UK Insurance Ltd v Holden & Anor

On Saturday 12 June 2010, Mr Thomas Holden, the First Defendant and a mechanical fitter employed by the Second Defendant (“Phoenix”) was working overtime at Phoenix’s premises. The day before his car had failed its MOT due to corrosion on its underside. Having completed his first piece of work that day he asked his employer if he could use the loading bay at the premises to do some work on the car which would hopefully enable it to pass the MOT. His employer agreed. His intention was to weld some plates onto the underside of the car to deal with the corrosion.

He disconnected the car battery so there were no live circuits which the welding equipment might interfere with. He then used a fork-lift truck to push the car up on its side so that he could get at the underside. He used a grinder first to prepare the underside and then successfully welded a plate under the driver’s side. He then reconnected the battery, started the car and moved it round the other way before disconnecting it again and lifting it up once more but now with the underneath of the passenger side exposed. He started to weld, but then his phone went and he stood up to take the call. As he did so, he saw flames inside the car. What had happened was that sparks from the welding had ignited flammable material inside the car including the seat covers. The fire spread and set alight some rubber mats lying close to the car. The fire then took hold in Phoenix’s premises and adjoining premises and substantial damage was caused before it was extinguished.

Phoenix’s insurer was AXA. It has paid out to Phoenix and the owner of the adjoining property in excess of £2m. Being subrogated to Phoenix’s rights, AXA has now made a claim against Mr Holden in the name of Phoenix for an indemnity in respect of the sums it has paid out (“the Claim”). If Mr Holden has any insurance in respect of the Claim, it is only by reason of his ordinary car insurance effected with the Claimant, UK Insurance Limited (“UK”) (“the Policy”).

As a result UK has brought this action, seeking a declaration that the Policy does not cover the Claim. AXA, though Phoenix, denies this and counterclaims for the indemnity. Although Mr Holden is named as First Defendant he has played no part in these proceedings. He is not personally at risk since Phoenix has undertaken only to recover such sum (if any) as can be recovered from UK.

The trial of this action took place without any live evidence and for present purposes, the facts as recounted by Mr Holden in his witness statement dated 16 October 2010 (though signed on 9 January 2015) summarised above, can be taken as agreed. The issues which have arisen are ones of analysis and law.

Before descending into detail it is worth summarising broadly the rival contentions as follows: UK contends that the Policy insured Mr Holden in respect of third party claims resulting from an accident involving his car while being driven or used on a public road or other public place. Here, however, it was on private premises and moreover was not in any relevant sense being used; rather it was being repaired. Accordingly, the Policy does not respond.

As against that, Phoenix contends that, properly interpreted, the Policy covers accidents off-road as well. Further, repair to a car can be described as the use of it and thus the fire fell within the terms of the Policy because it was caused by such repair.

There are some subsidiary arguments, too, but they will be referred to in context below.

Clause 1a of Section A provided as follows:

“Cover for you

We will cover you for your legal responsibility if you have an accident in your vehicle and:

•

you kill or injure someone;

•

you damage their property; or • you damage their vehicle.

This cover also applies to any accident involving injury or damage caused by a trailer, caravan or vehicle which you are towing.”

Clause 2 states that such cover will also be provided for anyone whom the insured allows to use but not drive the vehicle for social or domestic purposes or anyone getting into or out of the vehicle, among others.

Clause 1 of Section G, dealing with territorial limits provides as follows:

General Exception 1 provides that the owner is not covered for the following among others:

“We will not cover any injury, loss or damage which takes place while your vehicle is being:

•

driven or used by anyone not allowed to drive it or used for any purpose not allowed by the Certificate...or Schedule or

•

driven by someone who does not have a valid driving licence or is breaking the conditions of their driving licence.

This exception does not apply if your vehicle is:

•

with a member of the motor trade for maintenance or repair;

•

stolen or taken away without your permission; or

•

being parked by an employee of a hotel, restaurant or car parking service.”

The Certificate (which forms part of the Policy) contained the following declaration: “I hereby certify that the Policy to which this Certificate relates satisfies the requirements of the relevant law applicable in Great Britain and Northern Ireland, the Republic of Ireland, the Isle of Man, the Island of Guernsey, the Island of Jersey and the Island of Alderney.”

As is well-known it is compulsory for car drivers to have third party liability insurance, and it is a criminal offence not to. The RTA 1988 (“the Act”) is the present governing statute. All motor insurance policies express themselves, one way or the other, to comply with the minimum third party liability insurance requirements of the Act. What this means is that whatever the detail of such insurance required by the Act, properly construed, the underlying policy contains such cover as well. In this case, that linkage is expressed by the certification on the Certificate set out above.

Section 145 of the Act, headed “Requirements in respect of policies of insurance_.” provides, among other things, as follows:

must insure such person, persons or classes of persons as may be specified in

The EU background to the Act is the series of Directives requiring member states to take measures to oblige the insurance against civil liability claims in respect of motor vehicles.

Article 3(1) of the First Directive (Council Directive 72/166/EEC of 24 April 1972) states that:

The Second Directive (Council Directive 84/5/EEC of 30 December 1983) provided that "The insurance referred to in Article 3(1) of [the First Directive] shall cover compulsorily both damage to property and personal injuries."

Article 1 a of the Third Directive (Council Directive 90/232/ EEC of 14 May 1990) as amended, provides:

The various Directives were then consolidated into the Sixth Directive (Council Directive 2009/103 EC of 16 September 2009) which provides as follows:

These Directives are not of direct effect in the UK but it is common ground that so far as possible, the Act must be interpreted so as to accord with them.

This clause is not altogether happily worded as will appear. First, it suggests that damage to property is only covered if it belongs to someone who has been injured or killed in the accident. Mr Eklund QC accepts that there is no basis for that limitation and the true cover is for personal injury or death or damage to someone’s property or damage to someone’s vehicle (or more than one of these). Thus the claimed damage to property here would be within the Policy if all the other elements are made out.

Second, the reference to “if you have an accident in your vehicle” is accepted as being too narrow. The Policy will respond if there is an accident caused by the vehicle even if the insured is not in it or driving it at the time for example where it is parked without the handbrake being on and it rolls down a hill into another car. Or if it is parked in a dangerous matter as a result of which another car collides with it. A better wording would be “if there is an accident involving your vehicle”. But at this point one needs to take account of the words used in s 145 (3) (a) of the Act. This refers to liability for damage etc “caused by or arising out of the use of a vehicle” [on a road] and Clause 1a must comply with this.

In my judgment therefore, the proper construction of the initial part of Clause 1a is “We will cover you for your legal responsibility if there is an accident caused by or arising out of your use of your vehicle and you kill or injure…etc”.

s145 (3) (a) requires liability insurance in respect of use of the vehicle “on a road or other public place”. Is the Policy so limited? Phoenix contends not, for the following reasons:

In response, UK contends as follows:

In the Policy, there is no express limitation to roads at all, whether in Clause 1a or anywhere else. That suggests that use of the vehicle elsewhere is covered. Mr Eklund QC really had only one answer to this which is that it would be commercially odd if the Policy was deemed to give cover which was more generous than it had to by reason of the Act. He asked rhetorically why any insurer would do that. In my judgment that goes too far. The fact that a policy has to comply with the minimum third party cover stipulated by the Act does not mean, objectively, that it cannot be more generous. If UK wanted to confine the Policy (or this part of it) to roads, it could easily have said so.

I am fortified in this approach by the decision of Burton J in British Waterways v Royal & Sun Alliance [2012] EWHC 460. There the question was whether the policy in issue covered liability for the death of the person driving the vehicle as well as third persons. There was no express limitation in the policy itself. Burton J rejected an argument that because the relevant provision of the Act had been so restrictively construed the same should follow for the policy. See paras. 31-35 of the judgment. I accept that in that case the policy at issue was already shown to be wider than what the Act required in some respects so it was perhaps easier to reject any congruence between them in other respects: nonetheless the decision shows, in my view that the fact that the policy may go wider than the Act is not to be regarded as surprising or commercially impossible. Here the difference is glaring: the Act contains in the central section an explicit limit to roads. It is not repeated in the equivalent provision in the Policy, Clause 1a.

In addition there is one indicator in the Policy itself that it is not intended to be confined to roads. That is in the Certificate itself. This states that it does not cover races, track days etc or 4 X 4 off road events. That suggests that other activities not on public roads that would otherwise be covered are not excluded.

I should record that Mr Davie QC made the additional submission that the effect of Clause 1 (b) of Section G, cited above, was to import directly, the First Directive into the Policy. I do not accept that. This provision simply lists a number of countries where the insured would have minimum cover if driving there, and some of those are listed by reference to having been approved by the Commission as being compliant. However, in the light of my finding above, this argument is unnecessary.

Accordingly I find that the Policy here covered the location of the accident – assuming all the other requirements are made out.

That conclusion makes it unnecessary to consider the case of Vnuk in relation to s145 (3). But lest the matter go further and in deference to the arguments canvassed before me, I deal briefly with the point. Unvarnished by Vnuk, s145 (3) could not have helped Phoenix because of its express limitation to roads. However it is said that Vnuk has changed all that.

In Vnuk, a reference by the Slovenian Court, the Claimant, Mr Vnuk was standing on a ladder in a barn when he was struck by a tractor with a trailer attached, which was reversing in the yard in order to position the trailer. The question actually posed was whether the expression “use of vehicles” in Article 3 (1) of the First Directive extended to the circumstances here even though it was not a road traffic accident. It appears that the Slovenian legislation akin to the Act was not expressly limited to roads although it was submitted that it had been so interpreted. The German and Irish governments submitted specifically that the compulsory insurance requirement contemplated by Article 3 (1) was limited to roads in any event.

The ECJ held first that the concept of “use of vehicles” must be given an independent and uniform EU interpretation. It then held that this concept must be understood in the light of the dual objective of protecting the victims of accidents caused by motor vehicles and of liberalising the movement of goods and persons within the EU. On that footing it cannot be taken that the EU legislature wished to exclude from protection parties injured by an accident that was caused by a vehicle “in the course of its use, if that use is consistent with the normal function of that vehicle”. See paras. 42, 49 and 56.

With regard to the instant case, the Court observed that the vehicle here seemed to be reversing for the purpose of taking up a specific position and that use seemed to be consistent with its usual function. However, that was a matter for the national court to determine. But it was possible that it was covered by the requirement of use of a vehicle which meant “any use of a vehicle which is consistent with the normal function of that vehicle”. See paras. 58 and 59.

It is true that the ECJ did not make any specific ruling about whether the ambit of the First Directive was limited to public roads, and that question was not squarely put by the referring Court. Nonetheless it must be implicit in this decision that in an appropriate case cover can extend further than where the accident happens on a road, otherwise there could have been no cover in the instant case. Moreover, and as recited by the Court, the background to the question was the submission that compulsory cover did not apply here because (a) this was not an accident occurring on a road and (b) the tractor was not being used as a vehicle anyway but rather as a machine. Moreover the Advocate General referred to it specifically, submitting there should be no limit by reference to roads: see para. 43 of his Opinion.

In the Court’s view the key question was whether the use was consistent with the normal function of the vehicle. If it was then, as here, cover was required even though private premises were involved. On that footing the Vnuk definition of “use” not merely applies to the term as employed in the relevant provisions of the Directives, but it must be read into the Act if possible. There is no difficulty about that since the Act deploys the word “use” and it is just a question of defining it.

There is more difficulty, however with any requirement that compulsory insurance may extend to places other than roads and public places as far as the Act is concerned since it expressly limits the obligation to roads and other public places. See s145 (3) (a) cited above. Mr Davie QC’s solution is to add the word “including” before “on a road”. The addition of that single word belies, however the scale of the change. It may be that extending the ambit of the Act in this way does not in practice lead to many more claims especially given the need to have the requisite use of the vehicle but nonetheless, this is potentially a significant addition to the scope of the Act and it is in a different league to importing the ECJ’s definition of “use”. While (as with the ECHR) the Court should if at all possible construe primary legislation to as to be compatible with the requirements of EU law, that does not require the Court to violate the essential meaning of the provision in question, so as to go against its “grain”. Or to put it another way, so as to “cross the line” between interpretation and amendment. See the decision of the Court of Appeal in Wilkinson v Fitzgerald [2013] 1 WLR 1776 at paras. 48-50 of the judgment of Aikens LJ. In my judgment, it is impossible to add the word “including” without clearly going against the grain of the Act – this is an amendment, not interpretation.

Since the interpretation advanced by Mr Davie QC cannot be undertaken, all I can do is to say that in my judgment s145 (3) (a) is incompatible with Article 3 (1) of the Third Directive as interpreted by the ECJ in Vnuk.

I should mention a subsidiary argument advanced by Mr Davie QC here. He said that it was not in fact necessary to amend s145 (3) (a) at all, because the operative provision was s145 (4) (b) which required not more than £1m cover for damage to property. He says that this sub-section makes no reference to roads and so when it comes to damage to property (as here) the requirement should be to insure against third party liability, whether arising on or off road. I regard that as a hopeless contention. All s145 (4) is doing is providing some further limits to the cover already mandated by s145 (3). It is not some alternative version of the primary cover required. The absurdity if it were otherwise can be shown by the fact that if correct, the Act requires more extensive cover for damage to property than for personal injury or death.

Since I have found that the Policy itself extends beyond roads, the Claim here cannot be said to be outwith the Policy simply because of where it arose. One therefore turns to the next question which is whether what occurred can be said to have arisen out of the “use” of Mr Holden’s car.

The starting point, in the light of Vnuk is that the accident must have been caused by or have arisen out of the “use” of the car by Mr Holden, in the sense that it was a use consistent with its normal function. In my judgment the question acutely raised is whether a repair to a vehicle in circumstances like this is such a use. The answer to that question not merely impacts upon the true scope of the Policy but also the Act – the concept of use must be the same for both.

It is also necessary to distinguish this threshold question as to whether there was the required use from the posterior question as to whether the accident was caused by or arose from such use.

I accept that the Policy sometimes employs the word “use”. In some cases this seems to be interchangeable with “drive”, but not always. See, for example, Clause 2 of Section A, cited above. However that does not answer the question of the meaning to be given to the word “use” and in particular, its breadth.

In my judgment, the Vnuk definition suggests some activity performed by the vehicle qua vehicle. Thus, carrying passengers or goods, transporting the driver to some destination, positioning a trailer or caravan, or parking, by way of some examples. On the other hand, sleeping in an ordinary saloon car would not in my judgment constitute such use because the normal function of a saloon car does not include providing accommodation.

But as a matter of impression and from first principles, I cannot see how undertaking a repair to a vehicle is in any sense a “use” of it. The repair (as here) may be necessary to make the vehicle roadworthy so that it can later be lawfully driven and thus used but that is a different matter. The thing being used is the repair equipment.

There is no English authority directly in point. Extensive reference was made in argument to the decision of the Court of Appeal in Dunthorne v Bentley [1999] Lloyds Rep IR 560. In my judgment, however, that case was not concerned with this question but the next one, ie causation and I deal with it below in that context.

On the question of “use”, Mr Davie QC makes two essential points:

In Gramak State Farm (1975) CanLII 427, the Claimant’s employee brought his car onto the Claimant’s premises to drill a hole in it so a trailer could be attached but while drilling the petrol tack was punctured leading to a fire which caused damage to the Claimant. Donohue J held that the expression use included the ordinary and well-known activities to which vehicles were put and this included the work being done here.

In Elias v ICBC (2002) 95 DLR (4^th) 303, a case similar to this one, the Claimant was repairing his wife’s car at his employer’s premises when a spark from a welder set the car on fire and then the building. The issue was whether this was “use” of the car. Boyle J, following earlier cases, held that the repair work “went to” the use of the vehicle and that prevention of deterioration by a family member is an integral part of use and that was so whether the repair was immediately necessary or not.

Elias was followed in Munro v Johnston (1994) Can LII 2676 where Drake J held that where a house was damaged by a fire started when part of the car’s exhaust had been removed from it and was being ground on the driveway, this was part of the use of the car. Repair work on a car was partt of its use. See also Pilliteri v Priore (1997) CanLII 12135.

The Australian cases go in a slightly different direction in my view. First, in Govt. Insurance Office v King (1960) 104 CLR 93, the car owner of the parked car had made some repair to the carburettor and in attempting to start it for some purpose (perhaps to test the repair) poured petrol onto the carburettor which caught fire and as a result one of the persons there threw away the can of petrol which caused a fire injuring a third party. The High Court held that this did not arise from the use of the car. Dixon CJ

In Govt. Insurance Office v Green & Lloyd (1966) 114 CLR 437, the vehicle was a lorry moved into position to transport away a hoist which was being loaded onto it when the hoist fell off and caused injury. The High Court held that this arose from the use of the lorry. The act of loading upon it was an indispensable step in the intended imminent transport of the hoist. Windeyer J perhaps went further in saying that use included “driving or doing something to it or with it that is incidental to its normal use” although of course he did not have the question of repair specifically in mind since it was not relevant.

In Clement v Clement (1984) 1 MVR 435, the Claimant’s daughter was injured when his car which he had driven onto his sloping driveway fell having been jacked up. Foster J held not that the work being done to the car was use, but rather that the accident arose out of the driving of the car onto the driveway and then its being parked there in a position which was likely to lead it to slip. On that footing it arose from the car’s use. He was referred to King and did not suggest it was wrong – rather that this case was different.

In Motor Vehicle Trust v Seeney (1984) 1 MVR 443 the Full Court of the Supreme Court of W. Australia held that injury caused when a man was struck by a vehicle which was being moved in the course of repairs, albeit manually and not by the engine, arose from its use. Kennedy J referred to King and said that there was no real difficulty when the car was just being prepared for use. That was not use. The difficulty comes when some part of its mechanism was being used in the course of repair or where it was being driven in the course of repair. Where the vehicle was moving on its own wheels and being controlled by the steering and brakes it was unrealistic not to regard it as being used.

In Dickinson v MVIT (1987) 163 CLR 500, the car had been parked with passengers inside it, including the daughter of the car owner. While her father was at a shop she played with some matches which started a fire as a result of which her sibling suffered injuries. The High Court held that the car was being used because it had been carrying the passengers and that trip had been temporarily interrupted to enable the driver to go to the shop. Use could include things other than locomotion.

Finally, in New Insurance Ministerial Corporation v Handford (1994) 35 NSWLR 187, the tipper tray part of a lorry had been raised to allow Mr Handford to get access to the

It seems clear to me that the Australian approach is overall, less broad than the Canadian one and in particular while some cases of repair might amount to ones where the accident arose out of the use of the car, these cases (unlike the Canadian ones) do not foursquare equate repair of a vehicle to its use without more. Moreover I do not accept that the essential reasoning of the High Court in King has been doubted in later cases or (as Mr Davie QC suggested) is somehow obsolete. The potential for all sorts of incidents which could affect third parties arising form activities involving cars was just as much there in 1960 as it has been later.

There are therefore different approaches to this question. Sometimes this may be masked where a generous approach to causation is taken so that the accident is traced back to some use which itself is uncontroversial.

In my judgment, the Canadian approach is too broad especially given that the definition of use which is directly relevant is that laid down in Vnuk. I am not therefore prepared to hold that repair of a vehicle whether to enable it to pass its MOT or for any other purpose amounts to its use without more. While it is normal, indeed necessary, for cars to be repaired from time to time, I do not accept that it is a “normal function” of a car to undergo repair. Driving it to test a repair is a different matter, as might be running the engine to test a repair or for some other purpose since at least some important part of the car is being operated in the usual way. Insofar as public policy is invoked in support of a broader definition this does not help at what might be described as “the outer limits” of the concept of use. A line has to be drawn somewhere and different legal systems may draw it in slightly different places. The paradigm examples – namely driving and immediately related activities like parking – are covered in any event and this is surely the source of the vast majority of accidents affecting third parties.

One additional argument made by Mr Davie QC is that the Policy clearly intends “use” to encompass repair because of the carve-out from Clause 1 of Exception G (cited above) for members of the motor trade who have the car for repair. I do not agree. That merely allows such persons to drive the car while with them, to give the paradigm example. It says nothing about the scope of the word “use”.

He also submitted that since the Policy required Mr Holden to keep the car in a roadworthy condition, anything he did (including this repair) which was to ensure he complied with it must constitute use. I disagree. That requirement is obviously needed however one defines “use”.

On that footing, the repair being undertaken to this car was clearly not using it. It was not being operated in any way at all but was immobile and indeed partly off the ground so that it could be worked on.

If I took that view, Phoenix’s alternative argument is that some other use of the car which is uncontroversial nonetheless caused this accident. This arises in the context of causation, to which I now turn.

The only English authority which is in point here is Dunthorne. Here the driver had parked her car on the road having run out of petrol. She ran across the road, having seen a colleague, but into the path of an oncoming car. She was killed and the driver of the other car, who was seriously injured, brought a claim. The question was whether the claim fell within the deceased’s insurance policy on the basis that the injuries were caused by or arose out of the use of her car. There was no real doubt about what the use was. It was driving and then parking up the car. The question was whether the later accident arose out of it. The Court of Appeal held that it did. Rose LJ pointed to the fact that the reason she was crossing the road was because of her recent use of the car and indeed it was conceded that if she had got out of the car to get petrol that would arise out of its use. It is worth noting that Pill and Hutchison LJJ both thought that the case was close to the line. And Hutchison LJ made clear that he was not accepting some general principle that every time a driver gets out of the car to seek assistance, that must be use of the car. That is obviously right. If a driver leaves the car at home and then later walks to a garage to buy a can of oil, in one “but for” sense, it could be said that without the car having been there and being previously used, he would not be walking to the garage but that would surely not be enough to make any incident in that journey arise from use of the car. As with what constitutes use within a particular definition, what is caused by or arises out of such use is a question of fact and degree in each case.

Here, in the alternative, Phoenix argues that the fire arose out of the use of Mr Holden’s car. Either because he had just been driving it, or because he drove it into the garage, or because he drove it in near mats on the floor which also caught fire, or because he would be driving it afterwards, and points of this kind. See paragraphs 34-37 of the Defence. This is a wholly artificial analysis in my judgment. The fire was caused by and arose out the allegedly negligent repair of the car by the use of grinders and welders without taking any precautions with regard to flammable materials in the car itself. For that reason these alternative arguments also fail.

In the light of my findings above, it has not been necessary to resolve a further issue as to whether the fire resulting from the repair constituted an “accident” at all. Had I found the requisite use and causation, I would have held that there was an “accident” here because it was a fortuitous or unexpected incident. The word cannot be confined to where the incident is a collision between two vehicles or one vehicle and a person or property. But as I say, this point does not strictly arise.

Accordingly the Policy does not respond to this particular claim and UK is entitled to the declaration sought. The Counterclaim must be dismissed. I am grateful to both Counsel for their assistance and the excellence of their submissions.