Eastern Suburbs Leagues Club Limited v Royal & Sun Alliance Insurance Australia Limited [2003] QSC 413

[1] MACKENZIE J: This is an action for recovery of moneys under a policy of insurance. The fundamental issue is whether that damage was caused by an event covered by the policy or an exclusion.

[2] During a storm on the late afternoon of 9 March 2001, water entered the licensed club premises of the plaintiff. There was an insurance policy in force with the defendant at that time. It is not in dispute that some water damage was occasioned by water entering through the roof prior to the inundation in circumstances covered by the policy. However, the major damage was caused by water which entered through or under doors in various parts of the premises.

[3] It is common ground that the storm caused rain of one in 100 year intensity to fall over the area where the club is situated. Norman Creek, which broke its banks during the downpour, abuts Brisbane City Council land which is contiguous to the north and west with the plaintiff's football field. The licensed club is situated to the southeast of the football field. A sealed car park abuts the field along the eastern side, and is immediately north of the licensed club. On the eastern side of the club is a street called Main Avenue. The vehicular entrance to the club is from that street. Another street, Sixth Avenue, intersects with Main Avenue immediately opposite that driveway. Vehicles entering the club grounds drive past the main entrance to the building itself on the way to the car park. There is an area for dropping off passengers as well as a roadway giving access to the car park. Four disabled persons' car parking bays are adjacent to the entrance to the main car park, at right angles to Main Avenue. The main entrance to the club building is for practical purposes also opposite Sixth Avenue.

[4] A detailed contour map of the area shows that the drop-off area and driveway just mentioned are about 0.6 metre lower than the floor level of the club. Water would enter the club when the level reached 4m AHD, or perhaps a millimetre or two less. The area where the driveway enters the club premises and adjacent to it in Main Avenue is the lowest point in the locality. Main Avenue slopes downwards in both the north and south direction approaching that point. Sixth Avenue is a sloping street, running down from a substantial ridge which abuts the whole area to the east. Panitya Street which runs north-south along the western side of the football field is generally slightly higher, except for one area to the south-west of the plaintiff's premises, than the floor level of the licensed premises. The football field itself is more elevated along the axis from goal posts to goal posts than at the touchlines. The difference is about 0.3 of a metre and at its highest point it is about 0.9 of a metre lower than the floor level of the club premises. There is also a low concrete wall topped with a chain-wire fence along part of the northern side, the eastern side and part of the southern side of the football field.

[5] In the car park, there are variations in level, with what is referred to by some witnesses as a ridge running in a generally north-east/south-west direction. Despite its description, it is not significantly higher than other areas in the car park. However its importance is that it would control the flow of water from in front of the club by preventing it from flowing to Norman Creek (the natural path being north-west to the end of the football field and then more westerly) until the water level reached 3.34m AHD. Correspondingly it would prevent water from Norman Creek flowing to the area in front of the club until a similar height was reached.

[6] Later there will be reference to people being evacuated to a grandstand. It is the old grandstand, west of the back of the club premises. The contour map shows that a person walking from the premises to the grandstand would walk progressively slightly uphill. The ground level at the grandstand is approximately 0.5m higher than the floor level of the club.

[7] An inspection of the locality was held. This has aided understanding of the oral and documentary evidence. A series of time lapse images from various security cameras in and around the premises was also in evidence. While the quality was at times not very good and it requires specialised equipment made available by the defendant to play it, it did provide objective information against which evidence of witnesses could be measured. It also demonstrated that the rain varied in intensity although at times it was clearly torrential. This observation is confirmed by evidence of recordings made in a BCC gauge not far away.

[8] Evidence was given by a number of people at the club about the events of the afternoon. They painted a picture of water rushing down Sixth Avenue and Main Avenue and accumulating in the area in and around the driveway into the car park in front of the main doors. I have no hesitation in accepting the description given by them in that regard. The video evidence confirms that there was a good deal of water accumulated in that area. There is no reason to doubt that, from the configuration of the area, it may have expanded in a generally northerly direction as events unfolded. There were variations in the evidence, as may be expected, about the times at which various things were observed by witnesses. In some cases what was said about events is difficult to reconcile with the evidence on the video. This is understandable, given the extent of their observations, which were mostly intermittent and the fact that some of the witnesses were somewhat preoccupied with the fate of their motor vehicles which were at that time in the car park which was progressively being inundated. There is nothing to suggest that any of the witnesses was doing other than their best to recall honestly the events of the evening. To the extent that the eyewitness accounts differ from what is shown on the video, the video is accepted as an objective record of events. It is not controversial that the times imprinted on the video are 15 minutes slow. Times referred to in the judgment will be to the actual time.

[9] Ms McLean, a senior employee of the club, first noticed heavy rain at about 5.15pm. Otherwise she did not pretend to have a precise recollection of time. She observed the rain falling for about 15 minutes and noticed the water pooling and rising up the wheels of cars in the disabled car park. She said that some time between 5.30 and 6pm she looked out to the car park and decided that there was no cause for alarm at that time in respect of her vehicle. However, she saw a motor vehicle in the disabled car park fishtailing. She said she did not see the direction of flow, but inferred from the angle the car reached it was from south to north. She observed the water coming in through the front doors although she placed the time at 5.45 to 6pm, earlier than the experts agree and the video shows was the time of entry. She also said that at about 5.30-5.45pm she saw the football field appearing soggy but not with water ponding on it. She also said that about 15-20 minutes after the water came in, she helped evacuate patrons through the rear of the building. She described the area there as soggy.

[10] Ms Smith, the Human Resources Manager for the club, had finished an interview with a prospective employee at about 5.10pm. She was hoping to finish work at 5.30. She said that she spent about 20-25 minutes watching the storm from the auditorium and saw water rising quickly up the wheels of cars owned by staff. After a while, she went to the reception area and helped to answer the phones. While there she saw water gathering in the car park and saw a car in the disabled parking area fishtailing, at a time she estimated as about 5.40pm. About the same time she saw a keg floating from the southern end of the car park towards the north (There is a loading bay on the southern side of the club and the video shows kegs stacked in that area). She said that the water came through the front doors at 6.10pm. She had gone to the loading dock shortly before this, in anticipation of flooding, when the water was lapping the stairs, to get rags to mop up with. She was told that water had come in through the loading dock and observed water in the area. She also saw water running towards the kitchen. She said she remained in the club for an hour to 1 ½ hours. Where she went out through the TAB, there was water, but not as much as was in the club.

[11] Mrs Hartley, a member of the club, said she was conscious of thunder from about 5 o'clock onwards. She thought it was about 6pm when the water came in the front door. She said that she believed that the water coming in was from Main Avenue and Sixth Avenue. She said she saw water coming through the closed door of the coffee shop about 5-7 minutes after the water had entered the front door. Whether it entered first at that time, or whether it was when she saw it first is ambiguous. Her husband, with whom Mrs Hartley was having their evening meal, gave evidence that he saw water coming in the front door and the door leading from the coffee shop to the car park area. Mrs Hartley said they went out of the club about 15-20 minutes after the water entered. The water outside was about up to her ankles. Mr Hartley suggested it was knee deep, on one view of his evidence, although he also said it was only slushy as they walked to the grandstand.

[12] Mrs Thomson was at the club with her husband. She watched the water rising from the auditorium. She said she saw water running down from hills and filling the area in front of the club. She did not see the water moving except to rise. Their car was one that was partly submerged during the incident. She thought it was about three quarters of an hour after the initial entry of water before they were evacuated. She said there was not a lot of water outside at that time, although it was wet.

[13] Mr Mahboop, a club member, said he heard there was trouble at the front door. His recollection was that there were a taxi and a car in that area with water up their doors. However in cross-examination he was not disposed to maintain that the taxi did not drive off. He then went to the coffee shop area where he saw water half way up over the hubcaps of his vehicle in the main car park. He said that water was flowing from Main Avenue and Sixth Avenue into the area in front of the club, into the car park and then towards Norman Creek. About 5-10 minutes later, he saw water lapping the front doors. He went back to check his car. At that time it was up over the bonnet. He said the water was "breaking up over the bonnet and it was still rising". He said that the water was flowing from the southeast to the northwest fairly strongly, heading towards the football field and the creek. He said the water was basically hitting the windscreen and blowing up it. He said that soon after that, water came in the coffee shop door. About one-half to three-quarters of an hour after the water first entered the club he was evacuated. He said that the water outside was much the same depth as in the club or perhaps a little less. He estimated it was 8 to 9 inches deep and thought it was beginning to go away by then. In cross-examination he maintained that the water was flowing towards the creek at the time he was looking at his car. However, he was not convincing in defending his position that there was very little water elsewhere in the car park at the time his car was going under water.

[14] The witness who gave the most detailed account of events was Mr Ford who was a committee member of the club. He said that the rain started at about 4.45pm. It was very heavy (a "cloudburst") by 5pm. He said that the water flowed into the driveway and spread around the north-east corner of the club premises. He said it was not a torrent but was spreading out. By 5.30pm he observed water in the area beyond the cricket nets, which are beyond the northernmost perimeter fence of the club's land and directly north of the football field. He also saw water in the touchline areas of the football field. He said that although the water appeared to be gradually coming up, it was not at that stage over the crown of the field. About 15 minutes later he looked again and saw that the volume on the field had increased.

[15] At about 6pm he saw water spreading into the car park "...increasing in depth all the time...". He said there was a gradual spreading, not a torrent of water. He did not look beyond the immediate area and could not say what was happening at the northern end of the car park. He said he had a clear recollection of water entering the club at 6.10pm. To his observation the surrounding streets were still contributing water to the accumulation in front of the club. He said he did not observe any flow; the pool of water simply seemed to be increasing in height.

[16] His observations were made out of a concern that the water might enter the club. His focus was principally on the water in front of the club and on the north side. He went at various times to various points to look out. Because of his role, his observations seemed more purposeful and likely to be more reliable as a whole than those of witnesses who were distracted by other concerns or merely interested bystanders. He said that at perhaps 6.20pm when the evacuation of the club was being contemplated, he saw the football field completely covered and the water about to overtop the wall on the southern side of the ground. When he went outside at the rear of the club, the water was about ankle deep.

[17] There was also evidence from a resident in Panitya Street, Mr Bean. Mr Bean's home is on the western side of the street and abuts the BCC land on the banks of Norman Creek. It is situated approximately behind the western grandstand on the football field. He and his wife were caught in the storm while at Garden City. They encountered flooding on the roads on the way home where they arrived about 5.40 to 5.45pm. Soon after arriving home Mr Bean noticed water from Norman Creek had entered the park and was about 50 metres north of his fence. That was about half the distance from the creek's normal channel to his property. He said the water was not very deep but was dirty, muddy and full of rubbish. He also noticed water flowing from the southerly direction off the football field.

[18] He went inside intending to see the 6 o'clock news. However, he did not do so because he went under his residence and noticed water rushing in. He said that suddenly, about 6:05 or 6:10, a 6 inch wave of water came down Panitya Street and entered the built in area under his house. The wave of water met the Norman Creek water, creating a wave-like effect. Having decided that he could do nothing about it, he went upstairs to his patio. He saw water coming from the football field, which was relatively clear, meet the dirty water from the creek. He described the turbulence that happened as a "prawn boil", occurring in an area marked as "DB2" in Exhibit 19, some distance into the BCC land south of the end of Panitya Street and north-west of the corner of the football field. He said that it had the effect of pushing the muddy water back in the direction of flow of the creek. He said it would have occurred about 6.15 - 6.20.

[19] He said that at about 6.20 - 6.25, he saw a demountable shed used as a bar at the northern end of the football field pushed off its supports by the water coming from the football field, and turned about 90 degrees. This occurred about 20 - 25 metres from the "prawn boil". He also said chain wire fencing was knocked down in the area of the "prawn boil". (The fences fell towards the club premises and the predominant amount of debris adhered to the side facing the creek. This was a common feature, even on fences still standing after the flood subsided).

[20] In cross-examination, Mr Bean took issue with the amount of debris shown in the photographs maintaining that there had been much more of it, including a quantity on the inside as well. He said that no creek water got on to the football field. He denied that at about 5.45pm, creek water was moving towards the clubhouse over the football field. He said that he was sure about times because he kept checking his watch. He said in re-examination that at 5.45pm the water was not up as far as swings in the park, which he marked as "DB4" on Exhibit 19.

[21] As a result of an order of the court before trial the expert hydrologists, Dr Weeks for the plaintiff and Dr Macintosh for the defendant, conferred and produced a report which distilled their separate reports into a form where what was common ground and what remained in dispute were identified. I was greatly assisted by each witness's analysis of the issues in their various reports and in their evidence. One issue upon which consensus was not reached was the level of flood waters from Norman Creek at critical times and therefore the extent to which they may have influenced the ultimate question to be decided in these proceedings. The variances are small; differences in methodology are one reason.

[22] Each expert used objective data derived from measurements of rainfall and recorded stream levels selected by each as an appropriate basis to assist in forming an opinion. In support of the validity of their conclusions, eyewitness accounts, principally in the case of Dr Weeks, and, in the case of Dr Macintosh, mathematical modelling and incidents said to be observable on video were relied on. This is not a case where either expert suggested that the techniques used by the opposing expert were inappropriate for the purpose. The emphasis, so far as the hydrologists were concerned, was largely on whether factoring in data from a gauge upstream from where Norman Creek relevantly broke its banks, together with a gauge about half a kilometre downstream from that point, as Dr Macintosh did, was likely to produce an unduly early time at which waters from the creek might flow towards the club house and contribute to its inundation and whether Dr Weeks' reliance on just the gauge downstream as the basis for calculation was likely to produce a time that was unduly late for the same event.

[23] There was a good deal of analysis of these issues in evidence and addresses. The time difference is quite small. Given that any opinion which involves assigning a precise time to events which are not measured on the spot is subject to a plus or minus qualification, I would not, in the circumstances, have been confident that a definitive conclusion even on the balance of probabilities could have been reached on the hydrological evidence alone. Because of the height the flood eventually reached, there would have been a point when the water from the creek would have played a dominant role in events, so that aspect of the evidence is not controversial. The timing is, for determination of the case, critical.

[24] Because of that, evidence from other sources tending to shed light on what actually happened becomes particularly important. On one hand there is evidence of eyewitnesses, summarised previously, with respect to the direction of movement of water and the build-up of water in various areas around the club house. On the other hand there are events recorded on the video. Those particularly focused on by the defendant were the sudden rising of water outside the front door of the club after a period when the level appeared to remain static, and objects apparently floating in a west to east direction in the main car park not long before inundation of the club premises occurred.

[25] With respect to the rising water outside the front door, the reference point is a tyre of a taxi parked in the driveway. After a period when it was standing with water partly up the tyre and apparently remaining at about the same level, the water can be seen rising quickly from about 5.55pm onwards. The water rises up the tyre of the taxi which then disappears from the area. This occurred after the peak storm water flow rate has passed. It suggests that some other influence was operating to cause what can be seen.

[26] With respect to the direction of flow of water, the critical part of the video is concerned with what appeared to be a brown object and a smaller white object floating between rows of cars in a generally east-west direction. Both experts viewed the video, which is in time lapse form, with a view to agreeing upon what it showed. It was agreed that both objects had a velocity of 0.3 metres per second. The brown object appeared at approximately 6.08pm and was visible for 30 seconds while travelling an estimated distance of 8 metres before it disappeared from view. The white object appeared 5 seconds earlier and was visible for 16 seconds while travelling an estimated 4 metres.

[27] Certain other facts are not in dispute as the Report on Outcomes from Expert Hydrologists' Conferences establishes. The peak runoff rates in the local storm water catchment would occur within 10 to 20 minutes. It was accepted that the highest rate of rainfall was at about 5.20pm. Peak runoff rates would have occurred at around 5.30 to 5.40pm. At the time the water began to enter the club (6.10 to 6.15pm) the runoff rate would have been "significantly less" than at the peak of the event (The evidence of the video camera suggests that the first entry of water was at 6.10pm with the water having reached substantially inside the club by 6.15pm). No other local catchment would have contributed to the runoff in the catchment in which the club was situated. Peak inundation occurred at 7.00pm. It was at a uniform level of 4.35 metres AHD about the site although it was slightly lower in the club itself.

[28] BCC information as to rainfall and creek heights (including the maximum height data) was accurate. The Caswell Street gauge was representative of the rainfall in the locality of the club. (It shows that about 24mm fell from 4.30 to 5.00pm, about 70mm from 5.00pm to 5.30pm and about 32mm from 5.30 to 6.00pm. It rained for some time after that but not as heavily).

[29] It was common ground in the evidence that the Caswell Street gauge was about 530 metres downstream from the relevant point where the water overflowed from the creek. It would take about 7½ minutes for water to travel from there to Caswell Street at the relevant time. The gauge at Preston Street, which measured only the highest water level reached as opposed to giving periodic readings (which the Caswell Street gauge did), is about 50 metres downstream of the exit point of water from the creek near the club.

[30] There was also evidence that was not disputed that the maximum water level recorded at Caswell Street was 3.66 metres AHD. At Preston Street it was 4.32 metres AHD. The maximum level at the club was 0.03 metres higher than at Preston Street. For the purposes of calculation it was therefore appropriate to add 0.69 metres to the level recorded at Caswell Street to obtain an equivalent level at the club. The lower AHD levels at Preston Street and Caswell Street reflect, of course, the fact that the relative level of the stream drops as it goes downstream.

[31] It is convenient now to address the relevant parts of the policy under which the plaintiff's claim is made. By virtue of clause 2 and clause 2.5 the policy provides that the insurer will indemnify the insured against damage occurring to property insured, subject to the damage not occurring in circumstances which are excluded by clause 6. Clause 6.2 provides that section 1 of the policy (which includes clause 2) does not cover damage to any property insured:

Provided that the Insurer will indemnify the Insured for any Damage to Property Insured caused directly by any circumstances not excluded under Section 1 of this Policy, notwithstanding that these circumstances may in turn have been caused by any of the circumstances referred to in Exclusions 6.2.1 to 6.2.16."

"Flood means the inundation of normally dry land by water escaping from or released from the normal confines of any natural watercourse ..."

[32] It was submitted on behalf of the defendant that the words commencing with "provided" in clause 6.2 were an exception to an exclusion. As such the onus lay on the plaintiff to show that the loss was proximately caused by one of the perils falling within the exception to the exclusion. On the other hand, it was submitted for the plaintiff that the defendant bore the onus of proving that loss to the plaintiff was caused by an excluded event.

[33] The effect of the words commencing with "provided" is that the insurer will indemnify the insured if a non-excluded event is the primary cause of the loss, but it was in turn caused by an excluded event. In the context of this case the proposition would be that the damage caused directly by entry of water into the club was not a flood but that its entry was in turn caused by flood. Whether the words have any application to the case will depend on characterisation of what the evidence establishes to have happened.

[34] The first submission, rather ambitiously made, on behalf of the plaintiff was that the proximate cause of the inundation was rain. As it was expressed in the plaintiff's counsel's address:

"... our primary position is that whatever view is taken of interaction between stormwater and creek water on the night in question the real cause of the inundation of the club was the rain, unusual storm. The rain fell to earth, it descended upon the club in different directions, some as local catchment water, some as entering the creek and then coming back towards the club. But either way, it was the highly unusual, one in a hundred year storm which produced both manifestations and both lots of water bearing upon the club. In our submission, the proximate cause was the rain, not the creek."

[35] I do not accept that proposition, which sought to gain support from Petersen & Shadomill Pty Ltd v Union des Assurances de Paris IARD (1997) 9 ANZ Ins Cas 61-366. The findings of fact set out below lead to that result. The policy in that case provided that loss and damage caused by water was covered, subject to an exclusion of loss and damage caused by flooding. An exception from the definition of general flooding related to flooding from drains, which was interpreted as meaning that loss and damage from that source alone was recoverable under the exception (77-304). The appellant failed to prove that a significant part of the water that inundated the premises had not come from elsewhere than a drain.

[36] In the alternative the plaintiff in the present case submitted that it was a case of flooding by stormwater. It was submitted that the water actually escaping from Norman Creek must inundate the insurer's premises and thereby cause the loss if the flood exclusion is to apply. It was submitted that, for the flood exclusion to apply, it was not enough that water escaping from Norman Creek acted as an impediment to stormwater runoff, preventing it from getting away from the club's building thereby leading to a rise in the level of stormwater ponding around the club house and eventual inundation by the stormwater. In anticipation of the defendant's reliance on Provincial Insurance Australia Pty Ltd v Consolidated Wood Products Pty Ltd (1991) 25 NSWLR 541, it was submitted that that authority was distinguishable because there was, in that case, no provision in the policy corresponding to the proviso.

[37] It was submitted that Mahoney JA in Provincial Insurance had held that loss was "occasioned by or happened through" the inundation of normally dry land by water escaping from a watercourse in circumstances where, by reason of the inundation of normally dry land by water escaping from a watercourse, other water was forced into the insured's premises and caused loss. The distinction in the present case was said to lie in the proviso because the direct cause of the loss would be the "other water" which entered the insured's premises, not the floodwater. Therefore the loss would not have been caused directly by a circumstance not excluded under section 1 of the policy since the inundation was inundation by stormwater. The flood exclusion would not apply notwithstanding that the stormwater was forced into the insured's premises by the inundation of normally dry land by water escaping from the creek.

[38] It was further submitted that if the initial inundation of the clubhouse was by stormwater and that in itself was sufficient to cause the damage, it did not matter that water escaping from the creek subsequently entered the clubhouse. It was asserted, correctly, that there was ultimately no dispute that the initial inundation, from whatever cause, caused the damage by reason of the location of the property damaged, although the actual extent of damage was contentious.

[39] Provincial Insurance was concerned with an identical definition of "flood". One issue in that case which does not arise in the present case is whether the channel from which the water escaped was of a kind described in the exclusion. However, Mahoney JA addressed an argument that the loss and damage was not of a kind "occasioned by" or which "happened through" a flood, as defined, in the following way at 564:

"The argument was, in effect, that the water which entered the insured's premises and caused the loss or damage was not shown ever to have been in the relevant canal or natural watercourse and therefore it was not water "escaping ... from the normal confines of" the watercourse or canal. Therefore, the argument suggested, the exclusion did not apply.

I do not think that that argument should be accepted. For the exclusion to apply, it is not necessary that the precise water which escaped from the watercourse or canal be identified as having actually entered the insured's premises. The exclusion is of loss or damage "occasioned by or happening through" the inundation of normally dry land "by water escaping ... from ..." a watercourse or canal. If, by reason of the inundation of normally dry land by water so escaping, other water was forced into the insured's premises and occasioned loss or damage, that would, in my opinion, be loss or damage "occasioned by or happening through" the escape caused by such a flood."

[40] Kirby P at 555 referred to the trial judge's factual conclusion that the water had undoubtedly come from the channel in question and to support given to that conclusion by the objective evidence of the kind of damage, the proximity of the premises to the channel and the "...inherent likelihood of that source spilling over and preventing the escape of runoff...".

[41] Einstein J applied the passage from Mahoney JA's judgment with approval in Hams v CGU Insurance Limited [2002] NSWSC 273; (2002) 12 ANZ Ins Cas 61-525 at 76-183. The concept of something being "occasioned through" some other circumstances was discussed in Mercantile Mutual Insurance (Aust) Limited v Rowprint Services (Victoria) Pty Ltd [1998] VSCA 147 especially by Callaway JA at [24]. The concept implies that there is a consequential or causal relationship, not necessarily a direct or proximate cause.

[42] As previously mentioned the defendant submitted that the words quoted from clause 6.2 commencing with "provided" had no application to the present case since the direct cause of loss was inundation of ordinarily dry land by the escape of water from Norman Creek. It was therefore unnecessary to engage in analysis of whether the inundation by runoff water was "in turn" caused by flood. In this connection, the defendant relied on Wayne Tank & Pump Co Ltd v Employers Liability Assurance Corporation Ltd [1974] QB 57; Elilade Pty Ltd v Nonpareil Pty Ltd [2002] FCA 909; (2002) 124 FCR 1, Hams v CGU Insurance Ltd, and Petersen & Shadomill Pty Ltd v Union des Assurances de Paris at 77-034 as authority for the principle that where there is more than one proximate cause of damage, one of which entitles the insured to indemnity and the other an exempted event under the policy, there is no obligation on the insurer to indemnify the insured.

[43] Particular reliance was placed on the decision of Angel J in Prosser v AMP General Insurance Ltd [2003] NTSC 80 which, the defendants submitted, was strikingly similar factually to the present case. Angel J made a number of crucial findings, summarised as a conclusion that water that inundated the insured's property was a combination of accumulated surface water and sinkhole ground water dammed up by flooding in the Katherine River. There were a number of effective or proximate causes of inundation. His ultimate conclusion was expressed as follows:

"[12] The plaintiffs have established that their loss was caused by an event within the terms of the policy viz. "rain falling naturally from the sky including rainwater run-off over the surface of the land". The defendant has established a concurrent cause was the Katherine River overflowing its banks both east and west of the plaintiffs' property trapping the rainwater and groundwater accumulating on the plaintiffs' property.

[13] There being concurrent effective or proximate causes, one covered and the other excluded by the policy, the defendant insurer is not liable on the policy: Wayne Tank & Pump Co Ltd v Employers Liability Assurance Corp Ltd [1974] QB 57; Peterson v Union des Assurances de Paris IARD (1995) 8 ANZ Insurance Cases 61-244; on appeal (1997) ANZ Insurance Cases 61-366 at 77,034; HIH Casualty & General Insurance v Waterwell Shipping [1998] NSWSC 436; (1998) 146 FLR 76 at 83, 84."

[44] The principle in Wayne Tank is well established. For the reasons that will be developed shortly, in my judgment the cause of the inundation was influx of water from the creek into the area where there was the existing pond of runoff stormwater which caused water to rise to a level at which the inundation of the club premises occurred. I do not accept that the runoff stormwater and the creek water can be treated as two discrete bodies of water, on the evidence in this case, for the purpose of the proviso. In other words, there is no basis for finding that there was inundation by the runoff water, as a discrete body of water, in turn caused by the flood water blocking its dispersal in a way that would have avoided inundation.

[45] I am satisfied that water from Norman Creek overtopped the ridge in the car park and thereafter contributed to the pond of water to the southern side of the ridge, causing the overall level of water in front of the clubhouse to rise. In that sense there are no sequential causes as envisaged by the proviso. Looked at in a different way, there was one body of water even though intermingling may have been only partial and perhaps quite small by the time the inundation occurred. It is therefore not simply a case where the blocking action of the creek water caused stormwater to continue to build up to a level where it inundated the premises because it could not flow away.

[46] There were three matters upon which the experts could not resolve their differences. The first related to the Norman Creek hydrograph. The difference in the methods of calculation were that Dr Weeks plotted water levels over time in reliance on levels at the Caswell Street gauge (500 metres approximately downstream) whereas Dr Macintosh used a weighted average which took into account the differences in distance between the Southeast Freeway gauge (about 3.7 kilometres upstream) and the gauge at Caswell Street. The process that Dr Macintosh adopted is explained in detail at T391-2. Dr Macintosh's method tends to result in a more rapid increase in flood levels in the creek than Dr Weeks' method. The time of inundation from the creek at the Club would be earlier under Dr Macintosh's method than Dr Weeks'.

[47] Calculations performed by Dr Macintosh had the water level reaching the floor of the club's premises at about the observed time. This implied that the flood levels in Norman Creek were the controlling influence at the time of inundation. Dr Weeks' approach led to the possibility that the water level in Norman Creek was lower than the floor level at the time of inundation. Since the water level in the creek would not have been sufficiently high to cause inundation, the initial inundation must on this view have been contributed to by local catchment runoff rates sufficient to raise the level beyond 4 metres AHD.

[48] There was also non-agreement on the mechanism of the initial entry of water into the club. Dr Macintosh was of the opinion that the flood level in Norman Creek was the controlling influence at the time of inundation. Dr Weeks was of the opinion that at the time of initial inundation water levels in Norman Creek were not high enough to cause it. There must have been contribution from local catchment run off rates to cause it. Dr Macintosh includes in his report of 2 April 2001 an "estimated events sequence" consisting of diagrams illustrating his opinion of the relationship between stormwater runoff and creek water overflow and the rates of flow of runoff and water overflowing from the creek at quarter hour intervals from 5 pm to 7 pm. Dr Weeks did not perform calculations of this nature. Asked about acceptance of Dr Macintosh's figures, he said they seemed to be of a correct order of magnitude. Subsequent cross-examination tended to reveal some scepticism based partly on eye-witness observations and on the proposition that is undoubtedly true that models only provide a certain amount of information. Dr Macintosh accepted this and also that the outcome produced depended on the quality of information inputted. Nevertheless Dr Weeks did not distinctly challenge the notion that the figures proposed were not unreasonable assumptions. According to the joint report, the values given in the diagrams are based on calculations carried out as part of the assessment.

[49] The third area where agreement was not reached was the interpretation of evidence of turbidity of the water. There was a good deal of evidence from eyewitnesses that the water that initially entered the Club appeared clean and contained only debris consistent with having come from the gardens at the front of the Club. There was also evidence that the amounts of sediment deposited in the Club after the water receded were small. It was common ground that there is a degree of subjectivity in relation to turbidity as an indicator of the source of water. It was common ground that neither expert had the opportunity to inspect the sediment deposits within the club prior to clean up operations. The joint experts' report records that inspections in the general area of the flood revealed differing levels of sedimentation. The point was also made that because of the likelihood that there was mixing of creek water with ponded run off, the sediment concentrations would be expected to have been less than that within other water escaping from Norman Creek. Each expert took the view that the degree of sedimentation supported his conclusion as to the kind of water that inundated the club.

[50] There are three things that can be said about this evidence. One is that there is inevitably a degree of subjectivity about the extent of turbidity. There is no reason to doubt that the water may have appeared clear to eyewitnesses in the club at the time of the initial entry of water into the premises. The quantity of fresh water in the ponded area would have substantially diluted any contribution from creek water. The second is that if some silt was deposited in the club during the incident, it is not decisive as to the nature of the water that entered the club initially since the most probable conclusion on the evidence is that at some time water containing a greater proportion of creek water entered the premises than at the beginning. The third is the fact that eye witnesses described the water in a way that suggests it was essentially clear water albeit with garden debris in it also not determinative of the critical issue in the case.

[51] Resolution of differences between the evidence of the eyewitnesses, which is essentially subjective, and the more objective evidence such as that on the video and the evidence of relative levels in the area is fundamental to the decision. The first important matter is the opinion shared by the hydrologists that peak run off would have occurred into the area in front of the club no later than about 5.40 pm. If the level of the water collected in that area had reached 3.3 metres and the creek had not yet risen to that level, the stormwater run off would have been able to escape to the creek. The apparent equilibrium demonstrated by the water shown in the video remaining at much the same height relative to the tyre of the taxi for a period suggests that either the water was getting away by spreading out and/or running away towards the Creek at a sufficient rate to accommodate any continuing stormwater run off. There would probably still have been some surplus water running into the area since it was still raining, although not as heavily as earlier.

[52] Then, at about 5.55 pm there is a quite sudden rise in the water level evident on the video. There is no visible evidence of the direction of flow in the area in front of the club. Nevertheless the phenomenon suggests that at about that time there must have been an inflow into the area in front of the club. Stormwater runoff is not a plausible explanation, for the reason given above. In the evidence of the hydrologists, the issue of flow reversal was explored in some detail. Perhaps in simplistic terms, the mechanism for that is that water from the creek reaches a higher level relative to the water flowing towards the creek and begins to assert a dominant role. Since the flow of water is predicated on a difference in water levels, the creek overflow begins to run towards the clubhouse once it has overtopped the ridge in the car park and is higher relative to the rainwater ponded around the club. Water which had been spreading from the pool in front of the club house would not have been able to continue to run towards the creek. The flow reversal would not necessarily be obvious as a visible flow but the fact that the water was rising would at least show that the process was at work in the absence of any other plausible explanation.

[53] The hypothesis that by the time the water began to rise, the flow reversal had occurred and water from the overflowing creek was contributing to the rise in level in front of the club is, in my view, supported by the video evidence of objects floating generally wast to east. This is consistent with the flow being other than in a generally more westerly direction which would be the case if the flow of water from the pond in front of the club towards the creek was the dominant flow.

[54] I have given serious consideration to the plaintiff's submissions that the evidence of the floating objects is too slender to rely on as evidence of the direction of flow. Those included that the field of vision was very narrow between the cars; it was not possible to see where the objects came from or where they went subsequently; there was no evidence whether something was happening beyond the angle shown by the camera which caused them to move in that direction. In particular it was suggested that a motor vehicle might have been moved and created a wash or that someone walking through water did something which caused them to be propelled in that direction. I also considered the possibility that the action of the wind might have had an influence.

[55] Some of these kinds of possibilities were put to Dr Macintosh who nevertheless maintained, for reasons explained particularly at T417, that his interpretation of the objects being debris moved by the flow of water was a reasonable one. Dr Weeks was not cross examined about the issue, but he had seen the video prior to giving evidence and offered no competing interpretation of the nature and movement of objects. The fact that he had not been cross-examined was the subject of a mild complaint, but the plaintiff conceded in addresses that it was unlikely that he could have added anything further with regard to the issue had he been asked.

[56] The evidence of eyewitnesses and particularly that of Mr Mahboop and Mr Bean has been analysed earlier. While I have no doubt that they were doing their best to recount accurately their recollection of what was happening at the times they stated, their interpretations of events, to the extent that they are inconsistent with the evidence on the video, must inevitably be discounted. In the case of Mr Bean, the physical evidence of the direction in which the fence fell and the side upon which debris was clinging suggests that his interpretation of the dynamics of the water was not what was happening.

[57] It is also necessary to address the plaintiff's reliance on evidence about the timing of entry of water into the club through the front doors and the coffee shop door. The argument essentially was that the fact that some witnesses said that water entered through the front doors some minutes earlier than through the coffee shop supported the conclusion that the direction of flow was towards the creek at that time. The evidence of timing depends on the preciseness of observation of witnesses whose focus was not necessarily on what was occurring out the coffee shop door. Further, the front doors were self opening doors, which the video shows did open on occasions when people passed within range of the sensors that activated them. Early in the process of inundation, that caused a substantial amount of water to enter. There is also evidence on the video of water building up outside the doors and seeping through between them, while maintaining a differential in height between the higher water outside and the level of the floor of the club. The coffee shop doors are conventional doors, configured as exit doors. The point to be made is that a more tightly sealed door would be likely to retard entry of water more effectively. This proposition was advanced (T 337) by Dr Macintosh to explain why, if the observations of Mrs Hartley and Mr Mahboop were precise as to timing, the evidence did not falsify his hypothesis. In the end, I am not persuaded that the eyewitness evidence outweighs the more objective evidence on the video.

[58] In my judgment, the combination of the sudden rising of the water in the ponded area at about 5.55 pm and evidence of movement at about 6.08 pm of the objects in a direction consistent with flow reversal having occurred leads me comfortably to the conclusion that on the balance of probabilities flow reversal had occurred and that the overflowing creek water was contributing to the level of the body of water in the vicinity of the club by the time that it inundated it at about 6.10 pm. The proviso in clause 6.2 does not apply, having regard to the finding of fact.

[59] It follows that the plaintiff fails on that aspect of the case, although it is entitled to recover the agreed sum for damage caused by rainwater penetrating the roof of the premises. It was common ground that $20,000 paid by cheque as settlement for the whole liability had been rejected.

[60] With respect to damages $17,468 was agreed as the value of damaged caused by water which entered through the ceiling and covered by the policy. There was an issue as to whether reinstatement or indemnity value was recoverable, so far as the defendant was concerned and whether, from the point of view of the plaintiff, what was being claimed was in any event only indemnity value. Apart from those issues, the defendant's submissions were limited to the poker machine bases, the post-mix drink system and the applicability of interest under s 57 of the Insurance Contracts Act 1984 (Cth) apart from those items the agreed quantum was $459,381.06.

[61] For the purpose of clause 4 of the policy the following definitions apply:

"4.1 Reinstatement Value is the cost necessary to replace, repair or rebuild the Property Insured to a condition substantially the same as but not better or more extensive that its condition when new.

4.2 Indemnity Clause is the cost necessary to replace, repair or rebuild the Property Insured to a condition substantially the same as but not better or more extensive than its condition at the time that the Damage occurred taking into consideration age, condition and remaining useful life."

By clause 4.3 the basis of settlement of items of the kind in dispute is reinstatement value.

[62] In the "reinstatement provisions" applying to such items the following appear:

"4.4.2 Such work must be commenced and carried out with reasonable despatch, failing which the Insurer shall not be liable to make any payment than the Indemnity Value."

"4.4.4 The Insurer shall not be liable to make any payment beyond the Indemnity Value of the Property Insured until a sum equal to the Reinstatement Value has actually been incurred by the Insured in the replacement, repair or rebuilding of the property, provided that where the Insured repairs or rebuilds any Damaged Property Insured at a cost which is less than the cost of the reinstatement but greater than the value of such property at the time of the Damage, then the cost so incurred shall be deemed to be the cost of reinstatement."

"4.20 The Insurer shall not be bound to replace, repair or rebuild exactly or completely, but only as circumstances permit and in reasonably sufficient manner and shall not in any case be bound to expend more than the applicable Limit or Sub-Limit of Liability."

[63] If the insured wished to avail itself of reinstatement value it had to carryout such reinstatement works with reasonable dispatch. If it did not do so it lost the right to claim reinstatement costs and was limited to the indemnity value basis of settlement. Reliance was placed on CIC Insurance Ltd v Bankstown Football Club Ltd (1997) 187 CLR 384, 401. With regard to the concept of the sum equal to reinstatement value actually being incurred, reference was made to the same authority at 403. Although obiter, the majority expressed the view that no payment beyond the amount which would have been payable under such a method of assessment was to be made until the sum equal to the cost of reinstatement had actually been occurred.

[64] For the plaintiff it was submitted that it was not asking for more than indemnity value. The submission focused on the definition of indemnity value in clause 4.2 and the proposition that on the evidence, the plaintiff had met that requirement.

[65] With respect to the poker machine bases, the evidence was that they were made of fibreboard laminated with veneer. The fibreboard tended to swell after immersion in water and remain distorted. The purposes served by the bases were firstly to support the poker machine itself. The second was to provide a secure area in a locked compartment to hold the coins until the machine was cleared. The plaintiff relied on evidence of Mr Smith, the club's maintenance supervisor, to the effect that if the bases were to be reinstated to their condition before the inundation, it would be more cost effective to replace them than send them to a specialist cabinet maker. In some instances, the plinths upon which the bases stood disintegrated during the clean up after the flood. Some had an appearance consistent with water damage to the veneer at the time of the inspection during the trial. Mr Smith also made the observation that the atmosphere of a club is adversely affected if it has shabby fittings, even though the bases maybe functional.

[66] Mr Smith said that he had done repairs on the bases of 30 to 40 machines, particularly to ensure that the doors were operable to allow the cash boxes to be cleared. There was no specific evidence that prior to the inundation the bases were significantly depreciated in value or in a significantly devalued condition. The fact is that a certain amount of work has been done on a number of them. Others appear to have continued to function adequately. I am satisfied that some damage has occurred as a result of the flood to the poker machine bases but I am not persuaded that there was a need to replace all of them. There was evidence that some unspecified number of second hand bases were purchased. It is assumed that they were suitable for the purpose, in the absence of evidence to the contrary. In all of the circumstances I am not persuaded that replacement of the 230 bases quoted for at $388 each represents the cost necessary to establish the indemnity value of the damaged bases within the meaning of the definition in clause 4 of the policy. It therefore becomes a matter of impression as to what should be allowed. I would allow $20,000 under this head.

[67] With regard to the post-mix system it is a system where pipes in a duct under the floor convey drinks from a chiller to various service points within the club. The lines conveying drinks are surrounded by lines containing the coolant liquid. Maintenance of the chilled temperature until the drinks arrive at the serving point is achieved by coating the lines with an insulating substance. The composite of lines and insulation is known as a python. There was evidence that when the insulating material becomes sodden, as would be likely to occur in a flood where water entered the duct, the insulating properties are adversely affected.

[68] Following complaints by customers about the taste of spirits, it was found that the duct in which the python ran had been flooded. It was pumped out and the lines flushed out. There was evidence that the insulation at least around the beer lines was not as effective as it had been because the temperature at the tap was more than 1 degree warmer than expected. There was also evidence that a rubbery taste could penetrate the lines if the insulation deteriorated. There was also evidence that because of the difficulty of completely drying out ducts the only practical solution is to remove the pythons and physically dry the ducts by pulling something through them to do so. There was also evidence that it would not be feasible or as effective to use a cheaper alternative solution proposed by another company.

[69] On the balance of probabilities I am satisfied that the insulation in the post-mix system was adversely affected by water penetrating into the duct and in the absence of evidence that there was any substantial depreciation in value of the pythons at the time of the flood, I allow $112,807 as indemnity value of the post-mix system by reference to a quotation in Exhibit 24. I note that complaint was made about an item of builder's margin and overheads in the quotation. The basis of the complaint was that the actual installation of the pythons would be subcontracted to an expert in the field. However, there is nothing to suggest that what was applied is not a normal mark up in similar circumstances.

[70] With regard to interest under the Insurance Contracts Act, s 57 provides that the insurer is liable to pay interest on an amount for which it is liable to pay a person from the day when it was unreasonable for the insurer to have withheld payment of the amount. The defendant resisted the plaintiff's claim for interest at the higher rate that would be payable on the basis that the defendant had not acted unreasonably. There was a serious question as to the club's entitlement to recover at all. Further, the amount claimed by the plaintiff in the statement of claim was substantially greater than it would obtain at trial even if successful on the disputed items. It was submitted that it was not unreasonable for the defendant either to defend the merits of the claim or the quantum. Whether a defendant has acted unreasonably is a matter of fact depending on all the circumstances of the case (Elilade Pty Ltd v Nonpariel Pty Ltd; Settlement Wine Co Pty Ltd v National & General Insurance Co Ltd (1994) 62 SASR 40). Had the plaintiff been successful on the merits, I would have declined to find that it was unreasonable for the insurer to have withheld payment.

1. I give judgment for the plaintiff in the sum of $17,468;

2. The parties are at liberty to make written submissions as to costs within 7 days of the delivery of judgment.