Flanagan v Bernasconi

The CGU policy and the Vero policy

1. In February 2011, Dr Flanagan renewed her Householders insurance policy with CGU, ensuring coverage for the period from 28 February 2011 to 28 February 2012. The Product Disclosure Statement ("PDS") for the CGU policy, under the heading "What is insured", provided that CGU would cover "your buildings and contents for any accidental damage or accidental loss". The PDS also provided that CGU would not cover the insured for any accidental damage or accidental loss caused by, relevantly, "a defect in an item, faulty workmanship, structural defects or faulty design". Under the heading "What you are required to do for us", the PDS relevantly stated, "You must take reasonable precautions to prevent anything which could result in a claim under this policy".
2. As the primary judge noted at [174], in 2011 the premium for the CGU policy was $8,440.59. On Mr Bernasconi's evidence, in late February 2012 he learned that the premium CGU was quoting to renew the policy was $15,414.94. He expected that Dr Flanagan would be unhappy about the increase and asked another employee of Nadic to look for alternative quotes: [175].
3. At a meeting on 15 March 2012, Dr Flanagan's evidence, which the primary judge accepted (at [187]), was that Mr Bernasconi recommended to her that she take out the Vero policy instead of the CGU policy. The Vero policy covered the "home", which included the house but also "any fixture or home improvement on the site" including, relevantly, "in ground pools". However, there was a difference between the CGU policy and the Vero policy which her Honour found that Dr Flanagan did not understand in accepting Mr Bernasconi's recommendation. (at [181]) Pursuant to the Vero policy: 1. an insured could only claim for loss or damage to home or contents, or both if, relevantly, "the loss or damage is not caused by any of the excluded events listed on pages 24 to 27"; and 2. on page 26, under the heading "Excluded events", item 12 identified "Events involving swimming pools including", relevantly, "the pool lifting or any area around the pool lifting".
4. Both parties called experts in the field of insurance, Mr Le Plastrier for Dr Flanagan and Mr Gribbin for the respondents. The experts agreed that policies of home insurance do not necessarily cover the risk of a pool lifting under accidental damage cover: [272]. They also agreed that no insurer would have offered cover for the risk of the pool lifting at the cost of the Vero premium. The terms on which such cover was provided would depend on the particular policy: [273].

How the pool came to be empty

1. The primary judge found that the pool had been empty since at least the end of winter in 2012: [62]. Dr Flanagan gave evidence as to how that came to be the case. According to her affidavit, affirmed 22 October 2019, in December 2011 Dr Flanagan had noticed a large increase in her water bill. In late January 2012, she had an on-site meeting with a plumber, Mr Houghton, who said that he would disconnect the water connection to the pool to stop it automatically refilling, and told her that it was "a relatively easy process to empty the pool and it will be easier to fix with no water in it". Dr Flanagan's evidence was that she said to Mr Houghton: "I am still waiting for Geoff to refinance and so I can't get to it right now. I will monitor my water bills for the next couple of months and we can revaluate (sic)? That way we can fix any issue as soon as I have the funds come in, and we are sure it is being caused by the pool."
2. According to Dr Flanagan, Mr Houghton continued to monitor the possible causes of the increase in her water bill and, in late May 2012, advised her that he thought it was a leak coming from the pool. After asking him to make inquiries of "Kev's Pools" as to whether she could empty the pool, Dr Flanagan recalled the following subsequent exchange with Mr Houghton: "Houghton: I have spoken to Kev's Pools and we can empty it. It is a 'commercial' type pool. They are designed to be emptied, serviced and refilled on a regular basis. [Flanagan]: Okay Robin. I know it was emptied before and I have seen the council empty swimming pools every year. I just don't have time to deal with this at the moment, given everything I am dealing with. Can we empty the pool? Houghton: Leave it with me. I will organise it."
3. In relation to Dr Flanagan's statement to Mr Houghton that she did not have time to deal with the pool given everything she was dealing with, her evidence was that during 2012 she was "under some stress and was time poor due to": "(a) my divorce from [Dr Oldfield] and attempting to reach a financial settlement in the Family Court; (b) [Dr Oldfield's] inability to refinance in accordance with our Heads of Agreement, NAB were threatening to foreclose on the property which would leave my children and I with no home; (c) having to work and take care of my patients; and (d) being a support to my children with [Dr Oldfield]."
4. Mr Houghton arranged for the pool to be emptied. The pool was then left empty, without Dr Flanagan having the cause of the leak investigated or repaired.
5. On 9 January 2013, Dr Oldfield sent Dr Flanagan an email about the pool, referring to it having been emptied and stating: "You will be aware of the danger in this with cracking due to outside compression and of its 'popping up' as the underlying clay dries out. We wil (sic) both remember Aden Foy's place in King St and the problems he had. I will hold you respnsible (sic) for any such damage it if is not at least half re-filled'" as such damage will reduce potential sale value…indeed being empty will also reduce value by detracting from the houses' (sic) appearance' I would like you to take action on this now please'"
6. Dr Flanagan sent the email from Dr Oldfield to her family law barrister on the same day, stating, inter alia: "…I think the most interesting part of this email is that [Dr Oldfield] wants the value of the house to be increased by spending our own money when the beneficiary will be the bank. …I do not wish to be side tracked by [Dr Oldfield's] ranting but if I do need to respond these are the facts. I received a $2000 water bill a few months ago. I had the plumber out to find out where the leak was an (sic) it was by elimination deemed to be in the pool. I have drained the pool and the problem with the excess water resoled (scil resolved). The pool is 25m long and 5 lanes wide and holds 160,000 litres of water. It is not in my financial scope to find the actual leak, pay for the repairs and then refill the pool and pay for the water and chemicals required nor can I afford to run the pool cleaning system both from the electricity point of view and also I need a new cleaner which due to the size is a commercial quality and will cost about $5000.00. I did get advice about the pool popping and was told it won't. Also I don't want the pool fixed at this point as it will make the option of selling the house less attractive to the NAB. …"
7. When Dr Flanagan was asked in cross-examination about the advice to which she referred in this email, she confirmed that she was referring to what Mr Houghton had told her in the exchange I have extracted in [18] above. As to Dr Flanagan's further statement, that she did not want the pool fixed, the following exchange took place in the course of cross-examination: "Q. That was your position at the time. You didn't want it fixed. A. At that particular point the bank was thinking - was - was being difficult, and so my spending our cash on that didn't seem a priority to me. … Q. The concerns you had about the bank taking enforcement action that you held in 2012 continued through to early 2013? A. Yes, they did. Q. You thought it would be a good idea to leave the pool empty because it would discourage the bank from taking that enforcement action. Correct? A. That was a side issue, but it was also a - it was a thought, yes. Q. It was enough of a consideration for you to tell your barrister? A. Yes. Q. What you didn't list amongst your considerations was the fact that some pool incident that Dr Oldfield referred to concerned a smaller pool. Correct? A. It was - they were different pools altogether. I didn't mention that, no. Q. You didn't make any mention of the hydrostatic valves, did you? A. No. Q. You didn't take any steps to have the pool fixed and refilled because you didn't think you could afford it. Correct? A. It wasn't a priority at that point. Q. And because you didn't want it full because you thought it would help you in keeping the bank from the door. Correct? A. I didn't wish to spend my money to assist the bank."
8. The primary judge noted at [196] that the pool did not lift as a result of the clay drying out (as Dr Oldfield had foreshadowed) but, rather, after two days of heavy rain. As the pool and the building that enclosed it had been demolished before the proceedings were commenced and experts retained, the only report about the incident which involved an actual inspection of the pool and surrounds was prepared on behalf of Vero by Josh Mitchell, an engineer employed by Mountford Prider Pty Ltd. In his report, dated 2 September 2013, Mr Mitchell expressed the opinion that "based on the available evidence, the pool and partial enclosure failure has been caused by rainfall collecting under the pool shell following a significant rainfall event and the emptying of the pool for maintenance". Mr Mitchell considered that the pool became buoyant because the pressure of water collecting in the pool shell excavation below it exceeded the weight of the pool.
9. Dr Flanagan and the respondents both retained expert engineers in relation to the cause of the pool lifting in March 2013, Mr Scott Clemmett and Mr Tony Dockrill. They ultimately agreed that it was probable that hydrostatic valves were installed as part of the original construction of the pool. Mr Clemmett stated that the pool was constructed so as to rely on three hydrostatic valves "to relieve the hydrostatic groundwater pressures that could build up whenever the weight of the pool shell and the water contained within was less than the pressure exerted by groundwater within the surrounding soil". Mr Dockrill provided a similar explanation of the purpose of hydrostatic valves, stating: "If the pool is empty and the surrounding ground water table rises above the valve level, the one-way valve will open and allow the water to enter the pool, relieving the pressure built up under the pool. The water will continue to pour into the pool, through the valve until the height of equilibrium is achieved."
10. Mr Clemmett stated that he would have advised that the pool should only be emptied for maintenance and left empty "subject to confirmation that all three of the hydrostatic valves had been maintained in working order". Mr Dockrill expressed the opinion that "a hydrostatic relief valve is defective if it does not allow ground water to flow into a swimming pool and relieve hydrostatic pressure from under the pool". In the Joint Report, dated 25 June 2001, Mr Dockrill agreed with the following scenario posed by Mr Clemmett: "If hydrostatic relief valves were installed as referenced in the correspondence from Bel-Air Pools, then it is probable that the valves were defective and did not perform as expected at the time of the pool failure."
11. In the course of their (brief) concurrent evidence, Mr Clemmett and Mr Dockrill agreed that the pool could be left empty provided the valves were "functioning correctly". They also agreed that if the pool was full of water, that would have acted as a means of keeping the pool anchored in the event the hydrostatic valves were defective and not operating. As to how full the pool would have to be in order to fulfil that anchoring function in that circumstance, Mr Dockrill stated (and Mr Clemmett agreed): "I did do some rough numbers previously in regard to it, and - and it - the pool would need to be approximately half full. In the event that the hydrostatic relief valve was faulty and did not function correctly, the pool would need to be about half full to prevent it from moving."
12. The primary judge accepted that the pool was designed and built to withstand hydrostatic groundwater pressure when empty, and that there was no risk of the pool lifting out of the ground due to that pressure provided the valves were functioning correctly: [215]. On the expert evidence, at the time the pool lifted there was an un-investigated risk that the leak identified by Mr Houghton was due to a defective hydrostatic valve, as a result of which the pool could not be left empty and should have been half-filled to ensure it did not lift: [222]. This is what Dr Oldfield had raised with Dr Flanagan in January 2013, but Dr Flanagan took no further advice "even though she had not emptied, serviced and refilled the pool, but had instead left it empty, by then for over 6 months, without investigating the cause of the leak": [221].
13. An issue arises on the cross-appeal as to the proper reading of a paragraph of the primary judge's reasons on this issue. At [198], her Honour stated (emphasis added): "None of the evidence established that Dr Flanagan gave false evidence about how the pool came to be left empty or that she believed there to be a risk in doing so and simply acted recklessly. To the contrary, she acted deliberately, for reasons which seemed good to her. That, in all of the circumstances she should have understood that in doing so she was running a considerable risk which would have affected her rights, had she been fully insured as she undoubtedly thought she was, was also in issue, to which I will return."
14. Relying on the emphasised passage, Dr Flanagan submitted that the primary judge had made a positive finding that recklessness had not been made out. The respondents dispute this characterisation. I will return to this below.

The allegations of negligence

1. The duty of care that Dr Flanagan alleged that the respondents owed was to exercise due care and skill in the provision of insurance broking services: SOC [26]. The Statement of Claim alleged that Mr Bernasconi and Nadic, as his employer at the relevant time, breached that duty by failing to take any of the "Reasonable Precautions" (SOC [28]), which were defined to include: "(a) consider the scope of cover and any and all exclusions provided by the terms of any Homeowner's Insurance policy which they proposed to recommend or refer to [Dr Flanagan]; … (e) point out to [Dr Flanagan] any particular terms or exclusions in any proposed home building and contents insurance policy that might be of concern or relevance to her circumstances; (f) in particular, point out to [Dr Flanagan] any policy terms or exclusions in any proposed Homeowner's Insurance policy which might have the effect of denying insurance cover to [Dr Flanagan] for damage to any of the structures on the Property, including the Pool and the Pool Enclosure; (g) advise [Dr Flanagan] of the availability in the insurance marketplace of Homeowner's Insurance policies that did not have terms or exclusions that she was or might be concerned about or not content with; …"
2. The respondents initially denied the allegation of breach of duty. However, at the conclusion of Mr Bernasconi's cross-examination and following consultation with their insurance expert, Mr Gribbin, Senior Counsel for the respondents conceded on instructions that a reasonably competent broker who held Mr Bernasconi's view as to the significance of the pool exclusion in the Vero policy would have drawn it to the attention of Dr Flanagan. The primary judge was satisfied that the breach of duty on which Dr Flanagan's case rested was established: [188], [237]-[244].
3. The central issue was whether the admitted breach of duty caused the pleaded loss and damage. The particulars of the allegation of causation (SOC [43]) included the following (emphasis added): "(a) but for the breaches [Dr Flanagan] would have sought Homeowner's Insurance cover which covered damage to the Pool and the Pool Enclosure, including the Pool Damage; (b) had [Dr Flanagan] sought insurance as described in (a) above she would have successfully taken out such insurance cover, and on terms sufficient to cover the total cost of repair or replacement of the Pool Damage; (c) the Pool Damage would have been covered by the Homeowner's Insurance which [Dr Flanagan] would have taken out, subject only to the deduction of an excess; (d) [Dr Flanagan] would have claimed on her Homeowner's Insurance policy and the claim in relation to the Pool Damage would have been accepted; (e) the Pool Damage would have been either repaired by the relevant Homeowner's Insurance or that insurer would have paid [Dr Flanagan] a sum of money sufficient to enable her to fully repair the Pool Damage; (f) in either of the events as set out in (e) above, the Pool Damage would have been repaired at no cost to [Dr Flanagan], (other than an excess), by no later than twelve months after the Pool Damage occurred; …"
4. In responding to SOC [43], the respondents' Amended Defence included the following: "The [respondents] deny paragraph 43 of the Statement of Claim and say that: a. [Dr Flanagan] will be required to prove that she could have obtained homeowner's insurance cover: i. at a premium that was no more expensive than the 2011/2012 premium for the CGU Policy at the time of renewal for the 2012/2013 year; and ii. on terms sufficient to cover, and which would have covered, the total cost of repair or replacement of the alleged Pool Damage; iii. and further, in circumstances where: 1. [Dr Flanagan] deliberately emptied the water from the Pool and failed to refill the Pool for several months prior to the alleged Pool Damage occurring; and 2. there were significant defects in the structure and/or design of the Pool and/or the Pool Enclosure."
5. As the primary judge stated at [255], the respondents' case was that Dr Flanagan's pleaded claim was an "all or nothing case which required proof not only that she would have taken out another policy, but that her claim would have been accepted by that insurer". In circumstances where Dr Flanagan had not led evidence that she would have stayed with CGU if made aware of the Vero exclusion, or taken out a policy with another insurer, the respondents submitted that she had not discharged the burden on causation: [257].
6. Dr Flanagan, on the other hand, argued that while she bore the onus of proving that she would have successfully taken out cover for the risk which the Vero policy did not cover, she did not have to prove that she would have taken out a particular policy, or what the terms and conditions of that policy would have been: [259]-[260]. Rather, the onus fell on the defendants to establish that other available policies would not have covered her loss, given the two particular features relied on (being the "defect in an item" and "reasonable precautions" clauses): [261]. It followed, as the primary judge summarised her submission at [262]: "Accordingly, unless 100% satisfied that Dr Flanagan's claim would have been not only refused by a notional insurer, but even in a contest over that in court, involving matters of interpretation of an insurance policy, she would certainly have failed either on the defect point or the reasonable precautions point on which the defence relied, then Dr Flanagan's case had to succeed. That was not established on the evidence."
7. Applying "common sense criteria" as described in Unity Insurance Brokers Pty Ltd v Rocco Pezzano Pty Ltd (1998) 192 CLR 603; [1998] HCA 38, the primary judge was satisfied on the balance of probabilities that the evidence established that Dr Flanagan suffered harm as a result of Mr Bernasconi's failures (at [270]): "No longer having the full insurance cover which she previously had, the claim which she made under the Vero policy was rejected because of the result of the operation of the exclusion, of which she was not aware and which the CGU and like policies did not contain."
8. Her Honour found that but for Mr Bernasconi's negligence, Dr Flanagan would not have taken out the Vero policy, and instead would have taken out insurance which did not contain an exclusion of the kind which the Vero policy contained. Further, Dr Flanagan would likely "have accepted [Mr Bernasconi's] views about the cover CGU offered" (being that the CGU policy was his preferred policy): [271], [283], [288]. Her Honour had earlier rejected the respondents' contention that given Dr Flanagan's financial circumstances in 2012, she did not have the means to pay the premium for the more expensive CGU policy: [113]; see also [290]. The primary judge thus concluded that "[n]ot only was the negligence causally connected with the breach of duty established, but it resulted in a loss of the kind in the contemplation of a reasonable person in Mr Bernasconi's position": [291].
9. The primary judge acknowledged, however, that it was also necessary to consider "the question of whether Dr Flanagan could have established her claim, had she had full cover under a policy such as that which CGU offered, given what is in issue in relation to the defect and reasonable precautions provisions": [292]. The respondents had argued that given what she understood at the time, Dr Flanagan's actions in emptying the pool and leaving it empty showed that she would not have acted differently if she had taken out the CGU policy: [295]. Dr Flanagan, on the other hand, submitted that the respondents had not met the onus of establishing this, or of establishing any defect in circumstances where the valves could not be inspected: [294].
10. Her Honour concluded that the hydrostatic valves had suffered a defect within the terms of the CGU policy: [300]. Her Honour stated at [302]-[303]: "Given the way in which the word is used in the CGU policy, I consider that it includes a mechanism such as a valve, which does not suffer from faulty workmanship or design, has no structural defect and has not been operated incorrectly, but which has stopped functioning as it was designed to do. In the case of a hydrostatic valve, the question of operation does not arise, given how it is designed to function in a pool. I am satisfied, on the balance of probabilities, that the valve did not function as it was designed to do, because it had developed a defect, with the result the lifting of the pool, because hydrostatic pressure could not be relieved."
11. Her Honour considered that as there was no problem with the pool leaking until late 2011, it followed "that the valve, during all that time, had no defect and was operating as it was designed: one way, to allow water to enter the pool": [304]. The pool developed a leak in 2011, which was identified in 2012, but its cause was not investigated after the pool was emptied: [304]. On the basis of the expert evidence, her Honour was satisfied that if the hydrostatic relief valve did not function correctly, the pool could not safely be left empty, and it was probable that at the time of the pool failure "the valves were defective and did not perform as expected": [310], [312]. Her Honour concluded that if the valve had not had "that defect" (namely, "allowing water to leak out"), the pool would not have lifted as it did and the pool and pool structure would not have been damaged. Accordingly, her Honour rejected Dr Flanagan's case that the CGU policy or a like policy did not exclude the damage: [315].
12. Her Honour was also satisfied on the evidence that Dr Flanagan "failed to take the reasonable precautions which the CGU policy and those which offered like cover, on the expert evidence, also required": [316]. After considering a number of authorities on the meaning of reasonable precautions in the context of an insurance policy, her Honour stated that "[d]etermining the scope of such obligations requires consideration be given to the purpose of the particular policy and whether a risk has been courted by recklessness": [326]. Although her Honour accepted that Dr Flanagan "believed" there was no problem with leaving the pool empty, her Honour did not accept the submission that the expert evidence supported the correctness of that understanding: [328]. Her Honour stated at [329]-[330]: "It must be accepted that in all the circumstances, having decided not to have the cause of the leak identified and repaired or the pool refilled, it would have been reasonable for Dr Flanagan to have taken the precaution of obtaining advice about whether the pool could still safely be left empty for an extended period. On the expert evidence, if she had done so, she would have been advised either to establish the cause of the leak and have the valve repaired if it had a defect, as the pool later lifting established it did, or to half fill the pool, given the risk of lifting which existed if the valve was not functioning properly to deal with increased hydrostatic pressure while it was empty."
13. Her Honour concluded that if Dr Flanagan had cover under the CGU policy or another policy which contained a term requiring the taking of reasonable precautions, "it could not have been established that she did take such precautions in order to prevent the risk of the pool lifting while it was empty from materialising": [334]. It followed that "the obligation falling on the insurer under the policy in respect of the damage to the pool and pool structure would not have been engaged": [335].

Issues on the appeal

1. Counsel for Dr Flanagan identified her grounds of appeal as challenging the following two conclusions of the primary judge: 1. her Honour's conclusion that cover would have been denied on the basis of the clause excluding cover by reason of a "defect in an item"; and 2. her Honour's conclusion that cover would have been denied on the basis that Dr Flanagan did not comply with the clause requiring her to take reasonable precautions to prevent a claim under the policy.
2. The first issue turns on the proper construction of the phrase "defect in an item" in the expired CGU policy, that being the policy that the parties and the primary judge relied upon as containing the terms of a notional policy (Ground 5), and whether the clause applied with respect to the hydrostatic valves installed in the pool (Grounds 2, 3).
3. The second issue involves two sub-issues, the first of which is a challenge to her Honour's finding that the leak in the pool, which prompted Dr Flanagan to empty it, should have alerted Dr Flanagan to the fact that the hydrostatic valves were not functioning correctly (Grounds 7-8). The second sub-issue is whether Dr Flanagan's omission to take any steps to investigate the leak involved a failure to take reasonable precautions to avoid a claim (Ground 9).

"Defect in an item"

1. I have set out the wording of the CGU policy above at [13]. Dr Flanagan embraced the primary judge's finding at [303], that the "valve did not function as it was designed to do, because it had developed a defect" which resulted in the lifting of the pool, and submitted that it did not follow from that finding that there was a "defect in an item" within the meaning of the exclusion clause in the CGU policy. That clause referred to a "defect in an item" (emphasis added) which, in her submission, required a "lack or absence of something essential to make the hydrostatic valve one of proper design and construction". What her Honour had found was instead a "defect in the condition of the item" (emphasis added), in which the product, which had once functioned properly, stopped functioning as it was designed to do. Dr Flanagan relied in this respect on Zurich Australian Insurance Limited v CSR Limited (2001) 52 NSWLR 193; [2001] NSWCA 261 ("Zurich Australian Insurance"), submitting that where used in the CGU policy, the context and surrounding words supported the conclusion that "defect" was to be construed narrowly to refer to issues of physical failure rather than a failure in performance.
2. Dr Flanagan submitted that construing "defect in an item" so as to encompass ordinary wear and tear would be repugnant to the commercial purpose of the contract. It would also be inconsistent with the terms that followed it in the clause ("faulty workmanship, structural defects or faulty design") which related to matters that would affect an item from the outset. Dr Flanagan relied in this respect on the principle of construction that in a list of words, a word of uncertain scope may take its character from those surrounding it, if they have a recognisable characteristic: Australian Aviation Underwriting Pty Ltd v Henry (1988) 12 NSWLR 121 at 126.
3. On the construction of the phrase for which she contended, Dr Flanagan submitted that the evidence did not demonstrate that any issue (if proven) with the hydrostatic valve was a "defect" of the kind for which coverage was excluded under the CGU policy (or another like policy). She submitted that the experts' conclusion that the valves were defective was because they were not performing at the time of the pool failure, though her counsel accepted that that proposition was not explicitly put to the experts. It followed, in Dr Flanagan's submission, that any defect in the hydrostatic valves could only be described as one of condition, and not a defect existing in the item from the outset which would have triggered the exclusion clause. In circumstances where the onus was on the respondents to show that the exclusion would have been triggered in the circumstances of the case, Dr Flanagan contended that the respondents did not discharge that onus.
4. In support of her submissions on onus, Dr Flanagan relied on Wallaby Grip Ltd v QBE Insurance (Australia) Ltd (2010) 240 CLR 444; [2010] HCA 9. In addressing the burden of proof in the context of a contract of insurance, the Court (French CJ, Gummow, Hayne, Heydon and Kiefel JJ) stated at [36] (citations omitted): "In The Torenia Hobhouse J observed that the 'legal burden of proof arises from the principle: [h]e who alleges must prove' and that the 'incidence of the legal burden of proof can therefore be tested by answering the question: [w]hat does each party need to allege?', by reference to the contract of insurance. In the present case it was necessary for Mrs Stewart to establish that a contract of insurance under the Act was in existence at the relevant time and that Pilkington was liable to her husband for his injuries. The first was admitted, the second was established by evidence. It followed that the claim was within the terms of the cover provided and the insurer's obligation arose. QBE had to do more than decline to admit that Pilkington was entitled to an indemnity greater than the statutory minimum, a matter which, in any event, had not been raised in the further amended statement of claim. It was required to establish what limit, if any, had been placed upon its liability to indemnify. It did not do so."
5. Dr Flanagan's reliance on this passage was misplaced. As the respondents submitted, this was not a case in which an insured was seeking an indemnity under an insurance contract. Rather, Dr Flanagan's cause of action was one in negligence, in which she bore the onus of establishing that, but for Mr Bernasconi's breach of duty, she would have taken out a policy that responded to her claim, and on which the hypothetical insurer would have paid out. Dr Flanagan thus needed to deal with all aspects of the hypothetical policy, including by establishing that the relevant exclusion clauses would not apply. I do not consider that the primary judge reached a contrary conclusion in [261]. Read in context, that paragraph formed part of her Honour's summary of Dr Flanagan's submissions, and was not a commentary on the merits thereof.
6. In any event, the question of onus is not determinative of the proper construction of "defect in an item", as it appears in the CGU contract of insurance. As with any written contract, the task for the Court is to determine the intention of the parties as expressed in the words in which they have recorded their agreement: HDI Global Specialty SE v Wonkana No 3 Pty Ltd (2020) 104 NSWLR 634; [2020] NSWCA 296 ("Wonkana") at [18]. That task is to be approached objectively, ascertaining the meaning of words "by reference to what a reasonable person would have understood the language of the contract to convey": at [19], citing Toll (FCGT) Pty Ltd v Alphapharm Pty Ltd (2004) 219 CLR 165; [2004] HCA 52 at [40]; Electricity Generation Corporation v Woodside Energy Ltd (2014) 251 CLR 640; [2014] HCA 7 at [35].
7. In Topfelt Pty Ltd v State Bank of New South Wales Ltd (1993) 47 FCR 226 at 237, Lockhart J described a defect by reference to the ordinary usage of that word, as meaning "a lack or absence of something necessary or essential for completeness; a shortcoming or deficiency; an imperfection". The ordinary usage of the word defect must also be considered in context as part of the phrase, "a defect in an item".
8. The Court of Appeal in Zurich Australian Insurance was concerned with a similar phrase, albeit in the context of a statute. Section 3 of the Motor Accidents Act 1988 (NSW) defined "injury" as personal or bodily injury caused by the fault of the owner or driver of a motor vehicle in its use or operation, but only if the injury was caused during, relevantly, "such use or operation by a defect in a vehicle". Spigelman CJ observed that a vehicle is not defective only because its operation in a particular manner may lead to injury, but "the manner in which it is intended to operate may determine whether there is a 'defect' 'in' the vehicle" (emphasis in original): at [68] (Mason P and Handley JA agreeing).
9. In the present case, the engineering experts explained the intended function of a hydrostatic valve installed in the pool. They agreed that the pool could have remained empty without it lifting out of the ground if the hydrostatic valve was functioning as intended. As her Honour accepted, the experts were agreed that at least one hydrostatic valve had a defect and had stopped functioning as it was supposed to: [310]. The lifting of the pool thus demonstrated the presence of "a defect in an item" (emphasis added).
10. Counsel for Dr Flanagan submitted that it did not follow from the fact that an item was not functioning as one would expect that there was a defect in the item. So much may be accepted, but it does not assist in the present case, where the item in question, a hydrostatic valve, was designed to perform a specific function and, on the opinion of the experts, it did not perform that function when the pool was emptied. As the respondents submitted, in circumstances where the valves were not available for inspection, it was open to the experts to conclude, based on their experience, that the valves were probably defective within the meaning of that term as used in the CGU policy. It was equally open to her Honour to rely on that expert opinion, notwithstanding Dr Flanagan's dispute as to the particular aspect of the valve's componentry that was defective.
11. There is also no requirement in the clause for the defect to be present in the item from the outset. Dr Flanagan sought support for an implied limitation to this effect on the basis of the other items listed in the composite phrase that followed the term "defect in an item" in the CGU policy, being "faulty workmanship, structural defects or faulty design". However, those same items appear in an earlier exclusion with an express temporal limitation, excluding accidental damage or accidental loss caused by "water entering the building because of a structural defect, faulty design or faulty workmanship when the buildings were constructed" (emphasis added). The absence of an equivalent qualification tells against its implication in the exclusion with which this ground is concerned.
12. Further, and in any event, there was no evidence to support a change in the condition or functioning of the valves at some time after their installation, such that the defect in the item only arose at a later point. There was force in the respondents' submission in this regard, that as the pool was full for a significant period of time "one would simply not know, during the whole of that period, whether such a defect existed". While the pool was full, the weight of the water would have counterbalanced any competing hydrostatic pressure. Even in circumstances where the pool was emptied in mid-2012, it does not follow from the fact that the pool did not lift before March 2013 that the defect was not then present. This ground of appeal should be dismissed.

The requirement to take reasonable precautions to avoid a claim

1. I have set out above the terms of the CGU policy that required Dr Flanagan to take reasonable precautions to prevent anything which could result in a claim under the policy. It was common ground that compliance with this clause was a condition precedent to cover.
2. Dr Flanagan contended that central to the primary judge's finding that she had not complied with this clause was her Honour's conflation of the leak (which was why Dr Flanagan emptied the pool), and the cause of the pool lifting once it was empty (Ground 7). She contended that from this premise, the primary judge erred in concluding that Dr Flanagan had failed to take reasonable precautions, in not taking the leak as a warning sign that the hydrostatic function of the valve had also failed (Ground 8). Finally, she submitted that the primary judge erred in finding that the exclusion clause in the policy was engaged, given the absence of any finding that Dr Flanagan's failure to take precautions was reckless, or that she knew the risk was the subject of insurance cover (Ground 9).
3. The parties accepted the general applicability of authorities which have considered similarly worded, albeit not identical, reasonable precautions clauses in policies of insurance. Both relied on what Diplock LJ stated in Fraser v B N Furman (Productions) Ltd [1967] 3 All ER 57 ("Fraser"), that a requirement for reasonable precautions to be taken in a policy of insurance cannot require an insured to take measures to avert dangers which the insured does not foresee (at 61): "What in my judgment is reasonable as between the insured and the insurer, without being repugnant to the commercial purpose of the contract, is that the insured, where he does recognise a danger, should not deliberately court it by taking measures which he himself knows are inadequate to avert it. In other words, it is not enough that the employer's omission to take any particular precautions to avoid accidents should be negligent; it must be at least reckless, i.e., made with actual recognition by the insured himself that a danger exists, not caring whether or not it is averted."
4. In Legal & General Insurance Australia Ltd v Eather (1986) 6 NSWLR 390 at 397 ("Eather"), Kirby P stated that a reasonable precautions clause is breached where the insured "acted recklessly, indifferent to the normal precautions that would be taken". As to what constitutes "normal precautions", his Honour observed at 398 that this depended on the circumstances, but did state: "The greater the value of the property at risk of loss, the greater will be the obligation to take stringent precautions. The greater the foreseeable risk of a loss occurring in the circumstances, the greater will be the obligation to take precautions. The greater the possibility of precautions being taken, the more readily will a court infer that they ought to have been taken."
5. In the same case, McHugh JA observed, in a passage discussing Fraser (among other cases), that a condition of this type in a policy of insurance meant that "the insured must be concerned to protect the property from loss or damage and must take such steps to protect the property as he thinks are reasonable having regard to dangers which he recognises" (at 407). In Vero Insurance Ltd v Power Technologies Pty Ltd [2007] NSWCA 226, Beazley JA (Campbell JA and Harrison J agreeing) described this and other passages from Eather as emphasising the necessity to establish actual knowledge of the danger, stating at [63] that: "Otherwise, the prerequisite for the operation of the exclusion, that with actual knowledge of the danger, the insured failed to take adequate measures or knowing that a danger existed, did not care whether the danger was averted, does not exist." [Emphasis in original].
6. In applying those principles to the present case, Dr Flanagan submitted that a valve may not have functioned as expected either because it was leaking and allowing water to escape from the pool (described as a "leaking failure"), or because it did not permit water to enter the pool to achieve hydrostatic equilibrium (described as a "hydrostatic failure"). The primary judge, however, wrongly proceeded on the premise that there was a link between those two failures. Dr Flanagan submitted that this error was evident in [304] of the primary judge's reasons, where her Honour referred to the valve operating as it was designed to do until late 2011 because "there was no problem with the pool leaking as had been the case when the valve had to be replaced before 2001" (at [304]) and further, at [314]: "Instead of operating one way to allow water into the pool to equalise that pressure, the valve was allowing water to leak out. The result of the defect eventually permitted the pool to lift, with resulting damage to the pool…"
7. Dr Flanagan submitted that the expert evidence did not support the finding that because the pool was not leaking before 2011, the valves were operating as designed. Further, and critically, it did not follow from the fact that there was a leak, which Dr Flanagan did not investigate, that the pool could not be left safely empty and was at risk of lifting. So to find, in Dr Flanagan's submission, conflated the leaking failure (which is what caused her to empty the pool) with the hydrostatic failure (of which she was not aware). It followed, Dr Flanagan submitted, that the primary judge erred in finding that upon discovering the leak in the pool, she courted the risk of the pool lifting by failing to take the reasonable precaution of investigating whether the hydrostatic function of the valve had failed and whether the pool could safely be left empty.
8. Dr Flanagan relied in this context on [198] of the reasons as constituting a finding that she had not acted recklessly. Instead, she believed there was no problem with leaving the pool empty, did not know of the risk of the pool lifting if it was left empty and the hydrostatic valves were not letting water in, and did not flagrantly court that risk or simply ignore it.
9. Dr Flanagan submitted that it was relevant in this respect that while Dr Oldfield had warned her of the danger of the pool cracking or lifting out of the ground due to an absence of groundwater, the damage was in fact caused by a relative excess of groundwater and associated hydrostatic pressure. The danger was not the risk of the pool popping up due to the ground beneath drying out, but rather the risk of the pool popping up because of the hydrostatic pressure beneath it. The fact that the valves were leaking did not indicate that they were also not letting water in. Further, and in any event, if she had known about the relevant danger and was required to seek advice (so as to take reasonable precautions), she would have been advised that there was no danger of the pool lifting out of the ground if it had three hydrostatic valves which were operating correctly.
10. The primary judge considered the scope of Dr Flanagan's obligation to take reasonable precautions by reference to the general observations of Kirby P in Eather, to which I have referred above, summarising the circumstances at [327]. As her Honour there noted, the subject of the claim was "an unusually large pool and pool enclosure for a home", which was built at considerable expense. Dr Flanagan and Dr Oldfield had only paid for the pool after it was emptied several times and a faulty valve, which was involved in a leak, was replaced. When Dr Flanagan emptied the pool in 2012, it was on the basis of advice that there was a leak. She then left the pool empty without investigating the cause of the leak (not wanting to spend money which might prove to be only to the benefit of the bank) in circumstances where the only advice on emptying the pool that she had received was that it was designed to be "emptied, serviced and refilled".
11. It is of significance in this context that in January 2013, Dr Oldfield raised a concern about Dr Flanagan leaving the pool empty for fear it would "pop up", and advised her to half-fill the pool. Dr Oldfield may have been wrong as to what would ultimately cause that risk to materialise, but his email undoubtedly put Dr Flanagan on notice of the risk. As the respondents submitted, the fact that Dr Oldfield identified but one mechanism by which the pool could lift if left empty was immaterial.
12. The primary judge stated that it could be accepted "that Dr Flanagan believed that there was no problem with leaving the pool empty as she did": [328]. The difficulty that Dr Flanagan had to confront was that she was on notice of a leak in the pool. When it leaked upon construction, it transpired that the problem was with the hydrostatic valves. In 2012 she was advised that she could empty the pool to investigate the leak, but on the basis that it should be emptied, serviced and refilled. Instead, Dr Flanagan left the pool empty, with an un-investigated leak.
13. As the respondents submitted, Dr Flanagan was aware of the risk of the pool lifting and had various means, including "obvious, simple and cost-free" options, to address it, including to make inquiries as to the risk and possible solutions, half-filling the pool, or replacing the valves. Instead, she acted with indifference to that risk, deciding not to address it for her own (financial) reasons. The primary judge was correct to conclude that such measures were not difficult, nor were they costly or beyond Dr Flanagan's means: [331]-[332].
14. Contrary to the manner in which Dr Flanagan sought to construe [198] of the primary judge's reasons, which appears much earlier in the reasons and not in the specific part dealing with the reasonable precautions clause, that paragraph did not contain a positive finding that Dr Flanagan was not reckless. All her Honour there did was find that the evidence did not establish that Dr Flanagan had given false evidence either about how the pool came to be empty "or that she believed there to be a risk in doing so and simply acted recklessly". As I have just noted, her Honour accepted that Dr Flanagan so believed, but as her Honour foreshadowed in the next sentence of [198], the issue was whether "in all of the circumstances, she should have understood that in doing so she was running a considerable risk which would have affected her rights, had she been fully insured as she undoubtedly thought she was". In light of this conclusion, which is consistent with the respondents' characterisation of the paragraph, it is unnecessary to consider Ground 2 of the respondents' cross-appeal.
15. It should be apparent from the above that Ground 7, which sought to contend that the primary judge's decision rested on incorrectly conflating the existence of the leak with a failure of one or more of the hydrostatic valves, does not relevantly assist Dr Flanagan. Consistently with her evidence, the primary judge accepted that the pool was emptied because of a leak. In circumstances where Dr Flanagan did not investigate its cause and instead left the pool empty (inconsistently with the only advice she received), Dr Flanagan courted the risk, of which she was aware, of the pool "popping up". Had she made appropriate inquiries, she would have been told the pool could only be left empty if the hydrostatic valves were functioning correctly. Based on past history, in the face of a pool leak, the valves could not be relied upon for any of the functions they were required to perform absent investigation. The primary judge was correct so to conclude.

Conclusion

1. The appeal should be dismissed. As I noted in the introductory section, given my conclusion on the appeal, it is unnecessary to determine the issues on the cross-appeal, and it should be dismissed.
2. I propose the following orders: 1. The appeal is dismissed. 2. The cross-appeal is dismissed. 3. The appellant/cross-respondent is to pay the cross-appellants/ respondents' costs of the appeal and cross-appeal.
3. Simpson AJA: I agree with Mitchelmore JA.

DISCLAIMER - Every effort has been made to comply with suppression orders or statutory provisions prohibiting publication that may apply to this judgment or decision. The onus remains on any person using material in the judgment or decision to ensure that the intended use of that material does not breach any such order or provision. Further enquiries may be directed to the Registry of the Court or Tribunal in which it was generated. Decision last updated: 04 July 2023