type of event that was to be feared and the blind flange was part of the system of protection against such a contingency. As it happened the introduction of a lethal quantity of gas into the pump required human intervention but the situation could equally have arisen as a result of some unexpected failure of equipment. A valve could have failed or someone could have for some unexpected reason accidentally opened the GOV. In the case of Wright what caused the second accident was that the lorry veered onto the opposite carriageway. This was held to be a different type of accident to what might have been expected from the fact that a car had been left stationary in the inner lane. It was not only a question of scale of injury. Only a lorry travelling at an excessive speed could have crossed to the opposite carriageway (or so it was held). Supposing a bus full of passenger had been travelling at a reasonable speed but collided with the stationary motor car killing a substantial number of people then although the scale of the claim against the car driver would have been materially different it is likely she would have been held liable if no recklessness was involved. As I have already said an escape of foreign matter is just the type of thing that would be expected if a flange is left loose. All that was probably totally unforeseen was the scale of the disaster. Even allowing for the negligence of Mr Vernon the accident would not have happened if the flange had been fitted as instructed. At the end of the day there were two causes of the accident both of which could properly be categorised as direct. One was the fact that the pump was re-pressurised at a time when not all safety measures were in place. The other is that when this happened the blind flange had not been fitted so as to perform its proper function of sealing off the relief line. The chain of causation stretching from the original fitting of the flange had not been broken. When condensate entered the pump at that point of time the actings of Mr Sutton were for the first time put to the test and the absence of a properly fitted flange allowed it to leak out. The effects of the negligence of Mr Sutton and that of Mr Vernon co-existed and they were both operative in connection with the accident.
One can feel quite sorry for Mr Sutton. He was engaged in a relatively simple operation and no doubt he thought that if he cut a few corners not much harm was likely to result. It may even be that he simply did not concentrate and do his work carefully. Unfortunately the worst that could have been conceived then happened because of his slackness.
The pursuers also made the point that the defenders had an obligation to plead a case of novus actus interveniens if this is part of their defence because they have the onus of proving it (S.S. Baron Vernon v The Canadian Pacific Railway Company 1928 S.C.(H.L.) 21). This point might have a degree of merit although, if so, how it would operate in a case like this is arguable. However I need not decide the issue on a technical pleading point because for the reasons I have set out above I am satisfied that the defence is ill-founded.
7.3. The Jagging of Condensate Injection Pump A
The pursuers’ case, of course, largely depends for its validity on the proposition that the accident was caused because Mr Vernon in attempting to start Condensate Injection Pump A tried to re-pressurise it by jagging at a time when PSV 504 was missing. If the pursuers cannot prove that Vernon must have been jagging the pump then subject to one matter their case must collapse. I introduce a note of qualification because there is the pursuers’ separate argument that if the cause of the accident is not established then they would still be entitled to recover under the indemnities.
One fact dominates this area of dispute and that is that there is no direct evidence that Mr Vernon jagged the pump. No-one claims to have seen him do so. Thus proof of the facts on this topic must depend on inference. There is of course the fact that Mr Vernon was faced with a need to get a condensate injection pump running and proceeded to the 68-foot level after declaring that he was intending to start Pump A. There is also the pursuers’ claim that the two-phase pattern of the gas alarms would only be consistent with the leak having been caused by a multiple phase process such as jagging the pump. The pursuers’ contention is that the initial gas alarm was as a result of a gas escape after the first stage of the jagging process. A second stage of jagging carries out a few minutes later caused a second escape of gas which not only caused the second flurry of gas alarms but also was fuel for the explosion.
The general layout of the 68-foot area is conveniently shown in the schematic number 13/49 of process. It will be seen there that the push/pull buttons relating to the opening and shutting of the relevant GOVs were situated on stanchions which placed them about a yard or so from the GOVs themselves. Each pump had in the same area a running control panel (shown in the schematic to the right of each pump). Each of these panels was about two feet from its pump. A pump is started by pushing the start button relating to the pump.
One fact that is not disputed is that at the time of the accident Pump B had tripped and Pump A had been taken out of production for a proposed planned maintenance. This meant that Mr Vernon would be concerned to get one or other of these pumps running as quickly as he could. Failing this there would have been a serious loss of condensate production. In respect of restarting the pump that had tripped the operator would there be dealing with a pump already pressurised. To restart the operator would reset the speed selector to manual and wind the signal off the pump. Then at the control panel JCP 57 the operator would reset the GOVs which relate to the pump. Then one would proceed to the GOVs and re-latch them which would have the effect of opening both GOVs. This would put the operator in a position to try and restart the pump by pressing the start button. If the pump started the manual speed controller would later have to be adjusted. Thereafter the pump could be switched back to automatic running. It was easier if there were two persons involved in restarting the pump but one person if he worked quickly could restart a pump in a few minutes.
It is plain that after Pump B had tripped Mr Vernon (who had come down to the pump from the Control Room) together with Mr Richard attempted to restart Pump B and failed to do so. Thus were it at all possible to bring Pump A back into commission it would have been an obvious step to take even if efforts to get Pump B restarted were continuing. After Mr Vernon had attended to his business with Mr Clark which he carried out from the Control Room, he returned to the 68-foot level. There he and Mr Richard were shortly thereafter joined by Mr Grieve although not long afterwards Mr Richard left them when a gas leak was reported to him. However for at least a short period there would have been three men available at the two pumps. Mr Grieve for his part seems to have been engaged in trying to restart the tripped pump and in particular he was engaged in switching the speed controller to manual, winding it back, and in resetting the GOVs. Of course each time the pump failed to start or tripped the procedures which Mr Grieve was attending to would have to be repeated. Mr Grieve did indicate that it was not all that unusual for a pump which had tripped to be stubborn about restarting. Once the pump does restart then a red control light goes on to show that the pump is back running. He was there with the intention of adjusting the speed controller of the pump. On the other hand the witnesses Henderson and Seddon stated that if it was the intention to bring a dormant pump into operation it would be expedient to get it re-pressurised as quickly as possible in case it proved to be needed.
We were told that if a pump has been isolated it was practice to put on the valve a tag saying "Do not operate". However we were told very little about the operation and effect of such a system. If a pump is being isolated then various ventilation valves require to be opened and these in turn have to be closed when the pump is restarted although at what stage is not entirely clear. The isolation procedures in respect of adjusting valves takes about 10 minutes and that of course is exclusive of the actual de-pressurising. To de-pressurise manual valves are opened and the hydrocarbon escapes to flare. That particular part of the process we were told takes about 5 minutes. Then the operator would isolate the lines that sends this pressure to flare. After that the operator opens the zero vent line to the atmospheric vent system. This runs out to the flare system without any other control valves. From start to finish the de-pressurisation procedures could take up to 2 hours. The defenders claimed that these procedures have some implications for the re-pressurising of the pump.
Even if a pump tripped and the other pump was pressured-up and on standby the first step would be to try to restart the pump that had tripped if this seemed feasible because this is generally easier that trying to start a pump on standby. The one fact that the operatives seemed to be clear about is that the pumps were somewhat temperamental both as to their propensity to trip and to their readiness to restart. A pump on standby would not only be pressurised but would be set at manual speed and wound back to facilitate bringing it back to operation. If the pump which had not tripped was available but not pressurised it would be necessary to re-pressurise it before attempting to start it. This was done by the gradual process called jagging and this involved opening the suction GOV for short periods on a number of occasions so that a limited amount of hydrocarbon was introduced to the pump at each stage and the damaging stress on the pump which would result if all the pressure was introduced at once could be avoided. During the jagging process the operator could see what was happening by means of a suction pressure control gauge that was accessible to him as he jagged. Even if the pump which had been de-pressurised had been de-pressurised and reserved for a planned maintenance if this programme had not actually begun the procedure would be firstly to try and restart the tripped pump and then to retrieve the other pump if the tripped pump could not be started. Preparing the second pump for re-introduction would be a matter of "major concern". As the witness Mr Henderson also said if the tripped pump could not be restarted, if it was thought helpful in relation to restarting that pump, electricians and instrument technicians would be summoned to help. In fact just before the explosion Mr Young, an Instrument Technician had been summoned to the 68-foot level. Thus the procedure which Mr Vernon is alleged to have followed would have been perfectly normal had PSV 504 not been absent. It would not be too difficult to suppose that he would have acted in the fashion the pursuers ascribe to him were the situation that he did not have in his mind that PSV 504 had been removed from its site. The situation I have been referring to would always be accompanied by at least a measure of urgency because condensate would still be filling the collection vessel. However some time in this respect could be bought by unloading and recycling the reciprocating compressors thus slowing the condensate flow. It should be perhaps noted that there was on the local control panel a green light which shone if electrical power to the pump was switched on.
No re-pressurising of a pump could take place without re-connecting the air supply (assuming that this had been disconnected) but this would only take a few minutes. The time required to jag the pump appears to depend to a degree on the practice of the operator who is doing it and the time seems to vary between about 30 seconds and 5 minutes. There does not seem to have been an established procedure in respect of the timing of a jagging process. However it would be possible to jag the pipe fully in some seconds if this was desired. The number of jagging procedures required would also depend on how an individual operator went about jagging. The objective of the procedure was obviously to avoid a re-introduction of pressure to the pump which was too drastic.
7.3.2. Events at the 68-foot Level
On the evening of the accident Pump B tripped just before 9.45pm. Upon receiving the alarm that the pump had tripped Mr Bollands alerted Mr Richard and at the same time Mr Vernon "grabbed his helmet" and went down to the pump. Mr Bollands did not find Mr Vernon’s immediate concern surprising for as he said if the condensate injection pump goes down the whole gas plant may trip so that it is essential to act quickly. Mr Bollands reckons that Mr Vernon had left the Control Room about a minute after the alarm had annunciated. A further minute at most would have taken Mr Vernon to the Control Room.
So far as Mr Bollands was concerned the next event was that the JCP Panel alarm went off and he notified Mr Richard. He asked Mr Richard if he had unloaded the reciprocating compressor but cannot recollect his reply. Mr Bollands estimates that about three or four minutes had passed between the pump alarm and the JCP Panel alarm. I think almost all the estimates of time have to be taken with some caution since it is difficult for witnesses to be accurate about such short times. I think the JT Flash drum can be discounted as the source of the explosion since this would have leaked into the 68-foot level. Moreover Mr Grieve walked passed the JT Flash Drum some short time before the explosion and noticed nothing remarkable other than that a high level alarm was indicating.
At the time the pump tripped Mr Grieve had been in the Gas Conservation Module and he had noticed the radio conversation about the pump between Mr Bollands and Mr Richard. Some minutes later Mr Grieve went down to the 68-foot level to see if he could give the men there a hand. All he can remember is that from the conversations over the radio it was clear to him that they were not getting the pump started. Nevertheless he regarded this as a normal routine problem. His estimate as to how long before the explosion he decided to leave the gas conservation module is from five to ten minutes. On the other hand he thought that he had only been at the 68-foot level for a few minutes before the explosion. Walking down to the 68-foot level and passing through Module C he noticed nothing unusual. This might suggest that there was no appreciable quantity of gas at the east end of the module at the time when he passed otherwise he might have smelt it.
Mr Bollands has Mr Vernon coming back into the Control Room about ten minutes after the pump alarm had annunciated. This timing must be approximately right considering what must have passed since the alarm. At the point of his return to the Control Room Mr Vernon seems to have supposed that the trouble relating to Pump B stemmed from a problem with the lube oil system. Mr Bollands in relating what then occurred said "He wanted to get the other pump going, he wanted to take it back off maintenance." If at that point Mr Vernon had tried to contact any of his superiors over the radio the likelihood is that his message would have been heard by others. The practice for operators was to keep their radios on all the time. Further if Mr Vernon had contacted his superiors by telephone from the Control Room Mr Bollands would have heard this. Mr Vernon’s intimation to Mr Bollands that he wanted to take Pump A back from Maintenance caused the latter no concern which of course is consistent with Mr Bollands having no knowledge of the position relating to PSV 504. When Mr Clark was contacted about the retrieval of Pump A for production according to Mr Bollands he was "happy" about that suggestion. After leaving the Control Room (possibly having met Clark and had a discussion with him) Mr Vernon returned directly to the 68-foot level. Between Mr Vernon’s return to the Control Room and the first gas alarm he would have had enough time to jag the pump especially when it is considered that at best timings must be rather approximate. Of course whether any jagging operation took a few seconds or 30 seconds and whether it comprised one operation of the push/pull button or more would depend on how it was done. Certainly it looks as if about 15 minutes lapsed between Mr Clark receiving his Tannoy call from Mr Vernon and the explosion. Thus even allowing for a measure of flexibility in the timing of incidents like the first alarm there was space for Mr Vernon to have jagged the pump. Indeed if the jagging at that point consisted of two consecutive short operations close to one another in time there would even have been time for this. The pursuers of course pitch their main case at there having been two quite separate jagging procedures. This must be the minimum they require since there were seemingly two distinct leaks causing two separates alarm responses. Indeed it is the staggered operation involved in the re-pressurisation of the pump that they say justifies their view that a leak during jagging caused the accident. However it is more probable, I think, that if Mr Vernon (or for that matter Mr Richard) jagged the pump he must have completed the first stage of jagging (the one causing the first low level alarm) before Mr Grieve came on the scene. There would probably not have been enough time left after Mr Grieve arrived and certainly Mr Richard went away soon after Mr Grieve’s arrival. Moreover had there been two distinct jagging operations when Mr Grieve had been present then the possibility that he would have noticed this happening must increase substantially. On the other hand if Mr Vernon or Mr Richard sought to re-pressurise the pump before Mr Grieve arrived on the scene we can only surmise just how the operation would have been carried out since there were no eye-witnesses. Different operators seem to have employed their personal and individual jagging techniques. Thus it is perfectly possible that if there was an attempt to re-pressurise the pump, at a point in time before the first alarm and before Mr Grieve arrived, two separate lots of hydrocarbon were admitted into the pump at that stage. It is equally possible on the basis of the evidence of Dr Davies that the first entry of material if it caused a leak would not have extruded sufficient hydrocarbon to have set-off an alarm. When the valve was opened again then the material that emerged may have been heavier gas and triggered an alarm. With the pump leaking Mr Vernon may have decided some minutes later that the pump was still not fully pressurised and at that stage introduced the gas that caused the second stage alarms. The pump had a pressure gauge so that the person jagging the pump could see if the pump had been adequately pressurised. If Mr Richard had been the person who first jagged the pump Mr Vernon may have looked at the gauge to see if the pump was satisfactorily re-pressurised. Given that within limits jagging procedures vary it would be difficult to be certain as to just what the jagging sequence was and the only indication that gives positive information is the pattern of the alarms. If the pump was jagged against some minutes after the first insertion of hydrocarbon to the pump this may explain why there was a delay between the first escape that registered and the second.
The tags appear to have been left with Mr Clark after he had signed them. This would follow from Mr Clark’s own account of events since he has Mr Vernon out of the Control Room before he arrived there. Thus as far as Mr Vernon was concerned Mr Clark had the tags and was arranging for the electricians to de-isolate the pump. Mr Clark’s instructions had never been cancelled. This all leads to the important conclusion that the arrangements to re-start Pump A were going ahead and indeed were in train when the explosion occurred. It should also be noticed that a decision to re-introduce Pump A did not mean abandonment of plans to start Pump B as soon as possible. Mr Clark in his account of his telephone conversation with Mr Vernon (direct or indirect) said that the objective of re-introducing Pump A was to secure more time to work on Pump B. As Mr Clark added there was obviously a problem with Pump B so we needed more time to investigate it. Thus at least the objective would not be to spend time with Pump B if this would be at the expense of getting Pump A started.
Mr Clark seemed to think that he had telephoned Mr Savage from his own office after his communication with Mr Vernon but before he went to the Control Room. Since Mr Clark arrived at the Control Room before the first gas alarm his contact with Mr Savage must have been some minutes before that first alarm. If Mr Clark is right and Mr Vernon was away from the Control Room when he arrived there then Mr Vernon may well have left the Control Room as soon as he had done all he had to do there by making his arrangement with Mr Clark. This too could demonstrate that there was a space of time between Mr Vernon’s return to the Control Room and the first gas alarm.
It can be observed that the electrical switchgear for Pump A was at the 107-foot level in the Utilities Mode. Of course Mr Vernon may not have realised how long it would take Mr Clark to have the pump electrically de-isolated so that he may have had the intention to press on with the de-isolation of the pump as quickly as he could reasonably do so.
The evidence of Mr Grieve as to what was happening on the 68-foot level at the time of the explosion is particularly important because he was one of the two survivors in that position (the other being Mr Young). Mr Grieve says that when he arrived at the 68-foot level it was rather noisy. This comment is not surprising because there is a lot of noise in the working areas of a platform and this is why the men working there often wear ear protectors. I mention this because it could have a bearing on a suggestion that perhaps Mr Vernon had any kind of elaborate conversation with Mr Richard. To get to his worksite Mr Grieve had ducked under the JT Flashdrum. Mr Grieve walked past the control panel JCP57 up towards Pump B. He noted that the level in the Flashdrum was relatively high and that Pump B was not working. However the higher level was still well short of the danger level. He also noted that the GOVs for both pumps were closed. Of course Mr Grieve may not have known of the intention to de-isolate Pump A and he may well have assumed that the focus was on Pump B which he knew had caused trouble. As he approached the pump he became aware of Vernon and Richard at the west side of Pump B and at the foot of the GOV on the south side. Now Mr Grieve according to his own recollection was at the 68-foot level for about two or three minutes before the explosion. He does not claim that Mr Vernon and Mr Richard were doing anything in particular when he first caught sight of them. If the explosions were caused by two distinct jagging operations and there was as witnesses describe a few minutes between the first gas alarm and the second flurry of alarms then one would expect a similar time difference between the two jagging operations.
By signals from Mr Vernon (which Mr Grieve describes as shouting or sign language) it was indicated to him that another attempt would be made to start Pump B. He cannot remember if he personally reset the GOVs but they seem to have been set and another attempt made to restart the pump. Mr Grieve makes an important concession for when asked what the others were doing at the time he says that it is a difficult thing to say if you see people going about functions that are done routinely and he states "I can’t say I was aware of exactly what was being done at the time". Given that Mr Grieve had just arrived at the scene I cannot say that this is surprising. However he does say that he had the impression that the others were in the vicinity of Pump B and a decision was made to have as he puts it "another go at starting B Pump as GOVs for B Pump were re-latched at that period". He cannot say whether it was Vernon or Richard who reset the GOVs for the pump. However that may suggest that the work on the GOVs was done when Mr Richard was still there for not long after Mr Grieve’s arrival Mr Richard departed. This would coincide with the message he received from Mr Bollands that a gas alarm relating to Module C had gone off. It is perhaps significant that the message received by Mr Richard which presumably Mr Vernon also heard over the radio that an alarm had gone off at the east end of Module C did not appear to have concerned Mr Vernon as might have been expected if he knew that he was doing something irregular affecting that area. Mr Grieve accepted that he personally had carried out some function at the Control Panel and this would have taken him to a position about half way between the two pumps. Indeed he recollects that he pressed the button for Pump B, watched the motor turn, stopped the motor when he took his hand off the button, and then proceeded to the torque converter to make sure that it was at manual with the speed wound off it. He claims that he never actually got a chance to check the speed controller before the accident. When he was asked if he was aware of anyone else at that stage he said that he remembers seeing Mr Richard called away. This was the last time that anyone saw Mr Richard because he did not re-emerge after the accident. The interesting point is that in response to the last inquiry he makes no claim to have noticed what Mr Vernon was doing.
In the course of the duties I have referred to Mr Grieve also had to give some attention to the local pump panel. This comprised a cabinet with a door and this door had to be opened to gain access to the button inside. There were also some lights within the panel that could if circumstances were appropriate indicate alarms. Indeed he noted that the alarm for the JT Flash Drum was indicating. The point of giving these details is that there was the potential for Mr Grieve having his attention occupied as he performed some of his duties. The defenders argue that it was never shown that Mr Grieve knew that the injection pump A was not on standby. Indeed he himself claims that he did not know the state of the pumps and I have no reason to disbelieve him about this. However I am not sure to what extent it matters. Indeed if the case was that Mr Grieve was not aware that Pump A was de-pressurised this may have been a reason why he would not readily have supposed that Mr Vernon was jagging the pump. In any event whatever Mr Vernon was doing Mr Grieve seems to have had no precise knowledge of this.
The defenders argue that whatever Mr Vernon had been doing when Mr Grieve arrived he was not at the GOV push/pull button. A little later he saw Mr Vernon and Mr Richard at the west side of the Condensate Injection Pump B towards the south end. He did not know what they were doing. Of course Mr Vernon had already indicated that he considered the lube oil system as being the source of the trouble with Pump B and it is possible that when Mr Grieve first saw him he was inspecting it. He also indicated that when he arrived Mr Vernon signalled or called out to him that they would go for a run on the B pump. Mr Grieve was not sure if he was the person who had reset the pump. The last Mr Grieve saw of Mr Vernon before the accident Mr Vernon was between the GOVs on the pumps. He did not say whether he was coming or going. Thus he had no idea what he was doing in that area. For that matter Mr Grieve does not specify what he himself was doing at that point of time. He thinks he may have been walking round the local panel. Of course it is consistent with this evidence that Mr Vernon was not proceeding to jag Pump A but was returning from having done so.
I must add that in relation to what was happening at the 68-foot level and elsewhere just before the accident the very short timescales remembered by witnesses cannot be taken too literally.
If Mr Richard had arrived in Module C after an initial jagging and after the first gas alarm had annunciated then by the time he arrived there may not have been any obvious escape of gas. Indeed two quite separate escapes of gas dispersing quite quickly and with an interval between them may explain why he sent no immediate signal to Mr Vernon or Mr Bollands that there was a serious gas escape. If there had been a steady leak of gas filling a substantial part of Module C Mr Richard might have been expected to pass an emergency message on to Mr Vernon or Mr Bollands at once. On the other hand a leak from the PSV, if there was one, would have decayed and much of the initial leak of gas would have been transported eastwards along the module. However if the explosion was caused by a second jagging process then this would have occurred so close to the explosion that Mr Richard may not have had time to contact the Control Room particularly since by that time Mr Bollands was heavily occupied with a flurry of alarms. On the other hand Mr Grieve said that Mr Richard left the 68-foot level about a minute before the explosion.
Mr Grieve last saw Mr Richard going up the stairs to the south side of the pumps (which leads to C Module.) This happened after Mr Richard had received a call which Mr Grieve says he did not listen to. Apparently radio messages are directed towards particular participants and Mr Grieve said that if the message was not for you one might not listen "if you were busy". He stated that you would carry on with what you were doing and not listen to the message. This may suggest that Mr Grieve was occupied when Mr Richard was called away. The message that Mr Grieve would have received was that an alarm had gone off in zone 3 of Module C so that it is not difficult to suppose that he was proceeding to the east end on that Module. According to Mr Grieve the explosion was about one minute after the last time he saw Mr Richard.
As far as Mr Young is concerned he cannot say much about what preceded the explosion. He came down the stairs to the 68-foot level after Mr Richard had left there. He had only taken a step or two after coming off the stairway when the explosion occurred. It is perhaps significant that Mr Young did not encounter Mr Richard on the stairway since this adds support to the view that there was a material interval between Mr Richard leaving the 68-foot level and the explosion. He thought that when he arrived Mr Vernon and Mr Grieve walked towards him to speak to him but the explosion occurred before they got together. Mr Young heard a strange sound before the explosion. Mr Grieve does say that as he was attempting to start Pump B he was watching the motor which may suggest a degree of pre-occupation. When the motor would not start Mr Grieve went round to the speed controller and it was then when the explosion occurred.
Mr Young went to the 68-foot level after receiving a radio message to go there. This is consistent with Mr Clark’s evidence that Instrument Technicians would be called to assist in starting a tripped pump that was resistant to restarting. It is perhaps significant that Mr Young was at the west side of the platform shortly before the accident because an idea that was canvassed was that the blue flash noted by Captain Clegg may have been caused by a trail of escaped gas along the west face of the platform. If there was such a trail it would appear that Mr Young did not notice it. When Mr Young arrived at the 68-foot level Mr Vernon and Mr Grieve were near the condensate injection pumps but they apparently moved towards him perhaps to get into a better position to talk to him above the noise. However before he got a chance to speak to them there was a sound like a quick escape of steam and this ended in a dull thud. Dr Mitcheson ascribed this noise to the gas moving at the explosion.
The defenders make the evidence of Mr Grieve one of the lynch pins of this aspect of their case. Their argument put simply is that Mr Grieve’s evidence makes no mention of the jagging of the pipe by Mr Vernon. This evidence they say was not the subject of cross-examination so that the pursuers must be taken not to have challenged it. The defenders therefore claim that there is uncontroverted evidence of the fact that Mr Vernon did not jag Pump A during the time Mr Grieve was at the 68-foot level which was the space of about three minutes before the accident. If the facts were that Mr Grieve’s evidence positively excluded the possibility that Mr Vernon jagged the pump during the period his evidence covers then clearly this would make the pursuers’ cases difficult. I may say that I have no trouble in concluding that Mr Grieve was an honest witness trying faithfully to remember what he had observed. I doubt whether the pursuers would disagree with that or otherwise they would no doubt have cross-examined Mr Grieve. I think the pursuers did not challenge Mr Grieve because they accepted that all he describes is precisely how he remembers the situation and he did accept the possibility of imprecision in his recollection. Mr Grieve did not give evidence that Mr Vernon did not jag the pump. What he did was to give evidence that he did not notice Mr Vernon jagging the pump and it was this that the pursuers must be deemed to have accepted. In fact going beyond that I think the pursuers must be taken to have accepted that what Mr Grieve is accurate so far as it goes. If this is so the pursuers had no need to challenge Mr Grieve because they accept what he says. The question of course is how far does Mr Grieve’s evidence actually go.
Mr Grieve states what he saw and that certainly gives little support to the suggestion that while he was present Mr Vernon jagged Pump A. Of course the question is whether what Mr Grieve says excludes the jagging of the pump. If Mr Grieve is effectively saying no jagging of the pump took place while he was there then that evidence if unchallenged would as the defenders contend be bad for the pursuers. However he in no way goes as far as that. All he is saying is that he personally did not see any jagging. However that must be taken against the whole context of his presence at the 68-foot level. The positions he gives as relating to Mr Vernon only cover discrete points of time in relation to what Mr Grieve was doing. Moreover he accepts that he was not concentrating on just what Mr Vernon was doing apart from taking a general interest in starting Pump B. Whereas these positions do not support specifically that Mr Vernon was jagging the pump they do not negate that suggestion either. After all we are discussing a procedure that would have involved Mr Vernon in moving a few yards and taking perhaps 30 seconds or so. Mr Grieve was not standing watching most of the time but rather was engaged in his own work. As he himself observes he may well not have registered the routine work of his colleagues if engaged in his own tasks. When he arrived at the 68-foot level his thoughts were not on jagging Pump A since he thought that pump was pressured up and on standby. Thus he would scarcely have expected Mr Vernon to be jagging the pump and could easily have missed that if it occurred. If the only evidence that Mr Vernon might have been jagging the pump was the eye witness evidence of Mr Grieve then I could not possibly have concluded that he was thus acting. However Mr Grieve’s evidence does not stand alone. There is good evidence that Mr Vernon went to the 68-foot level with the express intention of starting Pump A. There is other expert evidence that the jagging of the pump would be consistent with the circumstances of the explosion. I do not see the evidence of Mr Grieve as defeating the inferences about jagging that might be drawn from other facts and circumstances. It is true that if the pursuers had seen fit to ask Mr Grieve if he could possibly have missed what Mr Vernon was doing they may have unearthed strong support for their hypothesis or conversely unearthed additional information which would have defeated it altogether. However the same could be said for the defenders. I can only decide the case on the evidence actually put before the court and the inferences that can be drawn from that evidence as it stands.
The pursuers have pleaded a case that the noise heard by personnel shortly before the accident was consistent with the hypothesis that the accident was caused by a leak from PSV 504. However this case was developed somewhat half-heartedly compared with the effort that went into other aspects of the case. For example I did not have the benefit of expert evidence about acoustics. However there was some relevant evidence. Nevertheless it must be observed that there was evidence from Dr Mitcheson that locating the source of noise is a very subjective matter (as anyone who has tried to react to a foghorn will know).
Dr Richardson claimed that he had been present at a release when gas was leaked through holes and he describes the resultant noise as a very high pitched screech. This would tail off as the pressure decreased.
Witnesses who were present in the Instrument Workshop or the Tea-room heard a noise before they were conscious of an explosion. Other witnesses heard nothing significant but then they were often further away from Module C where the explosion is alleged to have taken place or in noisier areas of the platform. At least out on the Modules ear protectors are often worn.
Mr Young heard a rushing noise just before he was hit by a blast of gas but as I have said Dr Mitcheson thought that the likeliest cause of this was fast moving gas propelled by the explosion. On the other hand Mr Grieve heard a noise above his head sufficiently loud in the noisy atmosphere where he was working to cause him to fall on his knees (and he was wearing ear muffs).
There was some evidence about noise from Mr MacGregor who was a Mechanical Technician and was in the Mechanical Workshop Tea-room when the accident occurred. Mr MacGregor describes how he was sitting talking when all of a sudden there was a loud noise like a banshee screaming. This description requires a certain amount of imagination since the phenomenon he mentions is not one with which many of us are directly familiar. In any event the noise was said to have lasted about half a minute. The noise disturbed the company he was with. He thought that it must be due to trouble with the air start of the west crane. It is interesting that he associated it with pressurised gas such as the air in a crane. This does not appear to have been the same kind of noise that Mr Young described because Mr MacGregor says that conversation stopped and then started up again just before the explosion. A minute or two afterwards the explosion occurred. There was blackens and the lights and fittings fell down.
Mr Thompson was a Maintenance Rigger and he was in the same location as Mr MacGregor. He also speaks of a loud screeching noise. He was surprised by the volume of this noise. The explanation for the noise that was considered by this witness was that someone had got stuck in machinery so obviously we are dealing with a very unpleasant sound. He thought that the noise had lasted for about 10 seconds. He also thought that the noise was ten times louder than any sound that may have come from the air of the crane could be. Twenty or thirty seconds or so after the noise there was an explosion. Again the explosion was accompanied by physical phenomena like the collapse of the ceiling. It seems clear from the evidence that the explosion was accompanied by a fair degree of violent vibration and some of the damage after the explosion may be attributable to this. Mr Thompson accepted that the sound he heard could have been caused by steel ripping apart.
The witness Mr Cassidy was an Instrument Technician who was in the Instrument Workshop when the explosion occurred. He describes how before the explosion he heard a very, very, high pitched noise like metal grinding at high speed. It was a frightening noise such as he had never before heard in his life. The noise lasted a short time perhaps ten seconds. As soon as the noise stopped the workshop began to shake and he was knocked off his feet. Apparently when Mr Cassidy gave evidence at the Cullen Inquiry he said that the noise he had heard had lasted some minutes. This confusion in his evidence may not matter. In these proofs a lot of time had passed since the Cullen Inquiry and in any event in the circumstances experienced by these witnesses I would not take timings too literally. On the other hand the timings put forward by Mr MacGregor and Mr Cassidy were not challenged and were fairly consistent one with the other.
Mr Standen in his tests had found that escapes of gas through a leak could produce a high pitched noise.
The evidence given by these witnesses seems in general to negative any suggestion that the main cause of the explosion was an escape of hydrocarbon some minutes before the explosion (rather than seconds). Moreover the location of the witnesses might suggest that the likelier source of the noise was Module C and not Module B which was further away. Indeed the divers who had been working closer to Module B do not speak to having heard a similar noise. I do not think it can be doubted that the noise heard by the witnesses I have been discussing was related to the explosion and that gas escaping through a leak can in certain circumstances cause a noise.
7.6.Conclusions about cause of accident
7.6.1. Outline of the Pursuers’ arguments
The pursuers, as might be expected, were anxious to convince me that the accident had been caused by the explosion of a heavy gas which leaked out of PSV 504. They contended that the explosion occurred at the east end of Module C and that is consistent with the fact that the alarms which annunciated related to that end of the Module. Moreover the escape of gas according to the pursuers can be related to contemporaneous activities at the 68-foot level. Mr Vernon not only intimated to Mr Bollands an intention to start Condensate Injection Pump A but he had the opportunity to begin the procedures needed to secure that objective. Counsel argued that it would be highly co-incidental for gas to have accumulated at the east end of Module C and yet not be the cause of the explosion. It would be a further co-incidence that the activity of introducing a heavier than air hydrocarbon at the pump was unrelated to the accumulation of gas at the east end of the module at the same point of time. The pattern of the alarms support a limited release of gas followed some short time later by a further release. It was said that the evidence showed that the blind flange would have leaked if it were insecure and the amount of gas released would depend on the size of the leak and the conditions within the pump. Because the size of leak would be related to the extent of the re-pressurising of the pump one could have a small release succeeded by a much larger release. The general picture of the leak and its potential could be arrived at without reference to the computer work carried out by Dr Bakke or Dr Richardson and also without the detailed wind tunnel tests carried out by Dr Davies. The pursuers’ case is that the firewalls were breached and the breaches established were totally consistent with the site and extent of the explosion which the pursuers seek to demonstrate. Moreover it was said that if there was an explosion which breached the firewall then it could be readily inferred that fragments of the wall would be propelled under some force and that this phenomenon could itself damage the condensate line within Module B and cause it to leak. It was, so it was claimed, established that the large fire in Module B at an early stage of the accident was fuelled by condensate and this fact goes to show that the condensate line was damaged by the initial explosion. The pursuers were at pains to point out that they were not depending on any individual circumstance but rather on the way the circumstances fitted into one another and combined. It was claimed that the hypothesis that the explosion happened in Module B was in no way supported by the evidence. The fact that there were two flurries of alarms goes to support the inference that the escapes of gas occurred during a multi-stage process such a jagging. Dr Davies it was suggested confirmed that for only one gas alarm to have triggered at the first stage any appropriate release of gas would have to have formed a shrunken cloud. One release of gas during jagging if it encountered an insecure blind flange at PSV 504 would generate a gas cloud downwind of the valve. It was said that Dr Davies should be read as saying that depending on the size of any leak the time taken for the cloud of gas to form and trigger alarms would range from 15 to 40 seconds. Even without the modelling tests it was clear on the qualitative evidence that it would take only a small lapse of time for any escape from the blind flange to arrive at the point where alarms would be triggered. If the initial release was a limited release then by the time Mr Richard arrived at Module C to investigate it would have dissipated sufficiently to make the leak hard to identify. When Mr Vernon realised that Mr Richard had been called away to investigate a gas alarm the fact that he proceeded with his work suggests, so it was contended, that he did not connect an alarm in Module C with what he was doing. He would have made the connection more readily if he had been aware at that point that PSV 504 was missing. Indeed this was one of the factors I accepted as being relevant to forming a view about the state of knowledge of Mr Vernon. It was said that after Mr Richard went upstairs to investigate Mr Vernon would have carried on with his work and jagged a second time. A consequential leak would explain the second flurry of alarms. The greater the spread of the gas then the more likely it would be that not one, but a flurry, of alarms would go off. Although there were a number of detectors perhaps designed primarily to detect gas within the compressors they could also detect gas outside the compressor compartments. The experts estimated that the sort of cloud that emanated from a leak at the blind flange might occupy about one tenth of the volume of the Module and would be driven by the airflow to the east end. Dr Bakke considered that the kind of gas cloud that is being postulated could if it exploded develop an overpressure of between about 0.2 and 0.7 bar and that these pressures would endure for about 100 milliseconds or more. This has a bearing on the response of the C/D firewall to any explosion because being a stiffer wall than the B/C wall it would have a shorter natural period and thus would respond in a static rather than a dynamic way. The evidence was that the overpressures that the postulated cloud could generate upon explosion would have been more than enough to cause penetration of the C/D wall. It was submitted that the evidence shows quite clearly that both the B/C and C/D firewalls failed so that the latter wall being stronger if it failed it is not difficult to conclude that there would be more than enough overpressure to collapse the other wall. Moreover the C/D firewall was stronger than the A/B firewall so that if there had been in Module B an explosion sufficient to penetrate the C/D wall then certainly one would have expected the A/B wall also to have been penetrated. It was pointed out that the evidence was that the A/B firewall remained intact. Dr Richards calculated that with a hole of a certain size and an upstream pressure of 46 bar then a release rate of 180 kilograms per minute would be generated. Dr Davies calculated that a release of the order of 150 kilograms per minute could after being sustained for 30 seconds create a flammable cloud of about 45 kilograms. Dr Mitcheson and Mr Cubbage thought that that size of cloud was the size likely to have caused the particular explosion which occurred. I was asked to accept Dr Palmer’s evidence that with such an explosion the pressure pulse generated on the B/C firewall would have been of sufficient energy to cause a projectile capable of damaging the condensate line. Pursuers’ Senior Counsel saw the expert evidence of the pursuers to have a general consistency. It was contended that the pursuers can comfortably establish their hypothesis even without the evidence of Dr Palmer.
Dr Drysdale thought that the inventory of fuel oil that might have escaped from the small bore piping in the vicinity of the fire would not have caused a sufficient pool of fuel to have sustained the fire which later developed. He postulated that the most likely explanation for a continuing oil fire in excess of the available inventory was backflow from the MOL via ESV 208 it having failed to close. The ESV not only had the electrical power which normally drove it but it had a back-up nitrogen system. However the nitrogen was transmitted through a valve from the pneumatic connection and it was argued that these would be vulnerable to fragments propelled by an explosion. The pursuers urged me to accept Dr Drysdale’s hypothesis about the fuel supply. It should be noted that quite soon after the explosion men located near the Divers’ Skid noted oil running freely down the MOL. In relation to the explosion causing damage to small bore pipe work carrying oil it should be noted that at the time of the accident the prover loop was missing and since the prover loop normally sat on top of small bore pipework it would have provided some protection to the pipework which would at the time of the accident not have been in place.
If an explosive cloud leaked from the blind flange at the suction side of the site of PSV 504 then this, it was contended, inevitably reflected adversely on the care taken by the fitter, Sutton who had fitted the blind flange. A flange which has been properly fitted and tightened by the prescribed procedures should not leak hydrocarbon. Since the blind flange in question did leak the inescapable inference is that Mr Sutton failed to take reasonable care when he fitted it.
7.5.2. Outline of the Defenders’ Submissions
The thrust of the defenders’ attack on the pursuers’ case was that the pursuers had failed to prove the cause of the accident. The pursuers could not recover under the indemnities unless the cause of the accident were established. Without the cause of the accident being demonstrated the pursuers could not prove that they were under any legal obligation to settle with the victims and their families. It was not enough that the pursuers might claim that a particular suggested cause of the accident was a possibility. The circumstances of the accident had to be proved on a balance of probability. The defenders urged upon me the various rules and principles which I have set out above and which circumscribe the law of evidence.
According to the defenders because all the main witnesses did not survive the accident, and furthermore all the platform and its equipment rest irrecoverable at the bottom of the ocean, the precise cause of the Piper Alpha tragedy will never be known. At best the pursuers case is highly circumstantial. However if a necessary link in the circumstantial chain is shown to be untenable then the whole hypothesis derived from the circumstantial evidence must collapse. The pursuers seek to build up their case by relying upon coincidence based mainly on the fact that Mr Vernon expresses an intention to re-pressurise Condensate Injection Pump A. However there were numerous other hypotheses derived from coincidence which point to quite different results. Thus several days before the accident the platform had changed from Phase II to Phase I operation. This meant that the existing equipment required to cope with considerably enhanced operational pressures and the ageing equipment may not have been able to cope with this. There had been trouble with one of the Condensate Injection Pumps following upon the transfer to Phase I. It was curious that when the explosion occurred all three Centrifugal Compressors had tripped. The Prover Loop had been displaced at the time of the accident and this could have caused a hydrocarbon leak in Module B. On the day of the accident work had been proceeding in the area of the Prover Loop. The safety door between Module B and Module C may have been left open. The Flotta Terminal had registered a breakdown in production before the alarms and this too was more than co-incidence. The defenders made much of the blue flash reported by Captain Clegg and of the embarrassment Dr Mitcheson experienced when he dealt with this. The expert evidence, it was said, agreed that the blue flash described by Captain Clegg could not have been caused by an explosion in Module C. The pursuers had accepted Captain Clegg’s account of events which meant that they could not continue to maintain their circumstantial case. When Mr Grieve had approached the mouth of Module C after the accident he had not seen anything indicating a breach in the B/C firewall.
The defenders maintained that neither Dr Mitcheson nor Mr Cubbage had anything like the experience of hydrocarbon explosions enjoyed by Dr Bakke. Likewise Dr Drysdale was an expert on fire affecting solids and did not have the relevant experience of Professor Magnussen. The latter had suggested that both the original fire and the fireball were consistent with a pool fire in Module B and his evidence cannot just be discounted. No one had explained in satisfactory terms why if the fireball had been caused by the combustion of condensate released from the 4 inch line there had been a gap of 15 seconds or so between the explosion and the manifestation of the fireball.
The pursuers’ evidence about the alleged destruction of the B/C firewall was not satisfactory for the reasons indicated in the evidence of Professor Reid and Professor Stollery. At best for the pursuers a strong shadow of doubt had been cast over that evidence. In respect of their static analyses of the failure point of the firewall Professor Fenner and Dr Palmer had contradicted each other. Nevertheless Dr Bakke in his simulation had been asked to incorporate Dr Palmer’s failure value and if this was the wrong value Dr Bakke’s results were suspect. On the other hand Dr Palmer had relied on Dr Bakke’s pressure pulse in his own dynamic analysis so that the conclusions of these two experts are critically inter-related. Important factors including membrane stress, strain hardening, kinematic impossibility, collapse mechanisms and theory and flexural rigidity had not been properly taken into account by Dr Palmer and therefore in any event his conclusions cannot be accepted as valuable. Dr Bakke in his general evidence had expressed doubt as to whether the overpressure created by an explosion of 45 kilograms or so of hydrocarbon would have been sufficient to destroy the effectiveness of the firewall. Moreover he had noted with lighter structures than the partitions these had only moved away about 10 centimetres at the end of the explosive pressure pulse and this does not consist with the sort of movement in the firewall that the pursuers would require to create projectiles with sufficient energy to damage the 4-inch condensate line in Module B. In any event it is not possible to conduct an exercise such as ventured by Dr Palmer without a computer analysis.
The evidence of Professor Reid and Professor Stollery had to be considered in relation to projectiles and this evidence at the very least casts doubt on Dr Palmer’s calculations. He had not taken sufficient account of shock tube theory nor had he taken the support conditions of the condensate line into account.
The defenders did not accept that the A/B firewall had not been punctured by the explosion. If the C/D firewall had been punctured this could result from an explosion in Module B which was amplified by the process known as pressure piling. The C/D firewall was not necessarily seriously damaged by the explosion. As Professor Magnussen had explained the penetration of flame and smoke to the north face of the platform could be due to flame emanating in Module B being blown over the roof of the Accommodation Modules.
The defenders contended that the evidence of Dr Davies and Dr Richardson in fact destroyed the pursuers’ hypothesis. The important consideration was that any gas which might have escaped from the PSV blind flange would not have been heavier than air as Dr Davies’ modelling assumes but rather would have been neutrally buoyant. This is because of the flashing off of lighter ends when the hydrocarbon stream first enters the empty pump and also the accumulation of compressed air that has to be expelled from the relief line before liquid condensate can escape. In fact given the fact that the alleged first jagging would have produced a neutrally buoyant escape the modelling of Dr Davies shows that such an escape could simply not occur without a detector in the C2 zone triggering an alarm. This did not happen. Moreover Dr Davies acknowledged that the presence of scaffolding beneath the blind flange would have the effect of dispersing any escaped gas and this would prevent the formation of a compressed low lying cloud. such as was required to escape triggering the C 2 alarm. The evidence of Dr Bruun also casts doubt on the reliability of the modelling methods of Dr Davies.
The defenders also placed considerable stress on the fact that Mr Grieve was the only witness who could have seen Mr Vernon or Mr Richard jagging the injection pump but his uncontroverted evidence was to the effect that when he was present at the pump just before the accident he did not see this happen.
The defenders submitted that there was no evidence to suggest that Mr Sutton had not properly tightened the bolts on the blind flange. He was an experienced valve fitter. The bolt tightening operation was straightforward and there was no reason why he should have departed from standard procedures. Certainly none of the eyewitnesses suggested anything from which it could be supposed that he did so. Moreover he was not at fault in relation to any escape of gas due to a loose fitting flange. The combination of events that led to Mr Vernon attempting to start a pump when the blind flange was missing was totally unforeseeable. Moreover Mr Vernon’s intervention in the matter broke the chain of causation.
If Mr Vernon did cause the accident by attempting to re-pressurise the pump when the PSV was missing then his actions were reckless and amounted to wilful misconduct on his part. In any event the Company had been guilty of wilful misconduct in tolerating irregularities in the permit to work system and in failing to monitor and audit this and the handover system.
In a case of this dimension both the pursuers and the defenders developed many issues in their submissions and those I have outlined are only the more salient of these.
7.5.3. The Cause of the Accident
The Piper Alpha disaster is one of the worst tragedies, if not the worst, ever to have been suffered by the Oil Production Industry. To establish the cause of this accident is of critical importance to the parties in the cases before me but because of the extent of the catastrophe it is also of considerable interest to the industry and the public. Of course I have to make a decision about the matter within the limits of the evidence the parties have led before me and within the technical confines of the rules of Procedure and Evidence. However the issues surrounding the cause of the incident were explored before me in considerable depth (and also I may say at considerable length).
The defenders sought to persuade me that the elucidation of what caused the accident was a matter of such complexity and difficulty that no firm conclusion as to what happened could properly, or indeed safely, be pronounced. There is no doubt that such evidence about the accident as is available is considerably less unarguably decisive than I should have preferred. Many of the platform personnel who had direct involvement in the events alleged to have caused the accident (or the circumstance relevant to their determination) did not survive the accident. The result is that I am left to decide many mysteries which those most directly involved could have unravelled with likely authority had they been able to give evidence. The same position arises in relation to such parts of the structure and its equipment as could have cast light on what occurred. The platform and the equipment on it are, as far as the parties were concerned, irrecoverably lodged at the bottom of the North Sea.
A particularly difficult situation arises in respect of OPCAL’s Lead Production Operator, Mr Vernon, and the Valve Fitter, Mr Sutton, who was employed by Messrs Score. Both these men tragically died in the accident yet it is said that their fault was the cause of this dreadful event. In fact the pursuers’ case depends on this being established. Obviously it is a matter of great concern to be asked to blame an accident as serious as the present one on two workmen who do not even have the opportunity to explain themselves and give their version of what happened. I am well conscious that the situation I have just mentioned calls for particular care and that the employees in question cannot have blame for the accident attributed to them unless it is clear that this is justified. The core argument of the defence is that such evidence as I may find acceptable is so flimsy, the inferences needed to support the pursuers’ case so uncertain and complex, and the possible causes of the accident so multifarious that the only reasonable course available to me is to hold that the cause of the accident simply cannot be proved and indeed may never be proved.
To extract from the huge array of available evidence a convincing explanation for the accident is indeed a formidable task and it therefore would be tempting to follow the course urged upon me and declare that it is just too difficult to arrive at a firm view. It also has to be admitted that no single feature of the evidence points so strongly in one direction that the solution of the case could be found in it. However as so often is the case in such matters individual pieces of evidence themselves inconclusive can when viewed together paint such a clear picture that only one conclusion is possible.
In the cases before me there is factual evidence spoken to by witnesses and opinion evidence explained and spoken to by a number of highly qualified experts. The evidence I found most important and useful was that of the facts and circumstances spoken to by witnesses with direct knowledge of them. Much of the technical evidence spoken to by experts was somewhat speculative and often disputed. Had I had only the technical evidence available I should not have been able to arrive at any conclusions. However the technical evidence can itself acquire added significance if it fits in well with the facts established by the direct evidence of lay witnesses.
The most important and incontrovertible fact affecting this aspect of the cases is that some minutes before that explosion that triggered-off the accident the Lead Production Operator Mr Vernon signalled his intention to attempt to start up Condensate Injection Pump A. At the time the pump had been withdrawn from Production for a planned maintenance procedure and moreover the relief line from the pump was interrupted because PSV 504 been removed for calibration. The end of the relief line which had thus been opened was supposedly sealed by a blind flange that had been fitted by Score’s employee Mr Terrence Sutton. When he indicated that he was about to re-pressurise the pump Mr Vernon was in immediate control of the production processes. Moreover not only did he intimate an intention to re-pressurise the pump but there is little doubt that he took at least certain practical measures to execute the operation he intended such as having isolation tags signed-off and arranging to summon electricians to de-isolate the Pump. Before he could start the pump the empty pump chamber had to be re-pressurised by the admission of the main flow of condensate and this operation would have been carried out in a number of stages by opening the main suction GOV leading to the pump by a series of controlled operations know as jagging.
The important implication of Mr Vernon’s intention to start Pump A is that it is recognised as dangerous practice to start a pump with a relief line having only a single PSV (like the pump under consideration) if that valve has been removed. The main need to have a PSV present is that if the pressure in the pump is suddenly increased beyond its normal operating limits by a surge in pressure then the PSV will automatically open and relieve pressure by allowing some of the condensate to escape to flare. To jag a pump will introduce to it a certain sudden surge in pressure. Moreover it was practice on the platform that if the open end of any line was closed-off by a blind flange then pressure would not be re-introduced to the line without the blind flange being pressure tested. There was satisfactory evidence that if a blind flange was properly fitted with its bolts adequately tightened then it should be able to contain the flow material in the line without significant leakage. Presumably the practice of pressure testing (which seems to have been a mandatory requirement on the platform) was a double precaution to ensure that a particular blind flange had in fact been properly fitted. There is no doubt that the prevention of hydrocarbon escapes was perhaps the major safety consideration on the platform. At the time that Mr Vernon was proposing to restart the pump there was no suggestion that the blind flange on the relief line had been pressure tested. Thus the pursuers’ case begins with a situation that just before the time the accident occurred Mr Vernon was proposing to initiate a procedure that the witnesses recognised was irregular and dangerous. The connection in time between the explosion and Mr Vernon’s highly irregular act must make the re-pressurising of the pump a prime suspect as a cause of the accident.
Of course a declared intention to commit an irregular act would not alone be enough to prove that this act had occurred and caused the accident. However the fact that there were two distinct phases in the annunciation of gas alarms just before the accident indicates the likelihood that there were two distinct escapes of gas separated by some minutes. Dr Davies had no doubt on this point. The introduction of condensate by way of jagging the pump (assuming that the relief line leaked) would provide just the explanation for the two leak episodes. So far as the evidence goes nothing else was occurring on the platform at the time likely to result in a two stage leak. Defenders’ counsel suggested that an escape of gas from a relief valve could have created a two stage leak (such as for example when compressors tripped). This may be an interesting possibility but it was not explored with any of the expert witnesses. It is also to be noticed that Mr Richard must have arrived at the east end of Module C just before the explosion. He would it is to be assumed pass directly to the east end of the Module since that was where he had to fulfil the urgent task of tracing the gas leak. He was only there for a short time before the accident but nevertheless the fact that Mr Young did not pass him on the stairs combined with the evidence of Mr Grieve would suggest that he had at least a short span of time before he was consumed by the explosion. Mr Richard made no immediate report of gas over his radio. This may be consistent with a situation where the leakage was not continuous.
The airflow in Module C at the time of the accident was from west to east so that if there had been an escape from PSV 504 some of the detectors at the east end of the module are those that would be expected to initiate an alarm. The detectors at the said end of the module were precisely those that responded just prior to the accident. On the other hand none of the detectors in Module B signalled a gas escape in that module. I have already expressed the opinion that the evidence viewed as a whole does not suggest that all the relevant detectors in Module B could have been inoperational. Thus on the detector evidence alone Module B could be excluded as the source of the leak. However the evidence pointing to that result does not depend on the detectors alone. The probability is that the Oil and Water Operators were working in Module B until the explosion but they did not report any oil leak or escape of gas. Moreover the blue flash that was manifest when the explosion occurred is unlikely to have been caused by an explosion of gas flashing off a pool of crude oil in Module B. I accept the evidence of Dr Drysdale who on this matter was supported by Dr Mitcheson. The evidence of Professor Magnussen on the point was not very clear and somewhat tentative. There were no process upsets signalled by any of the process alarm equipment in Module B nor did the Oil and Water Operators on duty report such upsets. Moreover the B/C and C/D firewalls were definitely opened up by the explosion. Apart from there being incidental damage in the Control Room complex in Module D Mr Bollands and Mr Clark were swept across the room. Mr Bollands described the experience as being like the wall coming in. On the other hand the A/B firewall did not fail - at least not to any significant extent. The C/D firewall was the strongest of these firewalls. I do not accept that it was shown that a pressure piling mechanism could have played any part in the explosion. The involvement of such a mechanism would have required an accumulation of gas in Module B sufficient to cause an explosion there and then also sufficient to survive the dilution en route to Module C, and form an explosive cloud there. The gas alarm patterns do not bear out such a possibility. In any event there was no satisfactory evidence to show that pressure piling was even a possible factor within the open ended Module C.
There are a number of features in the evidence from which (when they are considered together) it can be concluded that the initial explosion originated in Module C. In the first place there was the blue flash spoken to by Captain Clegg which he claimed seemed to come from the mouth of Module C. Since he had firmed up his evidence on this since the Cullen Inquiry I should have been reluctant to place much store on this particular observation were it not that it fits in so well with other evidence. Whatever doubts there may be about the evidence of Captain Clegg I thought that he was an observant and fair witness so that his evidence can be accepted insofar as it is reinforced by other evidence. On the other hand it would not be surprising if a momentary observation such as he was reporting contained some inaccuracy. I have certainly no reason to doubt that he saw a blue flash and indeed it would be strange if such a detail as the colour of the flash had imposed itself in his memory unless it were well founded. I have also no reason to doubt that he had seen the flash low down in relation to the production modules. This is a detail that he seems to have been positive about both at the Cullen Inquiry and before me. The one aspect of his evidence that is transparently not satisfactory is his description as to how the flash had extended along the west face of the platform both to the north and south of Module C. Captain Clegg seems to have adjusted his evidence about the extent of the flash between the Inquiry and the Proof. In any event he described his evidence as being an "impression". It is difficult to explain a flash along the north west face of the platform in terms of an explosion at the east end of Module C. However it is equally difficult to accept that such a flash could have been caused by the explosion of a trail of gas stretching along the platform face. Firstly this in all likelihood would have been noticed by personnel who passed along the walkway just before the accident. It is not easy to see how a trail of gas would have spread to the south of Module B (assuming the gas had come from there) against the prevailing wind. It is not impossible that this could have happened but unlikely. Nor does that hypothesis fit in with what was observed at the time of the explosion. If there was an explosion along the west face of the platform this alone could not explain the explosion which caused the initial damage. It is difficult to see where a trail of gas along the platform could have come from if not the west end of Module B. If there had been such an escape in Module B then given the prevailing wind not all the gas would have emerged for the west end of the module and one would have expected gas alarms to go off. If the blue flash was caused by an explosion in Module B then one would have expected the accumulated vapour within the module to explode at the same time. As I have indicated it is unlikely that this would have been marked by a blue flash alone. The hypothesis of Professor Magnussen was founded on the supposition that the explosion and the fire seen in Module B were one continuous event. This runs directly counter to the evidence of Captain Clegg. He observed a blue flash which clearly was the initial explosion. It was followed by a shimmering silence which may have been the effects of vibration. In any event it was only a few seconds after the flash that, as a separate event a fire developed in Module B. Dr Mitcheson thought that this gap in time was a important factor in confirming that the explosion had been in C. What Captain Clegg described was literally a flash so that a misdescription of its extent would in no way be surprising. Moreover given the perspective he enjoyed in relation to the platform a flash coming directly out of Module C could quite readily have seemed to extend along the platform. I do not think it in any way destroys the hypothesis that the initial explosion took place in C which is in any event supported by other evidence.
There are a number of other factors suggesting that Module C was the location of the initial explosion. Captain Morton also seemed a reasonable witness although again it would not be surprising if his memory was not accurate in every detail. However what he first noted was what appeared to be smoke or dust coming out of the east end of Module C. There does not seem to have been a fire at the east ends of either B or C. Moreover the first thing noted by the witness Flaws was smoke rising above the east end of Module C. Thus again there is evidence that the first incident of the explosion was in Module C.
Dr Drysdale, Dr Mitcheson, and Dr Bakke all confirmed that a blue flash would be indicative of an explosion of a lean mixture and particularly of a cloud of condensate at the east end of Module C. If the visible manifestations of the explosion suggest that the initial explosion is consistent with a leak of condensate then it is condensate which would have escaped from any leak at the blind flange. The pursuers’ expert evidence was also to the effect that an explosion causing a blue flash was only consistent with a situation where only part of the module was filled with gas. This of course suggests a finite and limited source of fuel.
The evidence of Mr Grieve is important in relation to the location of the explosion. Firstly he describes in graphic terms how the explosion seemed to be just above his head. Like something dropping there he said. The location of the source of noise is admittedly often difficult to pinpoint but is perhaps another coincidence that as far as he was concerned the explosion seemed to occur above his head in the general area of the suspect PSV. Moreover some fifteen seconds or so after the explosion he noted a fireball burning in the roof space in a location not far from the location above it of PSV 504. Insofar as this was dealt with by the experts they thought it consistent with a fire in Module C which had been fuelled by heavy gas dropping through pipe penetrations. The defenders argued that this phenomenon was possibly caused by unburned gas pushed ahead of the explosive gases which had knocked Mr Young and Mr Grieve over. It was said that these might have accumulated in the roof space and then somehow been ignited. This theory was not really explored with the pursuers’ experts. Given that the explosive gases were moving at speed when they knocked Mr Young over it is not entirely clear how unburned gas pushed ahead of the flame would accumulate presumably relatively stationary in the roof space under consideration. The defenders argued that the penetration in the ceiling between the 68-foot level and Module C were probably sealed. I do not consider this to be an important point. Whatever their condition before the explosion this had caused considerable vibration that had dislodged much material such as lagging and presumably sealant. Thus I consider that the likely cause of the fireball (which occurred about 15 seconds after the explosion) was a leak of unburnt gas through the ceiling from a fire burning in the floor above. In any event it is a significant coincidence that this phenomenon should emerge from a situation quite close to the PSV site. An other important implication of the fireball is that if the gas dropped through the ceiling above this suggests that the fuel was a gas heavier than air such as a propane mixture.
Certain observed manifestations of the accident, although possible consequences of an explosion in Module B, are more consistent with the explosion having occurred in C. These include the damage to the Chantry Riser Gantry, the fact that Mr Elliot had been beaten to the ground in a position below module C at the time of the explosion, and the fact that MOL panels which had been situated in Module C had been projected onto the west face walkway. These matters would not be conclusive if each stood alone but taken together and against the background of other events they fit very well into the supposition that the explosion was in Module C.
Undoubtedly shortly after the initial explosion there was some fire in Module C but much more extensive fire in Module B. Of course there was more potential for fire in module B because it had much heavier inventories of oil and condensate. Just how these fires happened to be fuelled is difficult to determine. As I shall discuss, the fireball phenomenon in Module B seems to have been fuelled by an escape of condensate but to maintain the fire in module B there must have been and escape of oil. Dr Drysdale gave some views on how oil may have escaped from the MOL. He is an experienced chemical and fire engineer but not of course a mechanical engineer so that his views have to be regarded with much care. The fact is that there is no doubt that shortly after the explosion crude oil was escaping. Of course some of the lighter ancillary piping carrying oil may have been damaged. The persons in the diving skid saw crude oil running down the MOL as it passed through there. Of course once fire was established the heat itself could cause damage. There was some fire in Module A, on the skid deck and elsewhere on the platform. If it is evident that the explosion occurred in Module C then it is inescapable that subsequent fires were a result of this. Two unconnected accidents of the scale observed occurring at the same time is beyond credence. Upon the original explosion there would have been at least some projectiles when the firewall ruptured, considerable vibration which may have affected fuel lines and equipment and of course heat and new explosive forces.
Dr Drysdale and Dr Mitcheson convinced me that the likeliest fuel feeding the fireball phenomenon in Module B which occurred about 15 seconds after the accident is condensate. I accept their interpretation of Mr Miller’s photographs. I also accept Dr Drysdale’s calculations which point to the fact that there was likely to be just enough condensate in the condensate line to provide the fuel for the fireball (which was in any event the consequence of a limited escape of fuel). On the other hand Professor Magnussen’s attack on Dr Drysdale’s fuel calculations was not persuasive and in any event I think had been put together very much at the last minute. The defenders said that there was no explanation for the gap of about 15 seconds between the explosion and the emergence of the fireball (the point being made was that if the condensate line had been severed by a projectile there would have been no gap). The defenders therefore sought to persuade me that the initial cause of the accident was flashing gas from a pool of oil followed by a pool fire as the fire heated up. I have already made the point that the initial explosion and the subsequent fire appear to have been separate events. There is also the evidence of the witness Mr Wood. This witness observed a burst of fire from Module B with dimensions that must I think have been the fireball. However just before this happened he heard a loud metallic sound from Module B. This could have been due to equipment changing state or even condensate under pressure escaping from an enclosed space. The Nowsco tests confirmed that a leaking flange can create a noise because of gas escaping from a confined space. I think this evidence of Mr Wood confirms that there was some seconds delay between the explosion and the escape of condensate. Exactly how this came about will never be known. Perhaps the initial explosion damage combined with further local explosion or changes in the line support system or deterioration in the state of the damaged pipe from some cause was the explanation for a delayed escape of condensate.
The defenders relied upon certain evidence to suggest that perhaps the B/C firewall was not penetrated by the explosion. For example Mr Grieve approached towards the mouth of Module C after the accident and did not see any fire from Module B through the west end of the B/C firewall. However this kind of evidence is not very significant since by the times the very limited observations were being made there was so much smoke that precise observation was difficult. No one got really near to the mouth of Module C. I also found no attraction in the theory of Professor Magnussen that fire and smoke may not have penetrated through the C/D firewall to the north face of the firewall but rather could have been swept to the north from Module B over the accommodation module. This view did not get much support from the eye-witnesses who were present. The witness Swales gave a limited support to Professor Magnussen in that he saw some smoke and flame proceeding over the top of the platform. However it is obvious from other witnesses that there was heavy smoke emerging from Module D at the north face. Thus Mr Vernon before he died tried to enter the east part of Module D to gain access to firefighting equipment but was unable to do so.
For reasons which I have elaborated upon I do not think that the phenomena observed at the Flotta Terminal have any adverse reflection on the pursuers’ case on the cause of the accident. It should also be noted that there was no evidence that the B/C fire door was open during the night shift. There was no evidence that work in the area of the door was proceeding during the night shift. If the door had been left open by the day shift then one would have expected the Oil and Fire Operators to close it. Of course an escape of gas from B into C through the door would have to progress to the alarms at the east end of C to trigger the alarms there. Such gas would have had to be considerable in amount to survive dilution. Thus other gas alarms would have been triggered.
Dr Mitcheson and Mr Cubbage both thought that an accumulation of condensate at the east end of Module C was a possible cause of the accident. Dr Bakke confirmed the view that a local accumulation in the module was likelier than a gas cloud that filled more than a proportion of the module. If gas in Module C had extended materially to the west of PSV 504 then the progress of the gas through the Module would have initiated other alarms than those which responded. Moreover the explosive flame that came into the 68-foot level included hot gas. Mr Young got his eyebrows singed. However the gas that penetrated to the Control Room was cold gas and the Dr Mitcheson and Mr Cubbage considered that this indicated that the explosion had been caused by a limited amount of gas perhaps corresponding with what could have leaked from the pump. Dr Mitcheson considered that a mass of about 50 to 60 kilograms of gas would have been sufficient to cause the explosion witnesses described. This is important evidence because the defenders did not challenge it nor lead contradictory expert evidence. Indeed he thought these values represented an upper bound. If there had been more than a limited amount of condensate leaking then there would have been a more extended fire in the module. Dr Davies considered that to explain the alarm pattern the initial explosion must have been towards the eastern end of Module C. However he went further than that. His view is that an explosive cloud collected in the north east quartile of Module C would have set off different alarms to those which annunciated. Accordingly he thought that the south east quartile of Module C was the likeliest source of the explosion. This fits in with the theory that a leak from PSV caused the accident. Indeed none of the experts suggested other equipment at the south east of the module that could be considered a likely cause of the accident. Dr Davies with his modelling exercise eliminated an intrusion of gas from Module B. In any event such an escape would have triggered off alarms in B.
The evidence points unambiguously to the fact that the explosion occurred in the south east segment of Module C. This means that if PSV 504 had nothing to do with the accident there is yet another formidable coincidence. The defenders say that the closeness in time between the change over to Phase 1 Operation and the accident is also a coincidence. However the new procedure had been operating continuously for several days without apparent problem. The possible implication of the changeover was not explored with experienced witnesses such as Mr Wottge the Process Engineer. There was evidence to suggest that when the gas flow was altered when the reciprocating compressors were unloaded and recycled the flare arrangements operated and flaring increased. The experts were not asked to identify any other equipment at the south east end of Module C which night have been vulnerable after the changeover. Pump B had tripped on one other occasion after the changeover. It is unlikely that this pump was leaking since men were working next to it. On the occasion of the trip before the accident Mr Vernon suspected that this was due to a lube oil fault. If the pump was merely being overworked one might wonder why it did not restart. Certainly the centrifugal compressors tripped just after the first gas alarm. No expert evidence was led to suggest how these compressors could have caused a leak. The tripping of the compressors appears to have been a normal accompaniment of recycling and unloading the reciprocating compressors. The compressors tripped after the first alarm and if one was causing the leak it is perhaps odd that all three tripped just before the explosion. If the leak was internal to a compressor it would perhaps be curious to see all three showing alarms. The annunciation of alarms associated with all the compressors seems rather to suggest a gas source external to the compressors but common to them. The relief valves of the compressor were situated in Module B and if there had been leaks there we would have expected alarms and some indication of explosion or fire at the east end of Module B.
I agree with Dr Davies that it is quite possible that the detector G 103/1 was a candidate for the detection of a gas cloud escaped from PSV 504. Its ability to detect gas would depend on a fine balance arising from the exact position of the walkway, the position of the head of the detector, and its relationship to other structures that might have hemmed it in. If other factors pointed to it having been a detector that triggered then I see nothing to exclude that possibility. Indeed Dr Davies thought that because the escaped gas had a clear path to the detector being considered it was the likeliest candidate to give the first alarm.
Dr Davies gave evidence of his modelling exercises to test the pattern of gas releases needed to cause the alarm patterns that were observed. The defenders point out that he was asked to assume that the gas which leaked was a heavy gas and that is so. Dr Davies had no hesitation in accepting that the task he had undertaken was complex and his results had to be viewed with care. Indeed since the Cullen Inquiry he had modified his tests and results. There are difficult technical problems in such as exercise as he conducted such as scaling, calibration of test probes and making accurate allowance for the congestion within the Module and the consequent turbulence. Dr Davies gained his basic data from Dr Richardson, a fluid engineer who gave evidence about the structure of the pump, and its GOVs and PSVs as well as the relevant flows of hydrocarbon that one might expect under re-pressurising or leaking. The last two witnesses were very well qualified and able which does not of course mean that they were always able to arrive at the right conclusions in respect of the complicated problems they were dealing with. The defenders brought a specialist witness Dr Crofton about the mechanics of GOVs. His evidence was somewhat theoretical compared with that of the men who actually worked with the valves and also Dr Richardson had practical experience of valves which demanded respect. I did not find that Dr Crofton contributed anything to the case which affected the more general view that I can take from Dr Richardson and Dr Davies. I have no doubt that with a staged jagging of the pump which leaked the kind of mass of hydrocarbon which could have generated the explosion could have been generated. Moreover if the gas released were a gas heavier than air then two releases could create the alarm patterns observed. The defenders attack this approach on two grounds. The first was that the first stage of the jagging would inevitably be of neutrally buoyant gas. In an elaborate extrapolation from pieces of the pursuers’ evidence the defenders sought to persuade me that a staged escape from the PSV could not produce the alarm patterns that had been observed. They claimed that the condensate which would enter an empty pump would flash off at lighter ends. Moreover the air plug at the top of the relief line would a have to be expelled before any condensate leaked out. Their essential point was that the material that would be expelled during an assumed first stage of jagging would be neutrally buoyant gas. Such a gas cloud could not travel to the end of the module without rising and triggering a C2 zone alarm and in particular G101/1. This they say follows inevitably from the conclusions of Dr Davies. In the first place it has to be observed that the defenders’ theory was not put specifically to either Dr Richardson nor Dr Davies. Although both these experts had been asked to assume that the escape of gas would be heavier than air this assumption did not appear to trouble them. The behaviour of condensate when it enters the pump and thereafter leaks will be a delicate balance between pressure and temperature. If heavier ends flash off there will be mixing and stratification of the different gases. In any event the prospect of the gas cloud missing alarms will depend on the relative heights of the detector and the PSV leak. These heights were not absolutely clear. Obviously the pursuers anticipated the escape of heavier gases since they replaced some of their detectors at lower levels to account for this. Thus not only are the defenders constructing scenarios that were not put specifically to the pursers’ experts but the reality is that the observed phenomena suggest that the explosion was of a low lying material. I refer in particular to the fact that the blue flash was said to be low in relation to the modules and also the fireball observed by Mr Grieve. The material which escaped from the PSV indeed seems to have avoided setting off the C2 alarm and that could confirm that the defenders’ explanation of events is not inevitable and indeed I do not think that it is. Even assuming that the defenders’ postulation of what happens when the condensate enters the pump is correct it does not follow that the pursuers’ case will fail. As I have mentioned it cannot be assumed that the pump was empty when the first jagging which gave rise to a detectable leak took place. If the pump was jagged before the first gas alarm we do not know precisely how this was done. The first stage of jagging may have consisted of Mr Vernon or Mr Richard allowing condensate into the pump in say two applications of the push/pull button. This could have caused an expulsion of lighter gases which did not contain enough hydrocarbon to trigger an alarm. Then as the pump filled the condensate or heavier gas would begin to leak out and when it reached the zone three alarm it would trigger an alarm. The second stage of jagging would then consist of what must have occurred when Mr Grieve arrived at the scene. In this operation the latch may have been opened rather longer to be sure of fully pressurising the pump but that would still be an operation that would take less than a minute. The second operation perhaps releasing a greater weight of material would set off the second pattern of alarms. The defenders make another important point since they claim that Dr Davies accepted that his modelling results would be affected if there was a scaffolding below the site of the blind flange. This is not quite an accurate representation of Dr Davies’ posit. What he was saying in effect was that he could not accurately model the situation taking account of the scaffolding. He did attempt to model the scaffolding by placing a block beneath the presumed leak site but this he accepts was not very satisfactory. In effect he found himself unable to model the scaffolding without details of its structure, its location in relation to the blind flange, the equipment surrounding the scaffold, and the size and direction of the leak. It became too difficult to perform precise modelling taking into account the conjectural factors introduced by the scaffolding. However Dr Davies did not say that the presence of scaffolding made it unlikely that the alarms were triggered by an escape from PSV 504. In fact even in the knowledge that the material which had been introduced to the pump was said to be condensate and with the additional knowledge about the scaffolding he appeared to accept that leaks from the PSV could account for the alarm sequence. The scaffolding of course admittedly could have affected the accuracy of his modelling. There was however room for a combination of events that could have caused the alarms to go off and that was consistent with a phased attempt to re-pressurise the pump. Thus the main lesson that can be taken from the evidence of Dr Davies is that the pursuers’ hypothesis, if supported by other evidence which fits well together, is eminently possible.
The parties spent a considerable amount of the time occupied by this proof seeking to establish the structural strengths of the firewalls and the explosive overpressures that would be need to cause their collapse. I was never entirely sure what the pursuers’ intention was in undertaking this formidable exercise. One objective certainly was to seek to demonstrate that an explosion sufficient to destroy all or part of the firewall (and in particular the south-west end of the B/C firewall as viewed from C) would have generated projectiles with sufficient energy to breach the 4-inch condensate line in Module B. Another problem may have been that these are extremely extensive proofs and thus parties may not always have been sure what precisely their opponents were planning to do - particularly in reply. Thus the pursuers may have been seeking to protect themselves against evidence from the defenders to the effect that the pursuers’ hypothesis is impossible.
The experts were agreed that whatever the value of hand calculations in the exercise (and this itself was in dispute) a finite element computer modelling or at any rate an exercise assisted by computer modelling was likely to arrive at a more accurate result. The pursuers certainly recognised that because as I have earlier indicated their original intention was to lead evidence of such modelling. They had to abandon this plan but I think it is clear that the failure of the pursuers to approach these questions by way of a computer exercise was not due to any deliberate planning but rather due to faulty organisation. It also had to be noted that the introduction of a computer model does not remove every difficulty. It simply creates different difficulties. The experts agreed that however the matter is approached the dynamic analysis of structure raises many profound engineering problems. If computer modelling is used it is common to find questions raised about the appropriateness of the model or the adequacy of the input material and there was even a foretaste of this in these proofs. It cannot therefore be said with absolute confidence that a computer analysis of the firewalls in this case would have saved a massive amount of time. However I should definitely have expected such an exercise to save some time and probably even a lot of time. There was also a fair prospect that it would have produced more reliable results.
In these cases the pursuers led evidence based on manual calculation from Professor Fenner. He was a computational analyst and recognised his incapacity to do a complete dynamic analysis of the firewall by manual methods. He did do a static analysis of the firewall using beam theory and also he worked out comparative strengths of firewall components. I think it was largely in connection with the last-mentioned material that the pursuers were relying on him. They may also have hoped to use Professor Fenner to provide the foundation material for the exercise by Dr Cox. On the other hand the witness Dr Palmer did a complete dynamic analysis of the B/C firewall by hand calculation. The defenders attempted to diminish the value of such an exercise. However as Dr Palmer sought to remind me on a number of occasions, until recently engineers used to manage all their work by way of hand calculation. In any event the result of the pursuers’ approach is that I was assailed for many months with complex mathematics (at least so I judged it to be). Because much of this was difficult the witnesses producing the evidence required to spend much of the time explaining what they were doing and even with their expert help there were areas of the evidence which were close to being unintelligible to a non-scientist. A look at some of the oral evidence, the productions and the calculations on the flip charts will make this point perfectly obvious. When it came to submissions much of the detailed mathematics was not challenged. Whether this was out of deference to my faint antipathy to the material or because it was not considered helpful to the parties’ submissions I do not know. However I am left with wondering why much of the mathematics could not have been agreed in the first place since little of the pure mathematical calculation was challenged. This would have saved a lot of time. There may of course have been practical problems about such agreement.
The approach of the defenders to this area of the case has to be noticed. The defenders brought Professor Reid to challenge Dr Palmer’s methodology. He was a very capable witness and I am sure that, in particular, he was quite capable of producing his own quantitative analysis to illustrate the impact of the points he was making. However he had not been asked to do this and indeed only occasionally attempted it. Thus the defenders’ approach was to attack Dr Palmer’s methodology but they made no attempt to themselves establish the answers to the problems being considered.
Dr Palmer, Professor Fenner and Professor Reid were all in my opinion highly competent and exceptionally well qualified experts. That is not to say that their evidence often agreed. It is acknowledged that the theoretical structural problems in the kind of exercise I have to consider are very difficult and experts disagree over many aspects of the methodology. I can form certain broad opinions about the general approach of the witnesses. However given witnesses of such quality unless I can see a clear way through a particular problem I do not think it would be appropriate that I should attempt to resolve theoretical and also practical differences that have divided the engineering profession for years. I have set out above some of the detailed issues that arose in the proof but do not think it necessary or advisable to resolve the problems that arise other than in a broad manner and to the extent that I have already dealt with them.
Certain differences in the witnesses are significant. Professor Fenner had a narrow remit. He was mainly used for his comparative analysis of components and his results in these respects were not seriously challenged. Dr Palmer struck me as being very experienced in the resolution of practical problems so that his approach was to look for broad and pragmatic solutions. Professor Reid on the other hand was looking for solutions that were more scientifically rigorous so that he was inclined to be concerned about issues that in practical terms Dr Palmer thought could be looked at more broadly.
I should perhaps also add for good measure that I had no occasion to doubt the qualifications of Dr Bakke and Professor Stollery.
The important point is that if I conclude that on the basis of the other evidence I have considered it is clear that the re-pressurisation of the Condensate Injection Pump resulted in a leak at the blind flange and that this caused the explosion it is not critical whether the strength of the firewall or its capacity for developing damaging projectiles is established or not. If an accumulation of condensate at the east end of Module C exploded then I think that it is inescapable that all the damage to the platform that ensued after the explosion was triggered-off by it. Thus if there is evidence that the B/C firewall failed (and I am satisfied about that fact) then it may not matter precisely how strong the firewall is. If the firewall fails and this is followed by a conflagration it may not matter precisely how this has developed. Of course I would have to be persuaded that an explosion of the quantity of hydrocarbon that could have leaked from the blind flange would on the face of it be sufficient to damage the firewalls and set off fires in Module B. If the defenders were to have shown that an escape from the pump could not possibly have destroyed the integrity of the firewall then of course I should have to reconsider the other evidence pointing to the blind flange as the source of the tragedy but the defenders certainly have not done that.
It was pointed out that there is a difference between the results of Professor Fenner’s static analysis of the firewall and that of Dr Palmer. The argument was that the pursuers’ experts contradict one another so both should be disregarded. Professor Fenner brings out a failure value for the firewall of 0.007 bar as compared to the value of 0.1 bar produced by Dr Palmer. However we are not comparing like with like. Professor Fenner’s result was arrived at by way of a beam analysis. However Dr Palmer’s value was achieved as a result of a plate analysis. This clearly is a more complete methodology since it takes into account the effect of lateral support. Moreover Dr Palmer’s result is supported by Dr Bakke who expected that the failure pressure of the firewall may have been even higher than Dr Palmer found. However it has to be noted that Dr Bakke was only expressing a general impression and he never seriously challenged Dr Palmer’s results. Moreover in respect of the collapse of fairly complex and substantial structures such as the firewall then being himself a specialist in gas dynamics he was prepared to defer to the experience of a structural engineer like Dr Palmer. Dr Bakke also accepts that the calculation of an overpressure that might develop in a structure like Module C from the explosion of a specific quantity of hydrocarbon is a difficult and sensitive calculation. One very important factor is turbulence and this is very hard to assess accurately. The volumetric size of the cloud can affect the overpressure. The very source of ignition can affect the overpressure by a factor of five. Thus I do not think that Dr Bakke ever seriously challenged the general evidence of a witness such as Dr Mitcheson that the explosion of the 40 or 50 kilograms of gas that could emerge from a leaking injection pump could cause an explosion that would pierce the firewall. Indeed he supported their point of view and indicated that a cloud which occupied about one-tenth of the Module could generate an overpressure of 0.5 to 0.6 bar. He thought that the overpressures which the pursuers were basing their case on were rather conservative.
Dr Palmer was a highly experienced engineer. He was prepared to venture a conclusion for a dynamic analysis of the strength of the firewall. His results even by his own admission must be taken as very approximate. Professor Reid has expressed the view that Dr Palmer’s results are seriously unreliable but he has not himself offered quantification of this view. Thus I am left with only one complete analysis of the overpressure that would be required to destroy the integrity of the firewall. It is obviously not the only result that an experienced engineer might arrive at. However it is the only one I have and it comes from an expert well able to advance justification for his methods. It also has to be restated that many of the objections that were eventually suggested in relation to his methods were not put to him. Thus in deciding what reliance if any I can place in his work I must recognise that the approximate validity of his conclusions fit in well with the other evidence in the case. On balance I do not need to take any more from his evidence than that the destruction of the B/C firewall by an explosion of such gas as may have leaked from the injection pump is probably quite possible. Even if his calculations are not accurate there is general evidence which I accept that escaped material from the pump would generate sufficient pressure to destroy the integrity of the firewall.
The matter of the projectiles is more difficult. Again Dr Palmer calculates the projectile force that might be expected from such part of the firewall as is forced loose by the explosion. Professor Stollery gave evidence about shock tube theory but for the reasons which I have already discussed I do not think his reservations on Dr Palmer’s results are conclusive. Professor Reid is a particular specialist on the impact of projectiles and I accept that his views demand respect. However some of his opinions involve fairly difficult and controversial areas of his subject. Once again I am handicapped by the fact that much of the pursuers’ evidence was not put to Dr Palmer. Thus although on this branch of the case I can only have limited confidence in Dr Palmer’s views and could not rely upon them if they stood alone equally he is too experienced an engineer to have his views totally discounted. I have now arrived at the point where I can decide that the accident occurred because of an explosion in Module C. This explosion resulted from the fact that Mr Vernon jagged Condensate Injection Pump A at a time when the essential PSV was not in position. The explosion caused a substantial part of firewall B/C to break up. This was the cause of fire developing in Module B. I cannot make a finding as to precisely what energy fragments of the wall would have developed as they broke away. I am not sure that this matters. Just how the fragments impacted on the equipment in Module B will never be known. However I think it is clear that the impact of the fragments from the wall had sufficient force to do some damage. Dr Palmer with his experience must be right at least to that extent. I think it probable that the condensate line in Module B sustained some damage from the explosion. The same applies to the smaller bore oil lines. Just how damage to the condensate line operated to cause the escape of condensate cannot be known but with such an explosion the connection not only seems possible but it is shown to be possible by what actually happened.
The defenders contended that Mr Grieve was the only eyewitness present at the 68-foot level when Mr Vernon would have required to jag the pump if the second flurry of alarms is to be explained. That fact is true and the defenders therefore argued from it that since Mr Grieve did not see Mr Vernon jag the pump it could not have happened. I cannot accept that conclusion. Mr Grieve was himself occupied in helping to try and restart Pump B. He was not really aware what Mr Vernon himself was doing and said so. The jagging operation would have taken less than a minute and would have been carried out relatively close to where Mr Vernon is described as having been seen. It would have been quite surprising if Mr Grieve had noticed and remembered all that Mr Vernon did in a short period when Mr Grieve himself had detailed work to perform. Mr Vernon had indicated to others that he was intending to start Pump A. He was awaiting an electrician so that he could complete that objective. He could not do much before he had his electrician but it would have been natural that he would try and get the pump ready for starting so far as was possible before the electrician came. It was also to be expected that if he had changed his mind about using Pump A he would have instructed the cancellation of the electrician. There would be no point in de-isolating the pump and then requiring to electrically isolate it again. The preparatory work would include re-pressurising the pump. The jagging operation would only take a few minutes in total so that I see nothing peculiar in that at the same time as he dealt with pump A Mr Vernon should continue efforts to restart Pump B particularly when an instrument technician is coming on the scene.
If condensate came out of the Pump system during jagging then the only probable explanation is that Mr Sutton was careless. The evidence shows that blind flanges should not leak if they are properly fitted. Therefore I must conclude that Mr Sutton did not fit the blind flange properly. I do not know why he did not take care but there are various possibilities such as that he found the flange troublesome to fit in a confined space or that he expected the valve to be refitted before the end of his shift and therefore did not bother to do his job correctly. I have already found that the defenders’ arguments on foreseeability and novus actus interveniens are not well founded.
Insofar as Mr Vernon did not know or recollect that the PSV 504 was not in place I find that he was careless. However I cannot find that he was reckless or acting so as to show wilful misconduct.