Anderson & Ors v International Oil Pollution Compensation Fund & Ors [2001] ScotCS 34 (14 February 2001)

BAILII is celebrating 24 years of free online access to the law! Would you consider making a contribution?
No donation is too small. If every visitor before 31 December gives just Â£1, it will have a significant impact on BAILII's ability to continue providing free access to the law.
Thank you very much for your support!

[1] The claimants sue for compensation for damage to property sustained in consequence of the wreck of the tanker Braer.

[2] The claims were made in December 1995 and January 1996 by way of ordinary actions against The Braer Corporation, the owner, Assuranceforeningen Skuld (Skuld), the insurer and The International Oil Pollution Compensation Fund (IOPCF). These were among 239 such actions raised at that time.

[3] Skuld thereafter raised proceedings for limitation of its liability in terms of section 12 of the Merchant Shipping Act (Oil Pollution) Act 1971 (the 1971 Act). By interlocutor dated 25 September 1997 Skuld was found entitled to limit its liability to the sterling equivalent of 5,790,052.5 special drawing rights. Skuld duly paid its limitation fund into court and was exonered and discharged. The court then appointed claims on the limitation fund to be lodged in the limitation process.

[4] In order to expedite procedure, and after discussion with all parties concerned, the court issued a Practice Note dated 15 May 1999 the effect of which was that all procedure in the existing actions was brought to a halt and the limitation process became the vehicle in which all of the claims for compensation were thereafter pursued. In many of the claims the claimants adopted their pleadings in the earlier actions.

[5] 367 claims were lodged on the limitation fund. Skuld had no interest in the outcome of any of the claims. Skuld therefore did not compear.

[6] IOPCF, however, by reason of its contingent liability under section 4 of the Merchant Shipping Act 1974 (cf. Landcatch Ltd v IOPCF, [1998] 2 Lloyd's LR 552; affd. 1999 SLT 1208) has an interest in the outcome of each claim.

[7] All parties agree that in limitation proceedings such as these, each claimant has title and interest to object to other claims (Van Eijck & Zoon v Somerville, (1906) 8 F.(H.L.) 22).

[8] These claims ultimately relate to the liability imposed on the owner by section 1(1) of the 1971 Act

[9] This is a group of six similar claims in which there has been a conjoined proof before answer. In each claim the damage alleged is that of distortion or deterioration of asbestos cement roofing materials resulting from contamination by dispersants used to treat the spill. The parties lodged a joint minute agreeing that the evidence led in each of the claims, including productions lodged at the proof, should be evidence in all of them.

[10] IOPCF and Mr James L R Robb, Mr Brian Robb and Huxter Salmon Limited, who are claimants on the limitation fund, have lodged objections to each of these claims. The objectors accept that if dispersant used in the treatment of the spill was deposited on the roof of a claimant and if it caused damage to the roof, that damage would come within the scope of section 1 of the 1971 Act. The objection is that the claimants did not sustain the damage alleged or that, if they did, the damage was not sustained in consequence of the Braer mishap; that the compensation sought is excessive, and that in any event in the calculation of compensation there should be a deduction for betterment.

[11] During the proof, senior counsel for the claimants intimated that he proposed to lead no evidence in support of this claim. The claim therefore falls to be refused.

[12] This firm is a claimant for compensation for roof damage similar to that with which these cases are concerned. A proof in the claim has yet to be fixed. In this proof the claimants led the evidence of Mr John Henderson as to the state of the firm's roofs at three farms on the east side of the mainland. Their purpose was to support their circumstantial case. I have to take this evidence into account. It was competently led and both the claimants and the objectors rely on parts of it. On the other hand, my findings in these proceedings will anticipate some of the factual questions that will be in issue in the proof in Messrs Henderson's claim. In the circumstances, I emphasise that the facts relating to Messrs Henderson's buildings were not investigated at this proof as thoroughly as those relating to the claimants'. Messrs Henderson may have other and more detailed evidence to lead in support of their own claim.

[13] Mr Gordon Johnston raised an action on 4 January 1996 against the owner, the insurer and IOPCF. In the summons he averred that he was the proprietor of the dwellinghouse and garage to which the claim related. Part of the defence to that action was that Mr Johnston had only a one-half pro indiviso share in the subjects, the other one-half share being held by his wife. In reply to that, Mr Johnston averred that he sued only in respect of his one-half share.

[14] On 19 February 1999, after the court had made on order for claims on the limitation fund, a condescendence and claim was lodged in respect of the same subjects in the name of Mr Johnston and his wife Mrs Linda Jane Johnston. Mrs Johnston had not until then sued in respect of her own interest in the subjects. In its objection to the claim Huxter Salmon Limited took the point that any claim competent to Mrs Johnston was time-barred by reason of section 9 of the 1971 Act. In the closing submissions in this case counsel for the claimants did not attempt to argue that Mrs Johnston's claim had been timeously made. In my opinion the objection on this point is well founded. It follows that Mrs Johnston's claim must be refused as having been extinguished by the operation of section 9. It follows also that Mr Johnston can be entitled to only one half of any compensation that could be awarded for the damage alleged (cf. Eunson v The Braer Corporation, 1999 S.L.T. 1405).

[15] The locations of the claimants' properties, and those of Messrs J I Henderson, are an important element in the circumstantial case for the claimants. To put the evidence in context, I annex three maps to this Opinion. Annex 1 is a map of Shetland. Annex 2 is a map of the southern mainland, the area with which this case is concerned. Annex 3 is a map of the southern tip of the mainland in the vicinity of the wreck. These maps show the various places and landmarks that I shall mention.

[16] As counsel appeared to agree, this is not a case where the honesty of any witness is in doubt. There have however been numerous disputed issues of fact that have raised questions as the reliability of some of the evidence. In relation to the spraying operation, I have concluded that the evidence of certain of the claimants and their witnesses is not to be relied on, but that conclusion relates simply to the accuracy of their recollection. In relation to the roof defects, I have found some of the evidence of the claimant Mr John McLachlan and the claimants' witness Mr John Henderson to be unreliable. Their conviction that the incident was the cause of their roof problems led them to exaggerate about the previous condition of their roofs. I have serious reservations as to the reliability of the evidence of Mr Peter Walton, the expert witness for IOPCF and Messrs Robb, to the extent that he was a witness to fact. I shall give my reasons in due course. Beyond that, the main differences between the parties' witnesses related to matters of opinion and interpretation.

[17] On 5 January 1993 at about 5 am the Braer while on passage from Norway to Canada suffered a complete loss of engine power in a position about 15 kilometres south of Sumburgh Head, Shetland. The Braer drifted out of control towards the southern end of the mainland of Shetland. At about 11.20 am she grounded on rocks at Garths Ness on the west side of the south mainland. As soon as she grounded, her cargo of crude oil began to escape.

[18] On the day of the grounding and throughout the rest of the month the weather was fierce, even by Shetland standards. The winds came mostly from the south and south-west. The average windspeed was about 50 miles per hour. There were gusts of up to 100 miles per hour. The lengthy periods of sustained high winds during that month were the worst recorded in modern times. There was rainfall of about six inches in the course of the month.

[19] In southern Shetland winds gust and swirl. Two phenomena are typical of the area, namely "windballing" and "windshear." Windballing occurs when westerly winds rise over the cliffs and accelerate, creating a vacuum as they move over the land. The cliffs at Fitful Head can cause windballing in the area north of the Bay of Quendale. Windshear occurs when opposing wind forces collide causing air pockets. Windshear is a particular hazard for aircraft approaching Sumburgh Airport from the east. The extreme variations of wind in the area, and the eddying effects associated with them, are highly localised.

[20] By 6 January oil had spread southwards from the wreck across the Bay of Quendale and the West Voe of Sumburgh round to Sumburgh Head. On the west side of the mainland it had travelled northwards beyond Fitful Head. On the east side it had reached the mouth of the Pool of Virkie. By about 11 January it had spread extensively up the west side. Relatively little of the oil spread round to the east side of the mainland.

[21] Annex 4 consists of six maps showing the approximate extent to which the oil had spread each day from 6 to 11 January. They are reproduced from the Report of the Marine Pollution Control Unit, The Braer Incident : Shetland Islands, January 1993 (HMSO, December 1993). These maps are not scientifically accurate, but they usefully illustrate the general pattern of the spread of the oil during that period.

[22] On 11 and 12 January the Braer broke up. By 12 January virtually the whole cargo of about 84,000 tonnes of crude oil and all of her heavy bunker oil had been discharged into the sea.

[23] The Braer was carrying a light form of crude oil known as Norwegian Gullfaks. A small amount of the oil formed slicks; but because of the lightness of the crude oil and the force of the wind and the sea, most of the oil became mixed with the seawater instead of floating on the surface. The spraying of dispersants onto the oil assisted this process to a limited extent. The wind and sea conditions were such that oil, seawater and dispersants were blown into a fine spray that was carried eastwards over the south mainland. The combination of the extreme weather conditions and the lightness of the crude oil had the unexpected result that while there was relatively little contamination of beaches, there was extensive contamination of land and buildings.

[24] From 6 January the air in southern Shetland became contaminated with an oily spray. The spray covered roads, buildings, pastures and vegetation. It obscured the runway lights at Sumburgh. An oily film adhered to windows. Roads had to be sanded. The deposition of the spray was not uniform. Some parts of the southern mainland were free of it. The spray was deposited mainly around the western fringes of the southern tip of the mainland up to about Burra; but the wind carried some of it further inland.

[25] Skuld and IOPCF opened The Braer Claims Office (the Office) in Lerwick to deal with claims for compensation. In the months after the incident there were numerous claims for damage to felt roofs. The Office accepted these claims in principle.

[26] Later in 1993 Mr John Black, one of the present claimants, showed Mr Rae Tulloch, who was then advising the Office, some signs of distortion in his asbestos cement sheeted roofs. Thereafter Mr Tulloch and Mr Hurley, an official of the Office, mentioned to several householders and farmers the possibility that the spray could have damaged asbestos cement roofs. Some householders and farmers in the Quendale, Toab and Sumburgh areas reported signs of distortion and deterioration in asbestos cement tiled and sheeted roofs. The typical symptoms of which they complained were curling and surface damage. Further examples of such cases were reported on the west side of the south mainland. The problems were reported in sheeted roofs as far north as Bigton and in tiled roofs as far north as Burra.

[27] In late 1993 Mr Tulloch examined a number of asbestos cement roofs on behalf of the Office. He noticed that they showed signs of surface damage, erosion and discoloration.

[28] During 1994 and 1995 numerous claims for such damage were made with the Office. These claims appeared to show certain similarities in time, place and circumstances. In all of them the problems had first been noticed after the Braer incident. All of them related to properties located in the south mainland and, the claimants thought, in the general area affected by the windblown spray.

[29] The Office engaged Mr Gareth Halford, a technician engineer of Howard Humphreys and Partners Ltd (Howard Humphreys), to carry out a detailed survey of each of the claimants' roofs. Mr Halford carried out his survey during May 1995.

[30] Howard Humphreys had some samples from the claimants' roofs analysed to establish whether they had been damaged by oil contamination. In December 1995 Howard Humphreys reported, inter alia, that there was no evidence that contamination by oil was the cause of deterioration in any of the samples.

[31] The Office also engaged an expert on cement, Mr Peter Walton of BRE Limited. Mr Walton inspected a number of roofs in Shetland, including those of the claimants, in September 1995. He submitted his Report of those inspections to Howard Humphreys in October 1995. That Report was incorporated in the Howard Humphreys Report of December 1995. In the light of Mr Walton's advice and of the results of the tests, the Office rejected the claims. Mr Walton carried out a further inspection of the claimants' roofs in Shetland in March 1999. He submitted a second Report dated 30 April 1999. I shall refer to it later.

[32] Oil spill dispersants are blends of surfactants and solvents. They are similar to household detergents. The scientific distinction between the two seems not to be of any relevance in these cases. The surfactants in dispersants are dissolved in solvents so that the surfactants can be sprayed onto the oil. The solvents assist the surfactants to mix with the oil.

[33] Dispersants have an effect similar to that of detergents. They reduce the resistance of oil and water to mix together by altering the molecular forces, known as the interfacial tension, between them. The interfacial tension, or "skin", between oil and water is similar to the more commonly observed phenomenon of surface tension between water and air. Surfactants break down both interfacial tension and surface tension.

[34] When dispersant is sprayed onto an oil slick, the droplets penetrate and mix into the slick aided by the action of the solvent and the momentum of the droplet spray. The surfactant then diffuses through the oil layer. It forms a chemical bridge between the oil and the water and permits easier mixing by the action of the waves. This leads to the formation of oil droplets which go down into the water column.

[35] The spraying of the Braer spill was carried out from the air. The critical factors in aerial spraying are the size of the droplets that are sprayed and the speed and altitude of the aircraft. If the droplets are too large they fall too fast and pass straight through the thin layer of oil into the water column. If the droplets are too small they are subject to wind-drift and are blown off target.

[36] The dispersant is normally sprayed at a rate of about 10-20 litres/second in droplets of about 400-700 microns in diameter. The ideal size is about 600 microns. Once the dispersant is discharged from the aircraft, the wind affects the droplet size. This effect is minimised by flying the aircraft into the wind at as low an altitude and as slow a speed as is safe. The spray is released at a height of less than 50 feet and ideally at about 25 feet. It trails behind the aircraft in a carpet-like cloud which broadens out to about 10 metres before it settles onto the surface of the oil. The most concentrated deposition of the spray occurs nearest to the track of the aircraft. It consists of the larger droplets. There is also a long tail of spray-drift. This is a fine mist of the smallest droplets. It gives a much lower concentration of deposition and is more likely to be trailed by the wind away from the area of spraying.

[37] The operations controller directs and monitors the spraying from the control aircraft, which flies at a greater altitude. When the dispersant lands on the surface of an oil sheen, the sheen clears quickly. When it lands on thick oil, the oil turns a milky colour and gradually mixes into the water. By looking for these signs, the operations controller can see whether spraying is being carried out effectively.

[38] The spraying operation was carried out by six DC3 aircraft owned by Air Atlantique, contractors to the Marine Pollution Control Unit, and flown by experienced crews. Each DC3 could carry a load of 3.5 to 4 tonnes of dispersant. The dispersant was pumped by a variable pitch propeller which controlled the rate of release.

[39] Mr David Mason, the marine pollution and operations manager of Air Atlantique, directed the spraying operation. Mr Mason has long experience in aviation. He had 25 years' service in the Royal Air Force as a navigator and as a staff officer in search and rescue operations, latterly as a squadron leader. In his work with Air Atlantique he has had extensive experience of treating oil spills at major incidents. Mr Mason directed the operations from a Cessna 404 remote-sensing reconnaissance aircraft which flew at an altitude of about 700-1000 feet.

[40] Spraying operations were carried out on 6, 7 and 9 January. The DC3s took off from Sumburgh Airport westwards on runway 27. The western end of the runway was about 93 metres from the sea wall. The aircraft became airborne about 300-400 metres before they reached the sea. They flew into the wind at about 100-110 knots. The object of the exercise was to drop spray with a droplet size of about 600 microns onto the thickest parts of the oil. The aircraft were flown in a circuit, in groups of 3 or 4 at a time, each making several runs and discharging the dispersant at a rate of about 1 tonne/minute. Each run lasted for about 10-15 seconds. Each aircraft made about 10-12 runs before returning to reload.

[41] On Wednesday 6 January, between 11am and 3pm, the DC3s sprayed just over 100 tonnes of dispersant in an area extending from the end of runway 27 westwards to a point south of Broad Stack, southwards to a line running across the northern tips of Little Holm and Lady's Holm and northwards into the Bay of Quendale and round Garths Ness up to the site of the wreck. It appears that the principal area of spraying extended some distance into the Bay of Quendale. In the early stages of this operation there was an incident in which one of the DC3s discharged dispersant during take-off. I shall describe this incident later.

[42] On Thursday 7 January, the wind strength made spraying by the DC3s impossible; but a helicopter dropped a small quantity of about 750 litres of dispersant from pods onto an oil slick close to the wreck. The dispersant was dropped from a height of about 20 feet. The operation took about a minute. It was ineffective because the slick consisted of bunker oil on which the dispersant made little impact. It had to be abandoned when the weather deteriorated.

[43] On Friday 8 January the weather deteriorated further. The windspeeds were 36-39 knots gusting to 65 knots. No spraying was possible on that day.

[44] From 6 January there was public concern about the effects of the spraying operation and the use of the dispersants close to the land. It was expressed at the press conferences held at the emergency centre at Sumburgh Airport. Much of this concern seems to have resulted from the take-off incident. There was talk that members of the public might try to block the runway as a protest.

[45] In the morning of Saturday 9 January the emergency centre arranged a demonstration of spraying at Sumburgh Airport. The purpose was to reassure the public that the dispersant would not drift over the land. The demonstration was observed by the Member of Parliament, Mr Jim Wallace, and other public figures. On this occasion, two DC3s sprayed about 20 tonnes in an area to the south of Fitful Head approximately between Siggar Ness and Garths Ness. This operation was abandoned after 4 runs for safety reasons. At 1pm the wind peaked at 38 knots. For the rest of the day it was about 25-34 knots.

[46] No spraying was carried out after that because the weather and the sea conditions were effectively dispersing the oil. In all, about 24 sorties were flown and about 120 tonnes of dispersant were sprayed.

[47] The weather conditions were the most difficult that Mr Mason and the DC3 pilots had experienced. These problems were added to by the local wind phenomena to which I have referred. On several occasions, according to Mr Mason, the wind changed direction quite dramatically. Mr Mason himself experienced turbulent air around Fitful Head on the first day of spraying. Flying into the wind minimised side drift of the spray; but the variable wind conditions made some drift inevitable.

[48] The evidence raises a question as to the height at which the DC3s discharged the dispersant. Mr Mason said that on each spraying run the aircraft descended to about 25 feet above the surface. He accepted that some of the pilots may have flown at 40 feet but he ruled out the possibility that they sprayed from 50 feet or over. He said that if they had done so, he would have been informed.

[49] On the other hand, there was evidence that the aircraft released the dispersant at heights greater than that. Mr Barrington Stevens was assistant manager of the airport and was closely involved in the emergency operations. He said that he saw the aircraft in the West Voe of Sumburgh flying at about 50-75 feet; but it was not clear from his evidence if they were spraying at that time. The claimant Mr John McLachlan was employed as a fireman at the airport. He said that on 6 January he saw the DC3s spraying from West Voe to Garth Wick. Their loads were not reaching the target because the aircraft could not get down low enough. They released the dispersant and "it just blew back." The claimant Mr Geoffrey Brown said that on one unspecified occasion he watched a DC3 flying over the sea at a height of 60 to 100 feet. As the dispersant fell it was carried horizontally over the land. The claimant Mr John Black, who has a pilot's licence, watched the spraying operation on 6 January from a layby near to the runway. He estimated that two of the aircraft flew at a height of at least 150 feet.

[50] Mr Mason said that spraying from such heights would have been ineffective. Mr Mason was directly and professionally involved in the spraying operation. He seemed to me to have an accurate recollection of the operation and to be a reliable witness to the facts. If the aircraft had sprayed from a height of over 50 feet it would have been obvious to Mr Mason that the operation was ineffectual and there would have been a contemporary record of the fact. It is unlikely that experienced pilots would have failed on such an elementary point. Mr Mason said that on 6 January he saw the dispersant clearing the sheen rapidly and then acting on the thicker oil, which turned milky at the edges and disappeared into the water column. It would have been impossible to spray in the West Voe because of the wind direction and the restrictions on the DC3s' manoeuvrability within the Voe. On 9 January he did not see any spray being blown ashore. There is no reliable evidence from any other source to suggest that the pilots failed to carry out the spraying operation efficiently despite the difficult circumstances of the time. On this question I prefer the evidence of Mr Mason.

[51] On 6 January one of the DC3s released dispersant over land when the crew inadvertently switched on the spraying equipment as the aircraft was taking off. There was a conflict of evidence on this matter. Mr Mason said that the crew at once reported the incident to the control aircraft. According to information given to him by the crew, the accidental spraying must have lasted for less than 5 seconds. The crews were given a "forceful briefing" to ensure that the incident would not be repeated. It was not repeated. If it had been, he would have heard about it.

[52] Mr Stevens said that he watched some of the spraying operations from a north-west facing window on the third floor of the terminal. He had an unrestricted view towards the Bay of Quendale. He could see the DC3s take off. He saw only a small proportion of the flights. On three or four occasions he saw aircraft release dispersant before they had reached the end of the runway. He saw the wind picking up clouds of dispersant and blowing it north-eastwards across the land. Mr Stevens is not supported from any other source in his assertion that there were several such incidents. If there had been, it is probable that there would have been some contemporary record of them. In my view, Mr Stevens' evidence on this point is not to be relied on. The end of runway 27 was less than 100 metres from the sea wall at Sand of Hayes. In some of the sorties spraying would begin almost as soon as the aircraft had cleared the runway. Mr Stevens' vantage point was at the terminal, which was some distance away. In my view he must have been mistaken in his impressions. I conclude that there was only one such incident and that it occurred as Mr Mason described it.

[53] For some of the time the windblown spray contained dispersant. The dispersant was released into the air in three ways: (1) by being released from the DC3 in the take-off incident to which I have referred; (2) by being blown over the land after being released over the sea from the DC3s, and on 7 January from the helicopter; and (3) by being blown up from the surface of the sea in a mixture with oil and seawater.

[54] Taking Mr Mason's estimate that the accidental release on 6 January lasted for less than 5 seconds and assuming that the release was at the maximum rate of 20 litres/second, I conclude that the amount of dispersant released on that occasion cannot have exceeded 100 litres.

[55] There was a conflict in the evidence as to the amount of dispersant that was blown over the land during the spraying operation. Mr George Henderson was the Deputy Emergency Controller of Shetland Islands Council, which led the emergency effort. Mr Henderson is a retired police officer. He had 31 years' service. Latterly he acted as Deputy Chief Constable of the Northern Constabulary. On 6 January, Mr Henderson watched the spraying operation from Ward Hill on the road from Lerwick to Sumburgh. This vantage point overlooked the airport. He said that he saw "a substantial quantity" of the dispersant dropped from the aircraft being blown back over the land in a north-easterly direction and landing to the west of Toab. Mr John McLachlan, to whose evidence I have referred, said that when the dispersant blew back from the aircraft, most of it went onto the land. The public were being sprayed with it. There was considerable public annoyance. There is also the evidence of Mr Brown to which I have referred. Mr Stevens, who watched some of the spraying operations from the airport terminal, had the impression that little of the dispersant was reaching the surface of the sea, such was the strength of the wind.

[56] On the other hand, Mr Mason said that none of the dispersant dropped from the helicopter on 7 January and little of the dispersant sprayed from the DC3s failed to land on the oil. Mr Mason had reason closely to observe the spraying operation from the control aircraft. He said that it was obvious to him when the dispersant landed on the oil. On 6 January the dispersant acted effectively. On 9 January, when spraying took place further out to sea, he did not see any spray being blown ashore. In my opinion this evidence is to be preferred. It comes from an expert source who was directly involved in the spraying operation and was in a position to observe it with an experienced eye. There is some slight support for this evidence in a photograph in the Report of the Ecological Steering Group on the Oil Spill in Shetland (ESGOSS), The Environmental Impact of the Wreck of the Braer (Scottish Office, 1994, p 181), which shows a DC3 spraying over the sea near the airport on 6 January. The aircraft is in level flight close to the surface, and it appears that the dispersant is falling onto the surface in normal fashion.

[57] Cement is a hydrophilic material. It is porous. It absorbs water in its natural condition. Water penetration of cement is not harmful per se. Water is present in cement at the time of its manufacture and remains in it thereafter in various forms.

[58] After cement has been mixed with water and has cured, the water content in the set cement consists of water that is chemically bound with the cement, gel water that has not become chemically bound with it and external water that has become absorbed into it. The water that has not combined with the cement during the curing process is free to evaporate. When the excess water evaporates, it leaves a connected pattern of voids in the cement into which water can be re-absorbed. Because of the density of the cement and the surface tension of the water, water landing on the surface will take a considerable time to penetrate into the cement.

[59] When excess water leaves the cement at the time of curing, the cement shrinks. If it re-absorbs water, it will expand again, but not to the same extent. This "permanent drying shrinkage" is never again recovered. It occurs only once, namely at the stage when the mixture hardens into set cement. The amount of the permanent drying shrinkage depends on the moisture content in the manufacturing process.

[60] Asbestos cement is a composite of cement, water and asbestos fibres. The fibres normally constitute about 8-12% of the material. They are like solid particles embedded in the cement. They prevent it from cracking and increase its strength and durability. Asbestos too is hydrophilic. Asbestos cement sheets are less dense than asbestos cement tiles. They have larger pore sizes and can absorb and lose water more quickly.

[61] There is research evidence as to the effects of moisture on asbestos cement roofing materials. The principal reference cited at the proof is a report of the Building Research Station, Weathering Tests on Asbestos Cement Roofing Materials, published in 1947 (Technical Paper No. 29, by Dr F E Jones : "the Jones Report"). This was the most comprehensive study of the subject ever carried out. The Report establishes that asbestos cement absorbs moisture (pp. 19-21) and that when it does, the cycle of wetting and drying produces changes in the length of the sample. This phenomenon is known as "moisture movement." If one surface of the tile or sheet is wet and the other dry, differential expansion causes it to warp. This takes the form of a slight degree of curling or bowing and is only temporary. Curling in such cases can be up to about 5mms in tiles and up to about 50 mms in sheets. Changes in humidity can cause as much moisture movement as water. Normally the material will stabilise to a reasonably constant moisture level.

[62] The Jones Report also establishes that where materials that have been in situ for over eight years have been subjected to prolonged cyclic wetting and drying the curling becomes permanent. The conclusions of the Jones Report are substantiated by a report of the Building Research Station, Natural Weathering of Asbestos Cement Roofing Sheets, published in 1962 (Note No. D 853). It is normal, therefore, for curling to occur towards the end of the life of the materials. The extent of it can be fairly severe. If the material is old before it is first subjected to cyclical wetting and drying, the extent of the shrinkage will be less.

[63] Curling is also known to occur in newly laid tiles that have been stored in wet conditions. In such cases the tiles curl as they dry out. This is a temporary process. The tiles normally settle within a few months.

[64] There are other recognised causes of defects and of deterioration in such materials. Softening can occur as a result of a manufacturing defect where there is uneven mixing of the cement with the fibres. Acidic rainwater can also attack the surface of a tile or sheet and expose the fibres at the surface. This effect does not occur uniformly. It gives the roof an uneven appearance. Atmospheric carbonation is a cause of warping in thin sheets.

[65] Asbestos cement roofing materials can also crack or split for numerous reasons; for example, when they are walked over in situ; or are struck by windblown debris; or suffer differential thermal expansion or moisture movement when they have been fixed too tightly; or when the installer drives nails through them, or uses fixing bolts that are too large for the drill holes.

[66] Asbestos cement has been in common use as a roofing material in Shetland for many years. It was used in Shetland from the early part of the twentieth century. It was extensively used after the late 1940s. Many houses and agricultural buildings throughout Shetland are roofed with it. In normal use, it has a serviceable life of at least 50 years. It can last considerably longer. Some asbestos cement roofs in Shetland have lasted for over 70 years. Asbestos cement remained in common use in Shetland until about 1989-1992 when it became obsolete, mainly for environmental reasons.

[67] Asbestos cement tiles and sheets were popular as roofing materials in Shetland because of their durability. They did not curl to any significant degree in normal use. The cycle of wet and dry weather could cause minor and temporary curling, but this was not regarded as being a problem of any significance. Before the Braer mishap these materials had, in general, given no cause for concern.

[68] Each claimant pleads that the cause of the damage was the action of surfactants on the asbestos cement materials. This is in line with the final view advanced by the claimants' expert witness. Each claimant avers that the contamination "was confined to a localised geographic distribution on the Shetland Islands" (Cond 7) and that the damage in each case was sustained within that area. The claimants have not defined the boundary of the area.

[69] At the outset of the proof senior counsel for the claimants confirmed that their case was that the damage was done by dispersant and not by oil. The case was therefore based on section 1(1)(c) of the 1971 Act. This was in line with the averment made in each case that "the measures taken to combat the said pollution, in particular the use of surfactants, has (sic) damaged the asbestos ... " (Cond 7). During the proof counsel for the claimants raised the question whether oil could have been a cause of the damage. This line was objected to and was not persisted in. In his closing submission senior counsel for the claimants proposed a finding in fact to the effect that contact with oil from the wreck could have been one of the causes of the damage. Counsel for the objectors argued that that line was contrary to the case made at the proof. I agree. In my view, this case is about the effects, if any, of dispersants on the claimants' roofs and not about any other cause of damage. In any event, there was no evidence to support the finding in fact that senior counsel for the claimants proposed.

[70] The claimants' case on the facts is in two parts, the circumstantial case and the scientific case.

[71] The circumstantial case is to the following effect. Before the incident asbestos cement roofing materials had proved their durability by long use in the stressful weather conditions of Shetland. Curling in tiles and distortion, splitting, porosity and softness in sheets had not been recognised as problems of any significance. In each of the cases under review, there had been no such signs to any material extent before the Braer incident. The claimants would have noticed them if they had emerged. The roofs were subjected to an extraordinary event, namely the aerial deposition of oil, seawater and dispersant. Within about nine months to two and a half years after the incident, these problems emerged in numerous roofs in the south mainland, including the roofs of each of these claimants. In each case the roof was within a few kilometres of the areas in which the dispersant was released. The materials in these cases were of various ages, of various types and in various forms of use. Such problems were not observed in Shetland outwith the area affected by aerial pollution after the incident, nor on the mainland of Scotland. Even if isolated instances of such problems did occur, there was no concentration of such instances in any other single area. In each of these cases obvious explanations, such as manufacturing defects, could be ruled out.

[72] Counsel for the claimants submitted that the circumstantial case alone was sufficient for each of them to succeed.

[73] In briefest outline, the scientific case is that in each case the dispersant assisted the penetration of the asbestos cement by water and thereby caused the premature failure of the roof.

[74] The objectors dispute the circumstantial case on all of the essential issues of fact on which it depends. They also present a scientific case to the effect that the dispersant did not, and could not, cause the damage complained of.

[75] The onus is on the claimants to prove the facts relied on on a balance of probabilities. Although the objectors have led a proof on the facts and on the scientific questions, there is no onus on them to prove their objections. Even if I were not to accept the objectors' scientific case, it would still be open to me to hold that the claimants' case had not been proved (Dingley v The Chief Constable, Strathclyde Police, 1998 S.C. 548, Lord President Rodger at p. 601G-H; affd. 2000 S.C.(H.L.) 77).

[76] Each claimant must first prove two essential facts: (1) that dispersant was deposited in some significant quantity or concentration and for a significant length of time on the roof in question, and (2) that the roof shows the damage of which the claimant complains.

[77] There then arise factual questions crucial to the circumstantial case; namely, whether the damage, if any, is of a kind for which recognised mechanisms unrelated to the incident cannot provide a satisfactory explanation; whether the damage occurred, rather than was merely noticed, after the Braer incident, and whether the cases display a significant pattern of coincidence in time, place and circumstances.

[78] The next set of factual questions relates to the parties' scientific cases; namely, whether dispersant can cause damage to asbestos cement roofing materials and, if so, whether it caused such damage in each or any of these cases. These questions relate to the properties of asbestos cement and surfactants.

[79] Two major issues of mixed fact and law overshadow all of these questions, namely whether the circumstantial case alone, if proved, is sufficient for the claimants to succeed and whether the claimants' scientific evidence, taken at its highest, is sufficient to establish a causal link between the deposition of dispersant and the occurrence of the damage alleged.

VI THE DEPOSITION OF THE DISPERSANT
[80] Each claimant seeks to prove that the mechanism that led to the damage complained of was precipitated by the deposition of dispersant on the tiles or sheets. In my view it is for each claimant to prove that dispersant was deposited on the tiles or sheets in a sufficient quantity or concentration and for a sufficient length of time to trigger that mechanism.

[81] In the fierce and sustained weather conditions in the weeks after the mishap, a considerable amount of windblown spray travelled across the south mainland. The spray consisted at first of oil and seawater blown off the surface of the sea. After 6 January some of the dispersant deposited on the oil must have been taken up into the aerosol along with pure dispersant released over the land or blown onto the land after being released over the sea.

[82] The claimants have failed to prove by direct evidence that dispersant, in any measurable quantity, was deposited on any of their roofs. None of the materials in these cases has been tested for traces of it. It may be that gusting, sudden changes in wind direction and the local wind phenomena to which I have referred make it impossible in the absence of direct evidence to infer that any particular property received a deposit of dispersant in any measurable amount.

[83] Almost all of the dispersant sprayed over the sea landed on the oil and was mixed with the oil and seawater. Of the small amount of the dispersant that did not land on the oil, only a part would have blown onto the land with the sea spray. Of the dispersant that landed on the oil, only a part would have been taken up into the oily spray. There is no evidence as to the amount of the dispersant that was carried over the land from these sources; but in my opinion it must have been an extremely small proportion of the total amount of the windblown spray. I am unable to make any finding as to how far or in what directions any of the dispersant travelled. The claimants argue that it does not matter where the boundary of the deposition lies, because wherever oily spray was deposited on a roof the spray contained dispersant. The oily spray landed on all of the roofs in these cases. Therefore the roofs were all within the relevant boundary, wherever it was.

[84] The claimants' case on this point stands or falls on the evidence of the claimants' expert witness, Professor Willey. Professor Willey has had many years of research experience in materials science and since the late 1970s has specialised in the properties of asbestos. He has lectured at Paisley University and at the now Glasgow Caledonian University. He has been appointed as a visiting professor in the University of Strathclyde. He is managing director of ACS Environmental Services Ltd, a consultancy that specialises in problems relating to asbestos, mainly in its health and safety aspects. In that capacity he has practical experience of examining asbestos roofs in a wide range of conditions. Professor Willey proposed a model to explain how the dispersant in this case caused damage when it landed on the roofs. But before that model could come into play, he had to suggest an explanation of how the dispersant reached the roofs in the first place.

[85] Professor Willey's explanation was that whenever oily spray landed on a roof the spray must have included dispersant because it was "very probable" that the plumes of oil and of dispersant were substantially coincident, especially in areas close to the wreck, by reason of the fact that the oil and the dispersant were of similar molecular weight and would have similar particle sizes and size distribution. In cross-examination Professor Willey agreed that in the north-south direction the plume of dispersant would be more localised than the oil plume and that dispersant blown off the sea would not travel as far as dispersant released from the air. That, I think, undermines Professor Willey's theory on this point at the outset.

[86] I do not accept Professor Willey's contention. It is the merest speculation. In my view, it is demonstrably wrong. The wreck had been discharging oil for about 24 hours before spraying began. During that period, the oily spray had been blowing across the south mainland. Spraying took place on only three days and covered only small parts of the entire area of the floating oil. A small quantity of the dispersant was released over land and failed to reach the oil at all. There was no spraying of any of the oil that travelled to the West Voe of Sumburgh and round Sumburgh Head to the east side of the mainland. When spraying began on 6 January the oil had already spread to the west and south and, to a lesser extent, to the east. Only a small area of the oil was treated on that day. On 7 January only a minute amount of dispersant was dropped close to the wreck. On 7 and 8 January oil continued to be discharged from the wreck and to spread around the coasts of the south mainland. On 9 January a restricted spraying operation was carried out over a small part of the total area of sea affected and was abandoned after four runs. That was the end of the spraying; but the wreck continued to discharge oil for a further three days. In all, the spraying operations took only a matter of minutes; but the discharge took place over a period of about a week and the oil slicks spread out over a considerably longer period. For these reasons I consider that it is quite certain that the plumes of oil and dispersant were not coincident. In my view, only a small amount of the oily spray that was blown onshore contained any dispersant at all. Moreover, since the spraying was highly localised and of short duration, whereas the spread of surface oil was extensive in area and in duration, I conclude that the fact that oily spray landed on any of the roofs in this case does not justify the inference that that spray contained dispersant.

[87] IOPCF and Mr Robb led Mr Alun Lewis as a witness on this question. Mr Lewis is a chemist. He formerly worked for British Petroleum for whom he developed one of the commonly used dispersants. He is now an oil-spill consultant. He has an extensive record of scientific publications. His main experience is in the scientific assessment of marine oil spills rather than in the operational use of dispersants.

[88] Mr Lewis was not involved in the emergency. He took the facts set out in the ESGOSS Report and from that information attempted to calculate the quantity of dispersant that had been deposited on roofs in southern Shetland. He took three zones surrounding the wreck: an inner green zone, an intermediate blue zone and an outer red zone. His method was to calculate the area of each of the zones, to take the known amount of dispersant sprayed, to take into account the wind conditions recorded at Sumburgh Airport at the relevant time, to take into account the known droplet size and distribution and to apply his judgment. His judgment was based in part on his experience of flying in one of the DC3s during a test exercise in 1997 in calm conditions off Felixstowe.

[89] Mr Lewis calculated that the green zone extended to 150,000 sm. He calculated that at most 200-300 litres could have been deposited within it, at a rate of 40 to 50 mls/sm. He calculated that in the blue zone approximately 2 to 5 mls/sm could have been deposited. He calculated that over the red zone a maximum average rate of 1ml/sm could have been deposited and that beyond the red zone there was little likelihood of there having been any deposition at all. He considered that dispersant was likely to have reached the roofs of Mr and Mrs Brown and of Miss Fowlie, which are within the red and blue zones respectively. His overall conclusion was that only a small proportion of the 120 tonnes of dispersant that were sprayed reached the land and that the amount of dispersant deposited on each of the claimants' roofs, if any, was infinitesimal.

[90] I am unimpressed by this evidence. I was given no explanation of the basis on which the zones were delineated. Mr Lewis' methodology assumes a constant wind speed and wind direction, whereas the winds in the area were variable and unpredictable in both respects. It assumes that the winds blew over a flat surface. It takes no account of the topography of south Shetland. It assumes that the volume of seawater with which the dispersant was mixed in the Bay of Quendale was in a container. It therefore takes no account of the currents and tides within and around the Bay. The estimate of the volume is based on a calculation of an average depth of water within the Bay in the area in which spraying occurred. The method of calculation of this average or its degree of reliability was never made clear. There are further difficulties with this exercise in that Mr Lewis did not know the track of the aircraft in each of the sorties nor the wind conditions during the sortie, and appears to have assumed a uniform rate of release.

[91] Mr Lewis suggested that despite these methodological deficiencies his conclusions remained robust in view of the great margins for error that were built into his calculations. I do not accept that approach. In my view, if the methodology is fundamentally unsatisfactory, as it is in this case, the conclusion to which it leads ought not to be relied on, no matter the margins for error.

[92] Professor Willey suggested that Mr Lewis should have used a computer model to calculate the rate of deposition of the dispersant on the roofs. Mr Lewis argued that computer modelling would not have produced a more reliable result and would have involved unreasonable expense. He may be right; but that is a side issue. The primary question in relation to Mr Lewis' evidence is how realistic it is. I consider that Mr Lewis' simplistic mathematical exercise is based on assumptions that are either unproved or are obviously contrary to the facts. I am not prepared to base any findings in fact upon it.

[93] However this is a question on which the objectors bear no onus of proof. I am satisfied that the claimants have failed to prove that dispersant landed on any of their roofs in any measurable quantity and they have not attempted to prove that if dispersant landed on any roof, it remained on it for any significant length of time. This is a point of some importance in view of the high winds and heavy rainfall that were experienced throughout the relevant time. I agree with the view of the objectors' expert witness Mr Walton that there would have to be evidence on both points before one could begin to consider the scientific question whether dispersant can have the effects that the claimants allege. The claimants' failure in both respects is, in my opinion, fatal to the claims.

[94] There was conflicting evidence about the present state of the claimants' roofs and those of Messrs J I Henderson. Expert evidence was given on behalf of the claimants by Mr Rae Tulloch and Mr Michael Thompson FRICS.

[95] Mr Tulloch is the former senior lecturer in building construction, and latterly assistant principal, at Shetland College. He had 18 years of relevant experience in local government as a civil engineer and surveyor. For 14 years he was a building contractor. Mr Tulloch has also carried out maintenance to roofs for storm damage. He is thoroughly familiar with roofs in Shetland and with the practical aspects of asbestos cement as a roofing material.

[96] Mr Thompson has practised as a quantity surveyor in Shetland for over 20 years. He has extensive knowledge and practical experience of building and construction in Shetland. He has examined many roofs in Shetland. He is familiar with the use of asbestos cement roofing materials.

[97] Both of these witnesses were of the view that there were serious defects in the claimants' roofs and those of Messrs J I Henderson and that these defects could not be explained by the normal processes of weathering and ageing by which such roofs deteriorate over time.

[98] In addition, the claimant Mr Geoffrey Brown, who spoke to the condition of his own roof, was himself an expert. Mr Brown has been involved in the building industry for all of his working life. He came to Shetland in 1968 as a buildings inspector for the Property Services Agency. In 1980 he became a building control officer with Shetland Islands Council. He held that position until his retirement in 1996. He is thoroughly familiar with local conditions so far as they affect building and building maintenance.

[99] The condition of the other roofs was spoken to by the claimants Mr Black, Miss Fowlie, Mrs Johnston, and Mr McLachlan and by Mr John Henderson of Messrs J I Henderson. These witnesses spoke to the existence of serious defects in their roofs.

[100] IOPCF and Messrs Robb led evidence from two experts, Mr Andrew MacFarlane FRICS and Mr Peter Walton.

[101] Mr MacFarlane is a partner in the firm of D M Hall, Surveyors. He has wide experience in the maintenance and repair of buildings of all kinds and in the estimating of building costs, including those associated with asbestos cement roofs. He has no personal experience of construction in Shetland. Mr MacFarlane was an excellent and fair-minded witness; but he was at a disadvantage on this question, having made only one visit to Shetland, on 14 and 15 June 1999. In general, he found nothing exceptional in the state of the claimants' roofs on that occasion. He had no knowledge of the previous history of these roofs.

[102] Mr Walton is a senior scientist in the materials division of BRE Ltd, the former Building Research Laboratory. He is experienced in the study of cement and cement-based building materials and has specialist knowledge of the chemistry and the properties of such materials. He has almost 30 years of research experience in fibre cement composites. He has served on national and international standards committees and has special expertise in standards relating to fibre cement roofing products on matters such as strength, density and permeability. He has published numerous scientific papers. He has no experience of the construction industry in Shetland or of the practical problems associated with local climatic conditions.

[103] Mr Walton's evidence was in two parts. He first gave evidence of fact as to the condition of the roofs of the claimants and of Messrs J I Henderson, and by way of comparison gave his findings as to the condition of several other roofs in Shetland that he had selected as control properties. In my view, this part of Mr Walton's evidence has serious shortcomings both in the quality of the investigations on which it is based and in the reliability of certain factual information to which he spoke.

[104] The second part of Mr Walton's evidence consisted of his expert scientific opinions on matters relating to the properties of asbestos cement and on the question whether dispersant is capable of damaging it. This part of Mr Walton's evidence is important because it is the only expert evidence before the court as to the properties of cement and of asbestos cement building materials. For the reasons that I shall give, I regard this part of his evidence as cogent.

[105] The first of the shortcomings in Mr Walton's factual evidence is that his investigations were insufficiently thorough. He first visited Shetland in the autumn of 1995 to make a general examination of all of the asbestos cement roofs for which there were claims. He saw the roofs of seventeen properties on that occasion. He saw one of these, Mr and Mrs Brown's house, only from a neighbouring property that he was inspecting. His conclusion then was that the incidence of the curling that he observed in those roofs that were close to the wreck was not significant. Some roofs that were close to the wreck showed less curling than others that were further away. Some that were close to the wreck showed no curling at all. He concluded that there was no statistically significant evidence to link the defects to the incident.

[106] Mr Walton next visited Shetland in March 1999. On that occasion he inspected the claimants' properties and his control properties. His findings are recorded in his second Report. As he acknowledged, the findings of his inspections, even if accurate in all respects, are only snapshots of the state of each of the roofs on a particular day. This seriously limits the value of his study at the outset.

[107] Mr Walton categorised the nature and severity of the defects in each case. His conclusion was that in each case the alleged defect did not exist or, if it did, that it was one for which there was a straightforward and well-recognised explanation. In general, he considered that the curling in the tiles was to be explained by cyclic moisture movement and that the defects in the sheets were to be explained by age, weathering, faulty installation, inadequate maintenance, quality of the material and so on. I have serious reservations about this part of the study too.

[108] I consider that Mr Walton's second Report is inadequate and unconvincing so far as it relates to the history, and in some respects the condition, of the claimants' buildings. In assessing the value of it I have taken into account the constraints that were imposed on him or that he imposed on himself. He had a restricted remit, which was to report on the condition of each roof on the date of his inspection. He was required to inspect all of the claimants' properties and his control properties under the stress of a short deadline. He had to carry out fifty inspections within five days. It is not surprising that his inspections were brief and on occasions perfunctory. He agreed that his inspections were carried out more hurriedly than he would have wished. He was accompanied on these inspections by Mr Halford and a local roofing contractor.

[109] Some of Mr Walton's difficulties were of his own making. In several cases, for example those of Mr and Mrs Brown and Mr McLachlan, he declined the opportunity of a close-up inspection of the roofs. An important factual question related to the extent of the curling in each case. Mr Walton did not measure this. He founded his opinion on estimates made from his inspections or from photographs. In each case Mr Walton abstained from seeking important first-hand information from the claimants. He said that he wished to "remain aloof" so that he could make his own judgments uninfluenced by the opinions of the claimants. He left those matters to Mr Halford. On the other hand he took into account certain opinions of the contractor who accompanied him.

[110] Mr Walton's decision not to speak to the claimants had an unfortunate result on the question of the ages of the roofs. Since the ageing process could provide a possible explanation for some of the problems experienced in these cases, the age of the roof was a significant question to follow up. While Mr Walton considered it "unhelpful" to ask the owner how old each roof was, he nevertheless took into account his own and the roofer's estimates. In his Report Mr Walton gives three estimates of the age in each of these cases. The first purports to be the claimant's. In fact, it is the figure that he understands that his colleague Mr Halford was given by the claimant. In some cases, according to the claimant, that figure is incorrect. This is unfortunate because Mr Halford impressed me as a witness who had carried out his survey conscientiously and professionally. There may have been misunderstandings between him and the claimants. It may be that there was some confusion as to whether the age sought was the age of the roof at the date of the incident or the age at the date of the survey. Whatever the reason, in each of these disputed cases, I prefer the evidence of the claimant as to the age of the roof. The second estimate of age in Mr Walton's Report is that of the contractor who accompanied him on his tour of the sites. The contractor was not a witness. I have given no weight to this figure in any of these cases. The third estimate is Mr Walton's own. In none of these cases is his estimate less than that attributed to the claimant. In one of them his estimate coincides with it. In every other case it exceeds it, twice by a factor of two or more and once by a factor of more than three. In the course of his evidence, having referred to there being "a general feeling" that owners had given wrong ages for their roofs, Mr Walton conceded that his estimates of age were not accurate. Taking these considerations into account, I have relied on the claimant's estimate in every case.

[111] In addition, Mr Walton's failure to consult the claimants resulted in his making unfounded assertions of fact. He said that Mr Black's implement shed might have been roofed with second hand sheets and that the lower garage was roofed with second hand sheets. I accept Mr Black's evidence that he installed these roofs with new sheets.

[112] In a number of cases his Report imputes to the claimant a statement that he or she denied having made. For example, Mr Walton alleged that Mr Black had given ages for the roofs on his implement shed, his lower garage and his byre that were each significantly less than the true age. Mr Black denied having given these ages. Likewise, Mr Walton imputed to Mr McLachlan a statement as to the age of his silage shed which Mr McLachlan denied having made. Mr Walton records that Mr Brown said that the south face of his roof "never curls". Mr Brown denied this. Mr Walton also records that Miss Fowlie said that her slates "never curl at the back (north facing) only at the front (south facing)" (Report, 30 April 1999, p 100). Miss Fowlie denied this. In each case, I found the claimant to be credible and reliable on the point. For this reason wherever any statement of fact imputed to a claimant in Mr Walton's Report is at variance with the claimant's evidence, I have preferred the claimant's version.

[113] I have the impression that from the outset of Mr Walton's second visit to Shetland, his approach to the investigation was informed by a marked scepticism about the claims. I also have the impression that the cursory nature of some of his inspections and his making of factual judgments in preference to his seeking first-hand information from the claimants, originated in his conviction that the Braer incident could not be the cause of the claimants' problems. On the whole, his photographs of the roofs did not do justice to the claimants' complaints.

[114] Despite these strictures, I repeat that it would have been open to the objectors to lead no evidence at all. In the light of Mr Walton's evidence on the state of the roofs, all that can be said is that it lends no support to the claimants' case and to the extent that it is reliable merely weakens it further.

[115] I shall now set out my findings on the condition of the claimants' roofs. In almost every case the claimant's complaints were rejected by Mr Walton. I shall not in every case set out the points of disagreement in detail. I shall simply set out what I consider to be the significant features in each case.

[116] Mr Black farms at Brake. His farm buildings are roofed with asbestos cement sheets. He installed his roofs himself in the 1970s. He had experienced instances of storm damage over the years that caused leaks that were repaired in the normal way. Apart from that, he had not experienced trouble with the roofs before the incident. His roofs suffered deposition of oily spray. In September 1993, Mr Black noticed that some of the asbestos cement sheets were curling at the edges. The sheets curl with gaps of up to two inches. A finger and even a hand can be inserted in some of the gaps. There are signs of curling and bowing in both directions. These problems are illustrated in Mr Walton's Second Report (p. 236).

[117] Mr Black agreed that he had had no reason to inspect his roofs for curling before the incident. He would look at the roofs only if they had suffered storm damage. Apart from storm damage repairs he regarded the roofs as being maintenance free. He just happened to notice the problem in September 1993 around the time at which he made his claim for damage to a felt roof. He could not say when the curling started.

[118] This roof was installed in 1976-1977. It is the worst of the roofs. Unlike the sheets on the other roofs, these sheets are permanently curled. The roof leaks.

(ii) Implement shed
[119] This roof was installed in the late 1970s. Significant curling occurs from time to time. One sheet has cracked and been replaced. Some of the sheets have become soft and are easily cut with a knife. The roof leaks. In the Spring of 1994 Mr Black's son pressure-washed five of the sheets on this roof. They had been discoloured by the spray. Those sheets appear not to have curled.

[120] Mr Walton has noted that the sheets on this roof, like those on the silage shed/lean-to, have been laid "upside down." He says that "this gives a strong but erroneous impression of curling up ... and makes the roof more vulnerable to wind driven water entering the building" (Report, p 53). He says that the sheets on the silage shed and lean-to show moderate curling that looks more pronounced because the sheets are upside down (ibid., pp 54-55). He appears to have been unaware of the local practice of laying sheets with the downturned edge facing the prevailing wind. Mr Walton's comment on this evidence was that it did not seem logical and was not the manufacturers' recommended method. In the light of the evidence for the claimants on this point, I consider that in the conditions of Shetland this is a reasonable method of installation.

[121] The roof was installed in about 1972. The curling on this roof is also substantial. There are broken sheets, sheets with holes, sheets with cracks and sheets with parts missing. The cracking on the sheets has been caused by the use of square metal washers with the fixing bolts. The roof has sustained damage since the incident which has been repaired. This roof shows clear signs of dilapidation.

[123] The roof of the lower garage was installed in about 1973. Curling is prominent on this roof. It varies from time to time. A sheet has cracked; but Mr Black is unable to say if that occurred before or after the incident. In the case of the lower garage the lack of a gutter leaves it open to rainwater penetration in high winds. It is inevitable that it will leak.

[124] The upper garage was built in about 1981. The roof is in reasonably good condition and the distortion, if any, is slight. This roof does not leak.

[125] Numerous photographs of these buildings have been spoken to. My conclusion is that they have been constructed to a workmanlike, but not to the highest, standard. There are certain defects in design and construction and there has been no routine inspection or maintenance of the roofs. They exhibit numerous signs of failures in maintenance.

[126] Mr and Mrs Brown live at Orcadia, Virkie, which lies to the north of the airport. Mr Brown built their house in 1974-1975. He used Eternit "Thrutone" asbestos cement tiles which he laid himself. They carried a manufacturer's guarantee of 30 years. Mr Brown had a shed erected on the site for the storage of his building materials, including the Eternit tiles. The tiles were dry and flat on installation.

[127] In May-June 1992 Mr Brown built an extension to the house. In the course of building it he had to remove some of the tiles. These tiles were not pitted, curled or distorted in any way. Mr Brown was not aware of any problem before the incident. He had not inspected the roof for curling.

[128] The house and the roof suffered deposition of spray after the incident. In about October 1993, Mr Hurley called at the house to discuss Mr and Mrs Brown's claim for the cost of cleaning the exterior of the house and of repairing a felt roof. Mr Hurley advised Mr Brown not to pressure-wash the Eternit tiles because the Braer Claims Office had encountered cases where pressure-washing had caused problems to roof tiles. Mr Hurley drew to his attention signs of slight curling in the tiles and predicted that there would be further curling. Mr Hurley told him that the tiles had become porous. Thereafter Mr Brown noticed that the tiles curled in dry weather and that the surface of the tiles became pitted.

[129] Mr Brown noticed the problem again in about April 1994. At that date a surveyor for the Office, Mr James Smith, told Mrs Brown that the roof had deteriorated since his previous visit. Later in 1994, another surveyor for the Office, Mr Ingram, told Mr Brown that his and his neighbour's houses were the two worst that he had seen. The neighbour, Mr Thomson, replaced his roof at his own expense in about 1997.

[130] The tiles curl in dry weather. The extent of the curling is quite pronounced. It is about 1/2 to 5/8 inch. It is worst in summer when the surface of the tiles dries out. The curling is extreme in the north-facing rear slope of the roof. The curling is greater in extent than any that Mr Brown saw anywhere in Shetland in the course of his professional experience. In addition, the tiles have lost their factory-applied surface colour and have become pitted. This deterioration is worse than was to be expected from normal weathering. This is a minor problem in comparison with the curling.

[131] On the occasion of Mr Walton's inspection in March 1999 the curling was much less pronounced; but that was to be expected in the overcast and showery weather conditions on that day. The Report describes the condition of the roof as "discoloration/weathering : failed in patches." In his evidence Mr Walton tried to minimise the extent of the distortion. He said that normal cyclic moisture movement was quite sufficient to explain the curling that he saw. The curling is certainly cyclic, but the extent of it in dry weather is greater than Mr Walton has observed. In the summer of 1992 Mr Brown took photographs of the house. They appear not to show distortion of the tiles. Mr Brown also took photographs of his roof in about April-May 1994. These appear to show significant distortion.

[132] Since the curling in this case is cyclical, one cannot draw any hard and fast conclusions from the state of the tiles on any particular day on which they happened to be photographed or on which Mr Brown happened to be working on the roof.

[133] Miss Fowlie lives at Eastshore, Saltwater, Virkie, close to the airport and to the Bay of Quendale. She has lived in the house since 1984-1985. It had been refurbished in about 1982. The roof was about 10-11 years old at the date of the incident. Miss Fowlie was co-ordinator for an advisory committee of farmers and crofters. In this capacity she received reports of curling in asbestos cement sheets. In the summer of 1993 she noticed that the tiles on the roof of her house showed signs of curling. She had not experienced problems with her roofs before that. The problem became more noticeable in the following summer and increased thereafter. Miss Fowlie agreed that she had had no cause to look for curling on her roofs before the incident.

[134] The surfaces of the tiles have become rough. The tiles now curl and lift in dry weather on both slopes of the roof of her house. Two tiles have had to be re-nailed. Although it is difficult to detect curling on Mr Walton's photographs, there is other photographic evidence showing the extent of the problem (No 21/1 of Pro). This roof had many years of life left in it.

[135] In 1986-1987 Miss Sheila Fowlie and her former husband built an extension to her house in the course of which they removed a number of tiles from the roof. They decided to keep them as spares. They stored them in dry conditions in an inspection pit in the garage. In about March 1993 some tiles on Miss Fowlie's roof were damaged in a gale. Miss Fowlie replaced them with some of the spares. The replacements were therefore from the same batch as those which showed signs of curling after the Braer incident; but they do not show such signs. They have a different appearance.

[136] Mr Walton said that he did not follow up this line of evidence because of lack of time. He had to meet a deadline for submission of his Report. He added however that to assess the significance of this he would need to know a reasonable amount about the history of these tiles and the conditions in which they were stored. In view of the inadequacies of his own examination of his control properties I cannot follow why Mr Walton was so demanding in his requirements for information as to Miss Fowlie's spare tiles, particularly since he had the advantage of hearing her evidence on the point.

[137] In my view the possible significance of this evidence was a matter worth following up; but neither side did so. I can make no finding on the question whether the better condition of the replacement tiles is a result of their not having been exposed to the oily spray or is simply a result of their having been carefully stored.

[138] The sheets on the roof of Miss Fowlie's garage have curled to a significant extent. Large gaps have opened up at the overlaps and there are cracks in the middle of the sheets.

[139] Mr and Mrs Johnston live at Keldaberg, Cunningsburgh, on the east side of the south mainland.

[140] Mr and Mrs Johnston built the house in 1973. It is roofed with asbestos cement tiles. They noticed signs of leaking in their roof about six to eight months after the incident. It had given them no trouble previously. During the summer of 1995 they noticed that the tiles were curling at the edges. They saw a similar problem in the asbestos cement roof of their neighbour, Mrs Barbara Isbister.

[141] In October 1995 the Office sent Mr Ingram to investigate the claim made by Mr and Mrs Johnston for the cost of washing their walls and washing and repainting their roof tiles. To Mr Johnston's surprise, Mr Ingram told him that the roof was damaged. The tiles now curl and warp in dry weather and flatten in wet. There is also some cracking in the tiles. Mr Walton records that external examination of the roof gives no reason to think that it leaks, but I accept Mrs Johnston's evidence that it does. Her husband took Mr Ingram into the loft and showed him the signs of the leaks.

[142] This is the second worst of the tile cases after Mr and Mrs Brown's. The photographic evidence produced by Mr Johnston proves that these tiles suffer noticeable distortion. Mrs Johnston had had no occasion to check the roof for curling before 1995.

(ii) Garage
[143] The garage was built in 1978. It is roofed with asbestos cement sheets. Since the incident, Mr and Mrs Johnston have noticed that the sheets now curl slightly. The problem seems to be insignificant.

[144] Mr McLachlan farms at Exnaboe, Virkie. He has four buildings roofed with asbestos cement sheets. Before the incident, they were in reasonably good condition and generally watertight. After the incident they were all right for a considerable time. Then Mr McLachlan heard reports of trouble with asbestos cement. At this time Mr Tulloch visited the farm and told him that he suspected that there was going to be a problem with asbestos.

[145] In the summer of 1994 Mr McLachlan noticed slight curling in some of the sheets. Although there is now some degree of curling in the sheets, it is much less than in Mr Black's case. There are also signs of cracking, splitting and mild curling. The main problem in this case is that of water penetration. It penetrates at the screw fixings; at the laps or joins, at the lean-to flashing and in the middle of some of the sheets. In the years since he became aware of the problem, Mr McLachlan has noticed that the sheets have become porous and spongy, with damp spots in the middle of the sheets. There are splits along the valleys of the corrugations. The sheets let in water in numerous places. There is water penetration on every exposed face.

[146] Mr Walton estimated this roof to be "a good 30 years old." It was built in October 1982. The date is vouched by a letter from the builders. The side walls are also made of asbestos cement sheets. Although Mr Walton found no evidence of leaking in this roof when he viewed it from ground level inside the building, I find that the roof leaks. It is spongy after rain. There is also leakage on the side walls. The building lacks a gutter. There is considerable evidence of lack of maintenance in this building. There is a hole in the roof. The RSJ frame is rusted at its lower level. On one side the gutter has come off and has not been replaced. On the other side, the gutter has not been cleared.

[147] Mr Walton estimated that the roof was at least 40 years old. He suggested that the sheets may have been second-hand. I accept Mr McLachlan's evidence that this shed was built in 1947-1950 and was roofed with new sheets in about 1978-1979. The sheets curl in dry weather. There is also splitting along the length of the sheets. This became apparent about a year to a year and a half after the incident. There is also dampness and water penetration on the sheets shown by the wet patches on the insides of the sheets. The leaks are over and above those associated with the splits in the sheets. The degree of curling is slight. There is also slight softening of the sheets. There appears to be at least one defect of construction, namely the over-tightening of the fixings that has caused cracks along the ridges of the sheets. There is evidence of moss growth under one of the raised edges, which indicates that this gap at least is of long standing.

[148] This was built in about 1980. The sheets have become porous and spongy. There is some cracking in the sheets and the roof leaks. On the east side the gutter is missing.

(iv) Potato shed
[149] This is a lean-to building attached to the silage shed. It was built in 1989. The date is vouched by an invoice. Here too there is some cracking. The roof began to leak in about 1994. There is leaking at the laps and in the middle of the sheets. A damp patch on the inside wall is directly underneath the overhanging eaves of the silage shed on which there is no gutter. Water from the silage shed falls onto the roof of this shed. The flashing at the junction of the lean-to with the silage shed is leaking.

[150] In this case the claimant had no cause before the incident to examine the buildings for any of the problems of which he now complains. The claimant accepted that his buildings were in need of maintenance. He said that he had not done any maintenance in case he should be accused of "tampering" with the buildings. In my view, that is not a credible explanation for the present state of the buildings. In my view, the present condition of the roofs is consistent with their age, with poor workmanship or materials, and with lack of maintenance over a number of years before the incident.

[151] This firm has made a claim in respect of three sets of farm buildings at Burnbank, Sand Lodge and Vanlop. Mr John Henderson became aware of curling and cracking in sheeted roofs at all three sites about a year after the incident. The evidence about his firm's buildings was concerned with those at Sand Lodge, the only site that Mr Walton inspected. At Sand Lodge there are seven buildings, namely calves house, store, barn, byre, dairy, sheep house and implement shed. In the roofs of these buildings, and particularly on the roof of the barn and byre, the implement shed, the rear elevation of the store and the sheep house, there is extensive lifting of the asbestos cement sheeting, which in some cases has pulled the nails off the purlins. These roofs appear to show worse distortion than in any of the present cases. The worst example is on the roof of the dairy where it is possible to put a 250 mls paint tin into one of the gaps between the sheets. Mr Walton said that he had never before seen curling of this severity and could not explain it. On the roof of the calves house there are cracks and areas of crumbling. There are cracks in the sheets on the roofs of the barn, byre and dairy. To a greater or less degree there is leakage on all of the roofs other than the byre roof.

[152] Mr Henderson, who spoke to these defects, exaggerated on several material points. He claimed that the roofs had been "perfect" before the incident, had given him "absolutely no problem", but had now been "absolutely destroyed". He said that there was no poor workmanship in Shetland and that there was no need for maintenance. I am not prepared to accept that these faults occurred only after the incident. Messrs Henderson bought this farm in 1980. Mr Henderson has no knowledge of the ages of the buildings. It is obvious that the calves house and the store are of a considerable age. These buildings show signs of long-term deterioration. Several suffer from lack of maintenance. In the store roof the purlins have sagged slightly. The barn and byre roofs have an uneven appearance. In the implement shed some of the glass is missing from the windows. In this case there is evidence in each building of long-term deterioration referable to age and lack of maintenance.

[153] The essence of these cases is that within the space of few days these roofs suffered the deposition of dispersant and that the dispersant caused damage that became apparent over the next two and a half years. One of the critical problems for each claimant is therefore to establish that the roofing material in question now shows damage that cannot be explained by recognised causes such as normal ageing or wear and tear and that its occurrence can be linked to the incident by certain significant indicators. In my view, it is obvious that a proper approach to this problem should involve a control study.

[154] In a fax dated 3 November 1994 Dr Brian Dicks of the International Tanker Owners Pollution Federation Limited advised IOPCF on the problem of identifying the causes of the damage allegedly done to the roofs. At that stage oil rather than surfactant was thought to have been a possible cause of the damage. Dr Dicks proposed an assessment procedure to confirm that the property was within an area that could reasonably be expected to have been subjected to oil contamination; to verify that the contamination actually occurred; and to select comparative controls. The controls should be selected on as many as possible of the following criteria; namely, (a) that they should be unaffected by oil; (b) that there should be similarity of exposure to the elements, especially sun, rain, wind and salt spray; and (c) that there should be a similarity in the building materials. He suggested that whenever possible buildings should be selected with identical makes of materials, preferably from the same supplier and of similar age. Dr Dicks suggested that the larger were the number of controls the better.

[155] In my view, the essentials of this methodology remained valid once dispersant rather than oil was alleged to have caused the damage. This methodology would have involved a programme of monitoring of the claimants' properties and of the control properties. The evidence for the claimants was not based on a systematic methodology of this kind. That, I think, is a fatal weakness in the claimants' case.

[156] But before any control study could even begin, it would be necessary to identify the limits of the deposition of the dispersant so that properties could be identified that were indisputably within or beyond its reach. The claimants have failed on this essential point.

[157] From 1993 until the end of 1996, Mr Tulloch inspected many roofs all over Shetland. In his judgment, some roofs in the south mainland had aged prematurely and displayed curling which, in certain weather conditions, could be extreme. He had not observed this phenomenon before the Braer incident. He had found nothing comparable in other areas. He agreed that before the incident he had never inspected asbestos cement roofs for curling. He could not identify the cause of the damage but he was confident that in each case the problems occurred after the incident and were a consequence of it. In his view, if asbestos cement roofs in Shetland had shown such signs before the Braer incident, that material would have ceased to be used in Shetland long ago. He considered that the damage had been caused by the Braer incident and he had advised the Braer Claims Office to that effect.

[158] The difficulty in the evidence of Mr Tulloch is that his approach is unscientific. He has not systematically monitored the roofs and taken periodic photographs and measurements. He has not taken samples for analysis. He has not identified and monitored control properties. When he examined roofs in Shetland in the years before the incident, he was not specifically looking for evidence of curling and distortion. He seems to have relied on his general impressions as to the state of the roofs before and after the incident; but before the incident he was not familiar with any of the roofs with which these claims are concerned.

[159] Mr Thompson accepted that cyclic curling commonly occurs in asbestos cement roofs. While he could offer no technical explanation, he was sure that normal cyclic curling was not the problem in these cases. In his view, this was a more serious problem. He had seen around 5000 roofs over the last 20 years, of which about one-third were covered with asbestos cement materials. He had never seen distortion to the serious extent that some of these roofs display.

[160] Mr Thompson spoke of having seen a roof at Unst, in the north of Shetland, which on any view was unaffected by the Braer incident. This was similar to the roof of the claimants Mr and Mrs Geoffrey Brown. He saw it on a reasonably good day and it was entirely flat. Since there is so little evidence about this roof, and since its location is not even specified, I have given no weight to this evidence.

[161] Mr Thompson was an impressive and moderate witness; but the main difficulty in his evidence lies in the fact that, like Mr Tulloch, he did not carry out any systematic and detailed study of the roofs or monitor them in different seasons and weather conditions; nor did he attempt to carry out a control study. Like Mr Tulloch, he had no knowledge of the state of the claimants' roofs before the incident. He did not inspect any of the claimants' roofs until May 1999 when he was brought into the case. Before then he had never had occasion to investigate tiled or sheeted roofs specifically for curling.

[162] This was the first occasion on which Professor Willey had been asked to consider the question of distortion in asbestos cement roofing materials. There are three serious limitations to the value of his study. The first is that, because of the claimants' limited budget, he did not inspect any of the claimants' properties and had to rely on photographic evidence supplemented by what he understood to be the facts reported by the claimants. The second is that for the same reason he carried out no research, and no laboratory experiments, and took no samples of any of the materials in these cases, to test the hypothesis that he advanced. The third is that Professor Willey's hypothesis depends on certain technical questions regarding the properties of cement, of asbestos cement building materials and of oil spill dispersants. All of these are matters in which he has no special expertise and in which he was not supported by witnesses from any of the relevant scientific disciplines.

[163] Professor Willey considered that on the published evidence as to the aftermath of the mishap it was likely that surfactants had been blown extensively over the southern end of the mainland. He suggested that the evidence established that after the Braer mishap curling of asbestos cement tiles and sheets had been observed in areas of Shetland that had been affected by the pollution. That phenomenon had not been seen previously in those areas; nor had it been seen in other parts of Shetland that had been unaffected by the pollution, nor on the Scottish mainland. In Professor Willey's view, this called for an explanation.

[164] I do not accept that the claimants have proved that dispersant was deposited on any of the roofs in these cases in any significant amount and for any significant length of time. But even if they had proved that, there would have remained the question whether the dispersant was linked to the damage alleged. On that question Professor Willey acknowledged that while an explanation was called for, he himself could not offer one that was supported by any form of proof. He supported the methodology proposed by Dr Dicks to IOPCF. He expressed regret that Dr Dicks' suggested research method had not been followed through. He was merely in the position of suggesting proposed models to explain these phenomena, all such models being subject to verification by appropriate scientific investigation.

[165] Mr Brown had first hand knowledge of the condition of his roof in the years before the incident. Mr Brown was an impressive witness. He was honest, moderate and reasonable in his assertions. He was convinced that the defects in his roof were out of the ordinary, were not readily explicable and had not become apparent until after the incident.

[166] The other claimants who gave evidence had experience of their roofs in all seasons and weather conditions. The burden of their evidence was that the problems now complained of had not become apparent until some time after the Braer incident and had not been encountered before it.

[167] Mr MacFarlane said that there were other roofs in other parts of Shetland and elsewhere in Scotland exemplifying curling to an extent similar to that shown in these cases. He spoke of some unspecified private houses on the outskirts of Lerwick which, so far as he could judge from a distance, had tiles showing a degree of curling similar to that on Mr and Mrs Johnston's property. At various other properties, all unspecified, located on the Sumburgh-Lerwick road, at Bixter and in Lerwick itself he saw curling on sheeted roofs varying from 3 mms up to 7-10 mms. He also spoke to having observed curling on tiles in varying degrees from minimal up to the extent seen on Mr and Mrs Brown's roof at locations, again unspecified, in South Lanarkshire and West Lothian. He had also observed curling on sheets in Dumfries and Galloway, Glasgow and West Lothian having a displacement varying from minimal up to 75 mms. Mr MacFarlane did not know anything of the age or previous history of any of the roofs in the present cases or of any of the other properties to which he referred. His evidence about the latter is fairly general and in some cases it may not relate to asbestos cement materials at all. For these reasons I have not given any weight to this part of his evidence.

[168] Mr Walton's investigative method was to inspect the properties, to allocate them to three zones, to tabulate the defects found on his inspections and to assess whether the results were statistically significant.

[169] Zone 2 in this study covered Bigton and South, and was the area closest to the wreck. Zone 2 covered Scalloway, Trondra, Burra Isle and West Mainland. Zone 3 covered Cunningsburgh and Sandwick. The criterion by which these zones were delineated was not explained at the proof. Mr Walton did not delineate the zones. He understood that they were meant to represent the areas that had been affected, in different degrees, by the oily spray. Mr Walton considered it to be virtually certain that the zones did not relate to the areas of distribution of the dispersant. Since the evidence is not clear on this point, I cannot assess the significance of whether a particular property is within or outwith one of the zones. There is no basis for knowing whether those properties that were within a zone received dispersant or whether those that were outwith it did not. In the absence of an explanation of these matters, this part of the exercise lacks cogency. But there is another more practical difficulty about this exercise, namely that the categorisation of the defects depended on the findings of Mr Walton's inspections and his interpretation of the data recorded in his inspection sheets. I do not regard his inspection findings as reliable.

[170] Mr Walton also referred to his having seen curling of a generally similar nature on other roofs in England, for example on tiled roofs in Birmingham and on sheeted roofs at a car park in St Albans and at a stable in Bedfordshire. Since the roofs in Birmingham may have been affected by a manufacturing defect, and since he did not inspect the other properties in detail, had no knowledge of their ages or of any surrounding circumstances and had not identified the materials from which they were made, I have given no weight to this evidence.

[171] This is a chalet-style house of recent construction. Mr Walton viewed this house through a zoom lens from a distance of about 75 yards. His photographs of it show curling in the roof tiles. He does not know of what material the roof slates are made, contrary to the statement in his Report (at p. 281) that they were made of asbestos cement. There is at the least a doubt as to whether this house was roofed with asbestos cement materials at all. In view of its age, it is likely that the roof tiles are of fibre cement. There has been no published experimental work on such slates comparable with that set out in the Jones Report. I am not prepared to assume that a fibre cement compound has the same properties as an asbestos cement compound or is a valid comparator for the purpose of the present cases.

[174] This too is an old roof that shows a medium degree of curling. Mr Walton did not examine this roof thoroughly or offer any explanation for the curling.

[175] Mr Walton also referred to other control properties in his Report to which he did not refer in detail in his evidence. I understood his evidence to be that these roofs, like those that I have just discussed, simply show examples of the complaints made by the claimants in these cases. None of these had been thoroughly investigated.

[176] This is a modern building situated at Scord, several hundred yards south of the end of runway 27 at Sumburgh. It is unfortunate that Mr Walton failed to establish when it was built. It would not have been difficult for him to find out. On a balance of probabilities I hold that it was built before the Braer incident. This building was therefore in close proximity to the area in which dispersant was sprayed. According to Mr Walton, it is roofed with asbestos cement tiles, but Mr Walton could not confirm that this was the case. The evidence establishes that on the date of Mr Walton's inspection the roof tiles showed no signs of curling or distortion.

[178] These are asbestos cement sheeted roofs in the vicinity of Miss Fowlie's and Mr McLachlan's properties. They appear from Mr Walton's photographs to show no defects. Here again, Mr Walton did not inspect the roofs nor enquire as to their history.

[179] I consider Mr Walton's investigations of the control properties to have been so inadequate that I am not prepared to place any reliance on this line of evidence. In the light of the other evidence and my conclusions as to the reliability of the witnesses on this issue, I conclude that, with the exceptions of the roof of Mr Black's upper garage and the roof of Mr and Mrs Johnston's garage, each of these roofs exhibits signs of deterioration. In some cases, namely those of Mr and Mrs Brown, Miss Fowlie and Mr and Mrs Johnston, the deterioration appears to be premature. In the others, namely those of Mr Black and Mr McLachlan, it appears to all intents and purposes to be consistent with the normal ageing process and with deficiencies in construction and maintenance.

[180] Each of the claimants is sincerely convinced that the defects occurred only after the incident. I am not persuaded that that is the proper conclusion to draw. Almost all of the evidence on this point related to the dates at which the defects were first noticed. In the aftermath of the Braer incident householders and farmers all over south Shetland were alerted to the state of their roofs. Many roofs had been covered with the oily spray. Soon after the incident it was commonly known that the spray had damaged felt roofs and that the Braer Claims Office was accepting liability in principle in such cases. Owners took a closer look at their roofs and became aware of defects that they might not otherwise have noticed. By the end of Mr Halford's survey in May 1995 of roofs for which claims had been made, the number of roof claims had risen from 40 to over 180. It is not surprising that when the present claimants and their witnesses first observed defects in asbestos cement roofs they should have associated the defects with the incident.

[181] The claimants and their witnesses accepted that before the incident they had not inspected their roofs for signs of any of the defects to which these cases relate. With the exception of Mr Black, the claimants noticed the defects only after the defects were pointed out to them or after they had heard of the theory that the spray had damaged asbestos cement roofs. I am not prepared to accept the argument that the defect must have emerged after the incident because the claimant would have noticed it if it had emerged before then. An owner who had no reason to suspect that there was such a problem would not be as alert to such defects as he would be once the problem was at the forefront of his mind.

[182] In these cases there is no evidence, other than the say-so of the claimants, that these roofs did not display their present defects before the incident. In the tile cases Mr and Mrs Brown have produced a photograph of their roof taken before the incident; but that photograph simply records the state of the roof on an isolated occasion. Since the defect in that case is cyclic such a photograph has limited evidential value. In the sheet cases there is strong evidence that the defects are of long standing. In my view, in the cases of Mr and Mrs Brown, Miss Fowlie and Mr and Mrs Johnston, it is as probable as not that the defects existed, or the processes leading to them had begun, before 1993. In the cases of Mr Black and Mr McLachlan the probability is that the defects existed for a considerable time before the incident. The claimants have therefore failed to establish this crucial point on a balance of probabilities.

[183] In my opinion, the claimants have failed to prove that there is a significant pattern in the defects reported in these cases.

[184] The claimants have failed to prove that there was a significant coincidence in time. Even if one assumes that all of the defects first emerged after the incident, all that is left is evidence that the defects were first noticed at a series of dates extending over the years 1993 to 1995. There appears to be no significant pattern in these dates. Nor is there any apparent relationship between the date of a defect's being first noticed and the proximity of the roof to the site of the wreck or to the areas of spraying. One might suppose that if the spraying operation was the common cause of the defects, the effects of it would have emerged at about the same time. There was no cogent evidence to suggest that even if they did not emerge until the various times at which they were first observed, that pattern nonetheless supported the claimants' case. The claimants have offered no experimental evidence to suggest a time-scale within which damage of the kind alleged would manifest itself.

[185] The claimants have failed to prove that there is a significant coincidence in place. There are other roof claims from south Shetland that have not come to proof; but the present claims were presented on the basis that, as a group, they showed the necessary similarities in location. The weakness in this part of the claimants' case is that it assumes, rather than proves, that the condition of these roofs represents a state of abnormality made significant by their location. The claimants have not established the boundary of the area within which the allegedly significant damage has occurred. If one looks at a map of Shetland, or even a map of that part that lies south of Lerwick, it may appear that the claimants' properties are in a cluster in an area close to the areas of spraying. On the other hand, if one delineates areas all points in which are within, say, one, two, and three kilometres from the nearest point at which spraying took place, there is no obvious pattern. In my view it is unlikely, in the absence of any better evidence on the point, that all of these roofs form a significant cluster when one considers that while the buildings of Mr and Mrs Brown, Miss Fowlie and Mr McLachlan are about one kilometre from, and to the east of, the nearest point of spraying, Mr Black's buildings are two kilometres or more from, and to north of, the nearest point of spraying, Messrs Henderson's buildings at Sand Lodge are about 15 kilometres to the north and are on the east side of the mainland, where no spraying took place, and Mr and Mrs Johnston's buildings are about 20 kilometres to the north and are also on the east side of the mainland.

[186] The claimants have failed to provide a context in which the cogency of the circumstantial case can be assessed. The evidence proves, and I think that it is acknowledged on the claimants' behalf, that in the south mainland, and particularly around the Bay of Quendale, there are numerous asbestos cement roofs that show no signs of deterioration and that curling, cracking, softening and the like had been known to occur in asbestos cement roofing materials for many years before the incident. These cases are based on the averment that the damage was caused by dispersant. It is therefore for the claimants to prove inter alia that their properties were within the area or areas in which dispersant landed and not merely within the area or areas in which the oily spray landed. The court has been given no suggestion by the claimants as to a geographical boundary within which it would be appropriate to examine the problem. There is simply no reliable evidence to show what were the probable outer limits of the deposition of the dispersant. It therefore becomes impossible to assess what proportion of the asbestos cement roofs affected by it are displaying symptoms of deterioration, or to assess whether there is a pattern of any kind in the incidence of such cases. If such a boundary were defined, there could be an assessment of the proportion of similar roofs showing or not showing defects within the area of deposition of dispersant and of the incidence of roof defects in similar roofs of similar ages and in similar types of location in comparison areas elsewhere.

[187] By reason of the claimants' failure to identify an area or areas in which dispersant was deposited, even approximately, it has been impossible for them to adduce a proper control study. Their witnesses have referred to properties in places such as Unst that cannot conceivably have been affected by the spray; but the evidence about such properties has been so unspecific that I have been unable to place any reliance on it. On the other hand, when one comes to the properties in the south mainland that have been considered in this proof, it is impossible on the evidence to say whether some of them were within the area of deposition of dispersant, and may therefore have been damaged by it, or were outwith that area, and are therefore more properly to be regarded as controls.

[188] In the absence of an identified boundary of the area of deposition of the dispersant the claimants' fallback position is that the court should infer that, wherever that boundary may have been, each of the claimants' properties and that of Messrs Henderson must have been situated within it. They ask the court to make that inference from the presence of the defects, which they say did not exist beforehand, and from the consistency of those defects with a particular scientific hypothesis. That, I think, is to approach the problem from the wrong direction. In my view, the correct approach is to establish the boundary first of all and then to see whether the property in question lies within it. It is only when the area of deposition of dispersant has been defined that one can make proper use of evidence of control properties in order to assess whether the defects in these cases are out of the ordinary run of defects for which well-recognised mechanisms provide a satisfactory explanation, and can be linked to the dispersant.

[189] I have been critical of Mr Walton's evidence as to control properties; but there is no onus of proof on the objectors. In my view, the claimants have failed properly to support the evidence as to the claimants' properties with comparative evidence of control properties. Even if no other evidential difficulties have confronted the claimants this would, I think, have been fatal to the circumstantial case.

[190] The claimants have also failed to prove that the occurrences are significantly coincident in circumstances. All of these are defects that are known to occur in asbestos cement roofing materials. The problem for the claimants is to demonstrate that their occurrence in these cases has a special significance that is related to the incident. The defects complained of are not uniform, either in nature or in distribution. The symptoms include curling of tiles, with or without removal of the resin coating, some curling being cyclical and some permanent; and cracking, splitting, softening and porosity of sheets. In Mr Black's case the curling on one roof is permanent while on the others it is cyclical. In Mr McLachlan's case the problem seems to be that of leakage through the sheets. Moreover, in some cases, such as that of Mr and Mrs Brown, the defects appear to have occurred once and for all, whereas in others, such as those of Miss Fowlie and Mr McLachlan, the deterioration appears to be continuing. The same can be said of the distribution of the defects. For example, on Mr and Mrs Brown's roof the curling is worse on one slope than on the other. In Mr Black's case, one roof is more severely affected than the others. Another roof within sixty yards of Mr Black's property is not affected at all. Professor Willey suggested that the irregular distribution of the defects results from localised wind vortices. There is no scientific evidence to support this interpretation and I am not prepared to speculate that it is correct.

[191] The extent of the defects has no direct relationship with the proximity of the site to the areas of spraying. For example, Mr McLachlan's sheet roofs show only slight curling, whereas Mr Black's sheet roofs, which are situated to the north further away from the areas of spraying, show significant curling. Messrs Henderson's roofs, which show the worst single example of curling in asbestos sheets, are among the furthest away.

[192] I have come to the conclusion, for the reasons that I have stated, that the circumstantial case has not been proved on a balance of probabilities on any of the crucial facts on which it depends. The cases therefore fail on that account.

[193] Even if the claimants had proved the circumstantial case on all essential points, I would have held that it was not sufficient for them to succeed without scientific evidence as to the cause of the damage. The circumstantial case starts with the occurrence of an exceptional event. It then looks to the exceptional manifestations that followed that event, but had not preceded it, and to the improbability that these manifestations were not consequences of that event. In this way the claimants attempt to make good the deficiency in the scientific evidence by arguing that the circumstantial evidence itself points to the scientific conclusion that there was cause and effect.

[194] This approach overlooks the questions of onus and of the quality of the evidence that would be required to discharge it. The circumstantial case cannot be seen apart from the scientific evidence. It would not be enough for the claimants to demonstrate that a phenomenon that did not occur before the Braer incident, but did occur after it, happened only in an area in which wind-blown spray containing dispersant was deposited. One cannot deduce from that that dispersant must be capable of causing damage of the kind complained of in these cases and did in fact cause it. These are essentially scientific questions that the claimants must prove by appropriate scientific evidence (cf. Dingley, supra, Lord President Rodger at pp 551C-D, 557D-E; Lord Prosser at pp 602D-F; 615D-F; 615I-616A). The point need not be proved to the degree of certainty that a scientist would require; but there must at least be scientific evidence that would entitle the court to hold the claimants' cases proved on a balance of probabilities (Dingley, 2000 S.C.(H.L.) 77, Lord Hope of Craighead at p 86; 1998 SC 548, Lord President Rodger at pp. 600B-C; 601H). There is no presumption to assist the claimants (Wilsher v Essex Area Health Authority [1988] AC 1074; Rhesa Shipping SA v Edmunds [1985] 1 WLR 948, Lord Brandon of Oakbrook at p 951B-C).

[195] Without such evidence the circumstantial case is inconclusive. It leaves open the possibility that the occurrence of the damage complained of and of its alleged cause are merely coincidental (e.g. as in Dingley, supra, Lord Prosser at pp 615; 620F) and the possibility that there is some common cause of the defects other than the cause that is pled.

[196] If, contrary to my view, the circumstantial case had been proved, the question would then have been whether there was scientific evidence to support the conclusion to which it is said that the circumstantial evidence points.

[197] There is no published research evidence as to the effects of surfactants, or a combination of oil, seawater and surfactants, on asbestos cement compounds.

[198] Mr Walton carried out an informal test in which he put surfactant onto new fibre cement tiles, let it remain on them for a day and then washed it off. There was no effect. Mr Walton acknowledged that the test was not representative of the conditions to which the roofing materials were subjected in these cases. He did not regard the test as having produced a conclusive result. Moreover the test was carried out on a different type of material. I have given no weight to this evidence.

[199] In January 1997, at Professor Willey's request, the Institute of Occupational Medicine at Edinburgh University examined two samples of asbestos cement materials taken from locations in Shetland close to and remote from the site of the wreck. This was at a time when Professor Willey was seeking to establish that the dispersant had attacked the asbestos fibres. The samples were examined by electron microscopy and by a technique known as EDXS. Since the sample is not statistically significant and since the results were of doubtful validity, being at odds with published data on the composition of chrysotile asbestos, I have given no weight to this evidence either.

[200] In 1995 Howard Humphreys carried out some research on the question on behalf of IOPCF. Its findings are set out in the Report dated December 1995 to which I have referred. It investigated 17 roofs that were covered with asbestos cement materials. It found that there was a significant correlation between the strength of the material and its exposure to the elements and between the roof damage and the level of chloride in the material. The chloride is attributable to the salt-laden atmosphere in Shetland. Professor Willey considered that these findings were significant evidence of water penetration of the material.

[201] In the course of the proof there was considerable discussion of a letter dated 6 October 1993 written by Mr M G Rich, the Technical Director of Eternit UK Ltd. Professor Willey relied on this letter to support the proposition that unusually high moisture content is associated with curling in asbestos cement (Second Report, para 5). I need not quote the letter. It was written in response to a specific enquiry about oil contamination and power washing, which is not the issue in these cases. Mr Rich had not seen the materials in question and he did not deal with technical questions in any depth. Mr Rich was not called as a witness. I do not regard the statements in this letter as being material to the issue in these cases.

(c) Professor Willey's evidence
[202] In his first Report dated 11 June 1997 Professor Willey proposed a theory that the damage had been caused by a chemical reaction of the asbestos fibres to acidic attack by the surfactant which accelerated the weathering of the materials and caused the formation and propagation of micro-cracks. On further consideration of the matter, and after discussion with the scientific advisers of IOPCF, he resiled from that theory. He did so principally because he was informed by IOPCF's advisers that the surfactants were not acidic. There were also doubts as to the statistical value of the size of the sample analysed by the Institute of Occupational Medicine and as to the reliability of the test results.

[203] Professor Willey later suggested that there had been physical damage to the materials caused by crack propagation. This idea was further refined in his evidence at the proof.

[204] In section 7 of the second Report, Professor Willey starts from the fact that asbestos cement absorbs water. He next suggests that it is "quite reasonable to assume that the airborne plume of oil mist and the airborne plume of the dispersant would have been coincident - particularly in the areas closest to the Braer." For the reasons that I have given, I consider that the evidence suggests the contrary. Professor Willey then puts his proposed model concisely in the following way:

[205] His conclusion is that the dispersant, being designed to assist in the penetration of water into a medium, must have assisted in the penetration of the oil, water and dispersant mixture into the asbestos cement and along the linked pathways of voids within the material. In the case of the tiles, it would accelerate the decay of the factory-applied resin coating. Water ingress along the linked pathways of voids in the material could be substantial. In these cases the tiles would become excessively wet. This process would enable the water to penetrate further into the material and would lead to differential moisture levels throughout its thickness. That would lead in turn to warping in the cycle of wetting and drying and to premature ageing of the material. The phenomenon of curling was likely to vary according to the weather and would be more pronounced in dry and sunny weather than in wet. In the case of the sheets the increased absorption of water, assisted by the high winds, would extend the existing micro-cracks and increase the penetration of water. This penetration would enhance the weathering effect. Such cracks occur naturally in any brittle material. The material breaks whenever stress is applied to the point of the crack. The propagation of such cracks was an aspect of fracture mechanics. It could be extremely rapid. It would cause substantial degradation of the asbestos cement material in the form of cracking, softening and curling.

[206] Professor Willey also suggested a development of his model to explain the deterioration of the asbestos cement sheets, with particular reference to those of Mr Black and Mr McLachlan. He suggested that the exposed edges of the asbestos cement materials tend to delaminate to some extent. Wherever there is delamination, water can penetrate into the asbestos cement matrix.

[207] Professor Willey summarised his overall conclusion by saying that because cement is hydrophilic the deterioration of the cement will promote cracks into which the water, assisted by the surfactant, can penetrate further, faster and in a greater quantity than it would otherwise have done. Professor Willey considered that the enhanced penetration of the water is evidenced by the high chloride content found in the asbestos cement samples examined in the Howard Humphreys report (ibid., para 4.3). On Professor Willey's revised model, it would be expected that chloride from the sea air would penetrate into the asbestos cement. The chloride would therefore be a symptom rather than a cause of the water penetration.

[208] None of the asbestos sheets in these cases had a resin coating. The resin coating on the tiles had already begun to deteriorate by reason of weathering. Since there was no evidence that a resin coating acts as an impervious barrier to the ingress of water into the cement matrix, I do not regard the resin coating as having any relevance to the occurrence of damage in these cases.

[209] Professor Willey accepted from the outset that he had no scientific support for this model. There was no evidence that any of the sheets in these cases was more extensively cracked than other sheets of similar age and type located beyond the range of any possible contamination. He was nevertheless of the opinion that his suggested model provided the likeliest explanation for the curling that had been observed in these cases. The introduction of fracture mechanics as an element in the model was not foreshadowed in either of Professor Willey's Reports. As in the case of his model to explain curling, Professor Willey accepted at the outset that he did not have scientific proof. He nevertheless considered that fracture mechanics provided the likeliest explanation for the defects that had occurred in the asbestos cement sheets.

[210] Mr Walton prepared two Reports in this case, in 1995 and April 1999 respectively. In May 1999 he drew up a critique of Professor Willey's first Report. In evidence he added further points of criticism of Professor Willey's second Report and of the developments of it in Professor Willey's evidence. His fundamental points were that there is no evidence in the scientific literature to suggest that there is any link between the action of surfactants, oil or seawater, or any combination of them, on asbestos cement materials and deterioration of the materials of the kinds alleged in these cases; that he himself knows of no such link, and that he cannot think how such a link could be made.

[211] Mr Walton is not an expert in asbestos but he is an authoritative expert in cement and in asbestos cement building materials. Starting from the undisputed fact that cement is a water-based material that readily absorbs water, Mr Walton can see no reason why water should require the assistance of surfactant to penetrate asbestos cement nor why a sudden ingress of water should degrade it. Professor Willey made some concessions on this general line of argument. I understood him to agree that the effect of surfactant on oil is a molecular process; whereas water penetrates cement naturally and without requiring any agent to assist the process. I also understood him to accept that the absorption of water is not per se harmful to cement.

[212] Mr Walton challenged Professor Willey's hypothesis on two main lines. First, he pointed out that, as is agreed, there is no research evidence to support the view that when water penetrates the cement pores it will propagate cracks. He considered this to be unlikely because the science of fracture mechanics comes in to play only when some force is applied to the material in question and because asbestos cement products, being made of a fibrous material, are not brittle enough for fracture mechanics to apply to them. A conference devoted to the question whether the science of fracture mechanics could be applied to such products had failed to establish the point. Counsel for the claimants objected to this line of evidence on the ground of no record. I allowed the evidence to be led under the usual reservation. I now repel the objection. This evidence was given in response to the introduction of the idea of crack propagation in the evidence of Professor Willey. Although this part of Professor Willey's evidence was not specifically challenged in cross-examination, it was a line of evidence that was not foreshadowed in his reports and was raised by him for the first time in the witness box. It was touched upon in re-examination. In my view, the objectors were entitled to lead this evidence.

[213] Second, Mr Walton considered that Professor Willey's hypothesis did not make sense. Surfactants are designed to be attracted to oil rather than water. Since asbestos cement is a hydrophilic material, there is no reason why surfactant should be attracted to it. There is no scientific evidence that surfactants have any effect on cement or on asbestos. The dispersant was specifically designed to act upon the oil molecules and not the water. If dispersant landed on water, its effect was wasted. The surface tension between water and cement was already very low. It was quite possible that the dispersant molecules would have difficulty in entering the cement pores. Moreover even if, contrary to his view, the surfactant assists penetration by water, the cement already has a significant water content from the manufacturing process and from the external water that it has absorbed. For these reasons the surfactant would not assist water penetration to any significant extent. Mr Walton also commented on the apparent contradiction in Professor Willey's model which relates in part to the idea of increased penetration through the existing voids in the material and in part to the idea of increased penetration by the creation of cracks in the material.

[214] Mr Walton also considered that the surfactant, being water soluble, would in any case be washed off the surface by rain. Even on the most pessimistic view of the amount of dispersant that fell on the roofs, and was almost certainly washed off by the rain soon after, one could conclude that the roofs were not affected by a low concentration of a substance that is not known to have any adverse effect on them. Even in high winds, the pressure of rainwater on a tile or sheet would be low. A 100 mph wind would cause pressure on the surface of less than 1lb/psi. Even if additional water were to penetrate the surface, its immediate effect would be to cause a tile to curl downwards, not upwards, and when it dried it would lie flat. It would take weeks or months for water to penetrate right through to the other side of the tile. The process might be quicker in sheets. He considered that it was highly improbable that the presence of windblown oil or dispersant on the edge of the sheets for a short time could accelerate a normal weathering process to the extent that a process normally taking several decades could take effect in two years or less.

[215] In his first Report Professor Willey suggested that water ingress into the cement matrix could be relatively substantial because of linked pathways in the matrix provided by the chrysotile fibres. Mr Walton disagreed with that on the basis of his own experience of examining micrographs of asbestos cement and on the basis of research evidence which showed that the voids around and within the fibres became filled with hydration products, such as calcium hydroxide, associated with corrosion of the fibres (Investigation of the "Corrosion" of Asbestos Fibres in Asbestos Cement Sheets Weathered for Long Times, Opoczky and Pentek, (RILEM Symposium on Fibre Reinforced Cement and Concrete, 1975)). It may be that nothing turns on this point because it seems to have related to Professor Willey's earlier idea of acidic attack by surfactants that he no longer maintains; but Professor Willey did mention at one stage the possibility that the bundles of fibres, like bundles of glass rods, created additional pathways for the penetration of water. If this is still a relevant issue, I prefer the evidence of Mr Walton, which is supported by research, to Professor Willey's unsupported suggestion.

(e) Mr Lewis' evidence
[216] Mr Lewis contributed to this controversy by remarking that he had never heard it suggested that surfactants are harmful to cement or concrete. They are commonly used to clean oil from cement-based surfaces such as concrete floors.

(i) Have the claimants suffered loss at all?
[229] It was argued on behalf of the objectors that even if the court were to hold that a causal link had been established between the dispersant spraying and the occurrence of the defects of which the claimants complain, none of the claimants has in the event proved that loss was sustained as a consequence. None of the roofs is at any greater risk of damage in strong winds and none is susceptible to leaking to which it would not in any event be susceptible as a result of other defects. For this reason no compensation should be payable to any of the claimants.

[230] I reject this argument. I accept the evidence of Mr Tulloch and Mr Thomson, and that of the individual claimants who spoke to the point, that the integrity of a roof is vital in the climatic conditions in Shetland and that distortion, whether in tiles or sheets, makes the roof vulnerable in conditions of high winds. Moreover, leaking is an additional problem to which they would not otherwise have been susceptible. In my view the claimants would have been entitled to compensation if their cases on the merits had been proved.

[231] If I had found for the claimants, I would have considered it reasonable in each case that the claimant should be compensated, subject to questions of betterment, on the basis of the cost of replacement. The objectors have suggested that in the case of the tiled roofs the claimants could solve the problem by painting the tiles. Some owners in Shetland repaint their tiles when the factory-applied finish wears off. I accept the view of Mr Thompson and Mr Brown that this is not a realistic solution. Repainting might deal with discoloration of the tiles; but it is not clear whether it would remedy the problem of curling. It would effect only a temporary solution. The material would have to be painted in situ. It would require a suitable weather window for the tiles to dry and then be painted, and then for the paint to dry. It would be impossible to match the quality of a factory-applied acrylic finish. In Shetland conditions, it would wear off in 4 to 5 years on an exposed surface and in about 10 years on a less exposed surface. The roofs that were painted would look patchy. Moreover, since in dry weather the affected tiles might be curled, painting them could lock them in the curled position.

[232] The damage is not merely cosmetic. In each case there is a risk that a tile or sheet will blow off in the extreme wind conditions that can be experienced in Shetland, thereby exposing the roof to the risk of being blown off. Although the roofs have survived for over six years since the incident, it is only a matter of time before they are damaged in hurricane force winds. There is also in each case a risk of damage to property and of personal injury. Moreover, there is a likelihood of water penetration, where it has not already occurred. In my view, if a roof is no longer watertight and cannot be made so by normal maintenance, it is no longer serviceable.

[233] It would be unrealistic to remove and replace only the affected tiles or sheets. That would be neither visually satisfactory nor economic. It is true that some tiles or sheets could be replaced; but I accept Mr Thompson's view that the saving would be cancelled out by the extra labour costs involved.

[234] In all the circumstances, I accept the evidence of Mr Thompson that the most practical and cost-effective solution in each case is to renew the roof. I conclude therefore that, subject to betterment, the cost of replacement would have been a reasonable and appropriate measure of compensation.

[235] The claims for the tiled roofs are based on the use of Decra as the replacement material. Mr Thompson recommends this. Asbestos cement tiles, if available, are no longer suitable. The nearest equivalent is fibre cement, which has a similar appearance. Whether fibre cement will be as long-lasting remains to be seen. For this reason there is local resistance to its use.

[236] In Shetland Decra is now the favoured roofing material in cases where asbestos cement tiles were formerly used. Decra tiles are metallic with a granular finish. They are light in weight. This reduces the cost of transportation to Shetland. They can be applied to any roof structure. Decra has similar characteristics of durability in exposed conditions to that of asbestos cement. It is also resistant to wind damage and has a similar guaranteed life-span of about 30 years. It has proved to be reliable in Shetland conditions.

[237] Since the difference in cost between fibre cement and Decra is marginal, I would have held that it was reasonable to use Decra as a replacement material in these cases.

[238] The evidence on valuation consisted of written estimates from local contractors and the professional assessments of the estimates by Mr Thompson and Mr MacFarlane. Mr Thompson is familiar with the special factors affecting costs and prices in building contracts in Shetland. He carried out a study for Shetland Islands Council which indicated that on average costs in Shetland were about 35% above the all-Scotland indices. This was largely because of the restrictions imposed by the weather.

[239] For the objectors Mr MacFarlane carried out a careful study in which he examined Mr Halford's measurements, Mr Walton's Report and the claimants' productions. From published information, Mr MacFarlane derived rates per square metre taking into account costs of labour and materials, overheads, weather, restricted daylight, island location and transport costs, and updated these to current prices. He used standard UK indices. He adjusted his primary data for the special features of Shetland by applying an uplift of 21%. This was a matter of valuation judgment based on experience. Having established approximate costs per square metre, Mr MacFarlane compared these with the estimates lodged by the claimants. He established that there was a licensed tip in Shetland for the disposal of asbestos materials. He then visited Shetland and examined all but one of the properties. The exception was that of Mr Black who was about to leave for the proof when Mr MacFarlane arrived.

[240] I commend both witnesses for their professional approach to valuation and for their lack of partisanship. Mr Thompson, for example, said in most cases that in his view the claimant's estimate was too high. Mr MacFarlane on the other hand readily acknowledged Mr Thompson's advantage of local knowledge. Mr MacFarlane has no direct experience of costs and tender conditions in Shetland. He had not compared published indices with actual tender evidence in Shetland. He preferred to start from UK price indices and adjust for Shetland conditions. In my opinion, local price tender information is a better starting point. Since Mr Thompson's evidence is based on a close familiarity with local tender conditions and not on what was essentially a desk study, I prefer his evidence to that of Mr MacFarlane and in cases of dispute I shall adopt his valuations of the claims.

[241] It is agreed in every case that a replacement roof will bring a degree of betterment. Counsel for the claimants proposed that depreciation on these roofs should be assessed on a straight-line basis over the period of their serviceable lives. A deduction for betterment should be calculated by assessing what would have been the serviceable life of each roof and calculating the proportion of it that predated the Braer incident. If the claimants were entitled to compensation they would ex hypothesi have had defective roofs since the date of the incident, or shortly thereafter. I think that that is a reasonable approach.

[242] The question then is what is to be taken as the serviceable life of an asbestos cement roof. The claimants' evidence shows that in general roofs of this kind have a serviceable life of at least 50 years and in most cases probably more than that. Mr Walton challenged the claimants' evidence on this point on the basis that manufacturers of this material guaranteed it for only thirty years and that there was "a feeling in the industry" that 40 years was a reasonable expectancy. That view is at odds with one of his own scientific references (Building Research Station, Note No D 853 (1962), supra, at p 14) and with the evidence of Mr Tulloch and Mr Thompson who are more familiar with the lifespan of these roofs in Shetland conditions.

[243] Counsel for the claimants suggested that betterment should be calculated on the basis of a serviceable life of 55 years. I accept that suggestion. It seems to me to be a realistic basis on which to proceed.

[245] Counsel for the claimants proposed that betterment should be assessed on the basis of the age of the roof at the date of the incident. Counsel for the objectors argued that the relevant date for this purpose should be the date at which the cost of renewal was assessed.

[246] Since compensation in these cases would be based on the cost of renewal, and since that cost has yet to be incurred, I agree with counsel for the claimants that compensation should be based on up to date estimates and that the date of the commencement of the proof is the appropriate valuation date. However, I do not agree that betterment should be based on the age of the roof at the date of the incident. In my view, a logical and consistent valuation is produced if the proportion of the renewal cost attributable to betterment is assessed on the age of the roof at the valuation date.

[247] Taking X to represent the cost of repair at the valuation date and Y to represent the age of the property in years at the same date, I would assess the value of each claim, after deduction for betterment, by the following formula:

[248] In 1994 the total estimate was £22,754 exclusive of VAT. The updated 1999 estimate was £24,344 exclusive of VAT. Mr Thompson considered that this figure, representing an adjustment of about 10%, was reasonable. I can see no reason why the roof of the upper garage needs renewal. I would have disallowed this item of the claim, the updated estimate for which is £5202 exclusive of VAT. I agree with counsel for the objectors that since the buildings are in use for the purposes of an agricultural business there is no reason why the compensation should include an element of VAT, which the claimant seems to be entitled to reclaim.

[249] The following table sets out my assessments of the ages of the roofs at the valuation date, the breakdown of the 1999 estimate, and my valuation of the claim after allowance for betterment.

[250] The cost of replacement is not contentious. Counsel for the objector accepted that a figure of £10,000 plus VAT, that is to say £11,750, would be reasonable. This is about the figure that Mr Thompson assessed.

[251] At the relevant date the roof was about 24 years old. I would therefore have assessed compensation at £6623.

Miss Fowlie
[252] The estimates in this case are £11,193 plus VAT and £10,000 plus VAT. Mr Thompson would allow £8000 plus VAT, that is to say £9400.

[253] I accept Miss Fowlie's evidence and assess the age of the roofs at the relevant date at about 17 years. I would therefore have assessed compensation at £6495.

[254] The 1999 estimate for the roof of the garage is £3665 plus VAT, namely £4308. Neither Mr Thompson, Mr MacFarlane nor Mr Walton could detect any significant problem in this roof. I would therefore have awarded no compensation in respect of this item.

[255] The 1999 estimate for the roof of the house is £14,843 plus VAT, namely £17,441. Mr Thompson considered that a reduction of 20% from the figure would be appropriate. That gives a figure of £13,952 inclusive of VAT. Counsel for the objectors accepted that that is reasonable. However since Mrs Johnston's claim is time-barred, compensation can be payable only in respect of Mr Johnston's one-half share in the property. The value of Mr Johnston's claim before betterment is therefore £6976.

[256] The roof was 26 years old at the valuation date. I would therefore have assessed compensation at £3678.

[257] The lower of the 1999 estimates includes a figure for the repair of a felt roof with which this claim is not concerned. With the deduction of that item, the total estimate is £38,412 exclusive of VAT. Counsel for the objectors accepted that Mr Thompson's adjustment to that figure is reasonable. That produces a figure of £33,803 exclusive of VAT. I agree with counsel for the objectors that since these buildings are in use for an agricultural business there seems to be no reason why compensation should include the element of VAT.

[258] The following table sets out my assessment of the ages of the roofs at the valuation date, the breakdown of the estimate as adjusted by Mr Thompson, and my valuation of the claim after allowance for betterment.

[259] For the reasons that I have given, I consider that claim No 155, to the extent that it is maintained by Mrs Linda Johnston, must be refused as having prescribed.

[260] On the merits, counsel for the objectors IOPCF and Messrs Robb moved that in the limitation action I should make a determination under section 5(2)(a) of the 1971 Act that no liability has been incurred by either The Braer Corporation or the pursuer to any of the present claimants, and that no amount would, apart from the limit, be due to any of them. I shall make such a determination in respect of each of these claims. In each case I shall sustain the objections and refuse the claim.

[261] Counsel for these parties also moved that in each of the payment actions from which these claims arise I should repel the pursuers' pleas, sustain the third defenders' fourth and fifth pleas and assoilzie the defenders from the conclusions of the summons. I agree that that will be the outcome; but I shall not accede to the motion at this stage. In this judgment I am concerned only with the limitation process. In accordance with the practice in other claims hitherto, the objectors can enrol appropriate motions in the payment processes once the present Interlocutors have been granted.

	[Home] [Databases] [World Law] [Multidatabase Search] [Help] [Feedback]
	Scottish Court of Session Decisions
You are here: BAILII >> Databases >> Scottish Court of Session Decisions >> Anderson & Ors v International Oil Pollution Compensation Fund & Ors [2001] ScotCS 34 (14 February 2001) URL: http://www.bailii.org/scot/cases/ScotCS/2001/34.html Cite as: [2001] ScotCS 34

Scottish Court of Session Decisions