England and Wales High Court (Queen's Bench Division) Decisions

IN THE HIGH COURT OF JUSTICE
QUEENS BENCH DIVISION

		Royal Courts of Justice Strand, London, WC2A 2LL
		30th October 2003

Mr Justice Morland : Judgment

1. This litigation arises out of the rail disaster which occurred near Selby on the 28^th February 2001 and in which ten people were killed and over seventy people injured.
2. Proceedings were brought by Great North Eastern Railway Ltd and by what was then Railtrack Plc against Hart, the driver of the vehicle which precipitated the disaster. The claims were for damage to rolling stock and the rail infrastructure respectively and for consequential losses. Claims against Hart have also been made in respect of the fatalities and the personal injuries and by a commercial loser, Freightliner Ltd.
3. Hart's insurers immediately admitted liability and have so far paid out upwards of £22 million to the various claimants.
4. I have to determine a claim for contribution brought under Section 1(1) of the Civil Liability (Contribution) Act 1978 by the insurers by way of subrogation in Hart's name against the Secretary of State (now entitled) for Transport. In this judgment I shall call the Part 20 Claimant "Hart" and the Part 20 Defendant "the department".
The Rail Disaster
5. In the early hours of the 28^th February 2001 Hart was driving a Land Rover towing a trailer carrying a Renault Savannah westbound along the M62 between junctions 35 and 34. He fell asleep. His speed was about 50 to 55 miles an hour. He was travelling in the nearside lane. The Land Rover and trailer drifted over to its nearside as it naturally would because of the road's camber. Ahead was Little Heck Bridge which takes the M62 over the main North – South railway line which was perpendicular to the M62. The time was about 6.10. a.m.
6. The Land Rover and trailer left the nearside lane at an angle of about 5 to 6 degrees and at that angle crossed the rumble strip, the hard shoulder and went along and down an embankment. At the bottom the Land Rover and trailer demolished a stretch of wooden fencing marking the boundary of a field, went through the railway fence, down the cutting and came to rest partly on the North-South track
7. Hart was not seriously injured; he alighted from the Land Rover and telephoned on his mobile phone to the emergency services. Within seconds the GNER passenger train from Newcastle bound for London travelling at a speed approaching 125 mph collided with the Land Rover. It became derailed and was hit by a northbound freight train.
Nature of the Evidence.
8. The way in which the evidence was produced in this case was unusual. Over 45 ring binders containing witness statements, expert reports, voluminous documents, and authorities were presented to me. I am grateful to Lord Grabiner and Mr Burnett for directing me towards essentials.
9. Only six witnesses gave evidence orally. Cross-examination was limited with the result that only less than three days was spent hearing oral evidence. Final submissions were completed in one day but I was left with the task of weighing up the documentary material and the written statements of witnesses relied upon by the Department. These witnesses were not called because Lord Grabiner had informed Mr Burnett that he did not propose to cross-examine them. Thus their statements remained unchallenged. It therefore remains to me to give such weight to them as I consider proper.
10. I also saw some video footage and of particular value some police photographs (exhibit 53 in Hart's trial) which shew clearly the track marks of the Land Rover and trailer from the point where the front nearside wheel of the Land Rover begins to cross the grass embankment (photo 7), where the wheels on the offside of the Land Rover and trailer miss the end of the safety fence (photo 8), along the side of and down the slope (photo 9) and the demolition of the wooden fence and place of entry into the railway cutting (photo 10 and 1).
11. I also had a view of the disaster scene. Next day in court I said:- "The slope of the embankment was very much steeper than I had imagined and I was extremely surprised, almost could not believe, that the vehicles had not toppled over".
12. My layman's conclusion was confirmed in the report of the road accident reconstruction expert, Mr Sorton, dated the 11^th April 2001 who said:-

"The path of the Land Rover on leaving the motorway can be very accurately determined from a series of tyre marks which ran from a shallow angle along the embankment backing the southern side of the road.

The first point at which the vehicle struck the kerb at the edge of the hard shoulder was around 50 metres east of the beginning of the safety barrier. Logically, the first tyre to strike the kerb would have been the front nearside tyre of the land Rover.

The angle at which the Land Rover/trailer combination moved along the embankment did not vary to a great extent. This is surprising given the steepness of the slope. What is even more remarkable is the fact that the Land Rover did not roll over. This vehicle has a very high centre of gravity and is far less stable than a conventional motorcar. It may well be that the presence of the trailer played some part in preventing roll over taking place, particularly if the trailer swung round and was literally crabbing along the embankment.

At the outset the tyre marks were fairly closely grouped together consistent with the vehicle/trailer running parallel to each other and without any significant rotation. As the combination moved down the embankment the distance separating the marks increased. It is probable that the combination began to jack-knife at that point.

The position of the Land Rover at the point of impact with the GNER train can be precisely determined.

The evidence indicates that the tailgate of the Savannah estate opened as the combination came to rest. That tailgate was struck a glancing blow by the passenger train and the whole assembly was then distorted and displaced in a southerly direction.

A large proportion of the Land Rover came to rest clear of the actual railway line. The front of the Land Rover overhung the eastern rail line."

13. Lord Grabiner in the course of his submissions seemed to suggest that there were a number of witnesses whom the department should have called, people involved in the design and building of the bridge in 1974, in the reconstruction of the approach safety fence in 1993 and the repair to it in 1995. These included a Mr Wilson and Mr Hawes who were in some way involved in 1974 and a Mr Moulds, an engineer employed by Humberside County Council, who were agents for the Department in 1993. However it is for Hart to establish his entitlement to contribution. I must decide whether it is established on the evidence that is before me together with the documentary material drawing such inferences as are proper.
The Safety Fence.
14. At the time of the crash and still today the total length of "safety barrier" in advance of the railway line "proper" is 62.7 metres (i.e. 9.2 metres of ramped down safety fence plus 33.5 metres of full height safety fence plus 20 metres of bridge parapet). 53.5 metres of this length of barrier is of full height and exhibits "normal containment characteristics". (See the letter dated 1^st October 2003 from Mr Runacres, Head of Advisory Services at the Transport Research Laboratory).
15. Mr Runacres's written evidence is unchallenged. He with colleagues including Mr Macdonald who gave oral evidence before me visited the scene of the crash on the 28^th February 2001.
16. Mr Runacres in a report dated the 5^th June 2003 said:-

"The Land Rover's nearside front wheel appears to have struck the kerb on the nearside edge of the westbound carriageway at a point approximately 50 metres in advance of the terminal end of a safety fence that is connected to the parapet of Little Heck Bridge.

The vehicle then appears to have proceeded at a shallow angle to the kerb before the offside wheels struck the kerb approximately 24 metres in advance of the safety fence terminal.

From each point of contact, rolling wheel tracks through the grass on the verge identify the subsequent course of travel. These marks were very clear at the time of my site inspection and are readily observable in the data obtained by the three dimensional scanning laser system, as well as in photographs taken of the scene around that time.

The wheel tracks were remarkably straight as they crossed the verge at a shallow angle (measured as being between 5 and 6 degrees to the normal direction of travel along the motorway).

Due to its shallow angle of travel, the vehicle combination passed by the terminal end of the safety fence relatively close to it. Had the safety fence been approximately 7 metres longer or more, then the vehicle combination would have struck it.

After passing by the end of the safety fence and travelling behind it, the vehicle combination descended the highway embankment, which is approximately 3.5 metres high at this point.

The vehicle combination broke through the wooden post and rail boundary fence between the highway and adjacent private land at the foot of the motorway embankment approximately 17 metres before the intersection of that fence with the similarly constructed boundary fence of the East Coast Main Line (ECML) railway, which passes beneath Little Heck Bridge.

The vehicle combination continued through the boundary fence of the railway, dropped down into the railway cutting and finally came to rest fouling the ECML, where it was struck by a passenger train travelling south to London.

In all, the vehicle combination travelled approximately 95 metres between leaving the carriageway and coming to rest on the railway.

In my opinion and experience there is nothing at all unusual in the design and layout of the pertinent section of the M62 Motorway and there is little to distinguish it from many other lengths of motorway and/or many other motorway bridges over railways elsewhere in the UK.

When considering the 6 year PIA data for the westbound carriageway between Marker Posts 147/0 and 148/0 only (i.e. the 1 kilometre length including Little Heck Bridge, which is adjacent to Marker post 148/1), it can be seen that no PIAs are recorded.

Therefore, it can be concluded that the PIA rate for the section of the westbound carriageway of the M62 Motorway around Little Heck Bridge during the 6 year period up to 31 December 2000 (i.e. up to two months prior to the material incident occurring) is significantly less that the national average for motorways and, in my opinion, does not warrant any cause for concern. "

17. The type of safety fence at Little Heck is constructed of sections of open box beams (OBB). If as contended for by Hart the safety fence had been longer up to 100 metres, it would have been constructed of Tensioned Corrugated Beam (TCB) certainly after 10 metres of OBB after connection with the parapet.
18. The relevant Department of Transport standard is TD 19/85 which superseded earlier technical memoranda etc such as H9/71, H9/73 and the Lay-out of Roads in Rural Areas either wholly or in part. Material parts include;-

"Definitions, types and uses of Safety Fences and Barriers

3.1 The objective of installing safety fences and barriers within a highway is to reduce the consequences of vehicles leaving the carriageway and entering areas where it would be unsafe to travel.

3.2 A safety fence is intended to absorb some of the energy impact caused by an errant vehicle striking it and to redirect the vehicle within a narrow angle to follow the line of the fence so that it does not gyrate or overturn

A safety barrier is intended to provide containment without significant deflection or deformation under impact, and to redirect errant vehicles along the line of the barrier in the direction of traffic movement.

3.3 Tensioned Steel Beam.

3.3.1 Tensioned Corrugated Beam (TCB).

3.3.1.1 This consists of "W" section beams, used single or double sided, attached to "Z" section posts by shear bolts and tensioned between anchorages. These safety fences may also be attached to angled brackets. This type of safety fence shall be used on central reserves and where other than short lengths are necessary on verges.

3.3.1.2 A TCB safety fence shall not be used:-

(a) Where the length of fence between anchorages is less than 45m,

3.4 Untensioned Steel beam

3.4.1 Open Box Beam (OBB)

3.4.1.1 This is a trapezoidal section beam, 150mm x 200mm and may be used either single- or double-sided attached to stronger "Z" section posts than those used in tensioned systems. OBB safety fences shall be used for providing protection at obstructions over short lengths and/or where space for deflection is limited.

Criteria for Provision of Safety Fences and Barriers.

4.1 General.

4.1.1 Safety fences shall be installed on all new trunk roads where speeds of 50 mph or above are allowed and for which the circumstances described in Paragraphs 4.2, 4.3.1 and 4.3.2 herein apply.

4.1.2 Where there are exceptional local hazards, involving either layout or the roadside, safety fences may be needed as described in Paragraphs 4.2 and 4.3.2.

4.2 Verges

(a) On embankments 6m or more in height;

(b) On other embankments where there is a road, railway, water hazard or other feature (e.g. a subway entrance) at or near the foot of the slope"

19. The important standard applicable at the time of the construction of Little Heck Bridge was B.E.5. of which 201(vi) says:-

"To prevent direct impact between a vehicle and the end of the vehicle parapet facing the traffic on the nearside, a safety fence shall be provided on each approach towards the bridge.

The safety fence shall be at least 100 ft long and shall continue the line of the traffic face of the parapet."

and of which 205(ii) says:-

"Where no effective alternative arrangements exist to prevent a vehicle from leaving the bridge approach and falling onto the railway below, a safety fence shall be provided on each approach towards the bridge, as in the cases of Group P.1 and P.2 parapets."

20. It should be noted 201(vi) prescribes a minimum length of "at least 100 feet long" i.e. 30 metres. It is a matter for the judgment of the designers and engineers as to whether the safety fence should be longer having regard to the site topography and the hazard ahead of the main railway line.
21. Technical memorandum No. H9/73 also applied at the time of construction of the bridge.

"Criteria for the Provision of Safety Fences.

2 Safety fences are normally recommended only on major roads where speed restrictions permit 80km/h (50 mph) or above and for which the circumstances below apply

A. Verges.

a. On embankments 6.0m or more in height.

b. On embankments where there is a road, railway, river, or abnormal hazard at the foot of the slope.

7. Tensioned beam safety fences are to be used in preference to untensioned types except: -

a. Where the length of fencing is less than 45m between anchorages

There are 2 approved types: -

1. The tensioned corrugated beam (TCB) developed by the Transport and Road Research Laboratory."

22. It was agreed by the experts that the motorway bridge was a "hazard of exceptionally high consequences" and that at least the standard prescribed by 201(a)(vi) in B.E.5 had to be provided. The difference in wording between "at the foot of the slope" and "at or near the foot of the slope" is immaterial.
23. The 1968 Ministry of Transport manual "Layout of Roads in Rural Areas" makes this clear.

"3.17 Safety Fences.

Safety fences are needed where it would be particularly hazardous for vehicles to leave the road…

They are necessary on both sides of roadways on embankment, where the height is 20ft or more, or where there is a road, railway or river at the foot of the slope…..

Safety fences should be designed to redirect a colliding vehicle without overturning, and to minimise the risk of rebounding with obvious danger to following traffic...

It is important that the ends of safety fences should be carried into the ground or turned away from the line of approaching traffic."

24. An earlier circular to Divisional Road Engineers No 23/60

"Guard Fences

1. It has now been agreed that at points of special hazard, guard fences (which should not be confused with pedestrian guard rails between carriageway and footpath) maybe provided…

4. Guard fences may also be used at points of special danger, e.g. where there is a road, railway or river at the foot of the embankment,…

5. The precise length of guard fence should be decided from site inspection but the installation should commence sufficiently in advance of the hazard to give protection to vehicles leaving the road at a narrow angle. Short lengths of guard fence should be avoided."

25. On behalf of Hart Mr Hodgson was called as an expert witness. He was a man of over 30 years experience as a civil engineer in the fields of highways and transportation but not particularly of motorways and motorway bridges.
26. This passage occurred during his cross-examination by Mr Burnett:-

"Q. You do not mention in your qualifications and professional experience any direct experience of designing bridges over railways?

A. No, I do not have any direct experience of designing bridges over railways.

Q. Or safety fencing leading up to them?

A. I have been involved in the design of safety fences on a number of occasions, but just over the years, as I have said, my involvement in detailed design has diminished.

Q. But my question was whether you had any direct experience of designing safety fences leading up to bridge parapets.

A. Yes, I have done that but not on a trunk road.

Q. Have you at any time been concerned to look directly at the standard applying to bridges and bridge parapets, that is BE5 and its successor?

A. I think this is the first time that I can recall of where it has been particularly associated with the approach of a bridge parapet."

27. A very large measure of agreement was achieved at meetings between Mr Runacres, Mr Macdonald, the department's experts and Mr Hodgson as can be seen from their memorandum of agreement of July 2003. (Bundle C1 pages 104 – 109) and in their Supplementary Joint Statement (Bundle C1 pages 327-334) dated the 29^th August 2003.
28. The sticking point between the experts has been Mr Hodgson's contention that that the approach safety fence to the bridge was too short. That contention's validity is crucial to Hart's claim for contribution.
29. My approach in answering the all – important question must not be that there is negligence because a longer fence would have prevented Hart driving on to the Railway Line, therefore the department was negligent or that because some prudent highway and bridge engineers would have had a longer fence. Negligence is only established if the length of safety fence actually provided was dangerously short or in other words by the shortness of the safety fencing the department did not exercise reasonable care not to expose unreasonably those on or whose property was on the railway line ahead.
30. See for example: Baxter .v. Stockton-on-Tees Corporation [1959] 1Q.B. 441 at page 459 – per Jenkins L.J.

"As to the hypothetical case against the county council, there is, as we have said, authority for the proposition that a highway authority constructing a road for the public use under statutory powers owes a duty to the public to take reasonable care to construct the road properly, so that it will be reasonably safe for the purposes for which it is intended to be used"

and at page 460

"It would not have sufficed for the purposes of a charge of negligence against the county council merely to show that the system of traffic signs or the lighting arrangements might have been improved upon."

31. In testing the validity of Mr Hodgson's contention the fact that as Mr Burnett submitted he did not suggest what length the safety fence should have been in either his first statement dated the 5^th June 2003 or in his second report dated the 15^th August 2003 raises serious doubts. It was not until the 1^st October 2003 that Mr Hodgson in a letter to his instructing solicitors with an accompanying plan suggested "that the safety fence provision ought to have been in the order of at least 100 metres long to afford a reasonable level of protection". He does not appear to have postulated any suggested length of safety fence at meetings with the department's experts. When cross-examined he described it as a "rough and ready assessment". If contrary to the general opinion and experience of other experts his suggested 100 metres length is roughly right, the minimum standard of 30 metres which has stood for over 40 years is wildly wrong and almost certainly several hundreds of bridges not only on motorways but also on major roads should have largely increased length in approach safety fences. I do not accept Mr Hodgson's contention.
32. What is a reasonably safe length of safety fence is as Lord Grabiner rightly put it "site specific". It is not just a question of applying the 30-metre limit unquestioningly. There has to be a site visit to assess the hazard and the topography. Although there is no evidence of a site visit in 1974 or before, in my judgment the probabilities are that one did take place before the length of the fence was finally determined and a site specific decision taken that the minimum standard was appropriate. That was the recognised procedure as can be shown by the variety of lengths of such fences in the U.K. (See Bundle C.1. page 66). Mr Barton whom I found to be a most fair and impressive witness said:-

"Q. Would you agree that the precise length of the safety fence must be determined by the prevailing circumstances at the individual location?

A. Yes, indeed.

My experience really starts in the early 1970s and certainly we would not design a bridge in my experience, since I have been involved, purely by reference to drawings in the office. That is not my experience"

33. The site has to be visited for the twofold purpose of assessing the length of the fence having regard to protection of the parapet and the nature and degree of the risk from vehicular egress.
34. Assuming that it is right for me to infer that there was an exercise of judgment as to the length of the approach safety fence in 1974 by professional highway and bridge designers and engineers, I agree with Mr Burnett's submission that the Bolam criteria apply and the fact that Hart's expert Mr Hodgson would have had the fence's length "in the order of at least 100 metres" does not establish that the professional judgment made in 1974 was negligent.
35. I prefer the evidence of Mr Runacres and Mr Macdonald to that of Mr Hodgson. They considered that the length of safety fence prescribed "fully complied with requirements of Standards and followed good engineering practice, given the prevailing site circumstances".
36. In reaching this conclusion I have considered the correspondence between Mr McMurdo of British Rail and Mr Elliott of the Department of the Environment in August 1974 following an incident near Leyland when a car left the M6 on approach to a bridge and ended up on a four line track.
37. Mr McMurdo wrote on the 6^th August 1974:-

"I have expressed my concern over the adequacy of the containment standards for crash barriers and bridge parapets where motorways and highways run close to, or cross over, railway tracks.

At Leyland, a car travelling along the M6 left the roadway, travelled down the motorway embankment and across the high speed electrified main lines before coming to rest on the slow lines. The point where the car left the road was some 60 yards short of the crash barrier erected at bridge No. 95 between Euxton Junction and Leyland Station. The incident, which, as you will appreciate, could have had very serious consequences, has raised the whole question of the extent of the crash barriers. So far as I am aware, there are no statutory obligations placed upon the D.o.E in respect of the provision of crash barriers which is apparently left to the discretion of individual highway engineers. I should be grateful for information as to the standards which are applied or the guidance given to these engineers to ensure a consistently acceptable level of protection."

38. Mr Elliott wrote on the 30^th August 1974:-

"The extent of the safety fence is a matter for the decision of the responsible Highway Engineer and the recommendation quoted as you imply, is advisory and not statutory."

39. Although I appreciate that nothing followed this correspondence and that the Leyland incident was almost certainly not brought to the attention of those responsible for Little Heck Bridge I do not consider that evaluation of the Leyland incident assists me. The layout at Leyland was very different from that at Little Heck. The M6 and rail lines crossed at an angle of about 30 degrees. The decision as to an appropriate length for the approach safety fence is site specific, a matter of professional judgment having regard to both the minimum standard and assessment on site inspection.
The Department's Duty.
40. Mr Burnett Q.C. for the department submitted that the department owed no duty of care to prevent the egress of a vehicle from the carriageway to avoid either physical injury or property damage to those off the highway.
41. He suggested that as a matter of law no such duty could arise. He said that support for that submission was to be found from the fact that there is no reported decision of a claim for such damage being recovered from a highway authority in such circumstances. Lord Grabiner Q.C. confirmed that a search of cases in other Common Law Jurisdictions revealed no relevant case until when I was writing this judgment he discovered Harbinson v Department of the Environment for Northern Ireland [1983] 9 NIJB, a decision of Lord Lowry, when Chief Justice of Northern Ireland, sitting at first instance. It was the trial of a preliminary issue on extraordinary facts.
42. A number of youths pushed a large heavy cylinder from the public highway from a roundabout into the infant plaintiff's garden. The cylinder struck her causing her severe injuries. The D.O.E., the Highway Authority, unsuccessfully sought to have the claim dismissed. Lord Lowry accepted the contention of Mr Montague, counsel for the plaintiff. At page 4 he said:-

"Mr Montague also contended that the failure to erect crash barriers at a roundabout would be evidence of negligence by the DOE, since it was foreseeable that road users could leave the road at this point. Once that factual proposition is accepted (and if the presence of a suitable barrier would have prevented this accident), the difficulty of foreseeing the precise way in which the injury was caused does not bar the plaintiff: Hughes .v. Lord Advocate [1963] AC 837, Harvey .v. Singer [1990] SC 155. Although this alleged failure is a fault of omission, the highway authority would not, if otherwise found negligent, escape under the cloak of non-feasance, provided it was responsible for introducing the roundabout."

43. Since receipt of Lord Grabiner's letter dated the 20^th October 2003 I have considered Mr Burnett's letter to me dated the 20^th October 2003. I am not persuaded that the views which I expressed during the hearing and which I now definitively set out as to the potential extent of a highway authority's duty are not correct.
44. Despite the absence of authority other than Harbinson I see no reason why in an appropriate factual situation a highway authority should not owe a duty of care and be liable in damages to those neighbouring the highway if the effective cause of the egress of the vehicle inflicting the damage was a danger on or adjacent to the highway created by the highway authority. Hypothetically there could be cases where a vehicle left the highway and caused damage without any fault on the part of the driver the effective cause being for example a dangerous mal-alignment of the carriageway or dangerously insubstantial bridge parapets or approach safety fencing. If such dangers were created by the Highway Authority, in such a situation there is no reason of policy why the law should not impose a duty of care on the Highway Authority not only to users of the highway but also to those who are or whose property is on neighbouring land. If the highway authority were not liable to those off the highway, such people could not recover the loss that they had suffered. In such a situation the imposition of the duty of care would accord with the speeches in the House of Lords in Caparo Industries Plc v. Dickman [1990] 2AC 605.
45. I would go further and say that, even if the vehicle leaves the highway partly through negligent driving and partly from a danger negligently created by the Highway Authority and causes damage to someone off and beyond the highway, that person could recover from the highway authority. If in Levine v. Morris [1970] 1 WLR 71 the facts had been slightly different with Mrs Levine, not a back seat passenger in the car, but a woman out walking her dog in a neighbouring field and the car after hitting the columns of the road sign had entered the field and struck Mrs Levine, in my judgment Mrs Levine could have recovered not only against the defendant driver but also the Ministry, as Widgery L.J. said at page 79F:-

"All motorists are guilty of errors of one kind or another, and I think it would be quite unreal if roads were designed on the assumption that no driver would ever err. Indeed, as Lord du Parq put it in London Passenger Transport Board v. Upson [1949] AC 155, where the quotation appears at p.176: "A prudent man will guard against the possible negligence of others when experience shows such negligence to be common"."

46. In the present case the department erected the safety barrier on the approach to Little Heck Bridge to guard against two eventualities; collision with the parapet and egress from the carriageway onto the railway line. If such eventualities occurred, in almost all cases they would be caused by negligent driving although they might be caused by a sudden heart attack, stroke or other non-negligent loss of control of the vehicle.
47. In my judgment at common law in building Little Heck Bridge on which the M62 motorway was to be carried and under which there was a main line railway track the department was under a duty to take reasonable care that not only the users of the motorway but also people and property who could foreseeably be on the railway track would not be exposed to an unreasonable risk of injury.
48. Clearly leaving a bridge parapet unprotected by any length of approach safety fencing or barrier would be a breach of that duty for two obvious reasons. The first is that a damaged parapet may collapse onto the rail track below. The second is that vehicles for a variety of reasons are driven off the carriageway and if that occurs either on a bridge with a damaged parapet or on or near the approach to a bridge vehicles may end up on the rail track with potentially horrendous consequences.
49. It then becomes a matter for the professional judgment of highway and bridge designers and engineers to determine what the length of the approach safety fencing or barrier should be.
50. Not surprisingly there are prescribed minimum standards, which have remained substantially unchanged from at least 1965.
51. The whole consensus of the evidence before me was that the approach safety fence had a twofold purpose; the protection of the bridge parapet and the prevention of vehicular egress onto the railway below.
52. Again the consensus was that the determination of the length of the approach safety fence was not a blind adoption of the minimum standard but an exercise of judgment following a site inspection.
53. The number of railway bridge approaches in the U.K. is 585 of which 356 have a range of lengths of safety barrier of less than 100 metres. The U.K. minimum standard length of 30 metres is as high or higher than European country standards other than Switzerland. It follows that if Mr Hodgson's rough assessment of 100 metres should be the criterion, the vast majority of bridges in the U.K. and Europe pose an unacceptable danger.
54. Another factor which in my judgment I can legitimately take into account in deciding whether Little Heck bridge with its 33.5 metres of full height safety fence was and remains a danger denoting negligence is risk assessment. I appreciate that this factor is ex post facto, that there was no formalised risk assessment in 1974, that current risk ranking tools may be primarily designed to determine priorities for works within financial constraints and that the data of past accident history is unreliable especially as it only details personal injury accidents recorded by the police. Nowadays risk assessment is a professional expertise. In effect it is recognised by statute (see Section 39 of the Road Traffic Act 1988). It matters not for this purpose that this Section does not apply to the department itself or its Highway Agency.
55. The evidence in relation to risk assessment was put before me in written form by the department. It was unchallenged. Although an assessment had been undertaken on behalf of Hart by a Dr Walker, his evidence was not adduced. The evidence or its conclusions are set out in the statement of Mr Barton who works in the Highway Agency H.Q. dated the 27^th March 2003. He said:-

"The risk-ranking tool developed for use on high-speed roads was used to rank the M62 Little Heck Bridge. It was decided to ignore the protective gabion walls alongside the toe of the embankment for the purpose of the ranking. The Highways Agency's Maintaining Agents, WSP Civils, have reported a risk-ranking score of 62 for this site. The low score, well below the agreed advisory threshold score of 70 (see paragraph 6.12 above), confirms that the scale of the Selby incident and the extensive media coverage which has raised the public expectation of some additional preventative measures being required at this site, the Highways Agency has decided to subject the site to a detailed site-specific risk assessment. Notwithstanding the low score of the risk-ranking tool the site was reassessed using the assessment procedure described above. The results did not justify any additional preventative measures either in safety or cost terms. The risk assessment procedure lists bands of scores based on the risk-ranking score set against a guideline of the indicative expenditure, that is, the spend that could typically be justified for a site with a score in that band. The guidance suggests that for sites other than those with very high risk-ranking scores, of 100 or more, expenditure should be assessed alongside other road and rail safety needs. For sites with risk-ranking scores between 80 and 99 the indicative expenditure that might typically be justified ranges from £200 to a maximum of £5000 at any one site. Typically for sites scoring less than 80 there is no indicative expenditure justified. However, the notes for guidance encourage engineers to use their discretion and go beyond the suggested indicative expenditure if there is a simple, cost effective safety benefit that could be provided.

The risk ranking process has confirmed the findings of the Highways Agency Working Group (HAWG) and the Health and Safety Commission's Working Group (HSCWG) that the M62 Little Heck Bridge site was typical of a site with a very low risk of a vehicle leaving the nearside and accessing the railway. The HSCWG commented in their report that the risk of a death caused by a vehicle leaving any road and being struck by a train was small compared to other railway risks and tiny compared to other road risks. The accident was caused by an extremely unlikely series of events that was initially triggered by a driver who fell asleep whilst driving. The incident has served to identify a far more significant problem at sites where single carriageway local roads cross railways and where accident records demonstrate the risk of incursion is appreciably greater than that for trunk roads and motorways.

The risk-ranking and further assessment of the Highways Agency's railway bridge sites was carried out as a direct result of and in the aftermath of the Selby Rail crash. There is no evidence of any particular concern being expressed about the safety of the Highways Agency's railway bridge sites or any specific accident data that pre-existed these tragic events. I believe that the results of the risk-ranking and second stage risk assessment procedure that have been used to assess the risk of errant vehicles accessing the railway line at Little Heck Bridge demonstrate that there is no justification in safety or cost terms for any additional protection measures. Therefore, it follows that the current length of safety barrier in advance of the parapet reduces the risk of an errant vehicle accessing the railway line to an acceptably low level for both road and rail users. Any additional protection measures would have to be justified on the basis of the need to allay understandable public concern following the tragic events at this site."

56. Much of the unchallenged evidence is highly technical on such matters as cost benefit analysis and probabalistic analysis given by economists, mathematicians and statisticians.
57. Among the statements read were the following:- Kate McMahon an economist in the Department of Transport, who concluded.

"there would be an overall reduction in safety if funds were diverted from other safety measures to safety barrier extensions with a lower NPV" (Net Present Value)"

58. Richard Bell, the Regional Structures Engineer for Network Rail's Eastern Region, who said this about Little Heck Bridge (bridge 8a):-

"At various stages throughout the development of the risk ranking tool trials took place at several overbridge sites to check that the tool was providing meaningful results when comparing one bridge site with another. The tool was applied to Little Heck motorway bridge and the risk at that site was found to rank at or near the bottom of the list when compared with other bridges. For example, I applied version 3 of the risk-ranking tool to 25 sites in North Yorkshire and summarised the results on 3 September 2001. Bridge 8A came 24/25 in risk ranking. I then also applied the Durham County Council developed tool (as it then was) to the same 25 bridges and Bridge 8A came out bottom of the list. In October 2001, version 7 of the combined tool was applied to the same 25 bridges plus 3 Durham sites to come up with a ranked list. On that occasion, Bridge 8A came out 25^th of 28. I consider this to be consistent with my judgement of the relative levels of risk at a large number of overbridge sites that I have seen. Even if the risk ranking tool had been in use prior to the accident at Great Heck, I do not believe that any physical works would have been deemed necessary at Little Heck motorway bridge."

59. Peter Stanton, the Safety Risk Strategy Manager of Railtrack, said:-

"During 2001, Railtrack, in conjunction with the County Surveyors' Society (CSS) (represented by Durham and North Yorkshire County Councils), developed a tool for the ranking of rail overbridges according to incursions risk. This tool. which comprises a set of 14 questions about the road, rail and road/rail interface elements of the bridge site, is currently being applied to all rail overbridges on the national rail network by teams from Network Rail and local highway authorities.

These bridges that achieve a score of 90 or more in the risk ranking exercise will be subjected to a second stage assessment to identify the costs and benefits of risk mitigation options. The figure of 90 was chosen after some research was undertaken by Railway Safety, indicating that 98% of the risk from incursions occurred at sites scoring 90 or more. The highest figure that I am aware of for any bridge risk ranked to date is 121. I understand from WS Atkins that the Highways Agency have provided a ranking score for Bridge EC2M/BA at Heck of 62, which suggests that no further work is required. The individual factor scores for factors 1 to 14 are 6, 1, 1, 6, 4, 1,5, 1, 1, 1, 2, 16, 5, and 12. From factor 5, it appears that the bridge has been scored with reference to the gabions put in place by the Highways Agency after the accident. The score for factor 5 could not exceed 10, so that an increase to the maximum of 10 for this factor would give a total of 68, still well below the level where we would be considering action was required."

60. The Gabions are square wire cages filled with stones placed in the area where Hart demolished the field and railway boundary fences.
61. Andrew Evans, Professor of Transport Safety at University College London, whose report was agreed, said:-

"The principal numerical conclusions in this report are that in Great Britain as a whole there are estimated to be an average of about 0.1 rail occupant fatalities and 0.5 road vehicle occupant fatalities per year in collisions between trains and road vehicles not at level crossings. This risk is distributed over many thousands of sites where such collisions are possible. Some sites will present greater risks than others, but it is clear that the risk at any one site must be very small."

62. James Catmur, a specialist in safety management and risk assessment, said:-

"It is my opinion that overbridge EC2/8A, with the level of protection provided at the time of the Great Heck accident, poses a low risk relative to other overbridges on the network".

"Overbridge EC2/8A has been allocated a low relative risk score using both the original Defendant risk ranking methodology and my revised version (see Section 7.3 and 7.4.2)."

"The risk at overbridge EC2/8A is less than one hundredth of the risk at overbridges where action is recommended (based on my version of the Defendant scoring methodology – see Section 7.4.1)."

"In my opinion, the site would not have been identified for further action as a result of more detailed risk ranking;"

"And therefore no further action would have been taken at this site."

63. Having considered all the evidence and the oral and written final submissions of Mr Burnett and Lord Grabiner, in my judgment negligence has not been established against the department in 1974.
64. Having reached the conclusion that the department was not negligent in failing to have a longer approach safety fence to Little Heck Bridge on the construction of the bridge in 1974, the consequence is that many of the issues debated do not need determination. However in deference to the submissions of Counsel I shall deal with them shortly.
Stovin .v. Wise [1996] AC 923.
65. Mr Burnett submitted that a duty of care in respect of the failure to exercise the statutory power under Section 67 of the Highways Act 1959 could only arise if Hart could show that the failure was irrational and the policy of the Act calls for the payment of damages (see Larner .v. Solihull Metropolitan Borough Council [2001] RTR 32).
66. He pressed in aid the passage of Lord Hoffmann at page 958:-

"In my view the creation of a duty of care upon a highway authority, even on grounds of irrationality in failing to exercise a power, would inevitably expose the authority's budgetary decisions to judicial inquiry. This would distort the priorities of local authorities, which would be bound to try to play safe by increasing their spending on road improvements rather than risk enormous liabilities for personal injury accidents. They will spend less on education or social services. I think that it is important, before extending the duty of care owed by public authorities, to consider the cost to the community of the defensive measures which they are likely to take in order to avoid liability. It would not be surprising if one of the consequences of the Anns case and the spate of cases which followed was that local council inspectors tended to insist upon stronger foundations than were necessary. In a case like this, I do not think that the duty of care can be used as a deterrent against low standards in improving the road layout. Given the fact that the British road network largely antedates the highway authorities themselves, the court is not in a position to say what an appropriate standard of improvement would be. This must be a matter for the discretion of the authority. On the other hand, denial of liability does not leave the road user unprotected. Drivers of vehicles must take the highway network as they find it. Everyone knows that there are hazardous bends, intersections and junctions. It is primarily the duty of drivers of vehicles to take due care. And if, in the case of Mrs. Wise, they do not, there is compulsory insurance to provide compensation to the victims. There is no reason of policy or justice which requires the highway authority to be an additional defendant."

67. In my judgment the dicta of Lord Hoffman have no application to the situation in 1974 when the department built the bridge. The question is whether the department owed a Common Law duty of care when the department built the bridge carrying the M62 over the main line railway to those on or whose property would be on the line to prevent egress of vehicles from the motorway when it was operational. The main railway line presented an exceptionally high hazard. In my judgment if the department had erected no approach safety fence or an unreasonably short safety fence, having regard to the foreseeably catastrophic consequences of the egress of a vehicle onto the railway line the department would have been in breach of duty. In my judgment Stovin .v. Wise has no application where the public authority has created the danger. The department would have been liable if I had found that the department had been negligent in providing the approach safety fence of the length that it was at the time of Hart's egress from the M62.
68. I accept Lord Grabiner's submissions on Stovin .v. Wise. That case decided that the failure of the Highway Authority to exercise its statutory powers to require the adjoining landowner to take the necessary action to get rid of the dangerous bank could not be used to found an action in tort for negligence. As Lord Grabiner pointed out Stovin .v. Wise was not a claim brought in common law negligence.
69. Lord Nicholls in his dissenting speech said at page 929:-

"The starting point is that the council did not create the source of danger. This is not a case of a highway authority carrying out road works carelessly and thereby creating a hazard".

Lord Hoffmann said at page 943E:-

"The judge made no express mention of the fact that the complaint against the council was not about anything which it had done to make a highway dangerous but about its omission to make it safer."

70. The Court of Appeal considered Stovin .v. Wise in Kane .v. New Forest District Council [2002] 1WLR 312. Simon Brown L.J. said at page 318

"28 Here, by contrast, the starting point must surely be that the defendants did create the source of danger. They it was who required this footpath to be constructed. I cannot accept that in these circumstances they were entitled to wash their hands of that danger and simply leave it to others to cure it by improving the sightlines. It is one thing to say that at the time when the defendants required the construction of this footpath they had every reason to suppose that the improvements along The White Cottage frontage would ultimately allow it to be safely opened and used: quite another to say that they were later entitled to stand idly by whilst, as they must have known, the footpath lay open to the public in a recognisably dangerous state."

71. I accept this submission of Lord Grabiner:-

"The footpath was not dangerous per se. The danger lay in conjunction of the end of the footpath and a blind bend on to which it exited, just as in the present case the danger lay in the conjunction of the motorway and the railway line. In both cases the danger had been created by the relevant public authority."

72. As I have decided that Common Law negligence has not been established against the department when the bridge was built in 1974, the Stovin .v. Wise point does not arise.
1993/1995
73. Lord Grabiner submitted that, even if Common Law negligence was not established in 1974 at the time of the creation of the bridge, it was established in 1993 and/or 1995 having regard to accidents which had occurred to the parapet and the approach safety fence and because in those years substantial reconstruction or repair to the approach safety fence had to take place. Here were opportunities to extend the fence with 45 metre lengths of TCB which would have added little to the costs of the works.
74. I do not accept that the accidents indicated that the approach safety barrier was dangerously short. There was no evidence that anyone had in fact driven off the motorway before the start of the fence although that was always a foreseeable possibility.
75. Lord Grabiner relied on part of a "Principal Inspection Report" on "Little Heck Railway Bridge" carried out by the Humberside County Council in August 1992. The full highly detailed report of 40 pages is to be found in Bundle H2.

"There are also areas in the carriageway face of the south parapet where the coating has been "pushed" outwards by the concrete (areas 25mm diameter approx.). This could be due to the continued expansion of the gelatinous silica gel disrupting the concrete (see photo ref. K). The situation does not appear to have deteriorated significantly since the previous principal inspection.

The connecting plates of the safety fencing to the parapets are all severely corroded and require attention. This work will be carried out as part of a major maintenance roadwork scheme programmed for early 1993. The south west corner of the concrete parapet has been damaged exposing a fixing bolt and requires repairing (see photo ref. M and N).

The safety fence adjacent to the parapet connector NW corner has been slightly damaged by vehicle impact and there is slight damage to the concrete parapet below the connector at this point. This damage consists of spalled concrete although there is no reinforcement exposed.

There is a section of solid aluminium infill panel missing (south west end) due to accident damage by an unknown vehicle (see photo ref. L.). This damage has been repaired since the inspection was carried out.

The steel parapet paint system and galvanising is breaking down in some areas and corrosion is in evidence and some localised remedial work is required

The addition of the solid aluminium panels will cause maintenance problems to the "hidden" sections of parapet and full inspection of these areas was not possible.

The base plates of the parapet posts appear to be the source of rust staining to the parapet plinths and, as such, require attention."

76. The collisions there referred to have no relevance to what should be the length of the safety fence. Safety fencing has a useful life of about 20 years. So, in 1993 the opportunity was taken to renew the safety fence at a time when the surface of the carriageway at the bridge was being renewed.
77. Originally it had been intended to replace the OBB with TCB (Variation Order 20^th September 1993). However, at a site meeting on the 1^st October 1993 attended by Mr Moulds who was probably the Bridge Engineer of Humberside County Council, the department's agent, the decision was taken to revert to OBB. See Mr Moulds diary entry which reads:-

"I discussed the problem with safety fence either side of gravel Pit and Little Heck bridges verges. Our site instruction asked for corrugated either side with a full height anchor at the parapets. Colin Philipson stated that tension corrugated beam must be a minimum of 45m long which is much longer than the original open box design. It was agreed with consultation with Andy Arendal to change the safety fence to open box but still have a full height open box anchor. Colin informed me later it would take 6 weeks to obtain the full height anchors to the new specification. It was agreed to heed the full height anchors specified in the old specification which there are in stock. The only difference is that the posts are socketted."

78. If the Variation Order had been complied with the total length of the Safety fence would have been about 70 metres (14.4 metres of OBB attached to the parapet + 45 metres of TCB at full height + 9 metres of anchorage). At this length the offside wheels of Hart's Land Rover and trailer would have struck the TCB and incursion onto the railway line would not have occurred.
79. Just before the trial started Mr Benstead, who had been for a short time but at no material time the Route Manager of the relevant part of the M62, discovered an accident probably in 1995. He was investigating why OBB manufactured in 1994 were in the safety fence renewed in 1993.
80. In his statement dated the 3^rd October 2003 he said:-

"I have, however, spoken to John Wharton who was Contracts Manager for Tarmac during the period of time when that company carried out all maintenance and repair work to safety fencing on behalf of Humberside County Council. He remembers an accident in which a heavy goods vehicle carrying a large electrical generator left the westbound carriageway of the M62 and struck the ramped terminal of the approach safety fencing to the Little Heck Bridge. The vehicle sustained a broken front axle but he believes that the driver was unhurt. Tarmac were required to remove part of the existing safety fencing to enable the vehicle to be pulled back up the embankment and removed from the highway by a recovery contractor probably acting on instructions from the Police. Following removal of the vehicle Tarmac repaired the safety fencing. Mr Wharton told me that all records from this time have been destroyed so he is unable to be more precise about the date of the repair. He was, however, sure that it was after the completion of the major maintenance work carried out by Tarmac on the eastbound carriageway. This therefore puts the date some time after April 1995 but before April 1996 when Amey Construction took over as maintenance contractor for the motorway.

A further source of information I have checked is the accident record for the period 1994 to 1996 and there are no recorded accidents (that is personal injury accidents) within 1.2 km of the bridge within the period. This is consistent with the fencing being damaged as a result of an accident in which there were no injuries that would have to be reported through the "Stats 19" procedure."

81. The department's power in relation to safety fences are derived from Section 66 of the Highways Act 1980 which reproduces Section 67 of the 1959 Act.

"(2) A highway authority may provide and maintain in a highway maintainable at the public expense by them which consists of or comprises a carriageway, such raised paving, pillars, walls, rails or fences as they think necessary for the purpose of safeguarding persons using the highway.

(4) The powers conferred by the foregoing provisions of this section to provide any works include power to alter or remove them."

82. Therefore Stovin .v. Wise would have been an answer to the claim based upon what happened in 1993 and 1995. Even if in 1993 or 1995 the department had owed a Common Law duty of care, the recorded accidents would not have indicated that the fence was too short and I would not have found the department guilty of negligence in 1993/1995 or thereafter.
Ex Turpi Causa.
83. Mr Burnett submitted that Hart's convictions for causing death by dangerous driving with concurrent sentences of 5 years imprisonment should preclude his entitlement to contribution under the 1978 Act.
84. I reject this submission. Almost all litigation arising out of a road traffic accident involves the commission of a criminal offence and almost all of those criminal offences do not involve specific intent but result from carelessness up to gross negligence. If ex turpi causa were to defeat a claim for contribution, it would seriously affect the whole basis of motor insurance. Whatever may be the position between joint tortfeasors both guilty of a crime of specific intent, in my judgment the ex turpi causa principle has no application to Hart's claim for contribution. I follow the judgment of Ferris J. in K .v. P [1993] Ch 140 at page 148H where he said:-

"In my judgment the ex turpi causa defence is not available as an answer to a claim for contribution under the Act of 1978. The specific purpose of that Act, as of the Act of 1935 before it, was to enable claims for contribution to be made as between parties who had no claim to contribution under the general law. To permit the ex turpi causa defence to be relied upon as an answer to such a claim would, in my view, narrow to a substantial extent the deliberately wide wording of section 6(1) of the Act of 1978 and would, in effect, make a claim for contribution subject to a condition precedent which is not to be found in the Act. Moreover, section 2(1) and (2) give the court ample power to fix the amount of the contribution at a level, including a zero level, which takes account of all the factors which, in relation to common law claims, are relevant to the ex turpi causa defence."

Apportionment.
85. If I had found the department negligent, I would have acceded to Lord Grabiner's submission that responsibility for the damage should be apportioned equally. In determining contribution both blameworthiness and causation have to be taken into account. If the department had been negligent, it would have been because they installed a dangerously short safety barrier having regard to the foreseeability of vehicular incursion onto the railway line with potentially catastrophic consequences. Egress from the motorway and consequent incursion onto the railway line could occur from many causes. It would have been the department's duty to prevent egress on approach to the bridge from whatever cause and consequent incursion onto the railway line. On the other hand Hart was the precipitating cause of the incursion onto the railway line. He was grossly negligent in driving when he had had no or almost no sleep in the previous 36 hours.
86. In the event, negligence not having been established against the department, the claim for contribution is dismissed.