JARDEN AND JARDEN v LUMLEY GENERAL INSURANCE (NZ) LIMITED [2016] NZCA 193

JARDEN AND JARDEN v LUMLEY GENERAL INSURANCE (NZ) LIMITED [2016] NZCA 193 [11 May 2016]IN THE COURT OF APPEAL OF NEW ZEALANDCA401/2015[2016] NZCA 193BETWEEN DAVID JOHN JARDEN AND JOANNEJARDENAppellantsAND LUMLEY GENERAL INSURANCE (NZ)LIMITEDRespondentHearing: 21 April 2016Court: Randerson, Stevens and Wild JJCounsel: K M Quinn for AppellantsP J L Hunt and K J Rowe for RespondentJudgment: 11 May 2016 at 3:00 pmJUDGMENT OF THE COURTA The application by the appellants to adduce further evidence on appeal is granted.B The appeal is allowed in part.C The judgment in the High Court is set aside only to the extent that it determined that the amount paid to the appellants by the Earthquake Commission for cover under the Earthquake Commission Act 1993 was $179,163. The correct figure was $123,850.D The appeal is otherwise dismissed.E There is no order for costs on the application to adduce further evidence.F The appellants must pay the respondent costs for a standard appeal on a band A basis with usual disbursements.G If costs in the High Court cannot be agreed, they are to be fixed by that Court.____________________________________________________________________REASONS OF THE COURT(Given by Randerson J)Table of ContentsPara NoIntroductionThe terms of the Lumley policyThe EQ cover issuesWhat was the amount of cover actually provided by EQC forearthquake damage?Under what circumstances does Lumley's liability for top-upcover arise?Was the Judge wrong to reject the Jardens' claim that specificitems of damage were natural disaster damage within themeaning of the Lumley policy?Damage caused by leaks from the roof of the house andgarageInternal walls of the house being out of plumbVoidsWindows and doors jamming and out of squareDamage to the garageResultFormal orders[1][6][12][13][17][29][33][53][65][71][73][76][77]Introduction[1] The appellants Mr and Mrs Jarden live on a lifestyle property just north of Rolleston. They built their house in 1998 but it suffered damage in the major Christchurch earthquakes on 4 September 2010 and 22 February 2011. Mr and Mrs Jarden had a residential insurance policy with the respondent (Lumley). This provided top-up cover for natural disaster damage in excess of the cover provided under ss 18 and 20 of the Earthquake Commission Act 1993 (defined in the policy as EQ Cover).[2] Despite numerous expert reports since the earthquakes, the claims brought by the Jardens for the damage suffered were not resolved. They issued proceedings in the High Court against the Earthquake Commission (EQC) and Lumley. Shortly before the hearing in the High Court, the Jardens reached a settlement with EQC, theterms of which we discuss below. The Jardens' claim therefore proceeded solelyagainst Lumley. After a hearing occupying some six days, Kós J made declarations to the following effect:1(a) The Jardens' house and garage were damaged in the September 2010 and February 2011 earthquake events in some but not all of the respects claimed.2(b) The damage was to be apportioned as between the two earthquakes on the basis identified in the judgment.3(c) Reinstatement repairs were required to restore the house and garage to the standard set out in the judgment.4(d) Lumley's liability to pay the Jardens for the repair costs was to ariseonce costs have been incurred and to the extent the costs exceeded the total payment made to the Jardens by EQC of $179,163. This payment was to be apportioned for each event on the basis of 90 per cent for the damage caused by the September 2010 earthquake and 10 per cent for the damage caused by the February 2011 earthquake.[3] Kós J reserved leave to apply on two issues: the remediation strategy for repairs to the lounge doors and windows, and the quantum of any liability onLumley's part in excess of the EQ Cover.[4] On appeal, the Jardens assert that the High Court erred in:1 Jarden v Lumley General Insurance (NZ) Ltd [2015] NZHC 1427 at [137].2 The claims accepted were identified at [107] of the High Court judgment.3 At [112]–[113].4 At [118]–[121].(a) Determining the amount of EQ Cover and when Lumley's liability fortop-up cover commences.(b) Rejecting their claim that specific items of damage amounted to natural disaster damage within the meaning of the policy, in particular:(i) Damage caused by leaks from the roof of the house and garage.(ii) Internal walls of the house being out of plumb.(iii) Cracks in the concrete slab on which the house and garage were constructed.(iv) Voids beneath the concrete slab for the house and garage.(v) Windows and doors jamming and out of square.(vi) Damage to the garage.[5] During the hearing before us, Mr Hunt for Lumley confirmed that, without any admission of liability, Lumley was prepared to repair the cracks in the concrete slab for the house and garage. Mr Quinn for the Jardens confirmed there was no dispute about the method by which this would be achieved. This disposes of (iii) above.The terms of the Lumley policy[6] It is necessary to set out briefly the relevant terms of the Lumley policy, particularly in order to determine the issue relating to the circumstances in which Lumley becomes liable for the cost of repairs in excess of the EQ Cover.[7] The house and garage are insured for full replacement value based on the areaof the buildings. The policy provides cover for "sudden, accidental loss to yourhome during the period of insurance". Loss caused by natural disasters is excludedexcept to the extent covered by a "natural disaster benefit". The natural disasterbenefit:will cover any loss under those Sections during the period of insurance that occurs: as the direct result of natural disaster, or as a direct result of measures taken under proper authority to avoid the spread, or otherwise mitigate the consequences, of a natural disaster.It does not include any loss for which compensation is payable under any Act of Parliament other than the Earthquake Commission Act 1993.[8] As to when Lumley's liability commences, the critical clause provides:For the home and contents insured under Sections 1 or 2, we will pay the difference between the EQ Cover and the cover under this Policy, on the condition that we do not pay:- for any excess which applies to the EQ Cover, and- until the Earthquake Commission has paid or agreed to pay the EQ cover[9] EQ Cover means: the insurance of the home provided by Section 18 and/or the insurance of the contents provided by Section 20 of the Earthquake Commission Act 1993.[10] The standard of repair if the insured elects to repair or rebuild is expressed as follows:We pay the costs actually incurred to repair or rebuild it to substantially the same condition and extent as when it was new, or at our option the cashequivalent, subject to the following [11] The Judge summarised the terms of the Lumley policy:5[18] In short, the policy provides top-up cover over and above EQC'sliability. It will not be payable until EQC has agreed to make its payment. Cover is per event, so damage attributable to each seismic event is accounted for separately. Nothing is payable unless: (1) EQC has agreed to pay the insured; and (2) there is an insurance liability in excess of EQ cover. If the5 High Court judgment, above n 1.insured has made the election to repair or rebuild, payments are madeagainst costs incurred unless the insurer elects otherwise.The EQ Cover issues[12] There are two matters to be resolved in respect of the EQ Cover:(a) What was the amount of cover actually provided by EQC for earthquake damage?(b) In what circumstances does Lumley's liability to provide top-up cover commence?What was the amount of cover actually provided by EQC for earthquake damage?[13] As to the first issue, the High Court Judge found that the amount of the cover provided by EQC was the total amount for which the Jardens settled with EQC:$179,163. The Judge declined to accept the Jardens' submission that the coverprovided was only $123,850 and that the difference between the two figuresrepresented legal costs of $55,313 paid by EQC to the Jardens' solicitors. The Judgewas not satisfied there was sufficient proof that the amount paid was split in the way the Jardens argued, despite the fact that the settlement agreement produced to theCourt recorded that $123,850 was paid to the Jardens' mortgagee and $55,313 was paid to the Jardens' solicitors.[14] On appeal, the Jardens applied to adduce further evidence: a memorandum signed in the High Court by counsel for the Jardens and the EQC recording the terms of settlement along with affidavits by lawyers representing the parties to the settlement. Despite opposition from Lumley, we allowed the application. The memorandum did not require leave since it forms part of the Court record. The affidavits simply confirm the contents of the memorandum and explain that the additional sum of $55,313 was for costs. We are satisfied that justice requires that the application to adduce further evidence be allowed.[15] We are satisfied that if the Judge had the full information we now have, he would have found that the amount of the cover provided by EQC to the Jardens was:4 September 2010 event $100,000.00GST $ 15,000.00$115,000.00Less excess $ 1,150.00 $113,850.0022 February 2011 event $ 10,000.00Total $123,850.00[16] The additional sum of $55,313 was for the Jardens' legal costs associatedwith the court proceedings. The amount EQC paid for the first event was the full amount of the cover available under s 18 of the Earthquake Commission Act 1993 (the ECA). The amount paid for the second event was below cap and reflects the relatively minor damage caused by that event fixed on the basis of 10 per cent of the maximum cover.Under what circumstances does Lumley's liability for top-up cover arise?[17] The parties agreed that Lumley's liability was the difference between theactual cost of repair and the EQ Cover provided by s 18 of the ECA. The parties differed as to how EQ Cover was to be interpreted for the purposes of assessingLumley's liability. The Jardens argued that Lumley's liability arises as soon as thetotal cost of repairs for natural disaster damage exceeds $123,850 (being the amount of the EQC settlement). To the contrary, Lumley submitted it was not bound by the settlement the Jardens agreed with EQC.[18] In support of its position, Lumley submitted that it was not a party to the settlement; EQC settled without admission of liability; the amount of the settlement could not be taken as prima facie reasonable as the Jardens submitted; the final cost of repairs was not yet known; the amount of EQ Cover as defined in the policy and whether it is under cap will depend on the apportionment of the cost of repairs between the two earthquake events; the court should determine the amount of EQCover and the monetary effects of apportionment between the two events unless all parties agree.[19] The High Court judgment on this point is not entirely clear but it appears to be consistent with Lumley's interpretation.6[20] Mr Quinn submitted that Lumley's principal insuring obligation arose fromits agreement to pay the costs actually incurred to repair the property. The natural disaster damage section of the policy was merely a refinement of the principal cover. The EQ Cover should be interpreted as being equivalent to whatever amount EQC agreed to pay or was ordered to pay under s 18 of the ECA. Lumley would be protected against the possibility of an insured deliberately settling with EQC for under-value by common law obligations such as the duty to act in good faith. Finally, Mr Quinn drew our attention to a proviso to Lumley's obligation to pay for natural disaster damage. To understand the context, we set out Lumley's obligation and the proviso:You receive the following:(a) For the home and contents insured under Sections 1 or 2, we will pay the difference between the EQ Cover and the cover under this Policy, on the condition that we do not pay:- for any excess which applies to the EQ Cover, and- until the Earthquake Commission has paid or has agreed to pay the EQ CoverProvided that the combined total of any payments under (a) above: by EQ Cover, and by this benefit for each Sectionis no more than the sum insured shown in the schedule for that corresponding Section.[21] Mr Quinn submitted that the proviso and its reference to "by" EQ Covermeant that the amount of the EQ Cover was to be fixed by the amount actuallyreceived from EQC. He submitted there was nothing in the policy to supportLumley's argument that the extent of the cover had to be agreed by the parties(including Lumley) or fixed by the court.[22] We are not persuaded by Mr Quinn's submissions on this point. We aresatisfied that the primary insuring obligation in respect of natural disaster damage is to be found in the policy provision we have set out at [20] above. Lumley'sobligation is to pay the difference between the EQ Cover and the cover under the policy. That obligation is subject to two conditions: Lumley is not liable to pay for any excess applying to the EQ Cover and Lumley is not liable until EQC has paid or has agreed to pay the EQ Cover.[23] Relevantly, EQ Cover is defined in the Lumley policy as the insurance of the home provided by s 18 of the ECA.7 Section 18 provides:18 Residential buildings(1) Subject to any regulations made under this Act and to Schedule 3, where a person enters into a contract of fire insurance with an insurance company in respect of any residential building situated in New Zealand, the residential building shall, while that contract is in force, be deemed to be insured under this Act against natural disaster damage for its replacement value to the amount (exclusive of goods and services tax) which is the least of—(c) the amount arrived at by multiplying the number of dwellings in the building (being the number determined in accordance with subsection (3)) by $100,000 or such higher amount as may be fixed from time to time for the purposes of this paragraph by regulations made under this Act.[24] The obligation on EQC under s 18 of the ECA is to provide insurance against natural disaster damage for the replacement value of residential buildings up to the maximum sum of $100,000 (exclusive of GST). It is not in dispute that the cover is per event so that, as the Judge found, the damage attributable to each event is accounted for separately. The sum of $100,000 is commonly described as a "cap"signifying that it is a maximum sum. The actual amount payable depends on the7 Contents cover provided by s 20 of the Earthquake Commission Act 1993 (ECA) is not at issue.extent of the damage and the amount of the replacement value. Replacement value is defined in s 2 of the ECA. It is clear from that definition that EQC's obligation inrelation to a residential building is to pay the costs "reasonably incurred" in respect of replacing or reinstating the building to a condition "substantially the same as butnot better or more extensive than its condition when new, modified as necessary tocomply with any applicable laws".8[25] The definition of EQ Cover under the Lumley policy is consistent with thelanguage of s 18 by referring to "the insurance of the home provided by s 18". This is to be interpreted as EQC's actual statutory liability, which is not necessarily the same as the amount that EQC pays out to the insured. The proviso relied upon byMr Quinn does not assist the Jardens' argument. It simply provides that thecombined total of the payments under the EQ Cover and the benefits provided in the natural disaster damage section of the policy may not exceed the total sum shown in the schedule for damage relating to the items at issue.9 The proviso does not alter the meaning of the term EQ Cover as defined.[26] The fact that Lumley's liability does not arise until EQC "has paid or has agreed to pay the EQ Cover" does not mean Lumley is obliged to accept withoutquestion that the amount paid or agreed to be paid is the amount of the EQ Cover as defined by the policy. Its obligation is to top-up any excess above the EQ Cover as defined. In many cases, it will be obvious whether the cost of repairs amounts to natural disaster damage as that term is defined by s 2 of the ECA and whether the amount EQC is obliged to pay is under or over the cap provided by s 18. However, Lumley is entitled to be satisfied that the amount paid (or agreed to be paid) by EQCdoes equate with EQC's obligation under s 18 of the ECA. This may become an issue in cases where EQC's liability is at the cusp of the maximum payable or where it is unclear the damage is natural disaster damage. In practice, EQC and Lumley (or other insurers) work together in order to resolve their respective obligations. It is only if matters are not agreed that the extent of their respective obligations will need to be determined by the court.8 Subparagraph (a)(ii) of the definition of "replacement value" in s 2 of the ECA. The definitionincludes other costs not relevant for present purposes.9 It was not suggested in the Jardens' case that the total payments were likely to exceed the amount shown in the policy schedule.[27] Finally, we accept Lumley's argument that its cover and premiums are fixed on the basis of EQC's actual statutory obligation and that it is not open for the policyholder unilaterally to alter the basis upon which Lumley's liability arises under the policy. Nor is it a workable proposition that Lumley's protection against an ownersettling with EQC for under-value should depend on common law duties of good faith. The policy is to be applied according to its terms.[28] In the present case, the final repair costs are still not determined and Lumley has yet to be satisfied about the correct level of EQ Cover in terms of the insurance policy. The monetary effect of the apportionment of repair costs as between the two earthquake events is still unknown. Those matters will have to be determined by the High Court under the leave reserved if not agreed.Was the Judge wrong to reject the Jardens' claim that specific items of damage were natural disaster damage within the meaning of the Lumley policy?[29] Mr Quinn submitted that the appeal was by way of rehearing and the Court was obliged to reach its own conclusions on the evidence.10 He accepted however that, in respect of factual findings, the conventional position is that this Court would only interfere if the conclusion reached in the trial court was not open on the evidence or this Court is satisfied the trial judge was plainly wrong in the conclusion reached.11 Mr Quinn submitted further that since Kós J had made no adverse findings of credibility against the Jardens or their witnesses, this Court was as well placed as the High Court to assess whether the total body of evidence on any particular issue was sufficient to establish the Jardens' claims to the requiredstandard of proof.[30] Mr Hunt agreed that the conventional position for an appellate court in respect of factual findings is as stated by Mr Quinn. However, he submitted an appellate court should only interfere with factual findings if it can be shown that the trial Judge had failed to use or has palpably misused his advantage.12 Counsel also10 Austin, Nichols & Co Inc v Stichting Lodestar [2007] NZSC 103, [2008] 2 NZLR 141.11 Rae v International Insurance Brokers (Nelson Marlborough) Ltd [1998] 3 NZLR 190 (CA) at 197 and Moodie v Agricultural Ventures Ltd [1998] 3 NZLR 129 (CA) at 132.12 Hutton v Palmer [1990] 2 NZLR 260 (CA) at 268.referred to this Court's observation in Rae that this Court would not reverse a factual finding unless compelling grounds were shown for doing so.13[31] In a general appeal, the obligation of an appellate court is as stated by the Supreme Court in Austin, Nichols.14 The appellate court is obliged to arrive at its own assessment of the merits of the case.15 However, it is only if the appellate court considers that the appeal decision is wrong that it is justified in interfering with it.16The onus is on the appellant to show that the appeal decision is wrong.[32] We deal with each of the disputed items of damage in turn.Damage caused by leaks from the roof of the house and garage[33] Mr Jarden's evidence, supported by that of his wife, was that the house wasbuilt to a high standard in 1998. Apart from a minor leak soon after they moved into the house, there had been no problem with the roof until after the earthquakes. In the house, they began noticing leaks in the kitchen, dining room, hallway and two of the bedrooms. In the garage, there was water staining on the timber by the ridging in the centre. Water had been running out of the building paper.[34] On one particularly wet night, water began running down the light fittings inthe dining room with the result that two of the lights "went bang". Buckets wereneeded to collect the water. The Judge found that this incident appeared to have occurred around 9 May 2011, two to three months after the second earthquake. On that date, EQC had a note of a telephone call by Mr Jarden to this effect:claimant called to notify us that the roof is leaking. and it has been leaking for the last couple of days. it has been raining tremendously and now the roof is severely filled with water. it has caused an electrical hazard and it is dangerous for the roof will collapse.[35] The Judge recorded that EQC's contractor subsequently removed the rooffrom the south-west side of the house and replaced it, using silicon along the ridging13 Rae v International Insurance Brokers (Nelson Marlborough) Ltd, above n 11, at 198.14 Austin, Nichols & Co Inc v Stichting Lodestar, above n 10.15 At [5].16 At [4] and [13].and valleys to stop the leaking. Mr Jarden was unimpressed with the quality of the repair and confirmed in answer to questions from the Judge:The roof never leaked before the earthquake, we never had a problem with it and particularly in the southerly, if we get a southerly or quite a strong wind it leaks like a sieve.[36] The Jardens' evidence about the leaking roof of both the house and garage was not challenged in cross-examination. Lumley accepted the roof was leaking. Its defence to this claim was that any defects in the roof were unrelated to earthquakes. It was more likely due to defective construction or wind damage.[37] We will review the expert evidence on this subject shortly but we first recordthe Judge's conclusion on this issue:17[101] Considering the evidence as a whole, I do not consider that the Jardens have established on the balance of probabilities that the leaks they complain of were caused by the earthquake. There are essentially three reasons for my conclusion. The first is the period of time that passed between the most damaging earthquake, 4 September 2010, and 9 May 2011 which is evidently the date (or very close to the date) on which the major leak was identified. Secondly, there is the lack of evident damage to the more brittle ceiling linings. Thirdly, the balance of the expert evidence is strongly against a conclusion that the burden has been discharged by theJardens. Mr Williams did not inspect the roof. Mr Scarry's inspection wascursory. The more thorough examinations undertaken by Messrs Cook, Turner and Thompson are in this instance persuasive.[38] Assessing the evidence for ourselves, we consider the starting point must be the unchallenged evidence of Mr and Mrs Jarden that there were no leaks from the roof for more than 10 years after the house and garage were constructed in 1998 and that there were no leaks until after the earthquakes occurred. That was so despite the buildings being located in a high wind zone. Lumley did not contend that the roof was nearing the end of its life at the time of the earthquakes.[39] The only expert evidence called by the Jardens on the topic of the leaking roof was from a structural engineer, Mr Scarry. The Judge noted that Mr Scarry had not been up onto the roof himself and had confined himself to a brief inspection inside the roof for about half an hour. Mr Scarry could not reach a definite conclusion on the cause of the leaking other than to say it was not inconsistent with17 High Court judgment, above n 1.the earthquake damage. Under cross-examination, Mr Scarry was largely reliant onMr Jarden's evidence and the coincidence that the water ingress followed reasonably soon after the first earthquake. He accepted there was no obvious signs of damage to the ceiling.[40] Despite the absence of firm independent support from Mr Scarry, we considerthe Jardens' evidence that there had been no leaks over the 10 years prior to the earthquakes was sufficient to establish a prima facie case that the leaks were caused by the earthquakes. On this footing, the evidential onus moved to Lumley to show the leaks resulted from causes other than earthquake damage.[41] Lumley called a number of expert witnesses. A structural engineer, Mr Cook, inspected the property on 21 February 2014 and again on 10 November 2014. He inspected the roof and ceiling of the property on both occasions and did not see any evidence of earthquake damage to either. In Mr Cook's opinion, any leaking at theproperty did not result from damage sustained to the roof in the earthquakes. Rather, he considered wind was more likely to be the cause of roof damage.[42] Mr Cook's observations upon inspection of the roof were summarised in his evidence-in-chief:(a) The trusses are level. If the roof had sustained serious damage in the earthquakes sufficient to cause leaking, the trusses would not be level and/or would be wet;(b) The roofing contractor used nails to fix the roof to the purlins. Today, screws would be used as they allow less movement, and limit any gaps between the roof, nail and purlin. The nails which were used can expand and contract in the heat hereby allowing the potential for water to penetrate into any gaps;(c) A number of the nails had missed the purlins;(d) There were large amounts of sealant that had been applied in the areas around the nails. Sealant is a temporary measure only with more permanent repairs needed. If the roof had leaked as a result of the nails not being installed correctly or bending, and sealant was applied, then the sealant would only be expected to last a short time;(e) The plastic header tank located above the living area was filled with some water and had a lid on it. If it was leaking to the extent claimed by the Jardens as a result of the earthquake, I would expect to see water staining surrounding the tank area. I did not observe any staining.[43] Mr Cook went on to give reasons for his conclusions as to the cause of the leaks. These were that steel roofing including fixings and flashings is able to sustain relatively large strains without damage; the more brittle and fragile ceiling linings which must have sustained similar strains to the roof cladding were undamaged; bracing construction showed that the building itself was designed for exposed and high wind loading; the bracing in the structure was 3.5 times the earthquake loading, suggesting that wind was more likely to be the cause of roof damage; while it was possible snow might have caused some problems it was unlikely the loads exceeded the design loading; the building had a lightweight roof and well-designed timber trusses; with a low seismic mass supported on well-braced walls and solid foundation the roof was unlikely to have sustained damage by earthquake; and the level of bracing and stiffness of the walls suggested they did not flex in the earthquake.[44] Mr Cook added that if it was determined that the roof had leaked as a result of earthquakes then he would recommend the valley iron be replaced and the nails be replaced with screws. It was not necessary to replace the entire roof.[45] Mr Cook maintained his view under cross-examination and expanded on it in answer to questions from the Court. Mr Cook was asked to provide the basis for his conclusion that wind was more likely to be the cause of damage. He repeated the reasons he had given in his evidence-in-chief and added:A. I think it's very unlikely the roof has been damaged by theearthquake, however wind will lift iron, will pop nails, will actually rip iron in portions off roof, off the roof and inside the roof cavity some of the hold-down nails have missed the purlins, sorry, therafters within there and so there's areas where there's potential forwind to get under and lift the roof off. There's also an area of, you know, and lift nails. There's also an area of the edge of the flashing, the sort of soft aluminium that forms down that looks like it's beenlifted up, either physically by a person or possibly wind has managed to catch it and lift it up off the roof as well and so in terms of the wind, I know in Canterbury last year there was some fairly significant wind events as well. Wind is a – you know, is probablymore likely to cause damage to the roof, looking at the, the design and expectations of the building performance than an earthquake.Q. That's because there are loose elements, or loosenable elements inthe roof –A. Yes.Q. That the wind can play with?A. That's right, that's correct.[46] In response to a further question from the Court as to whether an earthquake could also play with loose elements in the roof Mr Cook responded:A. Well, in the roof, it does, but the roof, the loading is shared betweenthe ceiling and the roof and it's – the timber is so flexible and strong, it can tolerate large, large stains and movement in that roof withoutany damage, so it's, it's possible but the – in general the condition of the roof and the position of the nails – there's not a lot of slottedholes in the roof around the nails. The movement suggests to me that the earthquake has not caused significant damage to the roof.[47] Mr Cook's conclusion that the leaks were not due to earthquake damage wassupported by two experienced builders called by Lumley. Mr Turner inspected the property on 16 January 2014 and 10 November 2014. He spent a substantial amount of time inspecting the roof and ceiling at the property including the garage. He did not see any evidence to support the contention that earthquakes had caused the roof to leak. In particular, there was no stretching around the nail holes which would have constituted evidence that the roof had undergone movement. Rather, there was evidence of poor quality workmanship in the roof installation.[48] Mr Turner said the Jardens' claim to EQC following the heavy rain inMay 2011 was consistent with the roof leaking as a result of poor workmanship and the age of the roof rather than earthquake damage. If the roof leaked only during heavy rainfall, this supported the conclusion that water was penetrating around the nails that had not been installed correctly or that the valley iron was unable to take the amount of water. Mr Turner did not locate any significant water damage or evidence of any water having recently entered ceiling space. If the house had moved significantly during the earthquakes, he considered the water header tank in theceiling space would have spilled over. There was no evidence of staining that might have supported that conclusion.[49] In cross-examination, it was put to Mr Turner that the evidence from the Jardens was that the roof did not leak before the earthquakes but did so after them. Mr Turner's response was to say he was unable to find the source of the leak or thearea that was leaking. He said that, ideally, the source of the leak should be identified before any repairs were carried out. However, there was no directchallenge to Mr Turner's conclusion that the likely cause of the leaks was faultyworkmanship.[50] The second builder called by Lumley was Mr Thompson who inspected the house on three occasions between 25 May 2012 and 10 November 2014. He made a thorough examination of the roof of the property including assessing the fixings, flashings and construction of the roof coverings. In his opinion, the defects to the roof that he observed were unrelated to any earthquakes. Rather, he considered they were caused by defective construction methods. These included raised and insufficiently installed nails; insufficient flashings and junctions; silicon building up to valley trays and nail penetrations to roofing sheets.[51] Expanding on this by reference to photographs, Mr Thompson said the nails to the ridge flashing had been installed incorrectly and had caused tearing to some of the roof coverings around the point where the nail penetrates. Many of the nails had moved and remained raised. In his opinion, this was the result of expansion and contraction of the metal coverings causing the nails to stay proud. In that event, water can track down the tread of the nails and into the roof cavity. Wind-blown water could also track into the house underneath the ridge and valley flashings. He only observed very minor water staining around the roofing nails. If there had been leaking to the extent described in the Jardens' evidence, he would have expected to have seen substantial water staining. Mr Thompson's evidence was not challengedon this topic in cross-examination.[52] Our independent review of the evidence in relation to the cause of the leaks is sufficient to rebut the prima facie conclusion that the leaks were caused byearthquake damage to the roof as we have discussed at [40] above. The evidence establishes that the damage to the roof and the resulting leaks is more likely to have been caused by wind, by faulty construction of the roof, or a combination of the two. The expert evidence called by Lumley is convincing and was not effectively countered by the Jardens. The appellants have failed to persuade us that the Judge'sconclusion on this point was wrong.Internal walls of the house being out of plumb[53] The Judge heard a great deal of expert and technical evidence on the vertical alignment of the internal walls of the property and ultimately rejected the Jardens'claim that this had resulted from the earthquake. It is perhaps surprising that so much evidence was devoted to this topic given that:(a) unlike the roof issue, there was no evidence that the internal walls were perfectly aligned before the earthquakes;(b) it was not suggested there was any structural damage to the internal walls;(c) any superficial damage to the internal walls such as cracks had been repaired by the appellants; and(d) the Judge accepted that any "leaning" of the walls was not visible to thenaked eye.[54] The Judge nevertheless undertook a detailed examination of this issue. Mr Cowie, an engineering surveyor called by the Jardens, had undertaken measurements of the walls of the house, using a vertical laser plumb. His measurements were not disputed by Lumley. They showed that a number of internal walls in the house had leans of between 7 and 18 mm, generally measured over a vertical height of 2.2 m.18 The New Zealand standard at the time allowed a tolerance of 8 mm over 2.4 m. As the Judge observed, the maximum variation from true18 Exceptionally, some of the vertical heights were 1.2 m, 1.9 m and 2.4 m.(18 mm over 2.4 m) varied from the standard by less than half a per cent (0.42 per cent).[55] Kós J accepted that the September 2010 earthquake event was very severe and that there could be little doubt that the internal walls of their house were severely shaken. The issue was what damage resulted. The Judge accepted Mr Cook'sevidence that superstructure performance in an earthquake is directly correlated to foundation performance.19 He observed that one difficulty confronting the Jardens on this issue was that the original pre-earthquake measurements were unknown. Mr Jarden's evidence was limited to an inference to be drawn from his assertion thatthe windows and doors of the house were in alignment prior to the earthquake. After reviewing the evidence in that respect, the Judge concluded it was difficult to make anything in particular of the evidence of the doors jamming and he placed little weight on it.[56] The Judge also reviewed Mr Jarden's evidence about the repairs to internaldamage carried out by Mr Jarden, his son and a builder friend of his. This includedpatching up wall cracks and nail holes as well as planing doors. Mr Jarden'sevidence was that the cracks were visible, affecting the paint finish, but not with a measurable gap. The Judge recorded that Mrs Jarden described the cracks as "quite fine".[57] The Judge found as a fact that such cracks that appeared in the fabric of the internal walls and ceiling must have been fine, probably less than 2–3 mm in width. Kós J took into account several factors, including: that an EQC inspector who hadvisited the property in December 2010 had only noted "door jamming" in the familyroom; the absence of any other reference to the damage Mr Jarden referred to in his evidence; the absence of photographs taken of the pre-repair state of the walls; and the fact that no evidence was called from Mr Jarden junior or his builder friend.[58] The Judge then carefully reviewed the evidence of Mr Cook and Mr Davey for Lumley and Mr Scarry for the Jardens. In essence, the Judge accepted19 Based in part on Duncan Henderson "The Performance of House Foundations in the Canterbury Earthquakes" (ME (Civil) Thesis, University of Canterbury, 2013).Mr Cook's evidence, supported by that of Mr Davey, that the absence of significantcracks in the internal walls and ceilings indicated that major realignments of theinterior walls had not occurred. Mr Davey's evidence included discussion about theextent of displacement of the house needed in order to cause the internal walls to be displaced by up to 18 mm from vertical. He said that displacement of the house would need to be in the 30–40 mm range to achieve that outcome. Yet the GeoNetrecorders showed that the maximum displacement the Jardens' house would havesustained in the September 2010 earthquake was less than 15 mm. He said this wasa long way short of Mr Scarry's estimate that a displacement of the house by 25 mmwould have been necessary and much less than his own estimate of 30–40 mm. Mr Davey also supported his evidence on displacement effects by reference to a 1993 Building Research Association of New Zealand report.20[59] As the Judge said:21[84] This expert analysis for Lumley was really not answered by the Jardens. Mr Scarry accepted that the repairs undertaken by Mr Jarden could be described as cosmetic. He suggested that gib linings could be remarkably resilient. He tended to place more weight on the doors jamming, but I am unable to share that view.[60] Kós J also considered in detail conflicting evidence between Mr Scarry and Mr Davey as to the significance of the internal walls leaning in different directions. Again, for reasons he canvassed in detail, he preferred the evidence of Mr Davey.[61] The Judge also took into account Mr Turner's evidence that, despite the issueby the local council of a code compliance certificate for the building, the practical reality was that building inspectors never checked the vertical alignment of the wallsagainst the standard. In 1998 when the house was built, Mr Turner's evidence wasthat builders did not use lasers to check vertical alignment. Whether a wall was plumb was normally determined by a combination of the naked eye and a spirit level. Mr Turner also disputed the evidence of Mr Williams, a builder called by the Jardens, that if the wall framing had leaned to the extent measured by Mr Cowie after the earthquakes, then, at the time of construction, the installation of the gib20 Building Research Association of New Zealand Report on Racking resistance of Long-sheathed Timber-framed Walls with Openings (Study Report No 54, Wellington, 1993).21 High Court judgment, above n 1.board sheets and the hanging of doors would have been impractical. The Judge'sconclusion after reviewing the evidence was expressed in these terms:22[90] To conclude, on the balance of the evidence I find it unlikely that the internal walls have moved from their original position to anything like thedegree that Mr Cowie's measurements might suggest. First, I cannot assumeon the evidence that the walls were true originally. Secondly, I cannot assume, either, that the internal walls were within the NZS 3604 8 mm standard. Thirdly, I am unable to find that any particular wall has moved from within to without standard as a result of either earthquake event.[62] Mr Quinn challenged the Judge's conclusion that the leans were not visible tothe naked eye. He drew our attention to an observation the Judge made during the evidence after a site visit. The Judge is recorded as saying that he thought he could see a lean in a wall near the fireplace. But the Judge is also recorded as saying he was not sure whether this was an optical illusion. Mr Quinn also referred to Mr Scarry's evidence that the wall tilt was readily apparent to the eye, "especially atthe corner of the storage room". Counsel also mentioned Mr Cowie's evidence of "observed" leans but we are satisfied that, in context, Mr Cowie was using thisexpression to refer to the leans he measured, not that he could see them with the naked eye.[63] There was remarkably little evidence on this particular topic. But we are satisfied that if leans in the internal walls had been readily apparent to the eye the Jardens themselves and the experts would have made something of the point. We would have expected that the point would have been put to Lumley's experts but itwas not. We conclude little, if any, attention was given to this point at trial and we do not give weight to it in our analysis.[64] The other points made by Mr Quinn effectively invited us to prefer the evidence of Mr Cowie and Mr Scarry to that of Lumley's experts. We have reviewedthe evidence ourselves. We are not persuaded that the Judge was wrong to reach the conclusion that he did on this issue. We are satisfied that there was sufficient evidence to support the Judge's findings.22 High Court judgment, above n 1.Voids[65] The concrete slab foundation for the house and garage was poured on tailings used as hard fill. The tailings comprise river run gravel or stones sized between 20 and 60 mm in diameter. Mr Scarry's evidence was that slabs using tailings in this way had in some situations been left with voids beneath the concrete as a result of the earthquake events. As the Judge noted, no direct evidence of these voids was produced. Rather, Mr Jarden and Mr Cowie said that they had detected hollow sounds when tapping the floor of the slab in some places. Neither had used ground-penetrating radar nor core holes in order to establish whether voids existed. Mr Cowie conceded that tapping the concrete slab was an imprecise method of determining the presence or absence of voids. He accepted it was good practice to use ground-penetrating radar and intrusive coring in order to establish whether slab voids existed. He also accepted that to rely solely on floor level variations and/or slab cracking was, in his experience, a very imprecise method to determine voids. The evidence was that it would only cost approximately $3,000 to test for voids using ground-penetrating radar.[66] Against this evidence, Lumley called Mr Cook who, as the Judge noted, was the only geotechnical engineer to give evidence. His view was that voids usually occurred beneath slabs on grade where there was liquefaction. It was not in disputethat liquefaction was not involved at the Jardens' property. Mr Cook's evidence was that he had not seen any houses with voids in geotechnical situations comparablewith those relating to the Jardens' property. Mr Cook's evidence was that it was "very unlikely, very, very unlikely" that there would have been any significantsettlement of the tailings from shaking caused by the earthquakes. The only conclusive way of determining the existence of voids would be to carry out the kind of testing already discussed. But, in his view, there is no justification for that.[67] Mr Cook's evidence was supported by Mr Davey who gave detailed evidencerebutting Mr Scarry's evidence that alterations in the floor levels and cracking in theconcrete slab indicated that voids had formed beneath.[68] The Judge found that the Jardens had not proved on the balance of probabilities that there were voids beneath the concrete slab nor, if there were such voids, that they were caused by the earthquakes. He found the evidence produced by the Jardens to be inadequate and unpersuasive. They had not shown that such damage was more likely than not. It was unclear why the good practice endorsed by Mr Cowie had not been undertaken. The Jardens had the obligation to put their best evidence forward and they had not done so.[69] Mr Quinn submitted that the Judge was wrong to reach the conclusion that the Jardens had not proved their case on this point. He submitted there was direct evidence from Mr Jarden and Mr Cowie about the existence of voids, relying on their observations of the hollow sounds identified when tapping on particular parts of the concrete slab. He also pointed to Mr Scarry's evidence about the likelihood ofthe existence of voids under a slab of this type following earthquake events and Mr Scarry's evidence about the significance of slopes in some parts of the concrete slab.[70] Having reviewed the relevant evidence, we are not persuaded the Judge'sfindings on this point are wrong. The hollow sounds produced by tapping wereacknowledged by the Jardens' own witness to be an imprecise method of identifyingthe existence or absence of voids in this context. The position could have been put beyond doubt by the simple step of using the ground-penetrating radar or coring suggested by Mr Cowie. That was not done. The Judge was entitled to accept the opinion of Mr Cook as a geotechnical engineer in preference to the view expressed by Mr Scarry.Windows and doors jamming and out of square[71] Mr Jarden gave evidence that all windows and doors in the house worked well before the earthquakes but were out of alignment after them. However, whengiving a detailed list of damage to the property, the Jardens' witness Mr Scarry didnot identify any issue with doors and windows other than those opening from the northern side of the lounge onto the patio. Those doors were jamming and windowsalongside them were out of square. Lumley's witnesses accepted that was so andthere is no issue in relation to those specific doors and windows. Both Mr Cook and Mr Turner gave evidence for Lumley that apart from the patio area, all other windows and doors opened and closed freely.[72] The Judge made no specific findings on this topic other than to note Lumley'sacceptance that the doors and windows in the patio area were damaged in the earthquake and should be repaired. No doubt the Judge had in mind that there wassimply no expert evidence to support Mr Jarden's evidence that other doors andwindows were affected.23 We are not persuaded there was any justification for a finding on this topic beyond that made by the Judge for the windows and doors in the patio area.Damage to the garage[73] In relation to the garage, Kós J found:24[106] Finally, there is the garage. Mr Scarry's observation of damageincludes the walls being out of plumb, distortion to the garage door frames, significant sections of the brick veneer being loose, and cracks in the garage concrete slab floor. Forensically, the garage is the house in miniature. The evidence is, however, even less adequate. Consistent with my findings in relation to the house (and accepted indeed by Lumley) I find that the earthquakes damaged the brick veneer exterior wall. That item alone within the garage has been established on the balance of probabilities to have been caused by the earthquakes.[74] Mr Quinn accepted the Judge's finding that the method of construction of the damage and the earthquake issues relating to the garage were effectively identical to those relating to the house. Mr Quinn submitted there was one respect in which the construction of the garage differed from the house construction: the garage had no ceiling and the roof was therefore exposed from the inside. He submitted there was unchallenged evidence that the garage roof was leaking internally and that leaks were evident on the roof trusses and other structural elements. Mr Quinn referred usin that respect to Mr Jarden's evidence.23 We note too that, when considering the issue of the internal walls, the Judge found it wasdifficult to make anything of Mr Jarden's evidence that some doors were jamming and placedlittle weight on that point: High Court judgment, above n 1, at [77].24 High Court judgment, above n 1.[75] Despite this minor difference, we are satisfied that no different considerations apply to the garage than those relating to the house. It follows for the reasons already given that no basis has been established to disturb the factual findings made in the High Court in relation to related issues affecting the garage.Result[76] The outcome is that the High Court judgment is upheld in all respects other than the finding about the amount the Earthquake Commission paid to the appellant for cover under the EQA. As noted in [28], the final repair costs are still not determined and the monetary effect of the apportionment of repair costs as between the two earthquake events is still unknown. As already indicated, those matters will have to be determined by the High Court under the leave reserved if not agreed.Formal orders[77] The application by the appellants to adduce further evidence on appeal is granted.[78] The appeal is allowed in part.[79] The judgment in the High Court is set aside only to the extent that it determined that amount paid to the appellants by the Earthquake Commission for cover under the Earthquake Commission Act 1993 was $179,163. The correct figure was $123,850.[80] The appeal is otherwise dismissed.[81] There is no order for costs on the application to adduce further evidence.[82] The appellants must pay the respondent costs for a standard appeal on a band A basis with usual disbursements.[83] If costs in the High Court cannot be agreed, they are to be fixed by that Court.Solicitors:Grant Shand, Christchurch for AppellantsMcElroys, Auckland for Respondent