[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 >> Baxall Securities Ltd & Anor v Sheard Walshaw Partnership & Ors [2002] EWCA Civ 9 (22 January 2002) URL: http://www.bailii.org/ew/cases/EWCA/Civ/2002/9.html

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

IN THE SUPREME COURT OF JUDICATURE
COURT OF APPEAL (CIVIL DIVISION)
ON APPEAL FROM TECHNOLOGY AND
CONSTRUCTION COURT
HH Judge Bowsher QC

		Royal Courts of Justice Strand, London, WC2A 2LL
		22/01/02

Mr Justice David Steel :

1. These proceedings concern a valley gutter on the roof of an industrial unit at Bredbury, Stockport, in Greater Manchester. The respondents to this appeal (the claimants below) were the occupiers of the unit, which was used for the storage of electrical security systems manufactured and distributed by them. The unit was divided in half, and each claimant occupied a separate part.
2. The building had been developed by Berisford Property Investments Limited ("Berisford"). Berisford had employed the appellants ("SWP") as architects. The appellants were the only active defendants below.
3. The respondents became occupiers of the unit in 1994. Their claim arose out of flooding of the unit on two occasions: first on the 29th May and second on the 4th September 1995. This was attributable to the inability of the valley gutter and its associated drainage system to cope with the heavy rainfall on those days.
4. The trial judge, His Honour Judge Bowsher QC, sitting in the Technology and Construction Court, held that the appellants were liable in tort in respect of the consequences of the second flood, but not the first. He held that the respondents were not contributorily negligent and accordingly awarded the agreed measure of loss in the sum of £612,153.02 plus interest.

The gutter

5. The building was simply a large shed, with twin-pitched roofs running length wise. These two pitches were separated at the inner eaves by a valley gutter. The valley gutter had a flat base with vertical sides. In the base of the gutter were outlets, leading to drain pipes. However, there was no seal between the lip of the gutter and the underside of the metal roof. Accordingly, if the gutter over filled, water would pour over the lip of the gutter into the building.
6. It was common ground between the experts that the valley gutter had a fundamental defect. It ought to have had, but did not have, overflows. The usual form of overflow would have been by way of wedge-shaped cuts in the top of the ends of the gutter, known as weir overflows. In the unit as built, weir overflows were impractical because the ends of the gutter were set against girders, forming part of the steel frame. In these circumstances, the alternative should have been upstanding pipes at intervals along the gutter, on the same principle as the overflow in a domestic lavatory system.
7. The purpose of overflows can be defined as follows: if the drain pipes became blocked, or if a rainstorm occurred which was heavier than that for which the drainage was designed, the overflows would take the excess water away without internal flooding of the building.
8. The respondents' case below was that, as a result of want of care on the part of the appellants, the valley gutter had two latent defects: first the absence of overflows and second a shortfall in design capacity for the rainfall to be expected in the area. The appellants' case was that there was no shortfall in design, the absence of overflows was patent and in any event both floods were attributable to blockages which should have been cleared by the claimants.
9. The trial judge found that the first flood was caused partly by blockages and partly by the absence of overflows which was a patent defect. Accordingly the appellants were not liable. As regards the second flood, the judge found that this was partly attributable to the patent absence of overflows and partly to a latent shortfall in design capacity. In those circumstances he held the appellants liable.
10. Putting the matter rather simplistically, the appellants first contend that the judge was wrong to conclude that there was any shortfall in the design of gutter. Second they say that the judge was wrong to conclude that there were two defects in the gutter, one patent and one latent. They say that there was only one defect, namely a propensity to overflow, which on any view was patent. On either basis they assert that they were not liable for the second flood.
11. There was also a cross-appeal whereby the respondents contended that the judge was wrong to find that the absence of overflows was patent. Accordingly the judge ought to have found the architects responsible for both floods.

Design history

12. The detailed design of the roof drainage was originally to be carried out by a specialist sub-contractor, FK Roofing. It was to be by way of conventional gravity drainage. However, in early 1990, concern was expressed about the potential for collection of methane underground and, as a result, arrangements were made for the installation of siphonic drainage, the design and installation of which was sub-sub contracted to Fullflow. Such systems are designed to ensure that the drainage pipes remain filled with water so that the water run-off produces a sucking effect.
13. As the judge found, the manufacturer's literature regarding siphonic drainage equally made it plain that "large capacity overflows are essential." However, the appellants certified the works as practically complete on the 17th September 1990. They also issued a making good defects certificate and a final certificate on the 28th June 1991 and the 3rd December 1992 respectively. At no stage did the appellants notice let alone comment, on the absence of overflows. It is not in issue that they were at fault in not doing so.

Rainfall intensity

14. In designing the roof drainage, FK Roofing selected a rainfall intensity of 75 mm per hour. There was no dissent, let alone instruction to the contrary, from the appellants. The judge held that, having regard to the expert evidence, there should have been. The experts (Mr Lynch retained by the appellants and Mr Woodward by the respondents) had agreed as follows:

i. The valley gutter ought to have incorporated weir overflows in accordance with BS 6367: 1983.

ii. Based upon calculations provided by FK Roofing Ltd, the adopted design criteria for the gravity rainwater system incorporated a rainfall intensity of 75 mm per hour. The building, as now occupied, ought to incorporate a rain disposal system designed for a rainfall intensity of 150 mm per hour.

History of occupation by respondents

15. Prior to executing a tenancy agreement, the respondents instructed surveyors to inspect the unit. They also instructed environmental engineers to advise on the fitting out of the building. On the 30th July 1993, Mr Meiklejohn of Castle Computer Suites Ltd, the environmental engineers, wrote to the respondents as follows:

"I spoke with your surveyor, employed to inspect the above on your behalf earlier this week. I voiced my concern regarding the state of the roof to the premises, in that the previous week whilst I was escorting prospective contractors around the site, I noted that the warehouse floor was flooded in many places with rain water.

Your surveyor explained that he had been upon the roof and noted that the valley guttering was blocked with debris, which could be the source of the problem…

My concern is that the matter is entirely settled before we commence any works on site, particularly those which entail the suspension of services from the roof structure. Otherwise it would appear reasonable to suggest later that any roof leakage problems related to the service installation works rather than a latent defect. My advice would be that you do not take on any commitments regarding the premises until this matter is resolved. To do otherwise could leave us in dispute as to who is going to pay for a new roof."

16. This, as the judge found, brought home to the respondents that there had been not just small leaks, but flooding of the building. The surveyors, Messrs. Lambert Smith Hampton ("LSH"), reported in August as follows:

"Rainwater goods: 4.1. Rainwater disposal from the main roof comprises a galvanised steel box gutter with bolted sections around the perimeter of the building and a similar central valley gutter. There is a large amount of debris and silt in the central valley gutter and this requires cleaning as soon as possible, since as we discuss later, there appears to have been problems with rain water penetration through the gutter system.

4.2 Weir outlets are provided in the event of blockage to the system....

4.4 We recommend that all the gutters are thoroughly cleaned and painted with bitumen paint. This should be undertaken on a regular basis, say every three to four years.

4.5 In view of the rainwater penetration, we recommend that the gutter joints are thoroughly checked and re-sealed if necessary.....

4.7 As previously mentioned, the gutter is particularly silted with debris and the rainwater outlets are partly blocked. We suspect that there has been a serious failure of the drainage system in the past, causing extensive flooding internally, as evidenced by the water staining on the floor. Such a system with small bore outlets is particularly vulnerable in the event of blockage and it is extremely important that the guttering is kept clear. We suggest, however, that the landlords investigate the source of the problems and undertake any remedial action that is required…"

17. As the judge observed, the reference to the provision of the weir outlets appears at first blush to be positively misleading. However, it may be that such overflows were indeed observed but on the perimeter guttering. In any event, the judge found that the absence of overflows in the valley gutter ought reasonably to have been discovered by the surveyors.
18. In fact the only suggestion made by the surveyors in a letter dated the 5^th August was that the potential landlords should agree to "investigate and repair water penetration along the line of the valley gutters.." Mr Meiklejohn adopted a similar line and in a letter dated the 6th August commented:-

"With regard to your surveyor's report, my only comment is that I know that the roof has been leaking extensively. I have discussed this matter with the surveyor after his first visit and then we were unable to decide what the precise fault might be.

I am not sure how we might resolve this matter but I am concerned that it is highlighted as a latent defect that remains the Landlord's responsibility until remedied."

19. Thereafter, a list of work to be carried out by the respondents and their landlords respectively was prepared for annexure to the agreement for lease, which was duly executed on the 10th September 1993. But, as Mr Meiklejohn spotted, there was in fact no reference in those annexures to the roof guttering. In due course, the lease itself was executed on the 18th January 1994. It did contain a provision imposing responsibility for latent defects on the landlords.
20. On the 14th September 1994, Mr Meiklejohn wrote to the respondents drawing attention to the fact that seagulls appeared to have adopted the roof, thereby causing the drain outlets to become blocked. "This only becomes evident of course when it rains. The gutter then fills up and overflows." A regular maintenance contract was accordingly recommended.
21. Some cleaning work was undertaken on an intermittent basis, but nonetheless the first flood took place on the 29^th May 1995. It was common ground that the rainfall that day was less than the design rate of 75 mm per hour. The judge concluded the explanation was that the gutter outlets were obstructed by a considerable quantity of waste material and, absent any overflow, the water escaped between the gutter lip and the roof into the building. The resultant damage to the respondent's goods stored in the unit amounted to £135,459.43.
22. Albeit even now the absence of any overflows still remained undetected, the respondents did at least introduce regular cleaning. Indeed, the valley gutter was cleaned only three days before the second flood, which occurred on the 4th September 1995. This time, the rainfall was heavier. The judge found (and it is not challenged) that it was between 75 mm and 150 mm per hour. He concluded that the flood was in part caused by rainfall in excess of the design rate and in part by the absence of overflows.
23. At last, the penny seems to have dropped with regard to overflows. Mr Meiklejohn wrote to the cleaning contractor the very next day:-

"Please cut out and form an outflow from either end of the valley gutter to the above. The purpose of this is to ensure that the rain water level in the gutter is limited to two thirds of its depth and so unable to overflow into the premises along its length. This work is to be carried out as soon as possible."

24. For reasons already explained, this proposal was impractical. Following further investigation, remedial work was carried out as described by Mr Meiklejohn in his letter of the 17th July 1997:-

"An additional dual gravity system was installed above the Baxall occupied half of the premises with its outlet raised so as to allow the existing siphonic system to operate. Over the area occupied by Norbain, a gravity system was installed immediately following the major roof leaks flooding: this was to simulate an end of gutter weir overflow and installed to the last bay at the end of the premises. Later another siphonic system was installed to supplement the existing one and again its outlets raised so as to permit the existing system to operate."

Cross Appeal

25. It is convenient to start by considering the cross appeal, which sought to challenge the judge's finding that the absence of overflows ought reasonably to have been discovered by LSH. This was not an immediately attractive submission given that no challenge was made to the finding that the architects ought to have realised that there were no overflows.
26. Of course, the architects were directly involved in the design and construction of the building. Thus, aware of the presence of the steel frame, they would not have been misled by any assumption, if such was made by LSH, that there were weir overflows that had been cut at the end of the valley gutter. LSH, in contrast, were approaching matters solely by way of visual inspection. But even without a history of flooding, no good reason has been advanced for not spotting the absence of overflows in the valley gutter. The appellants' expert gave evidence to the effect that it was reasonable for LSH to have missed the absence of the overflows because a torch light was needed to see the ends of the gutters. The judge was unimpressed by this contention, in the sense that he could not understand why a surveyor, exercising reasonable diligence, would not have done the same. LSH were obviously conscious of the need for overflow protection and it was in the valley gutter that such protection was primarily needed: any water coming over the lip of the perimeter gutter would not go into the building. In the circumstances, I agree with the judge.
27. But in any event LSH had the inestimable advantage of conducting their survey when aware that there had been flooding of the building by way of the valley gutter. The immediate explanation was an accumulation of debris in the gutter. However, as the judge observed, even a layman might have asked the question:- "Why did not the overflows prevent the flood when the primary drains were blocked?"
28. It was submitted on behalf of the respondents that the judge ought to have been very cautious before making a finding that the absence of overflows should have been ascertained by the surveyors in circumstances where:-

i. There was no evidence from LSH themselves;

ii. It was the view of Mr. Lynch, the appellants' expert, that LSH had not fallen short of the standards to be expected from a competent surveyor;

iii. The judge had not seen the building for himself.

29. In fact, the judge was at great pains to make allowance for the absence of any evidence from LSH. As for Mr Lynch's evidence, it is clear that the judge had formed a somewhat jaundiced view of it, given Mr Lynch's somewhat parti-pris comments on the use to which the architects might expect the unit to be put. Neither party invited the judge to make a site visit. It remained for the judge to determine the issue on the evidence before him. I see no reason to disturb the judge's conclusion.
30. Furthermore it is clear that, after the initial survey, LSH carried out subsequent inspections to check progress on the elimination of other leaks in the roof. On the 14^th January 1994, LSH wrote to the claimants informing them that they would thereafter assess progress by reference to performance of the gutter rather than by inspection. In May, the first flood occurred. Still, the penny did not drop. The mere fact the gutters were obstructed did not explain why the excess water was not carried away by the overflow arrangement. No argument was ever presented to the effect that, at this later stage, it remained reasonable that the absence of overflows remained undiscovered.
31. I find, in accord with the judge, that the defect in the form of the absence of overflows could and would have been discovered by the surveyors acting for the respondents if they had exercised reasonable care prior to the lease and, even more emphatically, after the first flood. In my judgment, the cross appeal thus fails.

Shortfall in design.

32. I turn now to the appeal. It remains common ground that any underdesign of the drainage system was not something that ought reasonably to have been discovered by LSH or indeed the respondents. However, the appellants sought to challenge the judge's finding that the design capacity of the siphonic drainage system was in fact inadequate. The thrust of this submission was not clearly foreshadowed in the notice of appeal but has been fully argued.
33. The basis of this challenge was the submission that, whilst the design rate of 150 mm per hour was appropriate, as agreed by the experts, such could and would have been properly afforded by the provision of overflows. Put another way, it was submitted, as I understand it, that the design rate should be viewed overall, and if the gutter, downpipes and overflows taken together had the capacity to cope with 150 mm per hour, that was an adequate design.
34. Although there is no finding on the matter, I am readily disposed to accept for the purposes of the argument that the gutter, inclusive of the overflow, was capable of coping with 150 mm per hour (or, at the very least, the rain rate at the time of the flood). But was this adequate as a design?
35. Two particular aspects of the evidence were relied upon in support of the appellants' case on this topic. First it was submitted that Mr Woodward, the respondents expert, had accepted in his oral evidence that a design based on 75 mm per hour was not open to criticism so long as adequate provision was made for overflows. The principal exchange relied upon was as follows:-

Q: So, can we agree on this? Had alternative means of taking the excess been provided, you would not regard the selection or adoption of 75 mm design rate as being unreasonable?

A: The 75 mm design rate would then only have applied to the initial siphonic system, which providing there were back up systems in place, would not have been unreasonable.....In other words, you can design to 75 mm rainfall intensity with a siphonic system, which will operate just as itself under normal [conditions]....Bear in mind we are on a site which I believe everyone acknowledges at 75mm you are going to have a storm which is in excess of that every once a year. Whatever you design to, you are going to have to accommodate that type of storm once every twelve months, so you would need an alternative system in place. The siphonic people have come up with a siphonic overflow system. In other words, they will guide you to a relatively low initial intensity, but they will then recommend that you have alternative provisions in place to cater for the excess....

36. In that passage, Mr Woodward is making reference to the second aspect of the evidence upon which the appellants relied, namely the remedial steps taken after the second flood. Mr Meiklejohn's letter of the 23rd April 1996, describes the remedial works carried out, including the provision of a siphonic overflow as specified in a quotation by Messrs Sapoflow dated the 20^th November 1995. The impact of the provision of this siphonic overflow system was described by Mr Woodward in his oral evidence as follows:-

"....In other words, at the time that quotation is being put forward, there is already a siphonic system in place designed to cater for 75 mm of rainfall intensity. This is a system going in over and above. They have worked on a rainfall intensity of 100 mm. This is an overflow system which will not cut in until the 75mm is exceeded. It is then giving you a capacity of a further 100 mm. Therefore the true design is 175 mm per hour, if I've read those documents correctly."

37. It follows, so the argument runs, that the design of the valley gutter is properly to be treated as a balancing exercise between the inherent capacity of the gutter and its drainage on the one hand and the overflow system on the other. Accordingly, there was, it is contended, no defect in the design. Whilst the experts are agreed that the rainwater disposal system should be designed for an intensity of 150 mm per hour, the design of the gutter inclusive of the overflow met that standard.
38. I have no hesitation in rejecting this novel argument. The assessment of the appropriate design rate has to be made without regard to overflow capacity. As the British Standard, referred to in this judgment, makes it plain, the weir overflows are to allow for flow "in excess of the design rate". Mr Woodward's written statement to that effect was not challenged nor could it have been. Indeed, the formulation of the agreement between the experts is only consistent with that proposition.
39. The passage in Mr Woodward's oral evidence is entirely consistent. It merely demonstrates that sequential arrangements for drainage can be furnished to make good the proper design rate without in any way detracting for the need for overflows. As regards to the Sapoflow quote, this in terms stated:- "Roof/gutter must have sufficient holding capacity, and overflows, to deal with extraordinary conditions which may exceed the design capacity of the system." Indeed, the remedial works were not confined to the installation of supplementary siphonic overflows. They included the installation of gravity overflows, inserted at the extreme end of the gutter "to simulate weir conditions, raised 50 mm above the siphonic overflow outlets, so as to act as standby": see Mr Meiklejohn's letter of the 23rd April 1996.
40. In my judgment, the judge was quite correct to treat the absence of overflows as a separate matter from that of design rate. Nonetheless as I will develop later in this judgment, the absence of overflows remains of great causative significance when having regard to the relevance of the failure to design the gutter to cope with up to 150 mm per hour.

Knowledge of propensity to flood

41. It was argued below, and tentatively suggested in the notice of appeal, that it was wrong in law to impose a duty of care on architects (in contrast to builders) owed to subsequent purchasers of property in regard to a latent defect. This proposition has not been urged upon us.
42. The appellants' alternative argument, as I understood it, was that a defect was patent (i.e. not latent) if the dangerous propensity is apparent, even if the actual nature of the flaw is not. In other words, it is not necessary to consider whether the respondents knew or ought to have known of the absence of overflows or the shortfall in design since they did know, so the argument runs, of the tendency of the valley gutter to overflow. Thus, it is contended, they knew of the defect. Neither at the time of the submission, nor during more mature reflection in preparing this judgment, have I found this approach convincing or even helpful.
43. The conventional starting point on the issue of the proximity required to establish a duty of care is Heaven v Pender (1883) 11 QBD 503, where, at page 510, Lord Esher said:-

"This [i.e. the rule as to proximity that he had just formulated] includes the case of goods etc. supplied to be used immediately by persons or a particular person, or one of a class of persons, where it would be obvious to the person supplying, if he thought, that the goods would in all probability be used at once by such persons before a reasonable opportunity for discovering any defect which might exist, and where the thing supplied would be of such a nature that a neglect of ordinary care or skill as to its condition or the manner of supplying it would probably cause danger to the person or property of the person for whose use it was supplied, and who was about to use it. It would exclude a case in which the goods were supplied under circumstances in which it would be a chance by whom they would be used or whether they would be used or not, or whether they would be used before there would probably be means of observing any defect, or where the goods would be of such a nature that a want of care or skill as to their condition or the manner of supplying them would probably not produce danger of injury to personal property."

44. This passage draws the distinction between the defect and the danger that it gives rise to. Lord Atkin commented on it in his speech in Donoghue v Stevenson [1932] AC 562 at 582:-

"I draw particular attention to the fact that Lord Esher emphasises the necessity of goods having to be "used immediately" and "used at once before a reasonable opportunity of inspection." This is obviously to exclude the possibility of goods having their condition altered by lapse of time and to call attention to the proximate relationship which may be too remote where inspection, even of the person using, certainly an intermediate person, may reasonably be interposed."

45. The emphasis is accordingly on defects which are latent in the sense that they could not be detected by such reasonable examination as the defendant to the claim might, if he had given any thought to it, reasonably anticipate would be conducted. This approach is confirmed by the decision of the Privy Council in Grant v Australia Knitting Mills [1936] AC 85. At page 105, Lord Wright said as follows:

"The principle of Donoghue's case can only be applied where the defect is hidden and unknown to the consumer, otherwise the directness of cause and effect is absent: the man who consumes or uses a thing which he knows to be noxious cannot complain in respect of whatever mischief follows, because it follows from his own conscious volition in choosing to incur the risk or certainty of mischance.

If the foregoing are the essential features of Donoghue's case, they are also to be found, in their Lordships' judgment, in the present case. The presence of the deleterious chemical in the parts due to negligence in manufacture, was a hidden and latent feature, just as much as were the remains of the snail in the opaque bottle: it could not be detected by any examination that could reasonably be made."

46. The concept of a latent defect is not a difficult one. It means a concealed flaw. What is a flaw? It is the actual defect in the workmanship or design, not the danger presented by the defect. (A good example of the distinction is contained in Nitrigin Eireann Teoranta v Inco Alloys Ltd [1992] 1 WLR 498.) To what extent must it be hidden? In my judgment, it must be a defect that would not be discovered following the nature of inspection that the defendant might reasonably anticipate the article would be subjected to.
47. There is, accordingly, a question here of degree. The consumer of a fizzy drink will not, in the normal course, bring in an expert to inspect the goods he purchased. In marked contrast, the buyer of a building almost invariably would. Certainly in the commercial context, a defect would not be latent if it had been reasonably discoverable by the claimant with the benefit of such skilled third party advice as he might be expected to retain.
48. The classic definition of latent defect in the field of carriage of goods by sea is that contained in Riverstone Meat Pty Ltd v Lancashire Shipping Company Ltd [1961] AC 807, per Lord Keith of Avonholm at page 872:-

"He will be protected against latent defects, in the strict sense, in work done on his ship, that is to say, defects not due to any negligent workmanship of repairers or others employed by the repairers and, as I see it, against defects making for unseaworthiness in the ship, however caused, before it became his ship, if these could not be discovered by him, or competent experts employed by him, by the exercise of due diligence." (A similar approach is adopted in the insurance field: see The Caribbean Sea [1981] Lloyds Rep 338.)

49. Returning to factual circumstances more analogous to the present, the appellants sought to bolster their argument that only an awareness of the danger rather than the defect was required by reference to Murphy v Brentwood District Council [1991] 1 AC 398. The appellants placed emphasis on certain passages in Murphy which they suggested supported the proposition that a latent defect is only to be treated as patent once it has actually been discovered. On that basis, the appellants submitted that the negativing of the duty of care could not possibly depend on actual discovery of the flaw (which would render the defence largely redundant) but simply on awareness of the danger presented by the flaw. In the present case, so the argument went on, the defect was indeed discovered, in the sense that the surveyors and their employers became aware of the propensity of the valley gutter to flood.
50. In making this suggestion, the appellants had in mind, for instance, the passage in the judgment of Lord Keith of Kinkel, at page 464:-

"However, an essential feature of the species of liability in negligence established by Donoghue v Stevenson was that the carelessly manufactured product should be intended to reach the consumer in the same state as that in which it was put up with no reasonable prospect of immediate examination: see per Lord Atkin at page 599; also Grant v Australia Knitting Mills ltd [1936] AC 85 103 - 105 per Lord Wright. It is the latency of the defect which constitutes the mischief. There may be room for disputation as to whether the likelihood of an intermediate examination and consequential actual discovery of the defect has the effect of negativing a duty of care or breaking the chain of causation (compare Farr v Butters Brothers & Co. [1932] 2 KB 606 with Denny v Supplies & Transport Co. Ltd [1950] 2 KB 374). But there can be no doubt that, whatever the rationale, a person who is injured through consuming or using a product of the defective nature of which he is well aware has no remedy against the manufacturer. In the case of a building, it is right to accept that a careless builder is liable, on the principle of Donoghue v Stevenson, where a latent defect results in physical injury to anyone, whether owner, occupier, visitor or passer by or to the property of any such person. That principle is not apt to bring home liability towards an occupier who knows the full extent of the defect yet continues to occupy the building."

51. I do not accept the appellants' analysis. It must be borne in mind that the principal issue before their Lordships in Murphy v Brentwood was the legitimacy or otherwise of their earlier decision in Anns v Merton London Borough Council [1978] AC 728. The factual background was that there had been a failure of the foundations to the plaintiff's house which resulted in extensive damage to the walls and pipes. As a consequence, the plaintiff had received less than the market value on sale. The significance of the latency or otherwise of the defect for the purposes of the decision is helpfully outlined in the headnote to the report. It reads:-

"Held: Allowing the appeal, that while the principle in Donoghue v Stevenson [1932] AC 562 applied to impose a duty on the builder of a house to take reasonable care to avoid injury or damage, through defects in its construction, to the persons or property of those whom he ought to have had in contemplation as likely to suffer such injury or damage, that principle as stated extended only to latent defects: that, where a defect was discovered before any injury to personal health or damage to property other than the defective house itself had been done, the expense incurred by a subsequent purchaser of the house in putting the defect right was pure economic loss...."

52. In my judgment, the decision in Murphy v Brentwood was dealing with matters of broad principle and does not detract from the proposition that a defect is not latent if it is discoverable by the exercise of due diligence whether or not due diligence was in fact exercised. Awareness of the danger (i.e. the propensity to flood) is not of itself only consistent with a flaw. The relevance of such awareness is to the ability to discover the actual defect by the exercise of due diligence.
53. In summary, I would put the matter in this way. Where, in the normal course of events, a surveyor would be engaged in a survey of a building for a purchaser, and, with the exercise of due diligence, that surveyor would have discovered a defect, that defect is patent whether or not a surveyor is in fact engaged and, if engaged, whether or not the surveyor performs his task competently. It follows that in my judgment, the judge was right to conclude as follows:

"108. In the present case, the precise defect had not been discovered before the floods, but there had been at least one previous flood evidenced by markings on the floor in the building. In my opinion, the duty of the defendants to the claimants must depend on the question, "Was there a reasonable opportunity of inspection of the drainage system and discovering the defects before the floods?" If the claimants had a reasonable opportunity of inspecting the drainage system and discovering the defects before they suffered damage, it would not be fair just or reasonable to hold the defendants liable for that damage nor would it be right to say that there was a proximate relationship between the claimants and the defendants.

109. There was a reasonable opportunity of inspecting the building before the claimants took a lease. It would be normal procedure for any incoming tenant to have the building inspected by the surveyor, and that is what they did. Although the claimants received warnings from both the surveyors and from Mr Meiklejohn of a danger, they were not told what was the precise problem. The surveyors could, and in my view should, have told the claimants that there were no overflows, and that overflows should be provided. The cost of overflows was very small and if the claimants had been advised to install them then I cannot think that they would have failed to do so.

110. If Lambert Smith Hampton had been more assiduous in the performance of their duties, the claimants would have been expressly warned of the absence of overflows and the floods would not have occurred. To what extent is the claimants' claim affected by the acts of their professional advisers? Are they entitled to say, as they might in response to a defence of contributory negligence, that they took skilled advice and are entitled to rely on that advice? I do not think that that is the right approach.

I do not think that it is fair just or reasonable that the extent of the liability of the defendants should depend on the assiduity of the surveyors instructed by the claimants. The claimants had the opportunity to discover the absence of overflows by reasonable inspection by professional advisers who might reasonably be expected to be instructed: whether that reasonable opportunity in fact revealed the defect is irrelevant. Because there was a reasonable opportunity to inspect, the defendants were not in a proximate relationship to the claimants so far as concerns defects which could have been discovered by that inspection, namely, the absence of overflows. But I repeat my previous finding that neither the claimants nor their surveyors could reasonably be expected to have discovered the underdesign of the drainage system."

54. In my judgment the judge's analysis is correct. Actual knowledge of the defect, or alternatively a reasonable opportunity for inspection that would unearth the defect, will usually negative the duty of care or at least break the chain of causation unless (as is not suggested in the present case) it is reasonable for the claimant not to remove the danger posed by the defect and to run the risk of injury: see Targett v Torfaen BC [1992] 3 All ER 27 per Sir Donald Nicholls V-C at p.37.

Liability for the second flood

55. It follows that the judge was right to find that that there was no duty of care in respect of damage caused by the absence of overflows. But was he right to find that, because there was a shortfall in the design of the gutter which would not be apparent on reasonable inspection, there was liability on the part of the appellants in respect of damage caused by the second flood?
56. The judge expressed his conclusions briefly as follows:-

111. d. The first flood was caused by a combination of blockages of the drains for which the defendants were not responsible and the absence of overflows. The underdesign of the system was not a contributing cause of the first flood. Since the claimants could reasonably have discovered the absence of overflows, the defendants no longer owed them a duty in that regard and therefore the claimants can prove no breach of duty in relation to the first flood.

e. The second flood was caused by a combination of the underdesign of the system and the absence of overflows. Blockages of the system were not a contributory cause of the second flood. While there was a reasonable opportunity to inspect which should reasonably have revealed the absence of overflows, it cannot be said that the opportunity to inspect should reasonably have revealed the underdesign of the system. The defendants are therefore responsible for one primary cause of the second flood, though not for the other. On ordinary principles of liability in tort, the defendants are therefore liable for the whole of the loss arising from the second flood…."

57. It was only at this last stage that my initial reaction was to part company with the judge. I put the matter that way, because, when the Court came to hand down its judgment allowing the appeal on this issue, it was faced with an application by the respondents that the appeal be restored for further argument. The theme of this application was that the respondents had not been afforded a proper opportunity to address the Court on the issue, it not having been raised, it was contended, with any sufficient clarity by the appellants.
58. It is, of course, open to the Court to vary or alter its judgment at any time before its Order is perfected: Re Harrison's Share Under a Settlement [1955] Ch 260 at 276. That jurisdiction remains despite the introduction of CPR: Stewart v. Engel [2000] 1 WLR 2268. It is a jurisdiction which can only be exercised in exceptional or special circumstances: Noga d'Importation v. Bacha [2001] 3 All ER 513. But a good reason would be if the applicant could argue that he was taken by surprise or otherwise did not have a fair opportunity to consider an issue: Re Blenheim Leisure (Restaurants) Ltd (3) The Times 9th November 1999 per Neuberger J.
59. The Court felt that there was some force in the respondents' submission that the issue of causation had become buried under the argument on the major planks of the appeal on which the appellants were unsuccessful. Accordingly, the parties were afforded an opportunity to exchange further submissions on the issue of causation so as to enable the Court to review the conclusion outlined in its draft judgment. We told them that we would convene a further oral hearing if we considered this to be necessary, but in the event we have found it possible to deal with these further arguments justly on the basis of the further written submissions we received.
60. Despite the torrent of material put before the Court in response, I unhesitatingly re-affirm my conclusion that the only effective cause of both floods was the absence of overflows, a defect that ought reasonably to have been identified and remedied by the respondents.

Overflows

61. The respondents sought to contend that there was no finding to the effect that, if overflows had been installed by the time of the second flood, no flood would have occurred. It is true, as indicated in paragraph 34 above, that there was no finding as to the precise flow rate that could have been absorbed by the gutter (inclusive of overflows) as designed. For the purposes of considering the argument that the adequacy of the overall design must be considered by reference to the cumulative ability of the gutter and the overflow system to absorb a specific rate, it was not necessary to do more than assume that the inclusive capacity was 150 mm / hr.
62. But for the purposes of causation, the judge found in terms at paragraph 110 of his judgment that if the claimant had been warned of the absence of overflows "the floods would not have occurred". This is scarcely a surprising conclusion:-

a) As recorded in paragraph 31 of the judgment below, calculations by reference to BS 6367 had been made in late 1989/early 1990 as to the capacity of the valley gutter. Albeit the design rate was 75mm per hour, the actual capacity was substantially greater. (Indeed, the calculations record that the run off from the roof was 4.6 litres per second against an outlet capacity of 6.9 litres per second.)

b) By virtue of this redundant capacity (as against the chosen design), the experts are agreed that, whilst the valley gutter incorporated 7 outlets, it only required the installation of 3 more outlets to discharge a rainfall intensity of 150 mm per hour.

c) The day after the second flood, when the need for overflows was finally hoisted in, Mr Meiklejohn ordered that weir overflows should be cut. Whilst this was not practical, he arranged for "3 x 100mm. gravity overflows to be inserted at the extreme end of the gutter to simulate weir conditions": see his letter of 23rd April 1996. Further remedial work was then effected as described in paragraph 24 above.

d) As found by the judge, the rainfall at the time of the second flood was somewhere between 75 mm per hour (well below the actual design capacity) and 150 mm per hour (slightly above the actual design capacity). Any shortfall would have been readily made good by the steps taken in the immediate aftermath of the second flood, let alone the supplementary siphonic overflow system (with a design rate of 100 mm per hour) installed shortly thereafter.

63. Against this background, it is wholly unreal to suggest that this court cannot confidently proceed on the basis that the second flood would have been avoided had overflows been installed.

One or two causes?

64. It has to be said that the judgment below gives rise to the surprising result that, whilst the appellants are not liable for the damage sustained in the first flood, because the respondents could easily have ascertained the absence of overflows, they are nonetheless liable in respect of the second flood which would equally have been avoided if overflows had been installed, the more so when the first flood highlighted the absence of overflows.
65. In their extensive further submissions, the respondents reiterated the argument, accepted by the judge below, that the second flood was materially contributed to by two causes - the absence of overflows and the underdesign - and that thus the appellants are responsible for the damage sustained since one of those causes, the underdesign, was a latent defect.
66. I regret that I am quite unable to accept this analysis. It has to be borne in mind that the deficiency in the design of the gutter solely related to the number of outlets (seven instead of ten). As explained above, it was no answer to the complaint about non compliance with BS 6367 that the overflows would make good the absence of sufficient drainage. The overflows are to allow for flows in excess of the design rate: i.e., as a matter of design of the building, water should not be allowed to cascade from weir overflows in the valley (or perimeter) gutter unless the rainfall exceeds the appropriate design rate.
67. But from the perspective of protecting goods stored in the building, it matters not how the rain water is taken away from the gutter: an overflow is as adequate as a drain. It is for this reason that I was at pains, in paragraph 40, to emphasise the causative importance of the absence of overflows albeit rejecting the submission that the appropriate design criteria had been met.
68. Given that weir overflows were impractical, there was no distinction in principle between the drainage arrangements and the overflow arrangements: they both involved gravity or siphonic piping systems. The point is exemplified by the remedial arrangements. The existing siphonic drainage system needed to be protected in the sense that, since it only worked efficiently when filled with water, the supplementary drainage had to be raised above the siphonic outlets. Accordingly, once it is accepted that the respondents ought reasonably to have discovered the absence of overflows, the inevitable remedial measures, whether properly categorised by way of drainage or overflow, would have avoided the second internal flood.
69. Put another way, the overflows were supplementary to the design capacity of the system. Their purpose was to provide relief both in the event that the gutter became blocked and in the event that the rainfall exceeded the design capacity. There were no less than two occasions when the absence of overflows ought to have been appreciated: first at the time of the pre-purchase survey and second in the aftermath of the first flood. If the deficiency had been realised, the remedial measures would have been taken, none of which involved re-design save in the sense of installing additional drainage capacity. Thus, in my judgment, the sole effective cause of both floods was the absence of overflows.
70. In my judgment it is not possible to say that another contributing cause of the damage in this case was the underdesign of the gutter. The remedy remained the same. The chain of causation between the architect's error in regard to the provision of overflows and both floods was broken. I cannot subscribe to the proposition that the links in the chain can be re-connected to the second flood merely because there was another error which rendered the provision of overflows the more important.

Conclusion

71. For my part, I would allow the appeal and dismiss the cross-appeal.

Lady Justice Hale:

72. I agree.

Lord Justice Brooke:

73. I also agree.
Order: (Appeal allowed; cross-appeal dismissed; Respondents do pay the Appellants 50 percent of costs; application for permission to appeal to the House of Lords refused). (Order not part of Approved Judgment).