Fluor v Shanghai Zhenhua Heavy Industry Co, Ltd

CONTENTS

On 7 October 2016 I handed down a judgment on the issues of liability in this case (“the Liability Judgment”) following a hearing held during February, March and April 2016.

This is a dispute about the quality of the fabrication of steel piles (“MPs”) and transition pieces (“TPs”) for an offshore wind farm in the North Sea. The Claimant (“Fluor”) is an engineering, procurement, construction, maintenance and project management company registered in England (and is a subsidiary of Fluor Inc, the US multinational construction company). The Defendant (“ZPMC”) is a well-known steel fabricator, based in Shanghai with a major manufacturing base at Changxing Island. ZPMC has an international reputation as a heavy duty equipment manufacturer, in particular ship to shore container cranes.

This judgment concerns the issues of quantum, which are made much more difficult because in June 2010 substantial elements of Fluor’s claim were settled by an agreement set out in two letters, known as the “waiver letter” and the “warranty letter”. Accordingly, this judgment is not concerned only with the assessment of the damages to which Fluor is entitled - an exercise difficult enough in itself - but also with the extent to which the entitlement to damages which Fluor might otherwise have had has been curtailed or extinguished by the settlement agreement.

At the hearing Fluor was represented by Mr Sean Brannigan QC and Mr Richard Osborne, instructed by Roberta Downey of Hogan Lovells, and ZPMC was represented, as before, by Mr Andrew White QC, Mr Sean O’Sullivan QC, Mr Mark Chennells and Mr Iain Munro, instructed by Mr Adam Harris of Pinsent Masons. As before, I am very grateful to them all, and their teams, for their very carefully prepared submissions and for ensuring that this difficult and complex trial ran so smoothly. I owe a particular debt of gratitude to the solicitors on each side who went to enormous lengths - seven days a week - to provide me with the documents that I needed when I needed them.

The following few paragraphs are taken almost verbatim from the Liability Judgment, but it is easier to repeat what I said than to expect the reader to go to that judgment.

By the terms of the contract made between Fluor and Greater Gabbard Offshore Winds Ltd (“GGOWL”), known as the BOP Contract, Fluor agreed to engineer, procure and construct the foundations and infrastructure to support 140 wind turbine generators (“WTGs”) which were to be installed in the North Sea approximately 26 km off the coast of Suffolk.

Each foundation consisted of an MP, which was a massive steel structure composed of individual “cans” of rolled steel plates which were then welded together to form a cylindrical column, typically 65 m in length, 5-6 m in diameter and weighing between 550 and 700 metric tonnes. Each of these MPs was to be driven approximately 32 m into the sea bed. At the top of each MP sat the TP, each one of which was approximately 23 m in length and 5 m in diameter. The TP provided the connection between the MP and the structure above it.

Fluor’s plan was to fabricate the MPs and TPs in Shanghai and then ship them to Vlissingen in the Netherlands. Vlissingen was to be a staging port, where Fluor would receive the MPs and the TPs after fabrication, carry out certain work to the MPs (“the base scope works”) and install the requisite electrical equipment in the TPs, including transformers, and then load the MPs and TPs onto special-purpose vessels for transportation to the site and subsequent installation.

In order to ensure the quality of the welds in the MPs and the TPs the BOP Contract provided for non-destructive testing (“NDT”), one form of which was ultrasonic testing (“UT”). UT involves several different techniques known as scanning patterns: these are discussed in some detail in the Liability Judgment and so I do not propose to go into any more detail about them in this judgment.

For present purposes it is sufficient to say that the BOP Contract and, in turn, the contract between Fluor and ZPMC, provided that welds were to be scanned by a UT technique known as E scanning. This form of scanning turned out to be unable to detect certain types of transverse cracks in the welds. Unfortunately, the welding carried out by ZPMC contained a significant and, on the face of it, an unacceptable amount of transverse cracking, particularly in the circumferential welds. This was discovered by GGOWL when it arranged for some random testing to be carried out at Shanghai in March 2009 and, a few weeks later, on some of the MPs in Shipment No 1 shortly after its arrival at Vlissingen.

It was common ground between the experts, and I found, that the extent of the cracking in the MPs was such that they could not reasonably be installed in the seabed by Fluor unless either the cracks were repaired or Fluor obtained an Engineering Critical Assessment (“ECA”) that the structural integrity of the MPs (and TPs also) was not adversely affected by the presence of the cracks.

On 3 June 2009 (or 4 June – the precise date is in dispute) GGOWL issued a Non Conformance Report (“NCR”) on the MPs in Shipment No 1. Similar NCRs were issued in respect of Shipments Nos 2 and 3 on 29 July and 7 August 2009, respectively. Although these NCRs, on their face, referred only to the MPs, the view apparently taken at the time by Fluor was that for all practical purposes they applied to each shipment in its entirety - both MPs and TPs. Another slightly unusual feature of the NCRs was that, although they referred to a need to revalidate the NDT of the welding, the method of revalidation was not stated. However, GGOWL soon made it clear that it would only accept the revalidation if it was done by a technique known as D scanning, instead of E scanning.

This gave rise to an allegation by Fluor, supported by ZPMC, that this method of testing was “extra-contractual” with the result that Fluor, and therefore ZPMC down the line, was entitled to be paid for it. This was disputed by GGOWL and so Fluor commenced an arbitration against GGOWL for appropriate relief. Fluor was unsuccessful: by a Partial Award issued in November 2012 the arbitral tribunal (“the Tribunal”) found in favour of GGOWL.

In this judgment it is not necessary to embark on any detailed explanation of the issues in relation to the NDT regime and the dispute that arose out of it: it is all set out in the Liability Judgment. All that needs to be said for present purposes is that E scanning and D scanning were broadly similar techniques. To the extent that it is necessary, I accept the evidence given by Mr Hans Ho, of Fluor, to the effect that if in the first place Fluor had been required to carry out D scans instead of E scans, that would not have been materially more time consuming than carrying out E scans. It is true that in order to perform D scans it might have been necessary to carry out some grinding or smoothing of the weld caps, but once that had been done it only required one pass with the probe along the line of the weld as opposed to carrying out separate passes on either side of it. The problem at Vlissingen was that the entirety of the MPs had to be retested and, where necessary, repaired before they could be loaded out for installation in the seabed.

The existence of the cracks in the welds, and the consequent issue of the NCRs, threw the project into complete disarray. Fluor had anticipated that it would be able to receive the MPs and TPs at Vlissingen and thereafter prepare and install them at sea in an orderly sequence. In fact, by early August 2009 it found itself having to accommodate three shipment loads of MPs and TPs at Vlissingen and with a massive inspection and repair operation on its hands. In effect, what was intended to be a staging port had to be turned into a major manufacturing facility.

Once it became apparent, in early June 2009, that the MPs (and possibly the TPs also) in Shipment No 1 were going to have to be retested and repaired, Fluor arranged to hire a substantial labour force to carry out the required testing and weld remediation. In addition, it had to lease extra space at Vlissingen, hire tents in which to carry out the work and buy or hire (or extend the hire of) additional plant and equipment - such as positioners to rotate the MPs (so that they could be worked on), scaffolding and cranes. ZPMC, for its part, also sent over to Vlissingen a substantial number of welders to carry out repairs, principally on the MPs.

The expense of all this was enormous, but matters were made worse because Fluor had to bear extensive charges by way of standby costs for vessels that had been chartered to carry out the installation of the MPs and the TPs but which were left largely idle.

Two further factors added to Fluor’s troubles. The first was that a very substantial proportion of the MP’s had “out of roundness” defects, namely that the upper rim of the MP was not within the prescribed tolerances (i.e. was not perfectly round) which meant that the cylindrical hammer which was to be used to drive the MPs into the sea bed might not fit properly.

The second problem, which emerged in late July/early August 2009, was that there was a defect in some of the transformers that were to be installed in the TPs. During pre-commissioning of these it was found that a few had a high resistance defect, suggesting some form of loose connection within the transformer. The defective transformers were returned to Waterford in Ireland where they had been manufactured, but after inspection and discovery of the defect, the manufacturer, ABB, was unable to say how widespread the fault might be. As a result, ABB advised the recall of all the transformers to Waterford for inspection and, if necessary, repair.

But fate was not smiling on Fluor. When the transformers were returned to Vlissingen it was discovered that they suffered from a different fault that had been caused during the process of inspection and repair in Waterford. As a result, all the transformers had to be returned, yet again, to Waterford to have the new fault put right. The first batch of (two) transformers that had had these “second round” repairs were not returned to Vlissingen until 5 January 2010 (Footnote: ¹ ) - way beyond the date when Fluor had planned to start installing them.

In the meantime, both Fluor and ZPMC had concluded (Fluor having by then engaged various experts) that the cracking in the welds did not actually impair the structural integrity of the MPs or TPs and, on 1 October 2009, Fluor announced its intention of resuming installation on site irrespective of whether or not the MPs and TPs had been repaired. It now appears that they were right: it is no longer asserted by anyone that the MPs and TPs were not good for their 25 year design life. It seems that they had been sufficiently overdesigned so that they were well able to withstand the likely imposed loads in spite of the cracks in the welds.

However, by this time the summer season was over and so Fluor had to resume installation in less favourable weather conditions and times when, in some cases, the specialist vessels required were no longer available and less suitable substitutes had to be used.

I found that the cracking in the welds was the result of breaches of contract by ZPMC. My conclusions were, I think, based on broadly similar reasoning to that adopted by the Tribunal in its First Partial Award. (Footnote: ² )

However, during the latter part of 2009 and early 2010 Fluor and ZPMC held a series of discussions with a view to settling the dispute between them arising out of the welding and to combining forces to recover their losses from GGOWL. These discussions culminated in a meeting in Shanghai on 20 April 2010. At this meeting Fluor presented ZPMC with the two letters to which I have already referred - the waiver letter and the warranty letter. The former, which concerned “the joint pursuit of the substantial additional costs both Fluor and ZPMC have incurred as a result of the MP and TP weld testing and repair protocol required by [GGOWL]”, provided that ZPMC would assign to Fluor its claims for the costs it had incurred arising out of the “enhanced testing and remediation of the MP and TP welds it has been required to perform” and that Fluor would in turn waive its claims against ZPMC for “the additional costs and delays it suffered as a result of NCRs 006, 008 and 009 issued by GGOWL”.

Under the terms of the warranty letter (in its final form) Fluor agreed to release the final payment under the contract to ZPMC in return for a 25 year warranty of all the unrepaired MPs and TPs.

The two letters were not in fact signed until after ZPMC had obtained the approval of its main board in June 2010, by which time some relatively modest alterations had been made to the terms of the two letters. (Footnote: ³ ) The detailed sequence of events is set out in the Liability Judgment and there is no need to repeat it here.

One of the central issues at the liability hearing was the effect of this agreement. ZPMC contended that Fluor had agreed to waive all its claims arising out of the weld defects in return for the assignment to it of ZPMC’s claims and the warranty.

Fluor’s position was the contrary: its case was that the effect of the settlement agreement was in fact very limited. It submitted that it did not agree to waive costs and delays that it would have incurred had there been no NCRs: it said that it did not suffer any significant costs or delays “as a result of the NCRs” because it would have had to carry out all or the bulk of the additional testing and remedial work in any event. This was a somewhat surprising argument, since no less than four senior representatives of Fluor had given evidence to the contrary before the Tribunal. In the arbitration Fluor’s case was that all the testing and remedial work was the result of the issue of the NCRs. In the Liability Judgment I concluded that on this point Fluor’s case in the arbitration was correct and I therefore rejected the arguments put forward by Fluor at the liability hearing.

The effect of this conclusion is to impose significant limitations on the damages that can be recovered by Fluor, in particular in relation to Shipment Nos 1 and 3. Shipment No 2 is in a different position because NCR 008 was not issued until 29 July 2009, one month after Shipment No 2 had arrived at Vlissingen. I shall discuss the consequences of this later in this judgment and, in particular, Fluor’s argument that the waiver did not apply in the case of any MP (or TP) on which work of testing or repair had begun before the relevant NCR was issued.

These are wide ranging and all I propose to do at this point is to give a summary of the major issues which arise. They are:

The extent to which these issues - or some of them - will arise will largely be determined by the approach that is to be adopted to the assessment of damages. I therefore turn to that question first.

The damages which Fluor would ordinarily be entitled to recover in respect of ZPMC’s breaches of contract would be the sum of money which would restore it to the position that it would have been in if the contract had been properly performed (subject, of course, to any applicable exclusions or limitations in the contract). As a principle, this is, unsurprisingly, common ground. However, the parties do not agree how that principle should be applied in this case.

It is perhaps worth recalling precisely what I said at paragraphs 501 and 502 of the Liability Judgment. It was this:

“501.

In my judgment, the waiver letter must be read in the light of the parties’ mutual understanding of the position, namely that the claims that were to be pursued against GGOWL were for the additional costs incurred in respect of the extra contractual work instructed by GGOWL. Those are the costs that form the subject matter of the assignment by ZPMC to Fluor and which were the subject of Fluor’s release and waiver at numbered paragraph 3 of the waiver letter.

502.

The corollary of this is that costs which did not arise out of or were not incurred in respect of enhanced testing and consequent remediation of the MPs and TPs required by GGOWL are not costs the claim for which was being assigned by ZPMC to Fluor. Similarly, such costs would not be included in the waiver given by Fluor.”

(Original emphasis)

Thus Fluor agreed to waive the costs of any delay and the loss and expense that it incurred as a result of or in carrying out the additional NDT and remedial work to each shipment following the issue of the NCR in relation to that shipment that was required by that NCR. I concluded also that Fluor is estopped from contending that the costs of carrying out work required by an NCR were not costs caused as a result of the NCR because Fluor would have carried out the work in any event.

Putting it another way, as a result of the settlement agreement:

In relation to delay and loss and expense, Fluor’s principal argument was that where, before the issue of an NCR, work of testing or repair had been started on a particular MP, any subsequent testing or repair of that MP must be treated as not having been required by the NCR. Accordingly, submitted Fluor, the costs of such testing or repair, together with the costs of any consequent delay, are recoverable.

I regard this argument as quite unsustainable. It is a submission that cannot stand in the face of the conclusion reached at paragraph 547 of the Liability Judgment, where I said:

“Accordingly, I find that it was the common understanding of the parties in June 2010 that the enhanced testing and remediation of each shipment required by the NCR for that shipment was to be regarded as a result of the relevant NCR. Whether or not Fluor privately took the view that it would have carried out the enhanced testing and remediation in any event is, for the purposes of the construction of the waiver letter, irrelevant.”

It is therefore irrelevant whether or not anyone had started to test an MP prior to the issue of the NCR for the relevant shipment, because once the NCR had been issued the MP in question was required to be tested and any defective welds repaired. The implausibility of Fluor’s argument becomes even more apparent when the only testing that had been done prior to the issue of the relevant NCR was not done by Fluor, but by GGOWL.

Since I reject this argument, it follows that what Fluor has described as its Cases 1 to 3, which are variants of the principal argument, are not legitimate ways of framing its claim for damages. Thus Fluor’s approach, where it involves identifying those MPs on which work of testing or repair had started prior to the issue of the relevant NCR and then taking that number as a proportion of the total number of MPs in the shipment, is one that cannot be accepted.

However, what Fluor describes as Case 4, which is generally confined to losses occurring prior to the issue of the relevant NCR, is in my view a permissible approach - in principle, at least.

It seems to me that the only way in which Fluor’s claim can be approached is to consider the position in which Fluor would have found itself if the MPs and TPs had been fabricated properly, that is to say without the widespread welding defects that gave rise to the issue of the NCRs and the need for repair. In the ordinary course of events that position would then be compared with the position in which Fluor actually found itself. In this situation the assessment of Fluor’s damages - albeit difficult as a matter of detail - would be a relatively straightforward, if complex, exercise.

However, as discussed above, the settlement agreement severely restricts the amount of damages that Fluor can recover. In addition, further factors that must be taken into account are events that may have caused delay, disruption or loss and expense that were not the result of a breach of contract alleged against ZPMC. I will give two examples. The first is the problem in relation to out of roundness of the MPs. The second concerns the defects in the transformers. Whilst the out of roundness might have been the result of a breach of contract by ZPMC, no such breach has been pleaded in these proceedings – because Fluor’s cause of action in respect of any such breaches appears to have been compromised during the negotiations for the final account.

If there had been no breaches of contract in relation to the welding, Fluor would have been able to carry out the base scope work to the MPs in accordance with its programme (subject to any delays caused by the out of roundness problems or other factors not the responsibility of ZPMC) and then install them in the seabed. For the reasons given below, the loss of the opportunity to do this has been waived in relation to Shipments Nos. 1 and 3.

In relation to NCR 006, there is no possibility that any MPs could have been loaded out and installed in the seabed prior to the issue of the NCR for the simple reason that the required vessel, The Stanislav Yudin, was not even mobilised until 16 June 2009, nearly two weeks after the issue of the NCR.

The result in relation to NCR 009 is the same, but for different reasons. Shipment No 3 arrived in Vlissingen on 1 August 2009. NCR 009 was issued six days later on 7 August 2009. For any particular batch of three or four MPs the period required (from the arrival of the shipment) to carry out the base scope work and then subsequently load out on to the barge prior to installation was significantly in excess of this - some 11 days. (Footnote: ⁴ ) So even if the MPs in Shipment No 3 had been sound there would not have been enough time for Fluor to carry out the necessary base scope work to any MP before the date when the NCR was actually issued. Thus the loss of the opportunity to install the MPs was a result of the need to comply with the requirements of the NCR.

In my judgment, therefore, in relation to Shipments Nos 1 and 3 Fluor can recover only those expenses which it had incurred, or which it had agreed to pay, prior to the issue (or taking effect) of the NCR as a result of the defects in the welds. If, in addition, the presence of the welding defects caused, prior to the issue of the relevant NCR, subsequent identifiable delay, the costs of that delay would also be recoverable.

In relation to NCR 008, the position is different. Shipment No 2 arrived at Vlissingen on 29 June 2009, just over one month before the issue of NCR 008 on 29 July 2009. If, during that period, Fluor could have completed the base scope work on some of the MPs in that shipment and installed them in the seabed, then that is an opportunity that it lost as a result of the defects in the welds of the MPs and TPs in that shipment: it was not an opportunity that it lost as a result of the issue of NCR 008. However, ZPMC contends also that the ability to install the MPs from Shipment No 2 during July 2009 was adversely affected by the issue of NCR 006 and so, to that extent, Fluor’s claim in respect of Shipment No 2 has been waived.

The claim based on the loss of the opportunity to install the MPs from Shipment No 2 is not one that Fluor advanced prior to the start of the hearing, but I raised it as a possibility during the parties’ opening submissions and both delay experts provided helpful supplementary notes addressing that way of putting the case. Damages are nearly always at large and, subject to the existence of the necessary evidence, a claimant is entitled to formulate its claim for general damages in any way that is reasonably open to it provided that the defendant is given a proper opportunity to meet that claim. The court must do the best it can to assess those damages and it is not required to apply the balance of probability test to every aspect of the measurement of the loss (Footnote: ⁵ ) . Of course, claims for special damages - that is, for specific expenses actually incurred as a result of the breach - must be pleaded and proved on the balance of probability, as must any points of causation or mitigation that are to be raised by the defendant.

By the conclusion of the quantum hearing, Fluor described this approach as the one “which both parties now favour”. I do not think that this correctly reflects ZPMC’s position (see, for example, paragraph 83 of its Closing Submissions), but it is a case that both ZPMC and Mr Morgan have addressed in considerable detail and I did not understand ZPMC to suggest that any injustice would result from allowing Fluor to reformulate its claim in this way. (Footnote: ⁶ )

However, in its “Reply Note” dated 7 September 2017 ZPMC accepted that, in principle, there was nothing objectionable in first considering, in relation to Shipment No 2, the position as if the contract had been properly performed and then making an adjustment to strip out the consequences of NCR 006 in relation to Shipment No 1. It then made the following submission:

“The potential problem with seeing this as a “two-stage” process is that there is an increased risk of falling into error by approaching the two “stages” inconsistently and thereby allowing Fluor to recover in respect of costs and delays which have been settled. There is a significant risk that, when one considers the temporal consequences of the unsettled periods, the temporal consequences of the NCRs are not fully taken into account.”

ZPMC added the following footnote to this submission:

“For example, the impact of matters like out of roundness or transformers cannot be determined in isolation at a “properly performed” stage: the consequences of the NCRs influenced how these issues were experienced and addressed.”

This is a submission that goes to the heart of this part of the case. I will take the problem of out of roundness. I accept that, if and insofar as remedial work to correct the out of roundness of any particular MP in, for example, Shipment No 1 could not be carried out until the additional NDT and any consequent repair work required by NCR 006 had been carried out to that MP, the cost of the delay so caused would be a result of the issue of the NCR and therefore excluded by the waiver letter. However, if it was the case that Fluor, distracted by the discovery of the cracking and the issue of NCR 006, did not address the out of roundness problem with the urgency that it would otherwise have given to it during June and July 2009, I do not consider that that can be said to have been caused by the issue of NCR 006. That NCR did not require Fluor to postpone its consideration of the out of roundness problem: that was simply a foreseeable result of ZPMC’s breaches of contract in relation to the welding, albeit one compounded by the issue of NCR 006.

For the sake of completeness, I bear in mind also the well-known principle, derived from Armory v Delamirie (1722) 1 Strange 505, namely that uncertainties should be resolved by making assumptions favourable to the claimant where the defendant’s wrongdoing has created those uncertainties. In Keefe v The Isle of Man Steam Packet Company [2010] EWHC Civ 683, Longmore LJ expressed the principle in the following terms, at paragraph 19:

“. . . A defendant who has, in breach of duty, made it difficult or impossible for a claimant to adduce relevant evidence [as to the consequences of a breach of duty] must run the risk of adverse factual findings.”

In my judgment, that principle may have some limited application to the facts of this particular case. Although I have held that Fluor has waived its right to claim the costs of testing and remediation following an NCR for the relevant shipment which required that testing and remediation, that does not mean that I cannot take into account, when assessing any loss suffered by Fluor as a result of the welding defects in Shipment No 2 prior to the issue of NCR 008 on 29 July 2009, the additional difficulties faced by Fluor when formulating that loss caused by the fact that the defective MPs from Shipment No 1 were still at Vlissingen and were also being worked on during July 2009 (albeit that most of the costs consequent upon those MPs being defective have been waived).

It is also important to keep well in mind the fact that Fluor did not waive the breaches of contract by ZPMC in relation to the welding, but only certain consequences of them: namely, the costs - in terms of both time and money - of complying with NCRs 006, 008 and 009. The extent of those consequences was analysed in detail in the Liability Judgment leading to the conclusions that I have set out above. This point was emphasised by ZPMC when it said, at paragraph 81 of its closing submissions, that the parties settled “particular costs , not causes of action ”.

In my view, this distinction provides the key to resolving one aspect of the different approaches adopted by the delay experts. Mr Anthony Morgan, ZPMC’s delay expert, has analysed in meticulous detail “what would have happened, in terms of timing, but for the unsettled consequences of the Judgment Breaches” (see the First Joint Statement of the Delay Experts, dated 6 March 2017, at paragraph 14). Taking Shipment No 2 as an example, Mr Morgan has considered how the weld defects in Shipment No 2 affected the overall progress of the onshore works at Vlissingen to the Shipment No 2 MPs and TPs during the period 29 June to 29 July 2009. In other words, how they affected the progress of that as built work and, in turn, the actual critical path of the MP/TP installation as a whole. This was referred to as the “but for” approach. (Footnote: ⁷ )

One aspect of this approach arises in relation to Fluor’s claim for the loss of the opportunity to install MPs in the seabed during July 2009 (prior to the issue of NCR 008) as a result of the weld defects in Shipment No 2. An important parameter in the analysis of this loss is the rate at which The Stanislav Yudin would have been able to install MPs during July 2009 (assuming favourable weather conditions). What actually happened was that for the first sixteen MPs installed, the rate was, on average, 1.7 days per MP (at least, I will assume this – it is in dispute). For the subsequent installation that rate improved to 1.5 days per MP (or less) – presumably, it was suggested, because by then the operators involved benefitted from the learning curve. (Footnote: ⁸ ) When considering the rate at which the first few MPs from Shipment No 2 would have been installed during July 2009 if those MPs had been delivered free of defects, Mr Morgan has taken the former figure. This is because, he says, that provides the best indication of the rate at which the first MPs would have been installed.

However, and on the assumption that there was such a learning curve, had the MPs from Shipment No 1 been delivered free of defects, the rate of installation of the first few of them would have been reduced as a result of the learning curve. As I have already said, Fluor’s claim for the loss of the opportunity to install the MPs from Shipment No 1 during June/July 2009 has been waived. However, if the claim had not been waived, the effect of the learning curve on the rate of installation of the first few MPs would have been a factor to be taken into account when formulating and quantifying the claim (the effect of taking it into account would have been to reduce the claim).

Since the claim in relation to Shipment No 1 has been waived, and with it the allowance that would have had to have been made for the learning curve, it must be wrong in principle to take the learning curve into account - effectively for a second time - when considering the claim in relation to Shipment No 2.

In my view, to start with the assumption that, when the time came to install the MPs from Shipment No 2 in mid-July 2009, no MPs from Shipment No 1 would have been installed is to mix the real with the hypothetical. In the hypothetical “counterfactual” scenario the starting point is to consider the position as if the contract had been properly performed - not just that part of it relating to Shipment No 2 - and then to compare it with what actually happened in the light of the breaches. In terms of the assessment of damages, the question of what costs and delays have been settled by the waiver letter falls to be taken into account at a later stage. That said, I accept ZPMC’s submission that when considering, for example, Shipment No 2, the court must be careful not to award damages in respect of any disruption to the work to Shipment No 2 that was a direct result of the issue of NCR 006 in relation to Shipment No 1.

I consider, therefore, that there is a fallacy underlying some aspects of ZPMC’s “but for” approach: it is that it assumes implicitly that Fluor has waived (or largely waived) ZPMC’s breaches of contract, rather than waiving certain costs that were the consequence of those breaches (being the costs incurred as a result of the issue of the NCRs). I should make it clear that this conclusion is not a criticism of Mr Morgan: I have no doubt that the approach that he adopted was one properly taken in the light of legal advice from ZPMC’s legal team on what are essentially questions – and difficult questions - of law and causation. Although, once one has arrived at it, the answer may seem reasonably obvious, the analysis that has led me to it is one that I reached only after considerable deliberation.

I should make it clear that this conclusion does not mean that the exercise carried out by Mr Morgan is of no value and must be disregarded: rather, it is just to say that any conclusions to be drawn from it must be checked against the underlying assumptions.

Mr Kampman was engaged by Fluor in January 2009 as a contracts administrator with responsibility for administering the two subcontracts which were then in place, one with Verbrugge Zealand Terminals BV (“VZT”), for port facilities and logistics on site, and the other with SEPAM Specialists Ltd. (“SEPAM”), who was responsible for the installation of the electrical equipment into the TPs. Mr Kampman’s mother tongue is Dutch, but he spoke fluent English and gave evidence without an interpreter.

I thought that he was a straightforward and honest witness, but his memory of the relevant events, which took place almost 8 years ago, was understandably hazy.

Mr Frampton was assigned as the Project Contracts Manager for the project, a position which he held from April 2008, when the BOP contract was signed, until about April 2012.

In my view Mr Frampton was a careful and honest witness, but he was asked a number of questions on matters such as “out of roundness” of the MPs, tolerances and so on, which were not his area of expertise and so much of this part of his evidence was little more than intelligent speculation. However, where the matters about which he was asked were within his area of competence, I accept his evidence save where I specifically indicate otherwise.

In 2009 Mr Johnston was a Senior Manager for Project Business Services (“PBS”), a department of which he became the Manager in May 2012.

Mr Johnston was a precise witness who was asked a number of questions about aspects of the claim such as the payroll burden calculations. Although he was responsible for the production of the relevant figures, he was unable to explain how some of the relevant percentages had been derived because it seemed that they had been provided by Fluor Corporation and so, I assume, were based on US market conditions with which he was not particularly familiar. One example was a mark up of 7.5% for liability insurance: Mr Johnston could not explain how this percentage had been arrived at. Whilst I have no reason to doubt Mr Johnston’s honesty, for these reasons I shall have to treat some of the figures that he has put forward with a little caution.

Mr Connolly is employed by Fluor as Director of Project Controls, a position that he has held since March 2004. He was one of those responsible for putting together the quantum of Fluor’s claim against GGOWL in the arbitration and he had a similar role in this litigation.

One of Mr Connolly’s tasks was to look out for issues which might give rise to a potential claim and, where appropriate, to assemble the figures and detail necessary to support the claim. He was very knowledgeable about the financial aspects of the project and he and those working for him had clearly done a great deal of work on issues such as adverse weather conditions.

At the time of the New York mediation in March 2013 Mr Connolly was asked, at very short notice, to provide figures for the claims that Fluor wished to “have on the table” at the mediation. In some cases much of the work had already been done, but in others it was unfinished and so he had to do a great deal of work in a short time in order to provide the numbers that Fluor’s senior management wanted.

Mr Connolly was cross-examined at some length and I thought that he gave his evidence honestly, although I think that on some occasions he was being pushed to the limit of what he could be expected to know. But on one particular occasion when he was faced with a difficult question, he paused and then said that he did not have an answer - he did not attempt to bluster his way out of the difficulty.

Although I have no reason to doubt Mr Connolly’s integrity, the circumstances in which he was required to put forward some of the figures does not inspire a great deal of confidence in their reliability.

Mr Eisman is employed by Fluor as a Project Director, a position he has held since 2001. He was asked to go to Vlissingen in July 2009 to manage the works on site. He was given the title of Site Manager. His job was not to investigate the reasons for the weld defects, but to make sure that there was a proper quality assurance and quality control structure and process in place at Vlissingen and organise the work of preparing the welds as quickly as possible.

I thought that Mr Eisman was an impressive witness, but some of his evidence was really no more than evidence of speculation or opinion. For example, his evidence as to how Fluor would have dealt with the out of roundness problem in July 2009 - which was before his arrival at Vlissingen - was essentially speculation. Similarly, his evidence as to how Fluor would have dealt with the transformers had there been no defects in the MPs or TPs was, essentially, evidence of opinion - albeit based on his experience of many years of employment with Fluor. I am not prepared to disregard such evidence simply on that ground, but it does diminish the weight that can be given to it.

Mr. Paul Bruno is employed by Fluor Corporation as Managing General Counsel, having joined the company in 2001 since when he has led Fluor Corporation’s dispute resolution team. He became involved in this dispute in mid-July 2009.

Mr Bruno was closely involved in the discussions that led to the settlement agreement with GGOWL in 2013. Although I thought that he was an honest witness, I did not find his evidence in relation to those discussions to be of much assistance for the reasons that I give later in this judgment.

Mr Ayres gave evidence at the liability trial and in the Liability Judgment I said that I thought that he was “an impressive, truthful and careful witness”. I am not prepared to resile from that view, but in the light of his evidence at this hearing I consider that he gave evidence in the arbitration that was not in fact correct. I am quite satisfied that those on the ground at Vlissingen in June and July 2009 considered it very likely that further examination of the TPs would reveal similar welding defects and that, sooner or later, the issue of an NCR on the TPs was more or less inevitable. In those circumstances I do not think that at the time anyone addressed his mind to the question of whether, for example, NCR 006 covered the TPs in Shipment No 1 because, by one route or another, further NDT and repairs of the welds of those TPs was going to be necessary.

As I explained in the Liability Judgment, the issue as to whether or not the TPs were covered by the NCRs was one that did not surface until after the evidence had been given in the liability trial. When at this hearing Mr Ayres was shown an e-mail from Peter Reilly, of Airtricity (but effectively GGOWL), dated 19 June 2009, in which Mr Reilly referred to “waiting until all of the repairs have been done on a MP or TP and then releasing it to GGOWL”, Mr Ayres said that he thought Mr Reilly had “made a mistake on the “or TP” side of things” (Day 9/156).

I consider that Mr Ayres is correct in saying that NCRs 006, 008 and 009 were, on their face, confined to the MPs. He may also be correct in saying that GGOWL never gave an unambiguous instruction for further testing and repair of the TPs but, as I have just explained (and discuss further in the section dealing with the TPs below), it seems to have been taken for granted on the ground that no distinction was to be drawn between the MPs and the TPs. Further, whether it was entitled to or not, GGOWL appears to have made it clear that it would not lift the NCRs until the TPs, as well as the MPs, had been similarly tested and repaired. In these circumstances, Fluor, rightly or wrongly, assumed that the NCRs covered the TPs as well as the MPs, or were to be treated as doing so, and acted accordingly. Thus its case in the arbitration was that the NCRs did extend to the TPs and the Tribunal agreed with it.

I think that Mr Ayres believed at the time that the evidence he was giving in the arbitration was true but that now, having had the opportunity to conduct a more considered appraisal of the contemporaneous documents, he has reached the view that the NCRs did not extend to the TPs and that GGOWL never gave an instruction that they should be treated as doing so. He was, properly and understandably, embarrassed by and apologetic for this change of position. Part of his difficulty, which was understandable, was that by the time he prepared his evidence for the liability trial, the events in question had taken place seven years earlier and so his recollection of those events was substantially coloured by what he saw in the contemporaneous documents that were provided to him.

Mr Hans Ho also gave evidence at the liability trial. He was Fluor’s project manager in Shanghai with responsibility for overseeing the fabrication of the MPs and TPs.

He accepted that when the final account was being negotiated there was no suggestion that there might be a claim against ZPMC in relation to the costs of repairing TPs. He was also questioned in some detail about the claim for additional NDT costs in Shanghai. Essentially, what was put to him by Mr O’Sullivan was that the claim had been prepared by simply taking the NDT costs actually incurred by Fluor and subtracting the amount that it expected to incur by way of NDT costs. The point behind this was that it was said that Mr Ho’s approach overlooked the fact that the use of D scans should have been part of the original scope of the NDT work and therefore should not have been included as part of the extra costs claimed because it was work that would have been carried out in any event if the contract had been properly performed.

Mr Ho’s answer to this was a simple one, namely that D scans and E scans take the same time and so the substitution of one for the other would make no difference. What caused the increase in expenditure, he said, was the cost of the extra manpower required to carry out 100% NDT, as against a 2-3% spot check which was what he said was anticipated originally. This seems to me to be a satisfactory explanation and I accept it.

Mr Feng Gao also gave evidence at the liability hearing. His evidence was relatively uncontroversial.

Mr Chen Bin also gave evidence at the liability hearing. He dealt, amongst other things, with the repair of TP 31 in the latter part of September 2009 at Vlissingen. He said that it was not possible for other subcontractors to work on the TP whilst weld repairs were being carried out. He said that it was his understanding, based on conversations with employees of Fluor at Vlissingen, that the TPs had to be investigated and repaired as well as the MPs. The only difference was one of timing.

The court heard from four experts. Mr Robert McKibbin and Mr Anthony Morgan were the delay experts, and Mr Stephen Ross and Dr Franco Mastrandrea were the quantum experts; in each case for Fluor and ZPMC, respectively. They are all good and highly experienced experts who were eminently well qualified for their roles in this case.

Usually, at this stage I would make some general observations about each of them. However, in this case I do not consider that would be either appropriate or helpful. That is for this reason. This litigation is fraught with difficulties in relation to both the formulation of the claim and issues of causation. These are primarily matters of law for the lawyers, and not for the experts. However, in approaching their task and preparing their reports the experts have inevitably been largely guided by the lawyers instructing them as to the approach that should be taken. That is not a criticism: it is simply a reflection of the features of this particular case.

In these circumstances it is difficult to know the extent to which the approach taken by any one of the experts is a reflection of his own view or is one that has been guided by input from the lawyers, or a combination of both. Although each of the parties has, to a greater or lesser extent, criticised the approach taken by the other side’s experts, those criticisms, if justified, cannot necessarily be laid at the door of the expert in question. In these circumstances I consider that it is best to refrain from making any general observations: I will deal with any particular points relating to the evidence of the experts as and when they arise in the course of this judgment.

The key events took place as follows:

For the purposes of this case, the parties have split 2009 into six time periods, as follows:

The MPs were to be driven into the seabed by a hammer mounted on a vessel borne crane. The hammer took the form of a cylindrical sleeve, about 2.8 m long, with heavy “shoulders” at its upper rim which took the force of the impact. As I understand the procedure, the sleeve was lowered so that it fitted just over the top of the MP and then released so its shoulders, acting as a hammer, struck the top rim of the pile. Thus there had to be a fairly tight tolerance for the external diameter of the top 2.8 m or so of the MP in order to ensure that it did not jam on the inside of the hammer sleeve. The specified diameter of the top of the MP was 5100 mm. Although the internal diameter of the hammer sleeve was originally specified as 5126 mm, it was in fact 5130 mm.

The tolerance initially specified for the diameter of the top of the MP was +3 mm and -9 mm, which meant that the diameter could not exceed 5103 mm at any point. However, the increase in the internal diameter of the sleeve meant that the tolerance for the diameter at the top of the pile could be increased by 4 mm to 5107 mm. In a Fluor internal e-mail dated 17 June 2009, summarising a conversation with Seaway Heavy Lifting (“SHL”), the owners of The Stanislav Yudin, the latter was recorded as saying that the hammer supplier had installed piles with a 20 mm clearance, with the result that there should not be a problem with a pile diameter of 5110 mm and, further, that there might not be a problem with diameters up to 5114 mm.

In a later internal e-mail dated 25 June 2009, it was said that Fluor would have to achieve a diameter of less than 5110 mm “and if possible 5107 mm”.

It has been suggested by ZPMC that if the diameter at any point exceeded 5107 mm, there was a risk that the sleeve might jam on, or even fuse to, the top of the MP – the latter event was described by Mr Summers, of Fluor, in an internal e-mail dated 19 June 2009 as one that would be “catastrophic”. Apart from the fact that, if this happened, it would have to be cut off, a potentially more serious problem was that the vessel carrying the hammer would become tied to the MP which could put it at risk if the weather turned rough. However, the documents do not suggest, and I do not consider, that there was any risk of this happening unless the pile diameter was well in excess of 5110 mm.

In a letter dated 17 June 2009 SHL required Fluor to

“issue Hold Harmless agreement to SHL in case MPs with the diameter beyond 5107 mm are delivered offshore.”

It then said:

“SHL will confirm whether or not the MPs can be installed safely based on final DC Reports.”

I assume that “DC” refers to “dimension check”. This does not suggest that SHL was not prepared to drive a pile that exceeded the stated tolerance under any circumstances, but that it would only do so subject to a dimension check and then at Fluor’s risk. If it were otherwise, the indemnity would extend only to the wasted costs of loading out the MP and taking it to the site - costs which I assume Fluor would have to bear in any event if it loaded out an MP that exceeded the tolerance agreed with SHL.

The agreement that was subsequently made between Fluor and SHL contained the following provisions:

“. . . [SHL] shall in no event be obliged to install Oversized Mono Piles or to use its equipment beyond levels acceptable to [SHL]. If [SHL] initially accepts to install Oversized Mono Piles, this will not waive [SHL’s] right to stop the installation of Oversized Mono Piles at any time and at [SHL’s] sole discretion . . . ”

I assume that “Oversized Mono Piles” were MPs with a diameter in excess of 5107 mm. For present purposes, the relevance of this provision is that it shows, first, that the parties contemplated that piles with a diameter in excess of 5107 mm might be presented and accepted for installation; and, second, that the decision whether or not to install or to continue to install a particular oversized pile rested exclusively with SHL. Fluor, therefore, could not require SHL to install an out of tolerance pile. To the extent that Mr Eisman was to be taken as saying that Fluor would have been prepared to take the risk of installing out of tolerance piles, this could only have happened if SHL, too, was prepared to take that risk.

Although in evidence Mr Eisman described SHL as “difficult”, I conclude from this material that if an MP had a diameter which did not exceed 5110 mm at any point SHL would have been prepared to drive it using the original hammer, Fluor having provided the indemnity that SHL required. There would have been no good reason not to, because this would have provided the 20 mm clearance that was apparently acceptable to the hammer supplier. Further, in the light of what actually happened (as explained below), I think that it is likely that in particular instances SHL, might have been prepared to drive an MP, using the original hammer, with a maximum diameter up to about 5115 mm.

It is clear that many of the MPs in the first three shipments - and in Shipment No 2 in particular - did not meet, or appeared not to meet, a tolerance of 5100 mm + 7 mm/ - 9 mm. There were three aspects to this.

First, the weight of the steel in the MPs meant that when laid horizontally their profile deflected so that they became “squashed”: more technically, they became distorted such that their horizontal diameter became greater than the vertical diameter.

Second, there was a degree of “waviness” in the circumference at the top of the MP which meant that in the case of some MPs the diameter was in places out of tolerance.

The third aspect occurred with, I think, only one MP, and that was the fact that the longitudinal seam of the top can had not been properly welded so that the circumference at the top of the MP was greater than it should have been. This was repaired and it is not necessary to say any more about it.

I am satisfied that the first problem was an elastic deformation that put itself right once the MP was vertical - because the force of gravity would then be imposed along the line of its axis and not across its diameter. This was Mr Eisman’s view and I consider that he was correct. There was no credible evidence to the contrary.

The “waviness” in the circumference of the MP presented a more difficult problem. The first difficulty was that of measurement: the elastic deformation of the MPs when they were horizontal made it very difficult to measure accurately the extent of the “waviness”. As Mr Eisman put it: “the whole thing was pretty loose in its engineering reliability” (Day 8/97). There was nothing to contradict that evidence, and I accept it. The second difficulty was that it was not elastic deformation and required heat treatment and/or the fitting of a device called a stiffener ring in order to rectify it. Heat treatment had been used by ZPMC in Shanghai to remedy any out of roundness discovered during manufacture but, as the measurements taken at Vlissingen demonstrated, not always with sufficient success.

Out of roundness (or lack of ovality) was recognised as a problem from at least March 2009, when the fitting of a “jacking spider or collar” was suggested as a possible solution (see the e-mail from Mr Fuller to Mr Dekker dated 23 March 2009). By the end of May 2009, after Shipment No 1 had arrived, it was recorded that there were not enough MPs within tolerance for there to be three available to load out onto The Stanislav Yudin’s barge: as a result, the first load out – previously scheduled for 2 June 2009 - had to be postponed pending review of the out of roundness measurements (Fluor’s Site Weekly Report of 1 June 2009).

It is not clear from the evidence to what extent Fluor had carried out by this time any detailed investigation into the type of stiffening required to rectify out of roundness. However, two or three days later, NCR 006 was issued and brought the weld defect issues to the fore. I find that, quite understandably, at this point Fluor’s primary concern was the problem presented by the cracking in the welds and the need to carry out the revalidation of the NDT required by the NCR and the consequent remedial work.

It is quite obvious to me that the position in which Fluor found itself following the discovery of the extensive cracking in the welds was quite different from the position that it had been when Shipment No 1 arrived at Vlissingen in May 2009. During early June 2009 Fluor’s preoccupation, quite understandably, was the need to make - as a matter of urgency - the arrangements necessary to facilitate the required testing and repair of the welds. That is not to say that the out of roundness problem was forgotten: my impression from the evidence as a whole is that it was seen as an issue of secondary importance to that of the cracking in the welds. This is unsurprising - out of roundness did not affect the structural integrity of the MPs, only the method of installing them.

Faced with an out of roundness problem that was significantly more serious than that anticipated, (Footnote: ⁹ ) in early July 2009 Fluor ordered a second hammer with a slightly wider internal diameter in order to accommodate a greater degree of tolerance (+ 17 mm) in the diameter of the MP (“the first replacement hammer”). However, this hammer was not due for delivery until the end of July, and was not in fact delivered and fitted until 30 July 2009 (immediately prior to the installation of the first MP). This meant that if any MPs were to be installed before the arrival of that first replacement hammer, they would have to be driven using the original hammer.

If the MPs and TPs had not suffered from the extensive cracking that gave rise to the issue of the NCRs, then I consider that it is reasonable to assume that Fluor would have taken the following steps in relation to the out of roundness issue:

For the avoidance of any doubt, I should make it clear that I do not consider that the delay in taking these steps was the result of the issue of NCR 006 and the particular testing and consequent repairs that it required: on the contrary, it was the result of the discovery of the extensive cracking in the welds which, once discovered, had to be repaired unless and until appropriate expert advice had been received that such repairs were not necessary. It would, for example, have made no difference in my view to the degree of priority given to the out of roundness problem if NCR 006 had not been issued on 3/4 June 2009, but, say, two or three weeks later. The out of roundness problem had nothing whatever to do with the state of the welds, but the latter affected the degree of priority that was given to it.

The design solution for the stiffener ring was not produced until mid-July 2009. In the absence of the problem with the cracking in the welds, I consider that this would probably have been requested and produced about a week or so earlier than actually occurred. Similarly, given the serious financial consequences of any delay in installing the MPs and the fact that over a third of the MPs in Shipment No 1 appeared to have a maximum diameter in excess of 5125 mm, I consider that Fluor would have placed an order for a larger hammer about two weeks earlier than it in fact did. I am therefore prepared to assume that, absent the extensive cracking in the welds, the first replacement hammer would have been available about by mid-July 2009.

In relation to heat straightening, I would have expected Fluor to carry out trials. I bear in mind the fact that at this point in time there were no welders from ZPMC on site at Vlissingen, and so ZPMC’s prior experience of heat straightening would have been of limited value. Nevertheless, I would have expected Fluor to have found, by the beginning of July 2009, a method of heat straightening that could be used to remedy the piles that were only slightly out of tolerance - by which I mean with a maximum diameter of up to 5115 mm - so as to bring them within a tolerance of about + 10 mm (i.e. to a maximum diameter of about 5110 mm). For the reasons I have already given, I find that this would have been sufficient to enable the MPs to be installed in the seabed using the original hammer: indeed, in the light of what happened I consider that an MP with a maximum diameter 3-4 mm greater would probably have been installed also.

At this point I should digress to consider the effect of the out of roundness of the MPs and its contractual significance in a little more detail. I consider that the delivery of MPs whose roundness was out of tolerance was probably a breach of contract, but Fluor has not made any allegation of such a breach in these proceedings - for the very good reason that any disputes about out of roundness were resolved when ZPMC’s final account was agreed. Accordingly, to the extent that any out of roundness problems were the result of breaches of contract by ZPMC, Fluor’s cause (or causes) of action in respect of those breaches has (or have) been compromised. So, when considering what would have happened to Shipment No 2 if there had been no weld defects, any loss or delay caused by out of roundness has to be absorbed by Fluor and therefore taken into account - by which I mean that, in terms of out of roundness, when considering the hypothetical scenario the MPs in Shipment Nos 1 and 2 must be taken to have been delivered in the condition in which they were actually delivered.

I propose to approach the question of what would have happened to the MPs in Shipment No 2 - if the welding had been properly performed (the counterfactual scenario) - in two stages. The first is to consider how many MPs from Shipment No 2 would have been available (in terms of tolerance) for installation by 29 July 2009. The second is to consider, in the light of this and other factors affecting installation, such as weather, how many MPs Fluor could actually have installed by 29 July 2009.

However, before I do so I should deal shortly with one point raised by Fluor, namely that it “did in fact proceed with installation and take the risk of the hammer sticking once piles started to become available”, referring, as I understood it, to the original hammer: see paragraph 119(e)(v), at page 80 (when read in the context of the paragraph as a whole). If this is a reference to the original hammer, then it is not correct: as I have already mentioned, the first MP to be installed (IGI 05) was driven on 31 July 2009 - immediately after the first replacement hammer had been fitted to The Stanislav Yudin. (Footnote: ¹⁰ )

In an e-mail dated 17 July 2009 one of Fluor’s project engineers reported on measurements on the top of the MPs from Shipment No 2. He said this:

“IGE 02 - within tolerances

IGH04 - slightly outside tolerances but considered acceptable by Fluor

IGI 06 - slightly outside tolerances but considered acceptable by Fluor”

The tolerances against which these MPs were being measured were those appropriate to the initial hammer, that is to say the revised tolerances of + 7 mm and - 9 mm. according to this e-mail, the two “slightly outside” tolerance MPs exceeded the specified diameter by + 8 mm and + 9 mm, respectively, and it was noted that for these the need for a stiffener ring and/or heat treatment was “possible”. The diameter of each of these two piles, therefore, did not exceed 5110 mm at any point. (Footnote: ¹¹ ) The remaining 11 piles were described as “not OK” but no dimensions were given. A note to the table stated that the remaining MPs had been “sorted based on best to worst, as reported by ZPMC”: this suggests that some were probably only slightly out of tolerance but others significantly so. This was in fact the case.

Of the three MPs listed in the previous paragraph, IGH 04 and IGI 06 were unloaded on 3 July 2009, together with two others, and IGE 02 was unloaded on 5 July 2009. It seems reasonable to assume that this sequence of loading out was not affected by any possible presence of defects in the welds in Shipment No 2, so that it is the sequence that would have been followed in any event.

In the light of the conclusions reached above, I consider that it is reasonable to assume that any out of tolerance MPs that did not have a maximum diameter that exceeded 5110 mm could have been installed using the original hammer subject to Fluor giving an appropriate indemnity to SHL, which I find it would have been prepared to do. But even if I am wrong about this, I find that they could have been remedied using heat treatment alone so as to bring them within the required tolerance. However, I find also that in some cases SHL would have been prepared to drive an MP with a maximum diameter of up to 5115 mm.

The next question is what type of remedial work would have been required to the remaining 11 MPs of Shipment No 2 in order to bring the out of roundness to a tolerance of less than +/- 10 mm - on the assumption that on being given the appropriate indemnity SHL would have been prepared to install any MP the diameter of which at any point did not exceed 5110 mm.

ZPMC submits that in July 2009 Fluor adopted a policy of fitting stiffener rings to every MP whether required or not (as reflected by a communication from Mr Ayres to Mr Pashley on about 23 July 2009), and so this is what it should be taken to have done if there had been no defects in the welding of the MPs of Shipment No 2. Whilst I accept that this was or may have been Fluor’s approach at the time, I do not consider that it would have necessarily adopted the same approach if the MPs of Shipment No 2 had been free of weld defects requiring repair so that the only factor affecting installation was out of roundness.

Fluor submits that it is unfair to suggest that if the welding had been sound Fluor “would have approached the question of precisely how to deal with the out of roundness with the same urgency and dispatch (sic) as it did in a situation where that question was not one on the critical path” (Closing Note, paragraph 123).

I consider that Fluor is right. Common sense and experience suggests that if the only issue affecting the suitability of MPs for installation was out of roundness, Fluor would have devoted all its energies to the resolution of that problem. As it was, in early June 2009 Fluor - having been hit with NCR 006 - rapidly became aware of the full implications of the welding defects - which included the assumption that Shipment No 2 would be no better than Shipment No 1 (see the first witness statement of Mr Ayres, paragraph 11.3). This knowledge would inevitably have led Fluor to concentrate its resources on resolving the problems with the welding.

However, I accept ZPMC’s point that the stiffener ring solution would have required a proper engineering evaluation and I do not consider that it would be unfair to Fluor to assume that that would not have been achieved until about the end of the first week in July 2009 (Footnote: ¹² ) (ZPMC submitted that the court should not assume that a robust solution to the out of roundness problem would have been found until 14 July 2009 - but in different circumstances I consider that it would have been of little earlier). In the meantime, I would have expected Fluor to have taken urgent steps to develop a satisfactory method of dealing with MPs that were only slightly out of tolerance by heat straightening.

Having done so, I consider that Fluor would have attempted to install every MP that it could have persuaded SHL to accept - that is, using the original hammer all MPs whose maximum diameter at any one point did not exceed 5110 mm and, in some cases MPs with a maximum diameter of up to about 5115 mm.

I reach this conclusion in part by reference to what actually happened. The following MPs were installed without having had any treatment (using Mr McKibbin’s batch numbers):

I am not surprised that SHL was prepared to install an MP with a maximum diameter of 5122 mm: that corresponds broadly with Fluor’s assessment that it might be possible to drive an MP with a maximum diameter of 5114 mm using the original hammer (ie. 4 mm in excess of the stated tolerance). This suggests either that SHL was more relaxed about the tolerance than might have been expected or that the measurements are simply unreliable (as Mr Eisman suggested). However, I am a little puzzled by the fact that an MP with a maximum diameter of 5129 mm was driven, even with the first replacement hammer. I suspect, although there is no evidence about it, that SHL’s dimension check may have shown a lower figure. However, even if its true maximum diameter was less than 5129 mm, it seems likely that it was still well in excess of 5117 mm.

For the reasons which follow I find that, if there had been no defects in the welding of the MPs in Shipment No 2 (but that both Shipment Nos 1 and 2 suffered from out of roundness problems), out of the MPs in Shipment No 2 Fluor would have had available for installation by 29 July 2009 at least six, and probably eight – if (as I think likely) the first replacement hammer had been delivered by mid July 2009.

My reason for this conclusion is as follows. I find that, on the basis of the dimensions set out in the screenshot attached to the e-mail from ZPMC’s counsel dated 4 August 2017, (Footnote: ¹³ ) three MPs in Shipment No 2 had a maximum diameter of 5110 mm or less (and so could have been driven using the original hammer without any heat treatment), and a further two had a maximum diameter of 5115 mm or less - in relation to the latter, I consider it likely that SHL might have been prepared to drive them without any treatment or, alternatively, that they were brought within tolerance using heat treatment. Of the remainder, I consider that at least one (probably IGG 03) could have had a stiffener ring fitted in time to install it by 29 July.

Of the MPs with a maximum diameter of 5110 mm or less, one was in batch 1 and another was in batch 3. Of those with a diameter of 5115 mm or less, one was in batch 1 and the other in batch 2. That makes four MPs in all in the first three batches.

Of the remaining eight MPs in Shipment No 2, five of them had a maximum diameter of less than 5122 mm and so I consider that, with appropriate heat treatment, their maximum diameters could have been brought down to 5117 mm so that they could therefore have been installed using the first replacement hammer when it arrived. Two of these were in batch 1 and two in batch 2, so they could therefore have been installed by 29 July 2009. This produces a total of eight MPs in the first three batches suitable for installation during July 2009, together with IGG 03 which might have been fitted with a stiffener ring in time.

In partial response to this question, my finding in the previous paragraph imposes a ceiling of eight, possibly nine, MPs. What I must now consider is whether matters such as vessel availability, rate of installation, weather conditions and so on, would have reduced the number of MPs that could have been installed by 29 July 2009 still further. At this stage I will assume also that there was no delay to the installation of the MPs in Shipment No 2 in consequence of any delays to MPs in Shipment No 1. However, I will have to revisit that assumption later in this judgment.

The two delay experts have very helpfully prepared a number of scenarios as to what might have happened in July 2009 if the piles in Shipment No 2 had been free of weld defects. Mr McKibbin has prepared three scenarios and Mr Morgan no less than 21. Relevant parameters taken into account by both experts include matters such as: downtime owing to weather, the rate at which MPs could be installed, the length of time required for out of roundness correction (where required), to what extent out of roundness correction could be carried out in parallel with the base scope works, whether or not the installation of the jacket would have been postponed and the precise impact of the redeployment of The Stanislav Yudin on other work (a Gaz de France project, for which SHL on giving proper notice was entitled to deploy the vessel during the contract period).

For the purposes of his Scenario 3 Mr McKibbin has assumed that out of roundness issues were resolved using the method (either heat treatment or stiffener ring) that was actually used for each MP and I consider that this is the most realistic of his three scenarios. Of the 10 MPs in the first three batches of Shipment No 2, two had stiffener rings fitted. As I have already mentioned, ZPMC submitted that it should be assumed that stiffener rings would have been fitted to every MP. Whilst I do not think that Mr McKibbin’s assumption is entirely satisfactory, because the MPs were in fact driven with the larger hammer, I consider that it is to be preferred to ZPMC’s suggestion. As I have already mentioned, in the light of what actually happened it seems that SHL may have been prepared to exceed the acceptable tolerance by a further 4-5 mm. It seems to me that any alternative approach would be entirely arbitrary.

The delay experts have agreed that the planned approach was that each batch of MPs of Shipment No 2 could have had its base scope work completed and have been loaded out to The Stanislav Yudin’s barge, with sea fastenings attached, 11 days after the Shipment No 2 MPs had been loaded out to the quayside following their arrival at Vlissingen. Shipment No 2 arrived at Vlissingen at 21:30 hours on 29 June 2009, and the first four MPs were unloaded on 3 July 2009. These included two of the three MPs mentioned at paragraph 115 above. I take it that the delay in unloading the first MPs was, as Mr Morgan has noted, because the TPs had to be unloaded first, being distributed around the outside of the barge.

In an e-mail dated 4 August 2009 Mr Eisman noted that the current production level for fitting stiffener rings was “one MP every four days” and he accepted in cross examination that where MPs were not within the required tolerance, they would have had to be fitted with stiffener rings (Day 7/194). However, a further problem arose in relation to the design of the stiffener rings with the result that the fitting of stiffener rings was subsequently put on hold pending a new design solution. Mr Eisman agreed that this was so, but then said that the MPs that were not “so badly out of tolerance” could be rectified by heat treatment without the need for a stiffener ring - an observation that, although apparently in conflict with the evidence that he had given the previous week, I did not understand to be challenged (see Day 8/27) and I accept it. Heat treatment took less time than fitting a stiffener ring: in an e-mail from Paul Pashley dated 24 July 2009 he recorded having been told by Mr Ayres that it took 48 hours. This is the period taken by Mr McKibbin in his Scenario 3, and I see no reason not to accept it.

Save for the cases where an MP required the fitting of a stiffener ring, I consider that the planned approach was in principle achievable. Mr McKibbin has assumed, in his Scenario 3, that one of the MPs in the first batch, IGF 02, which required a stiffener ring, could have had that stiffener ring fitted between 5 and 10 July 2009. For the reason that I have already given, I consider that that would probably not have been possible because the design solution would not have been available by 5 July.

Mr McKibbin has concluded that, by 29 July 2009, eight MPs could have been installed in the seabed and a further three MPs could have been loaded out onto The Stanislav Yudin’s barge.

Mr Morgan’s Scenario 3 is the one most favourable to Fluor, in that he indicates that six MPs could have been installed in the seabed by 29 July 2009. However, Mr Morgan assumes that all MPs in Shipment No 2, save for the first batch of three, would have required the fitting of a stiffener ring (taking 4 days) and that only one stiffener ring could be fitted at a time. The effect of this, for batches 2 and 3, is that the time taken to complete the base scope work and out of roundness correction to each MP is pushed back by a further four days each time. Since I have rejected the suggestion that stiffener rings had to be fitted to all MPs that were out of tolerance, I regard this scenario as over conservative (that is, in favour of ZPMC).

However, I do not consider that the original planned intent of an 11 day period from load out from the incoming vessel to load in to The Stanislav Yudin’s barge would have been achieved in every case. This is for two reasons. First, the lack of a design solution for the stiffener ring until the second week of July 2009 meant that a stiffener ring could not have been fitted to any MPs in the first, and possibly second, batch. Second, the one MP in Shipment No 2 which was within tolerance, IGE 02, was one of the last to be unloaded on 5 July 2009, and so no base scope no work could have started on that MP until 6 July 2009.

In my view all these considerations preclude the adoption of any one of the many scenarios produced by the delay experts: thus I am not prepared to find that any one of those scenarios fairly reflects what would have happened if there had been no weld defects. However, I find that the scenarios are helpful because they do give an indication of the likely range of reasonably possible outcomes – being effectively from three to eight MPs installed by 29 July 2009. (Footnote: ¹⁴ )

I consider that the base scope work to MPs IGH 04 and IGI 06, which were loaded out on 3 July 2009, could have started on 4 July 2009, along with the base scope work to the other two MPs which had been loaded out at the same time. This is because, as Mr McKibbin notes, the MPs were offloaded directly onto Self-Propelled Modular Transporters (“SPMTs”), rather than offloaded directly onto the quay. He has assumed, in my view realistically, that once offloaded they would have been taken directly to the laydown area. Accordingly, I find that work could have started on a batch of MPs on the day following that on which they were offloaded from the incoming vessel.

If there had been no weld defects in the MPs in Shipment No 2, I find that by about 6 July 2009 (Footnote: ¹⁵ ) Fluor would have been in a position to begin heat treatment to the MPs of Shipment No 2 that were not, to use Mr Eisman’s words, “badly out of tolerance”. Further, I have no hesitation in concluding that Fluor would not have installed the jacket to the offshore substation when it did, but would instead have rescheduled that operation.

Fluor makes the bold submission that, for commercial reasons, Fluor would have proceeded to install the MPs regardless of the out of roundness concerns. Whilst I quite understand the financial imperatives, it was simply not within Fluor’s power to do this. Ultimately, the decision whether or not to install an MP was one for SHL (as the terms of the agreement between Fluor and SHL make clear). (Footnote: ¹⁶ ) I have already concluded that it is most unlikely that using the original hammer SHL would have been prepared to install an MP whose diameter exceeded 5110 mm by anything more than about 4-5 mm. In reality Fluor would have had to perform out of roundness correction to any MP whose diameter exceeded 5110 mm: it was not a question of choice. (Footnote: ¹⁷ )

Subject to any effect of delays to Shipment No 1, I accept Mr McKibbin’s opinion that three MPs from the first batch of Shipment No 2 could have been installed between 11 and 14 July 2009 (see paragraphs 1.3.7 and 1.3.8 of his Note on Counterfactual dated 10 July 2017). Mr McKibbin postulates that the next batch of three MPs would have been loaded on board The Stanislav Yudin late on 14 July 2009, and that at least one of these MPs would have been installed on 15/16 July 2009. The latter part of 16 July was interrupted by a spell of adverse weather from about 16:00 hours which persisted until the morning of 20 July 2009. Mr McKibbin assumes that from 09:10 on 20 July installation of the MPs would have recommenced and that the remaining two MPs from the second batch would have been installed within three days (on the basis of 1.5 days per MP), so that installation would have been complete by the morning of 23 July 2009.

On 24 July 2009 Fluor was notified that The Stanislav Yudin would be required for the Gaz de France works on 25 July, so that it would have to leave the Greater Gabbard site early that day (at about 05:00 hours). There is an issue as to whether or not a further two MPs could have been installed between 23 July and the time, early on 25 July, when The Stanislav Yudin was due to sail away to the Gaz de France site. Mr McKibbin postulates that The Stanislav Yudin’s barge could have left with two MPs at about 09:10 hours on 23 July and been transferred to The Stanislav Yudin later that day. The sailing time from Vlissingen to the site was 12-15 hours, so the barge could not have arrived much before midnight on 23 July.

Mr McKibbin has assumed that the next MP, which was to be installed in the site adjacent to the last MP, could have been installed during the early hours of 24 July 2009, leaving slightly less than 24 hours within which to install the last MP. That MP also was to be installed at a location which was very close to that of the previous MP, and Mr McKibbin’s conclusion (at paragraph 1.3.25) was that:

“Given the above, it may have been possible to complete installation of the MP on or only shortly after 24 July 2009.”

Mr Morgan’s Scenario 3, which I regard as having the most realistic assumptions of his 21 scenarios, shows that by the end of 29 July 2009 eight MPs would have been ready for installation, and that six would actually have been installed. Mr Morgan assumes that 9.5 days were available for transport of the MPs to the site and for installation, whereas Mr McKibbin considers that there were about 10.5 days. On Mr Morgan’s approach, only six MPs can be installed if a period of 1.5 days per MP is used throughout. On Mr McKibbin’s approach, it would be seven MPs using the same installation rate. It appears also that Mr Morgan has used discrete blocks of one day for the various activities (for example, loading out the barge and transport to site is treated as an activity lasting a whole day - even though the same time from Vlissingen to the site was not more than 15 hours). This is one of the factors that permits Mr McKibbin to assume that seven MPs can be installed during the period from 10-24 July, whereas Mr Morgan considers that only six can be installed during this period. Another difference is that Mr McKibbin assumes that in some cases the full 1.5 day period would not have been required. On these aspects, I consider that Mr McKibbin’s scenario is plausible and realistic and I accept it.

In the context of the time required for the installation of an MP, the other principal difference between the two experts is whether or not a further MP could have been installed between about 05:00 hours on 24 July and 05:00 hours on 25 July 2009 (when The Stanislav Yudin had to sail for the Gaz de France site). Here I prefer the view of Mr Morgan. Whilst Mr McKibbin may well be right that it would have been “possible” for a further MP to be installed during this period, that conclusion does not in my view provide the answer. I consider the question is not whether or not it would have been possible (assuming that even that is good enough), but whether or not SHL, the owners of The Stanislav Yudin, would have been prepared to take the risk of the operation taking longer than 24 hours and thereby delaying the departure of The Stanislav Yudin for the Gaz de France site.

Leaving aside the concern that a period of 24 hours would be too tight, it must be remembered that SHL would be looking at the problem prospectively – and that would involve taking a chance with the weather: if an unexpected storm blew up, the operation could have gone badly wrong. I consider it unlikely that SHL would have been prepared to take the risk of installing a final MP if it had only about 24 hours in which to do it. I doubt very much that it would have been willing to tolerate any risk of delay to the departure for the Gaz de France site.

For these reasons, therefore, looking at Shipment No 2 alone, I conclude that of the eight available MPs, only seven would have been installed by 29 July 2009 if there had been no problems with the welding to the MPs of that shipment.

When considering the damages to which Fluor is entitled one must start from the position that there had been no relevant breaches of contract by ZPMC. So, subject to the waiver, the court has to consider the position if the contract had been properly performed and then consider the position as it actually turned out as a result of the breaches.

This means that it is necessary to ask whether or not, absent the welding defects, the installation of Shipment No 1 would have proceeded as planned or, at least, proceeded in a manner that did not adversely affect the installation of MPs in Shipment No 2.

It is common ground that The Stanislav Yudin did not arrive at Vlissingen until 16 June 2009, some two weeks later than planned. There is a dispute as to when it would have been ready to install MPs because the certificate that was required to be issued by the Marine Warranty Surveyor (“MWS”) was not issued until 23 June 2009. It will be recalled that at the time when The Stanislav Yudin arrived at Vlissingen NCR 006 had already been issued so there was no question of installing any MPs at that stage. Fluor submits, quite understandably, that in these circumstances there was no particular urgency for the certificate from the MWS, since there were no MPs to install, and that the court should infer that it would have been issued sooner if it was needed. Fluor contends the certificate would have been issued almost immediately after The Stanislav Yudin arrived at Vlissingen and so it would have been ready to start work on 17 June 2009.

ZPMC submits that this is unrealistic. It submits that there were various matters that had to be attended to before the certificate could be issued and that, even during the survey on 23 June 2009, there was insufficient time to inspect the tug and that, later, it was discovered that further modifications were required to the barge. In my view, there is some force in these points and I consider that The Stanislav Yudin would not have received the certificate until about midday on 18 June 2009, so that it could not proceed to installation until that afternoon.

In relation to Shipment No 2, Mr McKibbin has assumed (and I have accepted) that The Stanislav Yudin could have left Vlissingen to install the first batch of MPs from Shipment No 2 on 10 July 2009. Fluor submits that, assuming that The Stanislav Yudin would have been available to start work on 17 June 2009, there would have been 23 days in which to install the MPs from Shipment No 1 before 10 July 2009. As a matter of arithmetic, this is obviously correct.

Fluor submits also that the BOP Contract required Fluor’s programme to accommodate an identified number of days per month by way of an allowance for adverse weather. In the months of June and July 2009 there were to be six days in each month. The duration in the 31 May 2009 programme for the installation of the MPs in Shipment No 1 was 23 days, so Fluor submits that this must have included the 6 days (or, on a pro rata basis, perhaps slightly less). ZPMC, by contrast, contends that the programmed installation periods did not include any time for adverse weather.

The experts have worked on the basis that the average installation time for an MP was 1.5 days (ignoring any additional time for the first few MPs that might have been referable to any learning curve). For a shipment of 14 MPs this is 21 days. Accordingly, I think it unlikely that a programmed period of 23 days included an allowance of even 4 or 5 days for adverse weather: it was 2 days at most.

If, as I have concluded, the certificate of the MWS would not have been issued until midday on 18 June 2009, The Stanislav Yudin would have had 21.5 days in which to install the Shipment No 1 MPs before 10 July 2009. This left almost no room for contingencies or any learning curve that might have extended the installation period for the first few MPs. In these circumstances, I consider it improbable that Fluor would have been able to install all the MPs from Shipment No 1 by 9 July 2009.

But in my view, the dominant factor affecting the potential installation of the MPs in Shipment No 1 in June/July 2009 was the out of roundness problem. The recorded maximum diameters for the MPs in Shipment No 1 were as follows:

For the reasons that I have already given, I consider that SHL would probably have installed the 6 MPs with a maximum diameter ≤ 5113 mm using the original hammer without any need for heat straightening. I have already concluded that, if there had not been the extensive welding defects, by the beginning of July 2009 Fluor would have found a heat straightening method that enabled the maximum diameter of the MPs to be brought within acceptable limits (save for cases where the degree of out of roundness was very significant).

I consider that it is reasonably possible that 4 or 5 of the MPs with a maximum diameter greater than 5116 mm could have been heat straightened so that they could have been installed on or before 9 July 2009. Conversely, therefore, as a matter of probability I consider that three of the MPs in that shipment would not have been ready for installation by 9 July 2009.

This means that Fluor could only have proceeded with the installation of the MPs in Shipment No 2 on the basis that I have already described if three MPs from Shipment No 1 were left on one side for installation at a later date. Thus in the counterfactual scenario in which it is assumed that there were no breaches of contract in relation to the welding, Fluor could have installed only four additional MPs (seven from Shipment No 2, less three uninstalled from Shipment No 1). That, in my judgment, represents the starting point for the assessment of the damages attributable to the loss of opportunity to install MPs during the period 29 June to 29 July 2009 resulting from the breaches of contract in relation to the welding.

However, in addition to the wasted costs of paying The Stanislav Yudin to lie idle in Vlissingen during that month, the inability to install an additional four MPs during June/July 2009 had another consequence. That was the risk of additional weather downtime (at Fluor’s cost) if the installation was put off to the winter months. Fluor’s quantum expert, Mr Stephen Ross, calculated that for every day of installation time lost during July and August 2009 Fluor would have to pay an additional 1.23 days in respect of time lost through bad weather if the installation had to be carried out between October 2009 and January 2010.

Dr Mastrandrea has produced a slightly lower figure of 1.02, but this takes into account adverse weather in July 2009. However, I agree with Fluor that this has already been taken into account when considering the number of MPs that could have been installed during July 2009. In short, I consider that ZPMC’s approach involves an element of double counting. Of the two, therefore, I prefer Mr Ross’s figure of 1.23 days.

In an earlier note prepared during the course of the hearing, Mr Morgan carried out a different exercise. He took the date of the installation of the last MP (being either the final MP for Phase 1, 28 December 2009, or the 69^th MP to be installed in during the first season, which was in fact part of Phase 2, 31 January 2010). If I had to take one of these, I would prefer the first approach, namely to count back from the installation of the final MP for Phase 1. On this basis, Mr Morgan concluded that 12 days of the time of The Stanislav Yudin would have been saved. This exercise, as Mr Morgan pointed out, took into account actual weather conditions as they impacted the as built schedule.

However, I consider that to adopt either of the approaches suggested by Mr Morgan would be to invest the hypothetical scenario with a degree of precision that is not justified in the circumstances - indeed, what would have happened in the hypothetical scenario would almost certainly have been different from the actual as built sequence of events. Whilst I accept that the court has to seek to adopt as precise an approach as is practicable when making an assessment of how many MPs could have been installed during June/July 2009 if the MPs in Shipment Nos 1 and 2 had been sound, I consider that overall the cost to Fluor of the loss of the opportunity must be considered on a broader basis (however, as it happens, in terms of the additional time caused as a result of having to install the MPs later in the year, the approach of Mr Ross and the first approach of Mr Morgan differ by very little). I therefore consider that Mr Ross’s approach is to be preferred and so I adopt his figure of an additional 1.23 days for every day that was lost in July 2009. On the basis of the loss of the opportunity to install four MPs, I consider that 6 days were lost during July 2009, so this produces approximately an additional 7.5 days of vessel time, making 13.5 days all told.

Mr McKibbin has also concluded that, in addition to the MPs installed in the seabed, a further three MPs would have had the base scope and correction works carried out and would have been loaded out onto the barge on 29 July 2009. Although, as I have explained, I do not accept Mr McKibbin’s sequence of events in his Scenario 3 in every respect, I am prepared to accept that at least one MP would have been ready for loading out onto The Stanislav Yudin’s barge before receipt of NCR 008 on 29 July 2009.

However, it seems to me to be an inescapable conclusion that this MP (or Mr McKibbin’s three MPs) would have fallen within the scope of NCR 008 and so Fluor would have had to unload it (or them) and carry out the testing and repair required by that NCR. The cost of unloading those MPs would be a cost incurred as a result of the NCR and therefore waived under the terms of the waiver letter. By contrast, if there had been no weld defects in those MPs the cost of loading them onto the barge would not have been wasted. Accordingly, instead of lying idle on that day the barge could have been properly used. Accordingly, I find that Fluor is also entitled to recover an additional 1 day of vessel time as a result of this, making 14.5 days all told. The daily rate for the The Stanislav Yudin has been agreed at €316,318.

Fluor also submits that, since the installation of these three MPs would have taken a further 4.5 days of the time of The Stanislav Yudin, it is entitled to recover this also as a wasted cost. I disagree. The reason why The Stanislav Yudin could not be employed in installing these MPs was the result of the issue of NCR 008. Thus the cost or loss has been waived.

The period of hire of The Javelin - the specialist vessel which was to be used for installing TPs - expired on 19 September 2009 and was not extended: an offer made by Jumbo (in a letter of 7 August 2009) to extend The Javelin’s operations to 15 November 2009 was not taken up (by Fluor’s letter dated 13 August 2009).

Mr Hans Dekker, who gave evidence at the liability hearing, was a key member of Fluor’s senior management. He was instructed to take direct control of the project in July 2009, with the title of “Project Director”, and, in particular, to deal with NCR 006 and the issues that arose out of it. In his witness statement made for the arbitration, dated 27 February 2012, Mr Dekker said this (at paragraph 111):

“GGOWL’s issuance of NCRs 006, 008 and 009, and its demands that Fluor change its NDT procedure and retest and repair welds previously tested, had a profound and devastating impact on the Project. Not only did all offshore installation activities relating to the foundation works come to a halt, but even when we ultimately made the decision to cease the retesting and repair effort in Vlissingen, the construction schedule was destroyed. Once we decided to restart the construction phase, we had to remobilize contractors (such as SEPAM) and procure vessels to perform the work, or modify existing vessel contracts to account for the delays. Because offshore work can generally only be performed during certain times of the year due to adverse weather in the North Sea, much of the work that we had planned to perform during the 2009 construction season could not even begin post-October 2009 (e.g. TP installation), or, if it could begin then, it was at a much higher cost to Fluor (e.g. MP installation).”

By way of amplification of Mr Dekker’s observations in the last sentence of that passage, I should explain that The Javelin was not a jack up vessel, which achieved stability by extending legs down to the sea bed, but was instead fitted with a dynamic positioning system (“DPS”) involving a combination of propellers and thrusters. These were constantly adjusted by a computer in order to keep the vessel in the same position. However, vessels which use a DPS cannot tolerate bad weather as well as a jack up vessel. Accordingly, there were drawbacks in employing The Javelin to install TPs during winter months. A further difficulty, which was peculiar to the installation of TPs, was the difficulty of carrying out grouting operations in cold weather. (Footnote: ¹⁸ ) So once The Javelin’s contract period came to an end in September 2009, there was effectively no prospect of using her again to install TPs until March/April 2010.

What Mr Dekker said in that paragraph echoed what he had said in paragraph 25.1 of his witness statement dated 30 March 2012, where he referred to the NCRs as having brought the construction phase of the project “to a complete halt”. After citing that paragraph, I said at paragraph 422 of the Liability Judgment:

“Again, this is an example of evidence that bears the ring of firm recollection, not of a reconstruction of events based on carefully selected documents prepared by others. I therefore conclude that what Mr Dekker said in his witness statements in the arbitration reflected Fluor’s view of the position following the issue of the three NCRs. I do not accept the evidence to the contrary that Mr Dekker gave at the trial.”

Mr Eisman, in a witness statement prepared for the arbitration in February 2012, said, at paragraph 26:

“At that point (end of July 2009), there was still hope that, when Shipment 3 arrived in Vlissingen, it would not be subjected to an NCR and there would be a supply of TPs on which SEPAM could begin the fit-out works. Ultimately, that proved not to be the case, because shortly after Shipment 3 arrived on August 1st, GGOWL issued NCR 009, which required re-testing and repair of the Shipment 3 MPs and TPs as well. At that point, we were left with little choice but to demobilize SEPAM, except for a skeleton crew, which we utilized to perform miscellaneous support works for the NDT remediation efforts, such as monitoring of temporary power and lighting.”

Mr Helmonds, a consultant employed by Fluor who dealt with some of the contractual matters in relation to The Javelin on the instructions of Mr Hovermale, (Footnote: ¹⁹ ) made a witness statement in the arbitration dated 24 February 2012 in which he said this (omitting references):

“18.

We proposed to Jumbo that they could take the period from the contractual early start date of 1st July through early August (10 August 2009 mentioned by Jumbo), to use the Javelin vessel for other work . . . The rationale behind this proposal was to eliminate the certain costs for standby time in July and early August, while securing the TP installation vessel in September when we were expecting to have sufficient TPs available and MPs installed . . .

19.

When this proposal was developed and negotiated with Jumbo, we still anticipated that TPs would be available, and a sufficient number of MPs installed, to start TP installation on 11 August 2009, the date Jumbo would return from their “away period”. This new start date was based on the then anticipated MP, j-tube and inter-array cable installation commencement dates. Also, we had calculated that there would be 12 TPs ready to be installed as of 11 August 2009, with a consistent flow of TPs available thereafter to support a continuous installation of Phase 1 TPs.

20.

Based on input received from Jumbo, we assumed that 20 TPs would require 1 day/each for installation plus 3 days “learning curve”, or a total of 23 days to install once a TP installation could begin (this number was based on the contractual quantity to be installed prior to the negotiated “away period” in the Jumbo contract). So, we anticipated that the first 20 TPs would be installed by 2 September 2009 and that the remaining 22 “Phase 1” TPs could be installed thereafter at a rate of 1 day per TP (or less), as TPs were available.

21.

After lengthy negotiations, Jumbo agreed to this “swap” of installation times on 7 July 2009. This agreement was later formally memorialized in a modification to the Jumbo Contract.

22.

When after the agreement no TPs became available we had to honour this contract modification. But we saved the costs for approximately three weeks standby time, which Fluor would have incurred without this new agreement.

23.

Of course, by 11 August 2009, we could not install any TPs because there were no MPs installed on which to place the TPs, and because there were no TPs available to install. The Shipment 1 MPs were still being retested and repaired in response to NCR 006, and the TPs had been set aside for NDT re-testing and remediation. Moreover, in the interim GGOWL had issued NCR 008 (29 July 2009) and NCR 009 (7 August 2009) requiring the retesting and repair of the Shipment 2 and 3 MPs and TPs, rendering it unlikely (if not impossible) that any MPs or TPs would be installed in the 2009 offshore installation season. As a result, we turned down Jumbo’s offer in early August 2009 to extend the Javelin through 15 November 2009.”

(My emphasis)

Although, as I have noted, Mr Helmonds did not give evidence at the trial (because the contents of his witness statement were not challenged in any important respect), I see no reason why I should not have regard to this evidence and, indeed, to attach significant weight to it. What he said in the passages cited above was, in my view, entirely plausible and consistent with all the evidence that I heard.

In particular, what Mr Helmonds said is consistent with the passage in the statement of Mr Eisman quoted above and with the minutes of the Weekly Progress Meeting held on 10 August 2009, at which it was noted that all SEPAM works on the TPs had been put on hold until further notice, pending the decision on the TP repairs and that SEPAM’s direct workforce have been employed elsewhere under Fluor direction. It is consistent also with a Fluor Subcontract Monthly Report relating to SEPAM dated 24 August 2009 which, under the heading “Major Areas of Concern” noted that “All works on TPs has been stopped at the request of Fluor due to welding repairs on TPs”.

Since SEPAM was the subcontractor responsible for the electrical fit out of the TPs, the notes of the meeting of 10 August 2009 confirm that by or during the first week in August 2009 Fluor had recognised that there was no prospect of installing any TPs in the near future and so put the TP fit out on hold until the welding problems had been resolved.

Further, in early August 2009 the contemporaneous documents show that Fluor was also giving active consideration to returning the TPs to Shanghai for repair.

Mr Hovermale’s evidence to the Tribunal was to similar effect. At paragraph 111 of his witness statement dated 27 February 2012, he said:

“TP installation commencing in a meaningful way before the departure of the Javelin on 19 September 2009 became impossible with the issuance of NCRs 008 and 009. Although Jumbo had offered us the option of extending the Javelin through as late as 15 November 2009 (and potentially beyond that date), we had no expectation at the time the offer was made on 7 August 2009 that meaningful progress on MP installation would be made by that time, let alone TP installation.”

In the light of this evidence, which again comes from Fluor itself, and the contents of the contemporaneous documents I find that the decision not to extend the charter of The Javelin was a direct result of the issue of the NCRs, and therefore the loss represented by the inability to use that vessel between 19 September and 26 March 2010 has been waived.

In the absence of The Javelin if Fluor subsequently decided that it wished to install TPs in late 2009 or early 2010, it would have to find another vessel. That is what it did and the vessel was The Leviathan. The Leviathan was the special purpose jack up vessel that was to be used for installing the wind turbine generators (“WTGs”), but in the absence of any installed foundations it was lying idle. In order to install TPs The Leviathan had to be modified, but even then she would only be able to install TPs at the rate of one TP in about 3.5 days (as against 1 TP/day for The Javelin).

On 11 December 2009 The Leviathan sailed to Vlissingen, where she remained until 15 January 2010. During this period various modifications were carried out, including works required to enable the vessel to install TPs. The Captain’s Daily Report to GGOWL for 21 December 2009 recorded:

“Ongoing works on vessel in preparation for TP installation.”

This was recorded in subsequent Daily Reports as continuing until 3 January 2010.

On 28 December 2009 the last of the 42 Phase 1 MPs, IGI 09, was installed. By then there was no installable TP available because none of the second-generation transformer faults had been repaired (or, if they had, the first batch of repaired transformers was still in transit to Vlissingen). Similarly, as noted in the previous paragraph, the work required to enable The Leviathan to install the TPs was still ongoing.

On 3 January 2010 the Daily Report recorded that:

“Jetting modifications to jack up legs completed.”

These modifications were probably put in train as a result of an incident a few weeks earlier when one of the legs became stuck in the seabed. Otherwise, the only activity recorded was:

“General maintenance around The Leviathan.”

The Daily Reports for the period 4-14 January 2010 recorded continuing general maintenance and little else. The Daily Report for 14 January 2010 contained a reference to activities for the following day which included:

“Ballast for simulation trial for TP loadout”

And

“Head off to the GGOWF for jacking trials.”

The daily reports of Mr Paul Pashley for 12 and 13 January 2010 include notes that The Leviathan was in port “awaiting TP load-out” and that Fluor intended to carry out the first load out of the TPs on 28/29 January. The latter note also appeared in the report for 15 January 2010. However, in his report for 16 January 2010 he noted that Fluor intended to carry out the first loadout of the TPs “first week in February”.

Meanwhile, on 15 January 2010 The Leviathan left Vlissingen for Harwich. The following day, the Daily Report noted that, whilst alongside the quay at Harwich, the vessel was jacked up

“with 500 T of ballast on board to simulate 2 x TPs.”

According to the Daily Reports, for the next two days the vessel remained jacked up in Harwich. Mr Anthony Morgan interpreted this as a trial to see whether or not the vessel could carry two TPs, rather than only one. I am not sure that he is right about this, but in any event it does not matter because, if he is right, it could have been dispensed with.

Mr Pashley’s daily report for 18 January 2010 repeated the note about the intended loadout of the TPs during the first week in February, and this was repeated in the report for 20 January 2010. In his report for 25 January 2010, Mr Pashley noted that the first load out of the TPs was to be on 1 February 2010.

On 20 January 2010 The Leviathan left Harwich and sailed to the site of IGE 02. The Daily Report for 20 January 2010 records that whilst jacking up at IGE 02 legs 2 and 4 penetrated the seabed more than legs 1 and 3, causing the vessel to list. The Daily Report for 22 January 2010 noted that the vessel was on standby for “MGR trials” (MGR stands for marine growth removal). The Daily Report for 23 January 2010 recorded that The Leviathan loaded a ring removal tool from OSV Relume. The report for the following day recorded that the stiffener ring had been removed from IGE 02.

On 25 January 2010 The Leviathan remained beside IGE 02 where gouging and grinding operations were taking place. The Daily Report records that the following evening the vessel jacked down prior to leaving IGE 02 for Harwich in the early hours of 27 January in order to prepare for TP installation.

In response to a suggestion from the court that The Leviathan might have been used to remove the stiffener ring from IGE 02 because there were no TPs to install, Fluor submitted that the removal of a stiffener ring had to take place before a TP was installed on an MP otherwise there would be a risk that the ring would fall down and damage the J tube entry point or the internal cabling, which would be very difficult and expensive to repair once a TP (and subsequently a tower and WTG) had been installed on the MP. That submission I accept.

In relation to IGE 02, Fluor further submitted that it was essential that removal of the stiffener ring was carried out in January 2010 since IGE 02 was to be the third MP to have a TP installed on it. So far as the 31 May 2009 programme is concerned, that is correct: IGE 02 was to be the third TP installed. However, a table prepared by Mr Morgan (Table G-5 in Appendix G to his first report, dated 24 March 2017, at paragraph G.6.1) shows that the electrical fit out to the TP for IGE 02 was not completed until 16 February 2010 and was the ninth TP to be completed that month. I did not understand this table to be disputed. A revision to the construction programme dated 5 January 2010 provided for IGE 02 to be installed between 22 March and 5 April 2010. In fact, the Jumbo/Fluor Daily Progress Reports dated 29 and 30 March 2010 for the Installation of the TPs record that the TP for IGE 02 was installed by The Javelin on those days. Indeed, as Fluor itself pointed out, (Footnote: ²⁰ ) in early 2010 it seemed that The Javelin would not be able to begin the installation of TPs until late May 2010. This information suggests that the as built sequence departed substantially from the original as planned sequence and that, as at early January 2010, there was no particular urgency about the removal of the stiffener ring from IGE 02.

Further, a document unearthed by ZPMC following closing submissions indicates that the suggestion by the court about the timing of the removal of the stiffener ring from IGE 02 may have been correct. This was the meeting note of a Contract Progress Meeting No 21, held on 14 January 2010, at which, amongst others, Mr Hovermale was present. Item 21/3.5 recorded the following:

“Leviathan is due to remove stiffener rings from piles after it has picked up the required access frame work, this is a way of exercising the boat and its equipment”

(My emphasis)

There was no correction to this observation in the minutes of the next meeting.

In his daily report for 26 January 2010 Mr Pashley noted that Fluor had requested a final inspection by GGOWL on 27 January “with a view to loading out on 1 ^st February”. The report for the following day noted that TPs 19 and 24 had been commissioned, and that the final inspection was to be on 28 January 2010, and the first load out was to be on 1 February. These reports are entirely consistent with a statement in a PowerPoint presentation by Fluor dated 26 January 2010 that “We will have the first two TPs on Quay by 2/1/10”. (Footnote: ²¹ )

The vessel’s Daily Report for 28 and 29 January 2010 record that The Leviathan’s sailing on 28 January was cancelled as a result of weather conditions and that she remained at Harwich for the whole of the following day, 29 January 2010, during which she loaded the “stairs for TP access”.

The Leviathan left Harwich for Vlissingen at about 14:00 hours on 30 January 2010. On 31 January 2010 the Daily Report records that The Leviathan was preloaded and preparations were made for TP loading. At 17:00 hrs there was a meeting “re tp loading”. This is consistent with Mr Pashley’s daily reports for 30 and 31 January 2010, the former of which noted that concrete blocks had been laid out at the quayside ready to accept TPs 019 and 024 “tomorrow”. This shows that the two TPs were not by then waiting on the quayside, but were to be delivered to it the following day (31 January). His daily report for 31 January recorded that The Leviathan was “in port jacked up at quayside ready for load-out”. It then went on to say: “TPs 019 and 024 transported to quayside in readiness for load-out to the jack up vessel Leviathan tomorrow”.

The Daily Report for 1 February 2010 noted that at 19:00 hrs that evening the TP for IGI 04 was loaded onto The Leviathan. The Daily Report for the following day, 2 February, records that by 16:00 hrs:

“seafastening of all equipment complete certificate issued by LOC”

The report noted that the vessel would proceed the following day from Vlissingen to the site of IGI 04.

I have not seen the certificate which was issued by LOC at 16:00 hours, but the reference to sea fastenings being complete suggests that it was a certificate to confirm that the vessel could proceed to installation, not that it was ready to receive the TP (as Fluor appears to contend) which, of course, by then had already been loaded onto the vessel. Without a certificate from a marine warranty surveyor The Leviathan could not carry out the installation of TPs (see the witness statement of Mr Helmonds, at paragraph 3.11). However, Mr Helmonds does not say at precisely what stage this certificate would have been issued.

Fluor’s submissions on this point are confusing. At paragraph 375(g) of its Closing Note it refers to a “MWS certificate issued to allow that vessel to carry TPs”. The following sentence refers to cross examination of Mr Morgan about the date on which The Leviathan “was ready to install TPs”, although the actual question put to Mr Morgan was about when The Leviathan “was ready to accept TPs”, which is not the same thing. Later in the same paragraph it refers to the certificate as indicating the date when The Leviathan “ became available ”. I think that Mr Morgan may have understood the difference because his answer, albeit somewhat inelegant, drew the distinction between the vessel being certified “to do TP works” and the vessel being “ready” (which she would have to be before the captain could obtain the certificate). Fluor’s criticisms of Mr Morgan’s evidence on this point are not entirely fair.

If, as appears from the evidence and the documents to be the position, a certificate from the MWS was required before The Leviathan, once loaded, could proceed to installation, then it provides no pointer to the answer to the question whether The Leviathan waited for the TPs or the TPs waited for The Leviathan.

The Daily Report for 3 February 2010 records that The Leviathan left for IGI 04 late that evening. Pre-loading was carried out on 4 February and on the following day the MP was cleaned using the MGR tool, after The Leviathan had been jacked up in preparation for the installation of the TP. The Daily Report for 6 February 2010 records that the TP was installed on IGI 04 during the course of the morning and the grouting was carried out between 15:30 and 19:45 hours. After this The Leviathan returned to Vlissingen.

I see no reason to doubt the accuracy of the information in either the vessel Daily Reports or the daily reports of Mr Pashley, which has not been contradicted by any more reliable evidence, and so I accept it. (Footnote: ²² )

For completeness, I should mention that according to Fluor’s Weekly Engineering Progress Reports for the TPs dated 11 and 18 January 2010 (cited by Mr Morgan at paragraph 5.73 of Appendix G to his first report), two transformers had been returned to Vlissingen by 11 January 2010 and a further 11 transformers were returned during the following week. These transformers had to be installed in TPs before commissioning could take place. The period for this shown in the 19 May 2009 programme was 20 days. However, ZPMC now accepts that the first two transformers were in fact returned to Vlissingen on 5 January 2010 and so I have taken that date.

In my view, when all the information in the documents is taken together it points strongly to the conclusion that The Leviathan was engaged on some of the activities that I have described during January 2010 because everyone was well aware that no TPs would be ready for installation until the end of that month. The reference to The Leviathan being used to remove stiffener rings as a way of “exercising the boat and its equipment” really leaves no room for any other conclusion - particularly since there was no evidence to the contrary.

It follows that it was not the unavailability of The Leviathan that prevented the TPs from being installed before the end of January 2010, but the fact that the first TPs were not ready. By this time, of course, the 42 Phase 1 MPs had been installed - the last one having been driven into the seabed on 28 December 2009. However, although this issue has generated much discussion, I doubt if it makes much, if any, difference in the long run because what really mattered, so far as TP installation was concerned, was the availability of The Javelin or a similar vessel.

I find, therefore, that in early January 2010 - at the latest - the critical path lay through the fit out and installation of the TPs. The next section of this judgment discusses the reason for this.

Part of the electrical fit out of the TPs, which was to be carried out by an Irish company, SEPAM, involved the installation of the transformers manufactured by ABB in Waterford, Republic of Ireland. The transformer takes the power from the WTG and turns it into a high voltage electricity supply which is fed via the export cables to the grid. The transformer is a large piece of equipment, weighing about 8-9 tonnes.

As a result of the perceived need to test the TPs, as well as the MPs, following the discovery of the defective welds, the TP fit out programme was severely disrupted because it was not realistic - as Fluor alleges and I accept - to install electrical equipment in a TP which required (or might require) gouging and repair of the welds. I find that had such equipment been installed prior to the remedial work then, in order to avoid damage, it would have had to be removed or protected before the repair of any defects was carried out. The disruption of the TP fit out programme meant that the fit out operatives had little to do and so, in order to keep them occupied, they were directed to carry out some pre-checks on the transformers which were at that stage lying in a warehouse. This was Mr Eisman’s evidence (at Day 8/33) and I accept it.

It was whilst carrying out these pre-checks that, on the thirteenth or fourteenth transformer checked, an unusually high resistance was discovered, which indicated a likely loose connection inside the transformer. According to Mr Eisman, this was on about 27 July 2009 (Day 8/34). However, Mr Trevor Plummer, of Fluor, in a witness statement prepared for the arbitration dated 30 March 2012, said that the fourteenth transformer to be commissioned failed on 21 July 2009, and that the eighteenth failed on 27 July 2009. This latter failure was recorded in a contemporaneous e-mail dated 29 July 2009. I regard Mr Plummer’s account as being more reliable, since it was based on contemporaneous documents, whereas Mr Eisman’s evidence appeared to have been given from memory.

Mr Plummer went on to say that a further eight transformers were successfully commissioned without any sign of the resistance fault, but on 6 and 7 August 2009 a further two transformers (the 27^th and 28^th to be commissioned) were found to have a similar resistance fault. During the following week ABB sent two engineers to inspect the four failed transformers, which were then returned to ABB’s factory in Waterford (there were not the facilities at Vlissingen to dismantle the transformers in order to investigate and repair the fault). This was done on 20 August 2009.

The four transformers which had failed the tests were units 9, 10, 11 and 12. On 25 August 2009 a Mr King, an independent engineer engaged by Fluor, prepared a Supplier Quality Surveillance Report for Mr Hardie. This gave an account of the inspection of three of the four transformers at ABB’s factory in Waterford, under the supervision of a Mr Patrick Rohan, of ABB.

Part of the circuitry in each transformer consisted of flexible copper links that were bolted to a rigid copper bus bar. At the lower end of these links it was found that some or all of the bolts connecting the links to the fixed bus bar had worked loose, so much so that some were missing. The bolts were used with a clinch type nut which was designed to embed into the bus bar and it was clear that some of these bolts had not been properly tightened at some point in the past. In each case it was noted that the flexible copper links consisted of ten laminates, rather than eight laminates as shown in the design drawings. It appeared that the bolt length was not long enough to engage all the threads on the clinch nut when the ten laminate flexible links were used.

A unit from the factory floor was then opened. This had eight laminate links in accordance with the design. No loose connections were found. During the discussion following the meeting Mr Rohan explained that it was not possible to check the quality of the connections using resistance measurements: it could only be done by visual inspection. Mr Rohan said that it was ABB’s view that the connections had become loose because the flexible copper links were too thick for the bolts used. Significantly, Mr Rohan said that it was not possible to identify from the recorded manufacturing data which units had the 10 laminate copper links. Thus the extent of the problem in the remaining transformers could not be predicted. ABB therefore recommended that all units already delivered to Vlissingen should be returned to Waterford.

I can see no reason not to accept the contents of this report, which I understood to be common ground. Copies of the report were circulated within Fluor two days later. Mr Rohan’s own report, dated 31 August 2009, noted that scoring on the bolt thread indicated that only two or three threads were engaging the nut (apparently the clinch nut is obscured from view when tightening the bolt). Mr Rohan concluded that during transportation to Holland some of the connections had worked loose. Under the heading “Recommendations” he noted that “an apparently good resistance reading can still have some loose bolts in the connection”. Since this report appears in the trial bundle and has a Fluor disclosure reference on it, it must have been sent to Fluor by ABB (as ABB had apparently promised to do). (Footnote: ²³ )

Against this background ABB confirmed, in a telephone conversation on 28 August 2009, that all the remaining transformers that had already been shipped to Vlissingen - about 45 in all - needed to be returned to ABB for testing and repair if necessary. Realising that this was a serious issue, Fluor agreed to this course.

In his first witness statement Mr John Eisman said this, at paragraphs 7.3 and 7.4:

“7.3

The extent of the manufacturing fault could not be firmly established by ABB although the faulty transformers were consecutively numbered (ABB ID#9-12), which suggested a defective batch. The repaired transformers (the “first generation transformers”) were returned to Fluor in instalments. They had all been returned by 9 November 2009 and thus the need to send them away had no impact on SEPAM’s progress. Because the repaired transformers were returned to Fluor in instalments, by the time SEPAM started fit-out of TPs in mid-October 2009, there were sufficient transformers on site for SEPAM to commence their works without experiencing delay.

7.4

Had the TP works in Vlissingen not been on hold, our decision would likely have been quite different, particularly given that only four transformers were found to have suffered from this particular problem. Since the transformers are all the same, in all likelihood, we would have proceeded with the installation of the remaining transformers in the TPs and the offshore installation of those TPs thereafter. Had any of the transformers showed signs of problems thereafter, or needed to be switched out, we could have made arrangements with ABB to switch out such transformers offshore. Although not ideal, it would have been quite feasible from a technical and construction schedule perspective, and, given that schedule was Fluor’s utmost priority, it is certain Fluor would have elected to do just that rather than holding up TP fit-out and installation.”

It was suggested to Mr Eisman that Fluor really had no option but to accept ABB’s proposal to return all of the transformers to Waterford since there was no way of knowing how many of them might be affected. Mr Eisman’s response (at Day 8/48) was this:

“Based upon a timing of events that this happened, and what was going on on the field from an installation standpoint, it was pretty much a straightforward decision.”

Mr Eisman went on to say that had it not been for the welding defects the transformers would have been checked as part of the TP commissioning process, and it would have been then that the problem would have been discovered. It was only discovered sooner because SEPAM was carrying out pre-checks on the transformers. He said also that, since it was not until the thirteenth or fourteenth transformer had been checked that a problem was discovered, it was likely - if things had gone according to plan - that several TPs would have been installed on MPs out at sea before the fault was discovered. Had this been the situation, Mr Eisman said (at Day 8/49-50) that:

“there would have been a much more in-depth discussion about what do we do with the TPs that have transformers installed in them versus transformers that are still sitting on the ground.”

At paragraph 7.7 of his witness statement dated 2 February 2017 Mr Eisman said this:

“If Fluor had been able to proceed with MP and TP installation in the summer of 2009 (i.e. if we were not prevented from doing so by the weld defects), I believe that Fluor management would not have made the decision to send the transformers back to ABB for repair. The reason is simple. By the time the issue was discovered in late August 2009 Fluor would have already installed 2-3 batches of TPs (i.e. 12-18 TPs). ABB’s report was then not received for several weeks after this. Fluor’s priority would have been to maintain the installation programme, even if this meant that it was likely that Fluor would have to swap out the bad transformers offshore at a later date and at an extra cost. The alternative would have been for Fluor to allow the Javelin (the TP installation vessel) to sit idle and pay enormous standby charges, which Fluor was obviously determined to avoid, and Fluor would have sought to recover the replacement costs from ABB in any event. Had we not returned the transformer (sic) to ABB in September 2009, the second transformer issue would never have arisen as this was caused by ABB’s own remedial works.”

It is not easy to reach a conclusion as to what Fluor would have done on discovering that, owing to a resistance fault, three or four transformers had failed the commissioning tests, but I do not find it easy to accept Mr Eisman’s evidence - if that is what it amounted to - that Fluor would not have returned any of the transformers to Waterford but would have continued to install transformers in the TPs and then install the TPs at sea as if nothing had happened and take the risk of having to replace any transformers in situ if they proved to be defective. Whilst Mr Eisman might have been right to suggest that the fault was confined to one particular batch, this could not have been assumed and, indeed, I am not aware of any evidence that the transformers were manufactured in batches of four. There was no suggestion by ABB that the problem might have been confined to one batch of four transformers.

I accept that, following discovery of the first, and possibly the second, high resistance fault, Fluor would probably have put the transformer in question to one side and simply replaced it with another and carried on with the fit out of the TPs. At that point there would have been no need to do anything else, and the defective transformer could be inspected in due course.

However, once the same fault had been discovered in three or four transformers I would have expected Fluor to have put further installation of transformers into TPs on hold until the defective transformers had been sent back to Waterford for urgent investigation and identification of the cause of the problem. If, by then, some TPs with potentially defective transformers had been installed out at sea, I can understand that Fluor might have chosen to leave them there – at least for the time being.

Once it became clear that ABB had no manufacturing records which would enable it to say with any certainty whether the four transformers were the only ones with the inadequate bolts, and that a transformer fitted with the short bolts could produce an apparently good resistance reading, it must have been obvious to Fluor that, at least as a possibility, there might be transformers that had already been installed in TPs which had the same defect but in which the securing bolts had not by then slackened sufficiently to produce a high resistance fault.

The date shown in the 19 May 2009 programme for the start of TP installation was 21 July 2009, when the first batch of five TPs was to be loaded out. By the time of issue of the 31 May 2009 programme this date had moved back to 31 July 2009. On 10 June 2009 Fluor agreed with Jumbo to put the loading out of the first batch back to 11 August 2009. This was presumably as a result of the weld defects.

I have already concluded that, having regard to the out of roundness problems, in the absence of any weld defects it is likely that seven MPs in Shipment No 2 could have been installed during July 2009 with the hammer that had been ordered initially (and thereafter using the first replacement hammer on the basis that it would have been delivered in about mid-July). This would have been in place of the 5 MPs from Phase 1 Fluor planned to install by the end of July according to the 31 May 2009 programme (because the jacket was also installed during this period).

Mr McKibbin has postulated that Fluor could have had the first batch of six TPs ready for loading out by 19 September 2009. This is on the assumption that, if the TPs had been free of welding defects, fit out work could have started on 4 July 2009 and would have taken 11 weeks. (Footnote: ²⁴ ) A further batch of six TPs would have been ready by 3 October 2009, and the third batch of six TPs would have been ready by 17 October 2009 (commissioning of that batch having been completed by 10 October 2009).

I will assume, without deciding, that this counterfactual scenario is plausible. I will assume also that, by 10 October 2009, high resistance faults would have been discovered in two or three transformers in the third batch, following the discovery of one high resistance fault in either the first or second batch. (Footnote: ²⁵ ) In other words, I am assuming - in Fluor’s favour - that the realisation that the high resistance faults discovered during actual commissioning - as per the scenario - might present a serious problem occurred after fewer transformers were commissioned (about 12-15) than was the case during the pre-commissioning that was actually carried out during June, July and August 2009 (after the 27^th and 28^th transformers were checked). I have said that this assumption is in Fluor’s favour because it assumes that the transformers would have been taken from the warehouse for installation in TPs in a different order, such that the problem was discovered earlier during the commissioning process than was in fact the case. It is to be recalled that, as delivered, the transformers were identical and it was not until after a transformer had been installed in a particular TP that it became specific to that TP.

So, on the assumptions that I am making, by 10 October 2009 Fluor would have known that the high resistance fault that had been detected earlier was not an isolated event. The discovery of further faults in the third batch would have dispelled any assumption that the fault was a one-off problem, thereby disclosing a more serious state of affairs. As I have already indicated, in these circumstances I consider that Fluor would have decided to return the defective transformers to Waterford immediately for investigation.

Once those transformers had been tested by ABB at Waterford it would have become clear (just as it subsequently did) that the extent of the problem in the remaining transformers at Vlissingen could not be predicted. On the sequence of events that I have assumed, this could have been during the third week of October 2009. At this point I would have expected ABB to propose, as it did in the circumstances which actually prevailed, to recall all the transformers remaining at Vlissingen to Waterford for inspection and repair. I consider that Fluor would have agreed to this proposal, save in respect of the 12 transformers that were in TPs which by then - on this counterfactual scenario - had been installed out at sea.

I pause at this point to revert to what actually happened. In the event it took about five to six weeks for the remaining transformers to be returned to Waterford, repaired and shipped back to Vlissingen: this is not a precise figure – the transformers were returned to ABB in batches, the first four being returned to Waterford on 20 August and received back in Vlissingen on 21 September 2009 and the last transformers were sent back to Waterford by 9 November 2009. (Footnote: ²⁶ ) In the hypothetical weld defect free scenario, it would have taken longer than this because about a dozen transformers would by then have already been installed in TPs at sea, rather than lying largely unpacked in a warehouse or installed in a TP from batch 3 that was still on shore.

Reverting to the counterfactual scenario, on any view, I consider that the earliest date by which the first few transformers in which a fault had been found could have been investigated by ABB, repaired and returned to Vlissingen would be mid November 2009, with the next batch being returned about two weeks later - that is to say by the end of November 2009. This former date is therefore about 6-7 weeks after the date on which the first four transformers were in fact returned to Vlissingen following initial repair (21 September 2009).

Once the first three or four transformers had been returned to Vlissingen, they would have been installed in TPs and subsequently commissioned. That commissioning would have revealed the second round of faults caused by the faulty repairs at Waterford. Assuming that the preliminary work (such as secondary steelwork) was already complete in some of the TPs, I will assume, again perhaps generously towards Fluor, that the second-generation faults resulting from the repairs would have been discovered by the end of November 2009 (by which time the commissioning process should have been close to completion). By this stage, of course, on this scenario the installation of further TPs at sea would have been at a standstill since about mid-October and further installation of TPs would probably have been prevented by the prevailing seasonal weather conditions in any event (see paragraphs 170-171 above).

Assuming that Fluor realised straight away that these first three or four transformers would have to be returned to Waterford for further inspection and repair, I doubt very much whether they could have been shipped back to Waterford, repaired and returned to Vlissingen within less than 4-5 weeks (the period actually taken was about six weeks).

So the outcome of this analysis is that, even on the counterfactual scenario, the transformers would not have been returned to Vlissingen following repair of the second-generation faults until early January 2010. So even if I were to accept Mr McKibbin’s hypothesis, and making assumptions favourable to Fluor, on the face of it the effect on the project of the problems caused by the transformers would have been to prevent any further TPs from being installed until the end of January 2010 in any event.

However, this leaves the fact that 12 TPs would have been installed at sea before the fault was discovered during the commissioning of the third batch. But, of course, these TPs were fitted with transformers that were or might have been defective. In my view, Fluor could not possibly have left those transformers in situ when there was no means of knowing whether or not any of them had the original fault (ie. 10 laminates instead of 8 with a securing bolt that was too short). If any of them did have the fault, then there would have been a serious risk that during its working life the bolt would work loose. The submission, at paragraph 385 of Fluor’s Closing Note, to the effect that it is very unlikely that Fluor would have removed from TPs installed at sea any transformers that had passed commissioning, is in my view quite untenable and was not supported by any satisfactory evidence. (Footnote: ²⁷ ) To do that would be to take a huge risk of a transformer failing a few months or, worse, years later.

Mr Eisman said, and his evidence was backed up by material produced during closing submissions, that the transformers could have been replaced even though the TPs have been installed at sea. The TP had an access door the removal of which would have permitted the removal of the transformer and its replacement. However, this would obviously have involved the use of a suitably equipped vessel, but there was very limited evidence before the court as to how long it would take to swap out a single transformer at sea or what the vessel required to do it would cost. In its Closing Note Fluor simply described this option as a “seemingly less costly route” (paragraph 387(c)). The Leviathan would not be available, because by this time it would be installing WTG’s following the installation of the first TPs (as described by Mr McKibbin).

During his oral final submissions Mr Brannigan suggested that the transformers could have been swapped out at sea using the OSV Relume. She cost roughly €20,000 per day, had at least one crane that had the necessary capacity and was available to Fluor. Mr Brannigan referred also to evidence given by Mr Plummer in the arbitration that transformers had been swapped out at sea during a day. (Footnote: ²⁸ ) But since it was Fluor’s case that these costs would have been met by ABB (see Day 17/122), they would not have represented a loss to Fluor recoverable against ZPMC.

The position becomes even more complicated if one assumes (a) that the order in which the transformers were installed in the TPs (in the counterfactual scenario) followed much the same order as that in which they were taken out for pre-commissioning and (b) that Fluor would have adopted the course (put forward by Mr Eisman) and just carried on installing TPs with potentially defective transformers. On this basis, the potential extent of the high resistance fault would not have become apparent until mid or late November 2009 (assuming that there was still a vessel available to install TPs beyond 15 November 2009), when the 27^th and 28^th transformers would have been commissioned. This would mean that the second generation of repaired transformers would not have become available until, probably, some time in February 2010, by which time there would have been 30 or more TPs installed at sea, each of which would have to have its transformer swapped out.

This is not a state of affairs that was canvassed at the trial and I am quite unable to say whether it would have left Fluor in a better or a worse position.

So far as the transformer problem is concerned, my overall conclusion is that Fluor might well have been left no better off in the counterfactual scenario than it would have been as a result of the events which actually happened. However, this conclusion is only relevant if the NCRs did not apply to the TPs. If the NCRs did apply to the TPs, then Fluor’s inability to install any TPs before 2010 was the result of the issue of the NCRs and the problem with the transformers was simply another cooperating cause. I now turn to the question of whether the NCRs did apply to the TPs.

In paragraph 578 of the Liability Judgment I said that I had heard no evidence as to the extent, if at all, to which any of the NCRs required, either expressly or by implication, testing or repair of TPs. This was a point that arose for the first time in the course of closing submissions in that trial.

In paragraph 579 of the Liability Judgment I went on to say this:

“I must therefore leave open for further argument the issue of whether costs of testing or repairing any TPs was required - either expressly or by necessary implication - by any of the three NCRs and, if so, to what extent. This includes the question of whether or not it is open to Fluor to take this point.”

In December 2016 Fluor applied to re-amend its Particulars of Claim. One of the amendments was an assertion that the NCRs did not require work to any of the TPs. I allowed this amendment, albeit not without considerable hesitation.

Although Fluor now argues that the NCRs did not extend to the TPs, it has never disputed that the NCRs required Fluor to retest the welds by scanning on the top of the welds using D scans although this particular UT technique is not mentioned in any of the NCRs. NCR 006, for instance, stated that:

“. . . the examination results have to be re-validated and appropriate weld repairs made to ensure that all welds meet code requirements.”

There is nothing in the wording of NCR 006 or in the examination report attached to it that says anything about the type of “re-validation” required, although some examination reports mention that the NDT that gave rise to the report was by “D scan only”.

However, there is no doubt that Fluor understood that the “re-validation” had to be by way of D scan. In an e-mail from Mr Hans Dekker to Mr Smith, of Airtricity, dated 14 June 2009, Mr Dekker wrote:

“In the meantime we have instructed ZPMC to do d-scan only until we know more.”

It is therefore clear that the requirements of NCR 006, as they were understood by the parties at the time, were not to be found in the wording of the NCR alone: GGOWL soon left Fluor in no doubt as to what had to be done if it wished to clear the NCRs.

At paragraph 422 of the Liability Judgment I quoted an extract from paragraph 25.1 of Mr Dekker’s witness statement in reply in the arbitration where he said that:

“NCRs 006, 008 and 009 condemned all the TPs and all the MPs in Shipments 1, 2 and 3, which brought the construction phase of) this Project to a complete halt.”

I concluded that this reflected Fluor’s view of the position following the issue of the three NCRs. That was a finding of fact.

At paragraphs 431 to 473 of the Liability Judgment I recorded the contents of the meetings and communications between Fluor and ZPMC between August 2009 and early February 2010. There was nothing whatever in these meetings and communications to suggest that the parties were drawing any distinction between the costs of testing and repair of MPs on the one hand and the costs of testing and repair of TPs on the other. This was confirmed by the evidence of Mr Ho which was given at this hearing.

The opening paragraph of the waiver letter which was handed over at the meeting on 20 April 2010 was as follows:

“I am very pleased that we have reached an understanding on the joint pursuit of the substantial additional costs both Fluor and ZPMC have incurred as a result of the MP and TP weld testing and repair protocol required by Greater Gabbard Offshore Winds Ltd.”

(My emphasis)

Numbered paragraph 1 of the letter referred to the assignment by ZPMC to Fluor of the costs that ZPMC had incurred arising out of:

“the enhanced testing and remediation of MP and TP welds it has been required to perform as a result of the acts and omissions of GGOWL, including, but not limited to, GGOWL’s issuance of NCR 006, 008 and 009 relating to MP welds.”

This correctly reflects the possibility that the enhanced UT procedure that GGOWL required Fluor to carry out was not to be found only in the NCRs themselves.

Whilst it is correct that numbered paragraph 3 of the letter, which defined the scope of the waiver to be given by Fluor, referred to “the additional costs and delays it suffered as a result of NCR 006, 008 and 009 issued by GGOWL”, I consider that the parties clearly intended the scope of the waiver to mirror the scope of the assignment: it would be very odd if the parties had agreed that ZPMC was to assign to Fluor its claims in respect of the costs of testing and remediating the welds of the TPs without there being a corresponding waiver by Fluor of its corresponding costs in relation to the TPs.

But even if this by itself is not a sufficient ground to construe the letter in the manner contended for by ZPMC, I consider that it was the common understanding of the parties throughout the course of their negotiations in 2009 and 2010 that the weld testing and repair protocol required by GGOWL extended to both the MPs and the TPs and that the waiver letter must be read in the light of it. I accept that in the early correspondence Fluor’s claims referred to the costs incurred in relation to the MPs only, but that distinction was not maintained at any of the discussions which followed.

In addition, having regard to the fact that, having obtained the assignment from ZPMC, Fluor pursued in the arbitration claims in respect of the testing and repairs to both the MPs and the TPs, I consider that Fluor is now estopped from contending that the costs and delays that it suffered as a result of the issue of NCRs 006, 008 and 009 did not include costs arising in relation to the additional testing and repair of the TPs. Having regard to both the conduct and the common understanding of the parties, it would be quite inequitable to allow Fluor to go back on that common understanding that it shared with ZPMC at the time of the agreement.

In its First Partial Award, the Tribunal said, at paragraph 362:

“. . . the issuance by GGOWL of the NCRs constitutes in the Tribunal’s view an effective direction for the Contractor to comply with its requirements for additional testing and repair about which the Contractor complains. The discussions between the Parties subsequent to NCR 006 made it clear that GGOWL would only accept D scanning at 12 db as adequate to clear the NCR, and that the retest and repair requirements extended to TPs as well as the MPs.”

I have not seen or considered all the material that was before the Tribunal, but the evidence before the court in these proceedings shows that Fluor understood shortly after NCR 006 was issued that it would not be cleared unless Fluor carried out the re-validation of the NDT using D scans (as GGOWL’s inspectors had done) and, further, that GGOWL expected this to be done for all the welds, including the welds of the TPs. Neither the evidence of the witnesses nor the contemporaneous correspondence show precisely when this understanding arose in relation to each shipment, but I consider that in relation to NCR 006 it did so within a week or two of its issue and, in relation to the other two NCRs, probably less. What is clear from the evidence of Mr Dekker to which I have referred above is that the understanding existed when the work was being carried out.

Leaving all other points aside, the position now being adopted by Fluor could hardly be more unattractive. Fluor having fought tooth and nail to obtain a decision from the Tribunal that the NCRs required testing and remediation of the TPs, as well as the MPs, and having obtained a decision to that effect, a reasonable observer might think that something would be very wrong if Fluor were to be allowed to assert the contrary in these proceedings. In my judgment, Fluor cannot be permitted to do so because that would be an abuse of the process.

On 3 June 2009 a Mr Robert Bilham, of Parsons Brinckerhoff, sent Mr Peter Reilly, of GGOWL, an e-mail which was timed at 11:58. The date on the message is in the European format, and so it is reasonable to infer that it was sent at about midday GMT or, possibly, an hour earlier if it was sent from a server based on the continent. The text of this e-mail was as follows:

“Peter,

Please find attached our reports covering the visual and NDT inspections carried out in Vlissingen last week.

I will use Ian’s NDT report supported by the RTD report as backup to the NCR.

Regards

Bob”

Mr Reilly then forwarded that message, without comment, to Mr Hovermale at 13:03 on the same day. The attachments to the message were described as: “Visual Inspection report 1^st June 09”, “Monopile Underfill Locations” and “NDT Report Summary 1^st June 09”. There was no NCR.

Later the same day, at 16:50 local time, Mr Reilly sent Mr Hovermale a further e-mail in which he said:

“Bruce,

Please find attached initial documents re further NCR on Monopiles for discussion tomorrow morning 08:30 hrs, an up to date of (sic) NCR log is attached for completeness, you should also note that the power transformer failure is included.

Peter”

The attachments to this message were: the NCR itself, the RTD Interim Report, the NDT Report Summary and an extract from the NCR log (which recorded that the NCR had been issued on 3 June 2009).

In the NCR it was noted that the validity of the ultrasonic examination executed by ZPMC on all monopiles within the first shipment was in question; it then continued:

“the examination results have to be re-validated and appropriate weld repairs made to ensure that all welds meet code requirements.”

As I have already mentioned, a point which emerged towards the end of this hearing (I think for the first time as a live issue) was that the NCR did not specify what work was required in order to satisfy it and, in consequence, was not effective as an instruction at the time when it was issued. However, this point was addressed indirectly in the Liability Judgment, at paragraph 405, where I said:

“At some stage in June Mr Hardie had a conversation with a Mr Pashley, of GGOWL, to discuss what Fluor had to do in order to have the NCR released. Apart from providing the relevant paperwork in relation to the welds, Mr Pashley made it clear that Fluor would have to scan all the welds using D scanning. Mr Hardie said that it was not practicable to scan unground welds with the probe on the weld cap on account of the irregularity of the surface, but he was told that it could be done using copious amounts of couplant.”

ZPMC’s answer to this was to point to two e-mails from GGOWL, dated 2 and 6 April 2009, respectively, in which it said that it could not have confidence in the current ultrasonic inspections of the welds and that revised inspection procedures would be required. In the earlier e-mail of 2 April 2009 the author had said that:

“Transverse scans should be carried out from the weld cap provided the surface is smooth enough, which it should be on normal submerged arc welds of this size”.

ZPMC also relied on the following evidence given by Mr Hovermale to the Tribunal during the arbitration (which was given with reference to the e-mail of 6 April 2009):

“A. My view is the seed was planted here. I kept pushing back. What disturbed me about it is I had provided some input. In most cases if I provided a reasonable explanation the issue got closed. I think, however, in this case it would not go away. I had numerous discussions, you can see that in my witness statement, with Peter Reilly and others that: we want to re-test in Vlissingen, we want to re-test in Vlissingen. Then as soon as the re-tests occurred in Vlissingen, NCR hit us, so, at a minimum, the manifestation of it was in Vlissingen with the first NCR.

Mr Grove: (Footnote: ²⁹ ) So, in your view, it was a disguised direction right from the start?

A.

That is my view.”

During the liability trial Mr Hovermale was asked whether he stood by that evidence, and he said that he did (Day 6/108).

I can well understand, and I am prepared to accept, that when NCR 006 was issued, Fluor would have appreciated that in order to clear it Fluor would probably have to carry out the testing described in the e-mail of 2 April 2009. However, I consider that before taking any action in response to the NCR, Fluor would have been entitled to require GGOWL to state what type of re-validation it required in order to clear the NCR. It may well be that if Fluor had not done this, but had proceeded to follow the guidance given in the e-mail 2 April 2009, GGOWL would have been estopped from contending that such testing was not required by NCR 006 - but in my view that would not have detracted from Fluor’s right to have the instruction clarified if that is what it had chosen to do.

In a witness statement made in the arbitration and dated 27 February 2012, Mr Ayres said, at paragraph 47:

“On 5 June 2009, I was tasked with developing the process to address NCR 006 and get it released. My first step was to find the equipment, facilities and labour needed to carry out this work. I immediately asked our procurement people to look into obtaining additional rollers from Bode Positioners to support the MPs. We had already procured one set of rollers from Bode to support the China NCR works . . .”

In his first witness statement for the liability trial dated 24 July 2015, Mr Ayres said (at paragraph 7.18) that NCR 006 added a further layer of complication because Fluor

“now had to get GGOWL’s buy in to the testing and repair exercise which was already full steam ahead and we knew that this would be laborious (which it was)”.

In a witness statement made for the liability trial by Mr Hardie, dated 7 August 2015, he referred (at paragraph 12.32) to an entry in Mr Paul Pashley’s diary which referred to a meeting with Mr Ayres on 6 June 2009. That entry was, according to Mr Hardie’s statement, in the following terms:

“He also would like Martin Hoskins [Hickson, I think he meant] NDT to also work alongside his NDT men and come to some agreement with regards to the best way of identifying the defects and the way to go ahead.”

Taking this evidence as a whole, I am not satisfied that, on 4 June 2009, Fluor proceeded on the basis that it was to carry out testing as described in the e-mail of 2 April 2009, although it may well have appreciated that that is what it would ultimately be required to do.

This conclusion is reinforced by what actually happened on 4 June 2009. In his witness statement dated 24 July 2015 Mr Ayres said that on 3 June 2009 he met Mr Hardie and told him to instruct QIS to start testing the first three MPs as soon as possible and that, at paragraph 7.9, he:

“instructed Doug [Mr Hardie] to provide a copy of the ZPMC NDT procedure to QIS and to make clear to QIS that they were to test in accordance with that procedure . . ..”

He went on to say, paragraph 7.12, that this is what QIS started to do on the morning of 4 June 2009.

Mr Dove prepared a report on 4 June 2009 which provided “a summary of the current status of these defects up until 13:00 hrs, on Thursday 4 ^th June”. This report included the following:

“PB Inspection Report No 09IM027 (date 1 ^st June 09)

PB used RTD to UT inspect various, cap ground circ. welds, on the MPs: IGH 06, and IGI 04 . . .

UT Inspection

On Thurs 4 June, six UT operators, from QIS, began scanning the Circ. Welds of IGI 05, which were internally ground on the weld cap . . .”

The report concluded by listing seven “Proposed Actions/Recommendations”, of which five had nothing to do with the NDT procedures. The two which did were in the following terms:

“i)

send David Thomas, Fluor’s level III, UT consultant, to ZPMC to perform a detailed review of all NDE activities associated with the GGOWF Project.

vii)

Get Sonovation to rapidly scan large volumes of weld, leaving detailed sizing by slower UT”

Beyond this, I am not aware of any evidence which shows precisely what happened on 4 June 2009 in terms of NDT activity. Mr Dove’s report appears to suggest that the scanning carried out that day was confined to the scanning of the welds that had been ground. Assuming that this was in the form of D scanning, this would not be work that necessarily went outside the procedures already being carried out by Fluor and ZPMC - because D scanning was specified as the method to be used when welds were ground flush (see paragraph 288 of the Liability Judgment). That appears to be consistent with what Mr Ayres said he told Mr Hardie to do. However, so far as ZPMC was concerned, that would have been subject to the concession given by RFI 34 - which relieved ZPMC of the requirement to perform a D scan on a ground weld if that weld had been the subject of an E scan before grinding.

I am not aware that any similar concession was given by GGOWL to Fluor and so I cannot see any basis on which Fluor could have argued - as against GGOWL - that the testing described by Mr Dove was “extra-contractual” so that it was entitled to be paid for the work by GGOWL. On the contrary, it was an instruction to redo work that GGOWL contended had not been carried out properly the first time round. Accordingly, it was not work the costs of which fell within the scope of the waiver letter.

Further, if Fluor was under the clear understanding that it had to adopt the testing methodology set out in the e-mail of 2 April 2009, I would have expected to have found this reflected in Mr Dove’s report.

For these reasons I conclude that the work carried out by Fluor on 4 June 2009 was not work that was required by the NCR because it did not involve any NDT that Fluor was not already obliged to carry out. Accordingly, irrespective of whether the NCR was issued on 3 or 4 June 2009, I find that Fluor did not incur any costs in respect of additional testing on 4 June 2009 that fell within the scope of the waiver.

Since neither side contends that NCR 006 is to be treated as if it had been issued on any date later than 4 June 2009, I need not say any more.

The court has been provided with detailed and careful reports by the two delay experts, for which it is grateful - not least because they provide a ready source of reference for the dates of key events. There has been an extensive debate about the correct approach to delay analysis. Mr Morgan said, and I would accept, that a prospective analysis - in other words considering the critical path at any particular point in time as viewed by those on the ground at that time - does not necessarily produce the same answer as an analysis carried out retrospectively. The former is the correct approach when considering matters such as the award of an extension of time, but that is not the exercise with which the court is concerned in this case. I agree that some form of retrospective analysis is required.

However, I am not persuaded that the approach adopted by either expert is appropriate given the unusual circumstances of this case - particularly, the impact of the waiver given by Fluor. I held in the Liability Judgment that the carrying out of the additional testing and consequent repairs required by each NCR was the result of the issue of that NCR, whether or not such work would have been carried out in any event.

But, that said, the court must not lose sight of the fact that the dominant underlying cause of the problems that occurred at Vlissingen was ZPMC’s breaches of contract in relation to the NDT and welding. It was those breaches that gave rise to the issue of the NCRs. However, the effect of the waiver and warranty letter is that to a large extent the NCRs “trump” those breaches of contract by relieving ZPMC of the consequences that would ordinarily flow from those breaches.

Whilst the waiver given by Fluor applies only to costs and expense incurred as a result of the so called “extra contractual” testing required by each NCR, that does not mean that such costs and expense are limited to the cost of the actual work required by the NCR: if a cost was incurred or wasted directly as a result of the issue of an NCR, that cost has been waived. A particular example is the loss that flowed from the unavailability of The Javelin after 19 September 2009. The NCRs did not require Fluor to refrain from extending the contract period for The Javelin, but the practical effect of the issue of those NCRs was to make it impossible for Fluor to install any TPs until well after 30 September 2009, when it decided to disregard the requirements of the NCRs and to proceed with the installation of MPs whether repaired or not. In those circumstances, to extend the period of hire of The Javelin beyond 19 September 2009 would have been pointless. (Footnote: ³⁰ )

However, it must be remembered that the issue of the NCRs was not the only factor delaying progress during the summer of 2009. As I have already explained, the out of roundness of the MPs was the cause of delay in making them ready for installation, and the defects in the transformers delayed the fit out of the TPs.

Fortunately, I do not need to decide which expert’s approach is correct, or even whether neither is correct, because in my view the consequences of the waiver given by Fluor limit the scope of the inquiry that the court needs to undertake. Indeed, the way in which Fluor put its case on delay in its Closing Note (from paragraph 398 onwards, and in particular at section 4.2.4) makes the exercise unnecessary. For example, whether or not, in June 2009, the critical path lay through the fit out and/or installation of the MPs (which I prefer) or the fit out and/or installation of the TPs does not matter because in my view the course of the critical path during the relevant period was determined by the events of August and September 2009.

As I have already mentioned, in August 2009, following the issue of NCR 009, fit out work to the TPs of Shipments No 2 and 3 was put on hold (Footnote: ³¹ ) and Fluor took the decision not to extend the period of hire of The Javelin beyond 19 September 2009. I have already found that both of those events were a direct result of the issue of the NCRs. Accordingly, the direct consequences of those decisions, whether in terms of direct cost or delay costs, have been waived.

However, by the end of September Fluor had decided to go ahead with installation and to stop any further testing or repair of MPs and TPs. At that point, therefore, the NCRs ceased to have any direct effect on decisions taken by Fluor after that date. Since no further testing or repair of MPs was to take place, there was a supply of MPs ready for installation and thereafter the rate at which the MPs could be installed was determined by the efficiency of The Stanislav Yudin: the position is illustrated graphically in the table at paragraph 7.3.8 of Mr Morgan’s first report. The last Phase 1 MP was installed on 28 December 2009.

By contrast, on 1 October 2009 there were few, if any, TPs that were ready (or even approaching readiness) for installation and there was no suitable vessel with which to install them even if they had been ready. For TP installation work to progress there had to be (a) an MP installed in the seabed (with any stiffener ring removed), (b) an installable TP (although efficient installation would require a batch of TPs that matched the capacity of the installation vessel) and (c) an installation vessel. The return of The Javelin had been scheduled for 30 June 2010, but this was later brought forward to 26 March 2010 under the terms of what was known as the Early Start Agreement. The only alternative steps which Fluor could take, and did take, were to implement modifications to The Leviathan to enable her to install TPs. But I suspect that this act of mitigation was a gesture of hope rather than expectation, since it would (and did) take some time to install the appropriate equipment and fittings in The Leviathan and, even then, she was only able to install TPs at the rate of one every 3.5 days, if that. Thus to install 42 TPs would take at least 21 weeks, assuming suitable weather conditions. But during the period January to March 2010, suitable weather throughout was not a possibility.

I have already discussed the movements of The Leviathan in December 2009 and January 2010, from which it is quite clear that at some stage during the latter part of 2009 the installation of the MPs ceased to be on the critical path. This is because there was likely to be a significant time lapse between installation of the MPs and the subsequent installation of the TPs: as Mr McKibbin put it, at paragraph 6.9.5 of his first report, “the sequence of the MP installation became immaterial as it was clear . . . that all the Phase 1 MPs would be installed before any TP could be installed”.

Effectively, therefore, there was no way by which Fluor could improve upon the arrangements that had been negotiated with Jumbo – in the Early Start Agreement - by which The Javelin was to resume operations on 26 March 2010. Although I understand that Jumbo was willing to make The Javelin available earlier, for reasons that I have already given, installing TPs in winter months using The Javelin or a similar vessel was not a viable option. By the time the The Javelin returned on 26 March 2010, 18 Phase 1 TPs had been made available for installation (according to Mr Morgan’s table G-6), of which 9 had already been installed by The Leviathan. Accordingly, with the remaining 9 TPs available for installation, The Javelin could start installing TPs without any delay.

Mr McKibbin has concluded that there were 232 calendar days critical delay to the project between 20 May 2009 and 7 April 2010. Of this, he says that 127 calendar days were lost between 20 September 2009 (the day after The Javelin’s first operational period expired) and 17 March 2010, and a further 8 calendar days were lost between 18 March 2010 and 7 April 2010, making 135 days in all; although he said that the cause of this additional 8 days delay was “unclear”, he expressed the opinion that its most likely cause was ongoing disruption to the TP installation (see paragraph 6.11.2 of his first report). Mr McKibbin has not considered the delays to the project after 7 April 2010, which was the start of the IA cable installation. This is because his instructions were that Fluor was not making any claim in relation to delays to the project caused by ZPMC’s breaches of contract after that date.

I am not satisfied that the 8 days of delay that Mr McKibbin says were lost in March/April 2010 were the result of the breaches of contract by ZPMC. I regard Mr McKibbin’s attribution of the cause (namely, “ongoing disruption to the TP installation”), even if correct, as being too vague to justify a conclusion that this period of delay was a consequence of any breach of contract by ZPMC. However, for the reasons I have already explained, whilst I consider that the 127 calendar days that were lost as a result of the unavailability of The Javelin were caused by ZPMC’s breach of contract, the resulting cost has been waived.

The position in relation to the period of delay that is the subject of Fluor’s claim has not changed since Mr McKibbin wrote his first report, because at paragraph 402 of its Closing Note Fluor bases its valuation of the delays suffered in period (C) (prior to the issue of NCR 008) on the 135 days critical delay found by Mr McKibbin to have occurred between 20 September 2009 and 7 April 2010. Accordingly, there is no need for me to consider critical delay beyond this date.

In addition to its claim for the loss of the opportunity to install the Shipment No 2 MPs during July 2009, Fluor contends that the start of the fit out work to the TPs of Shipment No 2 was delayed by 23 days during July 2009 as a direct consequence of the defects in the welds. It submits that this meant that the work to those TPs were always going to be 23 days behind programme, whenever work to the TPs was resumed.

I have already found that the contract period for The Javelin was not extended beyond 19 September 2009 because, since the issue of the NCRs, there was no prospect of installing any TPs at sea within the next few months. As Mr Helmonds explained at paragraph 3.13 of his witness statement dated 23 December 2016:

“From the start, in October 2009, I was directly involved in the negotiation of what became known as the “Early Start Agreement” with Jumbo, which allowed for a Jumbo vessel to return to the Project in March 2010 to commence TP installation. The decision as to which vessel to supply and use was made by Jumbo. Jumbo was pushing to start the installation of TPs as early as possible as they had vessel availability. While we considered bringing on the installation vessel earlier, a number of factors/concerns militated against it:

(a)

We were concerned that due to the cold weather the grout (used to connect the TP to the MP) would not have enough time to set properly during the weather window which would make the connection between the TP and MP less stable and susceptible to shifting in later or unforeseen adverse weather conditions and put Fluor at risk of having to correct the grout outside the weather window.

(b)

Also, even if we were able to get some of the TPs installed starting in February, we could not begin IA cable installation until early April 2010 (the date that the IA cable installation vessel was contracted to mobilise). Therefore, it did not seem to us to be reasonable or beneficial to expedite TP installation, only to have the TPs sit idle for two months (or more) before IA cable installation could begin.”

In the ordinary course of events, and other things being equal, I might be prepared to accept Fluor’s submission that, if the 23 days had not been lost in July 2009, whenever TP work was restarted it would have been further advanced, and the acceleration required correspondingly less. (Footnote: ³² ) However, this does not sit easily with the above passage from Mr Helmonds’ witness statement, which I accept.

In addition, table G-5 in Mr Morgan’s first report gives the dates of start and finish of the electrical fit out to the Shipment No 2 TPs. This shows that the electrical fit out to those TPs was completed between 1 February and 26 February 2010. However, the length of time it took varied between 72 days and 183 days. Following completion of the electrical fit out, a further two weeks was required for commissioning and loading out prior to installation (see paragraph 1.6.2 Mr McKibbin’s Note on Counterfactual dated 10 July 2017). So it appears from this that, even if the TP fit out had been extended by 23 days for the Shipment No 2 TPs, that did not cause any delay to any other activity because, by that time, the TP fit out was effectively in float (because The Javelin was not available and the only installation vessel available was the much less efficient Leviathan). As Fluor put it, at paragraph 395 of its Closing Note, “a steady stream of TPs were available for the Javelin upon its return”. Accordingly, any delay that there might have been to the fit out of the TPs was absorbed by the float created by the delay in the return of The Javelin.

I am therefore not prepared to find that any interruption of the fit out of the Shipment No 2 TPs in July 2009 caused any material delay to any subsequent activity. However, I consider that Fluor did suffer a loss, in that during July 2009 its TP fit out workforce was not usefully employed on the fitting out the TPs and were redeployed on other tasks, which did not have to be done, such as the pre-commissioning of the transformers. Although, for example, the SEPAM workforce was demobilised as a result of lack of work, that did not happen until about mid August 2009.

In my view, the fair way to compensate Fluor for this loss is to take it into account when assessing the additional site costs incurred in July 2009 as a result of having a defective batch of MPs and TPs.

This part of the claim is graphically explained in paragraph 131 of Fluor’s Closing Note:

“The handling of MPs and TPs which could not be installed required very large additional storage areas and associated infrastructure (including crawler mats and gravel dykes, laying sand, gravel and specially made concrete supports for TPs). Fluor was forced to contract for additional self-propelled modular transporters (SPMTs) and cranes. The atmospherically-controlled warehouse was too small for the increased demands, and an additional warehouse was procured and air conditioning and atmospheric controls installed. Additional costs were incurred in respect of fork lift trucks and fuel for the additional generators required to provide power for the welding. As the number of subcontractors on site expanded dramatically, Fluor had to contract for additional storage, canteen and sanitary facilities to accommodate them and their materials, housed in a new contractor park. The division of the sums paid to VZT between works to MPs, TPs and site-wide costs (and various sub-categories of costs within those heads) is explained by Mr Ross in his first report.”

The sum claimed (on the basis that NCR 006 took effect from after 4 June 2009 and that the NCRs covered the TPs) is €2,343,287. However, after correcting for a possible currency error, the claim is broken down as follows:

MP storage. Although the figure for MP storage was not formally agreed, Dr Mastrandrea accepted in cross examination that a proper calculation of quantum to reflect the effects of the unwaived breaches should include the figures put forward by Fluor (Day 14/11-12). I therefore accept the figure of €101,924 claimed by Fluor.

Bode positioners and materials for roller sets.

The electrical earthing. This relates to electrical earthing in the MP storage area during period (C). This was disallowed by Dr Mastrandrea on the basis that it would have been incurred in any event once an NCR had been issued, a position from which he felt compelled to resile in cross examination. ZPMC now accepts that it would be wrong to exclude costs on the basis that they would have been incurred in due course when NCR was issued (paragraph 53 of its Closing Submissions). I therefore accept Fluor’s submission that the sum claimed is recoverable.

The MP repair dayworks. In cross examination Dr Mastrandrea agreed that some of the costs in respect of the dayworks should be allowed, but he was unable to say how much. On that basis, he allowed nothing. However, the experts have helpfully agreed the relevant daily rates for the appropriate periods. Mr Ross’s approach has been to consider the number of MPs being worked on during each operative period, which he has translated into a percentage of the agreed rate. For period (C) this is 52%, and for period (E) it is 19%. This seems to me to be a realistic approach and I accept it. Fluor is therefore entitled to recover the sums claimed.

Mammoet Cranes.

MP related costs. This claim is for the additional costs of storing MPs during the same additional period of weather related downtime. In principle, this seems to me to be a valid head of claim. In their Third Joint Statement, the quantum experts have agreed the figures for the cost of MP storage and associated dayworks (in respect of labour and equipment) for the months of August 2009 to December 2010. However, Fluor correctly points out that the relevant months must be October, November and December 2009 (because installation of the Phase 1 MPs was complete by the end of December 2009). In relation to MPs, the daily averages for each of these months were €25,411 (October), €25,422 (November) and €9,535 (December). Fluor adopts the average of these three figures, namely €20,065. For 7.5 days, therefore, this amounts to €150,488, and I consider that Fluor is entitled to recover the sum.

Sitewide costs.

I need say very little about this claim since most of the figures are agreed. It is a claim for testing by QIS. For period (A) Mr Ross has assessed the claim in the sum of €17,476, as against €14,867 put forward by Dr Mastrandrea (but as a figure only). Fluor says that the discrepancy between the figures results from a correcting invoice for the difference, which was issued by VDS under cover of a letter dated 22 June 2009. There did not appear to be any dispute about that.

The issue raised by ZPMC in respect of period (A) is that the testing was carried out on 4 June 2009 and therefore after the issue of NCR 006. Since I have now held that NCR 006 did not take effect until after 4 June 2009, this point falls away.

For period (C), Dr Mastrandrea has identified a figure of €63,677 in respect of NDT to TPs in Shipment No 2 - a figure which Fluor is prepared to accept.

Accordingly, Fluor is entitled to the sum of €81,153 in respect of this claim.

Fluor claims €23,795 for testing carried out by Sonovation up to and including 4 June 2009. ZPMC submits that about €11,000 is attributable to time when Sonovation was idle because it was awaiting equipment. Fluor submits that the waiting time was limited to about 11 hours out of 54 hours on site and that the claim is in respect of charges that were billed by Sonovation and paid. It submits that ZPMC should not be permitted to go behind this without very good reason, which it has not shown.

I am not persuaded that Fluor has paid any sums that were not due or that did not arise out of the defects in the welds. Accordingly, I find that Fluor is entitled to the sum claimed of €23,795.

It is common ground that SGS was engaged to provide NDT on 3 August 2009. In his third witness statement dated 23 December 2016, Mr Ayres said that this was to perform manual UT and Distance Gain Size (“DGS”) inspection, which Fluor believed would provide more accurate measurements of the indications being identified in the welds, which could then be used for the purpose of an ECA.

ZPMC disputes this. It contends that these costs either form part of the costs of pursuing the claim by way of arbitration against GGOWL or form part of the NDT required by the NCRs and have therefore been waived.

I do not recall the evidence of Mr Ayres about the reason for this testing to have been challenged, although given the limited time available for cross examination generally, that may be understandable. However, I see no basis on which it could have been challenged and so I accept it. In my view, this testing was not required by the NCRs and does not fall within the waiver.

Further, I do not consider that this cost can be regarded as part of the costs of pursuing the arbitration (or a potential claim against ZPMC) - no claim for privilege in respect of these tests (if there was one) has been upheld. Whilst I am prepared to accept that there was the possibility of litigation against someone in August 2009, I do not consider that the dominant purpose of this testing was for use in litigation. At best, I consider that was merely a possible purpose. By late July 2009 Fluor was beginning to reach the conclusion (Footnote: ³³ ) that the MPs and TPs were fit for installation without repair (or further repair) but it needed independent evidence to support it. Use in future litigation may have been a purpose for which SGS was instructed to carry out these tests, but in my view it was by no means the only purpose, let alone the dominant purpose.

In these circumstances, I consider that Fluor is entitled to recover the sum claimed, namely €85,314.

The costs which Fluor paid to VDS in respect of periods (A), (C) and (E) have been agreed, subject to one small point, in the sum of €584,993.

The point disputed by ZPMC relates to €7,277 in respect of the welding (not surface cleaning or grinding) for the Shipment No 1 MPs IGI 05 and IGJ 02. ZPMC submits that it is common ground that NDT only began on IGI 05 on 4 June 2009 and that there was no testing of IGJ 02 prior to the issue of the NCR, so it is inconceivable that any welding carried out to those two MPs could have related to cracking.

This seems to me to be a fair point, and so I will reduce the sum recoverable by Fluor under this head to €577,716.

There are three elements to this claim. The first is the cost of scaffolding used to enable Fluor to carry out MP repairs during periods (C) and (E). The second is the cost of internal scaffolding to the TPs to enable the necessary NDT to be carried out. The third is a claim for scaffolding costs in relation to SEPAM’s fit out work. This last claim is considered separately.

The difficulty facing Fluor in formulating the first two elements of the claim is that Mr Ross described the BIS invoices as complicated, lengthy, requiring translation (because they were in Dutch) and involved ancillary costs which were not attributable to particular MPs. As a result he did not feel able to extract the appropriate figures from the invoices and so instead he produced an average cost per MP and then multiplied that average cost by the total number of MPs in the shipments that were not yet the subject of an NCR.

Dr Mastrandrea, by contrast, has considered the BIS invoices in detail and extracted the costs referable to specific MPs, but where costs were not attributable to a particular MP they were excluded. Fluor submits that as a result ZPMC’s analysis omits any costs of matters such as general support, common supervision or the costs of related equipment.

I have to say that I do not regard either approach as providing a reliable indicator of the true level of costs. Whilst in these circumstances there is a temptation to split the difference, that in my view is equally arbitrary.

The figure that was eventually accepted by Dr Mastrandrea was €112,538 for all periods (at Day 14/157), but in its closing submissions ZPMC contended that this should be reduced by 10% to reflect the fact that scaffolding was also required for base scope works. This again is an arbitrary deduction.

The burden of proving the appropriate figure lies on Fluor, and it is unfortunate that it finds it too difficult owing to the nature of the invoices. In this particular instance, I do not consider that the approach adopted by Mr Ross is a satisfactory substitute. What I therefore propose to do is to allow this claim in the sum put forward by Dr Mastrandrea, €112,538, but without any further deduction.

Fluor submits that the need to carry out the additional NDT and consequent weld repair at Vlissingen required an enormous increase in manpower and therefore, in turn, necessitated a substantial dedication of management and supervisory resource which would not otherwise have been required.

On 19 June 2009 Fluor created a dedicated cost code on its timecard entry system which were said to be “associated with the NDT and repair works” (paragraph 201 (b) of Fluor’s Closing Note). In fact, the cost code was described as “MP & TP remedial work”, but no explanation was given in the Job Bulletin that set up this new code as to what was meant by that expression. On the face of it the cost code would include remedial work by way of out of roundness repair. Mr Ayres said that the new cost code was set up because he believed “that there would be a pending claim somewhere” (Day 9/169). He went on to say how difficult it was to get people to complete timesheets properly, let alone using the correct cost code (Day 9/172-173).

In his witness statement dated 23 December 2016, Mr Ayres said that the personnel performing work in Vlissingen “were advised to use this cost code as a means of tracking their time associated with the NDT and repair effort” (paragraph 7.2). In the next two paragraphs he went on to explain that later he went through the time records in Fluor’s accounting system in order to determine whether the time charges were appropriate and assigned to the proper cost code. Sometimes he had to adjust the allocation of an individual’s time to a particular code in order to reflect accurately what Mr Ayres knew that person to have been doing in Vlissingen. In addition, he identified individuals who had not recorded their time to the new cost code but whom he knew to have assisted in the “NDT and weld repair efforts”.

What this suggests is that the cost code was often not used correctly and that the method of correction was fairly rough and ready. I can find nothing in the evidence to suggest that any distinction might have been drawn between repairs that were the result of defective welding and repairs that were necessary to correct out of roundness. It seems to me that what Mr Ayres described as the “repair effort” could, in the minds of many operatives, have included any activity that constituted repair to an MP or TP for whatever reason.

I consider also that a significant amount of management time would have been required to deal with the base scope work in the absence of any problems with the welding. Take, for example, the case of Mr P I Hall: he is described as the quality manager prior to the arrival of Mr Mike King, when he became a quality superintendent under Mr King’s supervision. It appears from the spreadsheet prepared by Mr Ross that Mr King arrived during the first week in July 2009. During the last four weeks of July 2009 Mr Hall is recorded as having worked on MP and TP repair for 73, 64, 65 and 66 hours per week, respectively. So it looks as if all his working time, or almost all of it, for that period was being charged to MP and TP repair. This suggests that he spent no time, or virtually no time, during that month on the work which he had been initially employed to perform.

Although Mr Hall is not a typical case, I suspect that there is a number of individuals who have recorded some time against MP and TP repair that would have been spent in any event dealing with supervision of the base scope work.

In addition, and probably more significantly, I have no doubt that a considerable amount of management and supervisory time was spent on dealing with out of roundness issues. Shortly after the issue of NCR 006 the further NDT and repair work that Fluor was directed to carry out was clearly specified and everyone concerned knew what was required, even though doing it called for a lot of organisation. By contrast, out of roundness proved to be a much greater problem than had been anticipated and a ready solution was not available. The contemporary documents show that significant time and effort was put into the task of finding an appropriate method of repair. I would expect a manager or operative who had been dealing with out of roundness repairs to enter the time spent under the cost code for MP and TP repair - there being, so far as I am aware, no instruction to the contrary.

In these circumstances I consider that to take as the starting point, as Mr Ross has done, the total amount of hours for the MP and TP repair cost category is to overstate the claim. In my view, the starting figures have to be reduced to reflect work that would have been done in any event and time spent dealing with out of roundness.

By contrast, I do not see the logic of applying the VBA based approach adopted by Dr Mastrandrea. He says, paragraph 1.23 of his Note dated 6 July 2017 (references omitted):

“Fluor’s Claim 9 is founded on an express connection between headcount for NDT and repair, and on the individuals who make up the claim. On the assumption that all those individuals are properly allocated to that claim, in my opinion the VBA should achieve a reasonable overall evaluation for that group of individuals.”

I have to confess that I do not understand why Dr Mastrandrea says this. His percentage uplift for period (C) is, as before, 11.19%, but I consider that this is too low.

Doing the best I can, I consider that Mr Ross’s figures should be reduced by 30% to reflect the factors that I have mentioned. Accordingly, the figures at paragraph 33 of his Note dated 6 July 2017 should be revised as follows:

In its Closing Note Fluor invites the court, for the first time, to make a further award on a similar basis to the claim for the loss of the opportunity to install MPs during period (C). In my view, this would amount to double counting: the figures in the table above represent damages in respect of the additional costs actually incurred. To award a further sum on the basis of a scenario that never occurred would be to award damages for the same loss in two different ways.

Finally, ZPMC submits that this claim has been waived, but I do not agree: it is a claim for compensation based principally on the loss of opportunity in period (C). The sums recoverable are those set out in the table above.

SEPAM was the subcontractor responsible for the electrical fit out and final assembly of the TPs at Vlissingen. SEPAM’s work involved the installation of sensitive switchgear, communication racks, HVAC equipment, the transformers, ventilation and dehumidifying systems, cabling, navigation aids and ancillary steelwork.

Fluor paid SEPAM €276,776 in period (C), but this was paid pursuant to its existing contract with SEPAM: it was not paid as a result of NCR 006, although Mr Ayres said that he tried to put SEPAM employees to work in various capacities supporting the NDT and repair effort (witness statement dated 2 February 2017, paragraph 8.3). I have already concluded that it was reasonable of Fluor not to proceed with any electrical fit out work so long as there was a serious chance that the TPs would require NDT and consequent remedial work.

In my view, the result of the breaches of contract by ZPMC in relation to Shipment No 2 TPs, so far as the work done by SEPAM during period (C) was concerned, was that SEPAM’s work was disrupted with inevitable inefficiency and waste. There is no means of evaluating the extent to which the sums paid to SEPAM were wasted, it has to be a matter of assessment. It is clear that Fluor managed to keep SEPAM’s workforce occupied on work in relation to the project that was, I assume, mainly productive. Based on the evidence, the material in the contemporaneous documents and my own experience of disruption claims, and taking into account my conclusion at paragraph 294, I would assess that 25% of the sums paid to SEPAM during period (C) was probably wasted as a result of the disruption. However, the work involved the MPs of both Shipment No 1 and Shipment No 2: any loss in respect of productive work to Shipment No 1 MPs following the issue of NCR 006 either falls within the terms of the waiver or was base scope work within the terms of the original contract (for which there can be no recovery). According to Dr Mastrandrea, and I am not aware that it is disputed, the sum claimed in respect of period (C) does not relate to any period that precedes the week ending 19 July 2009.

Work on the TPs from Shipment No 1 was put on hold in mid-July 2009 (see the summary given by Mr Morgan at paragraphs 6.3.55 to 6.3.62 of his first report). Had there been no defects in the welds of the TPs in Shipment No 2, thereafter SEPAM would have been wholly deployed on the Shipment No 2 TPs but instead SEPAM was put to work on the MPs, building a mock-up TP and some exterior base scope work on TPs in order to prevent its workforce from being idle: see the first witness statement of Mr Eisman, at paragraphs 4.3 and 4.4.

Since the payments in respect of which this claim is made did not relate to any period before the week ending 19 July 2009, none of the sums paid to SEPAM for the remainder of period C can have been in respect of fit out work to TPs, as ZPMC suggests, so the payments made to SEPAM after this must have related to work on the MPs from Shipment Nos 1 and 2 and, possibly, some non fit out work to TPs. However, I find that one head of loss suffered by Fluor during this period was the loss of productivity that occurred as a result of the disruption to SEPAM’s work caused by Fluor’s reasonable decision not to employ SEPAM on TP fit out work during the second half of period (C), but instead to employ its workforce less efficiently on the MPs and other tasks. I assess this at €69,194 (€276,776 x 25%). Of the balance, €207,582, I am prepared to accept Fluor’s submission that about 50% of the work was in respect of MPs from Shipment No 2 (the other 50% attributable to Shipment No 1 MPs being irrecoverable). However, I consider that it is likely that some of this was in respect of base scope work and some of it was in respect of NDT or repair related work (Mr Kampman accepted that from the invoices alone he could not say which - at Day 5/72): of these, only the latter is recoverable. In the absence of any clear evidence, I can do no better than assess this split as 50:50, and so I consider that Fluor is entitled to recover a further €51,896 in respect of remedial work to MPs of Shipment No 2 during this period. That makes €121,090 in all.

In relation to period (E), after NCR 009, the total amount paid was €90,899. However, Fluor claims only 19% of this sum, namely €17,271. Adopting the same approach in relation to disruption, but allowing for the likelihood that the extent of it would be reduced as time went on and, I consider, by the beginning of period (E) to about 15%, Fluor is entitled to recover 15% of the total sum as wasted expenditure, namely €13,635. In addition, taking 19% of the balance (€77,264) and, again, adopting a 50:50 split, produces a further €7,340. This gives a total of €20,975. I appreciate that this is actually a little more than the sum claimed by Fluor but, as explained in paragraph 339 above, I have added an amount to reflect wasted cost through disruption, so the comparison is not like for like.

Fluor’s claim in respect of mobilisation or acceleration costs is dealt with below.

On 9 April, 5 and 7 May 2009, and before it became aware of the defects in the welds, Fluor entered into, respectively, a charter for the provision of a diving support vessel (The Relume) for 4 months, a contract for the services of divers for 120 days and a consulting services agreement for supervisors for the diving operations (on a daily rate). This charter and the contracts were subsequently extended from time to time until mid-2011.

At paragraph 7.5 of his witness statement dated 22 December 2016, Mr Frampton explained why these contracts were entered into. He said this:

“I was informed by Fluor management around August 2008 that we needed a dive vessel to accommodate and support divers to carry out the installation of J-tubes required to enable the IA cables to enter and rise up the MPs before the TPs were placed on top of the MPs. Of course, we also needed the actual divers, and specialist supervision of the divers.”

This evidence is confirmed by the terms of the contract for the diving services which, under the heading “Scope of Work” described the work to be performed as the installation of conventional J-Tubes. Although Mr Frampton’s evidence on this point was not accepted by ZPMC, as I understood it, I see no reason not to accept it and I do so.

Having mobilised on 27 May 2009, The Relume was not ready to carry out diving operations until 11 June, but of course no MPs had been installed by then and so the dive team could not start installing J-tubes; although at that stage Fluor was hoping that the delay before they could do so would not be very long. Mr Frampton said (at paragraph 7.6 of his witness statement) that Fluor then used the dive team on other work, such as diving on seabed anomalies that had been identified by a sidescan sonar. To this end, on 7 July 2009, Fluor entered into a contract modification with Red7Marine, the company that supplied the divers, for a remote operating vessel so that the divers could conduct seabed survey work. The first J-tube was not installed until 7 August 2009.

In the event, in addition to installing the J-tubes, the dive team continued to carry out seabed survey work and related activities for the duration of the project with the result, Fluor says, that it never commissioned a seabed survey - which it would otherwise have done, being a far less expensive option. In formulating this claim Fluor has given credit for the cost of such a survey in the sum of £2,013,300.

Fluor’s claim, put very shortly is that, in order to mitigate its loss caused by the presence of the defects in the welds and the inability to use the dive team without interruption to install J-tubes, it found other work for the dive team by using it for seabed survey work.

Fluor submits that it is entitled to recover the costs of the dive team that were spent on activities other than installing J-tubes (and any related work that formed part of its base scope work), after giving credit for a seabed survey that it did not have carried out, as damages for breach of contract. Fluor has prepared an analysis showing the number of days on which the dive unit was able to conduct base scope activities, such as installing the J-tubes, and those on which it carried out other activities not originally intended, such as seabed survey and similar work.

Before giving credit for an alternative survey (but after allowing 10 days of dive unit time that would have been associated with such a survey), the cost of the additional work carried out by the dive unit has been assessed at £2,821,513 for The Relume, £2,486,338 for the divers and £149,330 for the diving supervisors. After giving credit for the cost of the notional survey, the claim amounted to £3,443,881.

The claim as actually made in this action is set out in the first and second columns of the table below, but in the third column I have removed the amounts attributable to work that could not be carried out because of the issue of an NCR on the relevant shipment (being costs that have been waived).

The figures in the column headed “Amount” are those put forward by Fluor after an apportionment based on the number of MPs on which testing had been carried out prior to the issue of the relevant NCR. That is an approach that I have rejected, so what is left are the claims for Shipment Nos 2 and 3 which are said to be for periods (C) and (E), amounting to €654,514.

However, it is important to appreciate that the figures in the above table are based on the actual costs of the dive unit for each of the periods shown. To give an example of the arithmetic, I will take the period 5-28 June 2009. For this period the cost of the dive unit was £687,203 which, when divided by 14 (the number of MPs in Shipment No 1), produced an average cost per MP of £49,086. This figure was then multiplied by 6, being the number of MPs in Shipment No 1 “which had undergone testing through the date the applicable NCR was issued” (Footnote: ³⁴ ) , to give the sum of €294,516 in the second column.

In fact, there was a time lag between the installation of a particular MP and the subsequent installation of the J-tube for that MP. At the start of the installation of the Phase 1 MPs this was 5 days. However, as Mr Connolly explained, whilst it took an average of (he thought) 1.3 days to install an MP in favourable weather conditions, it took 2 days (he thought) in similar conditions to carry out the fitting of the J-tube. So the installation of the J-tubes would gradually drift further behind the installation of the MPs. Whether or not Mr Connolly’s estimates were accurate does not matter, because the general trend that he postulated was reflected in the programme: this showed that the first J-tube was installed about 5 days after the first MP, whereas the last J-tube was fitted about three months after the installation of the last MP.

Fluor’s claim for period (C) is based on the installation of all 14 MPs of Shipment No 2 between 29 June and 29 July 2009. However, a claim put in this way is unsustainable for two reasons. First, on any view the installation of MPs from Shipment No 2 could not have begun before 11/12 July 2009, and so the first J-tube could not have been fitted until a few days after that. Accordingly, the dive unit could not have been engaged on fitting J-tubes to Shipment No 2 MPs before mid-July 2009. Since the claim is based on the premise that the dive unit was carrying out survey work when it would otherwise have been fitting J-tubes, there can be no claim in respect of the Shipment No 2 MPs before mid-July 2009.

Second, I have concluded that if there had been no breaches of contract by ZPMC in relation to Shipment No 2, Fluor could only have installed seven MPs during this period. So, whichever way one looks at it, the claim (as presented) for the first 15 days of July 2009 has to be excluded (because J-tubes could not have been fitted to the Shipment No 1 MPs because by then they were subject to NCR 006). However, in reality if 7 MPs from Shipment No 2 had been installed between 11-29 July 2009, the J-tubes would have been fitted to those MPs between 20 July (the adverse weather would have prevented an earlier start) and early August 2009 - allowing 2 days per MP.

If it was Fluor’s intention to have a sidescan sonar survey of the seabed, the divers would not have been used on survey work (apart from the 10 days diving required to support that survey). On this basis I can see no reason why a claim based on the difference in cost between using the dive unit and a sidescan sonar survey is not, in principle, valid.

However, as I have indicated, it is plain that no such claim can be maintained in respect of Shipment No 1 because the dive unit was not mobilised until after NCR 006 had been issued. In relation to Shipment No 3, NCR 009 was issued before any MPs could have been installed, so the use of the divers on survey work (instead of fitting J-tubes to the Shipment No 3 MPs) on the days in respect of which the claim is made is a loss consequent upon the issue of NCR 009. It has therefore been waived.

That leaves Shipment No 2. ZPMC submits that Fluor cannot succeed unless it can show that it would have made a different decision in relation to the dive unit because of the progress of the Shipment No 2 MP installation during period (C). (Footnote: ³⁵ ) I think that this approach is too simplistic. The reason why the installation of the MPs from Shipment No 2 did not start on 11/12 July was the existence of the defects in the welds of those MPs. By this time the dive unit had been engaged and Fluor was committed to pay for it until about September 2009. Accordingly, Fluor’s options were either to let dive unit stand idle for several weeks or to use it for other work. There has been no suggestion that in opting for the latter Fluor acted unreasonably.

Fluor’s case that, if things had gone according to plan, it would not have used the divers to carry out survey work was supported, albeit somewhat weakly, by the evidence of Mr Connolly (see, for example, Day 6/134-135). Whilst it is true that there is no evidence that Fluor took any steps prior to June 2009 to arrange for a sidescan sonar survey, it is equally true that there is no evidence as to the date by which it would have been expected to do so (see the intervention by the court and Mr O’Sullivan’s response at Day 6/140-141). In these circumstances I do not consider that it would be fair to draw any inferences adverse to Fluor simply on the ground that there is no evidence of consideration being given to a sidescan sonar survey in early or mid 2009.

ZPMC’s case, advanced through cross examination (see, for example, at Day 6/136), was that in June 2009 Fluor had not given any thought to the possibility of carrying out an advanced sidescan sonar survey but simply assumed that the survey work would be done by divers. Since it is ZPMC who asserts this, it seems to me that it must be for ZPMC to prove it. In my view, it has not done so.

In these circumstances, I consider that Fluor has just made out a case that for some of the MPs in Shipment No 2 - seven on my finding - it reasonably used and paid for the dive unit to carry out survey work because no J-tube work was available. In relation to those seven MPs that was the result of ZPMC’s breaches of contract in relation to the welding, and not result of the issue of an NCR. Accordingly, that part of Fluor’s claim has not been waived.

In my view, Fluor is entitled to recover the additional cost of using the dive unit for 14 days to carry out survey work between about 20 July and early August 2009. The quantum experts have helpfully agreed a daily rate that reflects the net additional costs of using the dive unit for survey work in the sum of £25,527 per day. (Footnote: ³⁶ )

Accordingly, in respect of this head of claim Fluor is entitled to recover £357,378.

Gardline provided the Marine Mammal Observers who were required to accompany piling operations. Fluor entered into a contract with Gardline on 12 January 2009 for amongst other things, monitoring noise levels during piling operations and ensuring that mitigating actions were taken if marine mammals were seen.

I can take this claim very shortly because the parties have agreed a daily rate of £1,220 and, in addition, are agreed that it should be treated in the same way as the claim in respect of The Stanislav Yudin.

Since the period of delay that I have allowed in respect of The Stanislav Yudin is 14.5 days, Fluor is entitled to recover £17,690 for this claim.

As I have already mentioned, work to the TPs of Shipment No 1 was put on hold in mid-July 2009. I have already mentioned the minutes of the Weekly Progress Meeting held on 10 August 2009, in which it was noted that all SEPAM works on the TPs had been put on hold until further notice. Apart from a skeleton crew, SEPAM’s workforce was demobilised on 20 August 2009.

ZPMC submits that the demobilisation of SEPAM’s workforce was the result of the issue of NCRs 006, 008 and 009. Fluor submits, and I agree, that before the issue of NCR 009 it had been keeping SEPAM on site in the hope that there would be usable TPs to work on in the near future. It then says, at paragraph 326(b) of its Closing Note:

“The point at which this became impossible was the arrival of defective Shipment 3 TPs in Period E. Effectively; from the moment that this Shipment arrived (1 August 2009) the demobilisation became inevitable. On this basis, on a proper analysis, the cost is one attributable to a cause in Period E and crystallised shortly after.”

This paragraph echoes what Mr Eisman said in a witness statement submitted in the arbitration, dated February 2012, but recasts the emphasis. What Mr Eisman actually said in that witness statement, at paragraph 26, was this:

“At that point (end of July 2009), there was still hope that, when Shipment 3 arrived in Vlissingen, it would not be subjected to an NCR and there would be a supply of TPs on which SEPAM could begin the fit-out works. Ultimately, that proved not to be the case, because shortly after Shipment 3 arrived on August 1st, GGOWL issued NCR 009, which required re-testing and repair of the Shipment 3 MPs and TPs as well. At that point, we were left with little choice but to demobilise SEPAM, except for a skeleton crew . . . ”

The same point was made by Mr Dekker in the evidence he submitted in the arbitration, which I have already quoted above and which I consider accurately described the position at the time. For convenience I will set out again what he said (at paragraph 111) in his witness statement dated 27 February 2012:

“GGOWL’s issuance of NCRs 006, 008 and 009, and its demands that Fluor change its NDT procedure and retest and repair welds previously tested, had a profound and devastating impact on the Project. Not only did all offshore installation activities relating to the foundation works come to a halt, but even when we ultimately made the decision to cease the retesting and repair effort in Vlissingen, the construction schedule was destroyed. Once we decided to restart the construction phase, we had to remobilise contractors (such as SEPAM) and procure vessels to perform the work, or modify existing vessel contracts to account for the delays.”

To my mind it is an inescapable conclusion from this evidence, and the situation as a whole, that the demobilisation of SEPAM’s workforce in August 2009 was a direct result of the issue of NCRs 006, 008 and 009. The subsequent agreement by Fluor to pay acceleration costs to SEPAM had two causes. First, a recognition from the outset (May 2009) that late delivery of the TPs and increased secondary steelwork and coating works meant that some degree of acceleration of SEPAM’s works was required in any event. Second, the need to make up the time lost between June 2009 and remobilisation in October 2009.

The first of these two causes obviously had nothing to do with any breaches of contract by ZPMC. However, the second of them did. But since the second cause was also the result of the issue of the NCRs, any loss resulting from it has been waived.

For these reasons, I consider that this claim fails in its entirety.

These two claims related to costs incurred to other contractors, BIS and Abird, in support of SEPAM’s acceleration.

I can deal with them very shortly, because Fluor submits that they should be treated in the same manner as the SEPAM acceleration claim. Since I have rejected that claim in its entirety, these two claims must be rejected as well.

Brabant was engaged by Fluor to blast clean the TPs before and after certain fittings were welded to them, to design and apply corrosion protection coatings and, finally, to paint the TPs. It was engaged on a time and materials basis.

Brabant’s work was very sensitive to environmental conditions: for example, one cannot paint at temperatures below about 5°C and humidity needs to be controlled. As the work was pushed into the winter months, it was necessary to provide extra tenting, heaters, humidity control units and ventilation systems. In addition, Brabant was required to work and extra shift in order to improve productivity and this added a premium to the hourly rate.

Fluor claims the cost of the additional shift. At paragraph 91.5(a) of the Particulars of Claim, Fluor pleaded that the claim was made “in respect of the cost caused by the acceleration arising from the delay in starting this work and resulting performance in periods of adverse weather”.

Fluor submits that a proportion of this cost is attributable to events that occurred during the “unwaived periods”. I am not persuaded that this is correct; but if, contrary to this, I were to accept that the 23 day delay to the start of the TP fit out works that is said to have occurred in July 2009 during period (C) had some sort of knock-on effect on the work done by Brabant, then one has to ask what that effect would have been.

The period covered by the claim is 26 October 2009 to 10 May 2010. If the effect of the lost 23 days was to push back the work by the same period, then the result is that in the absence of that period of delay the work would have been completed in mid-April 2010. As Mr Ross accepted, from the point of view of weather conditions, this would not have made much of a difference (Day 13/88).

The reality of the situation, as it seems to me, is that the cost of introducing the additional shift was caused by the interruption of the works to the TPs that resulted from the issue of NCRs 006, 008 and 009. For this reason, this claim fails.

I have already discussed the principles which I propose to apply in relation to claims for delay, and so in the following sections of this part of the judgment I will deal in turn with the individual claims for delay related costs made by Fluor.

I have already dealt with the claim based on the loss of the opportunity to install MPs during period (C). In my view, any other way of putting the claim has been waived.

The Leviathan was the subject of the time charter for a period of 426 days from the completion of its mobilisation, which was 24 September 2009. The hire under the charter was $145,000 per day. It was intended to use The Leviathan for the installation of WTGs, but of course by September 2009 there was no possibility of installing any WTGs in the near future so the vessel was standing idle. She was therefore put to use doing other tasks including, as I have mentioned, the installation of TPs.

Fluor’s approach to the calculation of delay is as follows. Mr McKibbin has concluded that between 20 September 2009 (the loss of The Javelin) to 17 March 2010 there was critical delay of 127 days. Fluor submits (at paragraph 351(a)(vi)(3) of its Closing Note) that:

“Those days were lost because the defects in the TP welds meant that no TP installation vessels were able to complete the first TP String until 17 March 2010.”

This is based on what Mr McKibbin said in his first report, at paragraph 2.5.2:

“It was obvious to Fluor by mid-August 2009 that there would not be enough MPs actually installed to warrant a possible extension of the charter period of the TP installation vessel and its charter period ended on 19 September 2009. At this point the critical path moved into the TP installation.”

In addition, Fluor submits that a further 8 days was lost between 18 March and 7 April 2010 because of on-going delay and disruption to the TP installation. I have already given my reasons as to why I am not satisfied that this period of delay was the result of any breach of contract by ZPMC.

In my view, the conclusions that I have reached in relation to the reasons for not extending the period of operation of The Javelin apply equally to The Leviathan. The effective cause of the inability to use The Leviathan for the installation of WTGs in September 2009 was that the issue of NCRs 006, 008 and 009 brought all construction work to a standstill until Fluor decided at the end of September 2009 to go ahead and install the MPs (and TPs) whether repaired or not. As I have already mentioned, fit out works to the TPs from Shipment No 1 were put on hold in mid-July 2009, and fit out works to the TPs from Shipment Nos 2 and 3 were put on hold in early August 2009, so there was no prospect of there being any TPs available for installation until well after the period of hire of The Javelin was due to expire.

Fluor submits that it is common ground between the parties that the conversion of The Leviathan to enable her to install TPs was a measure taken to mitigate its loss caused by the unavailability of The Javelin. Thus, on the usual principles, and subject to the steps being reasonable, the expense of that mitigation would be recoverable. Fluor submits that cancellation of The Leviathan charter was not an option because it had no right to do so, and so put The Leviathan to use as best it could. In these circumstances, I find that it was reasonable for Fluor to use The Leviathan in the way that it did.

However, I have held that the costs consequent upon the loss of The Javelin have been waived. Similarly, the loss resulting from the fact that there was no possibility of The Leviathan installing WTGs in the autumn of 2009 was also a direct consequence of the issue of the NCRs, so that loss too has been waived. It must follow that the costs of any steps taken to mitigate those losses have also been waived.

On the assumption, which I have now held to be the case, that the NCRs extended to the TPs as well as the MPs, Fluor has formulated a delay case based on the assertion, which for present purposes I will assume to be correct, that the start of the work to the Shipment No 2 TPs was delayed by 23 days during period (C). That case (set out at paragraph 402 of its Closing Note) is as follows:

In my view this approach faces two difficulties. The first is that since any loss consequent upon the delay resulting from the unavailability of The Javelin - the 135 days - has been waived, it cannot be legitimate in principle to permit recovery on another basis which uses that 135 days as part of the formula to calculate the delay. The second difficulty is that I have already concluded that if there was 23 days delay to the start of the TP fit out work for Shipment No 2, it was not critical to the completion of the TP installation in March/April 2010.

In addition to the 26 days of critical delay, derived in the manner set out above, Fluor also seeks to recover damages for what it claims was a further 21 days critical delay to the installation of the MPs (see paragraph 409 of its Closing Note). This 21 days is derived from an analysis of the counterfactual scenario in period (C) which is set out at paragraph 83 of Fluor’s Closing Note. I have already dealt with this in some detail, but, by way of brief recapitulation, Fluor submitted that between 10 and 25 July 2009: The Stanislav Yudin could have installed 8 MPs, a further 3 MPs could have been loaded onto The Stanislav Yudin’s barge and a further 3 completed MPs could have been delivered to the quayside. Given an installation rate of 1.5 MPs per day, this, submitted Fluor, equated to a total of 21 days MP installation that never happened and were therefore wasted.

The fatal flaw in this formulation is that it confuses lost opportunity with critical delay. The counterfactual scenario for period (C) is an exercise to determine how many MPs could have been installed during period (C) if the contract had been properly performed, that is to say if the welding defects which gave rise to the issue of the NCRs had not been present. The loss to Fluor during that period was, in my view, the loss of the opportunity to install MPs in the seabed. Indeed, it has advanced a claim on that basis which has, in part, succeeded.

Whether or not that loss of opportunity also resulted in critical delay to the project requires a different analysis. Suppose that there had been no breaches of contract in relation to Shipment No 2, with the result that 14 MPs could have been installed during July/August 2009 taking 21 days. In the meantime, Shipment No 3 arrived and shortly afterwards was condemned by the issue of NCR 009 on 7 August 2009.

Assuming further that this 21 days caused 21 days of critical delay to the installation of the MPs, the result would have been that, in the absence of this delay, the last Phase 1 MP would have been installed on about 7 December 2009, instead of 28 December 2009 (when it was actually installed). Since at that stage the installation of the MPs was an activity that was in float, even with the 21 days delay, this would have made no difference whatever.

Fluor utilised a number of Windcat Mk 2 service boats, of which two, Windcats 7 and 8, were to be used in particular during the Phase 1 works. Fluor contends that the Phase 1 works were delayed by 232 days as a result of ZPMC’s breaches of contract. This 232 days consists of the 135 days that I have already mentioned, together with a further 97 days between 20 May and 19 September 2009.

I have already explained why the claim against ZPMC based on the alleged critical delay of 135 days between 20 September 2009 and 17 March 2010 cannot succeed. For the following reasons the claim for the earlier 97 days fails also. The loss consequent on the inability to install any MPs from Shipment No 1 before 1 October 2009 was the result of the issue of NCR 006 and has therefore been waived. Likewise with Shipment No 3 and NCR 009.

In relation to Shipment No 2, the delay caused as a result of the issue of NCR 008 on 29 July 2009 has also been waived. As I have explained in the previous section of this judgment, the claimed loss of opportunity to install 14 MPs of Shipment No 2 did not cause any critical delay to the installation of the Phase 1 MPs.

Fluor claims the same 47 days of critical delay to the Windcat boats (21 days in respect of period (C) and 26 days representing the contribution to the TP delays). Accordingly, this claim fails for the same reasons as the claim in respect of The Leviathan.

Fluor claims the costs of its management, supervisory and Home Office staff being deployed to the project for the same 47 day period of recoverable delay (21 days for MPs, 26 days for contribution to the TP delays).

This claim fails for the same reasons as the previous claim.

This is another claim based on the same 47 days of alleged critical delay, and so it fails for the same reasons as the previous two claims.

This claim is for an additional sum paid to BBC Chartering to defer the start of Phase 2. BBC Chartering provided Fluor with a general vessel equipped with craneage which was to transport the WTGs from the Siemens Wind Power facility at Esbjerg (in Denmark) to Vlissingen and Harwich.

This is another claim based on the same 47 days of alleged critical delay, and so it fails for the same reasons as the previous three claims.

The claim is for $2,705,375 and ¥727,884.

There is a dispute as to precisely what costs form the subject matter of this claim. ZPMC submits that the claim is in truth for work which Fluor would have had to carry out in any event but which has been dressed up as a claim for additional work. Fluor submits that the claim is for additional supervision and coordination of ZPMC’s work on Shipment Nos 4 to 8 that it had to carry out in order to ensure that the defects which had been found in Shipment Nos 1 to 3 did not occur again. Fluor emphasises that this head of claim is not based on the existence of defects in Shipment No 4 to 8: on the contrary, submits Fluor, the expenditure was necessary to ensure that such defects did not occur.

In his first witness statement in this litigation, Mr Hans Ho said this (omitting references):

“4.2

The way the Project was originally resourced in China - shown in the organisation chart dated 24 September 2008 - reflected the intention that Fluor would monitor ZPMC’s work at a high level. Accordingly, Fluor initially planned for a team of 12 personnel in total, comprising a welding engineer, a project coordinator and a team of inspectors from SQS, reporting to Jason Estabrook, the SQS Manager. In particular, ZPMC was responsible for quality control (“QC”) and Fluor would carry out quality assurance (“QA”) “spot checks”.

4.3

In the event, there were numerous and ongoing issues with the quality of ZPMC’s work so that it became necessary for Fluor to increase its QA team. Fluor progressively ramped up until, at its peak, the QA team numbered 132 inspectors from SQS (see the organisation chart dated 19 September 2009).”

One would have thought that this, standing by itself, was clear enough. However, in its Amended Particulars of Claim, dated 12 February 2016, under the heading “Construction Management and Supervisory Personnel to Oversee NDT in Shanghai” Fluor pleaded this, at paragraph 83.1:

“In response to the discovery of defects in the welds in the MPs and TPs in Vlissingen, it was also necessary for Fluor to commit significant internal and subcontracted resources to undertake further NDT, as well as additional QA supervision/coordination, in respect of weld activities in Shanghai.”

ZPMC questioned this and asked Fluor to explain how the costs claimed under this head of claim were caused by the breaches of contract identified by the court in the Liability Judgment. In response, Fluor stated that it had to incur the costs claimed “to support further NDT and QA supervision/coordination in Shanghai in respect of Shipments 4 to 8”. By repeating the reference to “further NDT” Fluor might be taken, as I think ZPMC thought, to be referring to the actual testing itself, rather than just the work of supervising it. However, a little later in its response Fluor referred to the further NDT on Shipments Nos 4 to 8 performed by ZPMC and to the increased resources that Fluor engaged to supervise and coordinate this effort, as described in Mr Ho’s first witness statement.

These responses provoked a further request by ZPMC, in which, amongst other things, it complained that Fluor had failed to particularise to what this “further NDT” was attributable. Fluor then said this:

“As must be apparent to ZPMC, the further NDT carried out on Shipments 4 to 8 included particularly D scans on welds. ZPMC carried out D scans, and Fluor incurred expense in overseeing that, because it had become apparent to ZPMC and to Fluor that E scans were insufficient to detect transfers defects in welds.”

This, I regret to say, was not a model of clarity either. However, when read in conjunction with the relevant paragraphs of Mr Ho’s first witness statement (to which there was a reference in the immediately preceding section of the answer) I consider that it was clear that the costs which Fluor was claiming were costs of supervising NDT, and not carrying it out.

ZPMC submits that there is a “conceptual problem” with Fluor’s approach, in that it amounts to saying that ZPMC should have to “pay for the fact that ZPMC, and as a result, Fluor, carried out D-scanning of S4-S8” in circumstances where D scans involved no variation of Fluor’s contract with GGOWL or of ZPMC’s contract with Fluor. Accordingly, ZPMC submitted that the performance or supervision of D scanning fell within the base scope work which Fluor would have had to carry out in any event.

In my view, this is an attempt by ZPMC to distort Fluor’s case in order to rebut it. I accept that if Fluor had set out its case rather more clearly on this aspect, ZPMC would not have been presented with this opportunity. However, in my view it is reasonably clear that the claim that Fluor is making under this head is one for the costs of additional supervision of the base scope work. I would have no hesitation in accepting ZPMC’s point if all that had happened was that Fluor had used the same number of NDT supervisors to check ZPMC’s performance of D scanning instead of E scanning. But that is not Fluor’s claim. What Fluor is saying is that in the light of the history of poor welding of the MPs and TPs in Shipment Nos 1 to 3, it had to take additional steps to supervise the NDT of Shipment Nos 4 to 8. As I have already indicated, I find that it should have taken no longer to carry out D scans than it did to carry out E scans. Accordingly, the change in scanning method should not, of itself, required any greater amount of time be spent on supervision. What required the additional supervision was the history of poor workmanship in relation to Shipment Nos 1 to 3.

In my judgment, Fluor’s response was a reasonable, proportionate and reasonably foreseeable response to the situation given the sums at stake. It could not take the risk of further welding defects in Shipment Nos 4 to 8.

ZPMC has an additional point. It submits that references to “increased extent” of NDT do not advance matters because “everyone accepts” that D scanning led to transverse indications being identified, which in turn led to repairs and then further NDT. The submission, in my view, misses the point. This claim is not concerned with the further NDT that would have to be carried out by ZPMC in order to carry out the base scope work, which is to what ZPMC appears to be referring, but rather the change in regime from one of spot checking ZPMC’s work to one of checking 100% of it.

Mr Ho explained in cross examination (at Day 10/41) that by “spot check” he meant a level of 2-3%, as opposed to the 100% checking that was subsequently carried out. It was clear from Mr Ho’s evidence (at Day 10/46-50) that the claim was for the cost of the additional labour required to achieve this, and I accept his evidence.

At paragraphs 22.1 to 22.113 of his Quantum Report in the arbitration, dated 1 May 2012, Dr Mastrandrea went through this claim in some detail. He reduced the sums claimed by Fluor in respect of outside subcontractors from $2,705,095 to $2,322,019, the larger part of his reductions concerning increases in rates for which he could see no justification. I am not satisfied that in this situation this is really the province of a quantum expert: there is no suggestion that otherwise the rates are unreasonable in themselves. However, Dr Mastrandrea has raised a point in relation to business tax, which appears to have been charged by some subcontractors and not others. He has reduced the sum claimed by about $47,850 to reflect this.

In his fourth witness statement, dated 22 December 2016, Mr Ho confirmed that Fluor had spent the sums claimed on additional NDT carried out by the relevant subcontractors. That evidence I accept. The figures claimed included the adjustment of an arithmetical error identified by Dr Mastrandrea in his May 2012 report, so its contents had been noted by Fluor. However, in a Quantum Schedule prepared for the hearing of the Counterclaim in the arbitration Mr Ross commented on the application of business tax by saying that there were instances where subcontractors rightly included taxes in the invoices. He said that the circumstances were reviewed by Fluor in advance of making any payments. No-one, so far as I am aware, has since refuted this explanation. In these circumstances, I do not propose to make the reduction suggested by Dr Mastrandrea.

In relation to particular employees of Fluor, the claim includes the time of Mr Ho himself and Mr Jason Estabrook. It seems to me that both of them, who featured on Mr Ho’s original organisation chart, would have been working on the project during the relevant period in any event, and so in each case I am only prepared to allow 25% of the sums claimed (at Appendix 19A: $36,684 and ¥323,384, less 75%: $27,513 and ¥242,538) to reflect the additional work to which they were put as a result of the increased NDT.

A further small adjustment was noted by Mr Ross in respect of costs attributable to Shipment No 3 in the sum of $7,943 (see paragraph 28.3 of his Quantum Reply Report). The claim advanced by Fluor for its own labour must therefore be further reduced by that sum to $2,669,919 ($2,705,375 - $27,513 - $7,943) and ¥485,346 (¥727,884 - ¥242,538).

This is a claim for the costs of maintaining the TPs during the period immediately following commissioning by Fluor of each WTG and prior to the issue of a WTG Foundation Works Commissioning Certificate (“WFWCC”) by GGOWL in relation to that WTG and its foundations. These costs arose because GGOWL refused to issue WFWCCs when Fluor said it should have done on the grounds that the welds in the MPs and TPs were defective. The evidence of Mr Ayres, which was not challenged on this point, was that GGOWL refused to issue WFWCCs only in relation to those structures where the MP or TP had not been repaired and in respect of which Fluor had refused to carry out offshore NDT of the MP/TP in response to GGOWL’s letters of 30 April and 23 December 2010.

Fluor was contractually responsible for maintaining each TP until the issue of the WFWCC in respect of the relevant structure. (Footnote: ³⁷ ) Fluor asserts that GGOWL’s refusal to issue the WFWCCs was the result of ZPMC’s breaches of contract in relation to the welding and, indeed, was a reasonably foreseeable consequence of those breaches.

This particular claim contains a major wrinkle. Not only are the sums claimed as direct costs, but also there is an alternative way of putting the same claim - on the basis that this was one of the claims included in the settlement of the arbitration and so its value should be taken into account when considering the cost of that settlement to Fluor. In this section of the judgment I will deal with the claim on its merits as a direct cost. I will address the alternative claim, and its implications, later in this judgment.

ZPMC submits that either (a) Fluor was in breach of contract and that was the cause of GGOWL’s refusal to certify, or (b) GGOWL was in breach of contract in issuing the instructions and/or refusing to certify with the result that Fluor would have had a claim against GGOWL. In neither case, says ZPMC, can a claim for damages succeed against ZPMC.

In addition, ZPMC submits that the loss is too remote or excluded by Article 8.3 of the Terms and Conditions of Purchase attached to the Purchase Order.

I consider that the first argument raised by ZPMC must be approached in stages. First, I am satisfied that Fluor’s decision on 1 October 2009 to install the MPs and TPs was a step that mitigated its loss caused by ZPMC’s breaches of contract. The fact is that if Fluor had continued with the testing and repair regime required by the NCRs, which in turn was caused by ZPMC’s breaches of contract, the loss it would have sustained would have been vastly greater than it was. Viewed with the benefit of hindsight - since there is now no suggestion that the MPs and TPs are not fit for their purpose - and, indeed, probably viewed from Fluor’s position at the time given that it believed the MPs and TPs to be sound, the decision to install the piles was an act of mitigation. By acting as it did, Fluor avoided a substantially greater loss. I find it difficult to see how anyone could assert that Fluor’s decision was unreasonable, even though it may have involved a risk.

However, although Fluor took the decision to install the MPs and TPs in the light of internal and expert engineering advice, it did not at that stage have an ECA in a form that could be presented to GGOWL, or indeed to any regulatory authority.

The arbitral Tribunal found that by manufacturing MPs and TPs with such extensive cracking in the welds Fluor was in breach of contract. It was in my view a reasonably foreseeable, if not an inevitable, consequence of the breach that GGOWL would insist on testing and repair until (a) such time as either all the MPs and TPs were repaired or (b) a convincing and authoritative ECA was produced by Fluor which demonstrated that the MPs and TPs were good for their service life.

For present purposes I will leave open the question whether the effect of such an ECA would have altered the fact that both Fluor and ZPMC had been in breach of contract, because there was no such ECA. But, assuming the MPs and TPs to have been sound as Fluor believed, the consequence of the decision to go ahead with installation meant that Fluor’s breaches would cause no further long term loss to GGOWL.

The second stage is to consider the consequences of Fluor complying with the demand to carry out the offshore testing required by GGOWL. In principle, it might have been able to recover the costs of it from ZPMC as a consequence of ZPMC’s breaches of contract. However, I have no doubt that any claim for such costs would have been met by the retort from ZPMC that Fluor had failed to mitigate its loss, since the testing was not necessary and that is what Fluor believed.

I accept that it was a likely and foreseeable consequence of Fluor’s decision to go ahead and install the MPs and TPs that GGOWL would protest, but in my view that does not alter the fact that by acting as it did Fluor mitigated the loss caused by ZPMC’s breaches. If GGOWL’s protest continued to the extent of refusing to issue the WFWCCs for so long as Fluor did not comply with its instruction to carry out the offshore NDT, it seems to me that Fluor’s refusal to comply with the instruction was merely a further step in mitigating the loss - even though it was likely, or at least foreseeable, that the refusal to carry out the testing demanded by GGOWL would result in the WFWCCs being withheld.

I reach this conclusion because I have no doubt that the cost of the further testing required by GGOWL, and the likely demand for further repair in the light of the results of that testing, would have greatly outweighed the likely cost of maintaining the TPs for the period during which GGOWL continued to withhold the certificates. If Fluor was right, as we now know it was, sooner or later it would have been able to persuade GGOWL that the MPs and TPs were good for their service life (with or without some form of warranty).

As I have already mentioned, it is not disputed that Fluor was obliged to pay for the maintenance of the TPs until the WFWCC was issued for the relevant structure. But not only was that the case, but I consider that it was also the case that this was foreseeable. Assuming that ZPMC had no knowledge of the terms of the contract between Fluor and GGOWL, it could have been expected to know that such contracts typically require the issue of a certificate when completion of a particular structure or, sometimes, part of the structure, is achieved. It would also be expected to know that if there was a serious complaint about the quality of the structure as built, the employer might refuse to take it over and withhold the relevant completion certificate. In that situation the contractor would normally remain responsible for the preservation and, if appropriate, maintenance of the structure until it could be handed over. I find that all those things would have been known to ZPMC, or at least would have been reasonably foreseeable to a sub-contractor in its position.

Having reached this far, the third stage is to consider whether the costs incurred in mitigation were reasonable - which I consider to be the test: see, for example, Compania Financiera “Soleada” SA v Hamour Tanker Corp Inc [1981] 1 WLR 274, per Templeman LJ at 285E-F.

For these reasons, subject to the effect of Article 8.3 of the Terms and Conditions of Purchase, I consider that Fluor is entitled to recover from ZPMC a reasonable sum in respect of the costs of maintaining the TPs during the period for which GGOWL continued to withhold the WFWCCs. I reject ZPMC’s “heads I win, tails you lose” argument to the contrary. My assessment of the reasonable sum is in the next section of this judgment.

Further, if I am wrong about my conclusion in relation to mitigation of loss, I would hold, for the same reasons, that GGOWL’s withholding of the WFWCCs was a foreseeable consequence of ZPMC’s breaches of contract. These conclusions dispose of ZPMC’s point based on remoteness.

I now turn to ZPMC’s argument based on Article 8.3. The article provides as follows:

“Indirect and/or consequential damages, or whatever the cause may be shall be explicitly excluded from this Purchase Order.”

In my view, there is a short answer to ZPMC’s argument based on this article. Fluor’s claim for the costs of the NDT and remedial work required by the NCRs was not a claim for damages that can be described as indirect or consequential: on the contrary, in the ordinary course of events two of the most direct types of loss that result from the sale of a defective article are either the diminution in its value or, if reasonable, the cost of carrying out the necessary remedial work (which may include the cost of investigating the defect). If, in order to avoid one or other of those direct losses, the purchaser mitigates his loss by taking some alternative steps, the cost of such steps would in my view be equally recoverable (subject, of course, to reasonableness): I do not see how steps reasonably taken in mitigation could be described as indirect and/or consequential damages.

Since I consider that Fluor’s action in installing the MPs and TPs without further investigation and repair was a step that mitigated its loss, the reasonable costs of taking that step must be recoverable provided that the loss avoided was a loss of a type that was both direct and foreseeable. I therefore reject the argument based on Article 8.3.

The sum claimed is £699,364, which includes general maintenance, electrical maintenance, TP Re-certification and a small sum, about £8,500, in respect of other works. The overall sum claimed includes 17% for overheads and profit (this being a rate agreed with GGOWL). This is one claim that has not been discussed between the quantum experts in the course of this action, so there is no measure of agreement as to its value.

Mr Ross has conducted a review of all these costs and has concluded that the sum claimed is reasonable, subject to a small adjustment which reduces the figure to £698,006.

In his Reply Report on Quantum, at paragraphs 42.17 to 42.108, Dr Mastrandrea makes a number of detailed points, in most cases raising a doubt about the sum claimed or the adequacy of its verification, but does not at any point suggest alternative figures, and concludes by saying that he is not able to arrive at an evaluation of this claim “with reasonable certainty”. He makes the point that there is a potential for overlap between the work that is the subject of this claim and the post-handover punch list works that Fluor would have had to carry out at the same time. This is true, but Fluor seems to have done its best to apportion the time between the different activities.

Fluor’s retort to this last point is that in the arbitration GGOWL had a very similar claim and adopted the same methodology of apportioning the extent of the use of the vessels and operatives to the different activities, and that this was an approach that Dr Mastrandrea was prepared to support. In these circumstances I propose to accept the figures put forward by Fluor unless ZPMC can point to particular items where there has been duplication: otherwise I can see no reason to reject Fluor’s figures.

A particular difficulty with the approach adopted by Dr Mastrandrea is that he does not suggest any alternative figures for any of the individual heads of claim. That may be quite understandable, but it does not go very far to assist the court. I am very reluctant to disallow or make reductions in respect of items where Dr Mastrandrea has declined to do so. Further, and more significantly, in the course of the arbitration Dr Mastrandrea agreed the figures for this claim as figures (that is to say, that he accepted that the sums claimed had been incurred, without accepting that ZPMC was liable for them), although at that stage the figure agreed by the experts was £478,882, not the sum presently claimed. I have no way of knowing which figures that were put forward in the arbitration were the same as the figures put forward in this action: all one can say is that there was clearly a considerable overlap.

Dr Mastrandrea took issue with a sum of £87,050, that formed part of the claim for the Anchor Point costs up to 31 January 2012, on the grounds that it was based only on an estimate provided by Siemens. Dr Mastrandrea said that this should have been at least an invoice for that sum and, preferably, proof of payment. The estimate from Siemens dated 2 March 2011 was for the re-certification of equipment on 10 TPs and equated to £1,674 per TP. Fluor applied this rate to all 52 TPs in Shipment Nos 1-3, which produced the figure of £87,050. In a report prepared for the counterclaim in the arbitration, dated 12 March 2013, Mr Ross said that he had been directed to certificates issued by Siemens’ sub-contractor, Lift-Rite, in relation to this claim. The trial bundle for this action contains some 200 pages of invoices, reports and certificates issued by Lift-rite relating to numerous TPs. Having said in his Reply Report dated 10 April 2017 (at paragraph 42.77) that this claim should have been capable of being corroborated by relevant invoices, which is perfectly reasonable, Dr Mastrandrea later (at paragraph 42.85) referred to the fact that he had seen invoices - presumably from Lift-rite - issued in July and August 2011 (and therefore in the period ending on 31 January 2012). However, he did not say at that point that he had looked at these invoices or at other documents in the trial bundle and found them wanting: instead what he said was that he had not seen the Lift-rite contract in order to confirm the daily rates which had been charged.

I have to say that I do not find evidence of this sort at all helpful. I accept that the claims in this action and in the earlier arbitration are of enormous complexity and that the two quantum experts have done a vast amount of work: between them they must have produced several thousand pages of expert’s reports and joint statements. Nevertheless, I would have thought it would have been possible for Dr Mastrandrea, in this particular instance, to say whether or not the invoices and other documents that he had seen supported the sum claimed or, if not in full, to what extent. He could then make a comment to the effect that he had not seen a contract in order to justify the rates claimed or that the costs appeared to be duplicated with another claim, and qualify his opinion to that extent.

At this point it is perhaps worth observing that Fluor is the UK subsidiary of a very sophisticated international contractor and, as the evidence shows, has the benefit of sophisticated systems for approving and placing orders, processing invoices, making payments and so on. It is inherently unlikely that, apart from the odd isolated instance, Fluor would have paid sub-contractors for work that had not been done or without having satisfactory documentation to justify doing so. I can more readily accept that Fluor, like many other litigants in this type of dispute, may be advancing claims in respect of work that in truth it would have had to have carried out anyway or where there is duplication or overlap between claims for different items, but that is a different thing and the court must be astute to ensure that any award of damages takes such matters properly into account.

Reverting to the claim under consideration, Dr Mastrandrea also raises the question of why in some cases work has been done at variation rates and not at cost. Rates for varied work are often higher than the basic rates in a tender for several reasons: for example, because the work may be carried out under disruptive conditions, it may involve disproportionate set up costs or it may simply be that the contractor hopes to recover through variations a level of profit that he was not prepared to include in his tender. Without going into the matter in fine detail I see no reason to assume - in the absence of demonstrated examples - that the rates claimed by Fluor are inflated. However, I have removed 15% to reflect the profit element of the mark up (see below), and that may go some way to meeting Dr Mastrandrea’s point that a profit element may have been included twice.

A point that Dr Mastrandrea makes that seems to me to be a good one is in relation to two items, one of which is a claim for £16,300 for personal protective equipment (“PPE”) and the other of which is for “SE and Permit issue” in the sum of £9,085; in each case the sums claimed are for items which are stated to have been already included in the relevant sub-contractor’s labour rate (see paragraphs 42.50 and 42.45). I therefore propose to disallow these sums.

I do not know how the 17% mark up for overheads and profit has been made up: it is a rate that was agreed with GGOWL and which was applied to most of the claims for varied work. The evidence of Mr Johnston in the arbitration showed that Fluor’s bid for this contract included a margin of 8.2% on the anticipated project revenue. In his first report in the arbitration, dated May 2012, Dr Mastrandrea concluded that this gross margin was made up of 2% in respect of overheads and 6.2% which he regarded as gross profit (paragraph 32.30). I can see no reason why Fluor should not be entitled to recover the cost of its general overheads as part of the costs of steps taken in mitigation, but I have some difficulty in seeing why it should be entitled to recover in addition an element in respect of profit. It is entitled to recover its costs of such steps, not their value.

Taking Mr Ross’s figure of £698,006, and then stripping out the 17% for overheads and profit leaves £596,586. Deducting the two sums set out in the penultimate paragraph leaves £571,221, to which must be added 2% for overheads - producing a figure of £582,645.

Whether or not Fluor is entitled to recover this sum is a question that I address in the next section of this judgment which deals with the settlement of the arbitration between Fluor and GGOWL.

GGOWL’s counterclaim, served in March 2012, was summarised in its opening submissions in the Arbitration as follows:

“What is the recoverable quantum of (a) the historical costs and (b) the future monitoring and testing costs?

GGOWL’s case is that it is entitled to (a) the historical costs of £19,868,927.61 plus €546,224.58; and (b) the costs of the proposed monitoring and testing regime at a cost of £44,233,706 (net present value) as damages by reason of:

(a)

Fluor’s breach of contract in manufacturing the structures with weld defects;

(b)

Fluor’s breach of contract in failing to repair the defects discovered in Vlissingen or to produce ECAs to justify non-repair;

(c)

Fluor’s breach of contract in installing the structures offshore in October 2009 in a defective state;

(d)

Fluor’s breach of contract in prohibiting GGOWL from carrying out further testing of the structures from August 2009 onwards;

(e)

Fluor’s breach of contract in failing to carry out offshore inspection and testing of the structures as instructed by GGOWL or to agree to joint testing.”

It alleged also that, if the MPs failed, the cost of replacing 10 MPs would be about £165 million and the cost of replacing all 32 unrepaired MPs would be about £370 million.

It seems that it was not effectively in issue that GGOWL owed Fluor £42.65 million (plus VAT) in respect of unpaid Milestone Payments and Variation Orders. The real issues between the parties were (a) the sum that should be deducted from this amount and, consequently, the amount of the net balance that would be payable by Fluor to GGOWL or by GGOWL to Fluor and (b) what form of warranty or indemnity Fluor should give in relation to the costs of future monitoring of the MPs and a failure of one or more of the unrepaired MPs.

The final result, as recorded in the Settlement Agreement, was that:

In these circumstances it seems to me to be clear that GGOWL’s claims in respect of future monitoring and potential pile failure were addressed principally by the extended warranty given by Fluor under the Settlement Agreement. On the face of the agreement, that did not involve any money changing hands but was a contingent liability assumed by Fluor.

Leaving aside the fact that this appears from the way the agreement as a whole was structured, Mr Ross had advised Fluor that the claim for future monitoring costs was vastly overstated. He had advised that, if an appropriate discount rate was applied, the appropriate figure would be £28 million. I regard it as most unlikely that the adoption of the same figure as the cap on the warranty is coincidence: it is far more likely that GGOWL was prepared to accept the force of Mr Ross’s point.

Fluor’s case is that the £32.325 million that was described as having been paid in respect of “Counterclaim Damages” was exactly that and represented a payment in respect of GGOWL’s claims for past and future monitoring of the MPs, which was put at about £65 million in all, so that it effectively settled those claims for about 50% of their stated value. This, submitted Fluor, was self-evidently a reasonable settlement. Further, Fluor submitted that it gave up (or sacrificed) various other claims that it had against GGOWL which, Fluor asserted, had a value of about £20 million (or £30 million - its position changed during the litigation). Since Fluor received a net payment under the settlement of about £10 million, its pleaded case was that it gave up about £20 million in respect of the sacrificed claims. But if this is correct, it means that Fluor in practice paid £52.325 million to settle a claim stated to be worth about £65 million. However, that does not of itself mean that the settlement was not reasonable.

Accordingly, Fluor says that it is entitled to recover approximately £50 million from ZPMC as the cost of settling GGOWL’s counterclaim in the arbitration.

ZPMC challenges this analysis at almost every point. First, it submits that the expression “Counterclaim Damages” was no more than a label attached, at the instigation of Fluor, to the sum of £32.325 million which was to be paid by Fluor under the agreement. ZPMC submits that when one looks at the claims that were really in issue when the settlement was reached an entirely different picture emerges. ZPMC is not suggesting that the settlement was not a bone fide settlement, but it does submit that no reliance can be placed on the expression “Counterclaim Damages”.

Mr Brannigan, by contrast, submits that its meaning is perfectly clear and that effect should therefore be given to it.

It is settled law that, in principle, C can recover from a contract breaker, B, sums that it has paid to A in settlement of a claim made by A against C in respect of loss caused by B’s breach of its contract with C.

However, C’s settlement with A must be an objectively reasonable settlement and, if it is, that sum represents the measure of C’s damages in respect of B’s breach of contract (assuming that there were no other heads of loss). Even if C can show that its settlement with A was at an undervalue, the settlement sum still represents a ceiling on the amount that it can recover from B.

This was established by the decision of the Court of Appeal in Biggin v Permanite [1951] 2 KB 314. The judgment of His Honour Judge Hicks QC in Royal Brompton Hospital v Hammond (No 1) (1999) 66 Con LR 42, which refers to the full transcripts of the judgments in the Court of Appeal, shows that, contrary to assumptions made in some other cases, liability was in fact in issue at the time when Biggin settled the claim by the Dutch government, albeit that Biggin regarded its prospects of success as virtually nil.

Biggin v Permanite was a straightforward case of a settlement between a buyer and a seller in respect of defective goods, in that case adhesive for roofing felt, which had been sold to the seller by the defendant: in other words, it was a claim for damages involving a single head of claim being passed down the contractual chain.

The position is more complicated where several heads of claim have been settled between, say, an employer and a main contractor but where the defendant subcontractor is alleged to have been responsible for only one of them.

This type of situation was considered by His Honour Judge Thornton QC in the difficult case of Bovis Lend Lease v R&D Fire Protection (2003) 89 Con LR 169. Bovis, as main contractor who had agreed to construct a shopping centre, settled a claim brought by the employer, Braehead. The dispute involved numerous claims and cross claims. R&D had been engaged by Bovis to complete one aspect of the fire protection works which were alleged to be defective. Another subcontractor, Baris, who had carried out the rest of the fire protection works sued Bovis for various additional sums it said were owed to it under its subcontract. Bovis counterclaimed for the cost of remedying work carried out by Baris which was alleged to have been defective.

However, Bovis produced no evidence of the circumstances surrounding the settlement with Braehead or as to the apportionment or breakdown of the sum that it had paid. In those circumstances, the court held, on a trial of preliminary issues, that in a multi-party or a multi-issue dispute, the court would have to decide (if it could) what part of the overall settlement was attributable to the relevant breaches of contract of a particular defendant. Any such allocation had to be reasonable and would then represent a ceiling on the damages payable by the relevant defendant.

Bovis took the position that it was entitled to recover the costs of remedying the defective work irrespective of the settlement. Further, it took this position in circumstances where there was no evidence that the defective work in question would be remedied. The court held that, so long as Bovis was contending that it suffered a loss by entering into the settlement caused by breaches of contract by the defendants, the burden was on Bovis to establish what that financial detriment was.

In reaching this conclusion His Honour Judge Thornton QC followed the approach taken by the Court of Appeal in Townsend v Stone Toms (No 2) (1984) 27 BLR 26. In his judgment in that case Oliver LJ said this:

“That [the payment into court] has to be taken into account in some way seems to me to be beyond doubt and it is, of course, always open to a plaintiff who wishes to accept a payment in or thinks that its acceptance may cause him some embarrassment in the matter of apportionment to request an amendment of the notice of payment in to apportion the sum paid among the causes of action in respect of which it is paid. In the ordinary way that would, I think solve any difficulty, though I would like to reserve the position - which of course does not arise here - where there may be grounds for asserting that the apportionment is collusive or not made bona fide.

Where, however, the party who has to bring the money into account himself provides no material to show how any apportionment should be made (or, as in this case, invites the judge to deal with it in a particular way) the judge has to do his best with what material he has, and the only material he had in this case was the claims themselves. What he had to ascertain was what the plaintiffs had lost, and to what extent that loss had been mitigated or satisfied by what had been received. There was really no other reliable way of doing this except by assessing the true value of the plaintiff’s claim against [the contractor], and comparing it with the £30,000 received.”

This case raises rather similar problems, not because Fluor has refused to explain how the settlement sum is to be apportioned, but rather because ZPMC does not accept the explanation given. Three questions are raised in this case. First, is the description “Counterclaim Damages” conclusive as to the subject matter of the payment by Fluor of the £32.325 million? Second, if not, what were the claims settled by that payment and what amounts should be apportioned to them? Third, how is that apportionment affected by the fact that Fluor gave up (or sacrificed) some other claims of its own against GGOWL?

There is a very helpful judgment of Colman J in General Feeds Inc, Panama v Slobodna Plovidba Yugoslavia [1999] 1 Lloyd’s Rep 688. In his discussion of the exercise which the court must do he said, at page 692:

“Unless the claim is of sufficient strength reasonably to justify a settlement and the amount paid in settlement is reasonable having regard to the strength of the claim, it cannot be shown that the loss has been caused by the relevant eventuality or breach of contract. That is not to say that unless it can be shown that the claim is likely to succeed it will be impossible to establish that it was reasonable to settle it. There may be many claims which appear to be intrinsically weak but which, common prudence suggests should be settled in order to avoid the uncertainties and expenses of litigation. Even the successful defence of a claim in complex litigation is likely to involve substantial irrecoverable costs. It is thus an every day event for shipowners or their P&I clubs to settle cargo damage claims based on allegations of bad stowage or unseaworthiness for well under 50% of the claim when the alternative explanation of the damage is the inherent condition of the goods or some other cause for which owners are not liable. Unless it appears on the evidence that the claim is so weak that no reasonable owner or club would take it sufficiently seriously to negotiate any settlement involving payment, it cannot be said that the loss attributable to a reasonable settlement was not caused by the breach by reason of which the goods are in a damaged condition.”

That was a case where the owners settled a claim by the cargo insurers for US$ 600,000, although it was likely that the cause of the damage to the cargo was not one for which the owners were responsible. This represented 25% of the value of the claim. Although the arbitrators held that to settle for 25% was reasonable in the circumstances, they concluded that the total damage sustained was contributed to by the failure of the owners to take proper steps for the care of the cargo once it was seen to be overheating. They concluded that, had it not been for this, the claim by the cargo insurers would have been for US$ 1.6 million, and not US$ 2.4 million. Accordingly, the arbitrators concluded that the amount of the settlement sum attributable to the breach by the charterers was US$ 400,000 and that was the amount they awarded. Colman J upheld the award.

ZPMC submitted that the law was accurately represented by the following proposition, taken from Foskett on Compromise, at paragraph 31-63:

“Where a settlement agreement does allocate sums to particular claims or causes of action, that will not necessarily be conclusive. The court will look behind the form or words used in a global settlement to ascertain the underlying position and to ensure that third parties’ rights are not unjustly affected.”

That passage was based on the judgment of HH Judge Humphrey LLoyd QC in Durabella v Jarvis (2001) 83 Con LR 145. There was a dispute between a contractor and a sub-contractor. The contractor, Jarvis, had entered into a settlement agreement with the employer on terms which expressly stated that no value was included in the settlement for the sub-contractor’s works. There was a “pay when paid” clause in the sub-contract and so the contractor asserted, as against the sub-contractor, that it had not been paid anything in respect of the sub-contractor’s works with the result that nothing was due to the sub-contractor. The settlement agreement had been drafted by counsel for the employer but included the terms required by Jarvis as set out in an e-mail from its solicitors. The Judge said that the relevant clause singled out Durabella’s work “and, exceptionally, places a figure on it, solely for Jarvis’ purposes, as Mr Ward had to admit.”

The Judge, unsurprisingly, was very critical of the contractor’s conduct and held that the relevant clause in the settlement agreement was a device intended to mislead readers as to the real transaction and that it had no evidential value. He considered that the issue of whether or not the contractor had been paid anything in respect of the sub-contractor’s works was a question of fact for the court to decide.

In my view, the task of the court when answering the questions posed above in the context of this case is to establish what proportion of the sum paid (or foregone) by Fluor (a) was attributable to breaches of contract by ZPMC and, of those, (b) was in respect of costs or delays which had not been waived. Having established that, the court must then consider whether or not that proportion of the settlement sum was reasonable, taking into account that Fluor had assumed a very substantial potential liability under the warranty - albeit one the risk of which was thought to be low.

It is clear that the settlement was intended to embrace all claims relating to the project that GGOWL had against Fluor and that Fluor had against GGOWL: therefore it seems to me that a necessary preliminary step is for the court to establish exactly what claims were, to use Fluor’s words, “on the table” at the time of the settlement.

Fluor reminded me of the authorities on the construction of contracts that were cited in the Liability Judgment. These are, of course, relevant when the court is concerned with finding the true meaning of a contract as between the contracting parties. Here the question is that of establishing the reality of the position as against third parties, which in my view is not the same exercise. The point can be illustrated by a simple example. Clause 11.1 of the Settlement Agreement provides that the parties will not in any future claim or proceedings “seek to re-open or challenge any decisions, awards or findings of the Tribunal in relation to the Proceedings”. Whilst the meaning of this, as between GGOWL and Fluor, is self-evident, ZPMC has not submitted, rightly in my view, that this prevents Fluor from challenging a decision of the Tribunal in litigation against ZPMC. Such a challenge may, in certain circumstances, amount to an abuse of the process of the court: but that is a question of fairness, justice and public policy, not of the true construction of the contract made between Fluor and GGOWL.

ZPMC submitted that when approaching the question of whether any part of the sum paid in settlement was in respect of costs that had not been waived attributable to breaches of contract by ZPMC, the court was entitled to take into account the prior negotiations between the parties. Fluor submitted that the authorities on the construction of contracts, to which I have just referred, did not permit this. In my view, that submission is misconceived. The extent to which prior negotiations can be taken into account when considering the terms of a contract is clearly established by the authorities. However, I do not consider that what is known as the exclusionary rule has any application in a situation where the court is considering whether a sum paid in a settlement between A and B, or part of it, was paid as a direct result of breaches of contract by a third party, C. When considering a global settlement, such as this one, the court is entitled, and in my view bound, to look at all the material available to it. Apart from anything else, this is a matter of elementary fairness. A third party, in this case ZPMC, should be liable only for the direct consequences of its breaches of contract, and not consequences that are the product - or said to be the product - of an agreement between two other parties into which it had no input.

By a Request for Arbitration dated 3 March 2011 Fluor referred to arbitration its claim against GGOWL in respect of the testing and remedial work that resulted from the issue of the NCRs. By its First Partial Award dated 13 November 2012 the Tribunal held that the adoption of the so called extra-contractual testing was an appropriate and necessary change of technique which did not involve any variation of the BOP contract between Fluor and GGOWL. Accordingly, Fluor’s claim against GGOWL failed.

GGOWL did not serve a counterclaim when it served its defence in the arbitration but instead reserved its right to do so at a later stage when it had completed its investigations into the structural integrity of the unrepaired MPs. Accordingly, the First Partial Award was concerned with GGOWL’s defence to the claim in respect of the extra-contractual testing and was concerned only indirectly with the question of the structural integrity of the unrepaired MPs.

A hearing of the counterclaim issues was set for April and May 2013. On 15 February 2013 GGOWL served particulars of the quantum of its counterclaim. In summary, this included past monitoring costs of about £20 million, future monitoring costs of about £44 million and a claim for the cost of replacing some of the unrepaired MPs/TPs (£165 million for 10 structures or £370 million for 32 structures).

A few days later GGOWL effectively abandoned its claim for the cost of repairing any of the MPs/TPs and instead advanced a claim for an indemnity against failure of the unrepaired MPs and TPs; however, it continued to pursue its claims for monitoring costs carried out to date (about £20 million) and future monitoring costs (about £44 million). Once this development had been communicated to Fluor via GGOWL’s solicitors, GGOWL and Fluor embarked on tentative negotiations through a mediator, Mr Eric Green.

A Fluor internal e-mail dated 16 February 2013 attached a list of Fluor’s outstanding claims which, according to its legal counsel, Mr Pepe, “would have to be resolved if there is to be a truly “global settlement” in this mediation”.

On 20 February 2013 GGOWL produced a table of issues entitled “Mediation Scope”. This sought a payment of about £66.3 million together with an indemnity against failure of the unrepaired piles. The figures put forward by GGOWL in the table were as follows:

Of these claims, GGOWL was clearly on strong ground in relation to its legal costs of defending Fluor’s claim in the arbitration: GGOWL was the effective winner in the first round of the arbitration, as reflected in the First Partial Award. Fluor’s liability for those costs would clearly be a substantial sum. Another seemingly strong claim was the Esbjerg claim which arose out of Fluor’s failure to collect the WTGs from Siemens Esbjerg in accordance with an agreed schedule, which gave Siemens a claim against GGOWL. This claim was apparently settled by GGOWL for £7.5 million, a figure representing about half the sum claimed (I am not clear why in subsequent negotiations GGOWL included it at only £7 million). Fluor’s case in the arbitration was that this liability was a consequence of the NCRs, and therefore not one for which Fluor could be liable. However, as ZPMC has pointed out, this was a position that became difficult to sustain in the light of the First Partial Award. On 4 March 2013 Fluor noted, in relation to this claim, that it should be “rejected at this value”: I do not know on what basis Fluor considered the settlement to be too high.

It is worth noting that at this point GGOWL was putting forward additional claims related to the project - items (6) to (9) - totalling £12.7 million. So far as the Siemens Delay Claim and the Lock off claim were concerned, Fluor considered that these should be rejected in their entirety or as a matter of principle. The claim for WTG storage at Harwich was accepted in principle, subject to substantiation. Fluor considered that the other disputed project claims (£6 million) should be rejected.

Subsequently, when it was proposed that any settlement should involve a clean break, GGOWL made it clear that the settlement would have to take into account the costs of completing the work under the BOP contract that remained incomplete. There was also the question of the legal costs of GGOWL’s counterclaim: its costs were said to be just over £5 million and Fluor had incurred costs of about £2.3 million. Although at this stage Fluor was reasonably optimistic about its prospects of defeating GGOWL’s counterclaim, this was not certain and I find that ultimately each party was prepared to bear its own costs of that part of the arbitration.

There were four major items of unfinished or disputed work: the transformers, the Tekmar seals, the Met Mast and various undisputed snagging items. Fluor was prepared to offer £2.95 million to settle the claims in relation to the Tekmar seals and the Met Mast, but nothing in relation to the transformers (on the ground that the contract would be novated).

In the meantime a similar claims marshalling exercise was being carried out by Fluor, whose employees were directed to produce a final and all-inclusive list of claims to be put before GGOWL in the mediation. A high level meeting was arranged for 28 February 2013 in order to discuss them. Several were claims that had never been submitted to GGOWL in the past.

On 7/8 March 2013 the mediation took place. Mr Pepe took detailed notes, running to about 20 pages, but I do not consider it necessary to refer to them in any detail because on 8 March 2013, following a second day of mediation, GGOWL summarised what it said were the final positions of the parties as follows:

“GGOWL last offer

1.

GGOWL to pay £41.6M for outstanding milestone payments (inclusive of interest) plus £1.5M for agreed VOs.

2.

Fluor to provide an Extended Warranty in the form of an extension to GGOWL’s rights under the BOP Contract.

3.

Fluor to pay £21M for inspection and monitoring costs (to be paid as incurred, on a 75/25 basis as between Fluor and GGOWL, capped at £28M at 2013 prices)

4.

Fluor to complete the remediation of the Transformers, Tekmar Seals and Met Mast Defects plus other undisputed Defects and snagging

5.

Fluor to pay GGOWL £30.35M to settle GGOWL claims for: legal costs (10.85), Siemens claims (7.0), other project claims (2.5) and historic Weld Dispute costs (10.0)

6.

GGOWL to pay Fluor £9.75M to settle Fluor claims for: Harwich (2.95), Adverse weather (1.8) and other Project claims (5.0)

Fluor last offer

1.

GGOWL to pay £41.6M for outstanding milestone payments (inclusive of interest) plus £1.5M for agreed VOs.

2.

Fluor to provide an Extended Warranty in the form of an extension to GGOWL’s rights under the BOP Contract.

3.

Fluor to reimburse GGOWL up to 100% of any inspection and monitoring costs incurred up to a cap of £28M on the occurrence of any valid claim under the Extended Warranty.

4.

The parties walk-away on their Project Claims.

5.

Fluor to rectify the Transformers, Tekmar Seals and Met Mast defects and GGOWL to complete the balance of defects and snagging in return for a payment from Fluor of £10M.”

The figures put forward by GGOWL, when quantified to include the monitoring costs, reflected a position where only about £1.5 million would change hands but Fluor would provide an extended warranty in respect of the unrepaired MPs and would complete the items of work identified at paragraph 4.

Fluor’s last offer shows that it was looking to receive about £33 million (if GGOWL agreed to complete the work) but, in return, would give an extended warranty and accept a contingent liability to pay monitoring costs of up to £28 million in the event of a pile failure.

GGOWL’s position, three weeks later, was that it accepted a liability to pay Fluor £42.65 million owed under the contract and that it would deal with future monitoring costs and potential pile failure by means of the extended warranty and monitoring costs (subject to the £28 million cap). However, it contended that its claims for the cost of completing the work, its legal costs of defending Fluor’s claim in the arbitration and its other project claims should be settled for a similar amount so that no money would change hands. GGOWL indicated that its legal costs and project claims amounted to about £20 million. At that time (early March) Fluor’s internal assessment of the costs to complete the project was about £20 million. However, as noted above, in its “last offer” Fluor offered to pay GGOWL £10 million to complete the work under the BOP Contract.

On 9 April 2013 the Fluor team prepared a draft e-mail to the mediator, Mr Eric Green, setting out Fluor’s position. The first paragraph dealt with the payment of the balance due under the contract of £42.65 million, the next three paragraphs dealt with the terms of the extended warranty and paragraphs 4-9 were as follows:

“4.

Fluor will pay GGOWL the sum of £26 million in full satisfaction of all GGOWL’s claims, including but not limited to, the legal fees, Siemens Delay Claim, Historic Costs, Esbjerg Claim, WTG Storage, etc

5.

GGOWL will assume responsibility for completion of all work remaining on the Project, including but not limited to, warranty claims and snagging matters and Fluor shall have no further responsibility [or] liability for such work.

6.

. . .

7.

. . .

8.

The parties will cooperate in the structuring of the final settlement agreement so as to satisfy their respective internal accounting, tax, and other requirements.

9.

Except as provided herein, neither party will have any obligation whatsoever . . .”

On 10 April 2013, in line with the draft, Fluor offered to pay £26 million in satisfaction of GGOWL’s claims, including its legal fees, the Siemens Delay Claim, the historic monitoring costs, the Esbjerg claim and the claim for WTG storage. This would have left Fluor as a net recipient of about £16.65 million.

However, at this point a sticking point emerged. This is what became known as the extrapolation argument. GGOWL’s position was that, in the event of a claim under the warranty as a result of the discovery of a weld defect in one MP, GGOWL would have the right to demand replacement of all the unrepaired MPs. Unsurprisingly, this was completely unacceptable to Fluor.

In the end the extrapolation argument was resolved in Fluor’s favour, but GGOWL then demanded a reduction in the net payment to Fluor. On 28 April 2013 GGOWL prepared a draft settlement agreement which included two important provisions. Clause 9 covered the “Fluor Settlement Payment”. The figure was left blank but expressed to be:

“… in full and final settlement of the following:

.1 all legal costs incurred by GGOWL in relation to the Claims and the Proceedings;

.2 [the Siemens claims];

.3 all costs incurred in relation to [the disputed defects and snagging items set out in Schedule [ ] of this agreement;

.4 [all historic costs]”

Clause 10 provided for a “GGOWL Settlement Payment”. Again, the figure was left blank but was expressed to cover:

“.1 the Fluor claim for delay for the Harwich work stoppage . . .

.2 [adverse weather]

.3 [other Project claims].”

What is significant about this draft agreement is that it shows clearly that GGOWL was still approaching the settlement in broadly the same way that it had done prior to the emergence of the extrapolation issue. In spite of Fluor’s forceful submissions I find that there is nothing in the draft agreement to suggest that there had been some profound change in the structure of the proposed settlement, as Mr Bruno asserted.

By 2 May 2013 Fluor and GGOWL were close to settlement. The dispute about the terms of the warranty had been resolved and there was a recognition by both parties that each would have to move about another £8 million to bridge the gap. In fact, Fluor was not prepared to go quite this far and was proposing a payment to GGOWL of about £32 million. That was effectively what happened, because under the Settlement Agreement Fluor’s payment was increased to £32.325 million.

I have already mentioned that it had an unanswerable claim for its costs of the arbitration up to the First Partial Award. These had been put forward originally in the figure of £16.3 million, but applying a one third reduction in order to reflect the likely recovery, this figure had been reduced to £10.85 million. That was a solid claim and I cannot see any ground as to why any further discount would be appropriate.

Similarly, and for the reasons I have already given, the Siemens Esbjerg claim was also a strong one, given the decision of the Tribunal in the First Partial Award. The claim of £10 million for historic weld dispute costs, which I understand was largely made up of historic monitoring costs, was sound in principle - since the presence of the weld defects had been found to be a breach of contract - subject to any argument about the quantum. This claim was originally put forward at a little under £20 million and so I would not have thought that £10 million was an unreasonable figure to place on it at the time of the mediation.

The so-called “other project claims”, on which GGOWL had placed a value of £2.5 million, are not susceptible to any informed evaluation on the material before me. For the purposes of a settlement I would have expected Fluor to attribute a fairly modest value to those other claims (bearing in mind that they had already been reduced in value from £12.7 million). In all, I would regard the settlement value of the claims being advanced by GGOWL as just under £30 million. (Footnote: ³⁸ ) In addition, once it was agreed that GGOWL would take over the completion of the work, that obligation had a value. At the beginning of April 2013 internal e-mails still suggested that Fluor estimated the costs to complete to be of the order of £20 million. However, this figure was put forward at a fairly early stage and the work was continuing as negotiations proceeded. Mr Bruno said that at the time of the settlement he understood (from Fluor internal documents) that the cost of completing the work was between £3 million and £4 million. I am prepared to accept this as a minimum figure, but not as one that would have been regarded as reasonable by GGOWL, who were, after all, about to assume the risk.

When Fluor opened its case in relation to the claims “sacrificed” in the settlement, it submitted that the claims could have been settled for “at least the £20 million pleaded by Fluor” (see page 211 of its Opening Note). The same assertion was repeated at several points in its Closing Note. But, as ZPMC pointed out, this conflicted with the case pleaded, by way of re-amendments in January 2017, in which Fluor alleged that the value of the claims sacrificed was “at least £30 million”. This shift in position over the period of some three months does not inspire confidence. I now turn to the principal claims relied on by Fluor.

The Harwich work stoppage claim arose out of a tragic fatal accident during the loading of a turbine blade when the root end frame became detached and swung down killing one person and seriously injuring another. Fluor has advanced an arguable case that, as between itself and GGOWL, the latter was liable for the consequences of the accident because it contracted to ensure that its subcontractors carried out their work safely and competently. However, ZPMC has produced a compelling argument to the contrary (see section A2, paragraphs 9-20, of Appendix 2 to its Closing Submissions). I do not consider that it is either necessary or appropriate to go into the detail of these arguments, because no one came close to doing so during the mediation. So far as I can tell, the issues were dealt with using a much more rough and ready approach. Balancing these factors, and putting myself in the shoes of a hypothetical assessor in GGOWL’s camp at the arbitration, I consider that Fluor’s Harwich claim would have been seen as having about a one third prospect of success.

The claim, as formulated by Fluor, was substantial - some £20 million. Fluor regarded the strength of this claim as being a major plank of its submission that its sacrificed claims were worth at least £20 million. However, it is worth noting that the actual period for which Fluor was effectively prevented from carrying out installation works was from 21 May to 14 June 2010, some 3.5 weeks. Yet the claim is based on a period of delay of 63 days, some 9 weeks. At first blush, therefore, I would expect the hypothetical assessor in GGOWL’s camp to conclude that it looked substantially inflated. If that person had looked a little more closely into the claim, he or she would then have realised that the claim had in fact been advanced on a prospective basis, and not in the light of what actually happened. This would simply reinforce the suspicion that the claim was inflated.

Looking at the position very broadly, I would expect this hypothetical assessor to regard the quantum of this claim is probably being worth about 50% of its stated value. Applying a discount of two thirds to that reduced value (to reflect my assessment of the prospects of success), would leave a settlement value of between £3-3.5 million.

The second largest claim is for the costs occasioned by adverse weather, although prior to the mediation it had never been submitted to GGOWL. Fluor’s witnesses were unable to provide any satisfactory explanation for this. It was said to have a value of nearly £18 million. ZPMC’s principal criticism of this claim is that under the terms of the BOP Contract two conditions had to be satisfied in order to make a claim based on adverse weather and that this claim only purported to satisfy one of them, as Mr Connolly readily admitted.

Fluor’s response to this is that it is only necessary to satisfy both conditions if the claim being made is one for delay; it submits that it remains open to it to make a claim for any other form of costs occasioned as a result of adverse weather by satisfying the first condition only. But even if this is correct, and I regard it as no more than arguable, Fluor has not advanced its claim in this way. It has simply added up the number of days when there was adverse weather and, by implication, asserts that on each of those days the relevant vessels would have been productively employed. However, Fluor has not attempted to show this. It is by no means self-evident that this would have been the case: indeed, from what I have seen of this case generally it seems that from time to time vessels were not productively employed for reasons unconnected with the weather (for example, because of the need to carry out repairs or maintenance).

ZPMC makes the further point that it is impossible to tell the extent to which this adverse weather claim may overlap with other claims. For the purposes of illustration, I can take an example from the present litigation: Fluor’s claim for the loss of the opportunity to install Shipment No 2 MPs during July 2009. That claim is made up in part of additional days that would have been lost through adverse weather as a result of having to carry out installation during winter months instead of during the summer. Any claim against GGOWL that had been put in a similar way would likely have involved duplication with the adverse weather claim.

In my view, there are some fairly formidable stumbling blocks in the way of this claim, which might explain why Fluor had never previously submitted it. Far from having a significant settlement value, as asserted by Fluor, I consider that it would have been regarded as little better than a nuisance claim. I would regard its settlement value as being between £2–2.5 million.

These two claims had between them a stated value of £38.178 million. The stated value of all Fluor’s sacrificed claims was said to be £42.229 million (paragraph 103 of the Re-Amended Particulars of Claim), so the remaining claims had between them a value of about £4 million. (Footnote: ³⁹ ) These included the TP maintenance claim, which I have assessed as being worth £582,645 (on a full liability basis). In its “last offer”, made on 8 March 2013, GGOWL was prepared to attribute a value of £5 million to the remainder of Fluor’s project claims (although at that point the adverse weather claim was valued at only £1.8 million). However, I consider this was both mistaken - claims with a net value of about £4 million could not possibly have a settlement value of £5 million - and in any event overgenerous: I consider that the settlement value of these residual claims, including the TP maintenance claim, was about £2.5 million, and certainly no more.

Taking all these considerations into account, admittedly on a very broad brush basis, I do not consider that Fluor’s claims had a settlement value much in excess of £8.0 million, taking the mid-point of the range for the two major claims (£3.25m + £2.25m + £2.5m). GGOWL’s allowance for those claims of £9.75 million was overgenerous.

The relevant terms of the Settlement Agreement were as follows:

“1.

DEFINITIONS

…

“Claims” means (in each case whether past, present, future, known and/or unknown, but subject to clauses 3.4 and 12.1) all and any actions, causes of action, claims, liabilities, obligations, claims for costs or expenses (including but not limited to legal fees and disbursements), damages, losses and demands of whatever nature which relate to and/or arise out of the Project and/or out of the Proceedings;

…

“Proceedings” means the legal action commenced by Fluor against GGOWL under claim number 111822 in the LCIA and the counterclaim brought by GGOWL against Fluor under claim number 111822 also in the LCIA;

“Project” means the project for the design, construction and operation of the Greater Gabbard Offshore Wind Farm;

…

3.

SETTLEMENT

3.1

The parties agree to settle the Claims and Proceedings and any further obligations and/or liabilities arising in relation to the BOP Contract and the Project, on a full and final basis on the terms of:

3.1.1

this agreement; and

3.1.2

the Extended Warranty Deed.

…

5.

PAYMENTS DUE FROM GGOWL TO FLUOR

5.1

GGOWL shall pay Fluor (collectively the “GGOWL Payments”):

5.1.1

the sum of £41,600,000 (forty one million, six hundred thousand pounds) including interest but excluding VAT (the “Milestone Payment Sum”); and

5.1.2

the sum of £1,050,000 (one million and fifty thousand pounds) including interest but excluding VAT (the “Variation Order Sum”) and;

5.1.3

the sum of £3,009,198 (three million, nine thousand and one hundred and ninety eight pounds) being the VAT payable on the sums in sub-clauses 5.1.1 to 5.1.2 (inclusive) above

5.2

GGOWL shall make the GGOWL Payments subject to and in accordance with clause 7 of this agreement.

6.

FLUOR PAYMENT

6.1

Fluor shall pay GGOWL the sum of £32,325,000 (thirty two million, three hundred and twenty five thousand pounds) (the “Counter Claim Damages”).

6.2

Fluor will pay the Counter Claim Damages subject to and in accordance with clause 7 of this agreement.

7.

PAYMENT ARRANGEMENTS

7.1

The parties acknowledge and agree that within 60 days of this agreement:

7.1.1

GGOWL is due to pay to Fluor the GGOWL Payments …

7.1.2

Fluor is due to pay to GGOWL the Counter Claim Damages …

7.2

The payments in clause 7.1 can be set off against each other, such that the only payment to be made within 60 days of this agreement is the payment by GGOWL to Fluor of £10,325,000 … plus VAT of £3,009,198 …. (the “Net Payment”).

…

8.

COMPLETION OF OUTSTANDING WORKS AND DEFECTS

8.1

GGOWL shall following the date of this agreement assume responsibility for the Works (as defined in the BOP Contract) and for completion of any outstanding works relating to the BOP Contract (the “Outstanding Works”) and for the remedying of any outstanding defects in the Works (as defined in the BOP Contract) (the “Outstanding Defects”) and for the completion and continuing operation of any obligations owed to Competent Authorities with jurisdiction or authority in relation to the Project.

…

9.

EXTENDED WARRANTY

Upon the date of this agreement the parties shall each deliver to the other the Extended Warranty Deed, fully executed and unconditionally released.

…

11.

WITHDRAWAL OF THE PROCEEDINGS

11.1

The parties acknowledge the decisions, awards and findings of the Tribunal on all matters decided by the Tribunal in relation to the Proceedings. The parties shall not in any future claim or proceedings seek to re-open or challenge any decisions, awards or findings of the Tribunal in relation to the Proceedings.

11.2

The parties agree within 14 (fourteen) days of the date of this agreement to issue a letter to the Tribunal appointed in relation to the Proceedings confirming that the Tribunal is released from its appointment, and that no further award will be made by the Tribunal in relation to the Proceedings.

11.3

Each party agrees to pay 50% (fifty per cent) of any outstanding fees and expenses of the Tribunal, the LCIA, IDRC and other costs of the Proceedings such as transcription costs, that are not covered by the payments on account of the Tribunal’s fees and expenses that have already been made to the LCIA. The parties agree that will be no order as [to] costs in relation to the Proceedings.

12.

RELEASE

12.1

In consideration of the terms and conditions of this agreement and the terms of the Extended Warranty Deed, the parties agree to waive and unconditionally and forever release each other, their parents, subsidiaries, affiliates and associate companies (included but not limited to their respective directors, officers, employees, agents, successors, assigns and heirs) from the Claims, from the Proceedings and from any further obligations and/or liabilities arising in relation to the BOP Contract and the Project, but any such waiver and release does not take away the rights of the parties to enforce (i) the obligations set out in this agreement (ii) the obligations and warranties set out in the Extended Warranty Deed and (iii) the indemnities set out in clause 17.1 of the BOP Contract.”

It was the evidence of Mr Bruno that the parties entered into a “global settlement”, and that part of that settlement included potential future and past monitoring costs (Day 8/177). A little later, I asked him this question:

“And no doubt if other numbers hadn’t been right, then the extended warranty just by itself wouldn’t have been acceptable [to GGOWL] because it had to be supported by other payments or settlements across the board. Is that right?

A My lord, you hit it right on the head. The reason why the amounts seemed to go up in terms of cash was because they were not satisfied with the amounts contingent in the warranty.”

A little later, Mr White asked Mr Bruno this (at Day 8/189):

“Q So Fluor’s position was that it only wanted a mediation if there was a complete global settlement of all outstanding issues, correct?

A A global settlement is what I desired as well.

Q This would include all of its possible claims?

A If we could sweep up and offer our claims and their claims as well as the litigation risk and put a price on it that was acceptable to GGOWL, that would be a good settlement.”

Although Mr Bruno was prepared to accept that the earlier offer by Fluor to pay £26 million was to be in full and final settlement all GGOWL’s claims, including but not limited to those for legal fees, the Siemens Delay claim, and so on, he was adamant that the £32 million paid by Fluor was paid specifically in respect of the counterclaim and excluded any other items such as attorneys’ fees. His reason was this (Day 9/48):

“Because of the occasions, circumstances and developments that occurred in the middle of April.”

This answer was a reference back to events that took place shortly after 9 April 2013 which he had earlier described (at Day 9/6-7) in the following terms:

“A There is a process under which it was increased to 32 million, which is the important ingredient that is missing from the April 9th communication.

Q What is that ingredient?

A There was a recognition, a realisation, some time in the middle of April, that what was contemplated with respect to the warranty was much different, if not completely different, by Gabbard and Fluor. We had to have the warranty with issues of trigger and breadth such that it would not be valued currently or it might trigger a loss. Gabbard’s position was that there was going to be an easy access to the 28 million by virtue of an easier trigger, and that it would be extrapolated across the entire . . .

Q Yes, the extrapolation issue.

A . . . 32 if one [MP] failed. So that was a sea change some time in the middle of April.

Q What has that to do with increasing the offer from 25 (sic) to 32 million?

A Once we got a more limited warranty negotiated, GGOWL wanted an increased amount of money upfront, and that is when the particularisation that you have shown me in the April 9th communication broke down and it became just dollars being traded.”

I am quite unable to follow the logic of the first part of this reasoning. It seems clear to me that GGOWL was disappointed at its lack of success in negotiating a wider form of warranty and, as a result, held out for a higher payment from Fluor (which is what Mr Bruno had indicated in the answer that he gave to me that is quoted above). It is quite artificial to suggest that all the remaining claims between the parties were simply settled for no consideration over and above a mutual release: in my view, the settlement ended, as it had started, by being a global settlement which did not particularise any of the individual heads of claim which formed part of it. As I have already mentioned, GGOWL’s draft settlement agreement was prepared after these “occasions, circumstances and developments” that occurred in mid April.

There was no oral evidence from GGOWL as to why the payment by Fluor under clause 6 was originally described in the draft prepared by GGOWL as “Lost Revenue Compensation Sum”, but I assume that this was a description which someone within GGOWL thought might assist GGOWL in its internal accounting. Mr Bruno’s evidence (which was supported by GGOWL’s evidence in the arbitration) was that the claim for monitoring costs represented not only the costs of carrying out testing, but also the loss of revenue as a result of the interruption of power generation during the tests. This may well be correct (it is consistent with evidence given on behalf of GGOWL in the arbitration) and it may be that it suited GGOWL to describe the compensation in terms of lost revenue so that it could be attributed to that head in its accounts - but that is pure speculation.

However, it then occurred to Fluor to alter this description to “Counter Claim Damages”, apparently without any resistance from GGOWL. I infer that the reason for this was that it was thought that this description might make it easier to recover all or part of that sum from ZPMC. Indeed, that description is the central plank underlying Fluor’s submissions at this hearing. The suggestion that these changes of wording completely altered the nature of the settlement is to my mind quite inconsistent with what actually happened. There would have to be a very compelling case to persuade me that this is what occurred.

But, as ZPMC has trenchantly pointed out, Fluor has not been consistent in its position. Fluor brought an action in the TCC against the designers of the Met Mast, in which it made the following assertions in its Particulars of Claim:

“61.2

Appendix 1 hereto is an estimate prepared by Fluor of the cost of a replacement met mast and installing it on the Met Mast’s existing foundation. The total cost is estimated to be £1,600,000 (the “Met Mast Replacement Costs”).

61.3

Fluor will not itself procure and install a replacement met mast, since on 14 May 2013, Fluor and GGOWL executed a settlement agreement (the “Settlement Agreement”), under which Fluor was relieved of any further responsibility under its contract with GGOWL including its responsibility for procuring and installing a replacement met mast and agreement was reached as to the final account between Fluor and GGOWL in respect of the project. Clause 6 of the Settlement Agreement provided that Fluor should pay to GGOWL the sum of £32 million as Counterclaim Damages, such sum including the Met Mast Replacement Costs. Fluor paid the said sum of £32 million to GGOWL on 14 May 2013, by the procedure stipulated in clause is 7 of the Settlement Agreement.

61.4

Accordingly, Fluor has lost and seeks to recover from:

(a)

[the first designer] as damages . . . that portion of the Met Mast Amount which relates to the transportation and installation costs of a new met mast. As shown in the breakdown at Appendix 1, these amount to £1,442,500; and/or

(b)

[the second designer] as damages . . . the Met Mast Replacement Costs.”

(My emphasis)

This claim was supported by a statement of truth signed by Mr Levy, in-house counsel to Fluor. Mr Levy was in close contact with Mr Bruno during the negotiations leading up to the Settlement Agreement and was involved in its drafting. The averment in that litigation against the designers - that the Met Mast Replacement Costs formed part of the £32 million described as Counterclaim Damages - is directly contrary to the case now being put forward by Fluor as to the meaning of the expression “Counterclaim Damages”.

Fluor made a further, but unrelated, point in relation to GGOWL’s legal costs of defending the claim. It submitted the £10.85 million claimed in respect of these costs could not have formed part of the £32 million paid by Fluor because the parties agreed to deal with those costs in a different way. Fluor relied on clause 11.2 of the Settlement Agreement which provided that the Tribunal would be asked to make no further award in relation to the arbitration. In my view, there is nothing in this point. Not only is it plain that GGOWL had a clear entitlement to its costs of the successful defence of the claim, but also the sum of £10.85 million was already the result of a discount of one third. I can see no reason whatever why GGOWL would be prepared to forego its claim for these costs. The agreement to invite the Tribunal to make no further award was no more than a provision to achieve an orderly closure of the arbitration.

Mr White, for ZPMC, said that the description of the payment by Fluor under clause 6 as “Counter Claim Damages” was merely a label and was irrelevant to any inquiry into what the payment actually represented. As I have already indicated, I have little doubt that, when making these amendments, both parties were taking advantage of the agreement that they had reached earlier in relation to cooperating in the structuring of the final settlement agreement so as to satisfy their respective internal accounting, tax, and other requirements. Of course, I accept that if when drafting a settlement agreement the parties choose to give a particular obligation a label (as opposed to a definition), that may constitute, if taken by itself, some indication of what the obligation was intended to represent. However, what I cannot accept is that the court is required to turn a blind eye to the surrounding circumstances if those circumstances indicate that a particular payment obligation was in fact in respect of something different to or more extensive than what is indicated by the label.

As I have already said, the figures that had been put forward by GGOWL on 15 March 2013 as the value of its various claims at the close of the mediation were roughly as follows:

In response to this offer Fluor offered nothing in respect of the legal costs, the Siemens claim, the other project claims and the historic weld dispute costs. Its only offer was that the parties should “walk-away” from their project claims and that Fluor would pay £10 million in respect of the defects and outstanding snagging items (but excluding the transformers, the Tekmar seals and the Met Mast defects). At that point, therefore, the parties were about £40 million apart.

As I have already said, on 10 April 2013 Fluor offered to pay £26 million in settlement of all GGOWL’s legal fees and project claims but on the basis that GGOWL completed all the remaining work. However, in addition, it was forgoing its own project claims which I have assessed had a settlement value of about £8.0 million, making a total cost to Fluor of £34.0 million. Had the true value of the costs to complete at the time when this offer was made been a £4 million, then Fluor’s offer would have been very close to matching the sum for which GGOWL had been asking. However, I doubt very much if GGOWL’s estimate of the costs to complete would have been as low as Fluor’s, particularly as it would probably have had to engage other contractors to carry out some of the work.

It was after Fluor made this offer that the problem with the warranty arose because GGOWL then demanded an entitlement to replace all the piles if a defect was discovered or if there was a failure in any one of them (the “extrapolation issue”). Whether this was a misunderstanding by GGOWL of the nature of the warranty that Fluor had been prepared to offer hitherto, or whether it was a tactical gambit to extract more money from Fluor, I cannot say. It appears that it was presented as the former, but I am left with the suspicion that it may have been the latter - but for present purposes it does not matter which.

It seems to me to be an unavoidable inference that GGOWL, having backed down on the extrapolation issue, was only prepared to conclude a global settlement if Fluor would increase its offer. Mr Bruno’s evidence about the amounts going up “in terms of cash” and that it was “just dollars being traded” seems to me to be entirely plausible. A settlement depended on Fluor bridging its side of the gap between the parties or at least going a long way towards doing so. ZPMC submitted that this was not the case, but I see no reason to doubt what Mr Bruno said on this point even if it was said for the first time during his evidence. In relation to the costs to complete, I consider that it is probable also that GGOWL was looking for a figure closer to the £10 million that Fluor had previously offered rather than the £4 million assessed by Fluor.

Under the terms of the settlement agreement Fluor’s offer to pay £26 million was increased by £6.325 million to £32.325 million. That figure was a global settlement: there was no attempt to break the sum down by reference to particular claims.

As I have already indicated, if the costs to complete were estimated at £4 million, GGOWL’s claims amounted to about £34 million. A month earlier Fluor had offered to pay £10 million for GGOWL to take on responsibility for completing the work, but by early May 2013 it could reasonably be assumed that the true figure would be lower than this. I consider that, by then, GGOWL would probably be prepared to accept between £6 million and £7 million in return for taking on the responsibility to complete the work - in other words, a further £2-£3 million.

Of Fluor’s payment of the further £6.325 million, I consider that £2.5 million was paid in respect of the costs to complete and that the balance - £3.825 million - represented a trade-off for the more limited form of warranty. The need to pay that sum was attributable to ZPMC’s breaches of contract.

To summarise the position, I consider that the reasonable settlement value of GGOWL’s claims as at early May 2013 was as follows:

If, as I consider likely, the claims sacrificed by Fluor had a settlement value of about £8.0 million, including the TP maintenance claim, then the value given by Fluor was about £40.325 million in order to achieve a compromise - a sum more or less equal to the reasonable settlement value of GGOWL’s claims that were being compromised. There cannot be, therefore, a separate claim for the TP maintenance costs.

Of the sum paid or foregone by Fluor, only items (4) and (6), totalling £13.825 million, were claims arising out of the defects in the welds that did not fall within the terms of the waiver letter. The small discrepancy of less than 1% between the value of each side’s claims can in my view be ignored. In these circumstances, if the settlement was a reasonable one, I consider that Fluor would be entitled to recover from ZPMC £13.825 million as damages for breach of contract attributable to the settlement with GGOWL.

Although, as I have already said, the extended warranty given by Fluor must be taken into account, it is not possible to place a value on it. However, it clearly represented a benefit to GGOWL and a substantial potential contingent liability to Fluor - albeit a fairly remote one, which I think is reflected in the fact that GGOWL was only prepared to settle for the more limited form of warranty if Fluor increased its cash payment. On the other hand, GGOWL had put forward a very substantial claim in respect of future monitoring which was abandoned in return for the warranty.

Overall, the settlement with GGOWL was the result of hard fought and protracted negotiations in which each side had the benefit of informed legal advice. None, or at least very few, of the claims to which a value had been attributed by either party was so weak as not to be taken seriously. The analysis that I have carried out shows that the reasonable settlement value of GGOWL’s claims was very close to the sum paid or foregone by Fluor, so on that basis alone that aspect of the settlement was self-evidently reasonable.

Taking all the circumstances into account, in particular the inclusion of the warranty, I have no hesitation in concluding that the settlement with GGOWL was a reasonable one. Accordingly, Fluor is entitled to recover damages in the sum of £13.825 million on account of the settlement.

If it is not already obvious, I should add that I have had careful and very detailed submissions from the parties in relation to the various ancillary claims. I have read those submissions with care, but I do not consider that the answers to the questions that I have to decide are to be derived from a minute and detailed analysis of the merits of those claims, still less to embark on a series of paper mini trials or evaluations. This is because the question is not what would have been the outcome if the particular claim had proceeded to judgment, but how robust the claim would have been seen to be - in terms of both liability and quantum - by a reasonably well-informed opponent. In settlement negotiations such as these, where there are numerous issues and claims, parties tend to take a broad and commercial approach. I was unable to conclude that any of the claims were plainly hopeless, although I regard Fluor’s adverse weather claim as coming fairly close to it. In this case the task has been made a little easier since each party attributed values to the principal claims being made by the other and, on the whole, it seemed to me that those values were as good a starting point as any.

Fluor claims very substantial sums in respect of experts and legal fees of obtaining an Engineering Critical Assessment (“ECA”) of the structural integrity of the MPs. The period covered by the costs is from June 2009 to May 2013. As Mr Bruno put it (in paragraph 4.6 of his first witness statement):

“Over the course of those four years, Fluor engaged the team of external experts identified in section 5 below. Those experts were primarily engaged to support Fluor’s efforts to obtain satisfactory ECAs to justify installing the structures with unrepaired defects and, later, in defending the counterclaim.”

The work in relation to the ECA can be split into several phases. The first phase was the work up to the point when “Fluor decided to ignore the NCR and install the MPs and TPs in their as delivered/unrepaired condition” (paragraph 323 of ZPMC’s Opening Submissions). The second phase was from 1 October 2009 to the issue of the arbitration proceedings in March 2011. In fact, Fluor has run these first two phases together and so has not distinguished between costs incurred before and after 1 October 2009. For this period Fluor claims 100% of the costs of the relevant experts as being attributable to the ECA.

The third phase was from then until the Tribunal’s First Partial Award in November 2012, and the fourth phase was from December 2012 until the settlement in May 2013. For the third phase Fluor claims a proportion of the expert and legal costs to reflect the fact that a proportion of those costs were attributable to the pursuit of the arbitration. For the final phase Fluor claims the costs in full on the ground that they relate exclusively to work performed “for the purpose of the ECA and/or defending GGOWL’s counterclaim” (Mr Ross’s first report, paragraph 42.2).

The reason given by Fluor for continuing to pursue an ECA after the decision to install the MPs was taken on 1 October 2009 is summarised in paragraph 328 of its Opening submissions in these terms:

“Fluor continued with the ECA work long after taking the decision to ignore the NCRs because it had (1) to defend its decision to install against GGOWL’s allegations regarding structural integrity (which eventually evolved into its counterclaim) and (2) to protect its reputation in the market, particularly with regard to having decided to install a product with known defects.”

ZPMC submits that the claim for the ECA costs has been waived. It relies on numbered clause 3 of the Waiver Letter, because it submits that the ECA costs are “additional costs . . . suffered as a result of NCRs 006, 008 and 009 issued by GGOWL”. In essence, ZPMC makes two points:

However, this paragraph of the Code is not referred to in NCRs 008 and 009, which stated that the integrity of the MPs was in question until “the requirements of clauses B405 and B406 of the code have been complied with”. As set out in paragraph 277 of the Judgment, paragraphs B405 and B406 concern the need for additional NDT when frequent repairs are required and the need for full length examination of all welds made using the same procedure.

As I pointed out at paragraph 495(1) of the Liability Judgment, the subject of the “understanding” underlying the waiver letter was the “joint pursuit of the substantial additional costs that both Fluor and ZPMC have incurred as a result of the MP and TP weld testing and repair protocol required by [GGOWL]”. It seems to me that the particular provisions of the letter have to be read in the light of this statement of the parties’ common understanding.

In my view, the costs of obtaining an ECA were not costs incurred as a result of the testing and repair protocol required by GGOWL, but rather were costs incurred in order to demonstrate that the imposed testing and repair protocol was misconceived and unnecessary. It was a step that Fluor was compelled to take in order to mitigate the loss caused by ZPMC’s breaches of contract. I therefore reject the submission that the costs of the ECA have been waived.

ZPMC submits also that the structural integrity of the MPs and TPs was an issue in the arbitration, even though GGOWL refused to issue its counterclaim until the Tribunal had resolved the issues about the validity of the NCRs. Not only did Fluor make the same point consistently and with persistence before the Tribunal, but also GGOWL’s defence of the validity of the NCRs involved the assertion that the welding was defective and, by implication (if not expressly), the MPs and TPs were defective also.

In the Liability Judgment I said, at paragraph 533:

“In my view, this evidence clearly establishes that, as a result of the cracking that had been found in the piles at Vlissingen, the only reasonable option available to Fluor was to carry out an investigation into the true condition of the piles and to establish the extent (if any) to which this might affect their performance in service.”

And, at paragraph 573:

“As a result of the presence of this cracking the MPs and TPs on delivery at Vlissingen were not fit for the purpose of being installed forthwith without either further examination, testing and repair or a satisfactory Engineering Critical Assessment indicating that repairs were not necessary and that the MPs and TPs would, from a structural point of view, perform satisfactorily in service for 25 years.”

I am prepared to accept ZPMC’s submission that the costs of obtaining expert advice in relation to fracture mechanics and the integrity of the welds was of central relevance to the issues in both the claim and the counterclaim in the arbitration. But that is not to conclude that the only purpose of obtaining that advice was that of pursuing litigation. I accept Fluor’s submission also, namely that (a) it had to defend its decision to install the MPs and TPs in the face of a challenge to their structural integrity and (b) to protect its reputation, given that it was installing MPs and TPs which were known to have cracks in the welds.

In my view each of these grounds was a direct consequence of ZPMC’s breaches of contract.

For all these reasons, I am quite satisfied that Fluor is entitled to recover by way of damages the costs of obtaining expert advice about the structural integrity of the piles as part of an ECA up to the end of September 2009. I consider also that it is entitled to recover any legal costs directly incurred in obtaining that expert advice.

It might be argued that once Fluor had been persuaded by expert advice that the MPs and TPs could be installed in the sea bed without further examination or repair, there could no longer be any need for an ECA. However, I think that such an argument overlooks practical realities.

Fluor could reasonably install unrepaired piles in the sea bed once it was satisfied in its own mind that it was safe to do so. However, the decision might be controversial and Fluor would have appreciated, as in fact it did, that it would have to satisfy not only itself but also GGOWL and the relevant regulatory bodies that the piles were sound and would perform satisfactorily in service for 25 years. In addition, it had a legitimate interest in protecting its own reputation. If support for this first ground were needed, it is provided by Mr Bruno’s evidence that he instructed Hogan Lovells in 2009 because the contract was subject to English law and he was concerned that GGOWL might apply for an injunction in the English courts to prevent the installation of the unrepaired piles (Day 8/116-117).

As to GGOWL’s position, in its Statement of Defence served in the arbitration (in July 2011) it asserted that no adequate ECAs had been provided by Fluor and that the advice that been disclosed at the time “could not be said to constitute an ECA or to extinguish the concern about the effect of flaws which Fluor had confirmed to exist” (at paragraphs 2.7 and 2.8). In these circumstances I consider that it was reasonable for Fluor to keep its experts working on the ECA until it was satisfied that it had a document that would satisfy the scrutiny of the most exacting regulatory authority.

Fluor’s case at this trial has been that it is entitled to recover the entirety of the costs of obtaining an ECA, including the associated legal fees. (Footnote: ⁴⁰ ) On the basis of an analysis of the invoices submitted by its experts, in many cases Fluor has claimed 100% of the costs as being referable to the ECA. By necessary implication, therefore, in relation to these invoices it has asserted that none of the work carried out by those experts was done for the purpose of pursuing the arbitration or for obtaining legal advice.

However, this position directly conflicts with that taken by Fluor in the arbitration itself when it claimed privilege for all the documents relating to the ECA on the ground that the dominant purpose of obtaining them was for use in litigation, whether contemplated or actual, initially against ZPMC and latterly against GGOWL. Fluor says that it had already provided GGOWL with copies of some final expert reports well before it issued its Request for Arbitration (see paragraph 515.13 (b) of its Closing Note). However, this seems to me to be of limited relevance: any subsequent waiver of privilege cannot change the purpose for which the document was initially produced.

By way of example of the position taken by Fluor, in a letter dated 13 December 2011 Fluor’s US attorneys, McElroy, Deutsch, Mulvaney & Carpenter (“MDMC”), wrote:

“There can be no doubt that by this time ZPMC’s welding, as Respondent well knows, was directly at issue between the parties. Respondent had already rejected Shipment #1 on 3 June 2009 due to welding issues, causing Claimant to seek damages that, at that time, it contemplated recovering against ZPMC through the legal process. To that end, Claimant then directed first TWI, and later Jan Ibsoe, to perform certain tasks and prepare reports to be utilised for the dominant purpose of substantiating the legal claims that Claimant was then contemplating asserting against ZPMC, which claims were later re-directed to Respondent as a result of Fluor’s determination that the Employer Required NDT and Remediation and the NCRs issued by Respondent were without justification . . .”

In a note prepared in the course of the arbitration in January 2012 Fluor made the following submission:

“In addition, it must be noted that this issue and the other set forth by Claimant (i.e. the relevance and admissibility of all weld NDT data) are still in dispute solely because Respondent wrongly refuses to acknowledge that the issue of the structural integrity of the foundation works is and always has been at issue in this case as it is presently pleaded. ”

GGOWL’s position was summarised in submissions made on 27 January 2012 in the following terms:

“In general terms the purpose of an ECA is to seek to establish whether indications/defects jeopardise the structural integrity or design life of a structure. This was in fact the purpose or at the very least one purpose of the ECAs undertaken by Fluor as can be seen from the contemporaneous correspondence and Fluor’s Amended Statement of Case . . .”

Fluor’s submission in response made these assertions, at paragraphs 9, 10 and 14:

“9.

Fluor’s dominant purpose in retaining TWI in June 2009 and later, Jan Ibsoe, was to investigate the welds and testing procedures at issue for the purpose of substantiating Fluor’s then contemplated legal claims against ZPMC based on what was then thought to be ZPMC’s defective welding. An additional and also critical purpose was to determine whether GGOWL’s allegations of improper welding and NDT on the part of ZPMC were valid and supportable. Only later did that purpose evolve into substantiating Fluor’s assertions that GGOWL’s criticism of the welds, rejection of the shipments, and requirement of enhanced testing were all improper and in breach of the parties’ contract . . .

10.

All of the above reasons as to why Fluor utilised Dr Jan Ibsoe and TWI after June 2009 make clear that Fluor’s dominant purpose in doing so was to directly aid Fluor in its contemplated legal action to recover the extraordinary costs it was incurring in re-testing and re-welding the MPs and TPs . . . .

. . .

14.

GGOWL’s allegation that many of the documents at issue related to the preparation of Engineering Criticality Assessments (“ECA”) misses the point. While GGOWL may wish to characterise the ECAs as having been prepared so as to “persuade GGOWL that defects/indications [in the welds] were tolerable“, the dominant purpose of all such documents was, in fact, to establish in the contemplated litigation that Fluor had fully complied with the Contract’s requirement that each of the MPs and TPs meet the required design life requirement of at least 25 years, thereby further discrediting GGOWL’s argument that enhanced testing and remediation was necessary. While GGOWL may not believe that this issue is relevant to the advancement of Fluor’s claim, the proper consideration under a litigation privilege analysis . . . is the purpose that the person who directed the creation of the document had in mind - not what his adversary subsequently claims it to be.”

Fluor’s claim for privilege was supported by an affidavit sworn by Mr Hans Dekker dated 15 February 2012. It was clear from this affidavit that Fluor had completely satisfied itself by 1 October 2009 that the foundation works required no further testing. At paragraph 32 of that affidavit, under the heading “Post 1 October 2009 Events” Mr Dekker said this:

“Because the decision we made to proceed and GGOWL’s response left no doubt that the Weld Dispute (as it had become to be known) would now be resolved in litigation, we continued with the retention of TWI and Jan Ibsoe for that purpose and put them under the direction of our in-house and outside counsel.”

(My emphasis)

So in that affidavit Mr Dekker was asserting that the expert work carried out by The Welding Institute (“TWI”) and other experts after 1 October 2009 had been obtained for the dominant purpose of pursuing the claim in the arbitration against GGOWL. That submission had already been made in the letter dated 13 December 2011 in which Fluor’s US counsel, Mr Pepe, wrote to the Tribunal in the following terms:

“To that end, the claimant then directed first TWI, and later Jan Ibsoe to perform certain tasks and prepare reports to be utilised for the dominant purpose of substantiating the legal claims that the claimant was then contemplating against ZPMC . . . Regardless of the fact that the Claimant’s litigation against ZPMC did not ultimately materialise and Fluor instead pursued arbitration against Gabbard, the fact remains that the draft reports now being sought by the respondent were prepared in contemplation of litigation and are accordingly subject to the protections of the litigation privilege . . . The Welding Institute’s role changed such that by mid-June 2009, as detailed above, it was actively engaged in providing the claimant with technical substantiation for its then contemplated claims against ZPMC. Additionally, the Respondent’s characterisation of TWI and Ibsoe as “technical advisers” rather than “experts” is both incorrect and immaterial. The critical point is at the time they prepared their reports, regardless of the label the respondent ascribed to them, such reports were prepared for the dominant purpose of substantiating claimant’s legal claims which at first contemplated by ZPMC but were later re-directed to Gabbard once Fluor determined that the NCRs were not justified.”

Faced with these submissions and the affidavit of Mr Dekker, on 29 February 2012 the Tribunal ruled as follows:

“The Tribunal is of the view that the Claimant’s assertion of privilege over the documents created on and after 1 October 2009 is made out. . . . The dominant purpose of the subsequent creation of documents was litigation. Although the Claimant may possibly have provided documents to the Respondent in an effort to seek the Respondent’s acceptance of its position, this was an ancillary aim and not the dominant purpose of the creation of the documents.”

(My emphasis)

In cross examination it was put to Mr Bruno that Fluor had maintained in the arbitration that evidence about structural integrity was an essential part of its case that the work it had been instructed to do was extra. In response, Mr Bruno said this (Day 8/148):

“A Structural integrity was not necessary to prove Fluor’s case against GGOWL in the arbitration. What happened is there was a notice of counterclaim in July, I believe it is 2011, and later a notice of a counterclaim not specified, and then finally a counterclaim. And I do recall a great concern about getting ambushed in the arbitration because all of the allegations of the counterclaim had been made right before the arbitration on the claim.

There were two problems, I guess. One is ambush and the other is splitting a case in the arbitration and having the counterclaim go later.

Q But it was part of its case, Fluor’s case, in support of the claim, that the work of testing and repair was not technically necessary because the MPs and TPs were structurally sound without being repaired. Do you remember that?

A No.”

The cross-examination continued in this vein (Day 8/150-155). I was unimpressed by Mr Bruno’s evidence on this issue, which I thought was partly evasive and partly a little disingenuous.

Thus Fluor asserted, through Mr Dekker’s affidavit and by way of submissions, that the purpose, or at least the dominant purpose, of obtaining further expert advice from 1 October 2009 onwards, in particular, that of TWI and Dr Jan Ibsoe, was to support its claim in the arbitration against GGOWL. It persuaded the Tribunal to accept that that was the case with effect from 1 October 2009.

Although Mr Dekker had given evidence at the liability hearing, Fluor did not call him during the quantum hearing to explain why the contents of the affidavit that he swore in support of Fluor’s claim to privilege, and the submissions made on the basis of it, was incorrect. In my view it would not accord with due legal process for Fluor to be permitted now to assert the contrary without, at the very least, providing a very good explanation supported by the evidence of Mr Dekker. In response to this Fluor submits that, since the test for privilege is an objective one, the court is entitled to go behind what was said by Mr Dekker and Fluor’s lawyers and to decide for itself the true purpose for which the documents came into existence.

This, in my view, is a thoroughly unsatisfactory response – at several levels. To begin with, the case cited as authority for this proposition, Guiness Peat v Fitzroy Robinson [1987] 1 WLR 1027, involved a single letter, in which the defendants gave notice of a claim to their insurers, that had been disclosed by mistake. Although the purpose of the writer of the letter was to comply with the terms of the policy, the court held that the dominant purpose was the insurers’ need to have information upon the basis of which they could take legal advice.

There are at least three reasons why this case does not assist Fluor. First, in Guinness Peat there were two parties who had an interest in the letter being written, the insured and their insurers, and both had (or may have had) different reasons for wanting it written. By contrast, in the present case there was only one party, Fluor, whose purpose is relevant.

Second, there was only one document and determining the purpose or purposes for which it was written was a relatively straightforward exercise and one that the court could undertake without difficulty. In this case there are many documents and they are “scattered across the Arbitration trial bundle, the trial bundle in the ZPMC liability proceedings, and the more recent disclosure in these quantum proceedings” (Mr Ross’s first report, paragraph 42.7). Not only that, but they are documents the contents of which are highly technical. It is likely to be far from easy for the court to determine, from their contents alone, the true purpose for which Fluor obtained them.

Third, in Guinness Peat the claim for privilege was maintained throughout (once the mistake was discovered). It was not a case where the originator of the document originally claimed privilege asserting that the document in question came into existence for the purpose of contemplated or actual litigation and subsequently, when the wind changed, asserted the opposite.

In the circumstances of this case it would be quite impracticable, if not impossible, for the court to embark on the exercise of examining all the documents produced by Fluor’s technical experts and to attempt to determine what was Fluor’s dominant purpose in obtaining each one. In my view, there is no realistic option but to start with the evidence given by Fluor, and submissions made on its behalf, to the Arbitral Tribunal as to the dominant purpose for which these documents came into existence.

But, leaving these considerations apart, for the reasons that I have now given, I consider that the overriding factor is that since Fluor contended in the arbitration that the dominant purpose of instructing the experts was the pursuit of the claim against GGOWL, that is not a position from which it should be allowed to resile.

At paragraph 498 of its Closing Note Fluor recorded, correctly, that the Court has asked for assistance with the following question:

"If the Court concludes that the documents that came into existence after 1 October 2009 were created for the dominant purpose of pursuing litigation against GGOWL, does this mean that no costs are recoverable at all in respect of the relevant reports and legal assistance or, alternatively, is Fluor entitled to recover only a proportion. If the latter, what is the appropriate approach?"

However, in the following paragraphs Fluor did not actually provide an answer: rather it submitted that:

“there undoubtedly is sufficient material for the Court to be able to reach a view as to what costs ought to be wholly recoverable/unrecoverable and, as Fluor has sought to do, to make an appropriate apportionment where there are costs that were incurred for multiple purposes and so may only be recoverable in part”.

This answer misses the point. If the dominant purpose for which a document came into existence is A, then any other purpose, B, is by definition subsidiary. If the costs of purpose A are not recoverable, but the costs of purpose B are recoverable, the costs that can be recovered must be less than 50% of the whole. Thus if purpose A is truly the dominant purpose, I cannot see how the costs of purpose B can ever exceed about 40% of the total - but it could, of course, be some lower proportion.

I am prepared to accept that, where the descriptions of the work covered by the invoices support it, a subsidiary purpose of obtaining the expert evidence after 1 October 2009 was to buttress or improve on the expert advice that had been given up to 30 September 2009 in relation to the ECA for the purpose of satisfying GGOWL and the relevant regulatory authorities that the structural integrity of the MPs and the TPs was not materially affected by the cracking in the welds.

However, the converse of this conclusion is that it must follow that the bulk of the costs of instructing the experts, and the associated legal fees, cannot be attributed to the ECA. As I have indicated, in the absence of any cogent submissions to the contrary, I propose to adopt the approach that where the descriptions of the work done after 1 October 2009 indicate that a significant proportion was in connection with the ECA, Fluor should be entitled to recover up to 40% of the relevant expert’s fees as being attributable to the ECA, although the proportion might be lower if the relevant invoices do not suggest that such proportion is justified: to allow more would be inconsistent with the claimed dominant purpose of providing expert evidence in support of the arbitration - a position that MDMC reiterated in October 2011 (see its letter to Nabarro dated 18 October 2011).

In addition, I consider that Fluor is entitled to recover a proportion of the associated legal fees. Since the lawyers instructed by Fluor on each side of the Atlantic have not adopted precisely the same approach when invoicing for their fees, I consider that each firm’s fees must be considered separately.

Fluor submits that the costs of defending the counterclaim, that is the costs incurred after November 2012, are entirely attributable to the welding defects and were not caused by the NCRs. I do not disagree. However, I agree with ZPMC that they are costs of pursuing the Weld Dispute Claims: it was all too likely that the pursuit of the claims for the costs of the additional NDT and repair work would not only be resisted by GGOWL but also met by a counterclaim in respect of the risk presented by the unrepaired MPs and TPs.

Further, by numbered clause 5 of the waiver letter Fluor agreed to pay ZPMC 50% of any award after deduction of any damages recovered by Fluor and “all costs incurred by Fluor in the Arbitration”. The final sentence of the clause provided that if those costs exceeded the amount of the award, the difference would be borne by Fluor. The “Arbitration” was defined as the proceedings in which Fluor was to pursue the Weld Dispute Claims against GGOWL by way of arbitration in London. In my view, that definition clearly includes any counterclaim that GGOWL might bring in response to the claim. Since Fluor recovered nothing by way of any award, the consequence is that it must bear the costs of the arbitration. Accordingly, I reject Fluor’s claim to recover the costs incurred in defending GGOWL’s counterclaim because, by the terms of the waiver letter, it agreed to bear them.

The overall sums claimed by Fluor are divided into three periods as set out in the following table:

As I have already noted, Fluor has not distinguished between the costs incurred up to 1 October 2009, when the decision was taken to install the MPs without carrying out any further repairs, and the costs incurred thereafter. In my view, it is necessary to consider these separately.

Unfortunately, the only way to do this is by reference to the analysis of the invoices from the various consultants set out in the appendices to Mr Ross’s Reply Report. I shall have to take each of the consultants separately. I should say that I have also read carefully the annotated version of Fluor’s Appendix 7 with which I was provided during the closing submissions.

The Welding Institute.

Jan Ibsoe/ABS. Jan Ibsoe was initially retained on his own account but subsequently joined ABS. Accordingly, these two groups of invoices can be treated together. On 17 February 2010 MDMC retained Dr Ibsoe as “a consultant and potential expert witness” in Fluor’s “anticipated arbitration” against GGOWL.

Lateral Wave Ltd.

Acuren Group – Mr Ted Hamre.

Materials Research Institute - Mr Ginzel.

MRW Advisory.

Other experts.

The fees of MDMC.

The fees of Hogan Lovells.

I find that the sums that Fluor is entitled to recover in respect of the ECA costs are as follows:

Fluor’s claim in respect of liquidated damages is now for 47 days of delay caused by ZPMC’s breaches of contract in relation to Shipment No 2, the costs of which have not been waived by Fluor, for which it had to pay liquidated damages to GGOWL in the sum of £8,074,647.

The contract between Fluor and GGOWL contained four milestones, two for each phase: Infrastructure Commissioning and WTG Foundation Works Commissioning. For failure to achieve these milestones, clause 8.8.1 of the contract provided for the payment of liquidated damages for delay as follows:

“Subject to Clauses 8.7 (Rate of Progress) and 17.4 (Limitation of Liability), if the Contractor fails to comply with clause 8.2 (Time for Infrastructure Commissioning) and/or Clause 8.3 (Time for WTG Foundation Works Commissioning), the Contractor shall be liable for and shall pay every 10 Business Days or allow on demand liquidated damages to the Employer:

(i)

in respect of the Phase 1 Infrastructure Works, at the rate of 900 Pounds Sterling per MW per day in respect of the aggregate of the Rated Power Output of each WTG, in relation to which a WTG Foundation Works Commissioning Certificate should have been issued according to the Construction Programme, to the extent that power from each such WTG cannot be exported due to a failure by the Contractor to commission the Phase 1 Infrastructure Works by the Phase 1 Target Infrastructure Commissioning Date, on each day from the Phase 1 Target Infrastructure Commissioning Date up to and including the Phase 1 Infrastructure Commissioning Date;

(ii)

at the rate of . . . in relation to which the WTG Foundation Works Commissioning Certificates have not been issued, for each day from the Phase 1 Target WTG Foundation Works Commissioning Date up to and including the Phase 1 WTG Foundation Works Commissioning Date;

(iii)

in respect of the Phase 2 Infrastructure Works, at the rate of . . . in relation to which a WTG Foundation Works Commissioning Certificate should have been issued according to the Construction Programme, to the extent that power from each such WTG cannot be exported due to a failure by the Contractor to commission the Phase 2 Infrastructure Works by the Phase 2 Target Infrastructure Commissioning Date, on each day from the Phase 2 Target Infrastructure Commissioning Date up to and including the Phase 2 Infrastructure Commissioning Date; and

(iv)

at the rate of . . . in relation to which WTG Foundation Works Commissioning Certificates have not been issued, for each day from the Phase 2 Target WTG Foundation Works Commissioning Date up to and including the Phase 2 WTG Foundation Works Commissioning Date,

provided that no liquidated damages shall be payable by the Contractor to the Employer under Clause 8.8.1 (ii) and 8.8.1 (iv) in respect of any WTG following the issue of a WTG Foundation Works Commissioning Certificate in relation to that utility.”

The rate per MW per day was £1,125 (it was increased from the £900 stated in the contract), and the Rated Output was 3.636: so the daily rate per WTG was £4,090.50.

The contract also had three different caps for liquidated damages: the Phase 1 cap (£44,955,683), the Phase 2 cap (£92,064,926) and an overall cap (£91,347,077).

The dates on which the various milestones were achieved for Phase 1, the delay for each milestone, and the date on which the cap on liquidated damages for the relevant phase would have been achieved, were set out in ZPMC’s Closing Submissions (after subsequent correction) as follows:

Mr McKibbin attributed 232 days of the delay to the Phase 1 Foundations Works to breaches of contract by ZPMC, of which he said 97 days were attributable to breaches of contract the consequences of which did not fall within the settlement. In its Closing Note Fluor submitted (on the basis of the counterfactual scenario) that 47 days of critical delay were caused by ZPMC’s breaches of contract in relation to Shipment No 2 the consequences of which had not been settled.

So if GGOWL had chosen to deduct liquidated damages at the contractual rate from, say, the Phase 1 Foundation Works Target Date onwards, the contractual cap for Phase 1 would have been reached after 262 days delay (17 January 2011) if all the WTGs in that phase were affected by the delay. So, ZPMC contends, however one looks at it, in relation to the Phase 1 Foundations Works, that the contractual cap for liquidated damages would have been reached irrespective of any delay caused by ZPMC’s breaches of contract.

Fluor’s position, as finally set out in a Responsive Note dated 18 September 2017 (omitting footnotes), was clear and is worth setting out in full:

“29.2.1

GGOWL was entitled to levy LDs under Clause 8.8 of the BoP Contract [2.2/442]. Insofar as it wished to do so, it was required to give notice as soon as reasonably practicable and not later than 21 days after becoming aware of the event or circumstance giving rise to the claim (see Clause 2.4.2 at [2.2/406]).

29.2.2.

GGOWL did elect to levy LDs, and provided notice of its claim by correspondence within the trial bundle. The letter relevant to this claim was dated 23 July 2010 and is contained in the trial bundle at [18.13A/5087]. It notified Fluor that GGOWL was levying LDs in respect of the Phase 1 Foundation Milestone (i.e. that which Fluor alleges was delayed by ZPMC's breaches) from the date that the milestone had been missed (1 May 2010) onwards to the date of the letter (a total of 84 days). The LDs levied in respect of this milestone were notified in the sum of £14,259,483.

29.2.3.

ZPMC is correct that LDs accrued against the Phase 1 Infrastructure Milestone prior to this (¶33 of its most recent note) but neglects to mention that GGOWL did not seek to levy them from the point that the Phase 1 [Infrastructure] Milestone was missed and instead levied LDs in respect of the Phase 1 Foundations Milestone. At the time that it did so, only £12,696,912 of LDs had been levied (i.e. only c25% of the cap was exhausted).

29.2.4.

So the actual position is that GGOWL did in fact levy LDs against Fluor in respect of the very milestone to which ZPMC caused delay.”

If one stops the clock at the date of GGOWL’s letter of 23 July 2010, by which time there had been 84 days of delay, that delay had been caused largely (if not wholly) by ZPMC’s breaches of contract. However, as I have already found, to the extent that such delay was caused by the issue of the NCRs (leaving aside whether or not it may also have been caused or contributed to by other matters for which ZPMC was not responsible), the costs of it - for which ZPMC might otherwise have been liable - have been waived by the agreement reached in June 2010.

But on the counterfactual scenario I have concluded that the breaches of contract in relation to Shipment No 2, the consequences of which were not waived, deprived Fluor of the opportunity of avoiding 15 days delay (for the purposes of liquidated damages, this has to be rounded up from the 14.5 days that I found) to the installation of the MPs.

Accordingly if, as at 23 July 2010, this 15 days of delay had been critical to the achievement of the Phase 1 Foundations Milestone, ZPMC would have been liable for it.

If, thereafter, as a result of matters for which ZPMC is not liable, further delays occurred to the achievement of the Phase 1 Foundation Milestone such that the cap for Phase 1 was reached, or even exceeded, that in my view would have been irrelevant. In that situation, Fluor would have incurred a liability for 15 days liquidated damages, for which ZPMC would have been responsible: subsequent events could not eliminate it.

I therefore reject ZPMC’s arguments based on the assertion (which may be correct as a matter of fact) that the cap on liquidated damages for Phase 1 would have been exceeded in any event. It follows, of course, that the position is even stronger in relation to Phase 2.

However, the insuperable difficulty in the way of Fluor’s claim for liquidated damages is the fact that I have already held that the installation of the Phase 1 MPs was in float by December 2009, because by then the critical path lay through the fit out and/or installation of the TPs (see paragraph 398 above). Accordingly, even if the installation of the Phase 1 MPs had been achieved 15 days earlier – that is, just before mid December 2009 – Fluor’s achievement of the Phase 1 Foundation Milestone would have been unchanged. For these reasons, Fluor’s claim for liquidated damages based on this 15 days delay must fail.

US $. I find that the sums recoverable are as follows:

GBP. I find that the sums recoverable are as follows:

Euro. I find that the sums recoverable are as follows:

Other currencies:

Fluor claims overhead and profit, but in neither its Re-Amended Particulars of Claim nor its Closing Note is a figure put forward or is there any identification of the heads of claim or items to which this should be added. The passages in Mr Ross’s reports referred to in Fluor’s summary of its claim (at Appendix 1) do not take matters much further.

I shall therefore direct that Fluor is to identify the sums in paragraphs 615 to 618 above to which it says overhead and profit should be added, together with the basis for doing so, and the amounts claimed for each item, to which ZPMC can then respond.

Having received these submissions, I will deal with any entitlement to overhead and profit; if necessary, in in a short supplementary judgment. I invite the parties to agree a procedure for doing this, failing which I will give appropriate directions.

Similar considerations apply to VAT, in respect of which I shall give a similar direction.

ZPMC seeks to recover the balance of the Warranty Bond, €23,409,750, less “the limited claims that fall outside the settlement” (Opening Submissions, paragraph 560). The bank paid this sum, which was the full amount of the bond, following a call made by Fluor on 17 March 2014.

I understand it to be common ground that Fluor must give credit for this amount (together with interest, if appropriate) against any recovery in respect of its claim.

Fluor invites me to make an order now for an interim payment. Taking the figures that I have set out above (ignoring the sums in CA$ and ¥), and adopting the prevailing exchange rate for the sum in €, produces the following figures:

The call on the extended Warranty Bond, made on 17 March 2014, was for €23,409,750. The parties are agreed that there should be an interim payment, but there is an issue about conversion of the sums in Sterling and euros into US$.

At paragraph 581 of the Liability Judgment I directed that there was to be judgment for Fluor with damages to be assessed. Accordingly, there is jurisdiction to make an order for an interim payment under CPR 25.7(1)(b).

By way of background, I should mention that I have not yet heard submissions on costs and so I do not consider that I should speculate on what the ultimate order in relation to the costs of this litigation might be, but it is certainly not obvious that the outcome might be an award in ZPMC’s favour which could justify a refusal to direct an interim payment or a reduction in its amount. Further, I note that CPR 25.7(1)(b) specifically excludes any assessment of a sum of money by way of costs from giving rise to an interim payment, which suggests that a likely order for costs in favour of one party or the other is not a ground for making (or refusing) an order for an interim payment.

ZPMC submits that CPR PD40B permits the order to be for the payment of a sum in a foreign currency or the Sterling equivalent at the time of payment. It opposes conversion of sums in other currencies into US$ on the ground that this option is not permissible. I accept this point.

On 21 March 2014 the Sterling equivalent of €23,409,750 was £19,581,556, so, for the purpose of calculating the interim payment, I will take that rate. It can be adjusted later for the final reckoning if that is necessary.

Accordingly, ZPMC is to pay Fluor £1,829,634 (being £21,411,190 - £19,581,556) and US$5,893,591 by 4 pm on 1 February 2018 (an agreed date).

In the light of my conclusions I will hear the parties on all questions of interest, costs and any matters relating to the currency or currencies in which judgment should be given.

I am very conscious that I have not dealt with every point that has been raised by the parties. I have received over a thousand pages of written submissions (excluding all the appendices) all of which I have read very carefully and I hope that I have now decided every point that matters (apart from those mentioned in the immediately preceding paragraphs).