England and Wales High Court (Technology and Construction Court) Decisions

IN THE HIGH COURT OF JUSTICE
QUEEN'S BENCH DIVISION
TECHNOLOGY AND CONSTRUCTION COURT

		Royal Courts of Justice Strand, London, WC2A 2LL
		25 October 2013

The Hon Mr Justice Coulson:

1. INTRODUCTION

1. On 1 August 2007, water was seen cascading out of the balconies and down the sides of a block of flats known as Holly Court on the Greenwich peninsula, in South London. The water came from two of the cold water riser mains within the Cores of the building which were leaking as a result of two separate but related failures. The damage that was done was extensive, and the eventual cost of the remedial works and other losses has been agreed at £4.75 million.
2. At the time of its construction and the subsequent flooding, the owners of the building were the claimant, Greenwich Millennium Village Limited ("GMVL"). They paid for the remedial works and resolved the claims made against them by the leaseholders of the individual flats. They also made a successful claim on their property insurers. Accordingly, this is now a subrogated claim in which GMVL seek to recover the costs caused by the flooding against some of those who were responsible for the original design and construction of the building.
3. Although the original design and build contractor was Laing O'Rourke ("LOR"), and GMVL's original agents were WT Partnership ("WTP"), neither is the subject of the subrogated claim^[1]. Instead, GMVL pursue a claim against the mechanical and electrical sub-contractors, Essex Services Group PLC ("Essex") and the designers of the mechanical systems, Hoare Lea ("HL"). Those claims are pursued in reliance upon individual warranties provided to GMVL by Essex and HL respectively, although there are also alternative claims in negligence against both defendants. Essex pass on the claim against them pursuant to the mechanical sub-sub-contract that they entered into with HS Environmental Services Limited ("HSE") and HSE, in their turn, pass on the claim to the labour-only mechanical sub-sub-sub-contractor, D G Robson Mechanical Services Limited ("Robson"). Robson are the only one of the defendants and third parties who did not, on any view, have any sort of design obligation.
4. Accordingly, in many ways this is old-fashioned TCC litigation, involving five parties, with different defences and claims up and down the line, depending precisely on where the parties sit in the contractual chain. It is also old-fashioned litigation in another sense: many of the factual issues are surprisingly basic, such as 'who over-tightened the nut?' and 'who closed the isolation valve when it should have stayed open?' However, despite this retro feel, it should be said at the outset that the trial was crisply conducted by all counsel and the process was greatly assisted by clear and cogent expert evidence provided with the rigour required by the newly modified CPR Part 35.
5. There was a third (and less welcome) way in which the underlying facts of this case seemed redolent of another age. Skilful cross-examination revealed a situation, both on site and off, which was at times comically inept: important obligations were inadvertently muddled as they passed on down the contractual chain; warranties were provided in relation to contracts that were subsequently changed or services which were never provided; and everyone was so intent on avoiding responsibility for anything meaningful that no single organisation had overall control of the design, installation, supervision, inspection, testing and acceptance of the boosted mains cold water system ("BMCWS"). To use a phrase common in the UK construction industry 20 years ago, it was death by sub-contracting.
6. This muddle resulted in pipework systems which, on completion, varied markedly from core to core, even though they should have been the same in each. Different valves could be found in different places on the risers where there was no logical distinction between the locations. Sometimes there was a valve in a particular place on the riser in one core, but no valve in the same place on the riser in the adjacent core. There was no proper paper trail dealing with the critical late addition of surge arrestors to the BMCWS which lies at the heart of this case, with the result that Robson, the labour-only plumber, was able to argue that it did not do the particular work now under scrutiny because there was no contemporaneous paperwork to suggest that it did. Most importantly of all, there was no specification or design drawing showing the detail of the surge arrestor installation, no record drawing showing what was actually installed; and no inspection report saying how and why that which had been installed was satisfactory. To crown it all, the Operating and Maintenance Manuals provided to GMVL did not refer to the critical surge arrestors (and their related components) at all. In those circumstances, it is perhaps less surprising than it might otherwise appear that the Millennium Village development was the subject of the catastrophic flooding on 1 August 2007.

2. THE ISSUES

7. There were two separate flooding incidents on 1 August 2007. The first occurred in Core 3, Level 2, when an isolation valve ("IV") on a riser branch failed. The precise element which failed was the nut that connected the downstream pipework to the downstream part of the IV. It is agreed by GMVL and Essex (although this view was not universal), that the cause of the first flood in Core 3 was due to this plastic nut being over-tightened, and the presence of debris in the thread. The over-tightening was evidenced by tool marks, probably made by metal wrenches called Stilsons, which should not have been used on the plastic nut.
8. GMVL allege that Essex over-tightened the nut prior to handover in August 2006, and was therefore responsible for the damage which led to the first flood. Essex deny liability, suggesting (principally through Robson) that the nut was over-tightened as part of GMVL's maintenance work in the period of almost a year after the handover and before the flooding.
9. The claim in respect of Core 3 is passed on by Essex to HSE who, in turn, pass it onto Robson. HL are not involved in the Core 3 claim at all, either by reference to their design or their periodic inspections. Robson deny that any of their men would have over-tightened the nut using tools, and conclude therefore that the nut must have been over-tightened after completion. They also maintain that the debris was too small to have been capable of being realistically excluded from the thread.
10. As a consequence of the flood in Core 3, the water tanks (which served both Core 3 and Core 2) emptied. The water in the riser in Core 2 therefore drained down. Once the system equalised, and water began to come back up the riser in Core 2, there was a separate problem. The riser had been fitted with a surge arrestor at the top of the riser, a kind of valve which allowed air to get into the riser and prevent a vacuum from being created if the water level fell. But below the surge arrestor in Core 2 was both a non-return valve ("NRV") and, a little further down the riser, an isolation valve ("IV"). The NRV, which was closest to the surge arrestor, prevented it from working because it stopped the air from getting into the riser to prevent a surge in water pressure. In addition, if the IV was closed on 1 August 2007, that also prevented the surge arrestor from working, again because it also stopped the air getting into the top of the riser.
11. In this way, a vacuum was created at the top of the riser when the water level fell, so that when the water came back up the riser, there was a vacuum at the top, and no air cushion to slow down the velocity of the rising water. As a result there was a 'water hammer' event: the renewed water surge put pressure on the pipework that it was not designed to withstand. Pressure caused a breach in the riser on the sixth floor (the weakest point and not otherwise significant) and was the cause of the catastrophic Core 2 flooding.
12. The principal issue in this case concerns causation. After significant changes of case by a number of the parties, the final position was this. Robson and Essex both submitted that the sole or primary cause of the flooding was the closed IV. HL and HSE argued that the sole or primary cause of the flooding was the NRV, although both they and Essex advanced a secondary case that both the NRV and the IV were effective causes of the flooding. GMVL have always maintained a case that both the NRV and the IV were the causes of the flooding, although some matters of emphasis in the presentation of that case changed during the trial.
13. Unsurprisingly perhaps, the parties have adopted different cases on causation because of their potential liability for these different causes. Thus, if the NRV was the sole/primary cause of the flooding then Essex admits that the NRV was a breach, as do HSE; but both parties do so knowing that this is a claim that they can pass on down the line to Robson. HL say that, if it was the installation of the NRV, then the claim against them should be modest, because it does not involve any design failure. Robson deny that they installed the NRV but if they are wrong about that then, subject to my other findings, they have no defence to the claim. That explains the vigour with which they disputed the causal relevance of the NRV.
14. On the other hand, if the IV was the sole/primary cause of the flooding then the position is more complicated. This is principally because, in order to establish liability against Essex and/or HL, GMVL need to show, not only that the IV was closed at the time of the flooding, but that it was closed when this part of Holly Court was handed over to them on 18 August 2006. That is very much in issue. If GMVL fail to prove that the IV was closed at practical completion, then it is said – particularly by Essex and Robson – that they have failed to discharge the burden of proof and/or that GMVL are themselves responsible for closing the IV (probably inadvertently) after handover. On behalf of HL, Mr Hickey candidly acknowledged in his closing written submissions that, in all the circumstances of this case, that would be an "unattractive result", which perhaps explains one reason why HL have a rather different case on causation.
15. In addition, GMVL allege that, even if the IV was open at practical completion, and was subsequently closed, this was also the responsibility of both Essex and HL because they should have specified/installed a tamper-proof IV and/or put up clear warning notices to the effect that, if the IV was closed, then the surge arrestor would not work. Such notices were put up, but only after the flooding.
16. HL, Essex and HSE all deny the claim in respect of the tamper-proof valve and/or the warning notices, in part because, in their different ways, they each deny that they had the necessary design responsibility. That in turn gives rise to arguments as to the precise nature of the warranties which Essex and HL provided to GMVL, and the precise scope of the 'back-to-back' nature of HSE's obligations to Essex. There is also a debate about whether or not such valves should have been specified anyway, and, if so, what causative effect (if any) they, or the absence of notices, could have played.
17. The experts are agreed that the original flood in Core 3 accounted for 8% of the costs and losses (£380,000), whilst the flood in Core 2 was responsible for the remaining 92% (£4,370,000). The total claim is thus agreed at £4.75 million, together with interest to be determined.
18. The structure of this Judgment is as follows. I identify the relevant contractual chain in Section 3 below. I give an outline description and history of the development of the BMCWS in Section 4. I detail the flooding on 1 August 2007 in Section 5. In Section 6, I address the cause of the Core 3 flood. In Section 7, I analyse GMVL's claims against Essex arising out of the Core 3 flood and, in Section 8, I go to deal with the consequential claims (Essex against HSE and HSE against Robson) in respect of the Core 3 flood. In Section 9, I address the cause of the Core 2 flood from a legal and factual standpoint. I then analyse GMVL's Core 2 claim against HL in Section 10, and undertake a similar analysis of GMVL's claim against Essex in Section 11. I apportion the claim between HL and Essex for the purposes of the contribution proceedings in Section 12 below. Thereafter, in Section 13, I go on to deal with the consequential claims (Essex against HSE and HSE against Robson) in respect of the Core 2 flood. I address Essex's allegations of contributory negligence in Section 14. There is a brief summary of my conclusions in Section 15 below.

3. THE CONTRACTUAL CHAIN

19. The Main Design and Build Contract
20. Pursuant to a contract dated 28 January 2005, GMVL engaged LOR as the main contractor in respect of Holly Court and Becquerel Court, which were part of the Millennium Village project. The contract incorporated the JCT (With Contractor's Design) Form, 1998 Edition. The contract obliged LOR to carry out and complete the Works and to complete the design of the Works (clause 2.1). The design warranty in respect of the latter obligation was at clause 2.5 and equated LOR's liability thereunder to that of an "appropriate professional designer".
21. The Employer's Requirements included the design of the mechanical pipe work which had been carried out by HL. It was schematic only.
22. As I have already observed, as far as I am aware no claim has ever been intimated against LOR arising out of or in connection with the flooding on 1 August 2007. I note that LOR were also engaged as the overall remedial work contractors.
23. GMVL's Contract with WTP
24. Pursuant to a contract dated 15 October 1999, GMVL engaged WTP to act as their agents in relation to the works on site. Their functions seem to have been largely administrative, although they included the monitoring of the contractor's design development and the inspection of the quality of workmanship. Again, WTP seem to have been closely involved in the remedial works after the flooding events and again, like LOR, it does not appear that any claim against them has ever been intimated by GMVL.
25. HL's Warranty to GMVL
26. Pursuant to a contract dated 20 October 2006, LOR engaged HL as its mechanical and electrical consulting engineer. Such an arrangement is common when a design and build contractor is obliged to complete the design commenced by that same consultant. The contractual services provided by HL included design development, the preparation of production information for construction purposes, the examination of the Trade Contractor's drawings and details, and visiting the site at agreed intervals, "to see that"^[2] the works were completed fully in accordance with the contract documents. This last is conveniently categorised as HL's 'inspection' obligation.
27. Appendix 5 of the contract between LOR and HL, which set out the latter's detailed obligations, the performance of which HL then warranted to GMVL, included the following:

"... Develop the Detailed Design from the approved Design in the Employer's Proposals in coordination with the other Consultants and Contractors and prepare Production Information for formal issue to the Employer in accordance with the Construction Programme and within the Cost Plan.

Provide and agree with the Employer and regularly review on a two weekly basis a deliverable schedule for work up to completion of Production Information itemising each piece of design information to be submitted…

Prepare Production Information of sufficient detail for construction purposes and in full coordination with the other Consultants and Contractors as reasonably required by the Employer, Production Information to include but not be limited to fully dimensioned and detailed drawings, specifications [in NBS format], schedules, descriptions of materials and finishes, workmanship and quality standards.

Provide comments to the Lead Consultant on all Contractor and Supplier submitted design and technical information.

Attend regular meetings with the Employer and other members of the Project Team as necessary in order to monitor the progress of the Works and the issue of Production Information to the Employer.

Review with the Employer the design deliverables program in conjunction with the Information Required Schedule and the Construction Programme.

Examine the Trade Contractors and Suppliers drawings and details with particular reference to tolerances and dimensional and spatial coordination, interfaces, finish, durability, appearance and performance criteria. Report to Employer.

At all stages of Production Information check and coordinate the design produced by the Consultants and Trade Contractors for compliance with the Client's Requirements and Employer's Proposals and advise the Employer in writing of any likely non-conformities.

Coordinate and answer Request for Information queries (RFI) from the Employer and/or the other Consultants or Contractors.

In liaison with the other members of the Project Team issue to the Employer the requisite number of record drawings, schedules, specifications and guidance on the operation and maintenance of the project and its services as required by this appointment.

Before obtaining the Employer's approval to proceed to the next Stage carry out a check to see that all drawings, specifications, bills of quantities, schedules or other documents have been produced in compliance with the Client's Requirements, Employer's Proposals, the agreed Cost Plan and Programme.

Attend Consultants and Contractors design team meetings on site chaired and minuted by the Employer in connection with the design of the Works and to a programme agreed with the Employer.

Respond to technical queries raised and update and reissue drawings and/or specifications accordingly for construction purposes.

Visit the site at agreed intervals during the construction of the Works in order to inspect the quality of the work and to monitor progress and see that the Works are completed fully in accordance with the contract documents."

28. In consequence of this contract, HL executed a warranty in favour of GMVL dated 30 October 2006. The warranty included the following provisions:

"IT IS HEREBY AGREED as follows:

1. In consideration of the payment by the Beneficiary [GMVL] of £1…the Consultant ["HL"] hereby warrants and undertakes to and with the Beneficiary that:-

1.1 In respect of all the Consultant's Services to be performed under the Appointment the Consultant has exercised and will continue to exercise the reasonable skill care and diligence to be expected of a properly qualified and competent consultant of the Consultant's discipline experienced in performing duties similar to the Services on projects of a similar size scope and complexity to the Project…

…

1.5 For the avoidance of doubt the Consultant's liability for loss or damage should be limited to such sum as the Consultant ought reasonably to pay having regard to their responsibility for the same on the basis that all Consultants shall where appointed be deemed to have provided to the Beneficiary contractual undertakings on terms substantially not less onerous than those set out in this Agreement on respect of the performance of their obligations in connection with the project…"

29. There are disputes of fact as to whether, part way through the works, HL were told by LOR (i) that they did not need to check Essex's drawings and/or (ii) that they did not need to carry out further inspections to check on the quality of the mechanical works. This is despite the fact that both these services were expressly included in the contract between LOR and HL. It was HL's case that these obligations were somehow also excluded from the warranty with GMVL. In my view, for a number of separate reasons, these disputes must be resolved against HL on the facts, although I have also concluded that they are immaterial to the outcome of this case in any event.
30. As to design, the suggestion was that, at the technical review meeting on 21 June 2005 LOR indicated that their design team did not need to make any comments on Essex's electrical drawings. HL argue that they took this to mean that they were excused their obligation to make comments on the Essex/HSE mechanical drawings as well.
31. First, in my view, that is not what the minutes say. There is nothing to indicate that HL were included in this specific reference to LOR's design team, and nothing at all in the minutes about a waiver of the obligation to check the mechanical drawings. Secondly, this alleged waiver was not included or referred to in any subsequent documents, which is very surprising if it really was the widespread waiver for which HL now contend. Thirdly, despite the alleged waiver of this obligation by LOR, HL continued to claim precisely the same sums by way of their contract fee. Fourthly, the effect of the alleged instructions appears to have been negligible, since HL were in fact making design recommendations throughout the relevant period (see Section 4 below). For all these reasons, I find as a matter of fact that there was no agreed deletion of HL's obligation to review/comment upon Essex's mechanical design drawings.
32. Furthermore, the checking or otherwise of Essex's drawings is not the primary basis of the design claim against HL. HL's alleged design failings were more fundamental than simply a failure properly to review the Essex drawings. It was HL who recommended the surge arrestors: it is therefore said that it was HL who were obliged to ensure that their recommendation was properly carried through. That has nothing to do with an obligation to review the Essex drawings.
33. Beyond all of that, even if I am wrong, and HL's contractual obligation to review/comment upon Essex's drawings was subsequently waived, that would not affect the position under their warranty to GMVL. HL had provided a warranty to GMVL that they would carry out particular services properly. This included the obligation to check Essex's drawings. It is impossible to see how HL could avoid that obligation to GMVL under the warranty without at least informing GMVL of the proposed reduction in scope. It was common ground that they did not do so. I reject the suggestion that the terms of the warranty could have been changed unilaterally.
34. As to inspection, it was Mr Hart's evidence that Mr Belcher of LOR said that HL no longer needed to inspect the mechanical elements of the work. I note that no record or note of this conversation was ever made, and no reference to it can be found in any contemporaneous documents. Mr Belcher had no recollection of it at all. The only positive evidence in support of it came from Mr Hart.
35. In my view Mr Hart was an unsatisfactory witness. As I explain when dealing with the facts in Section 4 below, he never seemed to realise that the decisions of his design department were critical to the satisfactory design and installation of the mechanical services, and that this was something which would impact on the ultimate owners of the properties themselves. Instead, his principal aim seems to have been to minimise HL's role and responsibility, whilst maximising their financial recovery, without ever telling the ultimate beneficiary – GMVL – that this was the case. In relation to the whole saga of the surge arrestors, he attempted to justify his failure to address the issue on the entirely inadequate basis that the project was already up and running by the time that the need for surge arrestors became apparent.
36. On the balance of probabilities, I reject Mr Hart's evidence that the alleged conversation with Mr Belcher took place. If HL were no longer inspecting the mechanical works, then that was such a fundamental alteration to HL's agreed role that it would inevitably have been referred to by somebody, somewhere, in the contemporaneous documents. It was too important a change for it to go unremarked. Thus the absence of any reference to it anywhere supports my conclusion that there was no such instruction or conversation. So too does the absence of any reduction in the payments to HL under the contract; another telling point, given that the contract sum breakdown showed that a large proportion of the sums to be paid to HL by LOR were for inspection services.
37. Further and in any event, it was plain that, as a matter of fact, HL continued to inspect the mechanical works; Mr Gilbert of HL carried out the critical inspection on 17 August 2006. HL cannot therefore argue that they did not in fact perform the inspection obligation which they had warranted to GMVL.
38. In addition, HL were also inspecting the mechanical works pursuant to a separate contract of engagement by WTP. That separate appointment was evidenced by a letter dated 21 August 2006 from WTP to HL which said:

"…we confirm on behalf of [GMVL] that your fee proposal dated 10 January 2005 in the sum of £77,500 has been accepted and that GMVL approved your appointment as Monitoring Services Engineer in respect of the Holly and Becquerel Court project.

As stated in our previous correspondence, our appointment as Employers Agent is extended to include the employment of Monitoring Consultants…" (Emphasis added)

39. Accordingly, it would appear that, whatever happened to the inspection obligation agreed with LOR, HL were separately engaged by WTP, who were expressly acting as agents for GMVL, to carry out these same inspection obligations. Again, therefore, the alleged modification (accepting, which I do not, that it occurred) made no difference to what HL were actually doing on site.
40. Finally on the issue as to the alleged modification of HL's inspection obligation, I repeat the point at paragraph 30 above: even if I am wrong and the modification occurred, it did not and could not affect HL's obligations under their warranty in any event. It was again common ground that this alleged modification had not even been raised with GMVL, let alone agreed by them as a variation to the scope of the warranty.
41. Essex's Warranty to GMVL
42. By a warranty made as a deed on 23 January 2006, Essex warranted to GMVL various elements of their performance of their sub-contract with LOR. Terms of that warranty included the following:

"IT IS HEREBY AGREED as follows:-

1. The sub-contractor [Essex] warrants and undertakes to the Beneficiary [GMVL] that it has carried out and it will carry out each and all of its obligations duties and undertakings under and pursuant to the sub-contract when and if such obligations duties and undertakings shall become due and performable according to the terms of the sub-contract. Without prejudice to the generality of the foregoing or to any terms and conditions implied by law the sub-contractor also warrants and undertakes to the Beneficiary as follows:

1.1 In relation to the carrying out and completion of the Works:

1.1.1 The Works will when completed satisfy any performance specification or requirement included or referred to in the Sub-contract

1.1.2 As part of the Project that the Works and the materials and goods used therein will correspond as to description quality and condition with the requirements of the Sub-contract and be of sound manufacture and workmanship

1.1.3 That insofar as the Works have been or will be designed by the Sub-contractor its servant agents sub-contractors or suppliers they will when completed comply with all statutory requirements and building regulations…

1.2 That the Sub-contractor has exercised and will exercise the proper skill and care to be expected of a competent contractor experienced in projects being of a similar size scope and complexity as the Project in:

1.2.1 The design of the Works insofar as the Works have been or will be designed by the Sub-contractor and its servants agents sub-contractors or suppliers; and

1.2.2 The selection of materials and goods for the Works insofar as such materials and goods have been or will be selected by the Sub-contractor its servants agents sub-contractors or suppliers."

43. As is conventional, the warranty provided by Essex to GMVL related to their performance of their sub-contract obligations to LOR. However, in this case, that has created a particular difficulty. The sub-contract attached to the warranty by LOR (and signed off by both LOR and GMVL) was not in the same form in which it was eventually agreed between LOR and Essex. Thus it is said that there was a mismatch between Essex's obligations to LOR as evidenced by the form of sub-contract attached to the warranty, and Essex's obligations to LOR, as set out in the sub-contract that was eventually agreed. Again, the mismatch is said to be relevant to design only. However, for the reasons noted in Section 3.5 below, I have concluded that this dispute was not ultimately of any significance.
44. The Sub-Contract Between LOR and Essex
45. By a sub-contract dated 11 April 2005, LOR sub-contracted the mechanical and electrical works to Essex. This incorporated the domestic sub-contract DOM/2 Articles of Agreement, 1981 Edition.
46. Both the version of the sub-contract attached to the warranty, and the version of the sub-contract eventually agreed between Essex and LOR, contained the following documents/obligations relating to design:

(a) Scope of Works

This document stated as follows:

"The following is a brief summary of the items included within the Sub-Contract Works but it is not exhaustive. The scope of the Sub-Contract Work is everything which is shown or described or can be inferred from the documents contained or referred to within this Sub-Contract Agreement together with anything which could reasonably have been foreseen generally including inter alia:

1.1 Developing the design of the Mechanical and Electrical works to suit the ongoing scheme. The Sub-Contractor shall take ownership of the design concept prepared by Hoare Lea.

1.2 Supply, delivery, off-loading of all necessary plant, equipment, materials (including all fixtures and fittings) and to provide all required labour and supervision to carry out the installation of the Mechanical, Electrical, Public Health, Plumbing Works associated Builders Works, to include but not limited to the following…

Cold Water Supply including Cold Water Tanks, Booster Sets, Isolation Valves as necessary…

Booster Pumps…

Testing Commissioning and final Certification…

Preparation and submission of fully compliant O and M Manuals…"

(Emphasis supplied)

(b) Clause 1 of the Appendix to DOM/2 – Part 16

"Design Responsibilities

The Sub-Contract Works and the Sub-Contract Sum therefore shall include for

(a) The design thereof and the integration of that design with the works and/or designs of others, including the services and the structure, as indicated in any performance and/or other specification and/or drawing and/or bills of quantities and/or activity schedule and/or any other relevant information or documents…

(c) Everything necessary for the design development in connection with the Sub-Contract Works. 'Design development' is defined as the progression and completion of any design, including any amendments and/or consequential works, required to make the Sub-Contractor's Proposals fully satisfy the Employer's Requirements insofar as they apply to the Sub-Contract Works…

(e) Everything ancillary to and necessary for the proper execution of the Sub-Contract Works, including anything that may reasonably be implied and/or inferred as necessary for the safe and satisfactory installation and operation of the Sub-Contract Works."

(c) Clause 5.3.1 of DOM/2

"To the extent that the Sub-Contractor has designed the Sub-Contract Works (including any further design which the Sub-Contractor is to carry out as a result of a Variation required by the Contractor) the Sub-Contractor shall have in respect of any defect or insufficiency the like liability to the Contractor, whether under statute or otherwise, as would an Architect or, as the case may be, other appropriate professional designer holding himself out as competent to take on work for such design who acting independently under a separate contract with the Contractor, had supplied such design for or in connection with works to be carried out and completed by a building contractor not being the supplier of the design."

47. Although, prior to the conclusion of the sub-contract negotiations, Essex became concerned about the scope of their potential design liabilities, the contemporaneous documents show that LOR were never going to agree that Essex should have no liability for design. That can be seen from the following exchanges:

(a) At the meeting on 8 December 2004, where the minutes record that "Essex confirm that they will be liable for the design of the sub-contract works."

(b) Mr Sampson's (of Essex) emails of 10 and 11 February 2005 which Essex expressly accepted that part of the sub-contract stated that "Essex will take ownership of the design concept prepared by HL." Mr Sampson's oral evidence was that, because of his agreement to it, this provision expressly remained within the sub-contract.

(c) The exchanges on 15 and 16 March 2005 in which Mr Sampson identified various matters that Essex remained unhappy about and Mr Johnson of LOR's response which again made clear that the question of design was not negotiable.

48. Eventually, as part of these sub-contract negotiations, LOR and Essex did agree that "the Contractor's design responsibility under the following sub-clauses [of the Main Contract] will not apply to this sub-contract Agreement." The provisions identified were the design warranty provided by LOR to GMVL at clause 2.5.1 of the Main Contract (paragraph 19 above). It was this modification that was not part of the sub-contract attached to the warranty, and on which Ms Sinclair QC relied to argue that Essex had no design liability.
49. In my view, whilst it appears that Essex persuaded LOR to agree that the precise warranty that LOR had provided to GMVL would not, in its turn, be provided by Essex to LOR, this made little or no difference to the substance of Essex's design obligations. The contract documents/provisions set out in paragraph 41 above showed beyond any doubt that Essex owed LOR a series of clear design development/co-ordination obligations. They also provided a warranty to GMVL in respect of those obligations. That was all unaffected by the single deletion, which related not to the performance by Essex of specific design tasks, but to the much narrower issue of whether or not Essex were prepared to provide to LOR a similar warranty to that which LOR were providing to GMVL.
50. I should deal here with one particular submission advanced by Essex, to the effect that there was a discrepancy between, on the one hand, the agreement that the design warranty in the main contract would not apply to the sub-contract and, on the other hand, the detailed sub-contract provisions which made plain that Essex had a design liability. Ms Sinclair QC argued that, in those circumstances, the hierarchy clause in the sub-contract operated to provide that the deletion prevailed over all other parts of the sub-contract. I do not accept that submission for two reasons.
51. First, I do not agree that there was a discrepancy. The detailed design provisions were clear. The agreed deletion simply meant that Essex were not providing a warranty to LOR in connection with those obligations in the same terms as that which LOR were providing under the main contract. There was nothing unworkable or contrary about that.
52. Secondly, if I am wrong and there was a discrepancy, the hierarchy clause did not provide any help. The term provided that:

"Matters Agreed between the Contractor and the Sub-Contractor are as scheduled in the Addendum to the Appendix to DOM/2…part 15 Schedule D2SC/99/AN1 attached. In the event of conflict between the Matters Agreed and any other Terms and Conditions included in the Sub-Contract then Schedule D2SC/99/AN1 shall prevail."

The argument is that the 'Matters Agreed' included the deletion of the warranty. But that is said to prevail only over the 'Terms and Conditions'. The critical 'Scope of Works' clause referred to at paragraph 41(a) above (which referred to the taking of ownership of the design by Essex) was not a 'Term or Condition' as defined, and was therefore outside the hierarchy provision in any event.

53. In my view, the decision not to incorporate within the sub-contract one provision of the main contract concerned with design did not mean that all the rest of the sub-contract provisions dealing with design could somehow be ignored. Accordingly, for present purposes, I consider that Essex warranted to GMVL the proper performance of their design development/co-ordination obligations. Furthermore, such design obligations were entirely in accordance with the contemporaneous documents and the oral evidence of Mr Sampson, noted in paragraph 42 above.
54. I also consider that this result is in accordance with commonsense. Essex have always accepted a design development/co-ordination role, notwithstanding their argument about the deletion. As a specialist sub-contractor, Essex would ordinarily be responsible for taking HL's schematic design and ensuring that it worked on site. That would mean that they were obliged to check that schematic design and point out anything that they considered might be wrong (an obligation expressly accepted by Mr Sampson in cross-examination). In addition, they would ordinarily be obliged to ensure that the design coordinated with the works themselves, so that the scheme shown in the drawings could be taken to a state of physical completion on site. These obligations would also mean that Essex had to accommodate and deal appropriately with any variations to the design emanating from HL, particularly those variations (like the surge arrestors) which were made late in the day and involved the modification of existing designs and installations. There was nothing in any of the documents to suggest that this ordinary or typical situation did not exist here.
55. Accordingly, to the extent that there was a dispute as to whether or not Essex owed any relevant design obligations to GMVL pursuant to their warranty, I conclude that, whichever version of the sub-contract was or should have been attached to that warranty, Essex owed relevant design obligations to LOR in respect of both design development and design co-ordination, and they directly warranted the performance of those obligations to GMVL.
56. The Sub-Sub-Contract between Essex and HSE
57. Pursuant to a sub-sub-contract made on or about 4 April 2005, Essex sub-sub-contracted the mechanical works to HSE. It appears that the sub-sub-contract documents were the same as the sub-contract documents as between Essex an LOR. In addition, there were the minutes of a post-tender interview meeting dated 31 March 2005 and attended by Essex and HSE. These included repeated references to that sub-contract.
58. In addition, in relation to design, there were the following particular references in the minutes:

"1.1…HS are to be responsible for the co-ordination of the Mechanical and Electrical Services both spatially and with the building fabric as agreed previously between Chris Phillips and Brett Cowlbeck…

7.1 HS accept that the sub-contract will be back to back with that of the contract between LOR and Essex. Where the contract refers to 'Employer', substitute with 'Essex Electrical group Plc' and where the contract refers to 'Contractor' substitute with 'HS Environmental Service Ltd'." (Emphasis added)

Further, at 7.14 of the Minutes, there were references to the provision by HSE of certain design drawings by particular dates. They were generically referred to as "design development/drawing production".

59. HSE have always accepted that they had a design role, emphasising that this was limited to design development rather than some fuller design obligation. They also relied on the terms of the sub-contract between LOR and Essex. In all the circumstances, I have concluded that HSE had the same design obligations to Essex as Essex had to LOR, as set out in Section 3.5 above. I find that that is what 'back-to-back' meant at 7.1 of the Minutes. That conclusion is also consistent with the wider evidence, which was to the effect that Essex specialised in electrical works, not mechanical; that they entered into a joint venture with HSE (who were mechanical specialists) both in relation to this project and others; and that the HSE operatives often wore Essex-branded work clothes, so that other people at site would not have known that they were from HSE, not Essex. In addition, in cross-examination, Mr Phillips of HSE confirmed his understanding that, whatever design responsibility Essex had, HSE took on too.
60. For these reasons, therefore, it is unnecessary for me to deal in detail with the dispute between Essex and HSE about the precise formation of the sub-sub-contract between them. But in essence, I accept Ms Sinclair QC's submissions, on behalf of Essex, that the sub-sub-contract must be construed as a whole. If that approach is followed, it immediately becomes plain that HSE had a 'back-to-back' obligation in respect of design development co-ordination. I do not accept the submissions of Mr Hargreaves QC, on behalf of HSE, that, in some way, the court should ignore the 'back-to-back' provision in the Minutes, ascertain the obligations without reference to it and only then consider whether the words 'back-to-back' made any difference. That is a curious approach to contract construction. Moreover, I did not derive any assistance from the line of authorities relied on by Mr Hargreaves QC in support of this submission, and identified in Skips S/A Nordheim v Syrian Petroleum [1984] QB 599. Those cases are concerned with the particular rules governing the incorporation of conditions in a bill of lading. In the present case there was no incorporation by reference and the back-to-back clause was clear and explicit.
61. Still further, I consider that this is again a rather academic dispute, given that HSE have always accepted an obligation to develop/co-ordinate the design. In my view that covers the limited design obligations in respect of which GMVL assert a breach.
62. The Sub-Sub-Sub-Contract between HSE and Robson
63. By a contract, evidenced by Robson's quotation of 7 December 2004 and HSE's order of 8 July 2005, HSE sub-sub-sub-contracted the labour element of the mechanical works to Robson. It was therefore Robson who actually undertook the installation of the cold water pipework on site (although there are issues as to whether they installed the particular features of this installation which went wrong). It is not suggested that Robson had any design obligation whatsoever. Robson did, however, provide what was noted on the order as "working supervision".
64. The sub-sub-sub-contract included an indemnity in respect of:

"Any liability, loss, claim or proceedings of whatsoever nature such as shall arise by virtue of the breach or breaches of this Subcontract Agreement by, or act, default or negligence of the Subcontractor."

4. THE BOOSTED MAINS COLD WATER SYSTEM

65. HL's design of the BMCWS in these buildings was not particularly complicated. It incorporated one feature which is now becoming increasingly common in such designs: the water tanks were located in the basement, rather than on the top of the building. This is popular with developers, particularly in London, because it allows the top floors to be turned into penthouse apartments. It does, however, require the cold water to be pumped under pressure up the risers, rather than relying on gravity. The use of pressurised cold water can cause 'water hammer', where a surge creates water pressure and a burst pipe can occur.
66. The basic HL design implied the use of traditional copper pipes. However in recent times and for various reasons, including cost, copper pipes have commonly been replaced by plastic piping. That is what happened here, although when and how the change from copper to plastic occurred is obscure. One feature of plastic piping is that the valves are screwed into the pipe work with plastic nuts and the manufacturers recommend that any necessary tightening is done by hand rather than by any other method. Clearly the use of large metal wrenches, like Stilsons, runs the risk of damaging the plastic.
67. The original HL design was schematic and formed part of the Employer's Requirements, one of the documents in the main contract between GMVL and LOR. In its original form, it contained no method or device to deal with the possibility of water hammer. HL's schematic drawing of the water services, dated 11 February 2005 omitted any such device. However, the following month, in revision C2, the HL schematic drawing was revised and air vents were shown at the top of each riser.
68. In early 2005, in a separate development (which was wholly unrelated to this project) HL as a group became aware of a potential flooding problem caused by pressure surges in BMCWS. On 11 March 2005, HL held an extraordinary meeting of their engineers to discuss this problem, which was triggered by the flooding at a building which they had designed, called Grosvenor Waterside, West London.
69. The minutes of that meeting noted at the outset:

"WG introduced the recent problem experienced by the pressure tank cold water system (BMCW) at Grosvenor Waterside i.e.:

28^th February 2005 was to be hand over

Mechanical Contractor running late

BMCW pumps started up 'in a hurry'

BMCW pipe joints failed, resulting in flooding of the building

Flooding = £100k's damage

HL working under abridged duties, so we cannot just walk away from this problem (already costing HL money to control the situation)"

70. The minutes went on to consider HL's potential liability for this problem. At paragraph 3 the minutes noted that, if HL were working for a contractor, "we carry the can and every aspect of the BMCW design should be addressed from pipes, to joints, to valves etc." Of course, in the present case, HL were working for LOR as the mechanical services consultant, and therefore, in their words, they "carried the can".
71. At paragraph 4 of the minutes of the meeting on 11 March 2005, the HL engineers were asked to consider BMCW projects that were currently in progress, and to ask themselves whether anything in the relevant HL design needed to be modified. The minutes concluded with the promise that further guidance on BMCW systems would be produced in due course, and distributed.
72. Mr Hart was one of the HL engineers present at that meeting and the only one to give evidence at the trial. I have already noted that I found him to be an unimpressive witness. He repeatedly tried to suggest that, because the work for GMVL/LOR was already up and running by the time that HL became aware of the very real risks of water hammer in the BMCWS, HL were not obliged to consider any modifications to the existing design, and that it was an issue only to be considered for new designs in the future. That is plainly wrong as a matter of law: a designer owes an ongoing obligation to review his design in the light of accumulating knowledge; to warn of risks of which he becomes aware, and to modify the design if appropriate. But in any event these minutes, which were HL's own minutes of their own internal meeting, make plain that HL knew full well that they had to reconsider their mechanical services designs on all ongoing projects and to see whether any modifications were required.
73. Following that meeting, on 18 March, David Collins of HL sent all electrical and mechanical engineers at HL a number of documents, including an option report to provide failsafe BMCWS.
74. Meanwhile, on the GMVL project, in April 2005 Essex/HSE produced their own BMCWS schematic drawing. Despite the fact that air vents had been shown in the C2 drawing produced by HL the previous month, no such air vents were shown in the Essex/HSE drawing. Essex/HSE's evidence, given by Mr Smith, was that this was a deliberate omission; that Essex/HSE were concerned about the problem about surge and water hammer, and deliberately left off the air vents so as to provoke a debate about whether something more substantial was necessary.
75. I am bound to say that I found that evidence impossible to accept. A designer who is concerned about a particular design feature will specifically raise it for consideration; he or she will identify the particular problem about which they are troubled; and will often endeavour to suggest a solution. They will not typically omit an element of the work that they have been instructed to carry out without saying why, or fail to articulate the problem in the design with which they are concerned. Essex/HSE's elliptical approach led to the bizarre situation whereby their April 2005 drawing, which did not show air vents as instructed, and which also did not show any other means of preventing surges or water hammer, was the drawing to which all of the relevant risers were constructed. Mr Smith expressly accepted that, in consequence, the pipework was being constructed to a design and to a drawing which he knew to be defective.
76. At the trial, Essex/HSE also purported to rely on a later Request For Information ("RFI") dated 19 August 2005 in which it was said that they challenged this aspect of the HL design. It was said that this RFI expressly raised the question of water hammer/surge. But no mention of either was made in the RFI; instead it simply asked:

"Do you require vents at the top of each BCW riser in Holly and Becquerel Court. These are not indicated on drawings. These vents stop the compression of air at top of risers (hydraulic shock)."

This could have been a shrouded reference to water hammer, but it was accepted in evidence that it did not raise the issue directly. In any event, HL replied and said that air vents were required. To demonstrate how little Essex/HSE were concerned about this aspect of the design, even when they got the reply to the RFI in September 2005, Essex/HSE still did not amend their design to show the air vents upon which HL had insisted.

77. However, although neither Essex nor HSE knew it, HL's internal organisation was warning that water hammer/surge in risers such as these was a potentially serious problem. Thus on 27 May 2005, Mr Collins of HL sent a further set of documents to all HL's electrical and mechanical engineers which discussed various options to provide a fail-safe BMCWS. The email read:

"In the aftermath of Grosvenor, you're probably all aware that I was tasked with developing a new S13 section, pro forma booster set equipment schedule, and generic boosted water schematic, to take on board lessons learned. Bill and Ed have reviewed these documents, and they are attached so that ALL OF YOU can use them for the next project with BMCW in it…"

78. This documentation recommended what was called a 'combined air release/vacuum breaker valve'. At the trial they were known as surge arrestors. The pro forma specification said:

"This device shall incorporate automatic air venting with the facility to decelerate approaching water when air is venting. In addition, the device shall permit vacuum relief when draining down pipe work.

The combined air release/vacuum breaker shall be as manufactured by Cla-Val UK Ltd…at a pressure rating suitable for the maximum BMCW operating pressure. Final selection of the valves shall be determined by the manufacturer.

Install a lock-shield isolating valve at the head of the riser prior to the air release/anti-shock device to permit future maintenance."

79. In my experience, in a building defect/failure case such as this, it is rare to find a contemporaneous document which unwittingly predicts the failure to come. But that is what this document did. It identified a particular type of surge arrestor which would allow the air into the riser and form an air cushion to stop a water surge. It recommended that an IV should be installed so as to allow the surge arrestor to be properly maintained, and recommended that the IV should be of a "lock-shield" type, that is to say an IV which was made much more difficult to tamper with or to open or close inadvertently. Yet this document never found its way to any of the other parties in this case, and two years later a surge at Holly Court caused significant flooding because the surge arrestor in Core 2 had been prevented from working by an otiose NRV and a closed IV that was not a lock-shield type.
80. On 15 June 2005 Mr Collins sent another email to HL's engineers, including Mr Hart, enclosing a revised generic boosted cold water schematic and a pump equipment schedule. However, the generic specification referred to in the previous paragraphs remained unchanged.
81. The potentially significant problem which HL had noted in relation to pressure surges, and which was addressed in detail in Mr Collins' emails, was not translated into any amendments of HL's design of Holly Court or Becquerel Court. Neither were surge arrestors raised by HL with LOR as a modification or variation to the design which should at least be considered. Other than the RFI, which did not raise the question directly, and was dealt with by HL's repeated insistence on air vents (not surge arrestors), neither Mr Hart, nor anyone else at HL dealing with this project, appears to have considered that the water hammer issue was of any relevance to the project at all. As to Mr Hart's failure to require or even recommend that LOR/Essex/HSE use surge arrestors with lock-shield IVs, he again said in cross-examination that this was because this was an ongoing rather than a new project, an argument that I have already dismissed as unsustainable (see paragraph 65 above).
82. Many months went by without the surge arrestor issue arising again and it might have been forgotten entirely had not Michael Prosser of HL, who visited the site on 8 November, expressed his concern in an internal email that Essex had not received "any further details with regards to fitting surge arrestors at the top of each riser on the BCW." He went on:

"If they are to be fitted, we really do need to be issuing an instruction soon. The tanks are now built and boosters onsite, these are however yet to be piped up, but this does not take too long. I am concerned that this issue does not get over-looked."

83. Mr Prosser's email went to Mr Delvin of HL, who immediately contacted the principal designer, Mr Hart. Mr Delvin asked:

"Are Essex going to put these in? This is urgent as the detailed design was carried out when we knew about the enhanced requirements. Can you talk to Essex and understand what they priced/are willing to do?"

Although Mr Hart purported to argue in his oral evidence that it was not for HL to require or recommend the installation of these surge arrestors, I consider that the language of Mr Delvin's email makes plain that, on the contrary, it was their responsibility, and that he knew it. It was this email which prompted Mr Hart finally to take action; until then, it was something that he had wholly overlooked.

84. On 9 November, Mr Hart emailed Mr Cowlbeck of Essex, with a copy to Mr Belcher at LOR. He said:

"We are aware of a number of incidents in recent times of major failures of boosted cold water systems resulting in significant flooding of apartments.

The prime reason for this would appear to be when the booster set comes back on after a period being off, either as a result of a power failure or as a result of a maintenance shutdown. Basically there is a surge of water as power is restored and this can burst pipes and joints. A number of these floods have been during construction and have resulted in significant insurance claims.

In order to reduce the likelihood of this occurring we now incorporate into our new designs a special valve that automatically allows air in and out (to prevent a vacuum or trapped air) and also contains a surge arrestor. We attach detail page of a suitable unit as available from Grundfos Pumps.

The scheme does not currently incorporate these valves, only AAV's [air vents] as it was designed some months ago. We think that you should seriously consider incorporating these valves, one at the top of each of the nine risers in Holly and Becquerel Courts. List price of each valve is £368. Allowing for a price discount of the sub-contractor and allowing for fitting we would expect the final cost to be circa £4,500-£5,000. We consider this to be money well spent."

85. The attached sheet from Grundfos referred to the Cla-Val anti-shock and surge anticipation valves. In the detail, there was reference also to an isolating ball valve. There was no reference to a lock-shield IV. The generic specification document prepared earlier by Mr Collins of HL was neither referred to nor passed on to Essex by Mr Hart.
86. As a consequence of Mr Hart's belated recommendation, on 16 November, Essex requested a sub-contractor instruction from LOR for the supply and installation of surge arrestors. No such instruction was forthcoming by the end of November, when Becquerel Court was completed. However, on 29 November, when HL witnessed the commissioning, they noted that a shock arrestor had not been installed at the highest point of the system^[3]. Although there had been an argument about whether or not the work was a variation, on 29/30 November 2005, LOR instructed Essex to supply and fit a surge arrestor to the boosted cold water system in Becquerel Court. It seems that it was decided that, originally, this instruction would be limited to one block (see Mr Cowlbeck's email of 21 November). Neither the type of surge arrestor, nor the need for an IV, was specified by LOR, so the only inference that can be drawn is that both the surge arrestor and the IV were selected by someone at Essex/HSE. HSE purchased all the relevant materials required for the installation of all surge arrestors.
87. The failure to take proper responsibility for the specification and installation of the surge arrestors was not limited to HL. Essex, and therefore HSE, were responsible for design development and, as I have found, that included ensuring that any later variations or additions, such as this, were properly designed and co-ordinated into the existing building. Yet Mr Smith, who was in charge of the HSE design department, never saw the HL email of 9 November 2005 and was not informed of LOR's instruction to install surge arrestors. He was not asked to look at the manufacturer's literature or consider whether an IV was required below the surge arrestors. He said he did not know who else could have made such decisions, but someone obviously did.
88. To compound this muddle, the design drawings were not updated to show the surge arrestors and the IVs. It appears that this additional work was dealt with orally. That was plainly unsatisfactory, particularly in circumstances where none of those involved on site had ever come across surge arrestors before.
89. The evidence was that the first surge arrestor was fitted (at Becquerel Court) on 4 December 2005 by Mr Lawrence of Robson. He said that at that point he was not very familiar with installing such systems, a lack of familiarity he shared with just about every other witness, factual and expert: surge arrestors were very new in 2005, and witness after witness confirmed their lack of experience and understanding of this new piece of equipment. The daywork sheet which was completed to record Mr Lawrence's installation on 4 December showed a total of 8 hours work. He said he thought that slightly less than half that time would have been spent on the installation and a little more than half in relation to the commissioning and testing.
90. Mr Lawrence was asked how the surge arrestor would have been installed given that the riser had already been completed. He struggled to recall this particular installation, although he confirmed the importance of following the manufacturer's instructions. He also recalled that the surge arrestor came with an IV.
91. On 8 December, HL confirmed that a surge arrestor had now been installed at the highest point of the system in Becquerel Court.
92. On 21 December, LOR instructed Essex to fit surge arrestors to Core 4 (which included Cores 4A, 4B, 4C and 4D. It was (correctly) recorded that this was a "design recommendation" by HL. It appears that Mr Hughes of Robson carried out this work, as noted at paragraph 14 of his witness statement. Unfortunately, there were a number of oddities about the carrying out of this work. First, unlike the fitting of the surge arrestor at Becquerel Court, it appears that Mr McAllister of Robson thought that this work was part of the original contract scope (as per paragraph 22 of his statement); accordingly, there were no daywork sheets or paper trail of any kind relating to who carried out the work and when. Secondly, there was a delay in the carrying out of this work so that, when Core 4 was connected on 14 March 2006, a surge arrestor had still not been installed in that Core. It was installed by Mr Hughes of Robson two days later on 16 March. There was no explanation for why there was such a long delay between the instruction from LOR on 21 December and the actual carrying out of the work.
93. Thirdly and most importantly of all, the installation of the Core 4 risers was not properly carried out. One deficiency, which is otherwise irrelevant for present purposes, was the installation of an IV and a pressure reducing valve on the main riser, rather than on the branch serving the particular flats on the floor. More worryingly, all of the surge arrestors were installed on those cores without any IVs at all. In my view, these deficiencies arose because there was no design drawing, no specification, and no-one undertaking the work – in this case, Mr Hughes of Robson – who had any familiarity with surge arrestors and how they worked.
94. Although Mr Ter Haar QC suggested that IVs were unnecessary in those locations because the whole system could easily be drained down when these surge arrestors were maintained (as the blocks were not as high as the other blocks, so the risers were shorter), I reject that argument. It would have always been easier to use an IV to isolate pipe work rather than drain down the whole system, no matter how high or low the blocks might be. The HL generic specification made plain that an IV was required in every location that a surge arrestor was installed, as did the different examples of the manufacturer's literature that I have shown.
95. These other defects matter because Robson argue that they could not have installed the critical surge arrestors in Holly Court (Cores 1, 2 and 3) as a matter of fact because (i) there is no paperwork linking them to that installation; and (ii) they would not have carried out such deficient work as can be seen in Core 2. But neither of those arguments stands up to scrutiny. There was no paper work in relation to Cores 4A, 4B, 4C and 4D either, yet it is clear from Mr Hughes' evidence that Robson installed those surge arrestors. In addition, there can be no argument based upon the generally high quality of Robson's work in circumstances where the Core 4 work was, on any view, badly carried out.
96. In early January, there was a potentially important exchange between Essex and HSE. Although the origin of the exchange appears to have been nothing more dramatic than a leak from a towel rail, I consider that it has wider significance because leaks to various fittings and pieces of pipework were becoming more common on site, and were to become even more frequent in the months ahead. Essex pointed out to HSE that the onus was on HSE to endeavour to try everything possible to resolve the problems of leaks. In his response of 16 January, Mr Phillips of HSE reminded Mr Emrich of Essex that "we both challenged HL's design approach to this system at a very early stage with regards to pressure, temperatures, the different rates of expansion between plastic and metal etc etc, and were basically ignored. That is why we were all so relieved when LOR agreed to remove any design responsibility from you/our contracts." For the reasons set out in Sections 3.4, 3.5 and 3.6 above, that was an erroneous view of the contractual situation, but the exchange also reveals that, although neither Essex nor HSE saw fit to tell LOR or GMVL about it, they were apparently suspicious of HL's BMCWS design from the outset.
97. On 19 January 2006, LOR sent Essex an instruction to supply and install surge arrestors in Holly Court in Cores 1 and 2. It does not appear that anything happened in relation to that instruction. Four months later, on 2 May 2006, HSE recorded that they had been instructed by Essex to supply and install surge arrestors in Cores 1, 2 and 3 to prevent hydraulic shock within the BMCWS. On the same day, LOR issued a further instruction to Essex which now included Core 3 of Holly Court.
98. In May, it became apparent that the BMCWS in Holly Court was repeatedly being filled up and then drained down because of problems of leakage, stolen equipment and the like. On 12 May, HL's Mr Prosser emailed WTP to say:

"I am a bit concerned that the Boosted Cold Water has been switched on in the Core 3 riser, yet the upper floors are incomplete and there is no surge arrestor fitted at the highest point in the riser. This is leaving this Core exposed to malicious damage, as well as increasing the risk of a pipework failure. I did mention it to LOR when I was on site (not Dave Byford) and they informed me it was being shut off (please check), but it is something which needs to be treated with caution, usually this is opened up when a block is being fully commissioned."

In response, Mr Amundsen of HSE said that they were only turning the water on for the purposes of filling and providing domestic services and that they were requested to leave water on to prove flushing and the like during the inspection periods. He promised that HSE would ensure that the BMCWS was isolated when they were not using it.

99. A similar exchange involving HSE took place toward the end of May. Mr Sofroniou of Robson complained that, once Robson had filled, tested and isolated the water supply to a flat, it was not their obligation to continually turn the water on in the morning and off at night on behalf of HSE. HSE explained to him that this was necessitated by flooding due to a tap being left running. Mr Sofroniou replied again to say:

"DGR acknowledged that nobody wants any more floods at Greenwich but stand by my comments that once we have field tested and isolated a flat to HSE's satisfaction our scope of work is complete. DGR should not be demonstrating any works to your client LOR, or have no involvement in the chlorination and therefore it is the responsibility of your sub-contractor to ensure that he isolates the BCWS when his works are complete…"

The problem with this stance was that it ignored the piecemeal way in which Robson's contractual workscope was being completed (see below).

100. Like the other instructions to install surge arrestors, the installation at Holly Court did not progress promptly. When HL inspected the mechanical system in Core 3 on 20 June 2006, they noted that there was no surge arrestor at the top of the riser. It was only on that day that Robson provided a quotation to HSE for the installation of surge arrestors for Cores 1, 2 and 3.
101. At that point, the paper trail goes cold. There is no piece of paper that indicates that HSE accepted Robson's quote, or that Robson carried out the installation of surge arrestors in Cores 1, 2 and 3. Mr Lawrence simply could not remember whether or not he had carried out that work. His memory was of course equally deficient in relation to Becquerel Court: but there, unlike here, there was a daywork sheet to act as a prompt.
102. There can be no doubt that, as a matter of fact, surge arrestors were installed in Cores 1, 2 and 3. Photographs show them in place. I conclude that they were installed in July or early August, because the quote was dated 20 June (so installation could not have been before then). As we have seen, all other surge arrestors were also installed very late in the construction process. On the balance of probabilities, I find that, although it was unclear which of their plumbers actually did this work, this work was carried out by Robson. There are a number of reasons for that. First, it was Robson's men who did the same work in other areas: Mr Lawrence in Becquerel Court, and Mr Hughes in Core 4. Secondly, there was no evidence of anybody else carrying out such work on this site; it was only Robson who even quoted to do this work. The fact that they did not appear to receive any written instruction under their sub-sub-sub-contract with HSE was an oversight but, as I have already noted, that did not prevent them from carrying out the same work in Core 4.
103. Thirdly, there is no evidence that HSE instructed any other party to install these items. Mr Prowse, the relevant HSE supervisor, made clear that he did not instruct anyone other than Robson to install these items. He added that he did not see anyone other than Robson installing surge arrestors on the site. Fourthly, Mr Lawrence of Robson changed his original witness statement – which said that he did not install the surge arrestors in Holly Court – to the much more vague statement that he did not remember installing such surge arrestors. That watering down of Robson's evidence on this topic provides at least some support for the suggestion that Robson's attempt to avoid responsibility for the critical installation was principally based on the simple absence of written records.
104. Fifthly, there was no other credible candidate who could have installed these surge arrestors and the corresponding IVs. Indeed, one of the less edifying aspects of this case was the willingness of Robson's witnesses to blame other contractors for the defective installation in Core 2 (both in respect of the surge arrestors and the NRV) when, on analysis, there was simply nothing to sustain such an allegation. In my judgment, to suggest that other sub-contractors, whose vans were sometimes on site, were going in and carrying out unheralded modifications and extra work to the BMCWS pipework that Robson were installing was both irresponsible and damaging to the credibility of Robson's case as a whole. Mr Sofroniou, the principal Robson employee who had intimated this case, was cross-examined about it and was quite unable to justify the suggestions that he had made in his witness statement.
105. The most important surge arrestor for the purposes of this litigation was the surge arrestor at the top of Core 2, in the riser cupboard at Level 8. It appears that this surge arrestor was itself properly fitted. There was an IV, although whether that was installed in accordance with the manufacturer's instructions^[4] is open to question. Between the IV and the surge arrestor there was an NRV. This served no useful purpose whatsoever. Although there was some speculation that it may have been required for a temporary water supply, or to act as a double check valve, there was no factual evidence to support either theory.
106. I make the following findings of fact about the NRV. First, I conclude that it was installed either before or at the same time as the surge arrestor and the IV. The surge arrestor was installed so late in the sequence that it seems highly improbable that the NRV was installed afterwards. Moreover, if the surge arrestor was in place first, then the utter redundancy of installing an NRV thereafter would have been even more apparent, which also suggests that it was installed at the same time or before. Secondly, and for the same general reasons noted in paragraphs 95-97 above in connection with the surge arrestor, I find that the NRV was installed by Robson. It was part of the riser system which Robson had installed, and there was no record to suggest that anyone else had installed, or been instructed to install, the NRV. Thirdly, I find that it was completely useless, and contrary to any sensible design intent. As Mr Lawrence said at paragraph 23 of his statement, its installation "makes no sense". Fourthly, I find that it entirely negated the purpose and function of the surge arrestor because it prevented the air (which the surge arrestor was designed to let into the riser), from getting beyond the NRV and into the riser.
107. Since the NRV was missed by so many people, I find that it was either ignored altogether, or that it was assumed to be some integral part of the surge arrestor mechanism with which no-one was familiar (except Mr Hart, who did not attend site at all). The latter explanation shows the dangers of introducing a new and unfamiliar design feature in such a half-hearted and undocumented manner.
108. In July or early August 2006, when these items of equipment were installed, some parts of the Core 2 BMCWS had been tested, whilst other parts had not. For example, as Mr Hargreaves QC demonstrated in his cross-examination of various Robson witnesses, although the riser had been tested back in May/June, because various elements of the work were going on in the individual flats thereafter, the pipework branches to those flats could not be formally tested until much later that summer. Thus, there was a lengthy period when the IVs on the risers and the branches would have been being turned on and off to accommodate testing on the one hand, and further works to the system on the other. That situation extended all the way into August 2006.
109. The critical Core 2 riser was the subject of a final inspection before handover on 17 August 2006. The witnessing engineers were Mr Dayal and Mr Gilbert of HL. The written notes of their inspection revealed that, even then, not all the services were complete and that, for example, a cleaner's sink and water heater had not been installed at the time of their inspection. Thus further work to the BMCWS was still required after this inspection. The fact that the Heat Interface Units (HIUs) were not working on some floors was also evidence of the incomplete nature of the pipework installation at that time.
110. The photograph taken by Mr Gilbert of the Core 2 riser cupboard on Level 8 was taken as part of the inspection on 17 August. In it, two important things can be seen. One is the NRV, which neither Mr Gilbert, nor anyone else, realised was performing no positive function and was preventing the surge arrestor from working. It was only the disclosure of this photograph during these proceedings which led Essex finally to admit that the NRV had been installed as part of their sub-contract works.
111. Secondly, and more importantly, the photograph shows the IV in the closed position. This too prevented the surge arrestor from working. One critical issue is whether or not the IV was closed at the time of practical inspection and handover, which occurred the following day on 18 August. The photographic evidence is plain: the IV was closed at that time. The only contrary evidence came from Mr Gilbert who suggested that, at the time of his inspection, but after he had taken the photograph, he had asked an Essex employee to open the IV.
112. I consider that there are insurmountable difficulties with this element of Mr Gilbert's evidence. First it was extremely vague. The Essex employee was not identified and nobody at Essex, HSE or Robson claimed even to have been present at this inspection, let alone to have carried out this task. Secondly, no mention is made of this event in Mr Gilbert's written report of the inspection.
113. Thirdly, Mr Gilbert's memory was generally very poor. That is unsurprising, given that he would have seen numerous cold water systems over the last eight years. But when he was cross-examined, he repeatedly said that he could not remember particular aspects of the design and installation at Holly Court or Greenwich Millennium Village. At one point, he said that his memory "was not as good as it used to be". He could not remember carrying out any other inspections of riser pipework on any other project, although he must have done so. All this must call into doubt his assertion that he remembered asking an unnamed man to open the IV on 17 August after he had taken the photograph. Whilst at one point he suggested that his inspection on that day was "memorable" there was, on his own case, nothing to justify such a description.
114. Fourthly, Mr Gilbert only made this assertion (about opening the IV) in his witness statement served on 11 April 2013, almost 7 years after the inspection, and just 2½ months before the trial. Such a delay, on such an important issue, stretches credulity to breaking point and beyond, particularly in circumstances where HL originally pleaded that they were unable to admit or deny that the IV was closed on practical completion.
115. Fifthly, Mr Gilbert admitted in cross-examination by Mr Stansfield QC that he knew that the fact that the IV was shown as closed in the photograph constituted a defect. It was therefore put to him that, if he had noticed that this IV was in a defective state and had asked for it to be rectified immediately (by the opening of the IV), he would have realised that there were potential defects in all the riser cupboards where there was a surge arrestor, and that he would have immediately gone round to ensure that the other IVs were also open. Mr Gilbert agreed that this is what he would have done if he had seen the IV was closed, but he admitted that, in fact, he had not undertaken any such course of action. Again that seemed to militate against his suggestion that he had recognised the closed IV in this cupboard as a defect on 17 August 2006.
116. There is one final point which I consider to be critical. On an inspection to see whether or not construction works have reached practical completion/handover, an inspecting engineer like Mr Gilbert will take a photograph to record what is there. Its purpose is to demonstrate beyond doubt what the engineer has seen on that inspection. An inspector in the position of Mr Gilbert will not ordinarily take a photograph of something that is obviously defective, if that defect could be cured, as here, by one turn of a handle and a subsequent test to ensure that the surge arrestor was working. If he had spotted the problem, Mr Gilbert would have opened the IV (or had the IV opened) and only then would he have taken his photograph. That would have been the case whether the photograph was a formal document or, as Mr Gilbert suggested, was just for his own purposes. Accordingly, for all these reasons, I conclude that the photograph is a clear record of the installation that Mr Gilbert saw and accepted on 17 August.
117. It is not inherently unlikely that this defect – the closed IV – existed at the time of practical completion. Here, at the top of the riser on Core 2, all that the closed IV was doing was closing off the last section of pipe and the surge arrestor beyond. That was not going to have any effect on the operation of the BMCWS unless and until there was a water surge and the surge arrestor failed. This meant that it was entirely plausible that the IV stayed closed from August 2006 to August 2007. Mr Marshall, Essex's M and E expert, expressly agreed with that proposition.
118. At the inspection on 17 August, Mr Gilbert and Mr Dayal did not have any form of record drawing or other document against which they could check the installation. However, it appears that, three days beforehand, a record drawing had been prepared by Essex (6130/M/001 Rev RD). Not only was this drawing not available at the handover/inspection but it was also not a record of what had been installed. First, it did not show any IVs below the surge arrestors. Second, it did not show the NRV in Core 2. Mr Smith of HSE said that it was a record compiled from other records that had been marked up on site but he could not explain what these records were and how the drawing could have been so wrong.
119. These omissions might have been made good in the voluminous O and M Manuals with which GMVL were provided. However the evidence about the roundabout production of these Manuals was depressingly similar to other aspects of this case. The work was sub-contracted to a specialist document-provider (E Documents) which had had no prior involvement in the project. The various parties in the contractual chain were forever being asked to comment, review and approve the work that E Documents did, but no-one actually involved in the design and construction of these buildings took an overall grip on the contents of the O and M Manuals. In this way, for the mechanical work at least, the O and M Manuals did not correlate to the design and installation on site. In particular, there was no mention in the Manuals of the surge arrestors, and neither was there any mention of the IVs, and the critical need to keep them open at all times save for when the surge arrestor was being maintained. It goes without saying that there was no mention whatsoever of the NRV.
120. However, despite this catalogue of omissions, HL confirmed to WTP on 18 August that "the works have been satisfactorily completed and generally comply with the Employer's Requirements." This was in accordance with their contract with WTP (see paragraph 35 above). On the same day they confirmed to LOR "that the elements of the Works which are covered by our scope of services have been completed in accordance with the design intent." I find that this was in accordance with their contract with LOR (paragraph 24 above) which had not been modified as suggested by Mr Hart (see paragraphs 31-34 above). As to the words used, Mr Hart suggested that this was a meaningless form of words which had been provided merely by way of contractual obligation. He even suggested that HL could legitimately provide that certificate to LOR without having to inspect what was actually on site. I am afraid that his answers typified his whole approach to this project, and I reject them. The written certificates from HL were important documents which gave the relevant mechanical works a clean bill of health. It follows from my analysis of the facts that, in my view, they should never have been issued.

5. THE CATASTROPHIC FLOODING ON 1 AUGUST 2007.

121. The various buildings were handed over to GMVL at various dates throughout 2006. It appears that, following handover, there were ongoing problems with the cold water pipework. Mr Crawford, the Village Manager, referred to these persistent leaks and said that the staff used to joke that the site should be called the "Greenwich Milleakium Village". To some extent, this is supported by the documentary records of the call-outs to deal with the problems in the flats where there are numerous references to leaks.
122. On the other hand, it does not appear that there were any significant flooding events, certainly nothing on the scale that was to occur on 1 August 2007. It seems that the leaks were attended to as a matter of routine and nobody thought that their nature or frequency were such that there was any wider problem with the BMCWS.
123. The design and build contract with LOR had a two year defects liability period ("DLP"). It appears that GMVL operated on the basis that, during that period, LOR or their sub-contractors would return to attend to any problems on the site. It was only when that two year period was up that a separate M&E maintenance contractor, NG Bailey, was appointed for the task. However, whilst LOR were contractually obliged to return to rectify defects, many of the events that occur in the first two years of the life of any new building would not be the result of defective work as such, but would instead require ordinary or preventative maintenance.
124. Presumably with that in mind, GMVL appointed Pinnacle to act as a general maintenance contractor. It appears that, pursuant to their contract, Pinnacle's maintenance obligations included the maintenance of all plant, including the cold water risers. But Mr Crawford, who was employed by Pinnacle, rejected the suggestion that Pinnacle would actually carry out extensive maintenance work to the risers.
125. These arrangements are revisited in Section 14 below, in the context of Essex's allegations of contributory negligence. But it is important to deal with the case made by Essex, HSE and Robson that the IV might have been closed after handover and before the flooding on 1 August 2007, as part of the maintenance work. In my view, on the evidence, such a case has not been made out.
126. I have already found that the IV was closed at the time of handover. In my view, there can be no doubt that the IV was also in a closed position at the time of the flooding on 1 August 2007. I make that finding for two reasons. First, Dr Sanders had been the first expert on the scene and was there about seven hours after the leaks were first discovered. He photographed what he saw in the riser cupboard on the 8^th floor and that showed that the IV was closed. Secondly, in a note prepared by Essex in the bundle at C10/90, they indicated what they had seen and noted on their inspection on the same day. That note also recorded that, at the time of the inspection, the IV was closed.
127. There is no reason to believe that anybody else had been in the Core 2 riser cupboard at level 8 (the top floor) before Dr Sanders and/or the Essex representative. This was because everyone was concentrating on the riser cupboard on the 6^th floor, where the water was flooding out of the BMCWS. It is fanciful to suggest that an unknown individual rushed into the riser cupboard two floors above the flooding and turned off an IV (which was not, on the face of it, anything to do with the flooding at all). Accordingly I find that the IV in question was closed at the time of the flooding. Thus, the most likely scenario is that the IV was closed at the time of handover 11½ months before; it stayed closed during the period between handover and the flooding; and it was closed when the flooding occurred on 1 August 2007.
128. The contrary suggestion that the IV was closed after handover, when it was the responsibility of GMVL, would require the IV to have been opened at some point after handover, and then closed again before the flooding, by those who had access to the riser cupboard. However, although it is plain that there were a number of people who had access to the riser cupboard in Core 2 on Level 8, there was no suggestion that anybody turned, or would have had any reason to turn, this IV from one position to another. There was no evidence of any problems in this riser cupboard that would have necessitated any such operation: the suggestion that the surge arrestor might have 'dribbled', so that closing the IV may have been one way of dealing with it, was entirely hypothetical.
129. Furthermore, if (as I have found) the IV was closed at handover, then GMVL, and those carrying out maintenance on its behalf, were entitled to assume that this was its correct setting (particularly as the O and M Manuals made no mention of these IVs at all). The maintenance contractors, porters and others working for GMVL would have had no reason to open the IV: even if there had been leaks and the like, that would not have led anybody to open an IV that had been closed at the time of handover. Accordingly, for all these reasons, I find that, on the balance of probabilities, the IV which was closed on handover remained closed until 1 August 2007.
130. On 1 August 2007, the two failures happened in sequence. On Core 3, a plastic connecting nut attaching a ball valve to the riser branch at Level 2 failed which allowed the joint on the downstream side of the valve to separate. As a result, flooding occurred in Core 3.
131. The escape of the water in Core 3 gradually caused the tanks to empty because the pumps pumped water more quickly than the mains supply could replace it. When the tanks were largely emptied, a low water level sensor was triggered which caused the pumps to switch off. Because the surge arrestor at the top of Core 2 was isolated by either the NRV or the closed IV or both, no air entered the riser to replace the water which drained away and a vacuum formed at the top of the Core 2 riser. With the pumps off, the water tanks refilled from the mains.
132. In time, the water rose sufficiently to disengage the low water level sensor. The pumps were fitted with an automatic restart facility so the pumps came back on and quickly came up to full speed. A column of water shot up to the top of the Core 2 riser, causing a water hammer event. There was a resulting spike in water pressure in the Core 2 riser. That pressure then caused a failure in the weakest part of the system which was a plastic flange attaching a ball valve to the pipework on the riser branch at Level 6. Thus the origin of the flooding was the riser cupboard on Level 6 of Core 2.
133. Large amounts of water flooded into the sixth floor and the floors below. It ran into the raised floors, out onto balconies and cascaded down the side of the building. The damage to some of the individual flats was extensive.
134. Amongst the eye witnesses who saw what happened was Mr Crawford, the Village manager, who said that he saw "a torrent of water cascading off the balconies of the upper floors down the front of Holly Court." He said that, even though the main water supply had been stopped from the street, the water which was held in the large tanks in the basement of Holly Court was still being pumped up through the BMCWS to the upper floors, so that it continued to escape through the breached pipes. Water continued to pour through the ceilings and down the walls of Holly Court until the pumps in the plant room were turned off.
135. Mr Bacon, who worked for GMVL as the head of their customer care department, was alerted to the news and drove to Holly Court. He said:

"As I was driving along the road towards the village I saw from about 100 yards away a wall of water pouring out from the sixth floor windows and balconies and cascading down outside of Holly Court like a waterfall."

He went into Holly Court at about 7 am on the morning of 1 August, immediately after the pumps had been turned off. He said that "water was running down the stairwells and cascading down the walls mixed with debris falling from collapsed ceilings." Mr Crawford also noted the debris falling from ceilings and said that it was lucky that people had not been injured.

136. Originally, the emergency works were carried out by LOR/Essex on a day work basis. Some tenants had to be decanted into hotels before they could be allowed back into the buildings. Thereafter, a more extensive remedial work programme had to be put in place. Some of the defendants/third parties (and in particular HL) complained that the remedial works were carried out in an inefficient and unreasonable way, at least after the initial emergency was over, with the result that the remedial work costs were far higher than they should have been. However, despite these issues, part way through the trial, the parties sensibly agreed the quantum of the claim at £4.75 million plus interest. Interest is also agreed.
137. I now turn to address the claims. I start with the cause of the Core 3 flood (Section 6 below) and then address the claims arising out of the Core 3 flood (Sections 7 and 8 below). I then analyse the cause of the Core 2 flood (Section 9 below) and the claims arising out of that second, larger flood in Sections 10, 11 and 12 below.

6. THE CAUSE OF THE CORE 3 FLOOD

138. The Experts
139. There were a large number of experts in this case. GMVL called three liability experts: Dr Neil Sanders and Mr Graham Cooper, forensic engineers, and Mr David Gosling, an M and E engineer. Essex called two experts: Mr Nick Davison, the forensic engineer and Mr Steve Marshall, the M and E engineer. HSE's forensic engineer was Simon Clarke and their M and E engineer was Mr Hugh Parry. HL called just one M and E expert engineer, Dr Michael Humphries. Robson called two liability experts: Mr Rod Newbery, a forensic engineer, and Mr Michael Carver, an M and E engineer.
140. Despite these numbers, the expert evidence was clear and focused. The original Joint Statement, one of the most helpful documents of its kind that I have seen, was signed off on 10 May 2013. It ran to 38 pages. It made plain what each of the experts thought on each of the main issues and demonstrated clearly the extent of the (relatively limited) disagreements between them. There was a subsequent Joint Statement dated 12 July 2013, relating solely to the Core 3 Flood, signed only by the five experts concerned with that issue. Much of what I say about the cause of the two floods is taken directly from these documents.
141. The Experts' Evidence on Core 3 Causation/General
142. The cause of the Core 3 flood is dealt with between pages 5 and 16 of the first Joint Statement, as well as the whole of the subsequent Joint Statement. In essence, the forensic expert engineers agreed that the evidence indicated that the riser branch IV connection at Core 3, Level 2 failed due to circumferential cracking of the plastic nut that had connected the downstream side of the IV to the downstream pipe work. There was no evidence of a moulding fault or any other manufacturing defect in the nut.
143. Dr Sanders and Mr Cooper (GMVL), Mr Davison (Essex) and Mr Parry (HSE) all agreed that the water was at a pressure that was well below the allowable working pressure of the valve, so the flooding could not have been caused by valve failure or problems with the pipe pressure. Although Mr Newbery suggested a different view, he admitted that he had not identified any pressure in his report, nor the time over which any higher pressure may have been applied.
144. The experts were all agreed that the nut in question showed evidence of tool marks, and so they concluded that a tool must have been used both on the nut and the body of the IV itself. They agreed that the use of such tools on plastic nuts not only contradicted the manufacturer's instructions, but could also have caused the over-tightening of the nut. Indeed, Dr Sanders and Mr Davison were agreed that, in all the circumstances, the initiation of the crack was probably caused by excessive tightening. Mr Parry also agreed that that was a possible cause.
145. There was also evidence of metallic debris in the thread of the failed nut. These particles were small. However, a number of the experts, including Dr Sanders, Mr Cooper, Mr Davison, Mr Clarke and Mr Newbery, agreed that the particles could be seen in the location where the crack initiated. Some, in particular Dr Sanders, considered that the inclusion of the particles in the thread of the nut may have made it harder to tighten by hand, and thus gave rise to the over-tightening.
146. As to causation, therefore, the majority of the experts thought that the cause of the failure was over-tightening, or debris, or both. Mr Newbery, Robson's forensic expert, took a different line. He said that there was no evidence of over-tightening and that the problem could have been caused by significant fluctuations in water pressure, the overall design of the cold water system, and other related matters. Whilst it was generally agreed that these could have been causes of the failure, the other experts were clear that there was no evidence that suggested that they were the causes of this particular failure in Core 3. Furthermore, just before he gave evidence, Mr Newbery served a third report which indicated that he no longer considered that one particular water hammer event had caused the cracking of the nut. It appeared that he relied more generally on pressure surges, and was moving away from his original view, and towards the view held by the other experts.
147. For these reasons, therefore, I consider that Ms Sinclair QC was correct to note, at paragraph 16 of her written closing submissions on behalf of Essex, that "the overwhelming weight of the expert evidence was that the initiation of the crack in the incident nut was the combined effect of over-tightening and the metal debris in the thread."
148. Analysis

6.3.1 Metallic Particles/Debris

149. I start with the metallic particles. I find that the presence of such particles within the thread of the nut was a breach of British Standard 6700 at paragraph 3.1.2.1, which requires that "all piping and fittings shall be cleaned internally and shall be free from particles of sand, soil, metal filings and chips." I also consider that the evidence showed that the installation was in breach of the manufacturer's instructions, which warned that the valve had to be "protected from harmful, physical influences including dust." The manufacturer, Georg Fischer, also required that the valve be removed from its original packaging only "immediately before installation."
150. It was suggested (principally on behalf of Robson) that it was impossible on a building site to keep nuts of this kind free from all debris. The difficulty with that argument is that, of course, both the British Standard and the manufacturer's instructions require it, which indicates that it is both practical and possible. Moreover, even if a piece of equipment is contaminated in this way, it can always be cleaned and the debris removed.
151. In this context, I was struck by the cross-examination of a number of the plumbers (particularly by Mr Hargreaves QC and Mr Stansfield QC) which explored the possibility that the valve may have been placed on the ground whilst supporting brackets were cut in the same location. This evidence led me to conclude that this was the likely cause of the debris, which would also explain why this debris (the kind of material produced by such cutting operations) was found on the valve as well as the nut. The same point was made by Mr Clarke, HSE's expert, during his cross-examination by Mr Stansfield QC: he concluded that this was "the most likely explanation" of how the contamination had come about.
152. In addition, during his cross-examination by Mr Hargreaves QC, Mr Newbery (the relevant Robson expert on this topic) gave particularly clear evidence as to the importance of the contamination failure:

"Q: It is emphasised both in the [manufacturers'] instructions and in the British Standard, which we will look at in a moment, that in joints of all materials, you are best advised to keep metallic debris out of them?

A: That is correct.

Q: It is well known in the industry and it is no surprise to anybody in the industry, is it?

A: No."

After the British Standard had been looked at the cross-examination continued:

"Q: …is there any ambiguity or difficulty with the words 'free from'?

A: No.

Q: When it says 'metal filings', are those precisely what are produced by these processes on site that you refer to?

A: Yes.

Q: So in a sense Mr Newbery, it is less about a hand lens and a light source, and rather more about not letting metallic debris in there in the first place, isn't it?

A: Ideally there should be no metallic debris, I agree."

153. Mr Newbery accepted that, on his own approach, the metallic particles were actually the cause of the initiation of the crack. He went so far as to say that the failure of the nut would "almost certainly" not have occurred without the metal particles being present. As already noted, the other experts said generally that the metallic debris was one of two causes (the other being over-tightening) of the failure, and that it was impossible to say which of those two was the dominant cause. A number of them, such as Mr Parry, agreed that both the debris and the over-tightening were required in order for the nut to have cracked and eventually split open. On the basis of all the expert evidence, therefore, I consider that the metallic particles were, at the least, a significant cause of the failure of the nut.
154. There is one final point in relation to the metallic particles. Both Robson and Essex stress that there were only 5 or 6 of these particles and they were less than 0.5mm in diameter. It is therefore suggested that the presence of debris did not amount to a breach of the relevant obligations and/or even if they did amount to a breach, they were in some way de minimis and I could put them to one side. I do not accept that. The argument was based on the judgment of Diplock LJ in Margaronis Navigation and Agency Limited v Henry W Peabody [1965] 2 QB 430 at 448, where the judge said that the question was whether the obligation performed was within that margin of error which, in the circumstances of the case, it was "not commercially practicable" to avoid. But in the present case, there was no evidence that it was "not commercially practicable" to ensure that there were no particles of debris within the thread; indeed, on the basis of the evidence, the strong inference was that it was possible to keep the particles from the thread; it was only in this one location that any debris was not cleaned away. Moreover, as Mr Stansfield QC submitted in his closing submissions, it is impossible to see how something could be de minimis, when it was an accepted cause of the relevant cracking in the nut.
155. For all those reasons, I find that the metallic debris was an operative cause of the cracking in the nut on Core 3.

6.3.2 Over-Tightening

156. As noted above, the majority of the experts considered that the other dominant cause of the failure was the over-tightening of the nut. Of course, when a number of experts are called to give evidence on a point of this kind, it is wrong for the court to resolve the issue simply on the basis that the overwhelming majority must be right. Issues like this can never be decided on weight of numbers alone. But it was plainly significant that, of the four forensic experts giving evidence about this failure (there was no claim against HL in respect of the Core 3 flood, so Dr Humphries was not involved), four of them (Dr Sanders, Mr Cooper, Mr Davison and Mr Clarke) considered that the over-tightening of the nut, consistent as it was with the use of a tool or tools on the nut, was the most likely cause of the failure.
157. At paragraphs 21-25 of her written closing submissions on behalf of Essex, Ms Sinclair QC demonstrated the numerous changes of mind on the part of Mr Newbery (the one expert who had not signed up to this agreement), and the various modifications of his view which led him to reach conclusions which were much closer to these other experts than had originally been the case. It is unnecessary to set them all out, but they include the following:

(a) Mr Newbery originally said there was no evidence of significant over-tightening but he ultimately agreed that it was "possible to cause over-tightening to cause damage to the thread while producing the tool marks seen on the incident nut";

(b) Although Mr Newbery originally suggested that over-tightening the nut would impart a pre-load which would render it more resistant to fatigue cracking, he appeared ultimately to accept that, whilst that might be true for a steel bolt, it was not true for a polymer nut like this;

(c) Although Mr Newbery originally disagreed that over-tightening could cause a crack which grew over time, he ultimately conceded the point and accepted that he had not appreciated the speed at which brittle fractures could occur in this material;

(d) Although Mr Newbery appeared to rely on the 200 or so visible progression ridges as indicating that the nut had been subjected to 200 or so high pressure events, he ultimately accepted Mr Cooper's explanation that up to 90% of the fatigue life of a plastic component would typically produce no progression ridges at all;

(e) Although Mr Newbery originally stated that the over-tightening/debris theory could not explain the propagation or the final failure of the nut and that each would require more substantial loads than would be generated in the ordinary operation of this system, he ultimately abandoned both of those positions.

158. Accordingly, weighing the expert evidence in the round, I consider that the majority view was correct and that the other significant cause of the Core 3 flood was the over-tightening of the nut.

6.3.3 The Pump System and Pressure Surges

159. I do not agree with Mr Newbery's view that, in some way, the general design of the pump system, or the potential problem of water surges, caused the nut to crack. Although he maintained that view (up to a point) in cross-examination, there was no cogent evidence to support it. There was no evidence of any such surges and no evidence that there had been any significant leaks in Core 3 before. There was no evidence of the relevant mechanism of failure. It seemed to me that this was merely a theory which was not justified by any detailed analysis of the evidence in this case.
160. I find that there was no cogent evidence to suggest that high pressure surges had occurred either before or after the handover of Core 3. On that topic I accept the submissions at paragraph 23 of Ms Sinclair QC's written closing submissions. Mr Newbery was driven to postulate pressure surges which imposed just enough stress on the nut, for just the right amount of time, to cause cracking but without causing brittle fracture, yet he was wholly unable to provide evidence that any such surges occurred. To the extent that reliance was placed on subsequent inspections in 2009 and 2012, which showed nuts which had suffered more severe external damage but had not failed, I consider that such findings are irrelevant to what happened up to, and on, 1 August 2007. The subsequent damage could have been caused in any number of ways. What matters are the pressure fluctuations before 1 August 2007 and I find that they were within the ordinary range of fluctuations of a BMCWS of this kind.
161. Accordingly, notwithstanding Mr Ter Haar QC's written final submissions on this topic (which presented this hypothetical case rather better than Mr Newbery had done orally), I reject the design/water pump/surges theory as explaining, even in part, the failure of the nut. Thus, for all the reasons noted above, I conclude that the flood in Core 3 was caused by a combination of the metallic particles in the thread of the nut and the over-tightening of that nut (the latter possibly being caused or contributed to by the former).
162. Summary
163. There were therefore two co-operating causes of damage, but for either of which the damage would not have occurred. In those circumstances, should it be relevant, it would be sufficient for GMVL to establish that one of those causes was the responsibility of Essex in order to establish liability: see Heskell v Continental Express Ltd [1950] 1 All ER 1033. In that case, Devlin J (as he then was) held that if a breach of contract was one of two separate causes of the failure/loss, and that each were co-operating causes, that was sufficient to establish liability. That approach was adopted by this court in Great Eastern Hotel v John Laing Construction [2005] EWHC 181 (TCC), and the authorities cited at paragraph 314 of the judgment in that case.

7. GMVL'S CLAIM AGAINST ESSEX ON CORE 3

164. The Metallic Particles
165. As noted above, I find that, on the balance of probabilities, the particles contaminated the nut when the valve was placed on or near the ground when (unrelated) cutting work was done. The presence of metallic particles in the thread of the nut was not only causative of the Core 3 flooding, but also constituted a failure to comply with British Standard 6700 and the manufacturer's instructions. For the reasons set out in paragraphs 140-144 above, I reject the suggestion that it was impractical or unrealistic to keep them all free of all metallic particles, or that in some way the particles were "de minimis".
166. Because the work failed to comply with British Standard 6700 and the manufacturer's instructions, the presence of the metallic particles in the thread of the nut constituted a breach by Essex of their sub-contract with LOR, and thereby constituted a breach by Essex of their warranty to GMVL. Although there was an argument as to whether the sub-contract and the warranty were co-extensive (which I have in any event resolved against Essex), that relates only to questions of design liability, not (as here) a matter of pure workmanship. Accordingly, I find that the presence of the metallic particles meant that Essex were in breach of the terms of their sub-contract with LOR and in breach of their warranty to GMVL. In accordance with paragraph 152 above, that would be sufficient to establish Essex's liability, even if the over-tightening was post-handover (and therefore not their responsibility).
167. The Over-Tightening of the Nut
168. The over-tightening was evidenced by the use of a tool on the plastic nut. The use of such a tool was agreed as a breach of the manufacturer's instructions. The factual issue was whether this over-tightening was carried out by Robson (on behalf of Essex) prior to handover, when it would constitute a breach of Essex's sub-contract, or by those responsible for the maintenance after handover, when it would have been GMVL's responsibility.
169. Although I reject the criticisms of the arrangements relating to the maintenance of these buildings after handover, even if they had been well-founded, they were based on GMVL doing too little by way of maintenance, rather than too much. That does not generally support a case that the over-tightening was due to GMVL's maintenance regime. More significantly, Mr Crawford was clear that his maintenance staff were not allowed to touch (and would not have touched) the valves at all. Instead, they would have passed such problems back to the contractors returning under the DLP which, ultimately, meant Essex (acting as LOR's sub-contractor). Mr Crawford, although in some ways an erratic witness, was adamant on that point. Furthermore, he scoffed at the suggestion that the porters, and those dealing with standard maintenance tasks on behalf of GMVL, would have set about the valves with large metal tools. He said that they simply would not have acted in such a way because they did not have the technical know-how and, perhaps more significantly, they did not possess these sorts of tools. I accept his evidence on that point.
170. It is right, and Robson's expert Mr Newbery gave a good deal of evidence on the topic, that after the flooding on 1 August 2007, and up to and including July 2012, a good deal of maintenance work was carried out at these buildings, and this work would appear to have included the use of tools on the plastic components (because numerous tool marks were seen), and other 'bodging' of various kinds. But this has little, if anything, to do with the question of when and how the damage to this particular nut was done. It is clear that, at some point after the flooding and the end of the DLP, a maintenance contractor would have come on board with a much more extensive workscope, and could well have done this wider damage then. But by the end of May 2010, GMVL had sold their interest in the site in any event.
171. In my view, it does not follow that, merely because there were tool marks on plant and equipment which were seen by Mr Newbery on inspections in 2009 and 2012, that such tools were used by/for GMVL prior to 1 August 2007: indeed, Mr Sofroniou expressly agreed that there was no evidence of the widespread use of such tools on the valves and other fittings when the wider inspection was carried out after the flooding on 1 August 2007. This was despite the fact that the inspection by Robson, on that occasion, had been very thorough, lasting three days. Mr Newbery gave similar evidence about the absence of such damage on his more limited inspections, also in 2007.
172. Still further, there were no signs of this IV having leaked before 1 August 2007. If a leak had occurred there, the evidence suggests that it would have been reported to Essex, who would have been required to repair it, free of charge. That would again point away from the likelihood that this damage was done by or for GMVL after the handover.
173. On the other hand, there was a good deal of evidence that the sort of metal tools, (such as Stilsons), that might have been used to tighten these nuts, were available to Robson's employees on site during the works. Mr Amundsen (who gave evidence on behalf of Essex and HSE) said clearly that "it would not surprise me, from my experience, for him [the installing plumber] to then give the valve an extra 'nip' or turn with a wrench or some other tool after hand-tightening it." And although Mr Byford of LOR said that such a tool should not have been used, he said that its actual use would always depend upon the individual concerned. The evidence was that the Stilsons were part of the typical equipment available to the Robson plumbers during the works on site.
174. I accept that many of the Robson employees who gave evidence, such as Mr Lawrence, said they did not use – and did not see other Robson employees use – such tools on the plastic nuts, and that they had been trained not to use such tools. But that does not mean that it did not happen occasionally, particularly if there was a nut that proved more difficult to tighten than it should have been. And although Mr Robson himself was adamant that such tools had not been used, his cross-examination showed that this assertion was not based on first-hand knowledge of these works on this site.
175. In all the circumstances, I consider it more likely than not that, although the Robson plumbers on site did not regularly use Stilsons to tighten the nuts, some of them were likely to have done so, particularly if there was a nut or other fitting that proved particularly difficult to tighten. It is no coincidence that the thread of this nut had been fouled by the metallic debris, which could easily have made it more difficult than others to tighten satisfactorily (see paragraph 163 below). In those circumstances, it was more likely than not that a Robson plumber succumbed to temptation and tightened up the nut in a way that was contrary to the manufacturer's instructions, using a Stilson or another large metal tool. Because the manufacturer's instructions said that the plastic nuts should only be hand-tightened, the use of the Stilson in this way was a breach of the sub-contract between Essex and LOR.
176. Although the experts could not say with any certainty when the over-tightening took place, there were some elements of their evidence which suggested that it took place at the time when the IV was installed. Dr Sanders said that there may well have been a link between the presence of the debris and the over-tightening, because the debris could well have made the nut harder to turn, and therefore may have prompted the use of the tools. Secondly, Essex's expert, Mr Davison, said that the displacement between the mating flanges (after the nut failed) showed that it might have been necessary to tighten the nut with the tools, because it would have been difficult to make a leak-tight joint by hand.
177. Accordingly, as to the factual dispute about who used the tool to over-tighten this nut and when, I conclude that, on the balance of probabilities, it was one of Robson's employees, during the carrying out of the original works. The over-tightening therefore occurred before practical completion/handover. It was unsurprising that the damage was not noted at practical completion/handover since, although the marks were visible on careful inspection, they would not necessarily have been apparent on the more limited inspection appropriate at that stage.
178. For these reasons, I find that the use of a tool to over-tighten this nut was a breach of the workmanship obligations which Essex had warranted to GMVL. They were therefore in breach of warranty. I note that Essex accept that, if the nut had been over-tightened prior to handover (which is what I have found), they were in breach of their warranty.
179. Summary
180. It accordingly follows that there were two interdependent causes of the failure in Core 3, and that each contributed to that failure. Both constituted breaches of their sub-contract on the part of Essex. Furthermore, if I was wrong to conclude that both the metallic debris and the over-tightening were competing causes, and that only one was a cause of the failure of the nut, that would not affect the result. That is because, in accordance with Heskell, Essex would still be liable for the flooding and the subsequent damage, even if they were only liable for one of the two operative causes. The same would also be the case if I was wrong to find that the over-tightening had occurred before practical completion: even if (contrary to my primary view) that had occurred after practical completion, the presence of the metallic debris was still a causative breach of contract. Again, therefore, in accordance with Heskell, even on this basis, Essex would be liable in law for the consequence of the flooding in Core 3.
181. The contribution proceedings do not require any apportionment between HL and Essex because no-one suggests that HL had any liability for the Core 3 flood. Accordingly, Essex are liable to GMVL for the full sum of £380,000.

8. THE CONSEQUENTIAL CLAIMS AGAINST HSE and ROBSON ON CORE 3

182. Essex pass on the workmanship allegations (metallic debris and over-tightening) to HSE pursuant to their sub-sub-contract. No point is taken as to the 'back-to-back' nature of HSE's obligations to Essex in respect of these workmanship allegations on which this case rests. Because these defects were breaches of the applicable British Standard and/or the relevant manufacturer's instructions, they render HSE liable to Essex under the sub-sub-contract. These same allegations are passed on by HSE to Robson. Again, the presence of debris and the over-tightening of the nut were a breach of the British Standard and/or the manufacturer's instructions, which meant that Robson were in breach of the terms of the sub-sub-sub-contract.

169. (a) Robson accepted that the over-tightening which I have found was a breach of contract. However, in connection with the debris, Robson did not accept liability, arguing that the training/instructions that they received from the manufacturer, Georg Fischer, related to the cleaning of pipe joints prior to gluing, rather than the cleaning of threads before the hand-tightening of nuts. However, that does not seem to me to offer Robson any sort of defence, given that the manufacturer's instructions about hand-tightening were available in writing, and given that precisely the same reasons for keeping the joints to be glued free from debris applied to the need to keep the thread of the nut similarly uncontaminated. As Mr Newbery, Robson's expert, confirmed, however a joint is being made, "you wouldn't want them [the particles] in there."

(b) In my view, this evidence (and similar views expressed by other experts) makes clear that the damage to the nut, and therefore the risk to the joint, was wholly foreseeable, an issue not raised by Essex or HSE, but raised by Robson. Furthermore, although Mr Ter Haar QC asserted, by reference to The Heron II [1969] 1 AC 350 at 385, that the type of damage which was foreseeable as a result of the debris would only occur in a small minority of cases, and so could not be regarded as arising in the usual course of things, I consider that that submission too was contrary to the evidence outlined above. It was foreseeable that foreign material in joints may cause all sorts of problems, including consequential over-tightening and failure. That is why, amongst other things, the British Standard prohibits such contamination.

183. Thus, it seems to me that the workmanship allegations in respect of Core 3 pass all the way down the contractual chain and rest with Robson. They were ultimately responsible for both the metallic debris and the over-tightened nut, the twin causes of the first flood in Core 3. They are therefore liable to indemnify HSE pursuant to the indemnity in the sub-sub-sub-contract (paragraph 57 above).

9. THE CAUSE OF THE CORE 2 FLOOD

184. The Relevant Law
185. The starting point for any consideration of causation in law is the "but for" test. This was stressed by Lord Nicholls in Kuwait Airways Corp v Iraqi Airways Co [2002] UKHL 19 at paragraph 72:

"This guideline principle is concerned to identify and exclude losses lacking a causal connection with the wrongful conduct. Expressed in its simplest form, the principle poses the question whether the plaintiff would have suffered the loss without ('but for') the defendant's wrongdoing. If he would not, the wrongful conduct was a cause of the loss. If the loss would have arisen even without the defendant's wrongdoing, normally it does not give rise to legal liability."

As Hamblen J put it more recently in Orient-Express Hotels Limited v Assucurazioni General SA [2010] EWHC 1186 (Comm) at paragraph 33:

"As a general rule the "but for" test is a necessary condition for establishing causation in fact."

186. It is important that the court does not depart from the 'but for' test without clear and proper reasoning. In Fairchild v Glenhaven Funeral Services Limited and Others [2002] UKHL 22, at paragraph 53, Lord Hoffmann warned against the tendency to deal with causation by appealing to commonsense "in order to avoid having to explain one's reasons. It suggests that causal requirements are a matter of incommunicable judicial instinct. I do not think that this is right." In the subsequent case of Sienkiewicz v Greif (UK) Limited [2011] UKSC 10, this was put in even starker terms by Lord Browne of Eaton-under-Haywood when at paragraph 186 he said, in relation to Fairchild:

"Although, therefore, mesothelioma claims must now be considered from the defendant's standpoint a lost cause, there is to my mind a lesson to be learned from losing it: the law tampers with the "but for" test of causation at its peril."

Lord Mance expressly agreed with that "lesson of caution". But it needs to be noted that, as Lord Phillips made plain in paragraph 1 of the judgment, both cases were governed by the 'special rule' applicable to mesothelioma cases.

187. If the court wants to depart from the 'but for' test, what replaces it? Some authorities talk about "common sense"; others talk about the need to make value judgments (see paragraph 70 of Lord Nicholls speech in Kuwait). In that case he went on to say:

"73…Even the sophisticated variants of the 'but for' test cannot be expected to set out a formula whose mechanical application will provide infallible threshold guidance on causal connection for every tort in every circumstance. In particular, the 'but for' test can be over-exclusionary.

74…In this type of case, involving multiple wrongdoers, the court may treat wrongful conduct as having sufficient causal connection with the loss for the purpose of attracting responsibility even though the simple 'but for' test is not satisfied. In so deciding the court is primarily making a value judgment on responsibility. In making this judgment the court will have regard to the purpose sought to be achieved by the relevant tort, as applied to the particular circumstances."

Similar references to the need for value judgments can be found in the judgment of Glidewell LJ in Galoo Limited and Others v Bright Grahame Murray and Others [1994] 1 WLR 1360, and in March v E and MH Stramare PTY Limited a decision of the High Court of Australia and reported at [1990-1991] 171 CLR 506.

188. One of the main deficiencies with the straightforward 'but for' test arises where there are two concurrent, independent causes of the loss. In Orient-Express, having referred to the 'but for' test, Hamblen J went on:

"However, there may be cases in which fairness and reasonableness require that it [the 'but for' test] should not be a necessary condition. This is most likely to be in the context of negligence or conversion claims, but I would accept that in principle it is not limited to tort or to particular torts. I would also accept that a case in which there are two concurrent independent causes of a loss, with the consequence that the application of the 'but for' test would mean that there is no cause of the loss, is potentially an example of a case in which fairness and reasonableness would require that the 'but for' test should not be a necessary condition of causation, particularly where two wrongdoers are involved."

This passage reflects paragraphs 72-74 of the speech of Lord Nicholls in Kuwait, where he expressly addresses the problems created by two wrongdoers or two separate causes of the loss. At paragraph 74 he said:

"The classic example is where two persons independently search for the source of a gas leak with the aid of lighted candles. According to the simple 'but for' test, neither would be liable for damage caused by the resultant explosion. In this type of case, involving multiple wrongdoers, the court may treat wrongful conduct as having sufficient causal connection with the loss for the purpose of attracting responsibility even though the simple 'but for' test is not satisfied."

189. I consider that this approach is also borne out by the text books. Thus, in the 20^th Edition of Clerk and Linsell on Torts, at paragraph 2-94, the editors note that: "Where there are two simultaneous, independent events, each of which would have been sufficient to cause the damage, the 'but for' test produces the patently absurd conclusion that neither was the cause. The only sensible solution here is to say that both caused the damage." A passage to similar effect can be found at paragraph 9-066 of the 9^th Edition Keating on Construction Contracts.
190. A distinction should be drawn between cases where there are two concurrent independent causes of the loss, dealt with above, and those cases where there are two co-operating causes, that is to say situations where two causes give rise to the loss but where each, on its own, would not have done so. The best-known example of that is Heskell v Continental Express (see paragraph 152 above). And some caution is necessary when referring, in this context, to cases in the construction field addressing the particular problem of delays where there are two competing causes, one of which is the contractor's fault and one which is the employer's fault. The usual answer, as Edwards-Stuart J pointed out in De Beers UK Limited v Atos Origin IT Services UK Limited [2010] EWHC 3276 (TCC), is that the contractor was entitled to an extension of time but not entitled to financial compensation. As the judge put it:

"…the contractor cannot recover damages for delay in circumstances where he would have suffered exactly the same loss as a result of causes within his control or for which he is contractually responsible."

But this decision turned on the relevant contractual mechanisms governing extensions of time and loss and expense. I do not regard it as providing any wider guidance on issues of causation, a view I note expressly shared by Mr Stansfield QC and Ms Sinclair QC.

191. With these principles in mind, I go on to address the relevant factual and evidential issues in respect of the Core 2 flooding.
192. The Relevant Questions
193. Each of the relevant experts, namely Dr Sanders, Mr Davison, Mr Clarke, Mr Parry and Mr Newbery agreed that the valve connecting the flange at Core 2 Level 6 had failed due to a sudden overload caused by over-pressure. They went on to conclude that the cause of that pressure was water hammer; which can occur when water is pumped rapidly up a riser into a vacuum caused by a previous loss of pump pressure when the water suddenly stops at the top of the riser. The sudden stopping of the rapidly-rising water column produces a high pressure spike when the kinetic energy of the fast-moving water is transferred to pressure energy.
194. On that basis, one relevant question in this case is: What caused the vacuum which created the water hammer? But, on the facts, I consider that that is overlain by a second question. During the progress of the works on site, HL expressly recognised that, unless something was done to prevent a vacuum being created at the top of the riser, a water hammer event might occur. Indeed, HL had had direct experience of just such an occurrence. Accordingly, they recommended a surge arrestor in order to prevent a vacuum occurring at the top of the riser. A surge arrestor was fitted, but it did not prevent the creation of the vacuum.
195. Thus, in the present case, there are two related questions. What caused the vacuum and why did the surge protector not prevent the creation of that vacuum?
196. The Relevant Findings

(a) The IV

197. The experts started, logically enough, at the IV, because that was below both the surge arrestor and the NRV, and therefore closest to the water coming back up the riser. If, as I have found at paragraphs 119-122 above, the IV was in a closed position at the time of the events on 1 August 2007, then the experts agree that it caused or contributed both to the occurrence of the vacuum at the top of the riser when the water level fell, and the failure of the surge arrestor. I accept that conclusion.

(b) The NRV

198. The position in relation to the NRV was originally more complicated although, by the end of the trial, I considered that those complications had fallen away. It was agreed that the NRV prevented air from passing from the surge arrestor down into the riser; indeed, it seems very likely that the surge arrestor and the length of the riser immediately below it were filled with water at the time of testing/commissioning; that water could not drain away because of the NRV, which stopped it from flowing downwards and back into the riser. Thus the mere presence of the NRV on the riser below the surge arrestor meant that the surge arrestor could never have done the job for which it was designed, and could never have prevented the creation of the vacuum at the top of the riser when the water level fell. On that basis, it also caused the presence of the vacuum and the failure of the surge arrestor.
199. The original complication arose because, in the experts' first Joint Statement at paragraph 2.8, they said:

"The NRV orientated to allow only upward flow would also have prevented air from entering the riser but with the IV closed there would have been no means of drawing, or allowing, air in to the riser anyway. In summary, with the IV closed, the NRV had no influence on the Core 2 leak."

However, beginning with Mr Hickey's cross-examination of Dr Sanders, it became apparent that this was an inaccurate summation of the position. True it is that, with the IV closed, there was another cause of the vacuum, but it was not right to say that the NRV had no influence on the Core 2 leak. On the contrary, the NRV also prevented the surge arrestor from working and also caused the presence of the vacuum. Indeed, on one view, the NRV was more important than the IV because the NRV permanently disabled the surge arrestor, whilst the IV only did so if/when it was closed; if the IV was open, it was causally irrelevant. There was considerable expert evidence in support of this conclusion, including these exchanges:

(a) Mr Hickey's cross-examination of Dr Sanders:

"Q: The reason why the air can't get into the riser is because of the presence of the NRV which you are asked to assume was present?

A: Yes that is right.

Q: That's why you say, in your scenario, whether the IV is opened, it would have happened anyway, and the principal reason for that, the dominant reason, is because the NRV prevents any air getting in to provide the cushion effect, am I right?

A: Yes that is right.

…

Q: A surge arrestor has been provided and it needs to do its job. Its job is to avoid water hammer, or at least deal with it, and it is prevented from doing its job, not because the key is being switched on or off, but because there is a non return valve there and that is preventing it from letting air in. It is effectively as if you have built a riser without surge protection, am I right?

A: I think I see what you are saying. I don't know that it is really for me to say which is the predominant or the main cause. The fact is that either or both of these features would have prevented the anti-surge valve from working.

Q: Well the IV becomes much more important if the NRV was not there, because then it really would matter whether it was closed or open?

A: Yes."

(b) Mr Hickey's cross-examination of Mr Gosling:

"Q: So what is it that a locked open IV would have done to avoid water hammer?

A: With the presence of the NRV there, and the isolating valve open, then it is likely that the floods would still have occurred…

Q: So we keep coming back to the central point: the thing that stops the water hammer is the surge arrestor, if it is allowed to do its job?

A: Yes.

Q: And the thing that predominantly, if not exclusively, stops the surge arrestor from doing its actual job is the NRV that should never have been installed?

A: Yes.

Q: You agree with that?

A: Yes."

(c) Mr Stansfield QC's cross-examination of Mr Marshall:

"Q: As you say in paragraph 2.9(a), the non-return valve would have prevented the anti-surge valve from allowing air into the riser?

A: Yes.

Q: So it would have prevented the anti-surge valve from doing what it was supposed to do?

A: Yes.

Q: In those circumstances, at paragraph 2.9(b), you agree that the water hammer that would have occurred would probably have caused a similar core 2 leak?

A: That is correct."

Subsequently Mr Marshall went on to say expressly that, even if the IV had been open, the NRV would have done the damage instead, and that such damage would have been very similar to (or exactly the same as) that which occurred.

(d) Mr Stansfield QC's cross-examination of Mr Parry:

"Q: With the NRV present below the surge arrestor, would you agree that air cannot enter the riser through the surge arrestor regardless of whether the isolating valve was open or closed?

A: I would agree with that…

Q: As far as the leak is concerned, the leak is going to occur when the riser fills, isn't it?

A: Yes. As a result of the fast filling of the riser and the sudden abrupt halting of the water column.

Q: So would you agree that it is not occurring because the IV is there but occurring because the riser has filled?

A: Yes, that is correct.

Q: When the isolating valve is closed, there are two blocks on the riser, aren't there?

A: Yes.

Q: So far as the air entering the riser is concerned, the first block is the non return valve?

A: That is correct.

Q: You would confirm that as a technical matter?

A: Indeed."

(e) Mr Hickey's cross-examination of Mr Newbery, in which (although Mr Newbery argued that there was no demand for air because the IV was closed), he also agreed that there were two blocks on the line and that the first physical block to air getting in was the NRV. Mr Newbery also agreed that, in allowing air to come in, the first step in that process may have been to open the IV, but that would not have made any difference, because of the presence of the NRV.

184. (a) Finally on the NRV, I accept Mr Stansfield QC's proposition that, since the experts' conclusion meant that the damage would have been the same whether the IV was open or closed, 'a degree of suspicion' (as he put it) must attach to the proposition that the closed IV was the sole effective cause of the flooding. To put the point a slightly different way, there was no leak at Level 8, where the relevant NRV and the IV were located; the Level 6 IV which failed would have failed whether the Level 8 IV was open or closed.

(b) It seems to me that, for all these reasons, it would be idle to suggest that the NRV was anything other than an independent and effective cause of the flooding. This is particularly so given that the experts also generally agreed that, if the IV had been open at the time of the flooding then, not only would the NRV have been the sole cause of the flooding but also that, on the balance of probabilities, the same flooding and the same damage would have occurred on 1 August 2007.

200. The 'But For' Test
201. Essex and Robson argued that, on an application of the 'but for' test, it was the closed IV which was the sole cause of the flooding. They relied on the experts' original agreement to that effect and the fact that, when the water level in the riser fell away, the vacuum would have been created from the closed IV downwards. They also relied on the fact that the water surge reached the closed IV and then was unable to go any further towards the NRV.
202. The fundamental difficulty with this argument is that the position of most of the experts changed during the trial, when greater attention was paid to the NRV. The facts and matters set out at paragraphs 182-184 above make plain that the NRV was causally relevant. The mere fact that it may have been the closed IV which was the point that met the water surge^[5] is immaterial from a causation point of view, just as it is immaterial that the actual flooding occurred two levels below: that was merely the weakest point in the system. Where the surge reached was just one aspect of the mechanism of failure.
203. There is a further difficulty in respect of the closed IV. In my judgment, the closed IV was in the wrong place. All of the technical literature demonstrated that an IV was only necessary at all to allow maintenance and cleaning of the surge arrestor. For that reason, the literature indicated that the IV should be installed immediately below the surge arrestor. Here, the IV was at least one metre below the surge arrestor. This odd positioning was wholly unjustified. If the IV had been put in the right place, i.e. immediately below the surge arrestor, then the NRV would have been below rather than above it. Thus, on this scenario, and adopting Essex and Robson's case, the NRV would have been the first point that the surge hit and the only cause of the damage. It cannot be right that this mis-location of the IV can be relied on by Essex and Robson, who were both contractually responsible for that mis-location, in order to escape liability for the flooding.
204. By contrast, HL and HSE both contend that it was the NRV which met the appropriate 'but for' test. This submission was put most forcefully in HSE's written closing submissions at paragraphs 138-142. In essence, Mr Hargreaves QC contended that the surge arrestor was disabled when it was installed above the NRV, so that the contingent position of the IV was always irrelevant.
205. I accept that the NRV disabled the surge arrestor, and could be said to be the most proximate cause of that disablement (since the NRV was closest to the surge arrestor). I also agree that the NRV permanently disabled the surge arrestor, whilst the IV did not have any effect on it at all, unless it was closed. But beyond that, I consider that any physical or temporal distinction between the NRV and IV is a little unrealistic. Although it is not entirely clear which part of the installation came first (NRV or surge arrestor and IV), I think it likely that the NRV was installed before or at the same time as the surge arrestor and IV. Furthermore, it does not seem to me to be appropriate to ignore the IV altogether, in circumstances where the vacuum would have started from the IV (and probably not the NRV).
206. On behalf of GMVL, Mr Stansfield QC relied on the other authorities referred to in paragraphs 173-175 above to argue that both the closed IV and the NRV were causes of equal efficacy. He maintained that this is one of those rare cases where the 'but for' test would lead to the patently absurd conclusion that neither was the cause, and that instead the court should conclude that both the IV and NRV were equal causes.
207. As part of HSE's alternative submissions, Mr Hargreaves QC also argued the same thing: that the NRV and the closed IV were causes of equal efficacy. In the course of his written closing submissions at paragraph 150, Mr Hargreaves QC makes what I consider to be a most telling point on this aspect of the case: that it is rationally difficult to approach causation in this case on the basis of "whether one looks 'up the pipe' or 'down the pipe', with a commonsense outcome that both must be equal and independent causes."
208. It seems to me that, on this central issue of causation, Mr Stansfield QC's primary submission (and Mr Hargreaves QC's, Ms Sinclair QC's and Mr Hickey's alternative submission) is to be preferred. On any sensible analysis of what happened here, there were two equally efficacious causes of the flooding. There were two obstructions which caused the vacuum and prevented the surge arrestor from working. As a matter of commonsense, it simply cannot matter that it was the closed IV which might be said to have been the proximate cause of the vacuum, and the NRV which might equally be said to have been the proximate cause of the disabling of the surge arrestor. They both caused the vacuum; they both caused the surge arrestor not to work. They were therefore equally efficacious causes of the flooding in Core 2.
209. Accordingly, based on this analysis, I depart from the 'but for' test, but only to adopt the well-known alternative approach set out in the authorities noted in paragraphs 173-175 above. I conclude that the liability issues must be resolved on the basis that there were two equally efficacious causes of the flooding: the closed IV and the presence of the NRV.

10. GMVL'S CLAIMS AGAINST HL ON CORE 2

210. 1 Did HL Warrant to GMVL the Performance of Any Relevant Design Obligations?
211. For the reasons set out in Section 3.3 above, I conclude that HL warranted to GMVL the performance of their obligations in respect of the design, specification and installation of the surge arrestors.
212. 2 Did HL Owe a Duty of Care to GMVL in respect of Any Relevant Design Obligations?
213. In the light of my conclusions as to the warranty, it is unnecessary to consider whether or not HL owed a separate duty of care at common law in relation to the design, specification and installation of the surge arrestors. However, for the avoidance of doubt, I consider that they did. This seems to me to be a classic case of an assumption of responsibility by a professional consultant. It was HL who identified the need for surge arrestors; it was HL who recommended their provision on site. It was therefore HL who had to design, specify and see to the proper installation of the surge arrestor system.
214. In assuming responsibility for undertaking those obligations, HL owed a duty of care to GMVL, in accordance with general common law principles.
215. 3 Were HL in Breach of Warranty and/or Duty in respect of Design/Inspection?
216. 3.1 The Relevant Factual Findings
217. At paragraph 99 above I have concluded that the NRV was installed by Robson, although the precise time when it was installed is unclear. I have found that the NRV had no business being there at all and prevented the surge arrestor from working.
218. At paragraphs 103-110 above, I have concluded that, on the balance of probabilities, the IV was closed on 17 August 2006, at the time of the handover to GMVL. At paragraphs 119-122 above I have concluded that the IV was closed at the time of the flooding on 1 August 2007. In addition, in those same paragraphs, I have concluded that there is no evidence that the IV was opened at any time between 17 August 2006 and 1 August 2007, and I therefore find that the IV remained closed throughout the relevant period. No-one noticed that the IV was closed because the only thing that the closed IV did was to prevent the surge arrestor from operating; since there was no evidence of any surge that needed to be prevented until 1 August 2007, the closure of the IV was both missable and missed. The rest of this analysis is based on these findings of fact.
219. 3.2 The Design Claim against HL: Analysis
220. Although the presence of the NRV and the closure of the IV were primarily matters of workmanship (and therefore the claim against HL was triggered principally by their failure to perform their inspection obligation properly), I consider that the claim in respect of the allegedly deficient inspection has to be considered in the context of HL's overall design role and in particular, their knowledge of the surge arrestor system, which in this case was not shared by any other party.
221. HL were the original designers of the mechanical system. In that role, HL had belatedly (and half-heartedly) identified the need for surge arrestors on this project. Put another way, HL foresaw what might happen on 1 August 2007 and took at least some steps to prevent it; they recommended the use of surge arrestors to protect the buildings from flooding due to water hammer. Just like the architects in Sahib Foods Ltd v Paskin Kyriakides Sands (a firm) [2003] PNLR 181, they had a particular culpability because "they knew the risk". The problem was that the steps that they took to deal with the riser were inadequate, which explains why, even though surge arrestors were installed, severe flooding still took place.
222. Their general performance of their design obligations was deficient for a number of reasons. First, they never 'took control' of the underlying design issue. Although Mr Collins sent out helpful generic documentation (paragraphs 70, 71 and 73 above) which made plain what the HL engineers were to do on any given site, Mr Hart and others employed by HL on this project were very reluctant to do anything about it at all. It was for that reason that the need for surge arrestors was not addressed on this site until November 2005, and even then it only arose because of the concerns of Mr Prosser and Mr Delvin on site, not those – like Mr Hart – in the HL design department in Poole.
223. Furthermore, HL's way of dealing with the whole issue had something of the shrugged shoulder about it; it was almost as if, although they knew that Mr Collins and the HL head office were keen on surge arrestors, those involved on this project did not really mind one way or another whether they were fitted on these risers. They never sent LOR or Essex any sort of specification, even though they had their own internal specification which contained a good deal of useful information. I consider that was an inappropriate attitude towards something which they knew (because Mr Collins was telling them so) was a very important element of the BMCWS.
224. The failure to take control of this aspect of the design, and to follow it properly through to a conclusion, can be seen in many of the subsequent events. The HL team working on this project never produced a design drawing which showed surge arrestors; instead, long after Mr Collins was saying that surge arrestors were required, they were producing drawings which simply showed air vents at the top of the risers. They never specified what sort of surge arrestors should be installed, even though Mr Collins had carried out research on that very topic, and produced a generic specification.
225. Most important of all, they did not specify that an IV was required below the surge arrestor and, despite Mr Collins' clear identification that such IVs should be a lock-shield valve (i.e. one that was difficult to operate), those at HL who were involved in this project did not ask themselves why Mr Collins was recommending such a valve, and they did not pass on his recommendation. In my view, this was an important failure. The IV was necessary for maintenance purposes, but since it would, if closed, wholly negate the purpose of the surge arrestor, I consider that Mr Collins was right to recommend a lock-shield valve, and that the HL engineers (principally Mr Hart) involved in this project were wrong not to insist on such a valve. Mr Hart accepted in cross-examination that he did not even consider the point, and he agreed that the suggestion that he ought to have done so "was not an unreasonable statement". The valve that was put in was easy to tamper with, thus risking the negation of the surge arrestor entirely.
226. The M and E experts, with the exception of Dr Humphries, agreed that HL should have specified a tamper-proof IV because of their own particular knowledge of the risks. However, in cross-examination, even Dr Humphries agreed that HL ought to have considered the type of IV (which they did not) and ought to have considered the consequences of the failure of the surge arrestor to operate (which they also did not). It appears that Dr Humphries' view was based on the proposition that the risk of flooding damage was not sufficient to require a tamper-proof IV but, on that issue, I prefer the evidence of all the other experts, who considered that it was.
227. In addition, Dr Humphries appeared to suggest that a tamper-proof IV was not necessary because the riser cupboard in which it was located would be locked. Again, all the other experts said that this was irrelevant, given that many people had access to the cupboard, and that therefore the isolating valve was at risk of being operated in an unauthorised manner. Although Dr Humphries disagreed with that, in cross-examination he agreed that any decision (as to the type of IV) that was made without checking that the riser cupboards were lockable would again be negligent. There is nothing to say that HL did check that the riser cupboards would be lockable. Accordingly, Dr Humphries' final position was much closer to that of all the other experts and, in the round, I am in no doubt, based on the expert evidence as a whole that the only proper conclusion is that HL were negligent in not specifying a lock-shield IV.
228. In addition, notices warning about the significance of the closed IV were not provided until after the flooding on 1 August 2007, which was of course too late. I consider that such notices were important because they would have spelt out in a clear way how and why the IV needed to be kept open, except during maintenance. That was particularly important here, when the IV was over a metre below the surge arrestor and would not necessarily look as if it had anything to do with it. There was no explanation in the evidence as to why these notices were not provided before the flooding.
229. Still further, HL did not design, specify or warn that the British Standard required that all these valves be labelled. None of them were ever labelled, so the significance of the IV (and its relationship with the surge arrestor) was not apparent to anybody looking in the riser cupboard, an omission again exacerbated by the sheer distance between the surge arrestor and the IV (with a redundant NRV between them for good measure).
230. I consider that these were all important design/specification failures on the part of HL. Given that surge arrestors were a very new piece of kit, the widespread lack of familiarity with their operation amongst all the contractors who gave evidence before me was understandable, but this imposed an even greater burden on HL (who were familiar with them). HL had to ensure that there was a proper design and specification; that the relevant parts were labelled; the consequences of keeping the valve closed were spelt out; and that the opening of the IV was made as difficult as possible, making a lock-shield valve vital.
231. For all these reasons, therefore, it seems to me that the fact that the IV was closed both at practical completion, and at the time of the flooding, was not exclusively a workmanship matter. It was also a matter of design and specification. This was a new design feature introduced by HL and they needed to 'see to it' (their preferred phrase) that its operation was properly understood before it was installed. They wholly failed to do so.
232. 3.3 The Inspection Claim Against HL: Analysis
233. It follows from what I have said in Section 10.3.2 above that, as the party responsible for the specification of surge arrestors in the first place, HL had a clear obligation to see to it that the surge arrestors were properly installed. They would see to this by carrying out proper inspections. Further, they did actually carry out the relevant inspections: indeed, on 17 August, they were the only party inspecting. Subsequently, they produced not one but two certificates, with different words, but each essentially saying that the works had been completed satisfactorily (see paragraph 113 above). Since the IV was closed so that the surge arrestor would not work, and since there was also an unheralded and otiose NRV which negated the function of the surge arrestor in a similar way, these statements of satisfaction were palpably incorrect.
234. In my view, there were two specific failures of inspection. The first was the failure to spot that the IV was in the closed position (see paragraphs 103-109 above). Indeed, Mr Gilbert took a photograph of it in that position, which I have already said was the best evidence of his failure to realise what the IV was and what it was doing. Mr Gilbert may well have been confused by the sheer distance between the surge arrestor and the IV: since he was not familiar with the surge arrestor system, he may not have thought that the IV had any role to play in its operation at all.
235. The experts were generally of the view that, if the IV was closed, the defect should have been spotted and rectified. Dr Humphries, the relevant HL expert, agreed that it was important that the IV was open when the building was handed over to GMVL, and that the IV should have been open in normal circumstances. Although he endeavoured to say that the closed IV may not necessarily have been noted as a defect, he agreed that the IV was not just any IV, because it was situated below the surge arrestor. He also agreed that the inspecting engineer had to pay particular attention to the safety systems of the building in his inspection. The surge arrestor was one such system. In the light of that evidence, and the evidence from the other experts which indicated that HL should have identified the closed IV, I am confirmed in my conclusion that HL were in breach of warranty and/or negligent in failing to spot that the IV was in the closed position.
236. But there was a second failure. Let us assume that Mr Gilbert had spotted the IV, spotted that it was closed, and realised that it needed to be open. He could have opened it: that would have dealt with that defect. But of course it would not have dealt with the fact that the surge arrestor was still disabled, because of the presence of the NRV. Why did Mr Gilbert not see the NRV, particularly since all the experts, with the exception of Dr Humphries, thought he should have done?
237. That was a question which the parties spent sometime puzzling over during the trial. The best that HL could do was to suggest that Mr Gilbert might have thought that the NRV was a union coupling. However, it seems to me that that argument fails at every hurdle. First, Mr Gilbert gave no such explanation in either his written or his oral evidence. Secondly, I find that nobody undertaking any sort of proper supervision or inspection of this riser cupboard could possibly have mistaken the NRV for a union coupling. Even though the top part of the NRV was covered in lagging, the amount of the valve that was visible below the lagging made plain that it was much bigger then a union coupling and had two circular screws (like a valve) and not a central circular screw (like a coupling). It also had two arrows marked on the side showing the direction of flow which were not particularly difficult to see, even if they were towards the wall. Dr Sanders said the arrows were visible when he first saw the NRV on site after the flooding. These arrows also demonstrated that it was a valve not a coupling.
238. Thirdly, as Dr Humphries agreed, the surge arrestors were an important part of the safety systems and it was important that they remained operational. Dr Humphries agreed that a reasonably competent inspecting engineer should pay particular attention to the safety systems, including the surge arrestors, and that a careful check was required.
239. Fourthly, Dr Humphries agreed that when he himself went to site, he could see that the arrows marked on the valve indicated beyond doubt that it was an NRV. There was then this exchange:

"Q: So anyone who had given it a moment's thought and said 'what's that? I had better look at it' would have appreciated that this is a non-return valve?

A: Absolutely. If you took the trouble to go up close to it and look at it you would know that it was a non-return valve, yes."

240. On the basis of that evidence, as well as the majority view of the other experts, it seems to me plain that a reasonably competent building services engineer should have identified the NRV. That view is also supported by the absence of labels, because if the valves had been properly labelled, it would have been expressly stated that that was an NRV and Mr Gilbert may then have challenged it.
241. For all these reasons, I conclude that there was no justification for Mr Gilbert's failure to spot the NRV. Had he done so, he would have realised that it simply should not have been there at all and, even if he did not know that it was preventing the surge arrestor from working, he would have ordered its removal.
242. It would be unfair and unnecessary to dwell for too long on the reasons for the flawed nature of Mr Gilbert's inspection, and his failure to spot the closed IV or the NRV. However, it seems to me that paragraph 44(b) of Essex's written closing submissions contains, in clear summary form, a number of reasons why Mr Gilbert carried out what Ms Sinclair QC fairly described as a "plainly incompetent" inspection. He had not attended site before; he did not familiarise himself with the British Standard or any drawings or method statements; he had no prior experience of surge arrestors or plastic pipework installations; and he had carried out only one inspection of riser pipework in the previous five years. For all those reasons, it seems to me that the failings on this inspection were entirely explicable.
243. For these reasons, therefore, I conclude that GMVL's inspection claim has also been made out against HL.
244. 4 Summary
245. I find that HL were liable to GMVL both under the warranty and, if relevant, in tort. This liability is in respect of design (generally, because of their failure to take charge of their own recommendation, and specifically, in relation to the failure to specify a lock-shield IV and labels). It is also in respect of inspection, both in respect of the IV and the NRV, for the reasons noted above. Whilst I accept that the inspection failures were more directly relevant to the cause of the flooding (because what mattered most was the failure to order the removal of the NRV and the failure to order the opening of the IV), I do not consider that it is sufficient to say that the case against HL is in relation to inspection only. I conclude that both aspects of the case against HL flow from HL's fundamental failure to take a grip of their own recommendation that surge arrestors be installed at this site.

11. GMVL'S CLAIMS AGAINST ESSEX

246. 1 The Relevant Factual Findings
247. The relevant factual findings summarised above at paragraphs 197-198 are repeated. Essex's liability to GMVL is analysed by reference to them.
248. 2 The Design Claim
249. I have found (see Sections 3.4 and 3.5 above) that Essex owed relevant obligations in relation to design co-ordination and development. Were they in breach of those obligations?
250. Pursuant to those obligations, Essex raised RFIs in respect of the mechanical works, and they produced drawings which purported to show the detail of the mechanical works to be installed. On that basis, they should have become fully involved in HL's belated requirement that surge arrestors be fitted, and ought to have produced a drawing showing those surge arrestors and the detailed installation that was required.
251. Essex's culpability in relation to design is compounded by two internal documents. The first, (dealt with at paragraphs 67-69 above) was the RFI, and Essex/HSE's subsequent claim that they raised this because they were concerned that there was a risk of water hammer. That means that, on Essex/HSE's own case, they had the foresight to realise that the design – for which they had a responsibility – needed to prevent a water hammer event, but lacked the organisation or management to raise that point expressly with LOR, HL or GMVL.
252. Secondly, there was the email analysed at paragraph 89 above. In that email, Essex/HSE talks blithely of their underlying concerns about HL's design of the BMCWS. Yet at no time did they identify those concerns up the contractual chain to LOR or GMVL. Again, therefore, in my view this compounds rather then lessens their design coordination failures.
253. I consider that Essex are liable to GMVL for the same design/specification failures as HL, analysed in the preceding section of this Judgment. In some ways, the general nature of their default in respect of design can be pinpointed by reference to a third internal Essex document: the calamitous record drawing supposedly completed just two days before the inspection on 17 August (see paragraph 111 above).
254. The evidence demonstrated the slap-dash way in which this drawing had been put together. It was said to be produced from information supplied by Mr Smith to Jeremy Dove of Crofton Design. However the nature and scope of this information was unclear. There was no record of it. The drawing indicated that Mr Smith had checked the accuracy of the drawing, but he told me in cross-examination that he did no such thing, and had not checked the drawing against that which had actually been installed. I consider that Mr Stansfield QC was right at paragraph 76 of his written closing submissions to describe this as a "haphazard approach to the development of the design." It certainly explains why, in other parts of the development which were not directly affected by the flooding, there was a wholly inconsistent approach to the design and installation of the BMCWS. Amongst other things, there were surge arrestors without IVs and, in Core 4, an IV for a surge arrestor which was located in such a position that it would cut off the water supply to the relevant flats. Such careless work was compounded by the absence of proper design and record drawings.
255. The specific allegations in relation to design concern the lack of a tamper-proof IV, the lack of labels and the absence of a method statement. As to the tamper-proof IV, I have already set out the background in paragraphs 204-206 above in which I conclude that HL were liable for failing to specify a tamper-proof IV.
256. As to Essex's responsibility for this same omission, it is right to note at the outset that it was Essex who chose the particular surge arrestor and IV which were used. Once they had received LOR's instruction, they could either have asked for further details of the equipment to be installed, or they could have taken on the responsibility to choose it themselves. They chose the latter course, but failed to give any real thought to the decision. As the cross-examination of Mr Smith made plain, the matter was never even considered by Essex:

"Q: Well, under the circumstances where the instruction doesn't specify every item in the design, and Essex decides to go ahead anyway, making its own decision regarding the design, in those circumstances you would expect the remaining design to be carried out by your design office, wouldn't you?

A: Possibly; nobody asked for any information on the details of the surge arrestors.

Q: You are anticipating my next question, but do you agree with me, first of all, that in the circumstances I have just described, you would expect the remaining design to be carried out by the design department?

A: Yes.

Q: But in this particular instance nobody contacted you or your department to ask for further details or determination of the design?

A: No."

257. That Essex were in breach of contract in this regard was confirmed by two other strands of evidence. First the expert evidence: all the experts (apart from Dr Humphries called by HL), agreed that a means of rendering the isolation valve tamper-proof should have been provided. In cross-examination, Mr Marshall, Essex's M and E engineering expert, agreed that a tamper-proof valve and/or a warning notice was necessary for the safe and satisfactory operation of the building. Secondly, Mr Prowse, one of HSE's site supervisors, gave unchallenged evidence that he had specifically asked other employees of HSE why a lockable valve was not being used. Mr Prowse said that he raised this because he thought that a non-lockable valve could easily be tampered with and would not be best practice. He confirmed that he had suggested a lock-shield valve because he was aware that the surge arrestors played an important part of the safety systems in the building.
258. Accordingly, on the basis of this compelling evidence, I conclude that Essex were in breach of warranty for failing to provide a tamper-proof IV.
259. Furthermore, despite the points made by Ms Sinclair QC in paragraphs 58-67 of Essex's closing written submissions, there can be no doubt that, as a matter of fact, Essex chose the relevant equipment and are therefore responsible for the consequences of that decision to GMVL. That was part of their design co-ordination/development responsibility and would render them liable, even if (which I do not accept, for the reasons noted in Sections 3.4 and 3.5 above) Essex's contractual obligations under the warranty were somehow different to those eventually agreed with LOR.
260. Finally on this topic, whilst it is certainly right that not every manufacturer recommended a tamper-proof IV in every case, the particular circumstances here – the accessibility of the riser cupboard, the novelty of the surge arrestor system, HL's awareness of the importance of that system to protect the building – meant that, if proper research had been done by Essex, a lock-shield valve was likely to have been recommended by any competent manufacturer. The problem was Essex's failure to conduct that investigation in the first place.
261. The second design allegation was the failure to label the valves. The British Standard makes plain that that was a clear requirement and Mr Amundsen, the HSE Project Manager, told me that valves should be labelled as a matter of course. He was unable to supply any explanation as to why that was not done here. And whilst Ms Sinclair QC argued that the BS suggested that the labels could have been shown on a diagrammatic drawing in the plant room in the basement, instead of on the valves themselves, that was an academic point because that did not happen either. Whether this was an allegation of design or workmanship, it seems to me that it was a clear breach of contract/warranty by Essex. Contrary to Ms Sinclair QC's submissions, I also consider this to be an important failure because if, (for example,) the NRV had had a label, it would have become immediately apparent that it was playing no useful purpose in the riser and was detrimental to the proper operation of the surge arrestor.
262. Finally, there was the absence of a method statement for the commissioning. Again that is either a matter of design or workmanship, but either way it was a further failure on the part of Essex. I am not, however, persuaded that any specific loss or detriment flows from that failure. The absence of a method statement was simply further evidence of the slap-dash way in which Essex/HSE approached this whole project.
263. 3 The Workmanship Claim
264. Essex originally denied that the NRV was present in the riser at practical completion in 2006. Of course, the disclosure of Mr Gilbert's photograph demonstrated beyond doubt that it was, and Essex amended their position accordingly. Now, Essex not only accepts the presence of the NRV at practical completion but also admits that its presence was a breach of sub-contract on their part. Essex were therefore in breach of the terms of their warranty with GMVL for installing and/or failing to remove the NRV.
265. Essex also accepts that if (as I have done) I find that the IV was closed at practical completion, then that too was a breach of their sub-contract and their warranty to GMVL.
266. 4 Summary of Claim Against Essex
267. Accordingly, Essex were liable to GMVL in relation to both design and workmanship pursuant to the terms of the sub-contract, and thus the warranty.

12. APPORTIONMENT BETWEEN HL and ESSEX (CORE 2 ONLY)

268. On Core 2, in consequence of my findings above, HL and Essex are jointly and severally liable to GMVL for 92% of £4.75 million (£4,370,000) together with interest. Because there are contribution proceedings between those two defendants, it is necessary for me to apportion liability between them pursuant to the 1978 Act.
269. In undertaking that task, I start with the proposition that the principal deficiencies in the present case were matters of workmanship, namely the installation of an otiose NRV and the failure to ensure that the IV was in the open position on handover. It is then necessary to identify the proper approach to apportionment between a contractor responsible for the bad work, and a professional man who failed to spot it and have it rectified. In his judgment in Carillion JM Limited v PHI Group Limited [2011] EWHC 1379, at paragraphs 247-258 Akenhead J analysed in detail the cases on this topic, and concluded that this apportionment will usually be in the range of 80% (contractor) and 20% (professional), up to 66% (contractor) and 33% (professional).
270. In the course of his closing submissions on behalf of HL, Mr Hickey submitted that Essex should shoulder the entirety, or the lion's share, of the apportionment because:

(a) They installed the NRV without justifiable reason;

(b) They knew that the IV should be left open at all times;

(c) They selected the type of IV and installed them.

271. In response, in her written closing submissions on behalf of Essex at paragraphs 94-107, Ms Sinclair QC maintained that, because of HL's close involvement in the recommendation of the surge arrestor system, and the widespread nature of their failings, (both design and inspection), HL's contribution in relation to the NRV should be at least 33%, whilst their contribution in respect of the closed IV and the failure to specify tamper-proof IVs should be 50% or more.
272. I broadly accept Ms Sinclair QC's submissions. As I have endeavoured to point out at paragraphs 199-203 above, I consider that HL carry a significant responsibility in this case because it was they who recommended surge arrestors in the first place and it was they who failed to ensure that the surge arrestors were given the best possible chance of working. Those failings were principally matters of inspection but also included matters of design.
273. Accordingly, I consider that the correct overall split is 60% (Essex) and 40% (HL). In my judgment that percentage split properly acknowledges HL's important role as an inspector which, because the surge arrestors were their recommendation in the first place, was always going to be at the upper end of the range identified by Akenhead J in Carillion. It also reflects their overall responsibility for the proper design and specification of equipment which they had recommended. They were, after all, the only one of these five parties who had had any experience of surge arrestors before. In the round, however, Essex's total share was always going to be greater than HL's because the crucial failures were matters of workmanship, for which Essex (as contractor) would always bear a greater responsibility than HL.
274. HL's warranty with GMVL included an apportionment provision. However, it seems plain that this provision does not operate because it envisaged apportionment between HL and other potentially liable consultants. In this case, no other such consultants have been identified. Moreover, for the reasons that I have given, even if they had been, a 40% overall liability seems to me to be a reasonable result in respect of HL's failings on Core 2.

13. THE CONSEQUENTIAL CLAIMS AGAINST HSE AND ROBSON ON CORE 2.

275. Essex's Claim Against HSE
276. Essex pass on the design allegations to HSE pursuant to their sub-sub-contract. For the reasons set out in Section 3.6 above, I consider that, to the extent that the Core 2 flooding rendered Essex in breach of their design coordination/development obligations to LOR, and therefore under the warranty, HSE were in breach of their 'back-to-back' design obligations to Essex. Those allegations must therefore rest with HSE: they cannot be passed on to Robson because Robson had no design liability.
277. However, on any view, the principal cause of the subsequent flooding was the failure to close the IV and the installation of the redundant NRV. These are both failures of workmanship. In consequence, Essex pass them on to HSE pursuant to their sub-sub-contract, and no point arises as to the 'back-to-back' nature of HSE's obligations to Essex in respect of workmanship. HSE accept that they are liable to Essex in respect of both these workmanship allegations. The remaining issue is whether HSE can pass these allegations on to Robson.
278. HSE's Claims Against Robson
279. The NRV
280. I have found as a fact that the NRV was installed by Robson. There can be no doubt that this constituted a breach of Robson's contract with HSE. I have also found that the NRV was an effective cause of the flooding. On the face of it, therefore, Robson are therefore liable to indemnify HSE for this breach pursuant to the contractual indemnity referred to at paragraph 57 above. However, this aspect of HSE's claim gave rise to a substantial dispute with Robson.
281. It was HSE's case that, regardless of the position in respect of the IV, because the NRV was one of the two independent effective causes of the flooding, Robson were liable to indemnify HSE pursuant to the contractual indemnity. In reality, this would mean that Robson were obliged to indemnify HSE for all the losses passed down the line by Essex to HSE: on this basis, HSE said that Robson should provide a 100% indemnity.
282. In support of that submission, Mr Hargreaves QC relied on ENE Kos 1 Limited v Petroleo Brasileiro SA (No.2) [2012] 2 AC 164. In that case the issue was whether a charterer was liable, pursuant to a contractual indemnity, to indemnify a ship owner in respect of a loss which flowed from two effective causes, only one of which fell within the scope of the contractual indemnity. The majority of the Supreme Court concluded that the charterer was so liable. Lord Clarke said at paragraph 61:

"The real question under [the contractual indemnity] is whether the charterers' order to load the cargo was an effective cause of the owner having to bear a risk or cost of a kind which they had not contractually agreed to bear. I use the expression 'effective cause' in contrast to a mere 'but for' cause which does no more than provide the occasion for some other factor unrelated to the charterers' order to operate. If the charterers' order was an effective cause, in this sense, it does not matter whether it was the only one."

283. On behalf of Robson, Mr Ter Haar QC submitted that this was an incorrect approach to this case, for two reasons. First, he said that, if HSE were liable to Essex for the two causes of the loss, they had to establish that Robson were also responsible for both in order to get home under the indemnity and that, if they only established one element of their claim down the line, the indemnity was not triggered (the 'twin causes' point). Secondly, he said that an indemnitee could not rely upon an indemnity when his own negligence or default was also causative of the loss, unless the clause in question was sufficiently clear so as to include such indemnitee negligence. His argument was that, since HSE were in breach of contract themselves, they could not pass on the claim to Robson under the indemnity (the 'own breach' point).
284. As to the 'twin causes' point, I consider that there are two complete answers to this argument. First the indemnity provided by Robson was in respect of 'any liability'. In my view, as a matter of construction, that would cover Robson's liability for the NRV, even if Robson were not liable for the IV. Secondly, I consider that I am bound to follow Lord Clarke's approach in ENE Kos 1, following as it does the Court of Appeal's decision in The Miss Jay Jay [1987] 1 Lloyd's LR 32. There, the insurers insured against an adverse sea but not defective manufacture or design. Both were found to be proximate causes of the loss, and the Court of Appeal upheld the first instance judge that the owners could claim under the policy. Mr Ter Haar QC sought to counter that by relying on a paragraph in Hudson's Building and Engineering Contracts at 10-031, but I consider that this passage is dealing with a very different issue, namely primary and/or secondary defaults. The passage does not touch upon the principle noted in ENE Kos 1, to the effect that if the indemnitee was liable for one of two effective causes, then that was sufficient to allow a full recovery.

255. (a) As to the 'own breach' point, I do not accept Mr Ter Haar QC's broad submission. If it were right, it would mean that, in circumstances where B was liable to A for a breach of contract, but had sub-contracted all its relevant contractual obligations to C and obtained an indemnity in respect of C's performance, B would never be able to recover pursuant to the indemnity. It would destroy the ability of parties to pass claims down the contractual chain. No authority was cited in support of the submission. To the extent that Mr Ter Haar QC relied on the same paragraph in Hudson, I consider that the cases cited in that paragraph are not relevant to this issue, being personal injury cases concerned with negligence, which forms no part of Essex's claim against HSE or HSE's claim against Robson.

(b) I therefore conclude that his submission is contrary to both principle and business commonsense. I should also add that, if this argument were right, it would give HSE a defence to the claim from Essex, but they do not advance any such defence. Moreover, Ms Sinclair QC, on behalf of Essex, submitted to me that the argument was wrong in principle. For the reasons I have given, I respectfully agree with her.

285. In those circumstances, it follows that, in my judgment, given my finding that the NRV which Robson installed was one of the two effective causes of the flooding, they are liable to indemnify HSE in consequence of their breach. That is so, regardless of the position in relation to the closed IV.
286. The Closed IV
287. I accept that it does not follow that, merely because HSE were in breach of their sub-sub-contract (because the IV was closed at the time of the inspection before handover on 18 August), such a situation was automatically Robson's fault. HSE must show that, on the balance of probabilities, Robson's men were more likely than not to have closed and/or failed to open the IV immediately before the handover on 18 August 2006. For the reasons set out below, I consider that, on the balance of probabilities, HSE have established that claim.
288. First, I have found (paragraph 95 above) that the surge arrestor and IV were installed in July or even early August 2006, which was after the commissioning of the riser itself. In my view, that makes it more likely than not that the IV, which was therefore installed at the same time into a 'live' riser, was installed in the closed position, in order to minimise the disruption caused by this late element of the work. If that is right, this IV may simply never have been opened.
289. Secondly, although Robson's case was opened on a basis which sought to minimise their role it was, on analysis, factually incorrect. It was said that Robson completed the commissioning of the Core 2 riser on 5 May 2006 and that, thereafter, they did not perform any work on that riser. But some further testing to that riser was carried out in June. And thereafter, Robson were repeatedly working in and around the Core 2 riser on the completion and/or commissioning of various items of work to the BMCWS. That explained, for example, the grumbling emails from Mr Sofroniou, to which I have referred at paragraph 92 above.
290. Thirdly, immediately before the inspection on 17 August, Robson were commissioning the pipework in the apartments on Level 8 of Core 2 at Holly Court, including (according to the contemporaneous documents) all run-outs to taps, and back through the HIUs to the BMCWS valve in the riser. They were also balancing the heating systems and commissioning the CHP/HIU units in all the Core 2 flats. Such work could easily have led them to close the IV (assuming now that it was open), particularly as they were opening and closing all the other IVs in the same riser cupboard. The photograph shows that the six IVs in the cupboard that related to the individual apartments were all closed at the time of the inspection, but those IVs must have been subsequently opened (by Robson) to allow a water supply into the apartments.
291. Furthermore, as the plumbers responsible for the installation of the surge arrestor and its IV, Robson knew or ought to have known that, if the IV was in the closed position, the surge arrestor would not work. If it was closed, they should have opened it. They had an ongoing obligation to see that the pipework was not left in a defective state, which included ensuring that this IV was left open.
292. At one point, Robson suggested that, rather than themselves, the likely candidate for closing the IV was Goodwater, who carried out chlorination and flushing services. But there was no evidence at all that Goodwater shut the IV and nothing to say that they would have had any reason to do so. Indeed, the Goodwater certification suggests that all the IVs were open at the time of chlorination. Mr Hughes, the Robson plumber working on Core 2, agreed that Goodwater had handed back the BMCWS to Robson on 15 August, in order that Robson could continue with its commissioning works.
293. Fourthly, I also agree with Mr Hargreaves QC's proposition that the fact that Robson were continuing to commission/test certain works (particularly in connection with the HIUs) after 17 August is also relevant to this part of the case. Mr Marshall expressly said that, when a plumber is working on a live system after handover, he has to ensure that the riser is safe and working properly before even starting on his own works. In this case, that would have been a further opportunity for Robson to open the closed IV.
294. Finally on this point, Mr Hughes, (paragraph 13 of his witness statement) accepted that, when Robson had completed its various works of commissioning, whenever that was, they had to leave the system operational. In his cross-examination by Mr Hargreaves QC (Day 2 Page 142), there was this exchange:

"Q: Then the commissioning of the Core 2 heating systems and HIU's and so on. We've already discussed, this was all work on a live system?

A: Yes.

Q: That system had already been tested and commissioned so far as you are aware Mr Kelly?

A: Mr Hughes.

Q: I am so sorry. Forgive me. So it was important to leave the system operational when you had completed all of these works wasn't it?

A: Yes."

295. For all these reasons I accept paragraphs 92-99 of Mr Hargreaves QC's written closing submissions. Either Robson installed the IV in the closed position and never opened it, or closed it in the period leading up to or shortly after 18 August, when they were opening and closing all the other IVs in the relevant riser cupboard; or they failed to ensure that it was left open at the end of all their testing and commissioning works. On any view, in July and August (and thereafter), Robson had been opening and closing IVs on a system which had been commissioned, so they had to ensure that their own operatives did not alter the position of any valves (a proposition expressly accepted by Mr Hughes and Mr Carver) or, when they left, they had to ensure that all of the valves were in the correct position.
296. Finally, I should add this. Mr Bilson of Robson was asked detailed questions by Mr Hargreaves QC about Robson being on site in August doing the commissioning and testing, work which meant that, at least from time to time, they were in the riser cupboards. It was put to him that he would have noticed a closed IV on the main riser. Mr Bilson said he would not necessarily have noticed that, because he was concentrating on something else. That response was honest, but it demonstrated the dangers in Robson's limited approach to completion and commissioning. It was not unrealistic or unreasonable to expect Robson to check the position of all the IVs in a riser cupboard when they themselves were regularly in those cupboards, repeatedly opening and closing the IVs, and when they had installed all the relevant IVs.
297. Accordingly, I find that liability for the closed IV has been passed on by HSE to Robson. Robson are therefore liable for the closed IV, as well as for the NRV. However, because of my earlier findings in relation to the NRV and the indemnity, that liability is, at least in one sense, academic.

14. CONTRIBUTORY NEGLIGENCE

298. 1 The Allegations
299. The allegations of contributory negligence are made by Essex alone. They were rather under-particularised, a complaint which Mr Stansfield QC raised at the outset of the trial. The allegations in respect of Core 3 were concerned with maintenance, and can therefore be rejected as a matter of fact at the outset, because I have found that the critical defaults in relation to Core 3 (the metallic particles and the over-tightening) occurred prior to handover. This was properly acknowledged by Ms Sinclair QC on behalf of Essex when she decided no longer to pursue the allegations in regard of Core 3.
300. On Core 2, the allegations of contributory negligence could be summarised as follows:

(a) If the maintenance procedures prescribed by the O and M Manuals had been carried out as required then the Core 2 flooding would not have occurred;

(b) Had GMVL acted reasonably in respect of maintenance, the Core 2 leak would have been avoided.

301. 2 Are These Allegations Open To Essex?
302. The first issue is whether or not these allegations were even open to Essex. In Vesta v Butcher [1996] 2 Lloyd's Rep 179, Hobhouse J (as he then was) identified three possible categories of a defendant's liability:

(a) Where the defendant's liability arises from some contractual provision which does not depend on negligence on the part of the defendant;

(b) Where the defendant's liability arises from a contractual obligation which is expressed in terms of taking care (or is equivalent) but does not correspond to a common law duty to take care which would exist in the given case independently of contract;

(c) Where the defendant's liability in contract is the same as his liability in the tort of negligence independently of the existence of any contract.

He decided that it was only in a category (c) case that an allegation of contributory negligence can operate to reduce the claimant's recovery.

303. Mr Stansfield QC contends that Essex's liability arises out of contractual provisions which do not depend on negligence, so that the allegations of contributory negligence are not open to them. In relation to the critical argument as to the defects of workmanship, Ms Sinclair QC properly conceded that point, accepting that the workmanship allegations did not depend on contractual obligations the equivalent of a duty of care at common law. However, she maintained that the allegations of contributory negligence could arise in respect of the design allegations.
304. In my view this gives rise to two major difficulties. First, Essex allege that their only duty in tort was to exercise reasonable skill and care in the selection of a sub-contractor. No breach of that duty is alleged. It is not a duty that is co-extensive with any contractual obligations (for design or anything else): Essex are liable in contract for the work that it sub-contracted, but in tort they are not. Contributory negligence is therefore not available to Essex: see Raflatac Ltd v Eade Ltd [1999] BLR 261, at 265.
305. Secondly, even if that is wrong, and the design obligations are in some way co-extensive with a duty in tort, it is the workmanship allegations which are central to GMVL's case against Essex. Since the damages recovered in consequence of those successful allegations cannot as a matter of law be reduced by an allegation of contributory negligence it would, in my view, be artificial to allow an allegation of contributory negligence to reduce GMVL's recovery, let alone to reduce it by the 50% advanced by Ms Sinclair QC, by reference to the existence of a case against Essex on design which was of considerably less significance to the causation of the loss.
306. No authority was cited to me in support of Essex's approach. On first principles, therefore, I conclude that Essex cannot reduce their liability to GMVL as a matter of law by reference to allegations of contributory negligence which can only be maintained because of a secondary case against them on design rather than the primary case on workmanship. Thus, as a matter of law, I find that the allegations of contributory negligence are not open to Essex. However, in case I am wrong about that, it is necessary to consider whether the allegations can succeed on the facts. I do that in two ways: by looking at the allegations as a matter of principle, and then as a matter of detail.
307. 3 The Maintenance Allegations: Principle
308. As already noted, although LOR was liable to return during the DLP to make good the defects from the moment the buildings were handed over, GMVL had a separate obligation to maintain them properly. Accordingly, if an important piece of plant wore out 18 months after handover, because it had not been properly maintained, then that may well not have been a defect under the building contract. It would have been a failure to maintain and therefore an arguable element of contributory negligence.
309. In the present case however, the allegations of contributory negligence do not depend on the wearing out or the failure to maintain items of plant or equipment. Instead, what is said is that GMVL's maintenance contractor, Pinnacle, should have spotted defects which neither Essex nor any other defendant, spotted in the design, specification or installation of their own works. That is a very different thing.
310. Essentially, Essex had to persuade me that GMVL were contributorily negligent because GMVL did not rectify the defects for which Essex are responsible. Ms Sinclair QC very properly conceded that she was aware of no authority in which a building owner was found liable for contributory negligence in such circumstances. It seems to me to be contrary to basic principle. A contractor cannot reduce his liability for defective work on the basis that the employer failed to spot that same defective work after handover. In one sense, it might be thought that such a contention was an unlawful attempt by Essex to rely on its own breach of contract.
311. At one point during her submissions, Ms Sinclair QC relied on my own decision in Trebor Bassett Holdings Ltd v ADT Fire and Security PLC [2011] EWHC 1936 (TCC), [2011] BLR 61, where I made a 75% reduction against the building owners on account of their contributory negligence, a deduction upheld subsequently by the Court of Appeal at [2012] EWCA Civ 1158. But in my view, that was a very different case. There the employer was contributorily negligent because it failed to take the advice of its own property department and failed to install sprinklers in a factory where there was a clear risk of fire. It was a failure to heed detailed internal advice. In the present case, there was no such failure; indeed, there was no unheeded advice, because Essex were unaware of the defects that they had caused.
312. In those circumstances, I do not consider that, as a matter of principle, this is a situation in which allegations of contributory negligence can be sustained in any event. But in case I am wrong about that too I go on to look at the detailed allegations.
313. 4 The Maintenance Allegations: Detail
314. In my view, for the detailed reasons set out below, the allegations of contributory negligence must fail.
315. Firstly, I consider that the problems at the heart of this case are relatively straightforward: they involve a surge arrestor that was incapacitated because of an NRV and a closed IV. Yet not one of those items of equipment was even referred to in the O and M Manuals. It is impossible sensibly to criticise GMVL's maintenance regime in relation to those specific items in circumstances where HL/Essex failed to include them in the O and M Manuals in the first place.
316. Secondly, the experts were agreed that, although the O and M Manuals should have stated that the IV must only be closed for maintenance of the surge arrestor, it did not do so. Had it done, on this line of argument, then there may have been a failure by GMVL to maintain properly. But, because the O and M Manuals were silent on this critical point, which was related of course to the failure to put in a warning notice or label to which I have already referred, this was a breach by Essex (and HL) and cannot now be relied on by Essex to found a claim of contributory negligence.
317. Thirdly, there was force in Mr Stansfield QC's argument that, by reference to British Standard 6700, where inspections were recommended at least annually, it would be difficult to criticise GMVL even if no maintenance inspection had been carried out here in that time, because the flooding occurred less than a year after the handover.
318. Essex's principal response to all of this was to say that, if a competent maintenance contractor had been appointed, that contractor would have produced its own detailed asset register and would have noted the presence of both the NRV and the closed IV. But again, it seems to me that this argument is not open to Essex, in circumstances where they were responsible for the presence of the NRV and the closed IV in the first place, and where the O and M Manuals were entirely silent on the critical matters to which I have referred. Although the experts expressed views about what a competent maintenance contractor might have done, it seems to me that that is a matter of fact for the court and, in the circumstances to which I have already referred, I am not persuaded that there was any failure by GMVL which could amount to contributory negligence. I am also bound to note (although I do not decide the point on this basis) that the asset register point was not pleaded.
319. For all these reasons, I consider that the allegations of contributory negligence must fail.

15. CONCLUSIONS

320. For the reasons set out in Section 6 above, I consider that the cause of the Core 3 flooding was a combination of the metal debris in the threads of the nut and the over-tightening of the nut, as a result of the use of metal tools before practical completion.
321. For the reasons set out in Section 7 and 8 above, I consider that, in respect of the Core 3 flooding, Essex were in breach of their sub-contract and their warranty to GMVL; that HSE were in turn liable to Essex for the same defaults; and that Robson were in turn liable to HSE for the same defaults.
322. For the reasons set out in Section 9 above, I consider that the causes of the Core 2 flooding were the various failures of design and workmanship in respect of both the closed IV and the NRV which caused the vacuum and prevented the surge arrestor from working. The presence of the NRV and the closed IV were separate, independent, but effective causes of the flooding in Core 2. The poor workmanship was the principal reason for the failure, but design issues were also of relevance.
323. For the reasons set out at Sections 10, 11, 12 and 13 I consider that both Essex and HL are jointly and severally liable to GMVL for the relevant failures. I have apportioned liability between them on the basis that HL were liable for 40% and Essex 60%. Essex can pass on their liability to HSE. HSE are entitled to be indemnified by Robson because of the latter's wrongful installation of the NRV. HSE have also made out their case that it was Robson who was responsible for the closed IV, so the indemnity would attach to that failure too.
324. For the reasons set out in Section 14 above, I reject Essex's allegations of contributory negligence against GMVL.
325. I will deal with all issues of interest and costs, and any other ancillary matters after the handing down of this Judgment.

Note 1   At one point, WTP were introduced into these proceedings as a Fourth Party, but no live claim is now pursued against them.    [Back]

Note 2   This expression was used at HL’s insistence, rather than the word “ensure”.    [Back]

Note 3   This was all too typical of HL’s erratic performance: they had had months to recommend/require the installation of surge arrestors and had not done so; they then sent a general recommendation for the surge arrestors and did nothing more about it, but less than three weeks later they were complaining about the failure to install them.    [Back]

Note 4   As I have said, the IV was at least a metre below the surge arrestor. There was no discernible reason for this. The manufacturer’s drawings indicate that the IV should have been immediately underneath the surge arrestor.    [Back]

Note 5   Mr Stansfield QC made clear that that is not accepted by GMVL and the evidence was not entirely clear-cut.    [Back]