[Home] [Databases] [World Law] [Multidatabase Search] [Help] [Feedback]
	England and Wales Court of Appeal (Civil Division) Decisions
You are here: BAILII >> Databases >> England and Wales Court of Appeal (Civil Division) Decisions >> Yorkshire Water Services Ltd. v Taylor Woodrow Construction Northern Ltd & Ors [2005] EWCA Civ 894 (19 July 2005) URL: http://www.bailii.org/ew/cases/EWCA/Civ/2005/894.html Cite as: [2005] EWCA Civ 894

[New search] [Printable RTF version] [Help]

IN THE SUPREME COURT OF JUDICATURE
COURT OF APPEAL (CIVIL DIVISION)
ON APPEAL FROM TECHNOLOGY AND CONSTRUCTION COURT
MR JUSTICE FORBES
HT-01-289, [2004] EWHC 1660 (TCC)

		Royal Courts of Justice Strand, London, WC2A 2LL
		19/07/2005

Lord Justice May:

Introduction

1. This is an application by Yorkshire Water Services Ltd for permission to appeal against the decision and order of Forbes J, sitting in the Technology and Construction Court, of 8th July 2004 dismissing their action against the defendants, Taylor Woodrow Construction Northern Ltd. Although this is a decision on an application, it is the considered reserved decision of a constitution of three judges, and may be referred to as such for the matters of principle discussed in paragraphs 19 to 32.
2. The claim related to works of improvement carried out by Taylor Woodrow for Yorkshire Water to sewage treatment works at Knostrop, near Leeds. The principal component of the works was the construction of a six basin Sequencing Batch Reactor ("SBR"). The works were needed to improve the treatment of sewage to enable Yorkshire Water to comply with revised effluent standards due to be required by the National Environmental Programme from 1st July 2000, and the Waste Water Treatment Directive from 31st December 2000. The effluent from the Knostrop sewage works is discharged into the River Aire.
3. Taylor Woodrow were engaged to perform the works under a written main contract dated 19th August 1998, which provided for the works to start on 8th December 1997. The accepted tender price was £12,458,537.19, which itself gives an indication of the scale of the works.
4. Biwater Treatment Ltd, the first Part 20 defendant, were the process sub-contractors appointed by Taylor Woodrow for the SBR. The judge had determined a preliminary issue as to the terms of the sub-contract between Taylor Woodrow and Biwater. Taylor Woodrow and Biwater reached a compromise during the trial of Yorkshire Water's claims. Biwater admitted liability to indemnify Taylor Woodrow against Yorkshire Water's claim and conducted Taylor Woodrow's defence of the action for them.
5. Elga Ltd, the second Part 20 defendant, had a sub-contract with Biwater to design, manufacture and supply the SBR. Elga are the UK associate company of US Filter Inc (sometimes known as Jet Tech), an American company that owns the design rights in the relevant type of SBR.
6. Taylor Woodrow, Biwater and Elga carried out the works, so that the plant passed with some qualification contractual takeover tests on 26th November 1999. Yorkshire Water took over the plant on the same day. The contract required further performance tests. These had not been undertaken by November 2000. It was then, and remains, Yorkshire Water's case that the plant as designed and constructed was defective. They seriously considered determining Taylor Woodrow's contract, but did not in fact do so. However, Taylor Woodrow, Biwater and Elga ceased to be operationally concerned with the plant in about November 2000. Yorkshire Water engaged others to carry out additional, or as they saw it, remedial works, including the construction of an external selector to overcome poor sludge settleability.
7. In these proceedings, Yorkshire Water by their amended claim form claimed payment of all sums due to them, being the cost of remedial work under clauses 34.8, 36.7, 36.11 and/or 35.10 and Schedule 10 of the contract. They also claimed declarations in support of their allegation that there was a defect or defects in the plant and to try to cover the fact that they had not incurred the costs of carrying out all the remedial work which they said was necessary. Yorkshire Water's pleaded monetary claim was for £12,010,640, which they said was the actual or prospective cost of remedial works.
8. Taylor Woodrow counterclaimed for costs said to be due to themselves and for a claim by Biwater against Taylor Woodrow for an extension of time and additional payments. Taylor Woodrow also had a Part 20 claim against Biwater which, as I have said, was compromised.
9. Biwater had a Part 20 claim against Elga, and Elga had a counterclaim against Biwater for £101,894.55, which was mainly the agreed outstanding balance of their subcontract price.
10. Subject to the ramifications of Yorkshire Water's claim against Taylor Woodrow, the disputes about the counterclaims and Part 20 claims were limited. The judge decided such issues as there were. They are not the subject of the present application.
11. The judge dismissed Yorkshire Water's claims in their entirety. In the round, Taylor Woodrow's, Biwater's and Elga's claims all succeeded. It was a comprehensive defeat for Yorkshire Water both overall and in detail. Yorkshire Water seek permission to appeal against this decision. There are some 19 proposed grounds of appeal, which between them would seek to overturn most of the judge's findings against Yorkshire Water both of fact and law. The first 7 grounds raise points of construction of the contract and are, so far as they go, points of law. The remaining grounds of appeal between them seek to reverse the main thrust of all the important findings of fact. There are incidental points of law in some of the grounds which Yorkshire Water say are important, and which affected the judge's findings of fact, such that these require to be reviewed.
12. Rix LJ considered this application on the papers. He spent three days reading them, at the end of which he granted a necessary extension of time and adjourned the application for hearing on notice. He gave these as his reasons:

"1. My provisional view is: that PTA could be granted on questions of law but that I am not currently persuaded that PTA should be granted on questions of fact or in any event without first hearing the respondents on notice.

2. As for questions of law, I am provisionally of the view that certain at least of the questions of construction might in themselves have a reasonable prospect of success and on that basis I should not be minded to pick and choose between them.

3. It is not clear to me, nevertheless, whether the judge's findings of fact are wholly determinative of the claim however all questions of law are decided. I think they are (see judgment at para 533), save possibly in cases where the questions of law other than the preliminary issues relate closely to the judge's findings, viz Grounds 6, 8, 10 and 16 (skeleton para 206). However even in those cases, it provisionally seems to me that the judge's findings overwhelm the issues of law identified.

4. In these circumstances, it would seem that no PTA should be given unless the court can be persuaded that the judge's findings should be reviewed."

Rix LJ drew attention to four particular points, which he suggested Yorkshire Water might consider concentrating upon.

13. The judge heard the action on about 35 intermittent hearing days between 21st May and 20th November 2003. I imagine that some of the breaks in continuity were because the parties required time to prepare the presentation of immensely complicated technical evidence and argument. The judge then delivered his judgment eight months after the conclusion of the hearing. This is not a criticism, only perhaps a commentary – mild in this case – on an overloaded litigation culture. The judge, no doubt, had much else to do. The judgment runs to 557 paragraphs covering 165 single-spaced A4 pages in what looks to be no larger than a 12 pt font. It is a masterpiece of analysis of highly complicated technical detail of a mechanical and biochemical nature.
14. This is an application for permission to appeal, not an appeal. The judge spent more than 13 months immersed in the detail of the claim. It was at one stage a matter for consideration how long it was necessary for this Court to devote to this application. The answer in general is, of course, as long as is reasonably and proportionately necessary to do justice to the parties. The provisions of rule 1.1 of the CPR are in point. At a directions hearing on 14th February 2005, the parties agreed and accepted Rix LJ's direction that the oral hearing should be confined to one day. The parties' joint view was that the Court would need five days pre-reading, which each member of this Court has had and taken. The parties have economically provided each member of the Court with only two cardboard boxes containing only ten lever arch files. I say "only", because I am sure that Forbes J had vastly more paper than this.
15. I have read the judge's judgment and the parties' long skeleton arguments, together with so much of the accompanying paper as appears to be central. I reckon, with some diffidence, to have understood in the round the main technical points dealt with by the judge and the parties' commentary on them. The very nature of our respective input necessarily means that I do not have anything approaching the judge's depth of understanding of the interlocking finer points of what in detail is an immensely complicated subject. I trust, however, that my understanding is sufficient for the court's present purpose. I do not propose in this judgment to recite vast chunks of technical fact or expert analysis simply to prove (or disprove) that I have understood it all. That would in my view be a disproportionate use of the Court's resources. I ask the parties to trust that, in so far as I do not do that, I do nevertheless have sufficient general understanding for present purposes. It will necessarily be imperfect, but, as I say, this is an application, not an appeal.
16. The respondents say that this is a hopeless application, because it cannot succeed in any sensible way unless permission is given to open all the facts; and that the judge's decisions of fact are simply not amenable to appeal. Such points of law as might possibly be regarded as arguable are indeed, they say, overwhelmed by the facts. Limited permission to appeal on points of law only should not be given because, without a wholesale review of the facts, it would get Yorkshire Water nowhere.
17. Yorkshire Water accept that some of the points of law are pointless if they cannot shift the judge's findings of fact. But they say that others are capable by themselves of leading to some, perhaps limited, recovery. They say that errors of law by the judge embedded in grounds 6, 8, 10 and 16 infected his approach to the facts, which therefore require to be reviewed. But Mr Sears QC on behalf of Yorkshire Water realistically accepted that there was little joy for Yorkshire Water, unless they were given permission to appeal on what he said were the central findings of fact.
18. I have concluded that this is simply not a viable proposed appeal, unless Yorkshire Water could persuade us that they should be permitted to reopen large swathes of the technical facts. This was Rix LJ's provisional view, and I agree with him.

The court's approach to findings of fact

19. The first question, therefore, is what should be the Court's approach to the application for permission to appeal against the judge's findings of fact.
20. We shall soon reach a time (if we have not already reached it) when those under the age of 30 will not know from direct experience that judges of the Technology and Construction Court used to be known as Official Referees. They were and technically still are judges nominated by the Lord Chancellor to deal with Official Referees' business under section 68(1)(a) of the Supreme Court Act 1981. The terms of that section were not changed when the Technology and Construction Court was established in 1998.
21. Appeals to the Court of Appeal from decisions of TCC judges come within Part 52 of the CPR and no specific provision relates to them. Subject to Rule 52.3(1), permission to appeal is required. Permission to appeal will only be given where the court considers that the appeal would have a real prospect of success or where there is some other compelling reason why the appeal should be heard (Rule 52.3(6)). A cynic might say that the need to re-open highly complicated factual and technical issues on which a highly competent technical judge has immersed himself for over a period of thirteen months and produced a masterpiece of a judgment was a compelling reason why the proposed appeal should not be heard. The cynic would not be far from the mark.
22. Rule 52.11 of the CPR provides that every appeal will be limited to a review of the decision of the lower court unless a practice direction makes different provision for a particular category of appeal – not relevant to the TCC – or the court considers that in the circumstances of an individual appeal it would be in the interests of justice to hold a rehearing. There has been judicial consideration about what under the CPR constitutes a review and what a rehearing. In truth, there is a spectrum shaded between the one and the other – see paragraphs 84-98 of my judgment in E.I. Du Pont de Nemours v S.T. Dupont [2003] EWCA Civ 1368 (with which Aldous LJ (paragraph 6) and Keene LJ agreed) for an extended consideration. Yorkshire Water do not ask for, and the court would never contemplate, a rehearing "in the fullest sense of the word" – see paragraph 31 of the judgment of Brooke LJ in Tanfern Limited v Cameron-MacDonald [2000] 1 WLR 1311 at 1317. But they do ask for a review, the extent and intensity of which would have to be enormous. Further, they do not, understandably, seek a retrial. So this court on an appeal would not only be asked to hold that the judge's decisions of fact were wrong. We should also be asked to decide what the correct decisions of fact should have been. In the present case, those would be two vastly different operations, the second much more extensive than the first. In truth, I do not think that Yorkshire Water could reasonably expect to do better on an appeal than to secure a retrial. This is a general comment on the nature of the matter, before looking at the detail.
23. It was section 54 of the Access to Justice Act 1999 which enacted that rules of court might provide that any right of appeal to a county court, the High Court or the Court of Appeal might be exercised only with permission. Part 52 of the CPR, which came into force on 2nd March 2000, created a uniform system of appeals and replaced the former provisions of the Rules of the Supreme Court relating to appeals. For most ordinary appeals from decisions of the High Court and from many decisions in county courts, there had historically been a right of appeal on both law and fact without leave. RSC Order 58 rule 4 applied to appeals from Official Referees. It provided that, subject to section 18 of the 1981 Act, an appeal lay to the Court of Appeal from a decision of an Official Referee on a question of law, and with leave of the Official Referee or the Court of Appeal on any question of fact. Thus the pre-2000 rules provided a more restrictive regime for appeals from decisions of Official Referees on questions of fact than for appeals from decisions of other courts.
24. The antecedent statutory history of the former RSC Order 58 rule 4 is to be found briefly summarised in the judgment of Lord Bingham MR in Virgin Management Ltd v De Morgan Group 68 BLR 32 at page 33 as follows:

"Rights of appeal from official referees have varied over the years … it is enough for present purposes to note that under the Administration of Justice Act 1932 an appeal without leave lay on a point of law but not any question of fact; that following the Administration of Justice Act 1956 the rules provided for an appeal without leave on a point of law or 'on a question of fact relevant to a charge of fraud or breach of professional duty'; and that under the present rule leave is required on any question of fact. Thus leave is now required even on questions of fact relevant to a charge of fraud or breach of professional duty which could previously have been the subject of an appeal without leave."

This last change was made in 1988. Lord Bingham continued:

"The reason why rights of appeal against the factual decisions of official referees have been subject to special conditions and restrictions is obvious. Official referees were originally appointed to conduct protracted and painstaking enquiries into complex and detailed factual issues. The investigation of multifarious claims and cross-claims arising out of construction contracts is one example. The taking of accounts is another. The assessment of damages following an established breach of contract is yet another. We agree with Waller LJ when he said, giving the judgment of the Court in Moody v Ellis (1983) 26 BLR 39 at page 46, with reference to the old rule which only allowed factual appeals relating to fraud and breach of professional duty,

'In our judgment the policy behind the rule is clear. In the ordinary way the official referee's decision on fact is to be final. He had to go into very complicated and detailed factual questions which take a great deal of time. As a matter of policy, when an official referee, with his great experience in such matters, makes a decision, it should ordinarily be final.'

It was clearly felt that the factual minutiae with which official referees were required to deal were not suitable matters for reconsideration on appeal. It was better that, for better or worse, their decisions on such matters should stand, rather as if they were arbitrators. If personal or professional reputations were at stake, that was a different matter."

25. Lord Bingham discussed the reason for this. He then considered the test that should be applied to an application for leave to appeal from a decision of an Official Referee on a question of fact. He concluded that the test for the grant of leave to appeal to the Court of Appeal in this context should be whether the ground of appeal which it is sought to argue has a reasonable prospect of success: if so leave should be granted; if not, leave should not be granted. This closely approximates to the now universal test provided by Rule 52.3(6) of the CPR. Lord Bingham continued at page 34I:

"Whether a ground will be regarded as having a reasonable prospect of success will of course depend on what the point is. A prospective appellant will have difficulty in showing that the test is satisfied if he is seeking to challenge:

(1) an official referee's findings of primary fact based on his evaluation of oral evidence, not because of any peculiar immunity enjoyed by official referees but because of the weight invariably given to the trial judge's factual conclusions unless and until they are shown to be wrong;

(2) the fine detail of an official referee's factual investigation: while the official referee's findings of this nature are no longer final, the Court of Appeal will not readily enter upon an enquiry of this kind;

(3) findings of fact falling within an official referee's area of specialised expertise, particularly if the official referee has had the advantage of inspecting the site or the subject matter of the dispute in question.

We do not suggest that these are no-go areas, only that the burden on a prospective appellant in these areas will be hard to discharge."

26. Having said that the task of a prospective appellant in showing that he meets the test will be more easily discharged if he is seeking to challenge a secondary inference drawn from the facts, and having referred to possible links between issues of fact and issues of law, Lord Bingham continued at page 35I:

"Leave to appeal on questions of fact should not, however, be granted simply because an appeal on legal grounds is proceeding anyway. Applications for leave should always be made and considered with reference to specific issues of fact. They should not be granted unless it appears that the challenge, if successful, would be likely to affect the official referee's overall decision …

Care must be taken to prevent applications for leave blossoming into dress rehearsals for the full appeals. Over a century ago Bowen LJ gave a salutary extra-curricular warning ((1886) 2 LQR 1 at page 10):

'To impose on the Court of Appeal the duty of determining whether leave to appeal should be granted is really to require them half to hear a case in order to decide whether they should wholly hear it.'

This need not, we consider, be so if it is constantly borne in mind that at the leave stage the applicant's only task is to show that a ground has a reasonable prospect of success and the court's only task is to decide whether or not that appears to be so."

27. I do not consider that the advent of Part 52 of the CPR has altered this court's approach to applications for permission to appeal from factual decisions of TCC judges. The fact that permission to appeal is now needed in many cases, where formerly it was not, does not predicate a more relaxed approach than formerly in TCC cases. As Lord Bingham said, a reasonable prospect of success will depend on what the point is. But he had expressed agreement with the policy behind the pre-1988 rule stated by Waller LJ in Moody v Ellis. The policy has been somewhat relaxed, and findings of fact such as Lord Bingham described on page 35 in Virgin Management are not no-go areas. The burden on a prospective appellant in these areas is nevertheless hard to discharge. In my view, the more complicated and technical the facts, the harder generally speaking is the burden. The reason again is obvious. The more complicated and technical the facts, the longer and more expensive would be this court's enquiry, whether by review or re-hearing, and the more disproportionate would be the whole exercise for the parties and the court alike. Importantly, this court would have the disadvantage of not having heard any of the witnesses, including the experts, give oral evidence. I venture to think that, at the extreme, some questions of fact may be so complicated and technical that they should only be investigated in detail judicially once, provided that the resulting decision is not palpably incompetent. That would not only apply to decisions of TCC judges. The facts in the present case are quite close to any extreme. So far from being palpably incompetent, Forbes J's factual decisions are, so far as I have been able to judge, of the highest quality. In so far as I have reached this conclusion upon no more than 5 days reading and one day of oral argument, care must be taken, as Lord Bingham said, to prevent applications for leave blossoming into dress rehearsals for a full appeal.
28. There are two recent decisions of this court, given after the introduction of Part 52 of the CPR, which bear upon this question. The more directly relevant comes from the judgment of Lawrence Collins J, with whom Thorpe and Latham LJJ both agreed, in Skanska Construction UK Limited v Egger (Barony) Limited [2002] EWCA Civ. 1914, 18^th December 2002. Lawrence Collins J said at paragraph 7:

"The appeal of Eggar and the cross appeal by Skanska are brought with the permission of a single Lord Justice. Under the CPR appeals lie with the permission of the judge or the Court of Appeal, from decisions of the Technology and Construction Court, as they did in the period after 1988 from decisions of the Official Referees. But the decisions of the Technology and Construction Court have special characteristics which affect the readiness of the Court of Appeal to reconsider them on appeal. First, the findings of fact often fall within an area of specialist expertise, where the evidence is of a technical nature and given by experienced experts, and which is evidence of a kind which judges of the Technology and Construction Court are particularly well placed to assess. Second, the conclusions of fact will frequently involve an assessment or evaluation of a number of different factors which have to be weighed against each other, which is often a matter of degree. Third, the decisions may deal with factual minutiae not easily susceptible of reconsideration on appeal. Fourth, the judgments will frequently be written on the basis of assumed knowledge of the detail by the parties and their advisors, and will not address a wider audience, with the consequence that the underlying reasoning may not always be readily apparent or fully articulated.

Consequently the recent pronouncements of this court on appeals against findings of fact apply with particular force to the decisions of the Technology and Construction Court. In particular this court will be reluctant to interfere with a trial judge, not only on findings of primary fact based on the credibility or reliability of witnesses, but also where conclusions of fact involve an assessment of a number of different factors which have to be weighed against each other and involve an evaluation of the facts."

"Recent pronouncements of this court" no doubt included Assicurazioni Generali v Arab Insurance Group [2003] 1 WLR 577, [2002] EWCA Civ 1642. Each of the parties to the present application has referred to this case in support of their positions. Mr Sears points to Clarke LJ's citation in paragraph 17 from the judgment of Mance LJ in Todd v Adams & Chope [2002] 2 Lloyds Rep 293 at 319 that:

"Where the correctness of a finding of primary fact or of inference is in issue, it cannot be a matter of simple discretion how an appellate court approaches the matter. Once the appellant has shown a real prospect (justifying permission to appeal) that a finding or inference is wrong, the role of an appellate court is to determine whether or not this is so, giving full weight of course to the advantages enjoyed by any judge at first instance who has heard the oral evidence."

29. Mr Streatfeild-James QC for Biwater and Mr Elliott QC for Elga point to paragraphs 12, 14-16, 21 and 22 of Clarke LJ's judgment and to paragraphs 196 and 197 of Ward LJ's judgment. An appellant has to establish that a relevant finding of fact is plainly wrong or that the judge exceeded the generous ambit within which there might be reasonable disagreement about the conclusions drawn from the evidence. Relevant further observations include Clarke LJ at paragraph 14:

"The approach of the court to any particular case will depend upon the nature of the issues [and the] kind of case determined by the judge.";

Clarke LJ's citation at paragraph 21 of the judgment of Robert Walker LJ in Bessant's Case [2002] EWCA Civ 763 at paragraph 26:

"How reluctant should an appellate court be to interfere with the trial judge's evaluation of, and conclusion on, the primary facts? As Hoffmann LJ made clear in the Grayan case [1995] Ch. 241 there is no single standard which is appropriate to every case. The most important variables include the nature of the evaluation required, the standing and experience of the fact-finding judge or tribunal, and the extent to which the judge or tribunal had to assess the oral evidence.";

Clarke LJ's reference in paragraph 22 to what Lord Hoffmann called the "penumbra of imprecision as to emphasis, relative weight, minor qualification and nuance" which will vary from case to case; and Ward LJ at paragraph 196:

"So the more complex the question, the more likely it is that different judges will come to different conclusions and the harder it is to determine right from wrong. Borrowing language from other jurisprudence, the trial judge is entitled to "a margin of appreciation"."

30. The technical complexity of the present case is well beyond the practical judicial experience of most judges. To talk in this appeal about primary facts and inferences, the credibility of oral evidence and other matters normally relevant to an appeal against relatively uncomplicated findings of fact has something of an air of unreality about it. The well-known principles are of course relevant and to be applied. But no appellate court would solemnly set about deciding in this case which of the judge's findings were findings of primary fact and which inferences. A broader approach would of necessity be needed. Most of the evidence was technical and complex. Most of the witnesses, including those giving factual evidence, were experts. Credibility was scarcely in issue in judging who was telling the truth. But credibility was centrally in issue in the sense of judging which strands of evidence, opinion and argument were persuasive. It was here that the judge had the unique advantage both of hearing the evidence, opinion and argument expressed, and of being able, through becoming steeped in the subject, to put together in 557 paragraphs of dense technical prose the interlocking pieces of a highly complicated jigsaw. Understanding the jargon is an effort in itself.
31. Two of the members of this constitution have recently participated in hearing and deciding an appeal – Thomson v. Christie's [2005] EWCA Civ 555 – from a decision of Jack J, which had some broad similarities with the decision in the present case. It was not a TCC case. It concerned the date and provenance of two Louis XV porphyry and gilt-bronze two handled vases. There was much technical evidence, including metallurgical evidence. The general quality of Jack J's analysis and evaluation of this mass of evidence and of his factual conclusions was equivalent to the quality of Forbes J's analysis and evaluation in this case. The technical complexities in the present case are in my view substantially greater.
32. Paragraph 141 of the judgment in Thomson has these passages:

"The judge rejected Ms Thomson's case that the metallurgical evidence established that the vases were made in the 19^th century. The cross-appeal challenges this finding. For this challenge to succeed would mean that this court would have to be persuaded to reverse a finding of fact made by the judge upon a very careful composite evaluation of a large body of oral evidence, mainly expert evidence. It is a commonplace that this court will only rarely disturb a judge's finding of fact reached after hearing oral evidence whose credibility is in issue. It may sometimes be somewhat less difficult to depart from an evaluative judgment of expert evidence, if the evidence is mainly in writing and its evaluation an intellectual process largely unaffected by personal credibility. But that is not this case. The judge's factual conclusion that the vases were made in the 18^th century was a composite amalgam of a large number of particular judgments where an assessment of the weight and persuasiveness of each of the experts as witnesses was of critical importance. … It would, I think, be a travesty of the appellate process to interfere with the judge's composite judgment here as to the date when the vases were made unless it were shown that he made one or more glaring and important misjudgments. This is scarcely likely for such a thorough and careful judgment and the cross-appeal does not, in my judgment, begin to do so. … Restating evidence and submissions which the judge rejected does not persuade me that the judge was wrong, when the judge's conclusions were supported, as they were, by the opinions of Christie's experts. I accept Lord Grabiner's general submission that some individual matters on which the experts expressed a view were scarcely matters for their professional expertise, but rather questions of common sense or logic. I accept in general that the court is as able to consider such matters as the experts. These would not truly be amenable to expert opinion at all. And I grant that there are examples of matters of this kind in the evidence, as, for instance, … But, even accepting that individual points such as these are amenable to judicial appellate evaluation whatever the expert opinion, no appellate court should cherry pick a few such points so as to disagree with a composite first instance decision which, in the nature of a jig-saw, depended on the interlocking of a very large number of individual pieces, each the subject of oral expert evidence which the appellate court has not heard."

Thomson was a full appeal, which nevertheless illustrates aspects of what I consider should be the court's approach to the present application.

The main contract

33. This was a design and performance contract.
34. Yorkshire Water started the proceedings on the wrong foot by formulating their claim as a claim for damages for breach of the contract. This disregarded clause 44.2 of the General Conditions which provided that, except in the case of termination of the contract under clause 41, Taylor Woodrow's liability to Yorkshire Water for any breach of contract should be limited to the expenses, charges, damages and reimbursements expressly provided in the contract. Yorkshire Water had to make their claim under the contract, not as a claim for damages for breach of it.
35. Clause 34 of the General Conditions provided for Yorkshire Water to take the plant over and for the issue by the Project Manager of a Take-over certificate. Taylor Woodrow and their sub-contractors carried out the works and satisfied the contractual conditions for the Take-over, which was certified retrospectively as having taken place on 26^th November 1999. Clause 34.8 required Taylor Woodrow to complete all items noted on the Take-over certificate as incomplete. If they failed to do so, the cost of others carrying out the necessary work was for Taylor Woodrow's account. Yorkshire Water had a claim under clause 34.8.
36. Clause 35 of the General Conditions provided for performance tests specified in Schedule 7 to be carried out and to be passed within a year of the Take-over Certificate. Schedule 7 provided for tests to demonstrate that the SBR plant was capable of operating so as to achieve a number of guarantees, including guarantees as to effluent quality (TP1), flow capacity (TP2), reliability (TP9) and sludge production and sludge quality (TP11). These performance tests were never carried out. They were not carried out because everyone initially, but Yorkshire Water latterly, reckoned that the plant would fail the tests. By the summer or early autumn of 2000, Taylor Woodrow (Biwater) considered that the plant was working properly and wanted the tests carried out. They never were.
37. Clause 35.10 of the General Conditions provided that, if the plant did not pass the performance tests within a year of the Take-over date, Taylor Woodrow should pay Yorkshire Water "liquidated damages" in accordance with Schedule 10. Quixotically, the parts of Schedule 10 relevant to TP1, TP2 and TP9 did not provide for liquidated damages. They provided for Taylor Woodrow to pay Yorkshire Water "the actual costs incurred" to achieve the standard of the performance tests. These performance guarantees were among those said to be "absolute guarantees" which must be achieved, and were not subject to liquidated damages.
38. Clause 36 of the General Conditions enabled the project manager to serve a defects notice on Taylor Woodrow at any time up to a year after the date of Take-over giving particulars of defects. Taylor Woodrow were required to make good the specified defect at their own cost, if it was attributable to their own breach of contract. If Taylor Woodrow did not do this reasonably quickly, Yorkshire Water might do so. Clause 36.7 provided for Taylor Woodrow to reimburse Yorkshire Water the cost of so making good the defect "if the purchaser has made good the defect". Clause 36.11 extended Taylor Woodrow's obligations relating to defects not discovered upon reasonable examination at the date of Take-over to 12 years after the date of Take-over or the date when the relevant plant was brought into use.
39. The project manager served a defects notice under clause 36 on 14^th March 2000. This was initially expressed in rather general terms, supplemented by more details in a letter of 7th April 2000. It was Yorkshire Water's case that Taylor Woodrow failed to make good the specified defect or defects and they claimed the cost of doing so themselves. Taylor Woodrow's first case was that the plant was not defective.

Proposed grounds of appeal – points of law

40. The judge decided a number of preliminary issues as to the construction of the contract. He decided some of these in a judgment of 2^nd May 2003. Others he postponed to be decided in his main judgment. He decided the first and most significant preliminary issue largely in favour of Yorkshire Water. Taylor Woodrow had contended that Yorkshire Water could not bring a claim under Clause 35.10, if the relevant performance tests had never been carried out. The judge decided (paragraph 33) that this was correct for those performance tests for which Schedule 10 provided for liquidated damages, but incorrect for those that did not. He said in paragraph 37:

"In my view, Yorkshire Water's Schedule 10 entitlement to recover "the actual costs incurred" etc. in respect of TP1, TP2, TP9 and TP11(b) will arise if "the contractor is unable to meet the requirements of Schedule 7, Performance Tests …" within the relevant time scale. It is not necessary for the performance tests to have been actually carried out and failed in order for a Purchaser to be able to prove the inability of a Contractor to meet the necessary requirements of Schedule 7 performance Tests in respect of any of the absolute guarantees within the relevant time scale."

41. The first ground of appeal seeks to challenge the judge's decision that Yorkshire Water were not entitled to recover liquidated damages, if the performance tests were not carried out. It is accepted that success on this ground would take Yorkshire Water nowhere, unless they can successfully appeal the judge's factual findings in relation to the plant's ability to pass the performance tests.
42. The judge decided (paragraph 41) that Yorkshire Water had no contractual entitlement to be reimbursed the cost of making good a defect under Clause 36.7, if they had not carried out the remedial works. The second proposed ground of appeal seeks to challenge this decision. Yorkshire Water concede that the judge has made specific findings of fact that the plant was not defective for all but one of the individual defects listed in the letter of 7^th April 2000. So this ground of appeal leads nowhere without a successful appeal on the facts. The exception is said to be the occurrence of fill-decant events, which the judge acknowledged had been a problem (paragraph 261). Yorkshire Water contend that a solution of a fill-decant inhibit was not contractually compliant. But the judge accepted evidence that the fill-decant inhibit had an insignificant effect on the treatment capacity of the SBR. This matter does not warrant permission to appeal on ground 2 without permission on grounds which challenge the judge's findings of fact. See also ground 16.
43. Proposed grounds 3 and 4 also relate to decisions about Clause 36.7. These are closely related to ground 2 and the same concessions and considerations apply to them.
44. The judge decided (paragraphs 54-67) that the expression "actual costs incurred" in Schedule 10 meant that Yorkshire Water could not claim costs under the relevant parts of Schedule 10 if they had not actually incurred them. He also said (paragraph 67) that if he was wrong about this, it would make no difference to the outcome of the proceedings. Ground 5 of the proposed grounds of appeal seeks to challenge the judge's construction here. It is accepted that success on ground 5 would not give rise to a recovery, unless Yorkshire Water can successfully appeal the decision of fact that the plant would have passed its performance tests.
45. One of the design requirements in Clause 0.2.6.3 of the Specification was that the plant should be capable of treating the design flow and load to the effluent standard specified, "with one basin out of service". It was Yorkshire Water's case that this required a capacity to operate on 5 basins indefinitely and under all specified flow and load conditions. The judge considered and rejected this construction (paragraphs 92 to 107). He construed the requirement in its contractual context to mean that the provision was to enable such necessary operations as maintenance and repairs to be carried out whilst the otherwise "6 basin plant" continued to operate. If he was wrong about this, he was again satisfied that the outcome of the proceedings would remain unaffected. He pointed to paragraphs 529 and 530 of his judgment, at the end of which he concluded that there was every reason to believe that the SBR would have been able to meet the Schedule 7 performance requirements without difficulty, even if it operated throughout on 5 tanks.
46. Ground 6 of the proposed grounds of appeal seeks to challenge the judge's decision on this point of construction. Yorkshire Water say that his conclusion was based upon his finding, in paragraph 529, that the only circumstances in which, if operated correctly as originally designed and set up, the SBR would have "struggled" to nitrify would have been when being operated with 5 tanks and receiving BOD loads at FFT and at minimum ambient temperatures which was a combination of events which he regarded as being highly unlikely to occur. That being so, he went on to find that, even when operating on 5 tanks, the plant would have been able to nitrify for a sufficient period to cover such a very unlikely event. Yorkshire Water contend that the judge ignored the fact that Taylor Woodrow were obliged to supply a plant which could treat the stated flow and loads in all seasons and that this, on their construction, required performance tests operating on 5 tanks in all conditions. If this is correct, they say that the judge's decision on the central issue would have to be reviewed. I understand this to be a submission that, if Yorkshire Water are correct on the point of construction, a review of all or most of the main facts becomes necessary.
47. I have carefully considered the judge's extended and careful reasons for reaching the conclusion he did on this point of construction. I have considered equally carefully Yorkshire Water's submissions (in particular paragraphs 38 to 44 of the first skeleton argument) why he was wrong. Essentially it is said that the words of Clause 0.2.6.3 have no time limitation and should be given their ordinary and natural meaning. Reliance is also placed on certain parts of Taylor Woodrow's proposals.
48. I do not find these submissions persuasive. The judge took account of the main parts of the tender proposals and Yorkshire Water's submissions about them in paragraphs 101 to 103 of his judgment. I find the judge's reasons for his conclusion (which I do not set out or summarise) entirely convincing. The submissions in the grounds of appeal simply restate points which the judge took into account. Further, this is not a point of construction which can be determined shortly by reading and construing a few words. The judge was in a far better position to construe this provision in its contractual context than this court would be. I would refuse permission to appeal on this ground.
49. It was Yorkshire Water's case that Taylor Woodrow were contractually obliged to design, supply and install an SBR that was equipped with a standard Jet Tech IDSC manifold. IDSC is an acronym – one of many in this case – for Influent Distribution and Sludge Collection, and would thus be the manifold through which the influent to be treated flowed into the SBR and sludge was taken out. The practical relevance of this issue is that it was the firm view of Dr Chambers, one of Yorkshire Water's experts, that the omission of the IDSC was a significant cause of what he believed to be the SBR's lack of selector capacity. The judge came to the conclusion (paragraph 481) that Dr Chambers was wrong and rejected his evidence on this. He gave his reasons in paragraphs 481 to 488 of his judgment. The reasons are summarised in paragraph 488. Essentially the judge accepted the evidence of Mr Norcross that Jet Tech had built many successful industrial SBRs without an IDSC and that the decision to use the chute feed was correct; and the evidence of Dr Horan that the dump fill constituted the best selector you could possibly build. The judge decided (paragraph 488) that the omission of the IDSC did not prejudice the SBR's selector capacity.
50. The judge decided that the contract did not oblige Taylor Woodrow to provide an IDSC. Ground 7 of the proposed grounds of appeal sought to challenge this finding, but the ground is not pursued because Yorkshire Water made no claim for the cost of providing an IDSC. They say that the negotiations and discussion concerning its need remain part of their factual case. They would no doubt want to challenge the factual conclusion that an IDSC was not needed.
51. Ground 8 of the proposed grounds of appeal concerns the judge's findings relating to the capability of the plant to meet the contractual performance requirements. This is centrally and almost entirely a challenge on factual grounds, but there is said to be an embedded point of law. It is said that the judge failed to take account of the fact that Schedule 7 required performance tests to be undertaken over a series of four 28 day periods to reflect the seasons over a maximum of 52 weeks after takeover. A factual finding that the plant would have been able to pass the performance tests in the Autumn of 2000 is not enough to disentitle Yorkshire Water from a remedy under Schedule 10. Apart from the difficulty that this is relatively a small point in a welter of factual analysis and decisions, it seems to me to misstate the main question under Schedule 10. This was, as the judge said, that, in the absence of the contractual performance tests which Yorkshire Water had chosen not to conduct, it was for them to establish by other means that the plant did not meet the Schedule 7 performance requirements. They failed to do so with reference (but not exclusively) to the summer and autumn of 2000. But it is implicit in the judge's findings taken as a whole that that period was, in the absence of persuasive evidence to the contrary, to be taken as representative. Further, the judge dealt with and rejected Yorkshire Water's case that the plant could not perform to contractual standards at maximum flow during cold winter months. This is a single sentence synthesis of many highly complicated paragraphs of the judgment. Relevant contributing elements of the composite conclusions are in paragraphs 513 and 514 (for microthrix parvicella) and paragraphs 527 and 528 (for nitrification and more generally).
52. Ground 10 of the proposed grounds of appeal concerns fill-decants. If I have understood this correctly a fill-decant occurs if a basin of the SBR is being filled with influent at the same time as treated effluent is being decanted. The influent goes in at a low level. The effluent is decanted from the upper level. The influent stirs up the sludge and is itself only partly treated. So there is a risk or likelihood that this will affect the effluent so that it will not be as clean as it should be.
53. Some fill-decant events occurred during the earlier part of 2000. The problem was solved by introducing a fill-decant inhibit, which is a computer modification which stops fill-decants happening by temporarily diverting the influent elsewhere.
54. The judge found (paragraph 261) that the fill-decant inhibit adjustment put an additional, but insignificant, burden on the high level filters during each diversion. Subsequent monitoring by Yorkshire Water showed that there was no obvious effect (paragraph 381). The judge accepted that the issue was something of a red herring. The volume of flow actually diverted between September 2000 and November 2001 was only 1% of the total flow treated during that period. The judge accepted evidence that additional software changes could have significantly reduced the frequency of fill-decant inhibits even further. It was agreed that no harm was actually caused to the high level filters or the humus tanks. The judge accepted Dr Cannon's evidence that fill-decant inhibits had an insignificant effect on the treatment capacity of the SBR. He was satisfied that the ability of the SBR to meet the Schedule 7 performance requirements was not diminished in any material respect by the modification to the control system to provide for fill-decant inhibits (paragraph 261).
55. Yorkshire Water's essential complaint of law is that the judge did not pay sufficient attention to contractual requirements. He should have found that stopping fill-decants in this way was not a contractually compliant solution because it diverted elsewhere influent which the SBR was supposed to accommodate. If a fill-decant had ever occurred during a performance test, the SBR would have failed the test. In my judgment, permission to appeal should be refused on this ground, because, even if there were a point of law which might be arguable, it is overwhelmed by the judge's factual finding both generally and on the particular topic of fill-decants. A fill-decant would not have occurred during a performance test because the fill-decant inhibit would have stopped it. Even if the solution was not strictly in accordance with the contract, its consequences were insignificant. I am not aware – I may be wrong – that Yorkshire Water claimed any money for work to eliminate the need for fill-decant inhibits alone. It is on the facts found by the judge a tiny point.
56. Ground 16 of the grounds of appeal is under a heading "The plant's ability to nitrify". It is mainly a challenge to complicated technical findings of fact. It was Yorkshire Water's case that the SBR lacked treatment capacity for a number of reasons (see paragraph 516) of the judgment. One of these reasons was the contention the SBR was not able to nitrify at peak flows when ambient temperatures were low. In very simple terms, Dr Chambers contended that full nitrification required a minimum sludge age of 15 days which was likely to be too long to be able to control microthrix parvicella. A minimum sludge age of 12 days would, he said, be insufficient. The judge said (paragraph 523) that Yorkshire Water's case here was entirely dependent on theory. He was satisfied that a sludge age of 12 days was sufficient for nitrification (paragraph 521). Moreover, Dr Chambers' criticism was limited to various occasions when the SBR would be operating on 5 tanks and receiving 136 tcmd. It was clear from Dr Horan's evidence that the criticism was subject to the further limiting factor that the risk of ammonia failure due to insufficient react time would only arise in winter, when operating at FFT and at or about peak BOD loads (paragraph 525). Dr Chambers took too gloomy a view of the difficulties of nitrification at FFT and full loads in winter (paragraph 528). The judge accepted Dr Horan's unchallenged evidence that, provided the SBR could achieve the appropriate SSVIs, it would achieve nitrification to the required standards when operating on 5 tanks when receiving average BOD loads at FFT. He also accepted Dr Horan's evidence that the SBR could operate for about a week (although Dr Horan acknowledged that it "would struggle"), if it were to be subjected to full flows (136 tcmd) with maximum BOD load at minimum temperatures. The judge was however satisfied that the likelihood of such a conjunction of events would be so remote as to be, to all intents, non-existent (Paragraph 527). He summarised his conclusions in paragraph 529.
57. In paragraph 530 he pointed to unchallenged evidence to the effect that the decant depth could have been increased to 2.06m and that there was no hydraulic restriction to prevent this from being done. This would have resulted in significant improvements in both selector and treatment capacity, and supported the judge's conclusion that Yorkshire Water had not proved that the SBR was unable to meet the Schedule 7 performance requirements. If there had been an increase in decant depth, there was every reason to believe that the SBR would be able to meet the Schedule 7 performance requirements without difficulty, even if operated throughout on 5 tanks.
58. All this is a prelude to the point of law said to be embedded in ground 16 of the grounds of appeal. It is said that the judge ignored the fact that compliance with the stated contract requirements was essential for the purposes of passing the performance tests no matter how unlikely it was that the relevant circumstances would ever arise. If Yorkshire Water are correct here, the judge's findings relating to the plant's ability to pass the performance tests are wrong and have to be reviewed.
59. In my view, this is scarcely a point of law at all. If it is, it is forlorn. The question was whether Yorkshire Water had established that the SBR would not have passed the performance tests. The judge found in fact that they had not established this. A possibility, based on a theory whose main premise the judge rejected, that the plant lacked the capacity for nitrification was so remote as to be virtually non-existent. That was in effect a finding of fact that the plant would in this respect have passed the performance tests. I would refuse permission to appeal on this supposed point of law.

Points of fact - Introduction

60. The remaining grounds of appeal all seek to challenge comprehensively the judge's findings of fact. I have explained what I consider to be the court's proper approach to this on an application for permission to appeal. In my view, Yorkshire Water do not begin to measure up to the very exacting requirements in this case for permission to be granted. They wish to be permitted to re-open many, perhaps most, of the judge's main findings, when the judge himself has given entirely convincing fully detailed reasons for rejecting their case and preferring Taylor Woodrow's and Biwater's.
61. A brief sketch of the history of the later stages of the contract and its aftermath and an outline consideration of the central issues will suffice to explain my conclusion that Yorkshire Water should not be given permission to appeal on factual matters.
62. What follows should be seen in the context that:

i) it was Yorkshire Water's main case that the SBR was defective because it lacked selector capacity and treatment capacity and that a main problem was poor settleability;

ii) the judge accepted submissions that the case was almost entirely concerned with Yorkshire Water's contractual entitlement to recover its relevant remedial costs pursuant to the material provisions of Schedule 10 of the main contract (paragraph 86);

iii) Yorkshire Water do not seek to appeal the judge's finding that it was for them to prove that the plant would not have passed the performance tests if they had been taken (paragraph 51 of Yorkshire Water's Further Skeleton Argument in this court); and

iv) Yorkshire Water wish to be permitted to contend in this court that the judge's central finding in relation to the plant's ability to pass the performance tests was wrong and should be reviewed (paragraph 47 of the same skeleton).

Outline of the judge's finding of fact

63. As I have said, Yorkshire Water took the plant over on 26^th November 1999. The judge considered evidence about the SBR's performance from takeover to January 2000 in paragraphs 231 to 236 of his judgment. He summarised his conclusions in paragraph 257. There was no doubt that the SBR's performance deteriorated rapidly after Yorkshire Water became responsible for its operation. There was a number of reasons for this, fairly and accurately summarised in paragraph 339 of Biwater's closing submission, in terms which the judge quoted at some length in 7 numbered paragraphs. These included mechanical failures, inexperience in operation, and the fact that MLSS levels were allowed to rise beyond those maintained at takeover. The judge agreed that the matters referred to provided ample explanation for the problem then experienced immediately after takeover. The evidence suggested that microthrix parvicella had not yet become a major factor in any sludge settlement problem, although there was some indication that it might become such a problem in the future.
64. The judge considered the period from February to April 2000 in paragraphs 259 to 293 of his judgment, with a summary in paragraph 294. It was during this period that fill-decant inhibit software was installed and commissioned. The presence of microthrix parvicella had started to become a factor causing a certain amount of poor settleability, but it did not appear to have definitely become a problem until much later, the winter of 2001/2. The SBR lost nitrification completely from 25^th January to 17^th May 2000. However, a strategic decision was taken to sort out the continuing problem of solids carry over, such that it would not be fair to attach too much significance to the loss of nitrification during this period.
65. On 14^th March 2000 Earth-Tech served a defect notice under 36.1 of the General Conditions. Yorkshire Water had come firmly to the view that the SBR had deficiencies. Biwater did not consider that there was any design problem with the SBR, but that it was simply a question of getting it balanced and operated correctly. The judge was satisfied that, broadly speaking, events proved Biwater to be correct (paragraph 277).
66. There was a workshop on 22^nd March 2000, which set an agenda for the next few weeks. On 3^rd April 2000, Earth-Tech wrote to Taylor Woodrow clarifying and supplementing its letter of 14^th March 2000. By early April 2000 the real issue had become settleability.
67. In summary (paragraph 294), the judge said of this period that some short lived improvement in the solids carry over had been achieved at the expense of nitrification. By the end of March 2000, there were increasing incidents of high TSS concentrations in the effluent, and for the first time there was a problem with "an abundance of filaments in the mixed liquors". The judge accepted Biwater's closing submission that:

"… microthrix may well have been affecting settleability during March and April. However, there were other factors involved as well: MLSS levels were running at between 2000-2300 mg/l; there were high levels of mechanical breakdowns; the SBR was still operating on 5 tanks; and further control set point changes were made (C1/121) all of which might well have contributed to settleability problems".

68. In May 2000, an urgent Recovery Plan was put in hand. Biwater's Dr Zuber took the lead in addressing the various performance issues. Broadly speaking, although there continued to be a number of problems, nitrification was regained by the end of May and the SBR produced good quality effluent during the summer and autumn of 2000 (paragraph 297). There was no doubt that the BOD load received in the year following takeover was low (paragraphs 299, 300). As the winter of 2000/2001 approached, the performance of the SBR deteriorated for reasons which the judge considered later in his judgment (paragraph 300). There was a design review by Elga/Jet Tech in the summer of 2000. There was no suggestion in it that the SBR suffered from any microthrix parvicella problem (paragraph 306).
69. The relatively successful operation of the plant during the summer of 2000 was achieved as a result of the 2000 Recovery Plan, which the judge considered in paragraphs 307ff. Its two principal features were increasing aeration and replacing a strategy of reducing the MLSS with one of increasing sludge age.
70. By early June 2000, the SBR was producing excellent quality effluent (paragraph 315). The next stage of the recovery strategy was to try to identify what was or had been causing the various operational problems with the plant (paragraph 317). The general thrust of advice given to Biwater by Mr Norcross and a USF/Jet Tech team at a meeting on the 4^th May included that the nature of the influent that was being treated by the SBR was significantly different from that which could have been anticipated from the information made available to tenderers (paragraph 319). It seemed to Biwater that the various operational difficulties had not been due to any deficiencies in the design of the SBR.
71. In paragraphs 322 to 345 of the judgment, the judge considered highly technical evidence and opinion under the heading "The "BOD Disguised as COD" and the COD: BOD Ratio". In offensively simple terms (which may be inaccurate), I understand the thrust of all this to be approximately as follows. There was (a) surprisingly low Dissolved Oxygen; (b) surprisingly high sludge yield; and (c) unexpectedly high TSS loads, characterised as "absolutely unheard of", and, in the case of the daily loads of about 26 tonnes in July 2000 "an unprecedently large number". This volume of influent TSS resulted microbiologically in low Dissolved Oxygen and high sludge yield, because the micro-organisms had such a feast of TSS that they used up the oxygen, and sank exhausted and in great numbers to the bottom.
72. The relevant experts all agreed that there were only three possible routes of entry for the high TSS loads. Both Dr Horan and Mr Eikelboom, discounting the other two, considered it likely that the high TSS loads were being added to the high level stream either above or below the primary tanks from other parts of the Knostrop works before becoming part of the SBR influent (paragraph 342). The judge was satisfied that Dr Horan and Mr Eikelboom were correct here. Despite Dr Chambers' insistence to the contrary, it was clear from the mass balance calculations set out in pages 4 to 6 of exhibit TW6 that it was simply impossible that solids' loads of the magnitudes recorded in the SBR influent could have originated from the sludge return liquors. The judge also accepted the force of the various other points made in paragraph 599(2) of Biwater's closing submissions (paragraph 343).
73. In the judge's view, the evidence clearly showed that the high TSS loads did not originate from the internal return sludge liquors, as suggested by Dr Chambers. By a process of elimination, the source must have been somewhere in the high level stream, although it was not possible to say precisely where. However, the fact that the precise source of the TSS loads could not be identified did not matter. It was sufficient to find, as the judge did, that the unexpectedly high TSS loads were not caused by any aspect of the SBR's design or of its operational system of control (paragraph 344).
74. On 10^th July 2000, the parties agreed to obtain an opinion from an independent party, Mr Boon, as to whether the influent quality was different from that which could have been expected at tender stage. He eventually reported on 25^th September 2000 (paragraph 346).
75. The judge summarised the SBR's performance in the period from May to August 2000 in paragraph 347ff. It came back under control in May and early June 2000 and was producing good quality effluent. It still had not been stabilised. But graphs annexed to Dr Horan's first report showed that there were no consent failures for ammonia from mid-June 2000 until towards the end of January 2001, and only 4 suspended solids failures from the beginning of June to mid-December 2000. There was a similar picture for effluent COD and BOD respectively (paragraph 348). Consideration came to be given to increasing flows and moving towards performance tests (paragraph 349).
76. Having considered questions relating to the high level and low level filters during the summer of 2000, the judge then proceeded to consider the period from August to November 2000. Notwithstanding the SBR's improved performance, by the end of August 2000, Yorkshire Water were considering terminating Taylor Woodrow's contract and appear to have decided in principle to do so. Their preferred strategy was to take control of the Knostrop plant as a whole (paragraph 366).
77. Meanwhile Mr Boon reported on 25^th September 2000. One of his conclusions was that there were significant and unexpected increases in the COD: BOD ratio indicating that the sewage was more difficult to treat than had been initially expected. It was probable that the high level feed contained a substance or substances that were non-biodegradable or that inhibited biochemical oxidation. He recommended that further tests should be undertaken (paragraph 368). A further investigation seemed to have established that there was an excessive sludge yield and that there were unresolved concerns about the treatability of the influent (paragraph 369). There was to be a Recovery Plan 2 involving investigations by Earth-Tech. Ms Henderson returned to be involved. She made a presentation which "did not withstand the critical analysis to which it was subjected by Mr Streatfeild-James' cross-examination". The judge accepted that, when the correct figures were used to repeat her calculations, the SBR was shown to have adequate capacity (paragraph 375). "Inevitably, to the extent that Recovery Plan 2 was based on and/or incorporated the calculations and conclusions of Ms Henderson's presentation, its calculation and/or conclusions were by parity of reasoning erroneous" (paragraph 376).
78. Between August and November 2000, the SBR continued to perform well and produce good quality effluent. Yorkshire Water were trying to stabilise it upon an erroneously analysed basis before their planned termination of the main contract (paragraph 380). The main concerns during the period were how to stabilise the MLSS (paragraph 381). Biwater were asked to state what set points should be used for sludge wastage. After some initial problems, their system appeared to have worked well. It took about an hour every day to carry out and its success was compelling evidence that the SBR could be operated satisfactorily if its flexible control system was managed correctly (paragraph 388). Yorkshire Water's decision to terminate the main contract coloured their overall approach. The judge was satisfied, for reasons given later in his judgment, that the SBR was perfectly capable of meeting the contractual performance requirements if operated properly.
79. By the autumn of 2000, Taylor Woodrow/Biwater were quite prepared for the performance tests to be carried out. The judge considered that there was every reason to believe that it would have passed those tests satisfactorily had they actually been carried out. In the event, it was Yorkshire Water (the party that had the contractual obligation to carry out the performance tests) that decided to hold back from doing so (paragraph 385). All parties were nevertheless making genuine efforts to optimise the SBR's performance with a view to increasing the flows and loads (paragraph 386).
80. On 2^nd October 2000 Earth-Tech wrote to Taylor Woodrow saying that it was considered inappropriate to certify further payments. Taylor Woodrow's and Biwater's reply to this included their opinion that the contract guarantees related to a wholly different influent from that which the plant was then receiving and that therefore the guarantees were not applicable to the then current situation (paragraph 390). In the result, no further payment was made, and, on 23^rd November 2000, Biwater gave notice to Taylor Woodrow suspending the execution of the works. This brought to an end Biwater's active involvement in the operation of the SBR (paragraph 397).
81. The judge's summary of the SBR's performance from August to November 2000 accepted Mr Streatfeild-James' submission that the general picture during the period was one of improvement and good performance. There were 6 important additional features given in paragraph 398 of the judgment. These included the absence of any direct evidence as to the presence of microthrix parvicella or its effect on the SBR's performance during this period. The judge then said at paragraph 399:

"Once more I find myself in agreement with Mr Streatfeild-James's analysis of the position during August to November 2000 (see paragraphs 357 to 358 of Biwater's closing submissions). As he pointed out, not only was the SBR's performance good throughout the period but, once Mr Molloy was in control of the sludge wastage regime, the SBR's apparent lack of stability in August was rapidly remedied and stability achieved. As a result, the SBR was now operating well, despite being affected by the dramatic variations in the TSS loads, the low BOD loads and the high COD:BOD ratio, all of which had the effect of making the SBR difficult to operate. However, as Mr Streatfeild-James observed, "these various difficulties could be – and were – being overcome" (paragraph 358 of Biwater's closing observations). Furthermore (and importantly), although it is clear that Taylor Woodrow/Biwater were perfectly prepared for performance testing to be carried out on the SBR from about late September 2000 onwards, it was Yorkshire Water that decided not to do so."

82. On 24^th November 2000, Yorkshire Water wrote to Taylor Woodrow giving notice under Clause 36.5 of the Contract requiring the defect notified in Earth-Tech's letter of 14^th March 2000 to be made good urgently. In the event that the judge found, as Taylor Woodrow contended at the time, that the notified defect was not a defect, Taylor Woodrow's failure to respond to this to Yorkshire Water's satisfaction provides no basis for a claim.
83. From about November 2000, Taylor Woodrow and Biwater ceased to be actively concerned with the SBR's operation. Yorkshire Water took control. There were problems. A significant factor in the SBR's rapid deterioration in performance in November 2000 was precipitated by a serious error made with regard to the anoxic multiplier set point (paragraph 405). This, with the inevitable time that it took to recover and the fact that winter was starting, caused a decline in the performance of the SBR. In addition, there was a lack of a real and complete understanding of the SBR system and, in particular, of the appropriate changes that needed to be made to the set points from time to time on the part of the various concerned Yorkshire Water/Earth-Tech operatives at the time. It was important to note that there was no evidence (nor was there any suggestion at the time) that any of the operational problems during the winter of 2000/2001 were due to poor sludge settleability caused by microthrix parvicella (paragraph 407).
84. Earth-Tech carried out investigations. These were still predicated on a basis mainly derived from Ms Henderson's erroneous presentation. There was a conclusion, in December 2000, that the SBR lacked adequate selector capacity. This was specifically linked to the absence of an IDSC (paragraphs 408, 409). Earth-Tech produced a draft report recommending remedial works. It was decided that the way forward was to adopt a large external selector tank to overcome the poor sludge settleability (paragraph 412). The original proposed size of this was almost at once regarded as insufficient (paragraph 416). Mr Maw recommended going through the design from the beginning, but was told not to do so. His recommended exercise did not take place. The judge accepted that the result was unfortunate, in that the external selector was designed without Yorkshire Water knowing whether there were a lot of filaments in the SBR, and, if there were, what species those filaments comprised (paragraph 421). Meanwhile, during the summer of 2001, in general terms the SBR once more performed satisfactorily (paragraph 425).
85. By 10^th December 2001, phase 1 of the remedial works was completed and the SBR recommissioned with its new external selector tank in operation. As the judge pointed out in paragraph 431 of his judgment, the SBR was now markedly different from that originally designed and installed by Taylor Woodrow and Biwater. Having regard to the extent and nature of the modifications, it could not be assumed or inferred that the SBR's poor performance after recommissioning was attributable to any defect or design deficiency in the original SBR, nor did it demonstrate that the original SBR would have been unable to meet the Schedule 7 performance requirements (paragraph 432).
86. Within a month of its recommissioning, it became apparent (despite its external selector tank) that the new SBR was suffering from serious problems caused by the presence of large numbers of microthrix parvicella (paragraph 433). For reasons which the judge explained, the phase 1 remedial works had apparently resulted in a significant reduction of the flow that could be taken (paragraph 436).
87. There was further remedial work (phase 2 (part 1)). The modified SBR process was recommissioned during November/December 2002.

The judge's analysis of the evidence

88. In paragraphs 453ff, the judge considered the performance and capacity of the SBR as originally designed and installed, emphasising that the case was almost entirely concerned with Yorkshire Water's contractual entitlement to recover its relevant remedial costs under the provisions of Schedule 10 of the main contract. It was not enough to show that at certain periods the SBR performed poorly in terms of its effluent, because this could be due to one or more of a wide variety of causes. The reasons for the poor performance had to be identified (paragraph 454). Yorkshire Water's "unanswerable point" was not as persuasive as it at first appeared.
89. In no doubt over-simplified terms, the so-called "unanswerable point" is as follows. The SBR did not perform properly before Yorkshire Water's remedial works. It did (eventually) perform properly after the remedial works. Therefore its poor performance before the remedial works must have been attributable to matters for which Taylor Woodrow were responsible under their design and performance contract.
90. I find this point no more persuasive than did the judge. Taylor Woodrow's no doubt over-simplified answer is: the SBR was capable of performing, and did in fact perform, quite properly during the summer and autumn of 2000 before Yorkshire Water stepped in and themselves caused problems. The SBR did not work properly for a considerable time after the Yorkshire Water remedial works. In particular, there was a major problem with microthrix parvicella which had not been encountered before Yorkshire Water stepped in. But importantly, the modified SBR was not the same plant as that Taylor Woodrow and Biwater had supplied, nor was its performance comparable. This made comparisons uninformative. As the judge put it in paragraphs 455 and 456:

"Although it can be argued that, in the light of the modified SBR's recent performance history (in particular the results of the "performance tests" carried out in February/March 2003), "full compliance" in terms of effluent quality can be and is now being achieved all the year round, this has only been shown to be the case in relation to the treatment of a differently constituted influent at reduced flows by a significantly modified SBR (in effect, a CASS with an external selector tank).

The matter only has to be stated thus to be apparent that it does not follow, from the fact that the modified SBR is performing satisfactorily with regard to a different influent at reduced flows, that the original SBR would have been unable to meet the relevant performance requirements, because it performed poorly in the winters of 2000 and 2001."

91. The judge proceeded with what he regarded as the essential task of identifying the cause or causes of the SBR's poor performance during the two winters. He gave, in paragraph 457, an extended summary of his general factual conclusions. I shall not set it out in full, but it repays full consideration. It included a restatement of his reasoned conclusion that, if performance tests had been carried out, he was satisfied that there was every reason to believe that the SBR would have passed them satisfactorily. It also included that there was no evidence of significant numbers of microthrix parvicella being present during the winter 2000/2001. He concluded (paragraph 458) that the two main matters to be considered in relation to the original operational difficulties were (i) the difficulty experienced in achieving and maintaining the design/target MLSS levels and (ii) the SBR's selector capability together with the related question whether microthrix parvicella was the cause of sludge settleability problems and, if so, to what extent.
92. As to the control of the MLSS, the judge accepted Mr Streatfeild-James' submission (Biwater's closing submissions paragraphs 383 to 387) that the high sludge yields were obviously caused by the high and erratic TSS loads that were received by the SBR during 2000, which exceeded all expectation and for which Taylor Woodrow/Biwater were not responsible. He was satisfied that it was the unexpectedly high TSS loads that caused the high sludge yields, which in turn made sludge wastage and the control of MLSS levels much more difficult (paragraphs 462, 463). The apparent phenomenon of inter-basin transfer of MLSS was not the result of any defect or any design deficiency in the SBR (paragraph 464). The judge expressed his summary conclusions in paragraph 465.
93. As to selector capacity and microthrix parvicella, a selector is concerned to provide competitive advantage for floc-forming microbes over filamentous species. Microthrix parvicella is a filamentous species, but is not selectable. Selector design is not an exact science. The judge accepted that where, as here, there was operational experience of a selector, the best evidence of its effectiveness was an analysis of whether the desired consequences of selection actually came about. He accepted that a calculation by Mr Eikelboom was sound in principle and that it indicated that the SBR did have adequate selector capacity. The microbiological evidence during the operational life of the SBR before the remedial works was sparse, and the design and installation of the external selector in the phase 1 remedial works went ahead without any consideration of microbiology (paragraph 476). The judge accepted that, to the extent that there was any microbiological evidence, it strongly supported the suggestion (and the theoretical calculation) that the original SBR did not lack selector capacity. Stated shortly, the existing microbiological evidence showed the following:

i) there was no evidence of any significant quantities of selectable filamentous micro-organisms in the MLSS during 2000;

ii) the only filamentous micro-organisms of which there is evidence are microthrix parvicella, which is not selectable; and

iii) there was evidence of floc-forming bacteria in the MLSS (paragraph 478).

94. The judge was accordingly satisfied that there was no evidence of the presence of selectable filaments during this period and good reason to consider that they were not present. In his view, the SBR's operational history during 2000 strongly suggested that it did have adequate selector capacity. He rejected submissions on behalf of Yorkshire Water to the contrary effect.
95. The judge then considered the issue relating to the IDSC, to which I have already referred. In short, his view was that the omission of the IDSC did not prejudice the SBR's selector capacity (paragraph 488).
96. He then considered the "Average Flow" issue, concluding that, all in all, the suggestion that the use of 1 x DWF instead of 1.3 x DWF for average flow resulted in a loss of selector capacity, which in turn caused poor sludge settleability, was simply not sustainable. There was no evidence of the presence of selectable filaments. The average flows actually experienced were below the design average flow of DWF and, if the problem had arisen, it could have been addressed by appropriate adjustment of the set points.
97. The judge then considered the issue of the ASP effluent. He was satisfied that there was nothing in the point which Yorkshire Water made. The fact was that the SBR did successfully treat influent that included the ASP effluent for significant periods of time from late 1999 until late 2001. He accepted Dr Horan's evidence on this aspect of the matter in preference to that of Dr Chambers and Dr Johnson, for reasons which he gave (paragraphs 494 to 496).
98. As to microthrix parvicella, it was clear that it was present in sufficient numbers to cause some moderate settlement problems in the early part of 2000. However, the presence of microthrix parvicella in activated sludge plants does not mean necessarily that there will be serious sludge settlement problems. It is only problematic when its presence increases beyond a certain critical mass (paragraph 497). The performance of the SBR was very good during the summer and autumn of 2000, which clearly indicated that microthrix parvicella was not present in sufficient numbers to cause any settlement problems during that period. Furthermore, the winter of 2000/2001 and the greater part of 2001 passed without Yorkshire Water identifying any filaments problem or seeing the need to carry out microbiological investigation (paragraph 498). Accordingly there was no real evidence that microthrix parvicella was present in sufficient numbers to cause problems with sludge settlement until the SBR had been recommissioned with its new external selector in December 2001, shortly after which microthrix parvicella was found to be present in large numbers. The judge considered that this proliferation of microthrix parvicella was largely due to the installation of the new external selector. He referred to the evidence of Mr Eikelboom. This included Mr Eikelboom's view that two particular problems at Knostrop favoured the proliferation of microthrix parvicella;

(1) the nature of the influent (high TSL and low BOD) and

(2) the anoxic fill.

The judge then dealt with the question of nitrification and sludge age and the relevance of microthrix parvicella to that question especially during cold weather, to which I have already referred.

99. The judge then concluded in paragraph 514:

"I have therefore come to the following conclusions with regard to the issues relating to microthrix parvicella;

(i) microthrix parvicella was present, but not dominant, for a period of time in 2000;

(ii) microthrix parvicella did not become dominant until the end of 2001, after completion of the Phase 1 remedial works;

(iii) the primary reason for its dominance in late 2001 was the installation of the external selector, which favoured the growth of microthrix parvicella;

(iv) the evidence about the presence of microthrix parvicella during the operational life of the original SBR is unclear;

(v) the unexpectedly high TSS loads and low BOD loads in the influent (and later the external selector) all contributed to the development of microthrix parvicella in the SBR; and

(vi) problems caused by a proliferation of microthrix parvicella were not inevitable, because the presence of microthrix parvicella in the SBR could have been controlled by the use of a suitable sludge age/F/M ratio that could have been achieved and maintained, particularly if the influent had been within the limits of the "likely variations in flow and loads" (see Clause 0.2.6 of the Specification) which, for the reasons given above, I am satisfied it was not."

100. The judge finally in paragraphs 515 to 530 of his judgment dealt with the issue of treatment capacity. I have already dealt sufficiently with the main lines of this in considering Ground 16 of the grounds of appeal.
101. The judge then expressed his conclusions with regard to the SBR's performance and capabilities in paragraphs 531 to 533 of his judgment as follows:

"The operational difficulties experienced in 2000 were caused by a number of matters that had differing effects over time. The initial problems with set points, mechanical failures, issues over the operation of the SBR and so forth were eventually replaced by a period of conspicuously good operation and stability, where the only matter of concern was MLSS control. MLSS levels were difficult to control because of the unexpectedly high TSS loads, which persisted throughout 2000 and for which Taylor Woodrow/Biwater were not responsible. However, these control issues were perfectly capable of being addressed and were successfully addressed by Biwater in September and October 2000.

Microthrix parvicella may have been present throughout 2000. To the extent that it was present it was not there in any significant quantity, although its presence would have added to the difficulties being experienced. However, I am satisfied that microthrix parvicella could have been controlled at acceptable levels by use of a suitable sludge age/F/M ratio. Such a sludge age could have been maintained throughout the year without jeopardising nitrification. I am satisfied that there is nothing in the performance of the SBR as originally designed and built to support the suggestion that the SBR was fundamentally defective or lacked adequate selector capacity. I accept Mr Streatfeild-James' submission that Dr Zuber's assessment of the operational difficulties was entirely accurate and correct (D3A/341-343): i.e. the SBR was not defective, but was difficult to operate because of the varying and low BOD loads, the high and variable TSS, the high COD:BOD ratio and the high sludge yields (Biwater's closing submissions, paragraph 442). Unfortunately, as Mr Streatfeild-James went on to observe, Yorkshire Water had decided to determine the contract in the belief that there were fundamental defects with the SBR and that led to remedial works being undertaken that were not, as it seems to me, actually required to remedy any defect or deficiency in the design of the SBR.

I am satisfied that, whether by reference to its operational history or to appropriate theoretical calculation, the SBR as originally designed and built did not lack either adequate selector or treatment capacity. Had Yorkshire Water proceeded to performance testing in the autumn of 2000 (as Biwater was anxious to do) and had the SBR been operated properly, there is every reason to believe that it would have passed the performance tests in question, particularly if the influent had been within the limits of "the likely variations in flows and loads", which had not been the case for much of 2000. Furthermore, the prospect of a successful completion of the performance tests would have been even greater if the set points had been adjusted so as to increase the decant depth, a measure that could have been carried out without the need for any of the remedial works actually carried out. As it seems to me, it follows from these findings that Yorkshire Water has not proved that Taylor Woodrow was unable to meet the performance requirements of Schedule 7 of the main contract and its case, therefore, fails. By a parity of reasoning, I am satisfied that the SBR was not defective as alleged in either the 14^th March Defect Notice (see paragraph 275 above) or the 7^th April Defect Notice (see paragraph 290 above). It follows that, even if (contrary to my construction of the relevant terms of the main contract) Yorkshire Water is entitled to rely on any of the other contractual terms identified in the Particulars of Claim as well as the Schedule 10 absolute guarantees, the outcome of these proceedings remains the same."

102. I have now produced in 38 paragraphs a (no doubt inadequate) racing summary of 303 dense paragraphs (231 to 555) of the judge's judgment occupying 92 closely typed pages. It is quite obvious that his factual conclusions are not, upon the principles which I have stated, amenable to appeal to this court. That statement scarcely needs further elaboration. There is nothing glaringly wrong with the judge's findings and analysis, nor are they remotely incompetent. On the contrary, it is a highly persuasive masterpiece.

Factual grounds of appeal

103. The essentially factual grounds of appeal are grounds 8 to 19 inclusive under the headings:

8. The capability of the plant to meet the contractual performance requirements.

9. The performance of the SBR.

10. Fill decants.

11. The control of MLSS and TSS.

12. Selector capacity.

13. Lack of an IDSC.

14. Lack of treatment capacity.

15. Microthrix parvicella.

16. The plant's ability to nitrify.

17. Treatment of ASP effluent.

18. The nature of the influent.

19. No heading, but it challenges part of the judge's finding in paragraph 533 of his judgment.

104. The shape of many of these proposed grounds of appeal is to summarise what the judge found, and then say that he was wrong because he failed to take account of material circumstances (which are set out) or because he drew incorrect inferences. Generally speaking, the grounds seek to reargue in this court matters which were advanced and unsuccessfully argued before the judge, and which, upon mature consideration and for convincingly explained reasons, he rejected. The repeated proposition that the judge failed to take material circumstances into account is, again speaking generally, untenable. A startling example of this is in Ground 15 (microthrix parvicella) where the judge is said to have failed to take into account 9 material circumstances set out in paragraphs 15.7, 7 of which are expressed in the Grounds themselves as "the fact (as he himself found) ..." (my emphasis). The very grounds themselves show that the judge did take these matters into account.
105. As to allegedly mistaken inferences, this comes nowhere near a case in which this court might consider re-examining inferences on the basis that an inference is a deduction from primary facts and intrinsically more amenable to appeal. In the present case, inferences were so interlocked in the jigsaw of technical fact and opinion as to be inseparable from them. As to the facts and opinions themselves, they were supported by evidence or opinion to which the judge referred and which he was entitled to prefer to other evidence or opinion which he rejected. There are even some instances where, as Mr Streatfeild-James or Mr Elliott submitted, the proposed appeal now seeks to challenge findings which were based on evidence which was not challenged before the judge.
106. The proposed factual grounds of appeal are supported by two long written submissions. Mr Sears in his oral submissions adopted a realistically economical approach. He said that paragraphs 2 and 3 of the second written submissions were a route map for consideration of the proposed appeal. Paragraph 2 summarises the significance of Yorkshire Water's points of law, which I have already considered. Paragraph 3 contends that Yorkshire Water have a real prospect of persuading the Court of Appeal that the judge's central finding that the plant would have passed the Schedule 10 performance tests in the year after take over was factually wrong or against what is said to be the overwhelming weight of the evidence. The summary reasons given for this contention are:

"3.1. The evidence shows that, as a matter of fact, the Plant would not have passed the Performance Tests if they had all been taken in the year following take-over.

3.2. The evidence shows that the Plant would not have passed the Performance Tests without the implementation of the fill-decant inhibit (which was not a contractually-compliant solution).

3.3. The evidence shows that, as a matter of fact, the Plant was quite incapable of operating successfully on 5 basins.

3.4. On a proper analysis, and contrary to the findings made by the Judge, it is quite clear that high levels of TSS and/or a low BOD load were not the cause of the main problem experienced in operating the SBR in 2000.

3.5. In any event, neither factor (i.e. high TSS levels or low BOD load) would excuse the Respondents from responsibility for the poor performance of the Plant.

3.6. The experts agreed that the main problem was poor settleability and that the most likely cause was the presence of significant numbers of Microthrix parvicella which, similarly, would not excuse the Respondents from responsibility for the Plant's poor performance.

3.7. The evidence shows that, once Yorkshire Water had carried out its remedial works:

3.7.1. the Plant was able to satisfy the performance requirements despite the presence of significant numbers of Microthrix parvicella; and

3.7.2. the Plant in its present state (i.e. after remedial works) suffers from none of the settleability problems which existed in the Plant as originally designed and built (i.e. the so-called 'unanswerable point').

3.8. For the reasons explained below, the Judge was unable to provide a satisfactory answer to the 'unanswerable point' despite his attempts to do so."

I have already sufficiently considered the fill decant inhibit point, the question of operation on 5 basins, and the so called "unanswerable point".

107. Mr Sears made oral submissions with reference to the findings about high TSS and low BOD loads and microthrix parvicella. The structure of the argument, in simple terms, was that the judge was wrong to conclude that poor settleability in 2000 resulted from high TSS and low BOD loads for which Taylor Woodrow and Biwater were not responsible. On the contrary, the problem, admittedly with hindsight, was the presence of microthrix parvicella. The respondents point out, with justification, I think, that the present emphasis on microthrix parvicella as the problem in 2000 is very different from Yorkshire Water's case at trial. However that may be, the judge found, after exhaustive consideration, that microthrix was not a significant problem until after Yorkshire Water's ill-conceived remedial works and that the problem then was caused by the external selector.
108. As to passing the performance test generally, Mr Sears points to paragraph 67 of the first written submission which relates that in their first (pre-trial) joint written statement some of the experts had, he said, agreed that, if performance tests had been undertaken, the plant would not have passed all the tests. This joint statement was not, however, the entirety of the evidence, or even the main evidence, which the judge eventually heard and considered, and Dr Johnstone, one of those who joined in the first joint statement, was not called to give evidence. The judge was obliged to consider and entitled to rely on the evidence which was called, including that of Dr Horan.
109. As to the high TSS and low BOD loads, Mr Sears points to extended passages in the first written submission, which I have read. He submitted orally that the judge was wrong to find that there was anything unexpected or unusual about the TSS. He pointed to a pre-contract schedule in support of this. Mr Elliott showed convincingly that this was not to be regarded as a basis for the SBR's design. Mr Sears more or less let this point go in his reply. Further, although there may have been some evidence and argument to support the case advanced by Mr Sears, there was equally evidence and argument, to which the judge referred, quite capable of supporting his conclusion.
110. Mr Sears submitted that the recorded high levels of TSS did not coincide with the times when the plant performed poorly and that the worst performances were at times when the TSS levels were not high. He submitted that there was a real prospect of persuading the court that the cause of poor settleability was microthrix parvicella. I am quite unpersuaded. Not only was the judge's analysis of the microthrix parvicella question convincing – essentially it was not a real problem before the Yorkshire Water remedial works and not Yorkshire Water's case at the trial – but Yorkshire Water's solution on the ground was to construct an external selector; microthrix parvicella is not selectable; and the microthrix problem arose after, and (on the judge's finding) was caused by, the remedial works. This is no doubt over-simplistic. But I fail to see how the shape of the basic facts presents a real prospect of Yorkshire Water persuading this court that the cause of the settleability problem in 2000 was microthrix parvicella. Add to this my view that the judge's analysis of the cause of high TSS and low BOD levels is convincing and was available to him on the evidence (see paragraphs 338 to 345 of the judgment) and this proposed line of appeal becomes, I fear, hopeless. In reaching this conclusion, I have taken account of and (I hope) understood the relevant parts of the two written submissions, including paragraphs 78 to 81 of the second submission to which Mr Sears specifically directed us.
111. Mr Sears submitted that the judge was wrong to accept Dr Horan's "knife-edge" solution to the nitrification/microthrix point, to which I have already referred, and the practicability of changing the decant depth. It was, said Mr Sears, a solution which evolved only at the trial. In my view, the respondents are entitled to say that it was not a knife-edge solution, but simply a matter for appropriate, skilful operation of the plant, which historically Yorkshire Water did not always achieve. It was, said Mr Elliott, a forensic fall back position for circumstances which the judge found would not occur in practice.
112. It is no criticism of Mr Sears to say that his oral submissions and my account of them only scratch the surface of the points Yorkshire Water would wish to make on an appeal. There is a great deal more material in the written submissions, which, as I say, would seek to re-open most of the judge's carefully considered findings of fact adverse to Yorkshire Water. But the more you look at it, the more it is obvious that the judge's factual findings are simply not amenable to appeal. To start with, as Mr Streatfeild-James pointed out, Yorkshire Water would have to shift all or most of the findings of fact in paragraph 457 of the judgment. They would have to shift the judge's findings about microthrix parvicella, including, as an example only, his finding in paragraph 480 that there was no evidence of selectable filaments during the relevant period and good reason to consider they were not generally present. The proposition that knowledge of microthrix parvicella was only emerging at the time does not square with the fact that it was apparently a known contemporary problem at another of Yorkshire Water's plants. Mr Streatfeild-James says that the judge's conclusion that the SBR had sufficient selector capacity was based on unchallenged or overwhelming evidence. So the first major plank of Yorkshire Water's case fell and the costs of the new external selector were not required for any selector deficiency in the SBR.
113. I have no doubt that Yorkshire Water and those concerned with their case in these proceedings felt, and may still feel, that Yorkshire Water's technical case, that the SBR was incapable of performing properly and achieving the performance tests, was both strong and correct. The judge, however, found otherwise for carefully considered and convincing reasons. Among all the welter of technical detail, an outstanding point is that the SBR, skilfully operated under Biwater's guidance, performed well, producing consistently good quality effluent, from about the end of May 2000 until Biwater ceased to be involved in about November 2000. It also performed reasonably well during much of 2001, subject to operating errors by Yorkshire Water, who had by then taken over. As to the skill needed to operate the plant, Mr Elliott made the point that it was not just a question of switching on the plant and sitting back. Yorkshire Water decided not to have performance tests when the plant was operating properly. They went down their own route and carried out wrong remedial works. There is no doubt another point of view, but the judge rejected it, being fully entitled on the evidence to do so.
114. I do not propose to go further in considering in greater detail the array of technical points of fact or opinion which Yorkshire Water would wish to raise. Justice and proportionality do not, in my view, require it. This is an application, not an appeal. Applications should not be dress rehearsals for proposed appeals, nor decisions on applications disproportionately extended.

Conclusion

115. I would dismiss this application.

Jonathan Parker LJ: I agree.

Sir Peter Gibson: I also agree.