High Court of Ireland Decisions

1. In these proceedings, which were initiated by Plenary Summons which was issued 19th August, 1988, the Plaintiff claims damages for loss it incurred in consequence of the flooding of its supermarket premises at Castle Street, Bray, Co. Wicklow on the night of 25th August, 1986 and in the early hours of the morning of 26th August, 1986 during the course of the storm which has come to be colloquially known as "Hurricane Charlie". That night the River Dargle overflowed its banks and caused extensive flooding in the town of Bray and, in particular, in the area known as Little Bray in which the Plaintiff's premises are located.

2. When the hearing commenced in this Court, the proceedings had been discontinued against two of the Defendants, the County Council of the County of Wicklow and Powerscourt Estates. In broad outline, the bases of the Plaintiff's claims against the remaining Defendants are as follows:

3. At the commencement of the hearing it was agreed by the parties that the quantification of the damages, if any, to which the Plaintiff is entitled should be deferred. This judgment is concerned only with the issue of the liability of the Defendants to the Plaintiff for the damage caused by the flooding of the Plaintiff's premises.

4. The evidence adduced at the hearing can be conveniently considered under the following headings:

5. Paddock Pond was located approximately 14.4 kilometres, measured along the channel of the River Dargle, upstream of Bray Bridge. The reservoir was formed about a century and a half ago by the construction of a dam across the northern end of a natural valley, which was fed by a stream. It was originally part of the Powerscourt Estate. Historical sources suggest that initially its purpose was ornamental. In later years it may have been used to power a sawmill downstream. However, such use had long ceased before the mid-1980's. By then it was in the middle of an afforested area and its only practical benefit to its owner would appear to have been that the roadway which was constructed on the top of the dam afforded the only means of access to the forestry in the area to the west known as Priest's Hill.

6. Immediately downstream of Paddock Pond there is a steep wooded gorge. A watercourse connected Paddock Pond to the Dargle through this gorge. The confluence of the watercourse and the Dargle is somewhat downstream of Powerscourt Waterfall. A number of landmarks and features along the river, downstream of this point and between it and the outskirts of the urban area of Bray, are of significance in the history of the storm. The first is a bridge at the main entrance to Powerscourt Waterfall, which was not damaged in the storm. A short distance downstream a masonry bridge, which most of the witnesses have referred to as Valclusa Bridge, was demolished by the current and pieces of masonry weighing several tons were carried up to a hundred metres downstream. Further downstream, at the confluence of the Dargle and its tributary the Glencree River, Ballinagee Bridge was demolished. Further again downstream, roughly halfway between Paddock Pond and the sea at Bray, at Tinnehinch Bridge near the Golden Gates, the most dramatic incident recounted by an eye witness, which I will outline later, occurred. Another tributary, the Glencullen River joins the Dargle a short distance upstream of the bridge known as Dargle Bridge, which was called Fassaroe Bridge by some of the witnesses.

7. Downstream of Dargle Bridge the river flows through an area known as The Slang and it then passes the factory premises of Lithographic Universal Limited, (the Lithographic premises) which are on the northern bank of the river. At the eastern boundary of the Lithographic premises a steel bridge gave access to Killarney Glen on the southern bank of the river. The steel bridge was swept away in the storm. Below the Lithographic premises the river flows past a housing estate, the Coburg Estate, built on low lying land on its northern bank. A short distance downstream of the Coburg Estate, the river flows past the People's Park which occupies the northern bank of the river from River Lane, which delimits the western boundary of the People's Park, down to Bray Bridge. From River Lane to Bray Bridge the People's Park is bounded on the south by the river from which it is separated by a river wall. It is bounded on the north by a public road, Lower Dargle Road. On the north side of Lower Dargle Road there are terraces of houses. One such terrace, opposite the area of People's Park to the east of the bandstand, is known as Vale Terrace. Another further east, is known as St. Bridget's Terrace. The houses Nos. 1 and 2 St. Bridget's Terrace are at a junction opposite the narrowest portion of the river, where there is a slight bend in the river. This junction is known as Haughton's Junction. From Haughton's Junction, Lower Dargle Road continues to the east to Bray Bridge, Sheridans Lane (called Brabazon Cottages and St. Laurence's Terrace on the Ordnance Survey map) leads in a north-easterly direction to Castle Street, and Adelaide Villas leads in the direction of Greenpark Road.

8. From River Lane east to Haughton's Junction, the People's Park is separated from Lower Dargle Road by railings four feet high which are set in a plinth which varies in height from nine inches to eighteen inches. From Haughton's Junction to Bray Bridge, the People's Park, which tapers off very considerably going from west to east, is separated from Lower Dargle Road by a masonry wall 1.2 metres to 1.3 metres in height.

9. From the very considerable body of evidence adduced as to ground levels certain broad propositions can be deduced. The lowest ground levels are in the area to the north of Haughton's Junction along Adelaide Villas and Sheridans Lane. In general, ground levels fall from the Coburg Estate east towards Bray Bridge; Lower Dargle Road falls towards Sheridans Lane and Greenpark Road falls towards Castle Street. To the east of Haughton's Junction Lower Dargle Road rises towards Bray Bridge.

10. The Plaintiff's store is part of a shopping complex near Bray Bridge at the junction of Lower Dargle Road and Castle Street. The threshold level at the entrance to the Plaintiff's store is 3.86m O.D. I should perhaps point out that all levels mentioned later are expressed in meters over the Ordnance Survey (Malin Head) datum. On the Castle Street frontage of the store there are two carparks which, in general, are below the level of Castle Street. Near its junction with Lower Dargle Road, Castle Street rises in the direction of Bray Bridge.

11. Lower Dargle Road and the area to the north of it, which is commonly known as Little Bray, historically has been part of the natural flood plain of the Dargle.

12. Bray Bridge is a three arched masonry bridge which gives access from Castle Street to Main Street. At normal levels the river flows through the northern and middle arches only. The southern arch has been variously described by the witnesses as a "dry" arch, a "relief" arch and an "overflow" arch.

13. Downstream of Bray Bridge, as it flows to the sea, the Dargle is bounded on the north by Ravenswell Road, which in turn bounds Bray Golf Course. On the southern bank of the river, some distance downstream of the bridge, there is a housing estate known as Seapoint Court housing estate.

14. Overall the Dargle catchment comprises 12,100 hectares. It is a very fast flashy catchment, being very steep with very little storage.

15. The dam at Paddock Pond was a gravity dam. It had two core elements. The main core element was a pillar of blue clay, which was probably six feet wide at the bottom and tapered off towards the top. The other core element, which was masonry, also tapered off towards the top. Between the two core elements there was ordinary earth fill and there was earth on the upstream and downstream faces and on the top of the cores. A roadway was constructed on top of the dam.

16. Four witnesses who resided in the vicinity of Paddock Pond testified as to the use and the condition of the roadway on the dam immediately prior to the storm. Their evidence, which was not controverted, established that there was a considerable intensification in the use of the roadway prior to the storm, when it was traversed by large trucks and heavy machinery in connection with major felling operations in the forestry on Priest's Hill. This led to a deterioration of the surface of the dam, which dipped on the left hand side as one looked downstream. Some weeks before the storm red slatey shale had been spread across the dam, but this did not improve the surface. Mr William McLaughlin (Mr McLaughlin), a Consulting Engineer called by the Plaintiff, was of opinion that the traffic described by the witnesses was likely to have caused damage to the dam, not merely the surface damage which the witnesses observed, but also some breakdown of the structure beneath.

17. All of the expert witnesses who addressed the issue of the failure of the dam were agreed that it is impossible to say precisely what the failure mechanism was. Mr McLaughlin thought that what happened was that the water over-topped the dam at its lowest point - the dip observed by the local residents - and that this would have led to erosion. The over-topping water would probably have cut into the toe of the dam and this would have worked its way back and there would have been a catastrophic movement of the majority of the material in the dam. Mr Alan Kennedy (Mr Kennedy), the Managing Director of Nicholas O'Dwyer and Partners, Consulting Engineers, who was called on behalf of Coillte, advanced a somewhat similar theory. In his view, one could not expect the dam to be level and a dip in the surface was to be expected. Initially as the water in the reservoir rose it would have trickled over the lowest part of the dam and it would have begun to gouge at the top downstream and tend to cut down into the dam. As the quantity of water increased it would have begun to eat into the bottom of the dam, and it would have damaged the bottom and undermined it downstream. Once this happened, the dam would have, as it were, "hinged" open and literally fallen over. Mr Kennedy was of the view that the eventual failure would have been catastrophic and immediate. He emphasised that the water flowing over the crest of the dam would not have had sufficient time to affect the core.

18. At the time of the storm, the only means in existence for releasing excess water from Paddock Pond was a stone culvert constructed in the dam near the top of the dam on its eastern side. In the past on the opposite side there had been a pipe passing through the bottom of the dam with a contraption on the upstream side for controlling flow into the pipe. By 1986 this contraption had been defunct for many years. On the downstream side the pipe had been "deadheaded" - capped off by a large bolt system.

19. On the upstream side the culvert was the traditional rectangular shape with a semi-circular top. While there is no reliable survey from 1986 which establishes the dimensions of the culvert, I accept the evidence of Mr Amhlaoibh O h'Aonghusa (Mr O h'Aonghusa), a Civil Engineer who was working in the State forestry service in 1986, who calculated the dimensions of the culvert on the basis of the photographic evidence which was adduced by the Plaintiff as measuring 1.35m wide and 0.675m in depth as a reasonably fair estimation of the dimensions of the culvert.

20. There is reliable evidence as to the condition of the culvert before the storm from one of the local residents, Mr Hugh Hackett (Mr Hackett). The culvert had silted up with gravel and it was obstructed by vegetation and forestry debris. Mr Hackett investigated the culvert on the downstream side a few weeks before the storm and he discovered lengths of timber varying from 8 feet to 20 feet in length which had obviously come through from the upstream side and had got stuck at the exit and had become tightly enmeshed. Judging by the rot and decay, some of the timber had been in the culvert for a considerable time. The photographic evidence adduced by the Plaintiff of the condition of the culvert after the storm corroborates this evidence. On the evidence, I am satisfied that the culvert was capable of taking some overflow in its pre-storm condition but that its capacity was considerably reduced because of the silting up and obstruction by stones, vegetation and forestry debris. It follows that I consider that it has not been established on the evidence that the location of the culvert in relation to the lowest level of the dam was such to render the culvert totally ineffective.

21. There is no evidence that Coillte's predecessor had any scheme in place for the periodic inspection of Paddock Pond and for the repair and maintenance of the dam and the clearing and maintenance of the culvert or that any such inspections or maintenance or repair works were implemented. On the contrary, the only witness who was called who was in the employment of Coillte's predecessor at the time was not aware of any such programme or activity.

22. The drainage construction works which were being carried out in the vicinity of Bray Bridge prior to the storm in August 1986 were being executed under the Bray Sewage Scheme Stage II Contract Number 2A by Uniform for the Council. The consulting Engineers in connection with the project were Messrs John B. Barry and Partners (Barry). Mr Liam Prendiville, a Civil Engineer, was the resident engineer in connection with the project and he was a temporary employee of the Council. He was involved with the project for six or seven years. Mr Stephen Ceillier, a Civil Engineer, was Uniform's site agent on the work at Bray. In broads terms, the works involved the laying of a sewage pipeline from east to west, extending from a point downstream of Bray Bridge along the southern bank of the Dargle as far as Bray Bridge, then passing through the southern arch of Bray Bridge and continuing along the southern bank for a short distance and then crossing under the river and emerging on the north bank at a point opposite the shopping complex in which the Plaintiff's store is located and then continuing through the People's Park from east to west and crossing under the river again near the western extremity of the People's Park. The works also involved the construction of manholes to facilitate inspection and maintenance of the pipeline at intervals along the pipeline. By August 1986 the pipeline had been brought within a short distance of the downstream (east) side of Bray Bridge. A short distance upstream, to the west, of the bridge a section of pipeline had been laid to the point of the river crossing and the pipeline had been taken across under the river. On the northern bank the pipeline was being laid in the People's Park. Immediately prior to the storm in August 1986 the following works were ongoing: preparatory works for the laying of the pipeline under the southern arch of the bridge; approximately fifteen metres upstream of the bridge on the southern bank the construction of a manhole; and on the other side of the river the laying of the pipeline in the People's Park.

23. The factual evidence to support the Plaintiff's allegation that the river defences were breached in the course of the drainage construction works and that the flow toward and through Bray Bridge was obstructed on the night of the storm by reason of temporary works and machinery was supplied by Mr Ceillier and Mr Prendiville, who had the benefit of some contemporaneous documentation, such as site diaries, to aid recollection of what had occurred over ten years previously.

24. In relation to the allegation that the river defences had been breached, the evidence establishes that a stretch of the river wall on the north side of the river which had separated the river from the People's Park was taken down on 9th June, 1986 and had not been reinstated by 25th August, 1986. Mr Prendiville's recollection was that the resulting breach was between five and fifteen metres wide. Its location was the point at which the pipeline crossing the river reached the northern bank of the river. It was necessary to take down the river wall at this point because it had been built without foundations and it was not possible to lay a pipeline under it while leaving it in situ. At the time of the storm there was construction work outstanding in the area of the breach and that work was not completed until the beginning of October 1986. The only precaution which Uniform had taken to protect the breach was the building of an earthen embankment within the People's Park. Neither Mr Ceillier nor Mr Prendiville had any clear recollection of the location, construction or dimensions of the embankment. It was obviously upstream of the breach and Mr Prendiville's recollection was that it was built at right angles to the river wall.

25. It was not contended on behalf of the Plaintiff that the river wall should not have been taken down to facilitate the laying of the pipeline but it was contended that it should have been reinstated at the earliest opportunity. It was also contended that a clay or earthen embankment was not a sufficient barrier against a possible flood, pending reinstatement, because heavy rain and water movement over an earthen embankment would remove it quickly. Mr. McLaughlin stressed that the wall removed was part of the river defences and that the location of the breach was one of the weakest points in the river defences. His opinion was that a satisfactory temporary protection would have been afforded there by sandbagging. Mr Liam Downes, who at the time of the storm was the Civil Defence Officer for County Wicklow, testified that sandbagging is only a protection against surface water and that it is not effective against flooding.

26. The evidence disclosed only two items of temporary works which would have impacted on the flow of the Dargle on the night of the storm. The first was a temporary road which had been constructed under the southern arch of the bridge to facilitate movement of machinery from east to west in the course of the laying of the pipeline. A photographic survey of the southern arch prior to the construction works reveals that to some extent the southern bank of the river extended under the southern arch. To create the temporary road Uniform excavated the soft silty material and replaced it with suitable material such as gravel from the river bed and some imported material. The resulting surface was at a lower level than the level of the pre-existing bank which extended under the southern arch, the lower level being necessary to create a depth sufficient to allow a crane to travel under the arch. However, the temporary road was wider, that is to say extended further north, than the pre-existing embankment had. Mr. Prendiville's evidence was that the temporary road was necessary to facilitate the drainage construction works. Mr. McLaughlin did not seriously take issue with this proposition. However, he did make the point that, if there was no further use for the temporary road, it should have been removed. He also made the point that the silty sand which had been under the southern arch prior to the construction of the road would have responded to the flood more readily than the material with which it was replaced - larger and more compacted material - allowing greater access for the flow through the southern arch during the storm.

27. The second item was trench sheeting which had been put in place immediately to the west of the manhole which was under construction upstream of the bridge for the purpose of retaining the soil during deep excavation of the trench or tunnel for the pipe line.

28. Mr. Cellier's recollection was that the trench sheeting was 5 m or 6 m in length and 2 m or

29. The construction works were "rained off" on 25th August, 1986. In the early afternoon, having learnt that there was a weather warning, Mr. Cellier walked the length of the works with the site engineer and took such precautions as he considered necessary to secure the site and Uniform's plant and equipment. On the evidence, I am satisfied that the largest and heaviest item of machinery which was on site, an NCK/C 54 crane, was moved back out of the line of the river. The only machine which was left on the upstream southern bank in the line of the river was a Komatsu excavator, which was a "hired in" machine, and for which Mr. Cellier did not have the keys and, accordingly, could not move. At some time prior to the storm, in the course of the construction works, the southern bank upstream of the bridge had been cleared of trees and vegetation and had been benched back to facilitate the construction works. On the afternoon of 25th August the Komatsu was tied to the benching.

30. Mr. Cellier testified that following the storm there was debris in the northern arch and the middle arch of the bridge. A large tree which had been caught in the northern arch was subsequently removed by Uniform. However, there was no debris in the southern arch after the storm.

31. Both the NCK crane and the Komatsu excavator were silted up and had to be dug out and cleaned. Nonetheless, I accept Mr. Ceillier's evidence that the crane was moved on the afternoon of the 25th.

32. On the basis of the factual evidence given by Mr. Cellier and Mr. Prendiville,

33. Mr. McLaughlin expressed the view that the manner in which the works were executed was substantially satisfactory, subject to the reservations he expressed in relation to the temporary road, which I have already outlined. However, in his view, once the river was in spate the direct flow to the south arch would have been blocked by the presence of the Komatsu, the trench sheeting and the works at the manhole.

34. One of the ironies of this case is that one of the objectives underlying the drainage project was to separate foul sewerage from surface water and storm water drainage, there having been hitherto a combined system, in the Bray area with a view to preventing or at least alleviating flooding in the Little Bray area in the future.

35. The Little Bray area had experienced flooding prior to 1986. There had been flooding within living memory in November 1965 and there were records of severe flooding in 1905, coincidentally on the 25th August in that year. However, the storm, which was caused by an offshoot depression from "Hurricane Charlie", which hit the Wicklow and Dublin areas on 25th August, 1986 was exceptional. It occasioned publication by the meteorological service of a supplement to the Monthly Weather Bulletin for August 1986. This publication from its opening paragraph to its concluding paragraph paints a picture of a summer which is best forgotten - a summer which was "a real wash out" with a thundery June, a July which was the dullest in over thirty years and an August which brought a succession of storms which caused flooding in many areas. Of Monday 25th August, 1986 it records that in the worst affected areas the storm was of a severity that would be expected only once in one hundred years on average and it had the "depressing distinction" that no sunshine was recorded anywhere in the country on that day.

36. That gloomy picture is reflected in the rainfall records maintained by the meteorological service. Over the twenty-four hour duration of the storm from 8.00 a.m. (0900 GMT) on 25th August to 8.00 a.m. (0900 GMT) on 26th August the recorded rainfall in the Dargle and Upper Dodder catchments ranged from 80 mm in low lying areas to at least 250 mm in the highest areas with a considerable amount of the catchment which was located at middle levels recording 150 mm to 200 mm rainfall. At Bray Garda Station 86 mm was recorded during that twenty-four hour period. Rainfall of that magnitude at a low lying station has a return period of one hundred years. At Glenasmole, which is just outside the Dargle catchment, 165 mm was recorded for the same twenty-four hour period, with a return period of at least one hundred years. The one day values recorded at Bray Garda Station and Glenasmole are put in perspective when contrasted with the rainfall recorded at the same locations during the November 1965 storm. At Bray Garda Station the total rainfall recorded over three days of the November 1965 storm was 48.7 mm and at Glenasmole 165 mm was recorded over three days.

37. The intensity of the rainfall on 25th August, 1986 was reflected in the rainfall recorded at Casement Aerodrome where moderate to heavy rain persisted for most of the twenty-four hour period commencing at 8.00 a.m., the highest six hour duration total having occurred between 3.00 p.m. and 9.00 p.m. on the 25th and the highest twelve hour duration total having commenced at around noon on the 25th. Although Casement Aerodrome is outside the Dargle catchment, Mr. Denis Fitzgerald, the head of the Climatological Division of the Meteorological Service, testified that the pattern of intensity represented by those records would have been reflected throughout the Dublin and Wicklow area and was consistent with the record at Glenasmole, where the total rainfall recorded for the six hours from 5.00 p.m. to 11.00 p.m. on the 25th was 85 mm. For the sake of clarity I should perhaps state that, although the Meteorological Service records refer to G.M.T., the foregoing times are expressed as B.S.T..

38. While the rainfall statistics speak for themselves, I think it is worth noting that

39. Mr. Fitzgerald, who did not strike me as a man given to hyperbole, resorted to epithets such as "exceptional", "quite remarkable" and "startling" when describing the recorded rainfall during the storm. Mr. Fitzgerald expressed the opinion that in the extreme conditions which prevailed on 25th August the actual rainfall could have been in the region of 10% higher than the recorded rainfall.

40. The overall picture which the data recorded by the meteorological service paints of the storm on the 25th and 26th August, 1986 is of heavy persistent rain which produced exceptional one day rainfall totals which, applying the methodology of the Flood Studies Report, to which I will refer later, produced a return period from one in fifty years to one in one hundred years. The rainfall was accompanied by strong to gale force winds. The storm followed a number of other active weather systems which affected Ireland and gave high rainfalls earlier in the month. There was very little soil moisture deficit even in low lying areas.

41. The eyewitnesses whose experiences on the night of the storm I will record later were unanimous that the storm was unprecedented in its ferocity. Between them they had very many years of experience of living in the Dargle catchment and none of them had previously experienced rain of such duration and intensity. The level of the Dargle was higher on that night and its velocity greater than had been experienced in living memory. In short, their perception of the event was that it was a major natural disaster.

42. The effects of the storm were not just localised in the Bray area. There was evidence of very serious flooding in Dublin City when the Dodder overflowed its banks flooding approximately three hundred and fifty houses, as well as public houses and business premises from Milltown down to Ballsbridge. The rainfall in the Dodder catchment during the storm was the most significant in recorded history.

43. Throughout Co. Wicklow, outside the Dargle and Dodder catchments, bridges were destroyed or damaged during the storm. Annamore Bridge, an early 19 century masonry bridge over the Avonmore was badly damaged. Avoca Bridge, a mid-19 century bridge, on the Avonmore was badly damaged and, upstream of that, the Lions Arch Bridge at Castlehoward, a masonry bridge, was completely destroyed. The Sally Gap bridge high up in the Avonmore catchment was demolished. Ballyward Bridge, a masonry bridge on the Liffey catchment upstream of the Blessington Lakes, collapsed in early September 1986 as a consequence of damage sustained during the storm.

44. In summarising the eye-witness evidence of the flooding of the Dargle I will start at Paddock Pond and follow the river down to Bray Harbour.

45. At the time of the storm Mr. Hackett had been living in Paddock Lake Cottage for four years. The stream which fed Paddock Pond passed his house. While, living at a high altitude, he was accustomed to heavy rain, he had never experienced rain of such intensity over such a protracted period as he experienced on 25th August. At about 7.30 p.m or 8.00 p.m. that evening he noticed that the ground outside his house became almost invisible. The water kept rising and rising very steadily all evening. At about 9.00 p.m. it started to enter the house. At about 10.20 p.m. the water just ran out of the house as if somebody pulled a plug. Mr. Hackett attributed the departure of the water to the dam burst which he became aware of the following day. However, the expert witnesses were agreed that the departure of the water from Mr. Hackett's house was attributable to the release of some local blockage, rather than the dam burst, because of the relative levels of Mr. Hackett's house and Paddock Pond and the dam.

46. Mr. Ó hAonghusa gave a very graphic description of the gorge downstream of the dam when he visited the area after the storm. He said that the erosion was really spectacular. The water from the dam had attacked the glacial drift on either side and cut huge clips in it thirty or forty feet high. It had tumbled down big trees. It had scoured the gorge and the bottom of it was bare rock which looked as if it was sandblasted. There was not a scrap of vegetation left, no moss and nothing even in the crevices. Everything had been scoured out. He had never seen rocks so well cleaned.

47. Mr. McLaughlin, who had aerial photographs taken after the storm, interpreted the aerial photographs as indicating that all of the channels and river sections through which the Paddock Pond waters passed were more severely damaged than the tributaries in the Dargle catchment, for example, the Glencree and Glencullen rivers.

48. Moving downstream past the confluence of the watercourse from Paddock Pond and the River Dargle, Mr. John Hunt experienced the demolition of Valclusa Bridge. Mr. Hunt's house is located beside a stretch of the Dargle between two bridges: upstream is the bridge at the main entrance to Powerscourt Waterfall, which he referred to as the Waterfall Bridge; and Valclusa Bridge is downstream. Mr. Hunt had been living at this location since 1963 and his house had never been flooded before the 25th August, 1986. Between 10.00 p.m. and 11.00 p.m. that evening, he became concerned. He heard a very unusual noise outside, which he believes was the sound of boulders coming down the river. When he investigated, he saw water starting to come into his yard. He secured his house and then went out to investigate further. He saw a lot of trees and material coming down the river. At the Waterfall Bridge, the water was flowing over the bridge and underneath it was completely swelled up. Water was coming from every direction right around his house. It seemed to come in a "gush" rather than a steady rise. It came into the house and, after about five or seven minutes, it left the house again. When Mr. Hunt investigated, he discovered that Valclusa Bridge was gone. The river was still very high.

49. Mr. Forrestal, who was the County Engineer for County Wicklow at the time of the storm, with responsibility for roads and bridges in the county, attributed the collapse of Valclusa Bridge partly to the failure of the dam at Paddock Pond. When the water from Paddock Pond came thundering down the valley, Valclusa Bridge was carried clean away and pieces of masonry weighing several tons were carried up to 100 metres downstream. Mr. Forrestal attributed the demolition of Ballinagee Bridge further downstream also, in part, to the failure of the dam at Paddock Pond.

50. Mr. Paul Barnwell was in Enniskerry on the night of 25th August, 1986. While endeavouring to find a safe route back to Bray, where he lived, he arrived at the Golden Gates near Tinnehinch Bridge. Visibility was appalling. As he was approaching Tinnehinch Bridge, there was a lot of water lying on the ground in what he assumed was a dip in the road. There was a car ahead of him going in the same direction. The car ahead proceeded through the water at a fairly slow pace and got through. Having assessed that his own car had as good a chassis clearance as the car ahead, Mr. Barnwell decided to drive through the water. When he was approximately half way through, there was a sudden surge and the car was lifted and appeared to shoot to the left hand side of the road. The water began to come through at the side of the car. Then there was a sudden major surge and the car was hit by a wall of water. The rear of the car swung back and it appeared to drop down. The road had actually collapsed beneath the car. Mr. Barnwell could not open the car door. The car was still in motion. It was moving with the run of the water. Fortunately, he was able to open the car window. At that stage, the water was up to window ledge level and, when he opened the window, the water started coming into the car. He climbed out through the window and clung to a tree. Eventually, his horrific experience came to an end when he was rescued by a man on a tractor. Mr. Barnwell's evidence was that initially when he drove through the flood, the water seemed to be relatively still but after the surge the water was moving. He timed the incident at between 10.30 p.m. and 11.00 p.m.

51. Moving downstream again, Mr. Paul Kavanagh was working in the Lithographic premises that night. Shortly after his shift finished, between 10.00 p.m. and 10.15 p.m., a Civil Defence officer advised the employees to leave the building. At that stage, it was possible to see from the Lithographic premises, which were directly on the north bank of the river, that the river was fairly high. Mr. Kavanagh and a colleague drove to Main Street, Bray, to apprise the factory manager of the position. As they drove along Lower Dargle Road, there was a lot of water on the road, but Mr. Kavanagh could not say whether it was from the river or from the shores (i.e. the drains). Because of the flooding on Lower Dargle Road, they veered up Greenpark Road, via Pearse Road. There was a lot of water along Greenpark Road - the depth of the water was well up to the wheel of the car. They turned on to Castle Street at the Plaintiff's premises. There was generally a lot of water around at the time, that is to say, between 10.15 p.m. and 10.30 p.m. Having collected the factory manager, they returned to the Lithographic premises via Upper Dargle Road, which was not flooded. They made slow progress around Castle Street. The water was rising all the time and they noticed a certain amount of water in the Coburg Estate as they went by. They got back to the Lithographic premises at about 10.30 p.m. to find that, in their absence, the river had reached the wall of the Lithographic premises and there was quite an amount of water in the yard, but not in the factory premises. They started raising electrical equipment off the ground. The water started to come in to the factory premises. They lifted shore covers within the factory premises in the hope that the water would escape through the shores. However, after a time, water was coming in and none was going out. Waste bins which they had put against the back door were floating around. They opened the back door. Forty gallon drums were floating around in the yard. They closed the door and locked it. The next day, it was buckled. They stayed in the factory premises for about forty-five minutes and left between 11.00 p.m. and 11.30 p.m. Mr. Kavanagh estimated that, at that stage, the water level within the factory premises was over two feet. The electrical equipment which had been raised was submerged.

52. Mr. Bernard Collins lived in Killarney Glen across the river and a little downstream of the Lithographic premises. At the time, vehicular and pedestrian access to his property was via the steel bridge to the east of the Lithographic premises. He was at home on the night of the storm. Between 10.30 p.m. and 10.45 p.m., his son came in with the news that the bridge had gone. Mr. Collins went down to where the bridge had been. The banks on the southern side of the river were starting to cave in and were going in stretches of ten to fifteen feet at a time. The water was beginning to come up over the gabions. At about 11.30 p.m., water had come into and was beginning to rise within the house. When the front door was opened, a wall of water came in, like a tidal wave, and ran ,via a thirty foot long hall, out the back door. It took three members of the family to close the front door. There was about 3.5 feet to 4 feet of water outside the back door. The force of the water was horrendous. The family decided to evacuate the house via the back door. The water was waist high and for a distance of about thirty yards it was level. Thereafter, the ground rose steeply up into the wood and the family escaped by that route. Mr. Collins stayed around until about 3.00 a.m. or 4.00 a.m. on the 26th August. The water had peaked by this time and was not rising anymore. The expert opinion was that the demolition of the steel bridge was not related to the arrival of the waters from Paddock Pond into Bray.

53. Mr. Liam Doyle (Mr. Doyle), a Civil Engineer employed by Barry at the time surveyed the River Dargle from Dargle Bridge down to Bray Harbour in the immediate aftermath of the storm. He described the devastation along the banks of the river from Dargle Bridge down to Mr. Collins' premises. At Riversdale House, which is located about 3 km upstream of Bray Bridge, the northern bank was severely eroded, a gravelled area around the house was destroyed by the flood and there was extensive collapse of the lower bank, walling and pathway. Further downstream, a masonry wall, part of the old river defences, had collapsed. Upstream of the Slang, there was scouring on both sides and the erosion by the flood was very severe and it was obvious that the river had widened the channel. At the Slang on the northern bank, a three metre length of footpath had disappeared into a hole on Upper Dargle Road. At the Lithographic premises, which are located about 2 km upstream of Bray Bridge, a river wall was undermined and a slab on which L.P.G. tanks were sitting was cracked and had started to give way but, fortunately, the gas tanks had stayed in position. The northern bank was severely eroded and an out-building, which had been part of the Lithographic premises, was totally demolished. The northern bank, opposite Mr. Collins' house, was severely eroded. In Mr. Doyle's opinion, the scouring at this location was caused by the high velocities caused by the restriction which the steel bridge presented before it was swept away. The southern bank was severely eroded near Mr. Collins' house and there was extremely severe scouring. A whole section of bank - four metres or five metres long - had been scoured out by the river. The gabions were damaged or destroyed on this stretch.

54. Of the three residents of Lower Dargle Road who testified, Mr. Anthony Crinnion, who at the time lived in Vale Terrace, lived furthest away from Bray Bridge. At 9.30 p.m. on the night, Mr. Crinnion left his house to move his father's car to higher ground. As he did so, he noticed a huge pool of water right outside his door going down Pearse Road. Driving along Lower Dargle Road, he noticed that water was surging up through the iron grids on the drains. In fact, it was cascading out of the drains. A neighbour's car was on top of one of the drains and the car was actually bouncing with the force of the water. Having parked the car at Main Street, Mr. Crinnion walked home across Bray Bridge. He noticed that a tree was stuck in one of the arches of the bridge, either the northern arch or the middle arch, but probably the northern arch. When he reached Maitland Street on the walk back, the water was about twelve inches deep. The water was constantly rising slowly but, at some time between 11.00 p.m. and 11.30 p.m., there was a massive sudden upsurge in the level of the water, which rose from what Mr. Crinnion described as "knee level" and "shin level" to half way up the thigh and this rise occurred within three or four minutes. The water continued to rise. At about 11.30 p.m., Mr. Crinnion's family were concentrating on getting their possessions upstairs. Water had not actually started to come into the ground level of the house because they had blocked the door. Eventually, however, it started coming in through ventilation blocks and in under the floors and was rising from that source. About 12.20 a.m. on the 26th August, they decided to "throw" the master switch and to go to bed. As he looked out from the first floor of his house on to Lower Dargle Road at about that time, all Mr. Crinnion could see was the top of the railings on his own wall and the spikes of the railings at the Peoples Park. Everything else was submerged in water. The next morning, it was clear that the maximum water level within the house had been about three to three and a half feet but by then the water had totally receded.

55. Mr. John Stirling lived at No. 2 St. Brigid's Terrace immediately to the west of Haughton's Junction. After 9.00 p.m., he noted that water was flowing along Lower Dargle Road and that it was flowing fairly fast into Brabazon Cottages. He decided to move his caravan to higher ground and did so. When he was returning home about 10.00 p.m., he was told by a Civil Defence officer that Lower Dargle Road was impassable. Nonetheless, he persisted along Lower Dargle Road and reached his home. At that stage, the water was flowing very fast. It was probably about two feet high. He started sandbagging the entrance to his house. While he was in the course of this operation, there was a sudden surge and the sandbags collapsed. He made for the house but the water was ahead of him. As he was climbing the stairs within the house, he thought the water was not going to stop. It went up about six to six and a half steps of the stairs. He timed the entry of the water into the house at around 11.00 p.m. but he was unsure of the precise time. From his first floor window, he saw trees floating down the river and possibly flowing down Lower Dargle Road. It was not possible to distinguish between the river and the flood outside the river. He observed that there was a fire engine parked at Haughton's Junction and it was being lifted and it was hitting off the wall. It hit off the wall two or three times and it seemed as if it was being forced around a corner. There were people on the fire engine at that stage. They were subsequently rescued by somebody who came along in an inflatable dinghy. Mr. Stirling said that the main flow was going in the direction of Brabazon Cottages and St. Laurence's Terrace.

56. Mr. Vincent Eaves lived next door to Mr. Crinnion at No. 1 St. Brigid's Terrace. Shortly after 9.00 p.m., Mr. Eaves crossed the Lower Dargle Road and he entered the Peoples Park through temporary wooden gates immediately opposite Haughton's Junction. He could hear the river, which was very loud. He walked across the Park to the river wall and he could touch the water which was less than a foot away from the top of the wall. When he turned to go back to his home, he noticed that there was a lot of water further down the Park, that is to say to the east, perhaps seventy feet away, which seemed to be flowing slowly up the Park, i.e. westwards towards him. Mr. Eaves, in the course of his testimony, indicated where the water stopped by reference to a map of Little Bray which had been put in evidence. The point he indicated was the source of controversy later during the hearing. I believe that Mr. Eaves pointed to an area within the Peoples Park opposite No. 3 Dargleview Terrace. Although it was dark and he could not see down towards the water, Mr Eaves' belief was that the water was coming from the direction of the breach in the river wall in his direction. When Mr. Eaves re-crossed Lower Dargle Road, there was only rainwater on the road. Later, at about 9.45 p.m. or 10.00 p.m., he decided to go to the Boghall Road to get sandbags. As he drove along Lower Dargle Road, there was a lot of rainwater but he was able to drive. The round trip took him about twenty-five minutes. When he got back home and stepped out of his van, the water on the road had risen to just over the edge of his wellington boot below his knee, which surprised him. He realised it would be futile to do sandbagging. He went into the house and the water seemed to be rising fairly quickly. He arranged for his family to move upstairs. Thereafter, the water rose to about a foot. He then went upstairs himself. The water seemed to rise very quickly and, in a short space of time, it rose from the first step of the stairs to the fifth step of the stairs. The next day he measured the level of the water inside the house at three feet nine inches. Mr. Eaves also observed the fire tender. He saw it reversing down Adelaide Road and getting stuck with the personnel on it. They could not get out because the water rose so quickly. The boundary wall of his own house was submerged. He saw a neighbour's garden seat floating pass. The flow was going in the direction of Brabazon Cottages and the neighbour's garden seat went in that direction.

57. Two of the eye witnesses were employees of the Plaintiff at the time of the storm. Mr. John Garland was the security officer in the Plaintiff's store at Castle Street. At 11.10 p.m. on the night of the storm, he was called to the store by the cleaning staff and he arrived there at about 11.20 p.m. When he got there, there were pools of water in the car park but he assumed that the pooling was the result of falling rain. He dealt with the cleaning staff's problem and left the store again about 11.35 p.m.. Mr. Garland testified that when he left the store at 11.35 p.m. he did not notice any flooding in Castle Street and there was no water within the store premises. However, he decided to return to the store. He drove home and collected wet gear and he also collected his colleague, Brian Flanagan. He returned to Main Street and parked his car in Main Street. He made a phone call to the local radio station from a call-box on Main Street and he went to the Civil Defence office nearby for information. He then walked across Bray Bridge. The river was exceptionally high. He had never seen it so high before. He looked across the parapet of the bridge into the river. He could see the archways of the bridge. The water seemed to be very high and very close to the top of the arches. He said the water was a couple of feet from the top of the arches. Mr. Garland recollected that there were briars and a tree trunk stuck in one of the arches and he believed that it was the arch nearest to the Lower Dargle Road. There was flooding on the Dublin side of Bray Bridge in Castle Street. He waded through water in the shopping complex car park which was just below the level of his knees. When he got to the Plaintiff's store, it was totally flooded. This was about 12.15 a.m. on the 26th. The level of the water within the store was about eighteen inches and it subsequently rose to about twenty-four inches. He stayed in the store overnight. At about 5.00 a.m., the water seemed to disappear out of the store, but gradually, not instantly.

58. Mr. Flanagan had been working in the store until 10.30 p.m. on the night of the storm. He had walked home. As he crossed the car park of the shopping complex, there were pools of water in the car park. As he crossed Bray Bridge up into Main Street, the rain was falling heavily and the water was rising in the river. He accompanied Mr. Garland back to the store later. When they got to the car park, they had to wade through about eighteen inches of water. On Castle Street, there was water to a height of two and a half to three feet. When Mr. Garland opened the door of the store, the store was flooded. The goods were floating around inside. The water was about twenty inches deep inside. Over the space of the next two hours, the water rose by another six or seven inches within the store. He also stayed in the store overnight. At about 2.00 a.m., he went upstairs to the canteen which afforded a view out on to Lower Dargle Road. The road was not visible: the water had come over the wall and was about ten feet high outside the wall and it was continuing to flow over the wall. The water levelled off within the store at about 3.00 a.m. or 4.00 a.m. and it started to subside at 5.00 a.m.

59. Mr. Downes, in his capacity as a Civil Defence officer for County Wicklow, checked the Lower Dargle Road at about 8.05 p.m.. At that time, water was coming through the main drains in the middle of the road and there was about two inches of water on the road. He checked the river and he considered it to be well within safe parameters. While it was quite high, it was two or three feet below wall level. At about 8.30 p.m., flooding was reported at the upper end of Lower Dargle Road. A decision was made to monitor the situation and to start evacuating people, if necessary. At around 9.50 p.m., the monitoring team reported a rapid rise in the water level. A decision was made to open up pre-designated centres for receiving evacuees. Between 10.00 p.m. and midnight, the water had risen to a such a level that some of the Civil Defence groups were cut off from direct contact with the Civil Defence headquarters. The Lower Dargle Road had become impassable, even for four-wheel drive vehicles which would normally drive through about two feet of water. It was decided that vehicles were useless and boats from the local sailing club were organised just after midnight. Evacuation by boat continued until 2.25 a.m. on the 26th when a decision was made to withdraw the boats because one of them capsized.

60. Bray Bridge was crucial to the Civil Defence operations that night and Mr. Downes regularly checked the bridge visually to see whether there was a fall or rise in the water. He said he had a good view looking over the parapet of the bridge and he looked over at both sides - to the east and to the west. His recollection was that there was water flowing through the southern arch and the central arch. A tree was obstructing the northern arch. The flow through the arches, with the exception of the northern arch, which was a bit restricted, seemed to be consistent with the arches not being obstructed. He observed no appreciable difference in the levels upstream and downstream. The water was coming at great force. The bridge was vibrating as a result of the water coming down. At one stage, the water appeared to him to be within two feet of the top of the arches and it could have been higher.

61. The Civil Defence log which was maintained on the night of the storm was produced. I am satisfied that the log contains an accurate record of communications from personnel working outside to headquarters. Of particular significance are the entries in relation to the fire engine which was observed by Mr. Stirling and Mr. Eaves. Three entries were made in relation to the fire engine. The first recorded that at 11.25 p.m. the officer in charge of the fire engine reported that the vehicle had become immobilised in six feet of water at Lower Dargle Road. The log recorded that a Landrover had been despatched. The second entry was made at 11.35 p.m. when it was recorded that the fire engine and crew were still stranded. The final entry recorded that at 11.40 p.m. the fire officers were told to abandon the vehicle. The log recorded that the vehicle was abandoned.

62. Another abandoned vehicle which Mr. Downes recalled was a Volkswagen Beetle which had been parked in Castle Street opposite the Plaintiff's store. It was totally submerged in water but from about 2.00 a.m. onwards it began to become visible again and this was an indicator to the Civil Defence personnel that the water was receding.

63. Of the eye witnesses of the flooding in Bray on the night of the storm, it was acknowledged by the expert witnesses that the most reliable observer was Mr. Prendiville. Not only did he have a professional interest as the resident engineer responsible for the drainage construction works, but on the night he was called out by and accompanied by the late Mr. Kevin Conway, the Town Engineer. Moreover, fortunately, Mr. Prendiville did not have to rely solely on his memory when testifying as to the event which occurred more than ten years previously; he had the benefit of his contemporaneous diary record and report on the event. Mr. Prendiville was in Bray from 10.45 p.m. on the 25th to 2.30 a.m. on the 26th and during that period he was observing the behaviour of the river and its impact on the surrounding lands and buildings and he was noting flood levels. The following is a chronological summary of Mr. Prendiville's observations:-

64. Commenting on the event, Mr. Prendiville described it as a natural disaster of enormous proportions. His opinion was that, having regard to the immense amount of water which ended up in the Little Bray area, the flooding of the Plaintiff's premises was inevitable and, if a storm of the same magnitude were to be visited on the Dargle catchment again, the same thing would happen

65. The task which the Plaintiff assumed of establishing on the balance of probabilities that the flooding of the Plaintiff's store was caused or contributed to by factors other than the natural effects of the storm, namely, the escape of water from the reservoir at Paddock Pond and the drainage construction works in the River Dargle near Bray Bridge, was a formidable one and the Plaintiff adopted an ingenious, if unusual, method of addressing the issue. In the Spring of 1994, the Plaintiff commissioned H.R. Wallingford Limited (formerly the Institute of Hydrology, a component of the Natural Environment Research Council in the United Kingdom) to do a computer model of the flooding of the Dargle catchment capable of simulating the effect on the Plaintiff's store of -

66. This work was carried out by Dr. Paul Samuels, a Chartered Civil Engineer and Chartered Mathematician, with H.R. Wallingford Limited who specialises in fluvial systems and computational hydraulics, with the assistance of Mr. James Dent, the Principal Hydrologist with H.R. Wallingford Limited, in relation to the meteorological input. Shortly after the storm, towards the end of September 1986, Mr. McLaughlin had embarked on a study of the effects of the storm on the Dargle catchment and some of the results of his surveys and observations were utilised by Dr. Samuels.

67. Before dealing with the evidence adduced by the Plaintiff on the causation issue, namely, the evidence of Mr. McLaughlin, Mr. Dent and Dr. Samuels, a few general observations are called for.

68. First, it is clear from the evidence that the Plaintiff's instructions to Dr. Samuels were that on the night of the storm there were two breaches in the river wall in Little Bray, one being the breach opposite the shopping complex in which the Plaintiff's store is located, and that that breach was represented as being twenty-eight metres wide. Dr. Samuels was also instructed that the southern arch of Bray Bridge was effectively blocked by temporary works and machinery. At the hearing, it emerged that those instructions were incorrect. In fact, there was only a breach in the river wall - the breach opposite the shopping complex - and that breach, as created by Uniform, was between five and fifteen metres wide. The only temporary construction within or underneath the southern arch itself was the temporary road. As I have indicated above, the evidence did establish the existence of the trench sheeting and the works at the manhole and the Komatsu excavator on the southern bank upstream of the southern arch on the night of the storm. The case made by the Plaintiff at the hearing, in the light of these revelations, was that the access of the flow toward the southern arch was obstructed.

69. Secondly, the only "hard copy" generated by the computer which was put in evidence was an isohyetal map, a connectivity diagram of the flow of the river from Powerscourt Waterfall to the sea and a graph depicting the observed and simulated flood levels at Little Bray. On my estimation, the duration of the oral testimony of Mr. Dent and Dr. Samuels was just short of thirty hours. Not overlooking that it emerged that the computer model had produced about five hundred megabytes of information which was stored on compact disc, I venture to suggest that when computer technology is employed as a probative tool, the information it generates must be capable of presentation which is more conducive to more efficient use of Court time. Insofar as there is a criticism implicit in this observation, it is a criticism of all of the parties because it emerged at the hearing that a suggestion from the Plaintiff's Solicitors to the other parties that experts reports be exchanged was rebuffed.

70. There are no flow gauges on the River Dargle and the experts who studied the effect of the storm on the Dargle catchment were constrained to use diverse recognised methods to quantify the peak discharge. Mr. McLaughlin, in his evidence, suggested an approach and produced results which were at variance with the results produced by the computer modelling. As I understand it, this aspect of Mr. McLaughlin's evidence is not relied on by the Plaintiff. Mr. McLaughlin, in the course of his study, surveyed the arches of Bray Bridge and took flood levels at various points from east of Seapoint Estate to east of the Lithographic premises. There was no controversy as to the accuracy of that survey or the levels taken. Mr. McLaughlin also devised a chronology of the journey of the waters released by the dam break from Paddock Pond to Bray Bridge based on the eye witness accounts and all of the experts were in broad agreement with his conclusion that the dam break occurred around 10.30 p.m. and that the first of the flood waters arrived at Bray Bridge about an hour later. However, the accuracy of the survey of Paddock Pond and of the calculation of its capacity by Mr. McLaughlin was hotly disputed by Coillte.

71. It is not disputed by Coillte that the water released by the failure of the dam at Paddock Pond joined the Dargle upstream of Mr. Hunt's house and flowed with the river down to Bray and into the Irish sea. What is in dispute is the volume of water which was released by the dam failure and the timing and shape of its arrival in Bray. In the years following the catastrophe, either Coillte or its predecessor in title rebuilt the dam, but the evidence did not establish the height of the rebuilt dam relative to the height of the dam which existed before the breach. In the course of rebuilding the dam, the culvert was destroyed and an alternative means of drainage was provided. Paddock Pond has never refilled. Four different estimations of the quantity of water in Paddock Pond before the breach were advanced, namely:-

72. On the basis of the evidence, I have no difficulty in concluding that, in terms of survey methodology employed, the survey carried out by Land Surveys (Lynwood) Limited was the most accurate of the foregoing surveys but, unfortunately, it is of no relevance because it represents Paddock Pond as it was in March 1994 and not Paddock Pond as it was at 10.00 p.m. on 25th August, 1986. Moreover, both that survey and Mr. ÓhAonghusa's survey were to a local datum and it is not possible to compare them to each other or to Mr. McLaughlin's survey and, in particular, it is not possible to compare the crest levels adopted by the respective surveyors. Mr. McLaughlin's surveying methodology has been criticised, with justification, in my view. The upshot is that there is no satisfactory evidence of the volume of water impounded in Paddock Pond prior to the failure of the dam.

73. In fulfilling the Plaintiff's commission, H.R. Wallingford Limited used two established software packages. The first was the Micro F.S.R. model, which embodies the calculations in the Flood Studies Report and later supplementary reports, in particular, F.S.R. 16. The Flood Studies Report was originally published in 1975. It was the result of investigations of methods of flood estimation for engineering design purposes carried out under the aegis of the Institute of Hydrology by that organisation and other organisations in the United Kingdom with the co-operation of other organisations among them, the Office of Public Works in this State. In simple terms, as I understand it, the Micro F.S.R. model converts recorded rainfall over a catchment into river discharge, which is the volumetric flow rate measured in cubic metres per second (cumecs). The second was the Isis River model, which simulates the flow of flood waters through a river system.

74. The Micro F.S.R. model was used to ascertain the run-off response of the catchment which, as I understand it, in simple terms means the percentage of the falling rain that ends up in the river. The data inputted into the model was obtained from the Meteorological Service and included hourly rainfall amounts and durations at Dublin Airport and Casement Aerodrome, which were considered important in assessing the distribution through time of the rainfall during the storm. In ascertaining the percentage run-off, the catchment was divided into seventeen sub-catchments. The run-off response, as indicated through the model, lay between a range of 36% to above 70% for upper sub-catchments. The average percentage run-off over the whole catchment was 62% and that translated into an average rainfall depth of 141.6 millimetres over the whole catchment. In terms of discharge, the result produced was a peak river flow at Bray of 220 cumecs, which result was inputted into the main model, the flow routing model.

75. The dam break flood was modelled separately and the flow rate produced at the exit of the gorge was also inputted into the main model. Dr. Samuels acknowledged that in quantifying the volume of water impounded in Paddock Pond before the flood, the greatest area of uncertainty related to the level of the crest of the dam of which there was no reliable evidence. Taking the crest level from Mr. McLaughlin's survey at 267.56m O.D., the impounded volume was 78,840 cubic metres. It emerged through cross-examination that this volume represented not only water up to crest level but it also included the depth of water flowing over the crest to which depth 11,600 cubic metres were attributable. The result produced by the model was that the flow through the breach in the dam was 244 cumecs and after attenuation the discharge at the exit of the gorge was 204 cumecs. In the model, the dam break occurred at 10.35 p.m. and Paddock Pond drained in fifteen to twenty minutes.

76. In the model, the culvert was represented as a structure measuring two metres by three metres and it was represented in the hydraulic code as a sluice. Two positions were simulated: one with the culvert totally unobstructed and the other with an opening of only 100 millimetres. The first position produced a flow of three cumecs and the second position - the obstructed position - which Dr. Samuels understood to reflect the reality on the night of the storm - produced a flow of 0.4 cumecs.

77. The main model represented the River Dargle from Powerscourt Waterfall down to the Irish sea at Bray. The representation of the geography of the river channel in the model was based on survey data produced by Apex Surveys Limited which produced in excess of forty cross-sections of the river channel between Powerscourt Waterfall and Bray, measurements taken by T.J. O'Connor & Associates, Consulting Engineers, Mr. McLaughlin's survey of Bray Bridge and, as regards the river section downstream of Bray Bridge, the 1 in 1,000 ordnance survey map. All of the relevant topographical features were taken into account, for instance, flows from tributaries, the collapse of Valclusa Bridge and Ballinagee Bridge, transient flooding of the flood plains adjacent to the river and suchlike.

78. I did not understand any of the Defendants to disapprove of the manner in which the geography and topography of the river was represented in the main model, but they severely criticised the calibration of the main model, which, in broad terms, means the assignment of values to the model parameters to match the real situation being represented. In particular, there was criticism of the manner in which values were assigned to the roughness coefficient (in hydrological terminology, Manning's n), and the weir coefficients. The calibration data at Dr. Samuels' disposal was Mr. McLaughlin's chronology of the journey out of the Paddock Pond waters to Bray, Mr. McLaughlin's observed maximum water levels from the Lithographic premises eastward, the eye witness observations of the flooding, in the case of the flood plain, principally Mr. Garland's account. In the calibration of the model which produced the results which I will record later, Dr. Samuels assigned a value of 0.03 to Manning's n above the Lithographic premises. He acknowledged that his visual observation of the river upstream of the Lithographic premises would have suggested an appropriate roughness coefficient of 0.04, but, having regard to the evidence of the actual transit time of the Paddock Pond waters, he concluded that the model was diffusing or attenuating the Paddock Pond waters more than had actually happened and he reduced the river roughness to match up with the chronology. The value assigned to Manning's n between the Lithographic premises and Bray Bridge was 0.044 and downstream of Bray Bridge a value of 0.040 was assigned.

79. In relation to the discharge coefficients for the weir flow over the river walls, Dr. Samuels testified that experience over many years had indicated that where there is flow over walls running parallel to a river, lower discharge coefficients are necessary and the values assigned in the model were within the range which he had used on other occasions.

80. It is convenient to set out in tabular form the results produced by the model of the maximum water level in the Plaintiff's store and the maximum flow at Bray Bridge for the various scenarios which I have outlined above which H.R. Wallingford Limited was commissioned to investigate and simulate. The results summarised in the following table are predicated on the assumption that the breach in the river wall created by Uniform was twenty-eight metres wide.

81. When it emerged from Mr. Prendiville's evidence that the width of the breach was somewhere between five metres and fifteen metres, the model was re-run and produced the following maximum water levels in the Plaintiff's store for the 1986 scenario: in the case of a five metres breach, 4.05m O.D.; in the case of a ten metre breach, 4.09m O.D.; and in the case of a fifteen metre breach, 4.13m O.D.

82. The significance of the maximum water levels for the various scenarios produced by the model was tested against the critical threshold at the Plaintiff's premises, that is to say, the lowest level at which water could get in to the Plaintiff's store from the St. Laurence's Terrace area at the onset of the flooding, which, as I indicated at the outset, was 3.86m O.D. On the basis of the tabulated figures -

83. However, the result produced for scenario (iv) was considerably lower than the maximum water level observed by the eye witnesses within the Plaintiff's premises, which, in accordance with Mr. Garland's evidence, was in the region of 4.46m O.D. and, according to Mr. Flanagan's evidence, was in the region of 4.55m O.D. As I understand it, the highest water level within the Plaintiff's store which the model simulated was 4.37m O.D. which was produced as a result of additional sensitivity analyses carried out by Dr. Samuels during the course of his cross-examination, which he was prompted to do because the water levels produced when the width of the breach was reduced pushed the model results further away from the eye witness observations. The figure of 4.37m O.D. was based on a ten metre wide breach and a recalibration which assigned a value of 0.040 to Manning's n between the Lithographic premises and Bray Bridge and assumed that storage within the built up area in Little Bray was only two thirds of the maximum storage available, and corrected an incorrect linkage of the flow path to Castle Street which was in the model originally. The level of 4.37m O.D. represented scenario (iv) and, similarly calibrated, the model produced a level of 3.80m O.D. for scenario (i), which Dr. Samuels stressed was still below the critical threshold of 3.86m O.D.

84. During cross-examination, Dr. Samuels was asked to apportion the 0.3 metre rise in the water level which, on the basis of the tabulated figures, the construction works alone accounted for between the effect of the breach, on the one hand, and the effect of the obstruction of flow to the southern arch, on the other hand. He attributed 0.2 metre to the breach and 0.1 metre to the obstruction of flow to the southern arch. However, he acknowledged that the figures have a range of uncertainty associated with them, which he put at plus or minus 0.1 metre.

85. In relation to the peak flow volumes which the model simulated, Dr. Samuels' opinion was that attributing 20 cumecs to the Paddock Pond contribution by comparing the result of scenario (i) with scenario (ii) was too low and that it represented too much attenuation within the model. The Paddock Pond waters contribution of 13.5 cumecs at Bray Bridge, indicated by a comparison of scenario (i) and scenario (iv), represented the residual flow at Bray Bridge, but the model produced a value of 30 cumecs for the Paddock Pond waters contribution upstream at River Lane. Dr. Samuels acknowledged that there was a margin of uncertainty in the range of plus or minus 15% in relation to the assessment of the peak flows.

86. Bray Bridge was represented in the model as having two arches only open and the third totally obstructed. Although Dr. Samuels acknowledged that this was an incorrect representation of the physical reality, nonetheless, he contended that such representation was a proper representation within the model for calibration purposes of the measured afflux at Bray Bridge, which he calculated from Mr. McLaughlin's levels. In broad outline, Mr. McLaughlin had observed a flood level of 3.70m O.D. fifty metres east of Bray Bridge and a level of 4.30m O.D. forty-five metres west of Bray Bridge, the difference between the two levels being 0.6 metre. Dr. Samuels adjusted this difference to 0.46 metres to account for the slope of the river. On the basis of a maximum flow rate of 233.5 cumecs, using Yarnell's equations, he found that the difference in levels which was 0.46 metre, was consistent with one third of the combined arch width of Bray Bridge, as surveyed by Mr. McLaughlin, having been effectively lost. Dr. Samuels attributed 0.15 metre of the measured afflux of 0.46 metre to normal afflux. The balance, 0.3 metre, he attributed as to 0.1 metre to the temporary road under the southern arch and as to the remaining 0.2 metre to the other factors which obstructed flow to the southern arch and, in particular, to the trench sheeting and the Komatsu excavator.

87. On the basis of the model simulations, Dr. Samuels expressed the following opinions:-

88. Dr. Samuels' very considerable experience in interpreting eye witness accounts of flood events was applied in calibrating the model. Moreover, that experience enabled Dr. Samuels to suggest the proper inferences to be drawn from what was observed by the eye witnesses. Dr. Samuels' interpretation of the evidence of Mr. Crinnion, Mr. Stirling and Mr. Eaves as to the rapid rise in the water level which they observed was that there was a sudden surge, which was entirely consistent with the movement of an organised body of water down the river. The only source of that body of water was Paddock Pond. Dr. Samuels rejected the proposition that the established intensification of the rainfall over the catchment in the previous six hours would account for the rapid rise in water levels at Haughton's Junction around 11.30 p.m.

89. In the course of his evidence, Dr. Samuels gave a broad picture of the course of the flood as represented in the model and he elaborated on certain aspects of it in the course of cross-examination. The following are some of the significant points which emerged:-

90. Two expert witnesses challenged the reliability of any conclusions drawn from the simulations of the flood carried out by H.R. Wallingford Limited. Professor Conleth Cunnane, the Professor of Hydrology at University College Galway, was called on behalf of the Council and his evidence was adopted by Uniform. Dr. Michael Bruen, a college lecturer in the Department of Civil Engineering at University College Dublin, who specialises in hydrology, systems analysis and computer methods, was called on behalf of Coillte.

91. Both Professor Cunnane and Dr. Bruen questioned the suitability of the software for the task for which it was employed by H.R. Wallingford Limited and they were very critical of its application to determine the cause or causes of the flooding in the Plaintiff's premises. The paucity of calibration data and the lack of independent verification of the model's performance, in their view, reduces one's confidence in the model.

92. In relation to the Micro F.S.R. model, in broad terms, the main criticism was that, while it is suitable for design purposes, it is not suitable for analysis of a single past event. In consequence, its application to the storm and flooding on 25th August, 1986 produced an under-estimation of the natural flood discharge. The Flood Studies Report recommends that when reproducing recorded events, the percentage run-off should vary with the catchment wetness index within the event, as happens in nature, but the Micro F.S.R. programme does not provide for this. The Flood Studies Report methodology, which embodies a constant percentage run-off throughout the storm, was tested in the Flood Studies Report itself and it was disclosed that it under-estimates peak flows for severe events and over-estimates for minor events, which is a consequence of the averaging inherent in the methodology. Dr. Bruen testified that, even when properly used in the design situation, the Micro F.S.R. on average under-estimates the peak where the percentage run-off is 60% or more by about 15%, reduced by 1.8% having regard to the modifications introduced in F.S.R. 16 and embodied in the programme. In the design situation, regard would be had to these limitations of the Micro F.S.R. and a margin of safety would be built in.

93. Professor Cunnane considered the average percentage run-off of 62% which Mr. Dent calculated and the value of 220 cumecs which Dr. Samuels ascribed to the peak discharge in the light of those limitations. His view was that having regard to Mr. Fitzgerald's evidence of the weather before the storm, the retentive capacity of the soil would have been minimal. The long persistent rain over twelve hours on a catchment the size of the Dargle catchment would have resulted in the soil being saturated early on and remaining so throughout the storm. Mr. Dent's calculation of 62% was a design value, not an actual value, and it was more likely exceeded. Professor Cunnane's opinion was that the average percentage run-off would have been at least 70% and he would not rule out 80%. Professor Cunnane estimated that, if the average percentage run-off was 70% and if the percentage run-off was allowed to vary throughout the storm in accordance with the catchment wetness index, a peak discharge of 303 cumecs would have been produced.

94. The criticisms of the Isis model focused on its application and, in particular, the manner in which it was calibrated and the lack of verification. The calibration runs were not able to produce the maximum water level observed in the Plaintiff's premises by Mr. Garland and Mr. Flanagan. Therefore, Professor Cunnane suggested one would have to be dubious about the applicability of the model to the hypothetical situations - scenarios (i), (ii) and (iii), because the outcomes for the hypothetical situations are limited by the fact that the model in its calibration phase was unable to produce what actually occurred. As to the Isis model's failure to produce the maximum water level observed in the Plaintiff's premises, Professor Cunnane attributed this to one or other or a combination of two factors: first, that insufficient water was allowed into the model through the river channel from the catchment; and secondly, that the mechanisms in the model for allowing water into the flood plain contributed to the lack of water in the flood plain. In terms of volume, Professor Cunnane estimated that the calibration run which produced a level of 4.22m O.D. in the Plaintiff's premises was lacking about 20,000 cubic metres of water.

95. Professor Cunnane contended that inappropriate values had been assigned to Manning's n in the model and there was no objective justification for the values assigned. A value of 0.044 was too high for the stretch from the Lithographic premises to Bray Bridge and even a value of 0.040 was too high for the channel near the Peoples Park. In Professor Cunnane's opinion, the values ascribed by Barry to Manning's n, in its investigations after the storm, were more in keeping with reality. Professor Cunnane illustrated the effect of adjusting the value of Manning's n by the following example. If a value of 0.044 produces a discharge value of 220 cumecs, assuming the water level remains the same, adjusting the roughness coefficient to 0.033 increases the discharge value to 293 cumecs, by approximately one third.

96. Professor Cunnane's evidence suggested that a certain degree of scepticism should attend making any deductions from a comparison of the model calculation of the maximum level of the natural flood and the critical threshold level at the Plaintiff's premises, given that the maximum water level produced by the model on a calibration run (4.37m O.D.) was less than the maximum observed water level, from which it might be concluded that the maximum water levels produced for scenarios (i), (ii) and (iii) were under-calculated. The need for scepticism was underscored by the following illustrations:-

97. Professor Cunnane disputed Dr. Samuels' calculation of the afflux at Bray Bridge at 0.46 metre. He produced results of 0.3 metre and 0.32 metre on calculating the afflux by two different methods. He explained that every bridge provides some form of constriction which causes some measure of afflux upstream. In the case of Bray Bridge, which has a total water way width of 27.8 metres for a river which upstream of the bridge widens to a width of 39 metres or 40 metres, a fairly severe constriction is presented by the bridge. On his calculation, with a flow of 250/260 cumecs, an afflux of 0.32 metre would be entirely attributable to natural processes - the constriction created by the piers and the abutments of the bridge alone.

98. Professor Cunnane assessed the factors which the Plaintiff alleges contributed to the afflux individually and in combination and expressed the following opinions:-

99. Professor Cunnane analysed the significance of the breach in the river wall in the light of the eye witness accounts. Because of the existence of the earthen embankment, water coming through the breach would have flowed in an easterly direction first, filling the space between the embankment and the steps at Bray Bridge. It would have only gone in a westerly direction when the water level in the river reached a height a little above the top of the embankment. Mr. Prendiville's evidence was that the embankment was scaled by 10.45 p.m. However, on the basis of Mr. Prendiville's evidence that the area of the Peoples Park between the temporary gates at Haughton's Junction and the river wall opposite the gates was not flooded at that time, the inference to be drawn is that the water had not reached this part of Peoples Park, insofar as it was travelling west at this time. The record of the fate of the fire engine and its occupants establishes that between 11.30 p.m. and 11.45 p.m. the water had reached a level of 4.75m O.D. at Haughton's Junction. The highest flood level which Mr. McLaughlin recorded opposite the breach was 4.4m O.D.. Professor Cunnane inferred that by this time water must have been flowing from Haughton's Junction towards the breach and back into the river. His deduction as to the effect of the breach from the facts was that water could only have been flowing from the breach westerly in the direction of Haughton's Junction from approximately 10.45 p.m. to 11.30 p.m./11.45 p.m. and he concluded that the breach played a very limited, if any, role in the flooding of the Haughton's Junction area and it would have had a minimal effect on the Plaintiff's premises. If there had been no breach, in Professor Cunnane's opinion, the Plaintiff's premises would have been flooded. Indeed, he suggested that the breach had an ameliorating effect in that from midnight onwards, when the water was still rising, the flow was going back into the river channel through the breach.

100. Professor Cunnane expressed the view that the Plaintiff's premises would have been flooded by the natural flood even if it had not been augmented by the waters from Paddock Pond. In his view, the volume of water required to inundate the Plaintiff's premises was available in the natural flood and he contrasted the volume of water impounded behind the dam in Paddock Pond, whether it was 45,000 cubic metres or 80,000 cubic metres, with the quantum of rain which fell on the catchment during the storm, which he estimated at 16,000,000 cubic metres.

101. Although Dr. Bruen did not comment in any great detail on the Isis model, he did challenge the reliability of the results it produced having regard to the limited data for calibration and the lack of verification. While he acknowledged that the Isis model has a very good reputation, a particular weakness in its application which he adverted to was that it was infected, as it were, by the deficiencies in the Micro F.S.R. model which resulted in the modelling exercise not producing a reliable estimate of the peak discharge during the storm. That the modelling exercise did not replicate the timing of the peak of the flood as observed by the eye witnesses impressed Dr. Bruen and confirmed his view of the unreliability of the results of the modelling exercise. Acknowledging that he was not in a position to say that the Paddock Pond waters had no effect on the flooding of the Plaintiff's premises, he expressed the view that it had no significant effect, the determining factor in his conclusion being the time lapse between the arrival of the waters from Paddock Pond into Bray, which was reasonably well fixed by the evidence, and the peak of the flood as observed by the eye witnesses and, in particular, Mr. Prendiville.

102. As I have indicated above, one of the factors which Professor Cunnane perceived as being responsible for the failure of the model to produce the maximum water level observed in the Plaintiff's premises was the manner in which the linkages between the flow in the river channel and the flow on the flood plain were treated in the model. Professor Cunnane's criticism of the treatment of the linkages was not merely that it resulted in the model producing too little water on the flood plain, but that it undermined the integrity of the modelling exercise. Weir coefficients were used to allow water to transfer from the river channel on to the flood plain. Flow over weirs has been understood for one hundred years and the weir coefficients are based on a fortiori hydraulic knowledge. Yet, in the modelling exercise, they were allowed to be adjusted in the calibration process. This was not necessary and should not have happened. Professor Cunnane also disapproved of the modelling procedure whereby the weir coefficients were allowed to interact with the value assigned to Manning's n for the channel. He could not understand why something that is physical in the channel, that is effectively the measure of the resistance of the channel to flow, should be allowed to vary while one is also varying the weir coefficients. In his view, this kind of "linking" is not satisfactory in a case where one is trying to prove an effect. The gravamen of Professor Cunnane's criticism was that instead of using hydrodynamic equations to calculate the flow rate from the river on to the flood plain, a subjective approach was adopted, which is not a suitable basis for demonstrating or proving that something has happened.

103. In summarising the eye witness accounts above, I have recounted Mr. Doyle's observations of the impact of the flood on the river channel from Dargle Bridge downstream two or three days after the storm. These observations were the start of an extensive survey of the river channel and the flood defences carried out by Barry following the storm. The survey lasted three months.

104. Following the flooding of Little Bray in 1965, Barry had carried out an extensive survey of the Dargle with a view to designing the river channel to take a flow of 200 cumecs. The design was implemented in 1969 when the channel itself was re-graded by dredging, a weir in the vicinity of Country Brook was removed and the river wall was underpinned as necessary. Following the post-1986 survey, Barry calculated the peak flow in the channel between Dargle Bridge and the Slang on the night of the 1986 storm at 285 cumecs using -

105. With the aid of the Office of Public Works computer, Barry also did flow profiling of the flood, inputting the same material into the computer. Various discharges were investigated, to as high as 320 cumecs and consistency with the flood marks was found. Barry's conclusion was that a flood in the order of magnitude of 300 cumecs had occurred in a channel which had capacity to carry in the region of 200 cumecs.

106. Using the Flood Studies Report methodology, Barry concluded that a flood with a return period in excess of one in one hundred years occurred in the Dargle catchment on the 25th/26th August, 1986.

107. Mr. Doyle's interpretation of what occurred on the night of the 25th was that the river stayed within its banks between Dargle Bridge and the Slang; that a flow of in excess of 285 cumecs came down fully contained within that stretch. However, from that point down, there was not the capacity to retain it totally within the channel and it spread on the right bank and on the left bank, coming out on the right bank at Mr. Collins' house and on the left bank at the Lithographic premises. It flowed through the Coburg Estate. The flow of in excess of 285 cumecs was trying to come down the river but was not finding adequate capacity in the main channel. So it over-topped the river wall and flowed through the Coburg Estate and down into Little Bray.

108. In the course of Mr. Doyle's evidence, two maps were produced which gave a striking picture of Little Bray on the night of the storm. On one, Mr. Doyle had denoted with shading the extent of the area flooded. It came as no surprise when Mr. Doyle testified that 520 properties in all were flooded in the Dargle catchment. On the other map, Mr. Doyle had indicated the maximum depth of water above street level at various locations, which showed levels as high as 1.5 metres at Haughton's Junction and at the northern end of Fairgreen Road and as high as 1.6 metres at the northern end of Adelaide Villas. On this map, Mr. Doyle had also indicated the peak flood level relative to the height of the river wall along the Peoples Park, which demonstrated that, apart from a relatively short stretch of about ninety metres roughly opposite the bandstand, the entire river wall must have been over-topped.

109. Barry also analysed the contribution of the Paddock Pond waters to the flooding in Bray. Their assessment of the Paddock Pond waters contribution to the peak flow in Bray was that it fell within the range of 28.5 to 42.5 cumecs. Their conclusion was that, while the Paddock Pond waters were a definite factor in the flooding at Bray, in the overall context of the duration of the flood and the levels achieved and when they were achieved, they were not a significant factor. In short, had there been no dam break, the Plaintiff's premises would have been flooded in any event by the natural flood. The duration of the flood and the fact that the waters kept rising until some time after 2.00 a.m. on the 26th supported that conclusion, it was contended.

110. In relation to the evidence of Mr. Crinnion, Mr. Stirling and Mr. Eaves of a sudden rise in the water levels, Mr. Doyle's interpretation was that it was the result of the over-topping of a significant extent of the river wall along the Peoples Park, which in turn was the result of the water in the channel rising at a steady rate to a point where it suddenly started discharging over the river wall.

111. In relation to the assessment of the peak flow on the outskirts of the urban area of Bray, there was a large measure of conformity between Mr. Doyle's evidence and the evidence of Mr. Kennedy, whose firm had been retained by Coillte in connection with this litigation to investigate the contribution, if any, of the collapse of the dam to the flooding of the Plaintiff's premises. Mr. Kennedy's calculation of the peak flow at a point just downstream of the confluence of the last tributary to join the Dargle, which flows from the Scalp, was 310 cumecs or 341 cumecs depending on whether he assigned a value of 0.040 or 0.035 to Manning's n. His opinion was that 310 cumecs represented the proper discharge. He also calculated the contribution of the waters from Paddock Pond within that figure on the basis of the rainfall recorded at Glenasmole, assuming a percentage run-off of 70%, which put the contribution at 29 cumecs. Mr. Kennedy also calculated the discharge between the dam and the location at which the watercourse from Paddock Pond joins the Dargle and came up with figures of 187 cumecs and 210 cumecs depending on whether he assigned a value of 0.045 or 0.040 to Manning's n. Apart from a rainfall contribution, which Mr. Kennedy calculated at nine cumecs, these figures represented only the waters from Paddock Pond. Mr. Kennedy's conclusion was that there was considerable attenuation of the Paddock Pond flood when it reached Bray.

112. Mr. Kennedy's interpretation of the flooding process coincided with that of Mr. Doyle. In his opinion, the peak flow coming down the river was in excess of 300 cumecs and, as the river channel had a carrying capacity of only 200 cumecs, flooding was inevitable. The peak flow was 50% in excess of the channel capacity and it was inevitable that it would over-top the river wall. Mr. Kennedy acknowledged that water from Paddock Pond did come down into Bray and did contribute to the flooding.

113. However, Mr. Kennedy was not persuaded by the eye witness accounts that there was a distinguishable surge attributable to the Paddock Pond waters. His view was that if there had been a surge in the river it would have been noted by Mr. Prendiville. He explained that the Paddock Pond flood did not "piggy-back" on the natural flood; it was subsumed into the natural flood so as not to be noticeable. Mr. Kennedy stressed that rainfall can give rise to a sudden rise in water level - to a surge. He testified that, if the Paddock Pond waters were solely responsible for the over-topping of the river wall at the Peoples Park one would have expected the water level of the river to rise, then the wall to be over-topped and, when the Paddock Pond flood had gone through Bray, for the level to settle back to a lesser level. However, the evidence of the eye witnesses was that the water continued to rise and rise long after the Paddock Pond flood had arrived and he concluded that the Paddock Pond contribution was obscured or totally overtaken by the natural flood. He calculated that it took the Paddock Pond flood about an hour to pass through Bray but he acknowledged that that calculation assumed a linear rate of passage rather than an exponential rate which would have occurred in reality and he conceded that the passage would have taken longer. Nonetheless, he stood by his thesis because, on the evidence, the peak flood occurred at 2.10 a.m.

114. Mr. Kennedy took issue with some of the propositions advanced by Dr. Samuels as to what occurred at Paddock Pond. First, he rejected Dr. Samuels' proposition that an earthen dam could have sustained a flow above crest level for long enough to create a head of 300 millimetres over the width of the dam, which is what Dr. Samuels' volume of 11,600 cubic metres represented. While Mr. Kennedy did not quibble with Dr. Samuels' calculation of the head, his view was that, given that the dam was an earthen dam, what Dr. Samuels had envisaged happening could not happen physically. Secondly, in Mr. Kennedy's opinion, even if the culvert had been operating effectively, the dam would still have failed. On the basis of the eighty-five millimetres of rain recorded at Glenasmole over six hours, he calculated that the inflow into Paddock Pond was four cumecs, whereas the capacity of the culvert was one cumec only, leaving three cumecs of inflow to pass over the dam, with inevitable failure.

115. Having outlined the evidence, I now propose considering the issues which arise between the parties, namely:-

116. It was submitted by Mr McCullough, on behalf of the Council, that, as the sanitary authority for the Bray area, it was under a statutory duty, by virtue of Section 17 of the Public Health (Ireland) Act, 1878, to cause to be made such sewers as might be necessary for effectually draining the Bray area and that on August 25th, 1986 the drainage construction works being carried out in the Dargle at Bray were being executed in pursuance of that duty and in the only manner in which they could reasonably be executed. That being the case, it was submitted, as a matter of law, the Council is immune from an action based on nuisance and is free from liability unless it was negligent in the exercise of its statutory duty and power. I am satisfied that this proposition is correct and that, in the absence of proof of negligence, no liability attaches to the Council or to Uniform, which executed the works under contract with the Council.

117. The Council and Uniform are charged with the same acts and omissions in the implementation of the drainage construction works as constituting negligence, the claim against the Council being premised on the fact that it planned and directed the works, and the claim against Uniform being premised on the fact that it executed the works under contract with the Council. The only matters which emerged from the evidence on which the Plaintiff could found an allegation of negligence against the Council and Uniform were:

118. It was urged by Mr McCullough on behalf of the Council that, in determining whether the Council or Uniform was negligent, the Court should judge the conduct of these Defendants in the context in which they were implementing the works and, in particular, the Court should have regard to the absolute necessity to implement the sewage scheme in Bray and to the fact that there was no evidence that there was any better or safer way of executing the works. Moreover, it was urged by Mr McCullough that in determining whether these Defendants are liable for the consequences of their conduct, the essential issue is whether those consequences were reasonably foreseeable by them prior to the storm, the contention being that they were not.

119. Mr Comyn, on behalf of the Plaintiff, submitted that the correct test on the issue of foreseeability is the test to be extrapolated from the following passage of the Judgment of Walsh J. in Kelly v. The Board of Governors of St. Lawrence's Hospital [1988] IR 402 (a jury action involving an allegation of negligence against a hospital in connection with the custodial care of a patient suffering from epilepsy and some automatism) at page 410:-

120. The criterion is whether the type of damage done, rather than its extent, is foreseeable, it was submitted. Mr. Comyn pointed to the flashy nature of the river, the previous experience of flooding in the Little Bray area and the evidence as to the vulnerable nature of the river flood defences at the location of the breach. He acknowledged however, that as regards the standard of care expected of a public authority carrying out public works, the Courts allow a margin of tolerance.

121. In my view, neither the Council nor Uniform fell short of the standard of care which they owed to riparians in Little Bray. As I am entitled to, I have had regard to the social utility of the conduct of these Defendants and the object of the works - to improve drainage in the Bray area. While a storm of sufficient ferocity to raise the level of the river sufficiently to flood riparian lands was, in my view, a foreseeable possibility, it was a very remote possibility, particularly, as within the previous twenty years the river channel had been upgraded following the 1965 flood to accommodate a 200 cumec flood, at the behest of the Council and following an intensive study by the consulting engineers involved in the sewage project. I have no doubt that the standard of care observed by the Council and Uniform was commensurate with the degree of possibility of flooding and that it conformed with the standard to be expected from a reasonable sanitary authority and a reasonable building contractor in the circumstances which prevailed. My observations in relation to the specific allegations of negligence against these Defendants are as follows:

122. Accordingly, I hold that the Council and Uniform were not negligent.

123. I have outlined in some detail the evidence as to the condition of Paddock Pond immediately prior to the storm. There was uncontroverted evidence that the surface of the roadway which traversed the dam was in poor condition immediately prior to the storm. I am satisfied, on the evidence, that the deterioration of the roadway was attributable to intensification of its use in connection with the forestry operations in Priest's Wood and the failure of Coillte's predecessor to have a proper care and maintenance programme in place in relation to Paddock Pond and its failure to adequately respond to the deterioration and lack of repair of the surface of the dam. I am also satisfied that it was Coillte's predecessor's failure to have a proper care and maintenance programme in place and to act in accordance with it which led to the capacity of the culvert being considerably reduced because of silting up and obstruction by stones, vegetation and forestry debris, as I have already found, on the night of the storm. The evidence discloses a deplorable lack of care and concern on the part of Coillte's predecessor as the owner, occupier and body responsible for Paddock Pond.

124. It is common case that the dam at Paddock Pond collapsed at around 10.30 pm on the night of 25th August and that thereupon the water impounded behind the dam escaped, rushed down the gorge and joined the waters of the Dargle. The crucial question is whether Coillte's predecessor's conduct in permitting the intensification of the use of the road on the dam, failing to adequately maintain the surface of the dam and failing to maintain and keep clear from obstruction the culvert, was a material element and a substantial factor in bringing about the dam failure.

125. In my view, the Plaintiff has not established on the balance of probabilities a causal link between the use and lack of proper maintenance of the dam prior to the storm and to the dam failure. In particular, the evidence of Mr McLaughlin, considered in the light of the theories advanced by its experts as to the manner in which the dam burst, was not sufficiently cogent to lead me to conclude that the traffic on the dam resulted in structural effects on the core which were of significance in the subsequent failure of the dam or that the condition of the surface facilitated erosion such as to be a material element in the failure mechanism. I consider that the proper inference to be drawn from the evidence is that it was the effect of the enormous volume of water flowing over the crest of the dam at the height of the storm which caused the failure. Mr Kennedy's theory as to the failure mechanism, in my view, is consistent with the photographic evidence of the condition of the dam in the immediate aftermath of the storm, from which it appears that the core was undamaged.

126. That inference leads to the question of whether the obstruction of the culvert was a material element and a substantial factor in the increase in the level of the water impounded behind the dam, eventually to a level at which it overtopped the dam. It is impossible, on the evidence, to quantify the diminution in the outflow through the culvert caused by the obstruction or to relate that diminution to the inflow into Paddock Pond with any degree of accuracy. Dr Samuels, on the basis of the assumptions embodied in the model as to the dimensions of the culvert and its effective unobstructed capacity, quantified the reduction in discharge as a reduction from 3 cumecs to 0.4 cumec. However, I am not satisfied that either assumption reflected the reality on the night. Mr Kennedy calculated the discharge capacity of the culvert in its unobstructed state at 1 cumec and the inflow into Paddock Pond at 4 cumecs. If these calculations were based on data which reflected the reality on the night, the obstruction of the culvert could not have been a material element or a substantial factor in the volume of water which was inevitably going to flow over the dam. Of course Mr. Kennedy's estimation may be no more an accurate reflection of what happened on the night than Dr. Samuel's.

127. Having regard to the evidence as a whole, in my view, it was the unprecedented intensity and duration of the storm and, in particular, the extraordinary degree of precipitation in the upland area of the Dargle catchment in the five to six hours before 10.30 p.m. on the 25th which caused the water in Paddock Pond to rise and rise and, even if the culvert had been wholly unobstructed, the culvert could not have played an appreciable role in keeping down the level of the water. It follows that the blockage of the culvert was not a material element or a substantial factor in bringing about the dam failure.

128. Accordingly, the Plaintiff's claim against Coillte in negligence fails.

129. The most frequently quoted statement of the rule in Rylands -v- Fletcher is the following passage from the judgment of Blackburn J. in the Court of Exchequer Chamber [(1866) L.R.1 Ex. 265] :-

130. The judgment of Blackburn J. was subsequently approved by the House of Lords (L.R. 3 H.L. 330), although in his speech Lord Cairns restricted the application of the rule to circumstances where there had been "a non-natural use" of the land by the defendant. The decision in Rylands -v- Fletcher itself is authority for the proposition that the creation of an artificial lake or reservoir is a non-natural use. The nature of the liability of a defendant for the escape of dangerous things has been clarified recently by the House of Lords in Cambridge Water Company Limited -v- Eastern Counties Leather Plc. (1994) 2 AC 264 in which it was held that forseeablility of harm of the relevant type by the defendant is a prerequisite of the recovery of damages both in nuisance and under the rule in Rylands -v- Fletcher . In the light of that decision the nature of the liability of a defendant under the rule in Rylands -v- Fletcher is summarised as follows in Charlesworth and Percy on Negligence, 9th Edition, at page 893:-

131. In my view, Coillte could reasonably have foreseen that, if the dam at Paddock Pond failed and the impounded water escaped, it would flow via the gorge and the watercourse into the Dargle and that damage in the nature of flooding of the riparian properties downstream would ensue. In my view, the circumstances of the escape of water from Paddock Pond on the night of 25th August, 1986 come fairly and squarely within the ambit of the rule in Rylands -v- Fletcher and, unless Coillte has established one of the excusing factors recognised as constituting a defence to liability under that rule, Coillte is liable to the Plaintiff provided the Plaintiff has established that the flooding of its premises was caused or contributed to by the water from Paddock Pond.

132. The defence which Coillte claims absolves it from liability is act of God. Mr. Keane, on behalf of Coillte, relied on Nichols -v- Marsland (1876) 2 ExD 1, which is described in McMahon and Binchy on Irish Law of Torts , 2nd Edition, at page 491 as the only reported decision in which the defence of Act of God has been successful, in support of his contention that the dam failure was caused by act of God. The facts in that case, as set out in the judgment of Mellish, L.J., were that the defendant was the owner of a series of artificial ornamental lakes, which had existed for a great number of years, and had never previous to 18th June, 1872 caused any damage. On that day, however, after a most unusual fall of rain, the lakes overflowed, the dams at their end gave way, and the water out of the lakes carried away the county bridges lower downstream. The jury found that there was no negligence either in the construction or the maintenance of the reservoirs, but that, if the flood could have been anticipated, the effect might have been prevented. Mellish, L.J. distinguished the case before him on the facts from Rylands -v- Fletcher stating:-

133. Having stated that the opinion of the Court of Appeal was that the defendant was entitled to excuse herself by proving that the water escaped through the act of God, Mellish, L.J. went on to say:-

134. Mr. Keane emphasised the similarity on the facts of Nichols -v- Marsland and the instant case and urged that Nichols -v- Marsland should be followed. He submitted that the decision of O'Hanlon J. in Dockeray & Another -v- Manor Park Homebuilders Limited , in which a reserved judgment was delivered on 10th April, 1995, the only recent Irish authority in which the defence of act of God in a similar context to this case was considered, is distinguishable on the facts. Having recognised that act of God or vis major could, in appropriate circumstances, afford a defence to a claim under the rule in Rylands -v- Fletcher , and having quoted a passage from McMahon and Binchy at page 492 to the effect that it would appear that "only the most extreme of natural phenomena" will afford a good defence, O'Hanlon J. went on to say:-

135. Mr. Comyn's response to Mr. Keane's submission was that in Greenock Corporation -v- Caledonian Railway (1917) AC 556 the finding of fact by the jury in Nichols -v- Marsland was thought to be wrong by the House of Lords, which approved of the decision in Kerr -v- The Earl of Orkney (1857) 20 D. 298, on which Mr. Comyn relied. In the latter case Lord Justice-Clerk Hope said:-

136. It is stated in Charlesworth and Percy at page 902 that to be an act of God an occurrence must be such that it -

137. The third requirement was adopted by the House of Lords in Greenock Corporation -v- Caledonian Railway , and, while supported by a line of authority, that decision departed from another line of authority represented by cases such as Nichols -v- Marsland and Nitro-Phosphate and Odam's Manure Company -v- London and St. Catherine's Docks (1878) 9 Ch. D. 503, to which I will refer later in another context - cases in which it was held that to constitute an act of God it was sufficient to show that the occurrence could not reasonably be anticipated or guarded against. The authors of Charlesworth and Percy state that, in order for the defence to succeed now, it must be proved by the defendant that it was impossible to anticipate the occurrence or to guard effectively against it and they point out that the defence was not established where there had been an exceptional storm; a rainfall of extraordinary violence ( Greenock Corporation -v- Caledonian Railway ); an exceptionally heavy snow storm; a very high wind; or an extraordinarily high tide. They go on to suggest that the only circumstances in which the defence would be likely to succeed in the United Kingdom would be if some catastrophe happened as a result of an earthquake, a volcanic eruption or a tidal wave of vast proportions.

138. The views expressed by the authors of Charlesworth and Percy are obviously informed by the supremacy of the decision in Greenock Corporation -v- Caledonian Railway over the decision in Nichols -v- Marsland . Apart from the distinguishing feature which I have already referred to, that the finding of fact in Nichols -v- Marsland was made by a jury, Nichols -v- Marsland was also distinguished in Greenock Corporation -v- Caledonian Railway on the basis that, having reference merely to the storage of water as in Rylands -v- Fletcher, it did not affect the question of liability for interference with the course of a natural stream as laid down in other authorities, such as, Kerr -v- The Earl of Orkney to which I have already referred (see the speech of Lord Finlay, L.C. at page 573 and the judgment of Lord Wrenbury at page 584). Another distinguishing feature was that Greenock Corporation -v- Caledonian Railway was an appeal from the Court of Session in Scotland and the issue was whether the extraordinary rainfall which caused the stream to overflow in that case was a damnum fatale under Scottish Law. In his speech Lord Finlay, L.C. stated at page 571 that the authorities justified the view of the law propounded by Professor Rankine in his work on the Law of Land Ownership in Scotland in the following passage:-

139. I have quoted this passage because it discloses that even under the law of Scotland the strictures of the concept of damnum fatale gave way to the existence of a state of things for a period of long prescription. Immediately before 25th August, 1986 the dam at Paddock Pond was of sufficient antiquity to be "fortified by prescription".

140. In my view, the principle to be derived from the line of authorities which includes Nichols -v- Marsland and the Nitro-Phosphate case is more in line with the current concept of tortuous liability under the rule in Rylands -v- Fletcher , as exemplified by the decision of the House of Lords in Cambridge Water Company -v- Eastern Counties Leather Plc., than the decision in Greenock Corporation -v- Caledonian Railway. Moreover, on the facts, the instant case is closer to Nicholas -v- Marsland than to Greenock Corporation -v- Caledonian Railway . Accordingly, in determining whether Coillte is absolved from liability for the consequences of the dam failure by reliance on the defence of act of God, I consider that the test to be applied is whether the storm on the 25th August, 1986 could reasonably have been anticipated or guarded against by Coillte. I am satisfied that the evidence shows that the storm did fall within the category of the most extreme natural phenomena and could not reasonably have been anticipated or guarded against, so that the defence of act of God succeeds.

141. Accordingly, the Plaintiff's claim against Coillte under the rule in Rylands -v- Fletcher fails.

142. It was submitted on behalf of the Council and Uniform that the acts and omissions which were alleged to constitute a breach of duty on the part of these defendants did not cause the flooding in the Plaintiff's premises and the resulting damage. Similarly, it was submitted on behalf of Coillte that the escape of water from Paddock Pond, whether occasioned by conduct on its part which it was alleged constituted a breach of duty or by an act of God or otherwise, did not in a legal sense cause the flooding and consequent damage in the Plaintiff's premises. Having held that the claims in negligence against the Council and Uniform are not sustainable and that neither the claim in negligence nor the claim under the rule in Rylands & Fletcher against Coillte is sustainable, strictly speaking, it is not necessary to address the issue of causation. However, as that issue absorbed by far the greater part of the hearing time, I propose expressing my opinion on it.

143. At the very end of a long trial an interesting issue arose on causation. It arose primarily out of Mr. Keane's submission that, as regards Coillte, the crucial question on the issue of causation is whether the dam break waters added to the peak of the flood in the Plaintiff's premises. Mr. Keane submitted that, even if the dam break waters had contributed at any stage to the flooding in the Plaintiff's premises, if the natural flood subsequently increased the level of water within the Plaintiff's premises, that is to say, if the dam break waters passed through and were replaced by natural flood waters, then the natural flood was the cause of the flooding. Mr. Comyn contended that this proposition is not correct as a matter of law and he cited the decision in the Nitro-Phosphate case in support of his assertion. That case was concerned with the liability of a defendant, a dock company, which, it was held had been under a duty at law to maintain a retaining bank around its dock at a height of 4 ft. 2 inches above Trinity high-water mark, but had allowed the retaining bank to be below that level. In November 1875 an extraordinarily high tide took place and the river feeding the dock rose to 4 ft. 5 inches above Trinity high-water mark, in consequence of which the water in the dock overflowed the bank and damaged the property of a neighbouring landowner. It was held that the extraordinarily high tide was an act of God. At first instance a declaration was made that the whole damage sustained by the plaintiffs was due to the defendant's wrong. On appeal, delivering the judgment of the Court, James L.J. said at page 526:-

144. James L.J. then went on to deal with a submission that the whole damage was not due to the defendants' neglect and stated:-

145. As was pointed out by Counsel for the Defendants, the Nitro-Phosphate case is cited in Charlesworth & Percy at page 913 as authority for the proposition that a person who diverts the course of a natural stream may be able to establish, by way of defence, that even if the diversion had not occurred, the plaintiff's damage would have been the same.

146. It seems to me that the ratio decidendi of the Nitro-Phosphate case is that, if it could be established that there were two causes of the damage to the plaintiff's property, the wrongdoing of the defendant and vis major, there should be an apportionment so that the defendant should only be liable for the damage attributable to its wrongdoing. The head-note in the report bears out this interpretation. Interpreted thus, the decision does not give rise to any conceptual difficulties in this post Civil Liability Act, 1961 era.

147. It was submitted that the Plaintiff's case against these Defendants on the issue of causation was tied entirely to the evidence of Dr. Samuel and, in particular, to his modelling of the flood event, and this was conceded, properly in my view, by Mr. Comyn. It follows that unless the Plaintiff has established as a matter of probability that the modelling exercise accurately mirrored what happened on the night of the storm and that the results of the exercise accurately reflected reality, the Plaintiff has failed to establish causation.

148. I have outlined above in considerable detail the criticisms of Professor Cunnane and Dr. Bruen of the modelling exercise and it is not necessary to reiterate them here. Suffice it to say that I am satisfied that the criticisms are well-founded and I am not satisfied that the results of the modelling exercise accurately reflected the actuality of the flood event, and, therefore, the conclusions drawn from the results do not stand up. In my view, the most striking and significant disparity between the model simulations and what we know of the actuality on the night is that the model did not, on any run, achieve the maximum water level within the Plaintiff's store which was observed by Mr. Garland and Mr. Flanagan on the night. In my view, Professor Cunnane's conclusion that the volume of water in the model was underestimated must be correct and, in my view, on the basis of the evidence of Mr. Doyle and Mr. Kennedy, it was grossly underestimated.

149. Not only does the evidence adduced by the Plaintiff fail to establish a causative link between the breach in the river wall and the complained of construction works, on the one hand, and the flooding of the Plaintiff's premises, on the other hand, but the evidence adduced by these Defendants indicates a high degree of improbability that such link existed. I am satisfied on the evidence that the peak flow was at least 285 cumecs and that it was probably higher and that the primary cause of the flooding was the inability of the river channel, which had a capacity to carry a flow of 200 cumecs, to accommodate the peak flow. Moreover, having regard to the evidence of the extent of the flooding, both upstream and downstream of the bridge, the extent of the river wall which was over-topped and the anecdotal evidence of the traditional source of flooding in Little Bray, the only reasonable inference which can be drawn is that what was happening in the vicinity of Bray Bridge was only of the minutest significance, if any, in the flooding process.

150. Mr. Prendiville's observations, in my view, are the most reliable record we have of what happened on the night and I accept Professor Cunnane's analysis of those observations as to the significance of the breach in the flooding of the Little Bray area. In my view, it has not been demonstrated that water flowing through the breach was a material element or a significant factor in the flooding of the Plaintiff's premises. In relation to the afflux at the bridge, whatever the correct measurement of the afflux, in my view, the Plaintiff has not established that the drainage construction works, rather than natural processes, were a material element or a substantial factor in creating it.

151. On the causation issue, the position of Coillte is different from that of the Council and Uniform. The water which had been impounded in Paddock Pond escaped, it joined the Dargle and it flowed through Bray and into the Irish Sea. In my view, the evidence establishes that the dam break waters did have a significant effect upstream and they were certainly an important contributory factor in the collapse of Valclusa Bridge and in Mr. Barnwell's frightening experience.

152. However, the issue which arises here is whether the dam break waters were a material element and a substantial factor in the flooding of the Plaintiff's premises. On that issue, in addition to the results of the modelling exercise, we have the evidence of the so-called "surge" and the inferences to be drawn from that evidence.

153. The criticisms of Professor Cunnane and of Dr. Bruen in relation to the modelling exercise are compounded in relation to the dam break waters because of the controversy, which I cannot resolve, as to the volume of water which escaped from Paddock Pond, which was a finite volume of water. There are other conflicts on the evidence which I cannot resolve, for instance, at what rate the dam break waters attenuated on the journey to Bray and how much of the peak discharge at various points, for instance, River Lane or further downstream at Bray Bridge, was attributable to the dam break waters. Dr. Samuels' estimate, on his evidence, was that at Bray Bridge it was 13.5 cumecs in a total discharge of 233.5 cumecs - about 6%. I appreciate that this is a crude approach and that I have rejected Dr. Samuels' evidence on this point. However, it is indicative and it constrains one to ask how could Coillte be held to be wholly responsible for the flooding of the Plaintiff's premises, which is the case that was made.

154. The manner in which the Plaintiff's claim against Coillte is formulated is that the results of the modelling exercise suggest that it was the presence of the dam break waters which triggered the flooding in the Plaintiff's premises, in that the results suggest that the natural flood alone would not have reached the threshold level of the premises. Leaving aside the question of whether the model accurately reflected the volume of water which was coming downstream, there seems to me to be a fundamental flaw in approaching the question of ascertaining responsibility for the flooding of the Plaintiff's premises in this way in the light of the evidence. In general terms, it is clear on the evidence that there was flooding in Little Bray before the arrival of the dam break waters and that the flood level continued to rise for at least an hour and probably longer after the dam break waters would have passed through Bray.

155. More specifically, on an analysis of the evidence as a whole, the following emerges:-

156. It is undoubtedly the case that for an hour around midnight the presence of the waters from Paddock Pond had some effect on the volume of water spreading out over Little Bray and on the level of the flooding in Little Bray but, in my view, it was manifestly not an effect of an order which would justify the conclusion that the flooding of the plaintiff's premises is solely attributable to the waters from Paddock Pond.

157. Moreover, it seems to me that, had the Plaintiff made the case (which it did not) that, on the authority of the Nitro-Phosphate case, there should be an apportionment between the damage done through the fault of Coillte (assuming for the purpose of this hypothesis that Coillte was at fault) and the damage in respect of which Coillte was blameless, it would not be possible to do such apportionment. The decision in the Nitro-Phosphate case is predicated on the premise that it would be possible to identify a substantial and ascertainable portion of the damage fairly to be attributed solely to a factor in respect of which the defendant in that case was blameless. In the instant case, in my view, the evidence of the eye witnesses of the event indicates that the storm, that is to say, the natural flood, was overwhelmingly the significant cause of the flooding in Little Bray. Taking the evidence as a whole, in my view, it is not possible to identify a substantial and ascertainable portion of the damage which could fairly be attributed solely to the effect of the presence of the waters from Paddock Pond.

158. Finally, in relation to the sudden increase in the level of the water at Lower Dargle Road and Haughton's Junction observed by Mr. Crinnion, Mr. Sterling and Mr. Eaves which occurred sometime between 11 p.m. and 11.30 p.m. on the 25th - the so called "surge", in my view, to infer that this was solely attributable to the arrival of the dam break waters would not be the appropriate inference to draw. No doubt the presence of the dam break waters had some effect. But it must also be assumed that the six hours of very intense rain which were just ending also had an effect. The incessant rain must have caused an inexorable rise in the level of the river and, on its own, may have increased it to a point at which it overtopped the river wall opposite Haughton's Junction. I was impressed by Mr. Kennedy's argument that, if there was a distinguishable surge which was solely attributable to the arrival of the Paddock Pond waters, one would have expected the flood level to settle back to a lower level after the Paddock Pond waters had passed through Bray, which we know did not happen. The probability is that it was a combination of both factors which caused the "surge" observed by the eye witnesses.

159. The Plaintiff had failed to establish its case and its action must be dismissed.