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

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

		Royal Courts of Justice Strand, London, WC2A 2LL
		9 November 2016

Mr Justice Coulson:

A: INTRODUCTION

1. On the morning of Sunday, 24 March 2013, a fire broke out at Mead House, Appleshaw, near Andover in Hampshire. The fire started in the roof void above the second floor rooms and caused significant damage before it was brought under control. Quantum has been agreed, subject to liability, in the sum of £685,000 plus interest.
2. The Defendant is a sole trader who has worked for over twenty years as a pest controller. In around May 2012 and again in November and December 2012, the Defendant was asked by the Claimant, the owner of Mead House, to deal with a rodent problem in the roof void. In the absence of any other means of access, the Defendant removed a number of halogen lamps, and inserted bait blocks through the openings thereby created into the roof space, before replacing the lamps.
3. It is the Claimant's case that the Defendant was negligent because he placed bait blocks that were touching or almost touching the halogen lamps, and alleges that it is more likely than not that the fire was caused by one of the halogen lamps setting fire to one of the bait blocks, which then spread to other combustible material. The Defendant denies negligence in the placing of the bait blocks or the methodology used. As to causation, the Defendant identifies a number of potential causes of the fire and submits that, on any view of the evidence, the bait block was only one of a number of possible causes of the fire, and less likely than others.
4. The form of this judgment follows the agreed List of Issues helpfully ordered by Jefford J at the PTR. I set out the facts in Section B. At Sections C, D and E I deal with certain disputes of fact before, at Section F, addressing the issue of negligence. Thereafter, I deal with the two principal (and related) issues concerning the fire: at Section G, I consider the issue of where the fire originated, and in Section H, I deal with the fundamental question of causation: whether, on the balance of probabilities, the Defendant's alleged negligence caused the fire. I am very grateful to both counsel for their helpful and efficient management of the trial.

B: THE FACTS

5. Mead House is a large residential house built in 1720. The main house is in conventional Georgian style with a front door and two windows on either side at ground floor level, and five windows at first floor level. The second floor, which is in the roof and was presumably designed as the servants' quarters or nurseries, has three dormer windows at the front, and other windows on each elevation. It has a total of four rooms, one in each corner, and a bathroom in the middle at the front. There is a rear extension which does not have a second storey.
6. Mead House was apparently the subject of major renovation works about 13 years ago, which probably included complete rewiring. The Claimant bought the house in conjunction with his children in September 2011. It was tenanted. In early 2012, the tenant complained of a rodent problem in the narrow roof void above the second floor ceiling. The Claimant contacted the Defendant, who attended on three separate occasions in around May 2012. Because there was no conventional access to the roof void, he removed halogen lamps from the downlighters in the ceiling of the second floor, pushed the bait blocks through the openings, and then replaced the lamps. He charged £35 for each visit. At the end of the third visit, it appeared that the rodent problem had been resolved.
7. In November 2012, following a further complaint of rodent activity, the Defendant visited again. Although he had advised the Claimant after his first visits to install a better means of access to the roof void, that had not happened, so he was obliged to install further bait blocks using the same method as before. It appears that he revisited on 3 December 2012 for a final time. He was told by a young man (who nobody has been able to identify) that no further noises had been heard in the roof void after the baiting exercise the previous month^[1].
8. Just after this, the Claimant took on new tenants, the Goolds, at Mead House. Prior to the commencement of their tenancy, he instructed an electrical contractor, JCR Electrical, to carry out an inspection of the electrical installation in the house. This work took two days and resulted in an inspection certificate. This showed that, amongst other things, the lighting circuits on the second floor had been tested and no problems had been found.
9. There was some debate about whether or not, as part of this exercise, Mr Rolfe of JCR Electrical had changed any of the halogen lamps on the second floor. His witness statement indicated that lamps had been changed in the kitchen/ dining room on the ground floor, but not elsewhere. Although Mr Rolfe was too unwell to come to Court to give evidence, so that his statement was not tested in cross-examination, there was no material which indicated that any lamps had been changed on the second floor. Mr Rolfe may, however, have removed and then replaced lamps on the second floor as part of his wider inspection; there is no evidence either way.
10. The Goolds rarely used the second floor of Mead House, reserving it for guests. In consequence, it was originally suggested on behalf of the Claimant that the lights on that floor were not really used until March 2013, when the fire happened. The Claimant's case in opening made a clear link between the lack of use of the second floor before 23/24 March, and the fact that, when it was used that weekend, a fire broke out.
11. The oral evidence demonstrated that this was an incomplete narrative. Mr Goold said that, from Christmas to March, there were about six occasions when they had weekend guests who stayed for a night on the second floor. Since the north-eastern bedroom on the second floor was the principal guest bedroom, it is a fair inference that, during those winter months, the lights were used in that room, at least for some time, with no adverse effects.
12. On the weekend of 23/24 March 2013 the Goolds again had guests staying on the second floor. Mr and Mr Grillo stayed in the north-eastern bedroom whilst their nine-month old son was across the landing in the south-eastern bedroom. On the Saturday, it appears that the lights in the north-eastern bedroom were not used for more than about twenty minutes. However, on the Sunday, the evidence was that the Goolds' young sons went up to the second floor some time after 7am and, as a joke, switched on the lights in the north-eastern bedroom. Mrs Goold said that it was unlikely that this was later than about 7.15 am. She said that she and husband were awake anyway because of their young son, and, although they stayed in bed, the lights remained on thereafter.
13. It was not clear when, if at all, they turned the lights off. The Grillos went downstairs for breakfast at about 9am and Mrs Grillo said that she was sure that they would have turned the lights off when they went downstairs because that is what they habitually did: "normally we are reasonably particular and we turn light switches off on leaving a room". Mr Grillo agreed with that. The post-fire photographs of the switches are inconclusive. I find that, on the balance of probabilities, the Grillos switched the lights off at about 9am when they went downstairs. On that basis, the lights in the north-eastern bedroom had been on for between one and a half and two hours.
14. Before they went downstairs, between about 8.30 and 9 am, Mrs Grillo said that she and her husband could smell something burning. They were not particularly alarmed, and thought that it was something to do with breakfast.
15. When they went downstairs, they passed Mr Goold on his way upstairs. He also noticed a burning smell which he likened to an incense candle. He went up to the second floor. He recalled that the door to the south-eastern bedroom was open, but that the door to the north-eastern was closed. He did not open the door. He checked that the smell was not emanating from an oil heater in the south-eastern bedroom and then returned downstairs.
16. Shortly afterwards Mr Goold went to buy a newspaper. Whilst he was away, at about 9.30, the intruder alarm sounded. Those in the house saw smoke outside the house and Mr Grillo went outside to see smoke and flames in the roof area. Foolishly, he ran upstairs to retrieve his car keys from the north-eastern bedroom. When he entered that room he said he could see the ceiling lights glowing orange and thought that there were flames behind the lights. He was subsequently to indicate to the investigators one downlighter in particular, although his oral evidence confirmed that the glow was not confined to one downlighter. He could also hear the roar of a fire above him. When he gave evidence he stressed that the roar was "incredibly loud".
17. Mr Goold, returning from buying the paper, also saw the smoke. In the meantime Mrs Goold had rung 999 at 9.33am. The Claimant, who lives next door to Mead House, also saw the fire and, in particular, saw smoke billowing from the north-eastern bedroom roof. The fire brigade attended and brought the blaze under control but much of the roof, and therefore the second floor, was damaged, in particular the north-eastern corner.
18. The day after the fire, Mrs Gosling – a fire investigator instructed by the Claimant's insurer – visited the site and had discussions with a number of people, including a telephone conversation with the Defendant. She visited again on 28 March. On 3 April she produced what was called a "Preliminary Appraisal Report". This document, which subsequently formed the basis of much of Mr Smith's expert consideration on behalf of the Claimant, was only disclosed shortly before the trial.
19. The following passages were of particular importance:

"Above the plasterboard ceilings of the rooms on the second floor was a layer of reflective-type insulation. The insulation comprised several layers of paper, foam sheets and foils. There was no other insulation in the roof space. The wiring for the ceiling lights was routed through the roof space but I found no evidence of any other cables or electrical items within the roof space. Within the roof space, close to the light fittings I found remains of the rodent poison blocks used by Mr Nightingale. All of the blocks were situated in close proximity to the edge of the hole into which the halogen downlighter was recessed…

I found charring to one of the rodent blocks in an undamaged section of the roof space and therefore it is clear that the charring was caused by heat from the halogen bulb as opposed to the fire. I do not know whether the heat from the bulbs would be sufficient to ignite the blocks or only to char them but testing might enable me to establish this."

20. In her preliminary conclusions Mrs Gosling said:

"I consider the most likely cause of the fire to be associated with the halogen light fittings; either an incendive electrical fault or nearby combustible materials (such as rodents, fly, insulation, timber or the rodent blocks) being ignited by radiant and/or conducted heat from the halogen bulbs. Given that Mr Nightingale was the last person to work in the roof space and there had been no reported problems before his visit it is possible that the cause of the fire was associated with his actions (such as placing one of the rodent poison blocks on top of one of the light fittings or moving insulation). …I will arrange to carry out tests with the halogen light fittings, insulation and poison blocks in an attempt to establish whether heat from the halogen bulb would be sufficient to ignite the poison blocks and/or insulation."

21. On 17 May 2013, having concluded his investigation, the fire investigation officer noted the evidence of "burning around the undamaged downlighters in the roof space" and concluded that the cause of the fire was "therefore ignition of insulation material in roof space from halogen downlighters".
22. Thereafter, four further tests were carried out by the parties in relation to the bait blocks. In chronological order, they were:

i) In June 2013, on behalf of the Defendant, tests were carried out during which a flame was applied to the bait blocks. This demonstrated that, contrary to the manufacturers' data sheets, the bait blocks were both flammable and ignitable.

ii) In September 2013 there was the first round of joint testing by the experts in which attempts were made to ignite bait blocks retrieved from site by placing them on top of or very close to halogen bulbs. A number of different tests were carried out but at no point did any of the bait blocks ignite. Some charring/smouldering was achieved.

iii) In January 2014, the Defendant's experts again carried out some unilateral testing at Edinburgh University. A total of twelve different tests were carried out in which the bait blocks were placed on or close to halogen lamps. No ignition was achieved, although there was again some evidence of charring/smouldering.

iv) In February 2014 there was a second round of joint testing. Again, however, the halogen lamps did not ignite the bait blocks, which on this occasion had been bought specially for testing purposes.

Accordingly, the tests which Mrs Gosling said in her first report she thought were important (paragraphs 19 and 20 above) were carried out over a period of 8 months, and showed that the hot lamps were not sufficient to cause ignition of the bait blocks but, as she put it in that report, "only to char them".

23. Subsequently, Mrs Gosling went on maternity leave, and her role as the Claimant's expert was taken on by a colleague, Mr Smith. This led to a certain amount of difficulty in respect of the source of some of the information in Mr Smith's report. In my view, the reasonable queries raised by the Defendant's advisers in respect of those sources of information were not always dealt with in a clear or straightforward way. However, the belated provision of Mrs Gosling's first reports, and my order that other factual material recorded by Mrs Gosling was also to be disclosed (but not her statements of opinion or advice), eventually removed most of the unnecessary mystery surrounding that aspect of the case.

C: ISSUE 1 – HOW MANY OCCASIONS AND WHEN DID THE DEFENDANT VISIT MEAD HOUSE?

24. The Defendant's invoices demonstrate that he attended three times during the first round of visits (with the last visit on 18 May 2012) and then twice in the second round of visits, the second visit on 3 December 2012. Although the visits were the subject of some cross-examination, it seems to me that the contemporaneous invoice is much the best record of when and how many times the Defendant visited the property.

D: ISSUE 2 – WHAT TYPE OF BAIT BLOCK WAS USED?

25. This was something of a peripheral issue. All the bait blocks found at the scene appeared to be 'Contrac' blocks, but of course many others were destroyed in the fire. The Defendant says that this is surprising because, for the November visit, he recalled running out of 'Contrac' blocks and using 'Storm' blocks instead, although one possible explanation which was consistent with both was that he used 'Storm' blocks in the traps in the rooms, and 'Contrac' blocks in the roof void.
26. It seems to me that, on the balance of probabilities, the Defendant used both 'Contrac' and 'Storm' blocks on his visits in May and November/December 2012 but the likelihood is that he used many more 'Contrac' blocks because those were his preferred choice, and those were what was found after the fire. The 'Storm' blocks may well have been confined to the traps in the rooms.

E: ISSUE 3 – HOW FAR FROM THE LAMPS DID THE DEFENDANT PLACE THE BAIT BLOCKS?

27. Before considering this issue it is important to deal with an antecedent matter, particularly because it is agreed by the pest control experts. This concerns the methodology which the Defendant adopted of putting the rodent bait into the roof void by unscrewing the lamps of the downlighters and putting them into the void through the opening thereby created.
28. Mr Madge, the pest control expert called by the Claimant, expressly agreed:

i) that accessing the loft void via the light fitting was a reasonable method for applying bait to an otherwise inaccessible void;

ii) that applying the bait through the light fitting was an acceptable, cost-effective route into sealed areas;

iii) that placing bait through the light fittings provided easy access into the roof void and it would be a fair assumption that non-target species would not be negatively affected;

iv) that if the blocks were inserted through the hole left by the light bulb then blocks would have been the only control option.

Accordingly, there was no issue between the parties as to the reasonableness of the methodology adopted by the Defendant in this case. By the end of the trial, the only live issue was whether he had carried out that methodology negligently, because he had put the bait blocks so close to the lights that they touched, or almost touched, the lamps.

29. The Defendant's written and oral evidence was to the effect that, having removed the lamps, he would have pushed the bait blocks back approximately two to three inches away from the opening. He recalled telling Mrs Gosling that he placed them close enough to the opening so that he could see or touch them in order to check on his next visit whether they had been moved by the rodents. He said that he would have "pushed the blocks away instinctively. I mean there's a lamp and there's a hole and I want [to put in] a block and I would instinctively push it away from the heat source". He did not place the blocks on top of the lamps because that would have been impossible: he had had to remove the lamps in order to get the blocks in in the first place.
30. In opening, it was the Claimant's case that the Defendant had put the bait blocks very close to the lamps, although in evidence Mrs Gosling unhelpfully refused to be drawn as to what this actually meant. In his closing submissions, Mr Sullivan suggested that the bait blocks, or at least some of them, had been placed "hanging over the edge of the hole" so that, when the lamp was put back in, the bait block touched or came very close to the lamp. That case appeared to be based on: i) Mrs Gosling's contemporaneous note of her discussion with the Defendant, which referred to the blocks being placed "v. close" to the openings; ii) the photograph after the fire of another block in the north western (or 'white') bedroom, which could be seen through the hole where the bulb would have been; iii) the evidence of the charred block apparently found in the south western (or 'blue') bedroom which fell out when the bulb was removed.
31. Before looking at those three strands of evidence, it is important to note that what the Defendant said he did when placing the blocks (paragraph 29 above) was entirely consistent with what Mrs Gosling told me she found when she inspected the scene. She had not always been able to see the bait blocks when the lamps were removed: sometimes she had had to "feel around" for the bait blocks once the lamp had been taken out. That was precisely consistent with how the Defendant said he had gone about the task. It is the complete opposite of hanging the blocks precariously over the edge of the hole, and then putting the lamp back in, an unpleaded suggestion that was not directly put to the Defendant in cross-examination, and which was contrary to common sense, and the Defendant's need to check later to see if the blocks had been moved or touched. On this basis alone, I reject the suggestion that bait blocks were placed touching or almost touching the halogen lamps.
32. In addition, I reject each of the three specific matters relied on by the Claimant to indicate the contrary, listed in paragraph 30 above. As to the "v.close" reference, that was generally consistent with what the Defendant said he had done in his oral evidence. It reads far too much into these manuscript notes to say that "v.close" somehow meant 'actually or almost touching the lamp'.
33. Unfortunately, the Claimant's desire to establish this part of his case ignored what Mrs Gosling had herself found and how she described it. Mrs Gosling's scene examination report quoted at paragraph 19 above, described the bait blocks as being in "close proximity" to the downlighters, which was accurate. At paragraph 3.13 of a report prepared by Mrs Gosling on 6 May 2016 (disclosed during the trial), Mrs Gosling had changed this to "very close proximity". At paragraph 3.14 of his report, Mr Smith had faithfully copied out what Mrs Gosling had said in her May 2016 report, which was not accurate.
34. As to the photograph of the bait block behind the downlighter in the white bedroom, this may not have been 2 or 3 inches from the lamp but, similarly, it was not in a position where it could have been touching or almost touching the lamp. It was not charred. Mrs Gosling agreed that it was "further back" on the insulation, which was itself much closer to the hole, and therefore the lamp. She refused to answer the obvious point that it was not close to the lamp, saying only it was placed "so that it could be associated with that [downlighter] as opposed to anything else". That was pointless sophistry: since the downlighter hole had been used to place the block, all the blocks "were associated with the downlighter rather than anything else". In short, the photograph contradicted, rather than supported, the Claimant's case.
35. As to the charred block, the fact that it fell out when the bulb was removed after the fire strongly suggested to me that something – a rodent, perhaps, or the effect of the fire - had caused it to move after it was installed by the Defendant: it could not sensibly have been placed by the Defendant on top of the lamp for the reasons previously given, and he vehemently denied leaving it there. Moreover, as set out in paragraphs 82-83 below, there was serious doubt as to where this charred bait block was actually found, given Mrs Gosling's admission in cross-examination that she had wrongly located it on her drawing.
36. In addition, Mr Smith and Mrs Gosling between them accepted that the Claimant's pleading was incorrect when it suggested, by reference to the same site drawing, that there were numerous locations in the second floor roof void where bait blocks were found after the fire very close to, let alone touching, the lamps. In the majority of cases, Mrs Gosling had had to feel for the blocks with her hands. Even taking the Claimant's case as its highest, there were only two bait blocks close to or touching the lamps, namely the two bait blocks identified in paragraphs 34 and 35 above. In total, Mrs Gosling identified 11 bait blocks in the less damaged part of the roof void. Accordingly, that meant that, even on the Claimant's best case, only two out of eleven bait blocks were very close to or touching the lamps. On that basis, of course, as Mr Smith properly accepted in cross-examination, it was statistically less likely, rather than more likely, that any bait blocks in the north-eastern bedroom had been placed very close to, let alone touching, the lamps.
37. On the basis of all the evidence, therefore, I find that the Defendant placed the bait blocks on the insulation or the boards about 2 or 3 inches beyond the hole, so that he could see or feel them when he returned. It would not have been possible to place them very far beyond that. He did not place them where they were touching or almost touching the bulbs. This was the best that he could do, given the access restraints.

F: ISSUE 4 – WAS THE DEFENDANT NEGLIGENT/IN BREACH OF CONTRACT?

F1: Proximity of Blocks to Bulbs

38. Given that no criticism was made of the methodology adopted by the Defendant, the principal issue on liability was whether the Defendant had placed the blocks so that they touched or almost touched the halogen lamps. In the light of my findings of fact at Section E above, the case on liability must fail. The Defendant did not place the bait blocks so that they touched or almost touched the lamps. He was therefore not negligent or in breach of contract.
39. For completeness, I should add this. Although the Defendant agreed that, as a matter of common sense, he knew that he should not place a block right next to a hot bulb, there was nothing to indicate that he knew, or ought reasonably to have known, that the blocks were flammable.
40. The 'Contrac' blocks are amongst the most widely used products of this type across the world. No possible criticism can therefore attach to the Defendant's general use of these products. Furthermore, the products are not described as being flammable or combustible. On the contrary, the product data sheets describe them as being non-flammable and not containing combustible liquids. Although subsequent tests in this case demonstrate that this information might well have been wrong, the Defendant cannot be criticised for that.
41. In the 'Contrac' block data sheet there is a reference to the product not being used near hot surfaces and that extreme temperatures should be avoided. That latter point can be discounted immediately: nobody suggests that the halogen lamps were at an extreme temperature. As to the reference to hot surfaces, the pest control experts differed as to its significance.
42. I have concluded that, on this topic, Mr Charlton's evidence – on behalf of the Defendant - is to be preferred. He said that the warning in the data sheet about use near hot surfaces was simply a warning about the risk of impairing the integrity of the product. The fact that the materials were expressly stated to be non-flammable, and not containing combustible liquids, would have meant that a reasonable pest controller would not have thought that the 'Contrac' blocks posed any particular fire risk.

F2: Other Matters

43. There were a number of other criticisms of the Defendant made by Mr Madge, the Claimant's expert, and pleaded in the Particulars of Claim. Most of these were, in one way or another, concerned with the absence of written assessments and other forms of documentation. The cross-examination of Mr Madge revealed that, as a sole trader, the Defendant was not obliged to produce them. In any event, I do not consider that, even if the Defendant had carried out detailed written assessments for all or even one of his visits, it would have made any difference to what he did or what he was reasonably entitled to do. In my view, these allegations should never have been made.
44. There is also a complaint advanced by Mr Madge to the effect that the Defendant did not remove the bait blocks in December 2012. Mr Charlton said that, as there was no risk to human health, it was not unreasonable to leave the bait in place. I accept that evidence; it seems to me to flow from the agreements between the experts noted at paragraph 28 above.
45. Finally, there was the suggestion that the defendant should have put wire through the blocks to help with retrieval. The evidence was that this would have been unusual. The wire itself would have posed a potential hazard. And it would have made no difference on the facts.

F3: Conclusions on Liability

46. For the reasons set out above, I find that the Defendant was neither negligent nor in breach of contract in carrying out the work in the way that he did. The experts are agreed that he was entitled to put the bait blocks through the openings. The bait blocks were not described as flammable or combustible. Placing them about two or three inches away from the downlighters, as I have found that he did, was not contrary to any of the information in the product sheets and was a reasonable course to adopt; indeed, even if he had placed the blocks much closer to the lamps, I reject the suggestion that he was negligent.
47. On that basis, of course, the Claimant's claim fails. However, because both parties accepted that questions of liability were, up to a point, bound up with questions of causation, and because I have reached a clear view as to causation, it is appropriate to go on to deal with the remainder of the issues between the parties.

G: ISSUE 5 – WHERE DID THE FIRE ORIGINATE?

48. Mr Smith, guided by Mrs Gosling's contemporaneous documentation, was firmly of the view that the fire originated in the roof void above the north-eastern bedroom. Mr Gow, the Defendant's expert, devotes a large part of his report to arguing round that issue. However, as Mr Sullivan rightly pointed out, at paragraph 5.6.8 of his report, Mr Gow concluded that "The most reasonable deduction to be made is that the fire originated in the attic/void above and around the north-east corner of the building".
49. Accordingly, I conclude that the fire originated in the void above the north-eastern corner of the building. That is also entirely consistent with Mr Grillo's evidence about the glow behind the lights and the fire roaring above the ceiling when he returned to the north eastern bedroom to pick up his car keys.

H: ISSUE 6 – DID THE DEFENDANT'S NEGLIGENCE CAUSE THE FIRE?

H1: The Relevant Legal Principles

50. In circumstances where there are a number of potential causes of a fire, it is not appropriate for the Court to work through each potential cause, identify the least improbable, and then fix that as the cause of the accident or fire. The House of Lords made that clear when overturning the original decision in The Popi M (Rhesa Shipping Co SA v Edmunds) [1983] 2 Lloyds Rep 235. It is always necessary for the judge to stand back and consider whether the claimant has proved, on the balance of probabilities, that their case as to causation was, on the balance of probabilities, more rather than less likely.
51. This was neatly summarised by Thomas LJ (as he then was) in Ide v ATB Sales Limited [2008] EWCA Civ 424 when he said:

"4. …In the vast majority of cases where the judge has before him the issue of causation of a particular event, the parties will put before the judges two or more competing explanations as to how the event occurred, which though they may be uncommon, are not improbable. In such cases, it is, as was accepted before us by the appellants, a permissible and logical train of reasoning for a judge, having eliminated all of the causes of the loss but one, to ask himself whether, on the balance of probabilities, that one cause was the cause of the event. What is impermissible is for a judge to conclude in the case of a series of improbable causes that the least improbable or least unlikely is nonetheless the cause of the event; such cases are those where there may be very real uncertainty about the relevant factual background (as where a vessel was at the bottom of the sea) or the evidence might be highly unsatisfactory. In that type of case the process of elimination can result in arriving at the least improbable cause and not the probable cause.

…

As a matter of common sense it will usually be safe for a judge to conclude, where there are two competing theories before him neither of which is improbable, that having rejected one it is logical to accept the other as being the cause on the balance of probabilities. It was accepted in the course of argument on behalf of the appellant that, as a matter of principle, if there were only three possible causes of an event, then it was permissible for a judge to approach the matter by analysing each of those causes. If he ranked those causes in terms of probability and concluded that one was more probable than the others, then, provided those were the only three possible causes, he was entitled to conclude that the one he considered most probable, was the probable cause of the event provided it was not improbable."

This passage was recently approved by the Court of Appeal in Graves v Brouwer [2015] EWCA Civ 595 at paragraph 24.

52. The importance of the judge standing back so as to ask the critical question as to whether the claimant's suggested explanation is more likely than not to be correct was emphasised by Toulson LJ (as he then was) in Milton Keynes Borough Council v Nulty & others [2013] 1 WLR 1183. He said:

"34. … Consideration of such a case necessarily involves looking at the whole picture, including what gaps there are in the evidence, whether the individual factors relied upon are in themselves properly established, what factors may point away from the suggested explanation and what other explanation might fit the circumstances. As Lord Mance observed in Datec Electronics Holdings Limited v UPS limited [2007] 1 WLR 1325, at 48 and 50, there is an inherent risk that a systematic consideration of the possibilities could become a process of elimination 'leading to no more than a conclusion regarding the least unlikely cause of loss', which was the fault identified in The Popi M. So at the end of any such systematic analysis, the court has to stand back and ask itself the ultimate question whether it is satisfied that the suggested explanation is more likely than not to be true. The elimination of other possibilities as more implausible may well lead to that conclusion, but that will be a conclusion of fact: there is no rule of law that it must do so."

53. More recently, the proper approach was summed up by Carr J in this court in Cooper v Thameside Construction Company Limited [2012] EWHC 1248 (TCC) as follows:

"114. It is therefore common ground between the parties in principle that in a competing theories case:

a) The court is entitled to engage in a systematic analysis of competing theories and eliminate one in favour of the other;

b) That, in itself, may lead to the conclusion that the preferred theory is more likely than not to be true; but

c) The court must always then stand back and ask itself the ultimate question as to whether or not the preferred explanation is more likely than not to be true."

That is the approach I adopt in this case.

H2: The Probable Cause of the Fire

H2.1: The Halogen Bulbs/Downlighters

54. In my view, based on all the evidence, the probable cause of the fire was the generation of heat by at least one of the halogen lamps in the ceiling of the north-eastern bedroom. To that extent, therefore, I accept the evidence of Mr Smith, the Claimant's fire expert. There are five separate reasons for that conclusion.
55. First, there was broad agreement between Mr Smith and Mr Gow that downlighters are a common source of fire. I was referred to various research papers which indicated that, although the situation was confused, because such fires were often ascribed to electrical faults, this form of words often meant that downlighters had caused the fire. Of particular interest on this topic was the paper on the Electrical Products website called 'Sowing the Seeds of Disaster'. Accordingly, I can readily conclude that problems with downlighters are a common source of fire.
56. Secondly, that general view is supported by an analysis of the type of halogen lamp that was used in the second floor ceiling of Mead House. There are two types of halogen lamp: those with an aluminium reflector, and those with a dichroic coated reflector. The former reflect visible light and most of the heat forwards out of the front of the lamp; the latter reflects visible light forwards whilst allowing most of the heat to pass out of the back of the lamp. The Electrical Safety Council notes this difference and expressly warns in its literature that "the use of dichroic reflector lamps in downlighters designed for use with aluminium reflector lamps only could create excessive heat within the luminaire leading to an unsafe situation and risk of fire".
57. The installation sheet for the downlighters that were used at Mead House stated that they were designed for "aluminium low voltage lamps". However, the tests carried out by Mr Gow demonstrated that what was actually used were dichroic lamps. Thus, I find that the particular type of lamp used at Mead House was itself a risk of fire. It was contrary to the recommendations of the Electrical Safety Council. This therefore created, in their words, "an unsafe situation and risk of fire".
58. Thirdly, the inherent problem with the downlighters and the incorrect bulbs within them could have been exacerbated by the underlying electrical faults that existed in Mead House. I deal with those in more detail in paragraphs 89-90 below. Although none of the particular problems that had been noted appear to relate to the second floor, it was clear from the Claimant's evidence that the house was renovated (not by him) throughout, some 13 years ago, and so, to the extent that they were generic, those problems could easily have extended to the second floor. They could potentially have exacerbated the problems of the downlighters and the incorrect lamps, a potential overlap outlined by the Sowing the Seeds of Disaster article referred to in paragraph 55 above.
59. Fourthly, in support of the conclusion that the probable cause of the fire was the halogen lamps, I note that this was the considered view of the fire investigation officer in his report of 17 May 2013 (paragraph 21 above). Furthermore, this is even more important than usual because, after the fire, the officer took away two heat-damaged downlighters and some damaged insulation. Although Mrs Gosling was aware that these items had been removed, it does not appear that she ever asked for them; Mr Smith did not either. So the officer's conclusion was based, at least in part, on evidence which Mrs Gosling and Mr Smith had never even examined.
60. Finally, a consideration of why this fire happened on the morning of 24 March 2013, and not before, also supports the conclusion that the lamps were the cause of the fire. In my view, only two things were potentially different or new that Sunday morning^[2]. One was the fact that the lights in the north-eastern bedroom had been turned on for a significant period. That suggests that it was the turning on of the lights, and then their being kept on for at least an hour and a half, so that they could get hot, which may well have been the underlying event which led to the fire.
61. Of course, on this issue, I accept the evidence of Mr Goold that this bedroom was used at various times between Christmas 2012 and March 2013. There is, however, no evidence as to the amount of time that the lights were on during these visits. And although Mr Gunning QC makes the point on behalf of the Defendant that these visits would have taken place in winter time, when the lights would probably have been on for longer, it must be remembered that this was a guest bedroom, used by weekend guests. It is most unlikely that these lights would have been on for very long on any normal visit, save for a short period in the morning and a short period before going to bed. It is, I think, a reasonable inference that the period of one and a half to two hours that the lights were on during the morning of 24 March was an unusual event.
62. For all those reasons, therefore, I conclude that, on the balance of probabilities, the underlying cause of the fire was the excessive heat generated by the inappropriate lamps in the north-eastern bedroom on the second floor of Mead House.

H2.2: Combustible Materials/General

63. Of course, the lamps would not themselves have created a flaming fire: they were the underlying cause of the fire because they directed excessive heat into the roof void over a lengthy period but, for a fire to be created, something else was needed. The experts are agreed that what was required here was some combustible material that was set on fire by the heat from the lamps. What, on the evidence, was the most likely combustible material in the roof void that would have been set on fire? I deal with the alternative possibilities in descending order (i.e. the most likely first and the least likely last).

H2.3: Insulation

64. In my view, the most likely combustible material in the roof void at Mead House that was set on fire by the lamp or lamps which overheated was the insulation. It was a strong possibility that the lamps, which were the subject of the Electrical Safety Council warning, set fire to the insulation that rested on or around them in the roof void. Again, there are a number of reasons for that conclusion.
65. First, both experts referred to the insulation as combustible, and both agreed that it could have been ignited by the over-heating lamp. They agreed that the 'space blanket' insulation^[3] burns freely when ignited. Secondly, the literature notes that ignition of insulation is a well-known problem. Thirdly, Mr Gow demonstrated to my satisfaction that the damage to the insulation shown in the post-fire photographs proved that, in places, the insulation was resting on top of the back of the bulbs, although they were not designed to be so covered. Fourthly, the insulation was everywhere in the roof void; it did not require a particular combination of local circumstances to catch fire because it was everywhere, albeit in different quantities and types.
66. Finally, and perhaps most importantly of all, the fire investigation officer concluded that the cause of the fire was the "ignition of insulation material in the roof space from halogen downlighters". Of course, the mere fact that the officer reached such a precise conclusion in this case is not of itself determinative, but he had removed evidence which showed "heat damage to insulation", and his considered conclusion is powerful supporting evidence that this was the most likely material ignited by the halogen lamps.
67. Mr Sullivan argued that the insulation (or the roof timbers – see Section H2.5 below) could not have been causative because its condition had not changed since 2004. That completely misses the point: a fire can start when one ingredient or element of the situation changes, even if everything else stays the same. Here the most likely change, as the Claimant argued and I have found, was the long period that the lights were on in the north-eastern bedroom that Sunday morning. Nothing else was required to change in order for there to be a fire that day, when there had not been a fire before.

H2.4: Other Detritus

68. There is also the real possibility that other combustible materials, such as wasp nests or cluster flies were ignited by the halogen lamps and gave rise to the fire. Although I do not regard this as being as strong a possibility as the insulation, it is still a highly credible alternative.
69. It is a credible alternative because it was common ground that this was highly combustible material. Moreover, the evidence made plain that there was detritus of this kind, in large quantities, in this roof void. The Defendant said that, in his first set of visits, when he had removed a number of the downlighters, "dead cluster flies fell down through the gap with them". The Claimant confirmed that that had happened and described it as "perfectly normal". The same thing happened again on the November visit and the Defendant apologised to the Claimant. More significantly, on the December visit, the Defendant removed one of the downlighters that had felt stuck on the first visit and "the remains of a wasps' nest fell down through the hole". The Defendant recalled pointing out to the young man in attendance that there was a wasp nest right over the downlighter.
70. In addition, the build-up of insect detritus was the only other condition (in addition to the usage of the lights on the second floor of Mead House) which, in my view, was subject to change, and therefore potentially likely to give rise to a fire on 24 March rather than before. Insects and nests come and go; they do not stay in the same condition from one week to the next. Accordingly, it is quite likely that, say, a wasp nest was created over one of the lamps in the roof void in March which was not there before. It therefore offered the only other plausible explanation for why the fire occurred when it did.
71. I accept that it is a less likely candidate than the insulation because (as contrasted with the insulation), the detritus was unlikely to have existed throughout the roof void. Thus not every lamp that was removed by the Defendant resulted in a downpour of flies or other material. But I consider that, for the reasons that I have given, there was a real possibility that the detritus to which I have referred was what was ignited by the lamp.

H2.5: Roof Timbers

72. A further potential source of combustible material was the roof timbers. It was common ground that, in places, the downlighters in the second floor of Mead House had been placed too close to joists, in breach of recommendations. Of course, whether the downlighters in the north-eastern bedroom may have been placed too close to joists, we shall never know, because the evidence was destroyed in the fire.
73. Accordingly, it is a possibility that the halogen lamps ignited the wooden joists in the roof area and that this was the cause of the flaming fire. However, because Mr Gow said that he thought that this was unlikely, I do not consider it further.

H2.6: Bait Blocks

74. I consider the possibility that the halogen lamps directly ignited the bait blocks to be nugatory. I consider that the alternative case, that the bulb heated the bait blocks and made them smoulder (but not ignite), but that this smouldering in turn caused other combustible materials to ignite, is no more than a theoretical possibility, and has been raised now because it is the only way in which the Claimant can find a link between the bait blocks and the fire. The bait blocks are therefore the weakest candidate for the role of causative materials, and much less likely a candidate than the other materials to which I have already referred. There are a number of reasons for these conclusions.
75. The first and most important reason is that Mr Smith accepted in cross-examination that, if the Defendant placed the bait blocks in the way in which he described, so that they were not touching or almost touching the lamps, they were not likely to be the cause of the fire. He also accepted that, in such circumstances, the fire was less likely to be anything to do with the bait blocks at all.
76. At Section E above, I have set out how and why I accept the Defendant's evidence as to where and how he placed the bait blocks. In those circumstances, given Mr Smith's proper concessions, I conclude that the bait blocks were irrelevant to causation. On that topic I should add that, contrary to the Claimant's case, the position in respect of the bait blocks had not changed shortly before the fire; on 24 March they were just as they had been left many months before.
77. Secondly, I am entirely confident that the bait blocks would never themselves have ignited, even if they were touching the lamps. The battery of tests referred to in paragraph 22 above demonstrates that the chance of such ignition was so unlikely as to be insignificant. After all, the experts carried out a large number of tests to see whether or not bait blocks actually placed on top of or very close to the lamps would ignite. A total of twelve separate tests were carried out on this premise. Some of these created a much worse case than could have existed at Mead House, with the bait block in direct contact with the lamp and with both covered by the two different kinds of insulation located in the roof. And still not a single bait block ignited in consequence.
78. In cross-examination Mr Smith agreed that, taking the results of the tests on their own, they demonstrated that the bait blocks were "very unlikely" to cause a flaming fire. He appeared also to accept that, as a result of the test data, he could not say that it was more likely than not that bait blocks caused the fire.
79. I accept that no test can ever fully replicate the actual conditions prior to the fire, and I bear in mind Mr Sullivan's detailed attempts to distinguish between each of the tests and the potential conditions at Mead House. But that can work both ways, because some of the tests, as noted above, created a situation which was much more likely to lead to ignition than the reality. Standing back, the fact that these tests repeatedly drew a blank must mean that the bait blocks were a very unlikely material to act as the source of the flaming fire.
80. The difficulties created by the test results meant that the Claimant, and Mr Smith in particular, had to put the residual case on causation in an extremely particular way. Accepting that the bait blocks would not be ignited by the lamps, the theory was postulated that the bait blocks would be heated by the lamps to such an extent that they would smoulder, and the smouldering bait blocks would then set fire to other nearby combustible material, such as the insulation or the detritus or the timbers.
81. Of course, the first difficulty for that theory is that, as Mr Smith accepted, the insulation/detritus/roof timbers could just as easily have been ignited by the over-heating lamp in any event, without the agency of the bait blocks. In addition, as Mr Smith also accepted, this theory meant that the key ingredient or trigger for the flaming fire was the other combustible material, not the bait blocks. On either basis, the bait blocks were entirely irrelevant to causation. It was therefore put to Mr Smith that he could not conclude, even on the balance of probabilities, that the bait block played any role in the fire. At one stage Mr Smith appeared to accept that, although at another point he did not. However, it seemed to me to be the logical consequence of his admissions. There was no forensic or expert basis for concluding that a bait block had been involved in the ignition of this fire; all the evidence was the other way.
82. The high watermark of the Claimant's case (and Mr Smith referred to it repeatedly in evidence) was the discovery by Mrs Gosling of a bait block in another part of the roof that was charred (paragraph 35 above). On the face of it, that might have indicated that, depending on where it was placed, or depending on where rodent activity left it, a bait block might be charred by a hot lamp. Although it did not advance the case as to ignition, it was said to assist the alternative (smouldering) case.
83. Unfortunately, the evidence about this charred block was unsatisfactory. Mrs Gosling's preliminary report identified a particular downlighter in the blue bedroom from which she said the block had fallen when she removed the lamp. That is the Claimant's pleaded case. However, in cross-examination, it became apparent that this was wrong, and the location identified by Mrs Gosling was incorrect. No alternative case was advanced as to where the charred block might have come from. She simply could not say. In a case about a large fire-damaged roof, it was, on any view, the slenderest possible basis for any case on causation.
84. In my view, this residual case on causation did what Mr Sullivan complained that the Defendant was doing on another aspect of the case: it piled one improbability onto another. Just as the existence of one charred bait block, which on this assumption was taken from another part of the roof, did not on its own demonstrate that the Defendant had been negligent in the way in which he had installed the blocks above the north-eastern bedroom, the same charred block could not support a case on causation in the face of all of the other factors which strongly suggested another answer.
85. Of course, I only have to decide whether or not the Claimant has proved its case on causation on the balance of probabilities. For the reasons which I have given, I conclude that he has not. But in considering that issue, I have also reached a firm conclusion as to what, in the circumstances of this case, was the most likely cause of the fire at Mead House: the ignition of the insulation by the inappropriate lamps in the downlighters.

H2.7: Summary on Combustible Materials

86. In summary, therefore, if (contrary to my findings) the Defendant was negligent or in breach of contract, his breach did not cause the fire. On the balance of probabilities, the fire was caused by the use of the wrong halogen lamps which directed heat backwards into the roof void. The most likely material that was ignited was the insulation. It was also a relatively strong alternative possibility that wasp nests or other inflammable detritus were ignited.
87. In my view, it is only a nugatory possibility that the bait blocks were ignited by the downlighters because, despite the repeated tests, they resolutely refused to ignite. The alternative case, that the bait blocks smouldered and set fire to truly combustible material immediately highlighted the likely irrelevancy of the bait blocks altogether; it was much more likely that the lamps had ignited the combustible material without the dubious agency of the blocks at all.

H3: Other Causes

88. There were two other potential causes of this fire. One concerns electrical faults; the other concerns rodent activity. I regard neither of these as likely causes of the fire, so I can deal with them briefly.
89. As to the electrics at Mead House, I accept that there were faults with the electrical installation. Ironically, the principal evidence of this was the conversation between Mrs Gosling and Mrs Goold, on the day after the fire. She referred to a whole raft of electrical problems, which not only included tripping and blowing of bulbs but the receipt of electrical shocks and the identification of blue sparks. That would appear to suggest some problems with the electrical installation at Mead House.
90. In my view, this was a possible but unlikely cause of the fire. There is no evidence to suggest that (aside from the halogen lamps) there was any electrical fault at Mead House on Sunday 24 March, let alone one which caused the fire. I accept the possibility that the fact that there were problems with the downlighters in other parts of the building could have exacerbated the already risky situation arising out of the use of the wrong lamps in the second floor. Beyond that, I consider that electrical problems were an unlikely cause of the fire.
91. It is well known that rodent activity can cause fires. In particular, rats and mice eat the insulation away from cabling which causes wires to come into contact, one with another, and create shorting and the like. The experts are agreed that this is a common cause of fire.
92. Although he originally denied it, Mr Smith now accepts that there was evidence in this case of damage to cable and electrical fittings as a result of rodent activity. This was therefore another possible cause of the fire. Again, however, it is difficult to be more specific. Furthermore, I have already referred to the evidence (paragraph 7 above) which indicated that, by December 2012, the rodent infestation had been dealt with. There was nothing to suggest that, thereafter, the Goolds noticed any such problems. I infer that they would have done if there had been rats or mice in the attic because they had half a dozen guests staying on the second floor between Christmas 2012 and March 2013. If, on the balance of probabilities, there were no rodents up there during that period, there were no rodents who could have caused the fire on the morning of Sunday 24 March.

H4: Summary on Causation

93. For the reasons set out above, I consider that the fire started because the halogen lamps projected excessive heat into the roof void. This was much the most likely condition that led to the fire; it was much more likely than an electrical fault or rodent damage.
94. The lamp had to ignite something. In descending order of likelihood, I consider that the combustible materials ignited by the lamp were the insulation, the detritus, and the timber joists. The Claimant's case on combustible materials, namely the bait blocks, was the least likely of all the candidates suggested in the evidence. The Claimant's case therefore fails on causation grounds.

I: CONCLUSIONS

95. For the reasons set out in Section F above, I do not consider that the Defendant was negligent or in breach of contract.
96. For the reasons set out in Section H above, I conclude that, if I am wrong and the Defendant was negligent or in breach of contract, that negligence or breach of contract did not cause the fire. The most likely cause of the fire was the incorrect halogen lamps directing heat backwards into the roof void and igniting the insulation. The bait blocks were the least likely of all of the various candidates put forward as the materials ignited (or caused to smoulder) by the lamp. Thus, even if I was wrong as to liability, the Claimant has failed to make out its case on causation.
97. I will deal with all consequential matters when this Judgment is handed down.

Note 1   The Claimant has continued to engage the Defendant from time to time, describing him (with not a little condescension) as “a perfectly competent rodent killer”.    [Back]

Note 2   The other is referred to in paragraph 70 below.    [Back]

Note 3   One of the types of insulation in the void, and the most prevalent according to the evidence.    [Back]