Kerr v William Morrison Supermarket Plc & Anor

In October 2004 the claimant owned a Land Rover which was fuelled both by petrol and liquid petroleum gas (“LPG”) On 26 October he suffered serious injuries to his hand when using an LPG refuelling pump at a filling station owned and operated by the first defendant in Barry, South Glamorgan. The defendants are the manufacturers of the LPG pump nozzle that was used for filling cars with LPG at the filling station.

The claimant issued proceedings against both defendants, alleging that his injuries were caused by their negligence and, in the case of the second defendant, its breach of statutory duty. In a reserved judgment given on 20 March 2009, HHJ Curran QC dismissed the claim against the second defendant but found the first defendant liable. He ordered that there be judgment in favour of the claimant in the sum of £17,000. The first defendant now appeals against the finding of liability.

It was common ground that LPG is a highly volatile and inflammable gas, which requires storage and transfer under high pressure. If released rapidly from such pressure into the atmosphere, LPG will vaporise. If it contacts, for example, the exposed skin of the hand of a motorist using an LPG pump at a filling station, injuries similar to frostbite injury may be caused. There is a single LPG pump at the first defendant’s premises in Barry. The nozzle consists of a pistol grip handle with an integral trigger, which can be pulled back and released by the hand which is holding the pistol grip. There is a rigid metal inner tube, containing an outlet valve and seals, which projects beyond the grip, and an outer barrel which is designed to revolve around the inner tube to ensure secure attachment to a vehicle’s filler inlet.

The nozzle does not incorporate a splashguard which would protect the hand of a user from discharges of liquid gas. The mechanism by which the nozzle is attached to the fuel inlet pipe of a vehicle is as follows. After removal of the filler cap on the vehicle, the driver first has to ensure that the outer barrel of the nozzle is rotated fully anti-clockwise. The inner tube of the nozzle must then be inserted into the vehicle’s filler valve, which has a bayonet fitting. The barrel must then be rotated clockwise to tighten the connection. The trigger is then squeezed and latched open to allow liquid fuel to pass into the vehicle’s storage tank when the pump is operating. This is done by pressing and holding down a separate pump-mounted control button. When the filling is completed the button is released, and the trigger pressed again to release the latch. The outer barrel is then rotated anti-clockwise and the tube removed from the vehicle’s inlet valve.

The claimant’s account, which the judge accepted, was that on the day of the accident he attached the filler pipe from the pump to his car and found that there was some kind of delivery problem. He only managed to pump some six litres into the car. The pump was not working properly. He went to the kiosk and the cashier merely reset the counter on the pump. He returned to the pump and started again. He was able to pump 33.56 litres into the vehicle. He then said that he took his finger off the button. As he attempted to disconnect the nozzle he could see a lot of vapour escaping. When he pulled the trigger it did not unlatch. He tried to twist the barrel head off. All the time, vapour was continuing to come out. When the nozzle finally came off, the vapour stopped escaping. There was evidence that other users had encountered difficulties with the pump, all of which the judge accepted.

Mr Margetts said that on three occasions between the beginning of 2003 and the end of 2005 he had heard and seen gas escaping after he had connected his vehicle to the pump. He said that he thought that on each occasion he had not locked the pump properly onto the car. He unlocked it and tried again but it made no difference. As a result of his concern arising from these experiences he only used the pump if he needed to obtain a limited amount of gas in an emergency.

Mrs Evans was injured on 25 February 2005 some four months after the claimant’s accident when she was using the pump at the first defendant’s station. She turned the nozzle on the pump in order to release it and liquid gas ran all over her right hand. She could not release the pump so she tried with her left hand, but liquid gas discharged on to that hand as well. She suffered cold vapour burns to both hands as a result.

It was not in dispute that Mrs Evans had used the same pump as was used by the claimant and Mr Margetts. The first defendant’s engineer’s job sheet for work done on 25 February 2005 after a reported gas leak states: “Changed nozzle barrel; replaced nozzle latch bush.”

There has been some debate in this court as to whether what was replaced was the barrel alone or the barrel and the nozzle. The judge found that it was the latter. The engineer was not called to give evidence. In my judgment the judge’s interpretation of the job sheet is sensible and one that he was entitled to make.

Mr Birmingham and Mr Gillam were the experts instructed on behalf of the claimant and the second defendant respectively. In his report Mr Birmingham said, inter alia:

“8.01.

In my opinion the Claimant’s injury was caused by the Defendant’s failure to provide a reasonably practicable safety measure in the form of a protective shield or splash guard to protect the Claimant from a foreseeable unwanted release of LPG which caused the frostbite or ‘cold burn’ injury to the claimant.

8.02.

In my opinion, the use of a simple splash guard or gauntlet similar to those shown in photograph 4 at Appendix 1, would have prevented the liquefied petroleum gas coming into contact with the Claimant’s right forearm causing frostbite or a cold-burn injury.”

The first defendant did not instruct an expert. The experts who were instructed both noted that the barrel of the nozzle which they had inspected was not the barrel or the nozzle at the time of the accident. The equipment was therefore not the same or in the same state as when the accident occurred. Their opinions were therefore confined to the design of the nozzle. They agreed that a nozzle of this type was suitable for its purpose and would have been reasonably safe when operated in the normal manner according to the instructions displayed at the pump. They also agreed that a small amount of LPG which escapes on normal disconnection of such a nozzle from a vehicle does not represent a significant hazard. They were unable to explain from where or how LPG might have spurted out so as to injure the claimant in the way that he described. It was possible that the LPG escaped from the claimant’s vehicle and not from the nozzle. It was reasonable to conclude that the nozzle was in reasonable working order at the time of the accident. The injuries that the claimant sustained were not consistent with the normal amount of gas discharged on connection. I shall have to return to the expert evidence a little later in this judgment.

The judge found that Mr Margetts complained to the first defendant’s staff on the three occasions of which he spoke, that the claimant complained on the day of the accident and that Mrs Evans complained some four months later. There was no sign of any documentary record of these complaints nor was there any evidence that the first defendant had made any risk assessment in respect of the pump. But certain relevant documents (previously undisclosed) were produced by the first defendant at the hearing. These purported to be risk assessments made “generically” in respect of all the first defendant’s filling stations. The first is dated October 2003, but it is unnecessary to refer to that. The second document is dated November 2006. It is left blank so far as entries in respect of any particular store were concerned. It contains the following words in the column for “additional information”:

“Gas may escape if the connection is not properly made due either to improper connection made by user, fault with dispenser nozzle, or fault with car connector.”

The judge asked Mr Taylor, the first defendant’s group health and safety controller, whether the entry in this column meant that the first defendant had had experience of faulty dispenser nozzles causing escapes of gas. The judge said that he was not impressed by the answers he received. Mr Taylor said at first that the entry was “over-egging the pudding”. Then he said that he did not believe that the entry meant that there had been such faults, only that there may have been. The judge described him as a “defensive” witness. He said that the entry suggested that experience at the first defendant’s premises somewhere in the country had informed the maker of this risk assessment, that liquid gas may escape if the connection is not properly made due to improper connections as the result inter alia of faults with the dispenser nozzle.

At paragraph 38 of his judgment the judge found that the nozzle which was in situ at the time of the accident was subsequently assessed as requiring replacement as a result of the leakage or escape. If the part had been found to be in a wholly satisfactory condition and not defective, the judge said that it was inherently unlikely that it would have been replaced. The escape of gas which caused the claimant’s injury came from one of two sources, the nozzle or the vehicle. In the judge’s view it could not have come from the vehicle. There was no evidence that the inlet valve on the vehicle was in a defective state. The judge noted that the car, unlike the nozzle, was still available for inspection.

At paragraph 41 he said:

“Logically, upon the whole of the evidence, it seems to me on the balance of probabilities cannot have been discharged from the tank in the car, which had, on the evidence, a functional one-way valve which would have prevented any gas escaping from it. The probability in my judgment is that the liquid escaped from the hose, as the trigger was held open. This might foreseeably happen if the user were panicking at an evident escape of gas. On the evidence, it might occur if the nozzle failed to un-latch. That possibility is dealt with at para 17 above, and it is made more probable by the reference in the job sheet to the replacement of the latch bush at p.273 of the trial bundle.”

At paragraph 45 the judge said:

“That discharge cannot be explained by any defect in the Claimant’s vehicle’s inlet valve. It can only have been the result of a defect in the First Defendants’ equipment. If such a defect permitted the escape of more than a minimal amount of LPG it was entirely foreseeable that the Claimant would be caused significant injury. The First Defendants were the operators of LPG filler equipment provided by them for use by the public. There was no intermediary such as a petrol-pump attendant provided by the First Defendants to offer assistance in the use of the equipment, who might be familiar with any idiosyncrasies of the equipment.”

At paragraph 46 he said that the precise defect and the mechanism involved in the leak could not be identified. This was probably because the original nozzle had been replaced.

At paragraph 47 he said that he was satisfied that the equipment was defective at the date of the claimant’s accident and that it had been defective since Mr Margetts had had the first of his problems in 2003. The defect may have been such as to cause only intermittent problems in use, but it was for the first defendant to ensure that it was sufficiently well-maintained to enable customers to be reasonably safe in using it.

At paragraph 51 the judge said that the cause of the claimant’s accident was:

“…a defect of some kind with the nozzle which may have been a combination of a defective latch bush and an ineffective seal in the nozzle, thus enabling gas to be discharged from the hose even after the pump was cut off. It is in my view reasonable on the evidence for me to draw the inferences that the First Defendants had made no adequate risk assessments of hazards which could foreseeably affect customers such as the Claimant, that they failed to heed or investigate complaints of previous relevant incidents, or reports of potential defects in their equipment before the occurrence of the accident to the Claimant, and failed adequately to inspect or maintain their equipment. Such failures were the cause of the claimant’s injuries.”

The judge went on at paragraph 52 to say:

“If I am wrong about that, and the precise cause of the accident simply cannot be ascertained, I am nevertheless quite satisfied that the accident was not the result of any defect in the Claimant’s vehicle, since there is no evidence of such a defect, nor was it the result of any misuse of the First Defendants’ equipment, as I accept the Claimant’s evidence that he used the equipment just as uneventfully as he had always done. In those circumstances I would be prepared to find that the res ipsa loquitur principle is applicable, in that properly-maintained LPG equipment, such as the nozzle, which is free of damage or defect, would not permit the escape of liquid gas in such quantity as might cause significant injury. The First Defendants’ equipment did permit the gas to escape. The effective cause of the accident was probably some act or omission of the First Defendants or of someone for whom they were responsible, which act or omission constituted a failure to take proper care for the Claimant’s safety. They have not shown the accident could have occurred without negligence on their part, as I have rejected the only real alternatives suggested by them, which are based upon fault on the part of the Claimant or a defect in his vehicle.”

The grounds of appeal are: 1) the judge was wrong to allow the claimant at trial to depart from the case as pleaded and presented before the trial; 2) the judge was wrong to find that the nozzle was defective. The weight of the evidence was that the most probable cause of the excessive escape of LPG was a fault in the vehicle; and 3) the judge was wrong to apply the principle of res ipsa loquitur.

The Particulars of Negligence pleaded against the first defendant in the Amended Particulars of Claim were:

“The first Defendant their employers or agents were negligent in [that] they:

(a)

failed to carry out any adequate risk assessment of the said nozzle before installing it for customers use. It is the Claimant’s case that the guidance notes produced by the second Defendant to the said nozzle specify: ‘Keep your hands away from barrel area as some liquid gas is released (this is normal).’

The Claimant’s case is that an adequate risk assessment would have uncovered the likelihood that some liquid gas would escape when customers were filling their cars whereby at the very least a splash guard should have been fitted to the nozzle.

(b)

failed to read or take into account the said guidance notes before or when installing the said nozzle for use at their said garage premises.

(c)

failed, when they knew or ought to have known that when LPG turns from gas to liquid the substance was likely to cause frost bite damage to skin and flesh, to fit a splash guard to the nozzle used at their LPG pumps.

(d)

failed to warn the Claimant whether by means or signs or otherwise that there was a likelihood that some LPG would escape and/or that any escaping LPG was likely to cause damage to skin and flesh.

(e)

Installed and used LPG nozzles at their said forecourt premises which were unsafe for the Claimant.

(f)

In the premises operated an unsafe system of work.”

Mr Griffiths submits that the pleading concerns no allegation of mechanical defect in the nozzle. The absence of a splashguard is a complaint of a design fault not a mechanical fault. The allegation of negligence and installing and using an LPG nozzle that was unsafe was an allegation of defect in the intrinsic design of the nozzle, not an allegation that it was mechanically defective. Mr Griffiths emphasises that the pleading contains no reference to res ipsa loquitur. In short he submits that the claim was advanced on the basis that the design of the nozzle was defective by reason of not having a splashguard fitted. He submits that the general allegations in paragraph 5(e) and (f) of the Amended Particulars of Claim were reasonably understood by the first defendant as being directed only to the preceding allegations, which were concerned with the lack of a splashguard. In its defence the first defendant pleaded at paragraph 6 that it was “a matter of undisputed fact” that the LPG pump was not defective at the time of the accident. Paragraph 7 alleged that the injury was caused by the second defendant’s design failure to provide a splashguard.

Mr Griffiths submits that the defendants were meeting a claim based on the design of the nozzle having been defective by not having a splashguard fitted. Indeed even at trial Mr Birmingham did not suggest that the accident was caused by a mechanical defect of the nozzle. In his initial report he said that the injury was caused by the failure to provide a splashguard. He made no mention of a defective nozzle. Mr Gillam’s report stated that it was reasonable to conclude that the nozzle was in reasonable working order at the material time.

Mr Gillam suggested that the LPG must have been caused to escape by the manner in which the claimant used or misused the nozzle during his attempts to dispense LPG into the vehicle or that LPG escaped from the vehicle and not from the nozzle. In their joint report Mr Birmingham and Mr Gillam said that they were not able to explain how LPG might have spurted out so as to injure the claimant. They agreed that it was reasonable to conclude that the nozzle was in reasonable working order at the time of the accident and that it was possible that the LPG came from the vehicle and not from the nozzle. As evidence of how the first defendant understood the claimant’s case, Mr Griffiths draws attention to his skeleton argument dated 2 February 2009. He refers to paragraph 2.1, which states:

“In headline terms, it is submitted that the essence of [the claimant’s] case, as against [the first defendant], is that if a protective guard had been fitted between the handle and barrel, then the injury would not have been caused. …

Therefore, it is submitted that in all the circumstances, the thrust of [the claimant’s] case is that there was an element of a ‘design flaw’, in that, the design of the nozzle should have incorporated or had otherwise been fitted with a splashguard.”

Paragraph 3 of the skeleton argument states:

“The essence of [the first defendant’s] case is that:

3.1

The LPG pump was not defective in its design and/or operation.”

The heading to paragraph 11 is:

“Only if the court finds that the manufacturer should have fitted a splashguard, should the Court then go on to consider [the first defendant’s] liability.”

The claimant’s skeleton argument for the trial is dated 4 February 2009. At paragraph 5 it states:

“It is the Claimant’s case that as a result of a defect in the LPG system a significant amount of gas was allowed to escape which directly caused the Claimant’s injury. It is the claimant’s case that whether or not the precise defect can be identified there must have been a defect in order to cause him injury. …

Further or in the alternative even if the defect cannot be identified these circumstances must fall within the principle of res ipsa loquitur.”

At paragraph 6 (c) it says:

“It follows that the nozzle system of the LPG system at the First Defendant’s premises provided a foreseeable risk of injury.

Having read the claimant’s skeleton argument at the start of the trial on 5 February, Mr Griffiths raised with the judge his concerns about precisely what the nature of the claimant’s case was. He explained that the allegations of mechanical defect and the reliance on the doctrine of res ipsa loquitur were new. The judge said to Mr Griffiths:

“If you perceive a departure from a pleading in terms of opening up cross-examination … you are of course entitled to object.”

It is accepted by Mr Griffiths that the evidence did cover the issue of mechanical defect. He did not object because he was confident that the expert evidence that had been adduced on this point did not support the claimant’s case. As must already be apparent, that confidence proved to be ill-founded, at any rate so far as the judge was concerned.

In his closing speech Mr Griffiths says that he focussed on the issue of design flaw, ie the lack of a splashguard, and the question whether the first defendant had any responsibility for that. But he also dealt with the mechanical defect point. Mr Prys Lewis raised the issue of res ipsa loquitur in his final speech. During the course of argument the judge said that he did not think that res ipsa loquitur applied, since there was more than one possibility as to why the accident occurred. The judge did not give any indication that he would decide the case on the basis of res ipsa loquitur. The result was that Mr Griffiths made no submissions even in reply on that issue.

Mr Griffiths did not submit to the judge at any time during his final submissions or at all that the judge should not deal with the allegations of mechanical defects because it was not within the scope of the pleaded case. Mr Griffiths was content to deal with the point on its merits.

I accept that the allegation of negligence by reason of mechanical defect should have been specifically pleaded. On the other hand Mr Griffiths should have asked for further information of paragraph 5(e) and (f), which in my judgment were unacceptably broad, generalised allegations which called out for particularisation. Whether res ipsa loquitur should strictly have been pleaded or not, the pleading in my judgment should have set out the facts which would put the defendants on notice that the principle was in play and that the burden was on them to show that the accident could have happened without negligence on their part. In my view the Amended Particulars of Claim did not do so. But it was clear when Mr Prys Lewis’s skeleton argument was served on the first defendant that the claimant was saying that: 1) the accident was caused by a defect in the nozzle which could not necessarily be identified but for which in the circumstances the first defendant was responsible; and/or 2) res ipsa loquitur was in play. If Mr Griffiths thought that the first defendant was prejudiced by the late introduction of these two issues, he should have objected to the admission of any evidence that bore on them and should have asked the judge in his final submissions not to deal with them on the grounds that it was unfair to do so.

Mr Griffiths took neither of these courses. In these circumstances I consider that no unfairness was caused to the first defendant. Indeed Mr Griffiths has not been able to identify any prejudice that the first defendant has suffered as a result of these two issues being determined by the judge. He says that he might have called the engineer to replace the nozzle on 25 February 2005, but he has been unable to tell us what that engineer would have been able to say. In all these circumstances I would reject the first ground of appeal.

Mr Griffiths relies on the evidence of Mr Birmingham and Mr Gillam that there was no defect in the nozzle and that the LPG escaped from the claimant’s vehicle. That conclusion was based on the fact that the nozzle had continued to be used normally for a period of about four months after the date of the accident before the nozzle was changed and generally on the very extensive use of the nozzle. Mr Griffiths also relies upon the experts’ joint statement that it was “reasonable to conclude that the nozzle was in reasonable working order at the time of the accident”. I should say that, as I read their evidence, that conclusion of the experts was based entirely on the fact that the pump had been in use for a substantial period of time and only a very small number of complaints of malfunction had been recorded.

Mr Prys Lewis submits that the judge was entitled to reject this conclusion of the experts on the basis of the evidence that he heard during the trial and on the grounds that the experts had incorrectly identified the date of the replacement of the nozzle as 2 February 2005, whereas the job sheet for the replacement was dated 25 February. The significance of the later date is that it post-dated the accident suffered by Mrs Evans on 24 February. It was this error which led the judge to say at paragraph 26 of his judgment: “The foundation for the opinion may have little substance.”

The judge was impressed by the fact that there had been instances of malfunctioning of the pump. He accepted the evidence of Mr Margetts and Mrs Evans as well as that of the claimant himself. I would accept the submission of Mr Prys Lewis that the judge was entitled to find that the equipment was defective at the time of the claimant’s accident and that it had probably been in a defective state since Mr Margetts first had problems in 2003. The judge was also entitled to say that, despite the reports of problems with the pump, there was no evidence that such complaints were recorded or dealt with. In these circumstances the judge was entitled to reject Mr Gillam’s observation in his report and in the agreed statement that the continued use of the nozzle for four months after the accident, when there were no reported problems or complaints, provided a basis for the conclusion that he sought to draw that the nozzle was in reasonable working order at the material time.

As regards the suggestion that the excessive escape of gas was as a result of a defect in the vehicle and not the nozzle, the position is this. First, the joint statements of the two experts merely identified the escape from the car as a possible cause: see paragraph 7. Both experts gave oral evidence on this topic. When Mr Birmingham was cross-examined by Mr Griffiths about it:

“Q. “It is possible that the LPG liquid that injured the claimant escaped from the claimant’s vehicle and not from the nozzle.” Now that’s different isn’t it to what you said just before?

A.

Yes it is possible.

Q: Yes. Do forgive me, I thought that before in that question actually, the other paragraph 13, that it wasn’t, it is possible is it? That must be right?

A: Yes, anything’s possible but I think that I am saying that in order for that to happen, the most unlikely set of circumstances would have had to have applied, namely that the check valve in the vehicle was faulty, which would indicate that would allow liquid or vapour to escape all the time.”

As I have said, the judge was impressed with the evidence of the claimant that he had had no other such problems.

It is true that when Mr Birmingham was later cross-examined on behalf of the second defendant, he said that, making certain assumptions, he would accept that the only source for the escape of the LPG was the car: see the transcript at page 204. But the assumptions excluded the possibility of escape from the pump. It seems to me, therefore, that in his oral evidence Mr Birmingham did not support the theory of defect in the vehicle. Mr Gillam gave a number of answers on the issue of vehicle defect. At transcript page 216 the following exchange took place during the cross-examination by Mr Prys Lewis:

“Q: Yes. In so far as Mr Kerr’s vehicle, we have heard evidence that he used it before the accident and has been using it since the accident, I think he’s changed his car now, but you are aware of no evidence, are you, that the seal on the car, in other words the dispenser on the car, is defective in any way?

A: Yes, it’s not a seal, it is just the end of the tube, yes. I’ve not heard of anything.

Q: No, there is no evidence before this court, is there, that there is anything defective about the car?

A: Well I’ve not seen anything, just that it’s a possible cause of this accident.

Q: Sure, I know it is possible but there is no evidence of that is there?

A: No, it’s one of the possibilities but there is no evidence.”

In these circumstances it seems to me that the expert evidence provided a very slender basis for a finding that the cause of the escape on 18 October 2004 was a defect in the vehicle. The judge was entitled to conclude on the balance of probabilities that the cause was some unidentified mechanical defect. He was entitled to regard as highly significant the fact that, first, the claimant had had no problems of escape of LPG from his vehicle during approximately two years of use and, secondly, at least two other customers of the first defendant had had similar problems with the same pump.

I have reached the conclusion that the second ground of appeal is not made out. The judge was entitled to make the findings that he did and they were not against the weight of the evidence and, accordingly, I would reject the second ground of appeal.

That is a sufficient reason for dismissing this appeal. I do not therefore find it necessary to consider the res ipsa loquitur question.

I agree with the judgment of my Lord, Dyson LJ, and I would dismiss the appeal for the reasons which he has given.

I also agree that this appeal should be dismissed, for the reasons given by my Lord.