FIRST DIVISION, INNER HOUSE, COURT OF SESSION Lord President Lord Sutherland Lord Coulsfield Lord Gill 0/1261/5/1990 OPINION OF LORD GILL in RECLAIMING MOTIONS in the causes CALEDONIA NORTH SEA LIMITED Pursuers and Reclaimers; against 1. 2. 3. 4. 5. 6. 7. LONDON BRIDGE ENGINEERING LIMITED; PICKUP NO. 7 LIMITED (formerly NORTHERN INDUSTRIAL & MARINE SERVICES COMPANY LIMITED); BRITISH TELECOMMUNICATIONS plc; WOOD GROUP ENGINEERING CONTRACTORS LIMITED; NORTON (NO. 2) LIMITED (In liquidation)(formerly EASTMAN CHRISTENSEN LIMITED) and DAVID JOHN PALLEN, Chartered Accountant, the Liquidator thereof; KELVIN INTERNATIONAL SERVICES LIMITED (formerly KELVIN CATERING LIMITED); and COFLEXIP STENA OFFSHORE LIMITED (formerly STENA OFFSHORE LIMITED) Defenders and Respondents: _______ Act: MacAulay, Q.C., Batchelor, Q.C., Hofford; Paull & Williamsons Alt: Currie, Q.C., Keen, Q.C., Wolffe; Simpson & Marwick, W.S. 17 December 1999 2 I The history of the actions The pursuers and reclaimers were formerly known as Occidental Petroleum (Caledonia) Limited (OPCAL). They were the operators of Piper Alpha, the North Sea oil and gas production platform which was destroyed by an explosion on 6 July 1988. A consortium of companies owned the platform. They were known collectively as the participants. The reclaimers were one of the participants. They operated the platform on behalf of the consortium. In accordance with normal practice in the North Sea oil industry, most of the services on board the platform were carried out by contractors. The defenders and respondents were contractors on the platform at the time of the disaster. These reclaiming motions relate to seven of the actions raised by the reclaimers against contractors for recovery of sums paid by the reclaimers in settlement of claims for damages for death or for personal injury arising from the disaster. In each case the reclaimers found on an indemnity clause in their contract with the respondents. Six of the actions relate to death claims. The seventh, the Coflexip Stena action, is a claim by a survivor. The actions were raised in September 1990. The records were closed on 29 January 1992. The conjoined proof before answer began on 3 March 1993. It ended, on its 391st day, on 31 October 1996. The reclaimers’ insurances were arranged in layers. In the death cases the reclaimers were fully covered against liability for claims. In the personal injury cases they had an uninsured layer. In these cases the shortfall in the reclaimers’ cover is known as the Oxy Gap. It amounts to about $19 million. 3 The nature and terms of the reclaimers’ insurance arrangements were not in issue on Record in this case and were not the subject of detailed evidence. The policies of insurance were not produced. Mr Michael Payton gave evidence for the reclaimers about the Oxy Gap. There was a clear issue about that in the pleadings. Mr. Payton described the reclaimers’ insurance arrangements in general terms. His evidence gave the respondents an opportunity to take the point on which the case was decided in their favour. Soon after the disaster the survivors and the relatives of the deceased intimated claims against the reclaimers and organised themselves in preparation for legal proceedings. In all about 800 claims were pursued. The solicitors involved formed groups to pursue the claims. The principal group was the Piper Alpha Disaster Group (PADG). It was formed at a meeting held on 21 July 1988. As it happens, all of the claims to which the test cases relate were settled with the PADG. It was apparent to the reclaimers that the claimants had access to advice from lawyers in the United States who specialised in multi-party litigations and disaster claims. These lawyers were experienced in the financing of such claims and acted on a contingency fee basis. It was part of their expertise to choose the most advantageous forum. By the autumn of 1988 it seemed almost certain that the claimants’ representatives would soon raise proceedings in the United States of America. In that event the likeliest forum would be in the courts of Texas. There was considerable urgency on the part of the reclaimers and their insurers to settle the claims before actions could be raised in Texas. In that jurisdiction such actions were tried by juries. It was notorious that juries in Texas awarded damages at much higher levels than those at which damages would be awarded in Scotland. There was a risk that in individual cases juries in Texas might make “runaway” 4 awards that would be extravagant even by Texan standards, or might indirectly award punitive damages. The courts in Texas took a wide view of their jurisdiction. They rested jurisdiction on points of connection between the defendant and the State of Texas. Jurisdiction could be established against a defendant who was not domiciled there. If the defendant contested jurisdiction and if the court found against him, the defendant could appeal on that point only after the jury trial was concluded. OPCAL had significant points of connection with the State of Texas. OPCAL had placed numerous contracts for the purchase of equipment with suppliers in Texas. Some of these were sole suppliers. The most critical connection was the fact that OPCAL’s entire production from the North Sea was sold in Texas through one of its associated companies. The contractors were unaware of this arrangement. Although at that stage the reclaimers could not assess the extent of the risk as accurately as they attempted to do at the proof, they proceeded on the basis that the risk of their being convened to the courts of Texas was substantial. On 27 July 1988 there was the first of a series of meetings between the consortium, their insurers and the claimants’ representatives. OPCAL’s priority was to achieve an early extra-judicial settlement of all the claims. In an attempt to head off proceedings in Texas OPCAL proposed that the claims be settled at “midAtlantic” levels. The mid-Atlantic level for each claim was to be assessed by taking the probable value of the claim in the Scottish courts and applying to it a standard multiplier. OPCAL’s proposal to settle at mid-Atlantic levels was conditional on there being unanimous acceptance of OPCAL’s offers. OPCAL first discussed these proposals with the contractors at a meeting held in London on 2 September 1988. It was made clear to the contractors that to achieve a 5 settlement of all the claims offers would have to be made at levels higher than Scottish levels. The contractors and their insurers declined to admit liability. At about this time the Piper Alpha Insurers Committee was formed. Although the contractors reserved their position on the multiplier to be adopted, they did not suggest that enhanced levels of damages should not be offered. Later the contractors’ solicitors wrote to OPCAL’s solicitors in terms that implied that a settlement at levels higher than Scottish levels would be reasonable. In December 1988 the contractors were represented at one of the settlement meetings. The contractors’ representatives were warned of the risk that actions would be raised in Texas. OPCAL and their underwriters thereafter agreed to make direct offers to the claimants in order to forestall litigations. By the end of 1988 the settlement formula had in general been accepted by the claimants in the death claims. In due course all of the claims were settled. After the agreement of general settlement terms, OPCAL’s solicitors gave advance notice to the contractors’ solicitors of the terms of individual settlements. The contractors were invited at that stage to put OPCAL in funds. The contractors’ solicitors did not reply to these letters. In discussions with the solicitors for the contractors the solicitor for OPCAL founded on the contractual indemnities. The claims were settled by the reclaimers with funds provided by their insurers. The reclaimers’ insurers thereafter looked to the respondents to pay the cost of the settlements, on the view that they were liable to the reclaimers under the indemnities. The respondents denied liability. During 1989 Lord Cullen held a public inquiry into the causes of the disaster. The parties to these litigations were represented at the inquiry. We have not been referred to the Report of the inquiry or to Lord Cullen’s conclusions; but the parties 6 agree that in closing submissions at the inquiry, counsel for the present respondents submitted inter alia that: “The initial explosion was probably caused by the ignition of condensate escaping from the pipework from which PSV 504 had been removed and to which a blind flange had been fitted by Score (UK) Limited … ” This was the cause of the disaster which the reclaimers sought to establish at the proof and which the respondents did not accept. These actions were raised and pursued in the name of the reclaimers but on the instructions of their insurers. The insurers considered that they were entitled to be subrogated to the rights of the reclaimers against the respondents under the indemnities. The conclusions were for payment to the reclaimers of the sums paid by them to the claimants in the settlements. The six cases relating to death claims were raised solely for the benefit of the insurers. The Coflexip Stena action was raised partly for the benefit of the insurers and partly for the benefit of the reclaimers to the extent that the reclaimers had paid the claim out of the Oxy Gap. In briefest outline, the reclaimers’ case was that the explosion occurred in module C of the platform. It was caused by the ignition of condensate that had escaped under pressure at the site of a blind flange that had been fitted when a pressure safety valve (PSV 504) was removed from the process pipework for maintenance; and that the escape occurred because a) Mr Terence Sutton, an employee of Score (UK) Limited, did not fit the flange properly and b) Mr Robert Vernon, OPCAL’s lead production operator, jagged the condensate injection pump when the pressure safety valve was missing. In each action the reclaimers pled that they had been made liable to the claimants by reason of their being in breach of regulation 32 of the Offshore Installations (Operational Safety, Health and Welfare) Regulations 1976 (SI No 1019). These breaches had been committed by employees of the 7 reclaimers; by Terence Sutton and Alexander Rankin, who were employees of Score (UK) Limited, and in particular by the reclaimers’ employee, Robert Vernon. The case against Terence Sutton was that he had failed to secure the flange properly. The case against Alexander Rankin was that he had failed to supervise Sutton, that he had failed to instruct Sutton to fit the flange securely and to check it, and that he had failed to check Sutton’s work after Sutton had completed it. The case against Vernon was that he had failed to check the valve site before jagging the pump. The respondents averred that the explosion occurred in module B. Their defence to the case pled by the reclaimers was that the accident was caused by the sole fault and wilful misconduct of the reclaimers themselves. The respondents denied that Vernon knew that the pressure safety valve had been taken out for maintenance. They averred that the reclaimers themselves had been guilty of wilful misconduct in respect of their breaches of the permit to work system and that, if the reclaimers had operated the permit to work system efficiently, Vernon would have known that the pressure safety valve had been taken out. Alternatively they pled that, esto Vernon did know that the valve was out, the accident was caused by his sole fault and wilful misconduct. The respondents denied that there was any fault on the part of Score (UK) Limited or its employees. At the proof the reclaimers did not attempt to prove their case of fault against Alexander Rankin. They led him as a witness. Counsel for the reclaimers did not suggest to him that he had failed in any duty. The respondents on the other hand did not lead evidence to attempt to prove that the explosion occurred in module B or that there was wilful misconduct on the part of the reclaimers themselves. Instead they sought to prove that there had been wilful misconduct on the part of Vernon. Their esto case therefore became their central line of defence. 8 Proof of the crucial facts as to the cause of the explosion depended entirely on circumstantial evidence. Module C was not recovered. Most of the key personnel, including Sutton and Vernon, perished in the disaster. Those who survived it were unable, by reason of the effects of it, either to give detailed evidence or, as in one case, to give any evidence at all. The Lord Ordinary found for the reclaimers on all material questions relating to the causes of the accident and on the existence of a contractual obligation on the part of the respondents to indemnify the reclaimers. Despite that, in each of the actions other than the Coflexip Stena action he found for the respondents and granted decree of dismissal. He did so on the basis of the first of two substantial points of law taken by the respondents at the stage of closing submissions. The first related to the nature of the claim maintainable against the respondents. The second related to the extent of the respondents’ liability to the reclaimers. In the Coflexip Stena action the Lord Ordinary found for the reclaimers on liability; but on the basis of the respondents’ second point he granted decree for only part of the sum paid out by the reclaimers. In that action the Lord Ordinary had to decide a further question as to the tax consequences of the decree in the reclaimers’ and a related question of interest. Neither of the two defences on which the respondents succeeded before the Lord Ordinary was expressly raised in the respondents’ pleadings. The first of the surprise defences related to contribution and subrogation. It was not specifically raised by counsel for the respondents during the proof. On 12 March 1993, the seventh day of the proof, Miss Lesley Gray, the solicitor for the reclaimers, gave evidence relating to the settlement. In the course of her evidence she stated that the reclaimers were insured for the cost of the settlements. On 4 and 5 May 9 1993, the nineteenth and twentieth days of the proof, Mr. Michael Payton gave evidence about the reclaimers’ insurance arrangements. Mr Payton was a partner in a firm of London solicitors who acted for the reclaimers’ insurers in the settlement negotiations (19/2984-9; 20/3082-4). The purpose of his evidence was to explain to the court the circumstances in which the Oxy Gap had arisen. Mr Payton also spoke to the settlement negotiations in general and to certain specific matters regarding tax relief that the respondents had raised in their pleadings. During cross-examination Mr Payton said that the actions had been raised for the benefit of the reclaimers’ insurers who were exercising their rights of subrogation (20/3189-90). It was not suggested to him on the respondents’ behalf that the insurers were not entitled to exercise such rights. That point was raised for the first time on the final day of the respondents’main submissions. Counsel for the respondents then submitted that the actions other than the Coflexip Stena action were irrelevant because the reclaimers were no longer in a position of loss; that the effect of the insurance payments was to extinguish the liability of the contractors under the indemnity clauses; that the insurers were not entitled to be subrogated to the reclaimers’ rights against the contractors; and that the insurers would at best for them have been entitled to claim contribution from the contractors, since they and the contractors had incurred the like obligation of indemnity. Counsel for the respondents submitted that the Coflexip Stena action was relevant only to the extent that the pursuers were suing to recover that part of the cost of the claim that they had met. The Lord Ordinary found for the respondents on this submission. He held that in the Coflexip Stena action the reclaimers were entitled to recover only part of the 10 sum sued for. In the other actions he held that the reclaimers no longer had “any title or interest to sue” (6/1437) and assoilzied the respondents. The second surprise defence also emerged in the closing submissions. Counsel for the respondents argued that the indemnities did not extend to the cost of the settlements so far as those settlements were made at levels in excess of Scottish values. The excess element was not in the contemplation of the respondents when they entered into the contracts. The respondents were unaware of the reclaimers’ special vulnerability to the jurisdiction of the courts of Texas. Moreover the disaster was a unique event which had resulted in the concerted prosecution of claims by skilled disaster lawyers. This had increased the levels at which the reclaimers had had to settle. The Lord Ordinary found for the respondents on this point too. He held that if the actions other than the Coflexip Stena action had been well founded, the reclaimers would have been entitled to recover only the amounts that would have been awarded as damages by the Scottish courts. In the Coflexip Stena action he allowed partial recovery restricted to the Scottish value of the claim. The reclaimers have reclaimed against the Lord Ordinary’s interlocutor on the two points on which the Lord Ordinary found against them. On the tax question raised in the Coflexip Stena action, the Lord Ordinary’s decision has been to the reclaimers’ advantage in the events that have occurred; but the reclaimers have argued that his decision is wrong in principle. They have reason to take the point because they fear that the principle favoured by the Lord Ordinary could operate to their disadvantage overall in the other cases relating to the Oxy Gap payments. The respondents have taken the opportunity to cross-appeal on 49 grounds, not all of which have been insisted in. Some of them overlap. In the Coflexip Stena case 11 the respondents have cross-appealed on a point concerning the adjustment for interest. This point arises from the tax question raised by the reclaimers. It relates to the treatment of interest on the adjustments to be made to the sum due in consequence of the decision on the tax question. In my view, the starting point in these appeals is the construction of the indemnity clauses on which the actions are founded. If the obligation to indemnify is held to exist in the circumstances of these cases, the next question is how far the obligation extends. This is the consequential loss point on which the Lord Ordinary found for the respondents. From that question there follows the question whether the loss should be assessed by reference to Scottish levels of damages or to some other measure. The Lord Ordinary found for the respondents on this point. The next question is as to the significance, if any, of the fact that the sums at which settlements were concluded were paid out under decrees of the Court of Session. The Lord Ordinary found for the reclaimers on this point. It is convenient thereafter to deal with the contribution/subrogation point, the question of the impact of tax, and the question of interest. There are then a series of questions raised by the respondents as to the cause of the disaster. These relate to the actings of Terence Sutton, the actings and state of knowledge of Robert Vernon, and associated scientific questions. These are questions as to the sufficiency of the evidence; but there is also a preliminary question whether it is appropriate or expedient for this court to interfere with the Lord Ordinary’s findings on any of the factual issues. The final question is one of mixed fact and law as to whether the actings of Robert Vernon constituted a novus actus interveniens. 12 II The scope of the indemnity clauses I agree with the analysis and the interpretation of the indemnity clauses that are set out in the Opinions of your Lordships. If the respondents’ submission had been well-founded, it would have been a complete answer to these actions. Counsel for the respondents insisted that in each case the point involved a pure question of construction. I think that the point might usefully have been tested at a debate. The interpretation of these indemnities was impeded to some extent by the seemingly haphazard variations in the wording of the clauses. In each indemnity there is what may be described as a core provision. It is set out in clause 17.1.c of the London Bridge contract. There are minor variations of it in the other contracts, but in my view its meaning is common to all of the cases. In analysing these clauses counsel for the respondents used the term “preamble” to refer to those words, in either the long or the short form clauses, that precede the core provision. In my view, it is not helpful to describe those words as a “preamble”. Whether the clause is in long or short form, they contain an operative provision of the contract. For the reasons given by your Lordships I consider that the Lord Ordinary was right in holding that in each case the indemnity has the effect for which the reclaimers contend. Counsel for the respondents pointed to what they described as an odd result of the indemnity if the reclaimers’ interpretation were correct; namely that the application of the indemnity would turn on the fortuitous circumstance that the death or injury was also caused by the negligence of a third party, to however small an extent, and in circumstances where the third party was not connected with the 13 contractor and was someone over whose activities the operators themselves retained some measure of control under the permit to work system. Counsel for the respondents also pointed to what they described as “rough justice” where the contractor was liable even though the fault was that of a third party. For these reasons counsel for the respondents argued that the reclaimers’ proposed construction was “inherently improbable”. There is a danger in the court’s reaching a view as to the inherent improbability of a contractor’s accepting a particular obligation when the court is not in a position fully to assess the commercial considerations that would influence the parties to the contract. It may be that the commercial realities of the case give a party an incentive to accept an obligation that might seem inherently improbable to an outsider unfamiliar with the industry. There was evidence of practical considerations that provided a rational explanation for the comprehensive indemnities imposed in these cases. This evidence showed that it was standard practice in the industry for the contractor to grant an indemnity for losses sustained by its own workforce whether or not the contractor was in any way responsible for the loss (10/1732-1733). Mr Crain, the Senior VicePresident of Union Texas Petroleum, one of the consortium, gave evidence that the “fundamental economics” of the business required each employer on the platform to take the risk for its own employees, no matter what happened, and that in practice the insurance cost associated with that was built into the contractual rates. There was evidence of other circumstances relevant to the operation of production platforms. OPCAL had only 38 personnel on board. Under regulation 32 (supra) they had more onerous responsibilities than any of the contractors (MacMillan v Wimpey Offshore Engineers & Contractors, (1991 SLT 515). OPCAL’S liabilities 14 were almost absolute. They had a direct responsibility in respect of any acts of the contractors or of their employees and had responsibilities in relation to safety matters such as emergency drills, fire precautions, and escape procedures. The other participants were under the same obligations as OPCAL, because they were concession holders, whether or not they had any part in the operation of the platform. In such circumstances it was not unreasonable, in my view, that the parties involved in the operation of the platform should reach a pragmatic solution regulating their respective liabilities for losses arising in that environment. Such an agreement could minimise the possibility of protracted litigations between operators, contractors, subcontractors and individuals. This would avoid what was described in one offshore case in the United States as a “multi-party donnybrook” (Fontenot v Mesa Petroleum Co 791 F (2d) 1207 (1986), at 1209, 1216). In my view, the contractual provisions in this case were designed to avoid such a result (cf. Daintith and Willoughby, United Kingdom Oil and Gas Law, 2nd ed, paras 1-845, 1-846; Sharp, Offshore Oil and Gas Insurance, pp 48; 104-110; 132-5; 171-2; 275-7). III Consequential loss These are actions for payment of the losses sustained by the reclaimers in meeting third party claims. They are presented on the basis that the respondents have failed to fulfil their obligations to indemnify. It seems clear that the indemnity in clause 17 covers the cost of a compromise of a claim made against the reclaimers. The reclaimers are indemnified against “any claim” arising from personal injury or death. Such an indemnity is an indemnity against loss sustained in consequence of a claim that has at least some prospect of success (Comyn Ching & Co v Oriental Tube Co Ltd, (1981) 17 BLR 47). The heart of the respondents’ argument is that these 15 losses were, to the extent that the settlements exceeded the Scottish levels of damages, consequential losses from which a specific clause of the contract exempted the respondents. Clause 20 of the London Bridge contract quoted by your Lordship in the chair is a suitable specimen of the relevant provision. Although the other contracts are not in identical terms, it is agreed that for the purposes of the present point there is no material difference between them, except in the Coflexip Stena case. In the Coflexip Stena case the provisions are entirely different. The argument sustained by the Lord Ordinary in the other cases cannot apply to that case. In each case the Scottish value of the claim was agreed; but the respondents put the reclaimers to the proof of the probable amount of a Texan jury award. The reclaimers led evidence on this question from two expert witnesses, Mr Wayne Fisher, a Texas trial lawyer, and Dr John Allen, an economist, whose evidence provided a range of values within which each claim would probably be assessed in a Texan court. The Lord Ordinary accepted that the settlement figures, all of which fell between the Scottish values and those in the Fisher/Allen range, were reasonable. In the London Bridge case, for example, the Scottish value of the claim was agreed to be £231,750. The value in the Fisher/Allen range was £1,690,023 - £1,895,182. The claim was settled at £683,853.71. The respondents’ argued that clause 19 of the contract distinguished between “indemnity” and “damages”. Clause 20 related to indemnity only. The words “Notwithstanding any provision herein to the contrary, in no event ...” signified that the indemnity given by the contractor was limited by reference to “consequential loss”. Consequential loss was a term which by judicial decision had acquired a recognised meaning. It was to be understood by reference to the second rule of 16 damages for breach of contract set out in Hadley v Baxendale ((1854) 9 Ex 341), as explained in Koufos v C Czarnikow Ltd ([1969] 1 AC 350). Accordingly, on a proper construction of the clause, the reclaimers could not recover any indirect or consequential loss arising out of the claims. The losses sustained by the reclaimers in paying the claims were not recoverable to the extent that the settlements exceeded Scottish values since they arose from special circumstances, namely (a) the reclaimers’ vulnerability to jurisdiction in Texas, of which the respondents had been unaware when they entered into the contracts, and (b) the intervention of experienced and well-financed disaster lawyers who pursued the claims professionally and vigorously. Their approach to litigation placed importance on preparatory procedures such as pre-trial depositions and discovery, and enabled the claimant to apply greater pressure on the defendant to settle than would be the case where a single claim was pursued on a more limited budget, perhaps by a less experienced lawyer. As a result, part of the loss sustained by the reclaimers was referable to the pressure to settle imposed by the disaster element. For these reasons, the mid-Atlantic uplifts in the settlements were different in kind from the basic Scottish values, being consequential, and were therefore not recoverable. Certain further developments of these arguments were rejected by the Lord Ordinary and were not renewed before us. The reclaimers argued that on a proper construction of clause 20, the exclusion did not apply to the indemnities given under clause 17. In any event, if clause 20 applied, the losses resulting from the enhancement of the damages to mid-Atlantic levels constituted a direct loss on the first rule in Hadley v Baxendale (supra). There is no mention in the respondents’ pleadings of their relying on clause 20 to exclude part of the reclaimers’ losses as consequential (cf. Lord Ordinary’s 17 Opinion, 5/1071). It was suggested for the respondents that it was for the reclaimers to plead and prove that the whole contractual provisions entitled them to recover the losses claimed. In my view, this is a disingenuous argument. If a party to a contract contends that a particular provision of it provides him with a substantive defence, it is his duty to give fair notice of the point. In my view, the respondents were fortunate that the Lord Ordinary entertained this line of defence. The matter is of some practical importance in this case, because there are indications that the reclaimers may have been put at a disadvantage. Because the point was sprung on them at the end of the proof, they were unable to explore in the evidence the extent of the respondents’ knowledge of OPCAL’s connection with Texas (5/1071), or of OPCAL’s vulnerability to jurisdiction in Texas. In essence the Lord Ordinary’s decision on this point adopted the argument for the respondents. There was ample evidence to entitle the Lord Ordinary to find as he did (a) that it was reasonable for OPCAL to compromise the claims and (b) that the claims were compromised at reasonable figures (5/1193-5). The Lord Ordinary found that, but for clause 20, the payments at mid-Atlantic levels would have been covered by the indemnity. The respondents do not challenge this conclusion. In my view, the respondents’ argument on the construction of clause 20 is unsound. The respondents’ first resort is to the case law on consequential loss. But on that question the case law gives limited guidance. In general “loss” denotes a loss which directly or naturally results in the ordinary course from the event giving rise to liability (Croudace Construction Ltd v Cawoods Concrete Products Ltd ([1978] 2 Lloyd’s LR 55, Megaw LJ at 62); but in every case the meaning of an exemption from liability for consequential loss is to be sought, in the first instance at least, in the terms of the clause itself looked at in the context of the whole provisions of the contract 18 (Saint Line Ltd v Richardsons Westgarth & Co Ltd [1940] 2 KB 99; Millar’s Machinery Co Ltd v David Way & Son 1934 CC 204, Branson J at p 207; Croudace Construction Ltd v Cawoods Concrete Products Ltd, (supra). Such a clause will exclude only such types of liability as are clearly brought within its scope (Croudace Construction Ltd v Cawoods Concrete Products Ltd, supra, Parker J at p 58L). In these cases there is nothing in the relevant clause to suggest that it applies to some only of the contractor’s liabilities under the contract. It applies to the whole indemnity clause. The indemnity clause can cover any loss sustained by the company, whether it arises from a contractual, delictual or statutory liability. In such circumstances the question is whether the loss is a direct loss, that is to say, a loss of a kind that the defendant should have realised was not unlikely to occur (Koufos v C Czarnikow Ltd, supra, Lord Reid at pp 382-383; Macgregor, Damages, 13th ed, 131-132; Chitty, Contracts, 27th ed, 26-023). If that is established, the court is not concerned with the question whether the amount of it was foreseeable by the party in breach (cf Wroth v Tyler [1974] 1 Ch 30, at pp 60-62; H Parsons (Livestock) v Uttley Ingham & Co Ltd [1978] 1 QB 791, at p 813; Brown v KMR Services Ltd [1995] 2 Lloyds LR 513). As the respondents accept, “loss” under clause 20 can cover the cost of meeting a third party claim for damages. There is no authority to suggest that such a loss is anything other than direct. On the findings of the Lord Ordinary, the midAtlantic settlements caused a direct loss to the reclaimers. If the loss is of such a kind, the amount of it cannot transform it, or any part of it, into a consequential loss. In short, the Hadley v Baxendale principle applies to the heads of damages and not to the quantum (Wroth v Tyler, supra, Megarry LJ at 60-62). In Koufos v C Czarnikow Ltd (supra), for example, the loss resulted from a movement in the market 19 price during the period of delay. The court was not concerned with the foreseeability of the extent of the movement in the price. In A/B Oljefabriker Karlshamms v Monarch SS Co Ltd (1949 SC (HL) 1) the loss consisted of the cost of trans-shipping the cargo from the port of discharge to the contractual port of delivery. This was held to be a direct loss. The court was not concerned with the foreseeability of the extent of that cost under wartime conditions. On this approach, once the loss is of an admissible type, the amount of it, and the foreseeability of the amount, cannot be an issue. On this view, there is no need for this court to construe the clause by reference to Hadley v Baxendale. But if the respondents are correct in their argument that the second rule in Hadley v Baxendale applies, the findings of the Lord Ordinary entitle us to conclude that the loss, so far as it arose from the mid-Atlantic uplift, was foreseeable. The findings establish that when the contracts were entered into the parties contemplated that it was not unlikely that claims would be pursued for damages at greater than Scottish levels. The Lord Ordinary found that the parties contemplated that a variety of foreign claimants could be involved, including those resident or domiciled abroad; that the claimants would be likely to litigate where they thought would be to their best advantage, and that there must therefore be some risk that the claims would be pursued abroad (5/1021-1022). There is nothing in the respondents’ argument that the disaster aspect of this loss represents a special circumstance. The respondents have failed to offer us any criterion by which we could assess when such a special circumstance existed, whether by reason of the incident itself or of the number of fatalities or injuries. I agree therefore that the reclaiming motion on this point should be allowed. 20 IV Is liability limited to Scottish levels of damages? : The respondents’ ground of appeal 21 The London Bridge action is a convenient example with which to deal with this point. In that case the solicitors for the reclaimers sent a letter dated 2 May 1989 to the solicitors for the deceased’s widow and daughter making an offer in settlement, explaining the offer and requesting that the acceptances of the offer be docketed at the end of the letter. This resulted in an extra-judicial agreement in terms of the offer. A Memorandum of Agreement dated 28 April 1989 entered into on behalf of the reclaimers, the participants and PADG is attached to the extra-judicial agreement. Part of the agreement was that Court of Session proceedings would be raised in order to give effect to the settlement (para 4.2.2), and that as soon as possible after the action was raised the parties would execute a joint minute in order to settle it by way of decree for the agreed sums (para 4.5.3). Thereafter the deceased’s widow raised an action against the reclaimers and others (12/291 of pro) concluding for the sum agreed in the settlement. In the summons liability was imputed to the reclaimers on the basis of common law fault and breach of statutory duty. The joint minute was then executed (12/293 of pro). It narrated that the action had been settled extra-judicially on payment of the principal sums apportioned to the pursuer and her daughter, with interest from 17 April 1989 and with no expenses due to or by either party. Decree followed in terms of the joint minute and was duly extracted (12/294 of pro). On 2 May 1990 the solicitors for the reclaimers’ insurers sent cheques to the reclaimers’ solicitors to enable settlement to be made (12/296 of pro). On 9 August 1989 a full release agreement (12/297 of pro) was signed by the deceased’s widow on behalf of herself and her daughter. 21 The reclaimers were advised that it was important that there should be a decree granted against them. The reasons were, inter alia, that if actions were raised curators could be appointed to children for whom claims were made; that this would protect the reclaimers against the risk that such claims might later be renewed in the courts of Texas on the plea that the settlements were not binding on the children; and that the raising of actions against the reclaimers would preserve the reclaimers’ rights of recourse against Score (UK) Limited, the employers of Terence Sutton, who appeared to have been responsible to some extent for the accident. Throughout these negotiations the respondents reserved their position. While accepting that the settlement was reasonable from the reclaimers’ point of view, they did not accept that the settlement was reasonable in a question between the reclaimers and themselves (Lord Ordinary, 5/982). When the friendly actions were raised against the reclaimers in order to give effect to the settlement, the reclaimers did not serve third party notices on the respondents in order to have the indemnity questions resolved in the same processes. In each of the contracts to which these actions relate there was a choice of law clause by which Scots law was declared to be the exclusive law of the contract. The respondents submitted (1) that payment made by the reclaimers to the victims and their relatives was made in consequence of a decree of the Court of Session. Accordingly, the only event or peril within the terms of the indemnity clause for which the reclaimers could relevantly claim indemnity was a liability constituted by a decree of the Court of Session; (2) that the respondents were entitled to challenge the quantum of that liability; (3) that in assessing whether the amount paid under the decree was excessive, this court could have regard only to Scottish levels of damages; and (4) that even if there were claims with reasonable prospects of success under the 22 law of Texas, any such claims had been extinguished at the latest on the date of the decree of the Court of Session, and were now no longer relevant. The argument for the respondents rested entirely on the fact that a decree of the Court of Session gave effect to these settlements. The respondents argued that the reclaimers could no longer seek an indemnity against a “claim”, because the claim had been extinguished by the passing of the decree (Comex Houlder Diving Co Ltd v. Colne Fishing Co Ltd 1987 SC (HL) 85, Lord Keith of Kinkel at pp 121-122). Instead, the reclaimers sought to be indemnified for the payments made in satisfaction of a decree of this court. The reclaimers had sought that decree for their own protection. Therefore the granting of it was not a mere matter of the mechanics of effecting the settlement. It was a matter of substance. In short, by opting for the procedure of decree in implement of the settlement agreements, the reclaimers had suffered the consequence that the respondents were now subject to a liability under only one particular head of the indemnity clause; and, since that head of indemnity arose from a Scottish decree, the quantum of liability must be assessed by Scottish standards of damages. The reclaimers submitted (1) that since they and the participants were liable to compensate the relatives, the respondents were bound to indemnify each of them in terms of clause 17 of the contract and (2) that since it was reasonable for them and the participants to settle the claims at the sums sued for, decree should be granted for those sums. In my view, in settling the claims, the reclaimers suffered loss to the extent of the sums paid out to the claimants. The decree was merely a step connecting the incident to the loss. It follows therefore that up to the point of payment the decree created on the part of the reclaimers a liability. When that liability was met by 23 payment under the decree, the reclaimers suffered a loss. Therefore the reclaimers have suffered a loss resulting, by a series of steps, from the event giving rise to the claim. The loss that they have sustained is therefore one for which, on the face of it, they are entitled to be indemnified. V Is liability limited to Scottish levels of damages?: The respondents’ ground of appeal 22 The London Bridge contract provides a suitable specimen of the contractual provision on which this ground of appeal is founded. It provides as follows: “5. GOVERNING LAW This contract shall be governed by, construed and interpreted exclusively according to Scots Law. The parties agree to submission to the jurisdiction of the Courts of Scotland to the exclusion of the Laws and Courts of any other country.” The Lord Ordinary found that in all probability the claimants would have established jurisdiction against the reclaimers in the courts of Texas; that in that event the law of Texas would have governed the procedure and the method of quantification of the claim, and that in each case there was a risk of there being a much higher award in Texas than the amount at which the claim was settled. In making these findings in fact, the Lord Ordinary held that the decisive factor was the relationship between the reclaimers and their associated company in Texas. On that basis, the Lord Ordinary held that the reclaimers, had they been convened to the courts of Texas, would have been “very vulnerable” to awards in the Fisher/Allen range (Opinion, 6/1373). The respondents submitted that as a matter of general law, and on a proper construction of the contract, the reclaimers were entitled to recover only Scottish levels of damages under the indemnities. Counsel for the respondents argued that it 24 was significant that the contract provided for the exclusive jurisdiction of the Scottish courts. The respondents could not have been brought into the dispute between the claimants and the reclaimers in any jurisdiction other than Scotland. The jurisdiction clause therefore excluded the theory that the parties contemplated that the respondents could be required to hold the reclaimers harmless in any foreign jurisdiction. Counsel for the respondents argued that a further consequence of clause 5 was that the quantification of the obligation to indemnify was always a matter for the Scottish courts. In matters of quantification the Scottish courts invariably applied only the lex fori (Anton, Private International Law, 2nd ed, pp 742; 749-750). It followed that, on a proper construction of clause 5, the party indemnified could not invoke foreign law in this court in order to prove that what he paid was reasonable. The Lord Ordinary therefore failed to give due recognition to both the general law and the provisions of the contract in hearing expert evidence, and giving effect to it, on the law of quantum in the jurisdiction of Texas and in making findings of fact as to what the claims would have been worth in that jurisdiction. The reclaimers submitted that the task of the Lord Ordinary was to determine what loss the reclaimers had sustained. The Lord Ordinary had made findings in fact as to the law of Texas and as to the probable level at which awards would have been made had that jurisdiction been invoked. Since there was a serious risk that that could have happened, the Lord Ordinary was entitled to take that risk into account in concluding that the settlements were reached at reasonable figures. In my opinion, the Lord Ordinary’s reasoning was correct. The Lord Ordinary had to decide what the extent of the reclaimers’ loss was. He then had to decide whether it was reasonable, in a question with their indemnifiers, for the reclaimers to have sustained it, particularly in view of the risk to which I have referred. The Lord 25 Ordinary was entitled to make findings in fact as to the law of Texas, both on jurisdiction and on procedure and quantification, as a basis for applying Scots law to the reclaimers’ claims. He concluded that there was a serious risk that the reclaimers could have sustained much greater losses had awards been made against them by Texas juries. He therefore concluded that the settlements entered into by the reclaimers were reasonable, if regard was had to that risk. His decision to that effect proceeded on a matter of Scots law, namely that the amount of the settlements truly represented the reclaimers’ loss and that it was a loss that they had reasonably incurred. In my opinion, the argument for the respondents, so far as it rests on general principles of law, is unsound. There is no authority for it; and it appears to be contrary to principle. It is not the law that a party who sues in Scotland to recover a loss, the amount of which has been determined by the standards of a foreign system to which that party was subject, can recover only the amount at which the loss would have been assessed by the Scottish court on the application of Scottish standards. If that were the law, the Scottish courts would be failing to give such a party proper redress. In my opinion, the respondents’ argument is also unsound so far as it is based on the terms of clause 5 of the contract. The argument confuses jurisdiction with proper law. While the contract provides that the present question is justiciable only in these courts, it does not provide that the present dispute is to be determined by measuring the loss at the value at which it would have been measured by this court. The argument for the respondents also confuses the idea of jurisdiction in a question between the parties to the contract with the idea of jurisdiction in questions between the reclaimers and third party claimants to whom the reclaimers are liable. 26 VI The contribution/subrogation question Each of these actions, with the exception of the Coflexip Stena action, came to grief on the fundamental point that the reclaimers’ insurers had no subrogated rights against the respondents but would have had at most a right to contribution from them for the sums that they had paid in meeting the claims. It is agreed that this point was not expressly raised in the respondents’ pleadings. Before finding in the respondents’ favour on the point, the Lord Ordinary remarks in his Opinion that the contentions of the respondents on the matter “came somewhat as a surprise seeing that the case had proceeded for 381 days without there having been a whisper about the question of contribution. That such a fundamental argument should only emerge at the last gasp of such a long proof prima facie does not seem in harmony with a legal system that prides itself on the availability of preliminary procedures for disposing of points that are purely points of law” (6/1423). I agree with that comment. I am not impressed by the respondents’ approach to this line of defence. In our system of pleading, each party must give fair notice of the assertions of fact that he will seek to establish in evidence and of the propositions of law on which he will found. By observing this principle, the parties focus the true questions in issue and give to one another a proper opportunity to prepare (Morrison’s Associated Companies Ltd v Jas. Rome & Sons Ltd 1964 SC 160, Lord Guthrie at p 190; Neilson v Househill Coal & Iron Co (1842 ) 4 D 1187, Lord Justice-Clerk Hope at p 1193). If a party wishes to raise a new question of fact or law after the record is closed, his proper course is to seek leave of the court to amend his pleadings in order to provide a foundation on record for that question (Lawrence v Sir Wm Arrol & Co Ltd. 1958 SC 27 348, at pp. 352-3). These principles ensure that neither party can take the other by surprise. Junior counsel for the respondents accepted that the respondents’ advisers were aware that the reclaimers were insured, but he said that they were unaware of the precise terms of the insurance or other details. He argued that the reclaimers’ pleadings were to the effect that they and the participants had made payments to the claimants. The evidence did not bear this out. It established that, with the exception of the Oxy Gap payments in the Coflexip Stena case, the reclaimers and their participants sustained no loss. It was therefore simply a question of the reclaimers’ having failed to prove the loss that they claimed on Record to have sustained. That argument is, I think, disingenuous. The respondents had the means at hand to find out what the reclaimers’ insurance arrangements were. They could have recovered the reclaimers’ insurance policies by commission and diligence. If the respondents had given proper notice of their intention to raise the point, the insurers would have had an opportunity to consider it and, if so advised, raise protective actions for contribution (e.g. as in Sickness and Accident Assce Association Ltd v General Accident Assce Corpn Ltd (1892) 19 R 977). When the point was raised their right to do so, if it had existed, would have prescribed. Junior counsel for the respondents remarked that it was not for the respondents to tell the reclaimers what their case should have been. But in my view the point is whether the respondents told the reclaimers what their own case was to be. They obviously did not. In my view, the respondents’ line of defence should have been raised in their pleadings and epitomised in a plea-in-law (eg as in Kelly v Edmund Nuttall Sons & Co (London) Ltd 1965 SC 427). It involved a point of relevancy. It should have been 28 resolved as a preliminary question, either at a debate on the procedure roll or after a preliminary proof before answer limited to that question. If the point had been sound, almost all of the proof before answer would have been a waste of time and expense and a waste of the administrative resources of the court. In my view, the respondents were fortunate that the Lord Ordinary entertained the point (cf. Lade v Largs Baking Co ((1863) 2 M 17). The short contention for the respondents is that by reason of there being two obligations of indemnity owed to the reclaimers, one arising from the respondents’ contractual indemnities and the other arising from the reclaimers’ insurance policies, the reclaimers’ insurers are not entitled to be subrogated to the reclaimers’ rights against the respondents. Payment by the insurers to the reclaimers meant that the reclaimers were no longer in a position of loss. The payment therefore extinguished the indemnity liabilities of the contractors. The insurers were not entitled to be subrogated to the reclaimers’ rights against the respondents because they had none. The insurers had at best a claim for contribution from the contractors, since they and the contractors had incurred the like obligations of indemnity. The reclaimers therefore sustained no loss for which the respondents are liable. Therefore, since the actions are founded on subrogation they are fundamentally irrelevant. The reclaimers argued that the liabilities of the insurers and of the contractors were different in kind. Having paid out on the reclaimers’ policies of insurance, the insurers were entitled to be subrogated to the reclaimers’ remedies against the contractors. The Lord Ordinary found for the respondents. The effect of his decision is that, although the reclaimers proved their case on the merits and established the preconditions upon which the indemnity liabilities arose, they can recover nothing 29 from the respondents. Despite having called upon the respondents at the relevant time to fund the settlements, the reclaimers have been held to have discharged the respondents’ indemnity liabilities by virtue of having funded them from the proceeds of their own insurances. The insurers in turn, despite the respondents’ own insurance cover for this liability, acquired only a right of relief pro rata against the respondents. That right has now prescribed. More generally, the decision implies that an insurer can never be subrogated to any contractual rights of indemnity held by the insured in respect of the same risk. This is a surprising result. It is not vouched by any case in Scotland or in England or by any of the textbooks. In my view, it is apparent on the fuller review of the authorities that we have been given that the decision of the Lord Ordinary on this point is unsound. The main submission for the respondents was that they had no obligation to indemnify the reclaimers because the reclaimers had already been indemnified and therefore had no title to sue. On this submission, it is unnecessary for the respondents to consider whether or not the reclaimers’ insurers had any right of contribution against the respondents. It is enough simply to show that the reclaimers had been indemnified. In my opinion, this submission is overstated and is incompatible with the authorities. Despite the extensive range of the arguments and of the authorities discussed at the hearing, the issue turns on one short question: namely, whether the indemnity obligation and the insurance obligation were co-equal or whether they were of different kinds, the contractual obligation being primary, the indemnity being secondary. The question can be put in another way as follows: whether the two 30 obligations are to be treated as being the same because they both involve indemnity or are to be treated as being different because they do not both involve insurance. If the former view is correct, the Lord Ordinary was right in holding that the reclaimers should have sought the remedy of contribution. If the latter view is correct, the Lord Ordinary erred, because the action was well-founded, being based on the right of subrogation held by the insurers by virtue of their having met the reclaimers’ claim under the policies. In that case, the insurers were correct to maintain these actions in the name of their insured. The contracts do not specify whether or not the contractors’ obligations have primacy over that of the reclaimers’ insurers. The decision on the point therefore depends on principle. The Lord Ordinary found limited help in the cases to which he referred, because in those cases the two obligations were identical in nature (Moss v Penman 1993 SC 300; Sickness & Accident Ass Assocn Ltd v General Accident Ass Corpn Ltd (1892) 19 R 977; Albion Ins Co Ltd v Government Ins Office (1969) 121 CLR 342) In my opinion, the contractual obligation to indemnify in these cases is primary and the insurers’ obligation is secondary. These reclaiming motions are governed by the principle established, in the converse circumstances, in North British & Mercantile Ins Co v London, Liverpool & Globe Ins Co ((1877) 5 Ch D 569, Jessel MR at pp 575, 578). In that case insurers of grain who paid on a loss were subrogated to the insured’s contractual rights against the wharfingers, whose liability was in substance one of indemnity (ibid, at p 576). Mellish LJ, having referred to the fact that the bailees’ contract made them absolutely liable to the owners of the grain in case of loss, said that: “This is not a contract of insurance so as to make the bailee himself an insurer, but it is really the terms of a contract of bailment” (ibid, at p 584). 31 Similarly in Darrell v Tibbitts ((1880) 5 QBD 560) the lessee had an obligation to the lessor to reinstate the insured property in the event of its being destroyed. This was in substance an obligation of indemnity. It applied irrespective of fault. The insurers of the lessors were held entitled to be subrogated to the lessor’s rights against the lessees (ibid, Brett LJ at p 563). There was no suggestion in that case that contribution had any application to the case. There are dicta of Lord Wright and Lord Porter in Larrinaga S S Co Ltd v The King ([1945] AC 246, at pp. 256 and 261) to the effect that a marine insurer who pays on a claim under the policy is subrogated to a contractual right of indemnity held by the insured owner against the charterer. These dicta have not been doubted in any of the cases or by any of the writers. They are in line with the classic exposition of subrogation by Brett LJ in Castellain v Preston ((1883) 11 QBD 380, at pp. 388-9; cf Simpson & Co v Thomson (1877) 5 R (HL) 440, Lord Cairns, LC at p. 42). That a contractual obligation to indemnify is primary and an insurance obligation is secondary is also supported by all but one of the United States cases to which we were referred (viz. Hart v Western Railroad Corpn (1847) H F & H 99; Hall & Long v Railroad Companies 80 US 367 (1871), Strong J at pp 370, 372-3; Chicago St L & N O Rly Co v Pullman South Car Co (1891) 139 US 79 at p 104; F H Vahlsing v Hartford Fire Ins Co 108 SW (2d) 947 (1937), Slatton J at pp 949-950; Consolidated Freightways Inc v Moore 38 Wash 2d 427 (1951); North Central Airlines v City of Aberdeen, South Dakota 370 F (2d) 129 (1966)). These are cases in which insurers were subrogated to a contractual right of indemnity owed to the insured. The only exception to which we were referred is the decision of the Supreme Court of California in Patent Scaffolding Co v Wm Simpson Construction Co (256 Cal 32 App 2d 506 (1967)). This case stands apart from any decision in Scotland or in England and from the preponderant view in the United States decisions. It was decided on a principle of equitable subrogation that has no place in Scots law. No other reported case has been brought to our attention in which an indemnity clause in a contract of a different nature was equiparated with a contract of insurance for this purpose. The Lord Ordinary recognised that the contractors were not insurers (6/14331436) but he held that in these cases the two obligations were on an equal footing (6/1436; 1434-1435). On the face of it, it seems unlikely that the reclaimers’ insurers intended to place themselves on the same footing as to liability as the contractors, so as to give the contractors the benefit of their insurances. If the reclaimers had been uninsured, they would have been entitled to sue the contractors on the indemnities. But the Lord Ordinary has held that because the reclaimers have taken out insurance, they have suffered no loss; and that the principle of contribution limits their insurers’ rights. In my view, the Lord Ordinary erred on this point. There is no equality between the indemnity obligations of the respondents and the obligations of the reclaimers’ insurers. There is a difference in kind between an indemnity contract and an indemnity clause in a contract for services (cf Parr’s Bank Ltd v Albert Mines Syndicate, ((1900) 5 Com Cas 116). One is a contract uberrimae fidei; the other is not. One is an obligation of indemnity which forms the subject matter of the contract of insurance (North British & Mercantile Ins Co v London, Liverpool & Globe Ins Co, supra, at pp. 576, 584-5). The other is an obligation of indemnity that is incidental to a contract of a different kind (cf Castellain v Preston, supra; North British & 33 Mercantile Ins Co, supra). One is subject to statutory regulation, as in the insurance industry. The other is not. While the obligation of insurance is an obligation of indemnity, not all obligations of indemnity are obligations of insurance. Moreover the obligation of a contractor to indemnify is a primary obligation owed to the other party, come what may; whereas the liability of an insurer is a liability of last resort in the sense that the insurer undertakes to indemnify the insured from loss if he should incur it and should fail to be indemnified by the party responsible. The insurer, unlike the contractual indemnifier, pays out under the policy subject to recovery in the insured’s name from other parties who have a liability for the loss. If a payment by an insurer discharged a contractual obligation to indemnify there would be no basis for the principle of subrogation. I conclude therefore that the two obligations are not on an equal footing. The effect of these contracts is that if an insurer pays he will be subrogated to the reclaimers’ right against the respondents under the contract. On this analysis the reclaimers cannot be regarded as having suffered no loss. The action was therefore properly raised in the name of the insured on the initiative of the insurers and for the insurers’ benefit. The idea that the obligations of the respondents and the insurers were on the same footing is even less plausible in this case where there is a multiplicity of insurances some of which apply at different levels of risk. In this respect, the insurances in this case are akin to the sureties in Craythorne v Swinburne ((1807) 14 Ves Jun 160). For all of these reasons, I agree that the reclaiming motions on this question should be allowed. 34 VII The tax consequences of the reclaimers’ losses I agree that this ground of appeal should be refused for the reasons given by your Lordship in the chair. VIII The adjustment for interest I agree that this ground of appeal should be refused for the reasons given by your Lordship in the chair. IX The evidential questions as to the cause of the explosion (1) Evidential problems and issues The Lord Ordinary had an extraordinarily difficult task in attempting to ascertain the cause of an accident which only a handful of material witnesses survived, where there was confused and contradictory evidence on the facts, and where he had to pass judgment on the actions of two men who perished in the accident. The Lord Ordinary had to make crucial findings on the events in module C and in the control room in the few hours leading up to the explosion. He had to rely on the exiguous eye-witness accounts of a few survivors supplemented by expert scientific evidence. He had to rely to an unusual degree on inferences from facts and circumstances. In the accounts given by the survivors there were gaps and inconsistencies. 35 Despite these difficulties the Lord Ordinary’s task remained that which confronts every judge of fact, namely to make findings on a balance of probabilities by reviewing all sources of relevant evidence and drawing from them such inferences and conclusions as he considered appropriate. The reclaimers’ case was (1) that Terence Sutton had failed properly to tighten the flange with the result that it could not contain the condensate within the pipe at system pressure, and (2) that Robert Vernon had thereafter jagged the pump while the PSV was missing, thereby propelling condensate to the flange and into the atmosphere. The respondents’ case was (1) that the reclaimers had failed to prove on a balance of probabilities that the escape occurred at the site of the flange; (2) that even if that had been proved, the weight of the evidence was not sufficient to support the further inference that the escape occurred because Terence Sutton had left the flange only finger-tight; and (3) that Robert Vernon had jagged the pump in the knowledge that the PSV had been removed and had thereby been guilty of wilful misconduct. (2) The function of an appellate court on questions of fact On any disputed question of fact the judge of first instance has the advantage of seeing and hearing the witnesses and assessing their credibility and reliability. An appellate court cannot make such an assessment with confidence from a transcript. In general, therefore, where the fact-finding process involves questions of credibility and reliability, an appellate court will not disturb the findings of the judge of first instance where those findings were open to him on the evidence (Thomas v. Thomas, 1947 SC (HL) 45; Benmax v Austin Motor Co Ltd, [1955] AC 370). An appellate court may substitute its own judgment on the facts where the findings of the trial judge are not 36 dependant on the advantage enjoyed by him in having seen and heard the witnesses (cf. Thomas v Thomas, supra, Lord Thankerton at p. 54; Benmax v Austin Motor Co Ltd, supra, Lord Reid at p. 376). In such a case the court is in as good a position as the trial judge to make a judgment of that kind. In my view, this is not a case where an appellate court should disturb the findings of the judge of first instance. In Thomas v Thomas (supra) the proof that was examined on appeal consisted of the evidence of 13 witnesses. The transcript ran to 393 pages. In this case the transcript runs to 64,515 pages. We have been referred to selected passages in the transcript consisting of the whole evidence of some witnesses and parts of the evidence of others. While counsel have referred us to the essential evidence on the factual questions debated at the appeal, we could not reasonably be referred to the evidence in its entirety and we have been given little feel for the overall shape of the factual case. In particular, we have not been able accurately to assess how the issues of fact developed over the course of the proof and the reasons why the parties chose to present their cases as they did. In these circumstances, I do not think that we can be confident that we are as well placed as the Lord Ordinary to make a judgment on the facts. Apart from that special feature of the case, there is the important preliminary question whether in making his findings the Lord Ordinary had to make judgments on the credibility or the reliability of witnesses. The respondents argue that in this case the Lord Ordinary did not decide any question of fact on the ground of credibility or reliability. Therefore all of his conclusions of fact are open to review by this court. I find this hard to accept, if only because the respondents in another part of their case suggested that the evidence of the control room operator, Bollands, was not reliable and that one of the indicators of this was the inconsistency of his evidence on 37 certain points with the evidence of Clark. This was a surprising turn in the argument because when Bollands gave a detailed and circumstantial account of the events in and around the control room on the night, the details of his account, and his credibility and reliability in general, were not challenged in cross-examination; whereas Clark was challenged on both lines. In my view, the Lord Ordinary had to make numerous judgments on credibility and reliability. He had to assess the evidence of the survivors given by them in the difficult circumstances to which I have referred. He also had to resolve inconsistencies and conflicts in the evidence. For example, on the crucial question as to what Robert Vernon knew about the status of PSV 504 the Lord Ordinary had to make at least two critical decisions. These related to (1) the inconsistencies between the evidence of Bollands and the evidence of Clark and (2) the reliability of Alexander Rankin. In resolving these inconsistencies, the Lord Ordinary was exercising what has been described as the “privilege” of the judge of first instance (Clarke v Edinburgh & District Tramways Co., 1915 SC (HL) 35, Lord Shaw of Dunfermline at p. 37). He had to make a judgment based on what he had seen and heard when those witnesses gave their evidence. These were pre-eminently decisions for the Lord Ordinary. Since it appears to be accepted on both sides that these questions were central to the assessment of Robert Vernon’s state of knowledge at the material time, it follows that, on those questions at least, this court ought not to interfere with the Lord Ordinary’s decision (Thomas v. Thomas, supra; Rieley v. Kingslaw Riding School, 1975 SC 28). If that is so, the only remaining question is whether or not there was sufficient evidence to entitle the Lord Ordinary to reach the conclusion that he reached as to 38 Robert Vernon’s state of knowledge. Counsel for the respondents argued that a number of facts and circumstances, singly and cumulatively, point to Robert Vernon’s having known the status of PSV 504 when he re-pressurised the pump. Counsel for the reclaimers sought to show that these facts and circumstances do not have the weight which they might at first sight be thought to have. In my view, it is unnecessary and inappropriate for this court to venture into an assessment of the evidence on the point. So long as there was evidence entitling the Lord Ordinary to find as he did, it is not the function of this court to attempt to make its own judgment on the matter. On the evidence to which we have been referred, it is plain that there was sufficient to entitle the Lord Ordinary to reach the view that he did, notwithstanding the evidence pointing to the opposite conclusion. If I am right in the foregoing views, it follows that the respondents’ submissions on this point must be refused. (3) The circumstantial case – the approach of the court The unusual evidential difficulties of the case raise the question whether there was sufficient circumstantial evidence to establish the reclaimers’ case on a balance of probabilities; or whether, in view of the deaths of key witnesses and the loss of real evidence in the disaster, the court should have concluded that the cause of the disaster was not established (cf. Woods v Duncan [1946] AC 401; Rhesa Shipping Co SA v Edmunds [1985] 1 WLR 948). When vital direct evidence is missing, the court should proceed with care in making findings in fact (cf. Woods v Duncan, supra; Rhesa Shipping Co SA, supra); but whether the court finds the case proved on a balance of probabilities or returns a not proven verdict must depend on the circumstances of the case. 39 If there is meagre evidence that can support only speculative interpretations, the party on whom the onus rests will normally be held to have failed. In Rhesa Shipping Co SA (supra), the trial judge considered that he had to choose between the two rival theories advanced by the parties. He considered that the plaintiffs’ theory was extremely improbable, but having ruled out the defendants’ explanation, he found in the plaintiffs’ favour. That was held to be an erroneous approach. The trial judge should have kept in mind a third option, namely that the plaintiffs had failed to discharge the onus of proof (cf. Lord Brandon of Oakbrook, at pp. 953-955). On the other hand, if there is a body of circumstantial evidence which, taken as a whole, bears out the pursuer’s case, the court is not bound to hold that that case is not proved merely because other parts of the evidence might possibly have supported other theories. In my view, the Lord Ordinary did not make this error. He concluded that there was a sufficient body of evidence to justify his making the crucial inferences. He adopted a theory of the accident for which he said that there was a “marked probability” (at 6/1443) leading to an “inescapable inference”. Moreover, whereas in Rhesa Shipping Co SA (supra) the trial judge felt compelled to choose between two improbable theories, in this case the Lord Ordinary recognised that he had the further option of holding that the pursuers had failed to prove their case (at 1/214-5). With that in mind, he nonetheless concluded that on the two critical factual issues the pursuers had proved their case on a balance of probabilities. In my view, the Lord Ordinary’s approach to the problem cannot be faulted. (4) Did Terence Sutton fit the flange properly? 40 Counsel for the respondents argued that on the basis of Rhesa Shipping Co SA (supra), a finding of negligence on the part of Terence Sutton, on the basis that he left the flange only finger-tight, could not be arrived at by a process of exclusion. That conclusion was improbable. The full facts could not be known by reason of the loss of the platform and the deaths of most of the relevant witnesses. This is an attractive argument, but in my view it is unsound. This was not a case where the relevant findings were made by a process of exclusion. In Rhesa Shipping Co SA (supra), the finding that the cause of the loss was a collision with a submarine was regarded by the trial judge as improbable, not least because no submarine was seen in the area before or after the incident. In this case, although there may well have been other possibilities that an examination of the wreck might have suggested, for example a rupture of the pipe or a failure of a joint, the circumstantial case suggesting that the flange was loose was not improbable in the sense in which the submarine theory was held to be improbable in Rhesa Shipping Co SA (supra). There was evidence that the explosion was caused by a leak of condensate. There was evidence pointing to the flange as the site of the leak. There was evidence that if the flange was flogged up it would withstand system pressure, whereas if it was only finger-tight, it would not. Significant evidence on this question was given by the witness Bagnall, which the Lord Ordinary accepted. Bagnall was a practical man who was familiar with the job of fitting a blind flange. Despite vigorous cross-examination, his evidence as a whole justifies the finding that the task of fitting the flange, difficult per se for one man, would be even more difficult in the confined space in which Terence Sutton had to work. In those conditions it should have been a two-man job. There was further circumstantial evidence that it was expected that the valve would be ready for re- 41 installation later that day. The process of re-installation would involve the removal of the flange. In my view, the Lord Ordinary was entitled to hold that the pursuers’ case on this question had been proved on a balance of probabilities. (5) Did Robert Vernon re-pressurise the pump; and did the re-pressurising cause an escape of hydrocarbons at the site of the flange? These two questions can conveniently be answered together. The starting point is the undisputed fact that when pump B could not be restarted, the logical next step was to consider recovering pump A from maintenance and to restart it. If one disregards for the moment the eye-witness evidence of Erland Grieve as to the actions of Robert Vernon, there is a question as to whether the circumstantial evidence was sufficient to entitle the Lord Ordinary to hold that Robert Vernon jagged the pump. In arguing this question counsel for both parties went into minute technical detail regarding the jagging operation, its sequence and its duration, and on the particular question raised by one of the reclaimers’ expert witnesses, Dr Richardson, whether the jagging operation had been begun, was then interrupted and thereafter resumed in the short period before the explosion. These are questions to which there are no clear-cut answers; but to decide the main question as to the causes of the explosion, it is unnecessary, in my opinion, that they should be resolved. There is a danger in taking too detailed an approach to the evidence on matters such as the detail and timing of the jagging procedure and the sequence of the alarms. While such questions are important, they form only part of a wider picture. It was for the Lord Ordinary to look at the circumstantial evidence as a whole and to decide what significance, if any, to attach to individual questions of detail. 42 There was evidence that minutes before the explosion Robert Vernon had evinced an intention to restart pump A. Between then and the occurrence of the first gas alarm he had both the time and the opportunity to restart the pump. Moreover, he took certain practical steps to that end (Lord Ordinary, 4/856). In order to restart the pump he would have to jag it for reasons that are not in dispute. The effect of jagging would be to propel the condensate along the relief line towards the site of the blind flange. The first gas alarm went off at location C3 which was in the vicinity of the site of the flange. Thereafter there was a second cluster of alarms at locations C2 to C5, all of which were in the same sector of Module C. When the explosion occurred Grieve, who was working at the 68 foot level, felt the explosion directly above him at a time when he was in the general area of the 68 foot level underneath the site of the flange (62/9323). There was also evidence that he saw a fireball within seconds of the initial explosion and that the fireball was in the area underneath the site of the flange (Lord Ordinary, 2/358; 2/362-363; 4/863). The fireball was consistent with there having been a release of gas which was heavier than air and which descended to the 68 foot level from the site of the flange. This in turn was consistent with there having been a considerable release of condensate in the area around the site of the flange. These adminicles of circumstantial evidence, in my view, entitled the Lord Ordinary to conclude on a balance of probabilities that the cause of the explosion was the re-pressurising of the pump and in consequence an escape of hydrocarbons at the site of the flange. On that view, it was unnecessary for the Lord Ordinary to decide why there was a single gas alarm at location C3 followed some minutes later by a cluster of alarms at locations C2 to C5. Dr Richardson offered an explanation for this. The 43 respondents have advanced substantial criticisms of that explanation, not least because it was not put to any of the operators who gave evidence about the jagging procedure. I agree that these criticisms are there to be made; but in my view it was nonetheless open to the Lord Ordinary to reach the conclusion that the re-pressurising of pump A caused the escape, whatever may have been the detailed sequence of events that brought it about. The next question concerns the evidence of Grieve as to the actions of Robert Vernon shortly before the explosion. Grieve was the only survivor who was present at the 68 feet level when the explosion occurred. He was therefore a particularly important witness. Grieve was found by the Lord Ordinary to be credible (4/839, 848). He was apparently reliable. He seemed to have a good recollection of events. He was experienced in the jagging operation. Counsel for the respondents therefore argued that it was incumbent on the pursuers, who bore the onus of proof, to examine Grieve directly on the question whether Vernon re-pressurised pump A. Grieve’s evidence was potentially the best evidence on that point. This is the context in which Grieve’s evidence, in the submission for the respondents, raised important questions of law and of fact. Counsel for the respondents suggested that the Lord Ordinary misdirected himself in his evaluation of the circumstantial evidence. The legal issue arises in the following way. Grieve was led as a witness for the pursuers. He spoke to having seen Robert Vernon shortly before the time at which the reclaimers suggest that the pump was jagged. Robert Vernon was at the west side of pump B, a position which was on the far side of the pump from pump A. Grieve was not asked by senior counsel for the pursuers whether or not he had seen Robert Vernon at pump A. Counsel for the 44 respondents, who had an equally vital interest in the point, chose not to ask the question either; but it was not a point on which the respondents bore any onus. The propositions for the respondents were (a) that by reason of the failure of counsel for the reclaimers to ask Grieve the vital question, the court could not be confident that the direct evidence of the Grieve would not have been damaging to the reclaimers’circumstantial case; and (b) that in view of Grieve’s evidence, no other circumstantial evidence could displace the inference that Robert Vernon did not repressurise pump A. In my view both propositions are unsound. As to the first, it may well be that where a party fails to lead a witness who would have spoken directly to a material issue of fact, the court cannot be confident that the evidence of the witness would not have been damaging to that party’s circumstantial case. But where the witness is led, and his evidence does not exclude the factual theory that that party seeks to prove, there is no rule that a failure positively to examine the witness on the point is fatal to a circumstantial case which on the face of it establishes the point. The most that can be said is that, where a party could have examined a witness on a vital question, his failure to do so may in certain circumstances be a factor for the court to take into account in assessing the cogency of the circumstantial case. But where there is circumstantial evidence pointing to the factual conclusion that the party seeks to establish, and the direct evidence does not rule it out, counsel for that party may choose to leave matters at that. If the respondents’ first proposition is unsound, the second is unsound also. lt is not the law that, if a witness could have been asked a direct question on a point of fact but was not, no amount of circumstantial evidence however compelling can establish the point. It is clear in this case that there was circumstantial evidence 45 sufficient to justify the conclusion that Vernon re-pressurised the pump. I cannot see why, because of the failure of counsel for the pursuers explicitly to establish the point in the eye witness evidence, the court should hold that the point was not proved. These propositions were at first put forward as propositions of law; but in my opinion they are at best canons of interpretation of evidence (Dickson, Evidence, para. 199; Walkers, Evidence, pp. 8 to 9; Stair Memorial Encyclopaedia, vol. 10, para. 777). Counsel for the respondents came to accept that. In its final form the submission for the respondents was that the failure of counsel directly to examine Grieve on the point must be taken to have impaired the circumstantial case. On this view, the extent to which the circumstantial case was impaired was a matter for the court to assess. At this stage the respondents suggested that the Lord Ordinary had misdirected himself in failing to address this problem specifically. In my view, there is nothing in this submission. It is clear that the Lord Ordinary was alert to the problem. If the Lord Ordinary had failed to consider the significance of Grieve’s evidence, or if he had considered his evidence to have no bearing on the question that he had to decide, it might well have been said that he had misdirected himself. That is not how the Lord Ordinary dealt with the matter. On a fair reading of the Opinion, it is apparent that the Lord Ordinary took account of the gap in Grieve’s evidence and considered its significance with care. He concluded, after a review of all of the relevant evidence, that the gap in Grieve’s evidence did not outweigh the circumstantial case or deflect him from the conclusions that he drew from it (Lord Ordinary, 4/838-840). In proceeding in this way, the Lord Ordinary correctly dealt with the point. If the Lord Ordinary approached the analysis of the evidence correctly, the only remaining question is whether there was circumstantial evidence to justify the 46 inferences that he drew on this factual question. The respondents rely on Grieve’s evidence to the effect that he saw Vernon at the faraway side of the pump and on the fact that Grieve was not asked if Vernon had been at pump A. There is evidence from Grieve that suggests that Vernon did not remain throughout at pump B. Grieve gave unchallenged evidence that when he last saw him, Vernon was standing between the two GOVs and the south side of the room (61/9299). Grieve’s evidence therefore does not exclude the possibility that Vernon jagged pump A (Lord Ordinary, 4/839). Grieve’s evidence supports the view that Vernon had the opportunity to re-pressurise pump A, as he had said he would do shortly beforehand (Lord Ordinary, 4/840). Moreover, Grieve was at the 68 foot level for only a short time before the explosion, a maximum in his estimate of 2 to 3 minutes. There was evidence entitling the Lord Ordinary to hold, as he did, that at the material time Grieve could have had his attention occupied with his own tasks at the 68 foot level (Lord Ordinary, 4/835; 837; 840). There was also the evidence from Grieve that if the tripped pump could not be restarted, one would more or less immediately go on to the standby pump (62/9309). Grieve believed that pump A was on standby. He was not aware that it had been taken out for maintenance. In my view the Lord Ordinary took a proper approach in assessing Grieve’s evidence and in relating that evidence to the other circumstantial evidence (Lord Ordinary, 4/838-841). If there had been no scientific evidence in this case, there would nevertheless have been a significant body of circumstantial evidence to support the findings of the Lord Ordinary, much of which is not in dispute. One may summarise the nonscientific circumstantial case in the following way. 47 The gas cloud that brought about the explosion was located in the south-east area of Module C. The site of PSV 504 was in this area. The only interference that had taken place with the process pipework had been the removal of PSV 504 and the sealing of the two ends of the pipe with flanges. That interference had taken place earlier that day. The site of the PSV was on the relief line from pump A. Shortly before the explosion Robert Vernon had said that he intended to bring back pump A from maintenance and to re-pressurise it. He had taken certain steps towards that end. The effect of re-pressurising would be to expel condensate at the site of the flange if the flange was not properly tightened. Shortly before the explosion a low level gas alarm was activated in Zone C3. This zone was in the eastern sector of the module where PSV 504 was located. There was then a pause which was followed by a flurry of alarms in Zones C2, C3, C4 and C5. Then there was a high level gas alarm in the area of one of the centrifugal compressors, and almost immediately thereafter there was the explosion. When the explosion occurred the witness Grieve was working at the 68 foot level in the area that was approximately underneath the site of PSV 504. Grieve heard the noise of the explosion, which he thought was directly overhead. In addition, a fireball came down to the 68 foot level at a point approximately underneath the site of PSV 504. On this basis it would have been open to the Lord Ordinary to hold that there was sufficient evidence to establish on a balance of probabilities that the cause of the explosion was a leakage of condensate at the site of the flange. The next question is whether this circumstantial case is affected by the scientific evidence. Several eye witnesses established that a scaffolding platform had been erected underneath the site of PSV 504 to enable the valve to be removed for maintenance and that the platform remained in situ on the night of the accident. The 48 reclaimers did not dispute this evidence. They sought to show that the pattern of the gas alarms that immediately preceded the explosion was consistent with a release of condensate at the site of the blind flange. 49 Dr M E Davies was one of the reclaimers’ expert witnesses. His remit was, on the basis of certain hypotheses put to him as to the nature of the leak, its location, the composition of the condensate and a particular sequence of gas alarms, to conduct wind tunnel simulations to establish whether the reclaimers’ theory of the cause of the explosion was consistent with the pattern of gas alarms reported to him and, if so, to determine the typical range of release rates of the condensate. Dr Davies listed his assumptions at the outset of his Report. Dr Davies’ conclusions were that on the assumptions set out, a small leak of vapour from the condensate stream, followed by a much larger release, must have occurred in the vicinity of PSV 504, leading to the accumulation of a large explosive mass, of 40 to 60 kgs, in the south-east of Module C. Dr Davies’ wind tunnel tests had been conducted for the purposes of Lord Cullen’s inquiry. Dr Davies spoke to his results at the inquiry. It appears that at the inquiry, if not before, it became clear to him that there had been a scaffolding platform underneath the site of the flange. It seems that the respondents did not challenge Dr Davies’ evidence at the inquiry. At the stage of the proof at which Dr Davies gave evidence, the presence of the scaffolding platform underneath the site of the flange had already been established. Nevertheless, Dr Davies did not run any further wind tunnel simulations based on the presence of the scaffolding platform. For all that we know, he may not have been asked to do so. In the result, Dr Davies’ evidence at the proof was based upon experiments carried out several years earlier which had not been updated and which were based upon numerous series of tests all but one of which assumed that there was nothing erected underneath the site of the flange. However, in one of his series of tests, series 50 27, a board was positioned below the leak source apparently to replicate the presence of the scaffolding on the night of the explosion. The result of this test did not support the hypothesis put to Dr Davies, nor did it support the reclaimers’ theory of the accident. It showed that with a scaffolding underneath the site of the flange the C2 alarms would come up before the C3 alarms. In cross-examination Dr Davies was asked whether his conclusions would be affected if he were to assume that there was a scaffolding platform under the site of the leak. He concluded that in that event there would be dispersion or diffusion of the gas. He said that he would not disagree with the contention that, with such a platform in place, the dispersion would result in a repeated gas detection at Zone C2 before alarms in Zone C3 were activated (139/19723-19726). He simply could not say whether, with a scaffolding in situ, a release that went downwards could have avoided the C2 alarms and have activated the C3 alarms first (139/19732). He had not carried out the necessary tests. Dr Davies concluded inter alia that test series 27 was not fatal to the pursuers’ case because it tested an extreme case of a single release at a high release rate in a downward jet directly onto the scaffold platform. On the assumption that there had been a low level release, he considered that the presence of the platform was not fatal to the reclaimers’ theory (139/19819). The gas alarm pattern reported to him was consistent with a gas release in the area of the module where PSV 504 was located and was not consistent with a gas release in any other part of the module. The gas alarm pattern gave some slender support to the theory of a two-stage release (Lord Ordinary, 4/873-874). It supported the idea of a single alarm, the first stage, on the view that the pump was then being partly re-pressurised. It supported the second stage alarm pattern on the view that by then a substantial release would have been 51 made affecting a multiplicity of alarms. The Lord Ordinary concluded that this theory was “eminently possible” (Lord Ordinary, 4/874). The general conclusions of Dr Davies were not dependent on any detailed distinction as to the precise order of the final flurry of low level alarms. Finally, and in any event, Dr Davies’ evidence was not inconsistent or incompatible with the two-stage jagging theory of the disaster for which the reclaimers contended. Counsel for the respondents commented on certain alleged shortcomings in Dr Davies’ Report and on certain deficiencies in his work; but he further alleged that Dr Davies was in breach of the duties and responsibilities of an expert witness in a civil litigation (cf. The Ikarian Reefer [1993] 2 Lloyds LR 68). In my view these latter criticisms of Dr Davies are undeserved. Dr Davies appears to have replied reasonably and moderately to a rather hostile crossexamination. To some extent, the shortcomings in his experimental work may be explained by the fact that at Lord Cullen’s inquiry the respondents did not raise the question of the scaffolding platform as a technical issue affecting his conclusions and did not between then and his appearance at the proof intimate that they proposed to raise that question. As I understand it, the question was raised with Dr Davies at a fairly late stage in the cross-examination. Nevertheless, it is obvious that Dr Davies’ Report gave no support to the reclaimers’ theory of the accident. As he made plain in the introduction to his Report, the assumptions on which he conducted his tests did not include the presence of the scaffolding platform. Since Dr Davies accepted that the presence of the scaffolding platform would have a significant effect upon the pattern of dispersion of the gas, it follows in my view that his experiments, so far as they did not replicate the presence of the platform, are of no value to the reclaimers. On the contrary, to the limited 52 extent that Dr Davies attempted to replicate the presence of the platform, namely in test series 27, his experiment suggested an effect that was not consistent with the pattern of gas alarms reported by Mr Bollands. Dr Davies’ evidence does not rule out the possibility that the pattern of alarms could be consistent with the reclaimers’ theory; but with the added assumption of a scaffolding in situ, his results support the view that the pattern of gas alarms is not consistent with the reclaimers’ theory. That, I think, is about as far as one can reasonably take the matter. However, as the Lord Ordinary rightly observed, the reclaimers’ case does not depend on the expert evidence “viewed as in a vacuum” (Lord Ordinary, 3/561). The correct approach, in my view, is not simply to disregard Dr Davies’ evidence as if it had never been given but instead to add his evidence to the general body of evidence before deciding on the crucial issues of fact. This means, therefore, that the quite significant evidence that he gives in relation to test series 27 and the effect of the scaaffolding platform upon the pattern of gas dispersion must be weighed against other circumstantial evidence which could be said to identify the blind flange as the source of the leak. Associated with the respondents’ criticisms of Dr Davies’ evidence there was a question regarding the significance of the molecular weights of the constituent hydrocarbons in the condensate and the differential effects that occur in the flashing process. I agree that that question does not raise a true question of law. For the reasons given by Lord Coulsfield, I consider that there was nothing in that tract of the evidence that would have prevented the Lord Ordinary from reaching his overall conclusion on the facts. 53 (6) Robert Vernon’s knowledge as to the status of the PSV The respondents’ case involves the proposition that a senior employee of the reclaimers, Robert Vernon, was guilty of wilful misconduct by re-pressurising the pump when the PSV for that pump had been taken out and the pipe work closed off with a blind flange that had not been tested at system pressure. The reclaimers, on the other hand, argued that Vernon was at worst negligent. On this question the Lord Ordinary found that Vernon did not re-pressurise the pump in the knowledge that the PSV had been removed. The Lord Ordinary held that Vernon was guilty of serious inadvertence (5/1122), but that he was not guilty of recklessness or irresponsibility. The respondents challenge this conclusion on two grounds: (1) that the Lord Ordinary erred in his method of dealing with the question and therefore that the court should look at the evidence afresh; and (2) that on the evidence, the proper finding in fact should have been that Vernon knew of the status of the PSV at the critical time. The respondents argue that the Lord Ordinary’s findings were findings that he was not entitled to reach, because it involved his discounting a substantial body of evidence that at various points in the evening Vernon would have been made aware of the status of the pump. The respondents point to a series of facts and circumstances which, singly and collectively, suggest that Vernon must have known that the PSV had been taken out. The respondents further argue that the Lord Ordinary erred in failing to make findings in fact on certain critical points relating to this question. They submit that he also erred in his approach to the factual evidence because he could not properly assess the strength of the inference that he drew from the facts and circumstances without at the same time considering the strength of other facts and circumstances that implied that Vernon had been made aware of the status of the 54 pump at various times that evening. Finally, the respondents argued that the Lord Ordinary’s reasoning process in his conclusion was flawed. In support of this submission, counsel for the respondents relied on what they described as an extraordinary concatenation of circumstances. They said that there were a number of factors which gave Vernon occasion to know that the PSV was missing and that, while it is possible that in any individual case he may not have received that information, it is in the highest degree improbable that from all of these, taken together, he failed to know that the PSV was missing. They referred in particular to the fact that at the handover he would have learned that the pump was missing; that he would have known about it when he suspended the permit; that in any event the suspended permits were lying on his desk during his shift; that he would have seen the scaffold in module C in the area of the PSV; and that he would have known from his routine walkabout that the PSV was out. For these and other reasons they said that the court should confidently conclude that he must have known that the PSV was missing. It is a legitimate starting point in a consideration of this part of the case to look to the body of evidence demonstrating the unlikelihood that Vernon, or anyone in his position, would have restarted the pump in the knowledge that the PSV had been taken out. The Lord Ordinary had ample evidence to justify that conclusion, for example in the evidence of Bollands and the evidence of Clark. Moreover, no witness in the case said that the pump would ever be run without the PSV in the circumstances which arose on the night of the disaster. The main witnesses on the point were Clark, the lead maintenance hand (67/10016) and Henderson (80/11880-11883). The control room operator, Bollands, said that it would have been quite a drastic step. He did not think that Vernon would have done it (63/9480). If he was going to do that, he never 55 discussed it with the control room operator (Bollands, 63/9565) or with any of his other colleagues. If he had been concerned at losing production, it was open to him to have referred that matter to higher authority (Bollands, 63/9565-9566: Murray, 105/15143-15144). There was no evidence that this was a wholly exceptional event or that when a pump had tripped in the past, to restart the other pump without the PSV was ever considered as a serious option. The respondents are correct in saying that Vernon was probably conscious of the need to restore production. That was one of his essential functions. The evidence establishes that if pump B could not be restarted and pump A was not restored to service there would be a fourfold increase in flaring and a loss of production of the order of 20%. However, there was no evidence before the Lord Ordinary to suggest that that was seen as a pressure upon operators to take risks. The tripping of pumps was an almost everyday occurrence, but there is no suggestion anywhere of any instruction, express or implied, from the reclaimers to their operators that in such circumstances the maintenance of production was a priority that superseded the priority of safety. More generally, there was no evidence that the management had any priority higher than safety. There was ample evidence, perhaps to be taken at some discount but available to the Lord Ordinary nonetheless, that Vernon was an exceptionally conscientious and committed worker (Lord Ordinary, 4/763). There was evidence that he was able, experienced and responsible (Lord Ordinary 4/723; Henderson 79/11816-11817; Murray 105/15143). In contrast, there was no evidence that he was the kind of man who would be capable of downright irresponsibility. Like all of the lead production operators he knew that it was wrong to run the pump without the PSV. There was no suggestion 56 that that had ever been done (Lord Ordinary 4/763). Vernon had nothing to gain by doing so and he had much to lose. Finally, there was the evidence that in the immediate aftermath of the first explosion Vernon did not seem to associate the explosion with anything that had been done by him (Lord Ordinary 4/722; 4/834). This would seem to be an unlikely reaction on the part of someone who had knowingly re-pressurised the pump while the PSV was missing. The respondents’ approach is to say that if all the facts and circumstances point to Vernon’s having known that the PSV was missing, the startling nature of what he must have done should not deflect the court from concluding that that is what he did. But it is equally open to the court to conclude that there was nothing known of Vernon to suggest that he would have made such a decision; that there was no evidence that anyone had ever done such a thing before; and that if Vernon had done that, he would have done it in the knowledge that he was endangering his own life as well as the lives of his colleagues on the platform. These considerations justify the conclusion that, whatever the cause of the disaster, the cause was not a deliberate decision by Vernon to start the pump made in the knowledge that the valve was not in place. From this it would follow that if Vernon was responsible at all for the disaster the responsibility resulted from, at worst for him, his negligence. 57 The Lord Ordinary’s suggested inferences from the evidence were that Vernon had learned earlier in the evening that the PSV was out, but had forgotten about that when he jagged the pump; or that he had never known in the first place that the PSV was missing. On the whole matter, I consider that the Lord Ordinary was entitled to hold that Vernon never knew that the PSV was missing. Whether the Lord Ordinary was right in holding that Vernon might have forgotten that the PSV was missing at the time when he re-pressurised the pump is another matter. If it is open to this court to make inferences from the evidence, I would favour the inference that Vernon never knew that the PSV was out. I am impressed particularly by Vernon’s reaction to the first alarm and by the unlikelihood that he, of all people, would have taken such a course. This, I think, is a reasonable inference to draw, particularly since the pump was then out for maintenance and would therefore not be part of the operating plant taking over by Vernon on that shift. This view of the evidence also involves the inference that the permit to work for pump A must have been cancelled before the handover took place. That too is a reasonable conclusion on the evidence. I am not impressed by the theory that Vernon knew that the PSV was out but had forgotten that when he jagged the pump. If it were open to us to make findings in fact on this question I would not be inclined to make such a finding. I find it hard to accept that if a conscientious worker such as Vernon had been told that the PSV was out, he would later have failed to remember that when he restarted the pump. Counsel for the respondents made further submissions to the effect that even if Vernon did not positively know that the PSV was missing, the indemnities would not apply if he was in a certain state of mind that fell short of actual knowledge. 58 The first such submission was that if Vernon was not actively aware, he was in possession of the necessary information that would have made him aware. This submission involves the proposition that wilful misconduct in terms of the contractual definition simply requires that the act itself, and not the disregard, should be intentional. On this view, if Vernon intended the act of re-pressurising the pump, and if in doing so he was in possession of knowledge as to the status of the PSV which, for one reason or another, he did not bring to bear to that decision, that was sufficient to constitute wilful misconduct. In my view this argument is fallacious. In terms of the contractual definition it is the “disregard” which must be intentional. If Vernon had been made aware that the PSV was missing, but nevertheless overlooked that knowledge at the time when he repressurised the pump, he cannot in my view be said to have intentionally disregarded good and prudent practice. The second submission was that even if Vernon never knew, his disregard was “conscious” because of his constructive knowledge. This submission depends on the propositions (a) that Vernon had a duty to know of the status of the pump and (b) that he had the means of finding out. The respondents rely upon Wilson and McLellan v Sinclair ((1830) 4 W. & S. 398) and on decisions in criminal cases such as Mackay Bros & Co v. Gibb (1969 JC 26) and Smith of Maddiston v McNab (1975 JC 48). These cases establish that in the circumstances of each, where the person involved had constructive knowledge of certain facts, he was deemed to have disregarded those facts. In my view this case is to be decided on the terms of the indemnity itself and not on any definition of knowledge laid down by courts in other legal contexts. The cases cited by the respondents, therefore, do not assist us. Even if Vernon was grossly 59 negligent in failing to know a fact which he had the means of readily knowing, he cannot be said on that account to have been acting with intentional and conscious disregard of proper practice. The last submission on this line was that Vernon was at least guilty of recklessness. Counsel for the respondents argued that if Vernon did not know of the status of the PSV the explosion was caused by his recklessness, and therefore that the indemnities did not have effect. His recklessness, on this argument, consisted in his failure to acquire the relevant knowledge. In making this submission, counsel for the respondents relied on a comment by the Lord Ordinary that if Vernon’s conduct could be characterised as “reckless” the indemnity would not apply, a point which the reclaimers appeared to accept. In my view, the concept of recklessness has no place in this question. The contractual indemnity is excluded if, in the circumstances of the case, Vernon was guilty of wilful misconduct as contractually defined. On that definition recklessness does not enter into the matter. Nothing short of an intentional and conscious disregard will suffice. Where, by reason of recklessness, the person concerned has failed to know the relevant fact, he cannot be said to have intentionally disregarded safety with that fact in mind. Even if recklessness were relevant to this point, the issue would not be whether Vernon was reckless in failing to make himself aware of the status of the PSV. The issue would be whether he was reckless when, having failed to make himself aware of it, he re-pressurised the pump. Counsel argued that recklessness is to be determined by the objective test laid down by the High Court of Justiciary in Allan v. Patterson (1980 J.C. 57). In my view this argument is misconceived. Allan v. Patterson related to the definition of 60 recklessness in a particular statutory context. It cannot, in my view, guide this court in the interpretation of the contractual indemnity with which we are concerned. X Novus actus interveniens The permit to work relating to Score’s maintenance of the valve provided that a blind flange was to be fitted to open pipework (Lord Ordinary, 4/793). The Lord Ordinary found that one of the functions of the blind flange was to prevent the escape of hydrocarbons from the system, however they reached the site of the flange. This finding is not challenged (Lord Ordinary 4/804-807; 820). The Lord Ordinary found that a properly fitted blind flange would have been leak-tight when subjected to the equivalent of system pressure; and that all combinations of flange and seal types would give what were described as gross leaks if the bolts were only finger-tight (Lord Ordinary, 4/803: Report, 13/79 of pro). We may therefore conclude that even if Vernon had re-pressurised the pump in the knowledge that the PSV was missing, there would have been no explosion if the flange had been properly fitted (Lord Ordinary 4/820). The respondents’ proposition is that the actings of Robert Vernon in re-pressurising the pump constituted a novus actus interveniens eliding any liability arising from negligence on the part of Terence Sutton. This argument proceeds on the assumption that the reclaimers’ theory of the accident has been established. This argument becomes a live issue only if Vernon was at worst negligent in repressurising the pump. An argument that it was not reasonably foreseeable by Sutton that a leakage at the flange could cause an explosion of the kind that occurred was advanced before the Lord Ordinary but was not renewed in this court. 61 The respondents accept that Sutton owed a duty of care to secure the flange because it was foreseeable that hydrocarbons could seep past an isolation valve in the system and in this way reach the site of the flange. But they argue that it is an entirely separate question whether Sutton’s negligence in failing to secure the flange is causally connected to the explosion. They argue that the immediate cause of that harm was Vernon’s re-pressurising the pump. They characterise that action as an entirely unexpected event constituting a novus actus interveniens. In defining a novus actus interveniens in The Oropesa ([1943] P 32, at p 39) Lord Wright said that “To break the chain of causation it must be shown that there is something which I will call ultroneous, something unwarrantable, a new cause which disturbs the sequence of events, something which can be described as either unreasonable or extraneous or extrinsic.” In Haynes v Harwood ([1935] 1 KB 146) Greer LJ put the matter in the following way: “… it is not true to say that where a plaintiff has suffered damage occasioned by a combination of the wrongful act of a defendant and some further conscious act by an intervening person, that of itself prevents the Court coming to a conclusion in the plaintiff’s favour if the accident was the natural and probable consequence of the wrongful act” (at p. 153). The answer to the question whether an individual’s actions constitute a novus actus interveniens must “be dictated by common sense rather than logic on the facts and circumstances of each case” (Knightley v Johns ([1982] 1 WLR 349, Stephenson LJ at p. 367). The proper approach on a question of this kind is for the court to take a practical view of the evidence and decide whether the effects of the initial negligence were elided by some subsequent event that substituted a new cause for the loss. The evidence in this case suggests that the overriding consideration in fitting flanges was to preserve the integrity of the pipework system, either from the ingress of 62 contaminants such as dust, or from the egress of hydrocarbons. Although the witnesses who were examined on the point referred to the possibility that hydrocarbons could reach the site of the flange in consequence of their seeping through the isolation valves, that does not mean that the flange did not protect the system from leaks of hydrocarbons in other circumstances, such as those that occurred here. In my view the respondents have taken too restricted a view of the nature of the novus actus. The relevant actus is not, as they appeared to suggest, the repressurising of the pump while the PSV was missing. It is the migration of condensate to the site of the flange, whether as a result of seepage through an isolation valve, or of the re-pressurising of the pump, or of some other cause. On that approach to the matter, the re-pressurising of the pump was not a novus actus interveniens in the sense in which that expression has been defined in the cases. It was merely one means by which the foreseeable event of the escape of hydrocarbons at the loose-fitting flange could be caused. The logic of the respondents’ argument would seem to involve the proposition that Terence Sutton had a duty of care to tighten the flange to guard against the possibility of the seepage of hydrocarbons through the isolation valve, but had no duty of care to secure the flange against the possibility that the pump would be repressurised. In my view that would be an illogical proposition. It is not a complete answer for the respondents to say that the only known risk against which the flange had to be secured was that of seepage through the isolation valve. It is sufficient to conclude that when he fitted the flange, it would be apparent to Sutton that the risk was that of condensate reaching the site of the flange and escaping to the atmosphere. On that view, it is not necessary that Sutton should have been able to foresee the 63 precise circumstances in which that could occur (Haynes v Harwood, supra, Greer LJ at p. 156). The loose flange created a source of leakage with a consequent risk of explosion. Many things might have led to there being condensate in the pipe at the site of the flange. But whatever these may have been, they were merely events which produced the result the nature of which Sutton could reasonably have foreseen Haynes v Harwood (supra) provides a helpful parallel. The negligent act of the driver in that case corresponds with the negligent act of Sutton in this case. The bolting of the horse was the consequence of the negligent act as was the escape of gas in this case. The cause of the bolting corresponds with the cause of the hydrocarbon reaching the flange. The intervention of the policeman, which was held not to constitute a novus actus, corresponds with the re-pressurising of the pump by Vernon (cf Maloco v Littlewoods Organisation Ltd, 1987 SC (HL) 37, Lord Goff of Chieveley at p. 77-8). XI Disposal I agree that these reclaiming motions and the cross-appeals should be disposed of as your Lordship in the chair proposes.