Liberty Syndicate Management & Anor v Campagna Ltd & Anor

The Claimants in this action are the corporate and sole member of various Lloyd's syndicates and their managing agent, respectively, to whom I shall refer to as "Liberty". Between 2003 and 2008 Liberty underwrote a form of latent defects cover, mainly for residential property. This was known as the Premier Guarantee Scheme ("PGS").

The First and Second Defendants, companies registered in England (on 17 June 2002) and the Republic of Ireland (on 1 July 2005), respectively, provide technical audit services with respect to newly constructed residential property and, in particular, they provided services under the PGS. I shall refer to them as "Campagna”, unless it is necessary to distinguish between them.

The PGS was originally launched in the United Kingdom. In 2003 Liberty decided to extend the PGS to Ireland and Campagna was engaged to provide the technical audit services in respect of newly built residential properties in Ireland that were covered by the scheme. The PGS was administered for liberty by MD Insurance Services Ltd ("MDIS"), who also acted as the coverholder with a binding authority on behalf of Liberty to issue policies.

The scheme as it operated in England involved MDIS selecting suitable developers for admission to the scheme by a process of assessment, which took into account matters such as technical competence and experience. This did not happen in Ireland, where the only criteria for assessment were essentially financial.

In Ireland, once the financial check had been carried out, a policy was issued to the developer with an Initial Certificate. This provided cover to any potential home owner who had put down a deposit with the developer in the event that the developer became insolvent during the construction period. The second stage of the scheme came into effect after the house had been built and sold to the home owner. There were two main elements to the cover. First, 5 years of cover with respect to the waterproof envelope. This was known as the Defects Insurance Period. Second, 10 years of cover with respect to “Major Damage” to the property. This was known as the Structural Insurance Period (“Major Damage” was a defined term).

The two elements of the second stage would not come into effect until Campagna had issued a Certificate of Approval in respect of the property, which it would do once the house had been completed to its satisfaction. Subsequently, when the property was sold - which was often within a day or two, Liberty would issue the policy to the homeowner.

Unfortunately, the PGS in Ireland has been a disaster from Liberty's point of view. There have been numerous claims under the policies in respect of damage to or lack of water tightness of properties built in Ireland and inspected and approved by Campagna.

In this action Liberty has sued Campagna for breach of contract, contending that its checks and inspections of houses covered by the scheme fell below the standard of a competent technical auditor and that those breaches of contract have resulted in the excessive number of claims made under the scheme. Liberty also makes the same claims in tort, but I consider that this adds nothing.

This hearing is concerned only with various preliminary issues relating to the scope of the duty owed by Campagna in respect of the developments in Ireland and alleged breaches of it. These issues are being determined in the context of defects in various sample properties from the first three developments to be insured under the scheme (“the First Developments”). This hearing is not concerned with the extent of the remedial works or any issues of quantum or, in general, of causation. However, where there are issues of direct causation that are very closely linked to my primary findings I have addressed them. The list of issues agreed by the parties and directed by the court is attached as Appendix A to this judgment.

The real issue between the parties is the extent, if at all, to which a competent technical auditor in the position of Campagna should have discovered the various defects in the construction of the insured properties. In each of the first three developments some of the defects were widespread throughout the same development. However, the defects found in one development are not the same as those found in the others.

To give an idea of the scale of the PGS, I have been told that Campagna has inspected about 65,000 properties for Liberty under the scheme in Ireland and the United Kingdom, for which it has received over €5.7 million in fees.

There are very significant differences in the scheme as it was operated in the United Kingdom and the scheme as it was operated in Ireland. I have already mentioned the difference in the approach to the assessment of potential developers in the UK and Ireland. Two other important distinctions are, first, that the systems of regulatory control of residential building in the United Kingdom and in Ireland are very different and, second, that the cover in the UK is wider but permits recourse against the developer during the first two years (which was not the case in Ireland).

The PGS evolved out of an earlier scheme which had been set up by a Mr Gary Devaney. In 1997 Mr Devaney formed MDIS. Between 1997 and 2003 the PGS was underwritten by Trenwick International Limited. Liberty became the underwriters for the PGS in place of Trenwick at the end of 2002. Campagna was set up by a Mr Robert Williams, who used to be employed by Carillion Specialist Services Ltd. In that capacity Mr Williams had acted as a technical auditor on schemes managed by Mr Devaney.

In 2002 Mr Devaney decided that he was going to sever his business relationship with Carillion and so he entered into discussions with Mr Williams with a view to Mr Williams setting up his own company so that he could continue to carry out technical audits in Ireland for what had now become the PGS. As a result, Mr Williams set up Campagna (UK).

Before the PGS was extended to Ireland the only Irish building guarantee product was a scheme known as HomeBond. The HomeBond scheme was not popular because the home owner could only make a claim under the scheme if all remedies against the contractor had been exhausted, and the scheme was very inflexible. The decision to extend the PGS to Ireland, in competition with the HomeBond scheme, was taken at a meeting in Dublin in 2001 at the offices of the Irish insurance brokers, Coyle Hamilton. That meeting was attended by Mr Devaney and Mr Williams, representatives of the brokers and an underwriter from Trenwick.

The regulatory system of building control in Ireland has been described as less rigorous than that which operates in the United Kingdom. Properties under construction in Ireland are not inspected by any officer from a local authority. Instead, at the completion of construction a certificate is prepared by a professional, usually an engineer, who may also have been retained by the developer to check the quality of the work, who certifies that the property substantially complies with the relevant regulations.

It seems that the view was taken at this meeting that methods of construction in Ireland were relatively straightforward, traditional and robust and should not lead to an unusual number of claims under a properly administered scheme. Whether or not this perception was correct is not a matter that I have to decide, it simply serves to explain how the PGS came to be extended to Ireland in the manner that it did.

However, the fact that Liberty had no recourse against the developer had, in my view, two important consequences. First, it increased the risk to Liberty because the absence of a potential liability on the part of the developer for defects would not encourage developers to make sure that work was done properly, whereas its presence would have the reverse effect. The second consequence, which flowed from the first, was that the developer’s immunity from suit (as against Liberty) might be expected to limit the extent to which a technical auditor might be justified in trusting the builder or developer to carry out the work properly.

Another important factor which drove the decision to extend the PGS to Ireland was the very substantial property boom that took place in Ireland from around the late 1990s. Whilst this clearly created or increased the potential market for a scheme such as the PGS, it also increased the likelihood of less experienced builders entering the market or of existing builders employing labour that was inexperienced or insufficiently skilled.

The role of a technical auditor, which is usually carried out by a qualified building surveyor or building control inspector, is to inspect a property at various stages during its construction in order to confirm that it does not have any defects in design or workmanship that might give rise to a claim under the scheme or, to put it another way, that the building when complete represents a standard risk from an insurance point of view. The number or frequency of inspections which would be carried out in relation to any particular property, and hence the likely overall efficiency of the process in terms of detecting defects, was a matter of negotiation between the technical auditor and the administrator of the scheme. Obviously, the more frequent the inspections the less likely it is that a defect would go undetected. One aspect of the underwriting judgement, therefore, was to decide how frequently the technical auditor should visit a property. The more frequent the visits, the higher the cost to the developer who, under the PGS, paid for the costs of the technical auditor. The underwriter had to balance the risk of claims against the level of premium and the cost to the client - being the premium plus the technical auditor’s fee. If the cost was perceived to be too high, the cover would not be purchased.

Another aspect that emerged clearly from the evidence of Liberty's witnesses was that the technical auditor was expected to provide a service to the clients, namely the developers. As Mr Devaney said in evidence (Day 3/82):

“From our point of view, one of the key things - and that was one of the important parts of why we set up the technical services department and why we changed firms of surveyors along the line - was actually to do with service and, by that - if somebody carried out an inspection, the house sale would depend on us being to issue the certificate of insurance. Now, provided that inspection was satisfactory for instance, then we would - we wanted to be able to receive the certificate of approval within a certain number of days otherwise the person may lose the sale. And the same with inspections and everything else. It was actually - the last thing a developer wanted was to be told somebody was coming to site or ask for somebody to come to site to look at a key stage and then they didn’t turn up. So it was important for managing the service provided to the developer.”

And, as Mr Gray, an underwriter at Liberty, put it in his witness statement, "Developers were happy to sign up for the PGS because they received high quality input from knowledgeable surveyors".Mr Byrne, who was then the Technical Services Manager at MDIS, gave evidence to broadly similar effect (at Day 4/24). MDIS was understandably concerned that if the technical auditors were tardy or inefficient or carried out their duties in an officious manner, developers would complain and that this would give the PGS a bad name. For that reason MDIS used to undertake site inspections of particular projects at random when they would check progress against the reports that they had on file and ask site managers of their views about the performance of the technical auditors (Mr Byrne, Day 4/14).

It is self evident that an officious technical auditor who was constantly seeking to delay work in order to make inspections or who would ask for completed work to be opened up so that it could be checked would not go down well with developers. I consider that it is implicit in Liberty’s wish that technical auditors should not give the scheme a bad name that it must be taken to have accepted that, to a reasonable extent, technical auditors would from time to time have to take a contractor's work on trust if they had no reason to doubt the quality of the work that was being carried out on a particular site.

Following the formation of Campagna (UK) in June 2002 it was appointed by MDIS as technical auditor for properties in the north of England and in Ireland. This required Campagna to carry out surveys of certain housing units to establish whether they represented a standard risk to Liberty and whether Liberty should provide certificates of insurance to the home owners under the PGS. Two other companies were appointed as technical auditors to cover other parts of the United Kingdom.

The documents by which Campagna were first retained by Liberty have not survived. However, by a letter dated 12 June 2003 Liberty wrote to Campagna in the following terms:

“LATENT DEFECTS INSURANCE - TECHNICAL AUDITS

"Contract"

Any contract where Campagna Ltd have been appointed by Syndicate 192 act as Latent Defects Technical Auditor.

"Standard"

The standard of skill and care expected of a reasonably competent Latent Defects Technical Auditor.

This letter is to confirm that Campagna Ltd agree to carry out their obligations in respect of any Contract to the Standard. Syndicate 190 agree that Campagna Ltd will be under no liability to Syndicate 190 whatsoever (whether in contract, tort (including negligence) or otherwise) for any injury or death (except where injury or death is caused by Campagna Ltd's negligence), damage or direct, indirect or consequential loss so long as the Standard is met."

On 22 November 2004 MDIS, acting on behalf of Liberty, entered into a formal agreement with Campagna which set out in much greater detail Campagna's obligations under the PGS. I shall refer to this agreement as the “2004 Retainer”.

The 2004 Retainer included the following provisions:

“1.

Definitions

1.1.3

"Certificate of Approval" means the certificate issued by the Site Audit Surveyor [Campagna] to Liberty confirming a satisfactory Survey of a Housing Unit

1.1.7

"Final Certificate" means the certificate issued by Liberty to signify acceptance of each Housing Unit for insurance following the issue of the Certificate of Approval by [Campagna]

1.1.9

"New Development" means a Housing Unit or a group of Housing Units located at the site noted on the Initial Certificate and for which an individual Final Certificate is issued for each Housing Unit

1.1.10

"Housing Unit" means the property described in the Final Certificate

1.1.12

"Technical Manual" means the functional requirements and performance standards issued by the Scheme Administrator and in force at the time when technical information relating to the New Development is first submitted to [Campagna]

1.1.16

"Survey" means checks and/or inspections carried out by [Campagna] at a Housing Unit

3.

Survey

3.1

[Campagna] will carry out a Survey of a property as advised by the Scheme Administrator [MDIS] in order to establish whether the property has been designed and/or constructed in a manner to enable the Premier Guarantee policy to be issued by Liberty.

3.2

In carrying out its Survey, [Campagna] shall:

3.2.1

Comply with and adhere to the Council of Mortgage Lenders’ Logic Test where applicable, a copy of which can be found at Appendix 1.

3.2.2

Comply with the technical audit procedures / standards for each type of property, which are provided by the Technical Services Department of the Scheme Administrator from time to time.

3.2.3

Comply with any Professional Code of Practice, statute, bye law, regulation, and/or such regulatory document which relates to [Campagna’s] business and duties under this Agreement.

3.2.4

Ensure that any surveyor who carries out a Survey has at least five years’ post qualification experience and is a member of the Association of Building Engineers or Royal Institution of Chartered Surveyors unless agreed otherwise with the Scheme Administrator.

3.3

Without prejudice to the above, [Campagna] shall use all reasonable skill and care to be expected of a competent surveyor when carrying out each Survey.

3.4

[Campagna] shall immediately notify the Scheme Administrator if [Campagna] is, or is about to, breach any of the conditions of Clause 3.2 and/or 3.3.”

By the time that the 2004 Retainer had been entered into, Campagna had already been auditing developments in Ireland for Liberty for over 18 months, including many of the properties in the First Developments, the subject of this hearing. These were Tayleur's Point, Mountgorry Wood and Derrycorris Drive.

During this period, in October 2003, MDIS issued a Site Audit Surveyors’ Guidance Manual to Campagna (it should be noted that this was not a Technical Manual as defined in the 2004 Retainer). It was preceded by a draft manual which had been issued in September. Section 5 of the manual was headed "Ireland". The introduction to Section 5 said this:

In the same section of the manual the action to be taken by technical auditors was described in the following terms:

The summary of cover for Ireland was described in the manual in the following terms:

‘Major Damage’ was defined as damage which requires a ‘defect’ in the design or workmanship of the ‘Structure’ which causes ‘damage’ to the New Housing Unit. The Structure included the foundations, load bearing parts of floors, staircases and associated guard rails, walls and roofs, roof coverings, external finishing surfaces (including rendering) necessary for the watertightness of the external envelope and floor decking and screeds.

The wording of the policy changed from time to time during the period whilst the First Developments were being constructed and there were subtle differences in the definitions of the insured perils. However, in broad terms the summary of cover in the manual was in my judgment sufficient to enable Campagna's technical auditors to know what they should be looking for during both the 5 year Defects Insurance Period and the 10 year Structural Insurance Period.

As to the First Developments themselves:

Although many of the issues are confined to defects in the sample properties, the sample properties cannot be considered in isolation because they are part of larger developments which were the subject of Campagna's inspections. Whilst Campagna was required to inspect and treat each housing unit individually, what happened in practice was that when visiting a development Campagna's technical auditors would look at and check progress on properties other than those scheduled for particular inspection. Thus when evaluating Campagna's discharge of its duties in relation to the sample properties, it may be necessary to consider what information was or should have been available to Campagna's technical auditors as a result of what they saw or should have seen at non-sample properties.

Both parties helpfully provided succinct and well presented submissions supported by authorities. However, I have to say that this is one of those cases which turns very largely on its facts and, for my part, whilst I find that the authorities are helpful in providing general guidance as to the correct approach, they do not really provide any assistance in relation to the particular issues that I have to decide.

In relation to the duty of inspection, Campagna relied on statements in two cases involving the duties of supervising architects. In the first, East Ham Corporation v Bernard Sunley & Sons Ltd [1966] AC 406 at page 443, Lord Upjohn said:

“As is well known, the architect is not permanently on the site but appears at intervals, it may be of a week or a fortnight, and he has, of course, to inspect the progress of the work. When he arrives on the site there may be very many important matters with which he has to deal: the work maybe getting behind-hand through labour troubles; some of the suppliers of materials or the sub-contractors may be lagging; there may be physical trouble on the site itself, such as, finding an unexpected amount of underground water. All these are matters which may call for important decisions by the architect. He may in such circumstances think that he knows the builder sufficiently well and can rely upon him to carry out a good job; that it is more important that he should deal with urgent matters on the site than that he should make a minute inspection on the site to see that the builder is complying with the specifications laid down by him.... It by no means follows that, in failing to discover a defect which a reasonable examination would have disclosed, in fact the architect was necessarily thereby in breach of his duty to the building owner so as to be liable in an action for negligence. It may well be that the omission of the architect to find the defect was due to no more than error of judgment, or was a deliberately calculated risk which, in all the circumstances of the case, was reasonable and proper.”

[Campagna’s emphasis]

The second, which was also relied on by Liberty, was a decision of HH Judge Coulson QC (as he then was) in McGlinn v Waltham Contractors [2007] EWHC 149 (TCC), where he said at [218]:

“(a)

The frequency and duration of inspections should be tailored to the nature of the works going on at the site from time to time: see Corfield v Grant 29 Con LR 58 and Jackson & Powell, at para 8–240. Thus it seems to me that it is not enough for the inspecting professional religiously to carry out an inspection of the work either before or after the fortnightly or monthly site meetings, and not otherwise. The dates of such site meetings may well have been arranged some time in advance, without any reference to the particular elements of work being progressed on site at the time. Moreover, if inspections are confined to the fortnightly or monthly site meetings, the contractor will know that, at all other times, his work will effectively remain safe from inspection.

(b)

Depending on the importance of the particular element or stage of the works, the inspecting professional can instruct the contractor not to cover up the relevant elements of the work until they have been inspected: see Florida Hotels Pty Ltd v Mayo (1965) 113 CLR 588 and Jackson & Powell, at para 8–241. However, it seems to me that such a situation would be unlikely to arise in most cases because, if the inspecting officer is carrying out inspections which are tailored to the nature of the works proceeding on site at any particular time, he will have timed his inspections in such a manner as to avoid affecting the progress of those works.

(c)

The mere fact that defective work is carried out and covered up between inspections will not, therefore, automatically amount to a defence to an alleged failure on the part of the architect to carry out proper inspections; that will depend on a variety of matters, including the inspecting officer's reasonable contemplation of what was being carried out on site at the time, the importance of the element of work in question, and the confidence that the architect may have in the contractor's overall competence: see Sutcliffe v Chippendale & Edmondson 18 BLR 149 and Jackson & Powell, para 8–242.

(d)

If the element of the work is important because it is going to be repeated throughout one significant part of the building, such as the construction of a proprietary product or the achievement of a particular standard of finish to one element of the work common to every room, then the inspecting professional should ensure that he has seen that element of the work in the early course of construction/assembly so as to form a view as to the contractor's ability to carry out that particular task: see the George Fischer Holdings case 61 Con LR 85. That accords with Mr Jowett's evidence in the present case, with which Mr Salisbury agreed.

(e)

However, even then, reasonable examination of the works does not require the inspector to go into every matter in detail; indeed, it is almost inevitable that some defects will escape his notice: see East Ham Corpn v Bernard Sunley & Sons Ltd [1966] AC 406 and Jackson & Powell, at para 8–239.

(f)

It can sometimes be the case that an employer with a claim for bad workmanship against a contractor makes the same claim automatically against the inspecting officer, on the assumption that, if there is a defect, then the inspector must have been negligent or in breach of contract for missing the defect during construction. That seems to me to be a misconceived approach. The architect does not guarantee that his inspection will reveal or prevent all defective work: see Corfield v Grant 29 Con LR 58. It is not appropriate to judge an architect's performance by the result achieved: see Jackson & Powell, at para 8–238.”

Campagna relied also on an observation by HH Judge William Stabb QC in Sutcliffe v Chippendale & Edmondson [1982] 18 BLR 149 (a decision given in 1971), in relation to the question of the extent to which a supervising architect could rely on the contractor. The Judge said:

“I think that the degree of supervision required of an architect must be governed to some extent by his confidence in the contractor. If and when something occurs which should indicate to him a lack of competence in the contractor, then, in the interest of his employer, the standard of his supervision should be higher. No one suggests that the architect is required to tell a contractor how his work is to be done, nor is the architect responsible for the manner in which the contractor does the work. What his supervisory duty does require of him is to follow the progress of the work and to take steps to see that those works comply with the general’ requirements of the contract in specification and quality. If he should fail to exercise his professional care and skill in this respect, he would be liable to his employer for any damage attributable to that failure.”

This case went to the House of Lords as Sutcliffe v Thackrah [1974] AC 727, where the decision of the Court of Appeal was reversed and HH Judge Stabb’s decision reinstated (but I understand that it was specifically reported in the BLR to give wider currency to the judge’s observations on the nature and extent of the architect’s duty).

These passages provide a salutary warning to a court to be careful to avoid reaching the conclusion that because a defect has been found to exist the person responsible for checking the work must have been at fault, or that it is unjustified for a person in the position of a technical auditor to place some trust in the contractor. In this case it is quite clear that Liberty knew from the outset that not every defect would be noticed and it was Liberty that took the decision as to the extent and frequency of inspection by Campagna for which it was prepared to pay. In addition, I remind myself of the danger of making assessments of Campagna’s conduct at the time with the benefit of hindsight.

There was much discussion during the trial as to whether the role of the technical auditor was more akin to that of a building control officer (in the United Kingdom) or that of a chartered building surveyor who might be engaged, for instance, to inspect a house on behalf of a potential purchaser.

It is, I think, common ground that the scope of Campagna's duty and the standard of care required must be determined in the context of Campagna's retainer. It was not suggested that the 2004 Retainer increased the standard of care that had been required under the initial letter of engagement. Further, and as is the case with any commercial contract, the 2004 Retainer must in turn be construed in the context of the commercial circumstances that gave rise to it.

One thing is clear and that is that Campagna did not contract to ensure that any building that its technical auditors inspected would be defect free. The duty of the technical auditors was to take reasonable care during the inspections that they were required to carry out to ensure that each building represented a standard risk from an insurance point of view. The technical auditors were not required to check for strict compliance with the Irish building regulations, although to the extent that the regulations represented good construction practice, the technical auditors would be expected to check for compliance so far as the regulations related to the risks insured by the PGS, that is to say in relation to weather tightness and structural stability. For example, the technical auditors were not required to check the quality of the insulation in the roof spaces of the houses, apart from checking that most of it was physically present (in accordance with the logic check).

In their written opening submissions, Mr Ben Quiney and Mr Richard Sage, who appeared for Liberty, submitted that whilst there is no body of law dealing specifically with the duties of a latent defects technical auditor, the position is generally analogous to that of a building surveyor. By contrast, in their written opening submissions Ms Kim Franklin and Mr Crispin Winser, who appeared for Campagna, submitted that the contractual standard of care owed by its technical auditors was that of a building control inspector and that Liberty was wrong to say that Campagna's technical auditors should be judged by the standards of an ordinarily competent building surveyor. Whilst Liberty’s position was pressed rather less strongly by the end of the hearing, Campagna’s submissions on this did not change.

For my part, I do not consider that Campagna's technical auditors can be characterised as either building control officers or building surveyors, at least not in the sense in which those roles are commonly understood. The role of a technical auditor, under this engagement (and perhaps generally), is a combination of these two roles. The technical auditor needs to adopt a consistent and methodical approach to his or her inspections but, unlike a local authority building inspector, is not required to be a master of the minutiae of the local building regulations. But the technical auditor must be familiar with standard building techniques in the country in which he is operating and be able to bring an experienced eye to the standards of workmanship evident in the building that he is inspecting. It is clearly the duty of a technical auditor to look closely at those details of construction that are critical to structural integrity and the general weather tightness of the building. However, unlike a building surveyor engaged to carry out a survey of an existing house for a prospective purchaser, whom the purchaser might expect to find any reasonably ascertainable defect of any consequence, it is clear that a technical auditor cannot be expected to find every defect in every house during his intermittent inspections: it is inherent in the nature of intermittent inspections that some defects, however noticeable at a particular moment, will escape detection. Indeed, this is expressly recognised in the policy which advises the home owner that "No inspection system can prevent or identify all defects and the Policyholder should carry out a thorough inspection of the New Housing Unit prior to hand over".

A building control surveyor is concerned to ensure that, when completed, the building complies to the letter with the applicable building regulations. By contrast, a technical auditor must be concerned to ensure, so far as he reasonably can, that, when completed, the building has been constructed properly and in a manner that will achieve the required degree of weather tightness and structural ability for several years, not just at the time of inspection. It is also relevant to note that in the 2004 Retainer Campagna’s technical auditors were required to “use all reasonable skill and care to be expected of a competent surveyor”, not the skill and care of a competent building control officer.

The position is, I think, put very well in Appendix C of the Site Audit Surveyors Guidance Manual (to which I have already referred) where, under the heading "The Basic Concept", it said:

“Surveyors are engaged by The Premier Guarantee to manage insurance risk and to ensure that new housing units are constructed to recognised standards. This includes a combination of desktop appraisal of design proposals, combined with on-site inspection of work in progress and work completed. If information required ensuring compliance with standards is not provided, defects identified on site or works not completed, then The Premier Guarantee may withhold the insurance certificate for the new housing unit in question."

Mr Mulrooney, the expert instructed on the half of Campagna, was cross-examined at some length on this passage and, in his closing submissions, Mr Quiney made much of what he described as the "difficulty Mr Mulrooney experienced with this document, and particularly with the word “ensure”". I think that perhaps Mr Quiney's criticism here is a little harsh. Whilst the extract that I have quoted might suggest that Campagna's technical auditors were under a duty to ensure a particular result, the manual must be read in the context of Campagna's terms of engagement - which imposed an obligation to take reasonable care, not an obligation to achieve a particular result. So what the manual is really saying is that Campagna's technical auditors must exercise reasonable care when carrying out their plan checks and site inspections so as to ensure - consistently with the exercise of such care - that new housing units are constructed to recognised standards.

This to my mind requires the technical auditor to adopt a critical and inquisitive approach when carrying out his inspections. However, I do not accept the suggestion that a technical auditor’s duty goes so far as to require him to make the contractor open up any concealed details that he has not had an opportunity to inspect: indeed, if it had been suggested that Campagna’s technical auditors should do this as a matter of routine I would have expected MDIS to have protested about it, unless there was some reason to think that work that had been concealed might not have been properly done or that the builder had deliberately covered up the work so as to prevent inspection. It is in the nature of intermittent inspections that not every stage of the work in every house will be seen. To this extent, therefore, Campagna's technical auditors would have been bound to trust the contractor to carry out his work properly. But I consider that this trust must be earned. During the initial visits to a new development a technical auditor must ensure that he sees examples of crucial details such as damp proof courses and cavity trays before they are covered up. Once he is satisfied that the contractor knows what to do and is doing it properly, it may be legitimate for the technical auditor to be less demanding when inspecting subsequent properties built to the same design.

Beyond this, I do not find it easy to find a form of words with which to describe the degree of vigilance that should have been exercised by Campagna when it was carrying out its duties as technical auditor under the 2004 Retainer (or under the earlier arrangements). Rather, I consider that it is probably more helpful to consider each of the defects in the sample properties in some detail and then to determine whether Campagna did or did not discharge its duty in relation to that particular defect. I hope that in this way it may be possible to derive a standard that can be applied to other situations.

A point that was strongly relied on by Campagna was the difference in the fee structure for the UK and that for Ireland. In the UK Campagna was paid £300 per inspection and in Ireland €75 per property. Ms Franklin submitted that the lower fee for Ireland was indicative of a reduced scope of the service in Ireland compared with that in the UK. The total fee received by Campagna for the First Developments was about €20,000.

I have to confess that at first I was impressed by this submission. The difference between the fee received in the UK and that for Ireland appeared to be stark. In fact, it turned out that it was not a comparison of like with like. Mr Devaney said in evidence that the fees paid to Campagna for the UK and Ireland were effectively the same. Unfortunately, there was no documentation to support Mr Devaney's assertion that the amount received per visit in Ireland was €210 (the assumption on which his evidence was based). In its closing submissions Campagna submitted that the position was as follows:

Whilst I accept that this is a significant difference, it is common ground that the scope of the inspections in the UK was wider (for example, the inspections in Ireland did not extend to drainage and much more detailed plans were typically provided in the UK). I therefore find it impossible to make a true comparison in terms of time spent per pound earned. However, I am prepared to accept that in absolute terms the fee in Ireland (per hour spent) may have been less than for the UK, but I do not consider that any difference is sufficiently great to have an impact on the level of the standard of care.

I have already mentioned and set out some passages from the Site Audit Surveyors’ Guidance Manual that was issued to Campagna in October 2003. These are obviously directly relevant to the performance of Campagna's duties.

In another passage from Appendix C to the manual (that preceded the extract quoted above), which was entitled "How to achieve a satisfactory final inspection (Ireland)", it stated that:

“. . . in order [to] ensure that building works are suitably complete; The Premier Guarantee has set out a standard that can be readily interpreted by all interested parties and will provide a consistent approach to assessing the completeness of a new housing unit. It is also hoped that this will minimise the occurrence of abortive visits and ensure that any potential delays associated with rearranging inspections can be avoided.”

The manual provided guidance as to what would or would not result in a satisfactory final inspection by setting out a sequential logic test. It said:

“If the answer to any of the standard questions is YES then that the item will be classified as CRITICAL and confirmation that a satisfactory final inspection has been carried out will be withheld until the relevant issue is resolved.

If the answer to the standard questions [sic] is NOthen the item will be classified as NON-CRITICAL and confirmation of this would be provided to the house builder . . .

It is important to understand that the pre-hand over inspection performed by The Premier Guarantee’s audit surveyors is concerned with the checking of warranty and associated items; it is not a snagging inspection."

The logic test was shown diagrammatically in the manual as follows:

The manual went on to give examples of the logic test in operation. I will quote just some of them

I have to confess to thinking that there are some inconsistencies between the logic test and the risks insured. It is not clear to me why a number of defects, none of which would fall within the cover, when taken together would justify withholding a Certificate of Approval simply because the necessary remedial work would result in significant disruption to the owner. To take Example 1, roof insulation is not part of the “Structure” and I have difficulty in seeing how its absence would be likely to cause damage to the Structure. However, its absence is easily noticed and I can see why it would make commercial and practical sense to withhold a Certificate of Approval if a large expanse of it was missing. Another anomaly, pointed out by Ms Franklin, is that humidity that is not caused by water ingress through the weatherproof envelope is not covered by the policy, whereas a mechanical extract fan from bathroom discharging into roof void constitutes a critical defect under the logic test.

It may be that these examples have created an element of confusion in the minds of some of Campagna's technical auditors and MDIS about the nature and extent of the final inspection. One example is to be found in the witness statement of Mr Stuart Gough, a technical auditor and director of Campagna. At paragraph 12 of his witness statement he said that "Campagna was not expected to consider building regulation issues as this was the role of a separate certifying professional" and that "Campagna was retained to ensure that the development did not pose an abnormal insurance risk and not to exercise a building control function".

However, at paragraph 18 of the same statement he said:

“Examples of items checked at final inspection stage included:

- Mastic around external door and window frames

- Weathering at thresholds

- General inspection inside including the uppermost floors in particular to ensure that there was no evidence of water ingress.

- A ground level check of chimney and abutment flashing where visible.

- General check that the unit was habitable. If kitchen units, sanitary ware was ex-contract and to be completed by the purchaser (as part of a PC sum) then Campagna would receive confirmation of this.

- A cursory look in the roof space to check that the roof insulation was in place.

- Check that adequate vents were in place in the external blockwork walls of timber frame construction

- Flue outlets were satisfactory.

By ensuring that the above items were completed satisfactorily, the requirements of the logic test (included in the site audit auditor's manual) would automatically be satisfied.

Campagna's auditors could not have seen the following at final inspections:

- Whether the roof coverings were watertight, unless there was evidence of water ingress.

- Whether the damp proof courses were correctly installed - unless there was evidence of water ingress or damp.

- The extent to which roof tiles were fixed unless they were missing.

- Whether the external elements were structurally supported - unless there was evidence of stress, deflection or failure.

An inspection of the roofspace would not normally have taken place. Defects in roofing felt and the connections of ducting to ceiling fans would not have been seen.”

There are two inconsistencies in this passage. First, whilst Mr Gough says in the first part of the passage that the final inspection would have included a cursory look in the roof space, he says a little later that an inspection of the roof space would not normally have taken place. Second, the example of a mechanical extract fan from a bathroom discharging into the roof void which is given in the manual as a defect that would a result in a failure of the logic test, is one that could not be discovered except by looking in the roof space. However, this is something that Mr Gough identifies specifically as something that would not normally be seen on a final inspection, even though he had previously described the final inspection as including all the matters that would be necessary to satisfy the logic test.

Mr Gough was a civil engineering graduate who had been employed by the Belfast City Council as a building control surveyor for some 12 years. I would not have expected him to have been in any doubt as to the nature of his duties. It must be remembered that a final inspection would almost invariably take place after any scaffold or other means of access to the roof of the property being inspected would have been removed. It seems to me that a check of the roof space of any house was an essential feature of a final inspection. For reasons that I will give later in this judgment, I do not consider that the absence of a suitable ladder with which to gain access to the roof space is a valid reason for not carrying out that part of the inspection: arrangements should have been made to ensure that a ladder was available.

The problems facing technical auditors in Ireland were compounded by the fact that the plans, which should have been checked before construction started or, at least, at a very early stage in the development, were seldom available at the necessary time and, when they were produced, were sometimes little more than drawings produced for the purpose of obtaining planning permission. Of the few drawings that I have seen amongst the documents in this case, there are virtually none which provide information about particular construction details such as cavity trays.This was a situation of which Liberty and MDIS were aware: in a note of a visit to MDIS in January 2006, Mr Lynch, Liberty’s underwriter at the time, recorded the site surveyors’ complaint that they could not get hold of detailed plans. Mr Byrne accepted in evidence that where the design information was either not provided or was not available, then the focus had to be on what was done on site (Day 4/44).

This meant, as most of Campagna's witnesses accepted, that the technical auditor would have to see or find out what was being done on site in order to satisfy himself that the details were appropriate and that the work was being done to a sufficient standard and suitable specification.

A good example of the difficulty caused by late production or absence of suitable plans and details is provided by the canopies at Derrycorris Drive. The plans, such as they were, consisted of little more than simple elevations and layouts of the houses. On the drawings prepared by the contractor who was to supply the timber frame, the canopy was shown as an unsupported projection. So whilst it was clear from the drawing that the canopy was not to be supported either by a post or by brackets, the precise method by which the canopy was to be fixed to the main structure was not shown. However, on a different plan there was a sketch of the front elevation of a Type A house that appeared to show brackets at either end of the canopy, but these were not shown on the side elevation of the house in the same drawing. This was a drawing that was evidently prepared for the purpose of obtaining planning permission, rather than with a view to showing construction details. In these circumstances, I do not consider that any significance can be attached to the fact that a bracket may have been shown on the front elevation of the house when it was not shown on side elevation on the same plan.

The promoters of the HomeBond scheme had produced a very thorough House Building Manual on building regulations and techniques in Ireland, and this was a document that was widely used in the industry and was in fact carried by the Campagna technical auditors. As the manual indicated, canopies in Ireland were typically supported in one of four ways. The first two methods involved the use of external support, either by a post or by what were referred to as “gallows” brackets (that is a right angle bracket with a diagonal bracing). The third method was a cantilever design, in which the canopy was supported on the joists at first floor level that projected through the external wall. The fourth method consisted of hanging the canopy by fixings through its framework to the external wall of the house. All these were illustrated in the manual.

A feature of the development at Derrycorris Drive was that the method of construction of the walls of the houses was changed at some stage prior to construction. The original intent was for a double leaf external wall consisting of an outer leaf of brickwork and an inner leaf of either brickwork or blockwork. The revised plan replaced the inner leaf by a timber frame consisting of a timber carcass reinforced by sheets of plywood (or similar material). With this arrangement it is not possible to construct the joists so that they project through the external wall. Instead, the joists are suspended from the timber frame. In my view, a competent technical auditor should have appreciated that a consequence of this change in the method of construction was that the canopies would have to be hung off the building and would not be supported on projecting joists. As anyone with a modicum of experience of construction techniques would be expected to appreciate, supporting a canopy on projecting joists is a much more robust form of construction then hanging it from the external wall: the latter arrangement relies entirely on the integrity of the fixings into the wall. Further, the change to the timber frame construction would make it impossible to attach the canopy by means of bolts inserted from the inside of the outer leaf of brickwork (which have the advantage that they cannot be pulled out), with the result that the canopy would have to be fixed to the wall using either chemical anchors or some very robust form of mechanical anchor (such as a bolt with a sleeve that expands as the bolt is tightened) inserted from the exterior of the building.

If a proper set of plans showing the detail of the construction of the canopies had been provided at the pre-construction stage, the technical auditor would probably have been able to see how the canopies were to be constructed and, in particular, how they were to be supported. But in the absence of any such plans or sketches, I consider that it was the duty of the technical auditor to find out from the builder precisely how the canopies were to be fixed to the building.

In the case of Nos 6, 11 and 12 Derrycorris Drive the architectural and structural details were not provided until after the houses had been completed. Accordingly, if the technical auditor was to know how the canopy was being supported to the building, he would either have to see a canopy in the course of construction or ask the builder for the proposed details.

In evidence, Mr Williams said insistently that canopies of this sort were a standard feature throughout Ireland and so he did not consider it necessary to make any enquiries since it was clear that builders were able to construct them without difficulty. Unfortunately, this approach ignores the fact that there is more than one way of supporting a canopy: the mere fact that canopies have been constructed on other buildings at other times by a method or methods that have proved satisfactory provides no guarantee that such a method would be used on this particular development.

The Site Audit Surveyors’ Guidance Manual states that, in general, inspections should be undertaken at foundation, pre-plaster and completion stages. I consider that Campagna's technical auditors should have carried out inspections at these stages as an absolute minimum (unless this was not possible because insufficient notice was given by the developer). I will consider these inspections in more detail below.

In addition, I consider that Campagna should make additional visits where necessary in order to inspect critical construction details which would not be seen at the three main inspection stages.

The agreed issues in relation to inspections are as follows:

With the benefit of the knowledge gained during the hearing, I consider that the wording of these issues is not entirely appropriate. However, in the next part of this judgment I will do my best to answer the questions.

I consider that the issue in relation to the Desktop Study (the plan check) is not susceptible to any useful answer because in the case of the great majority of house types there were no plans available at the time when the plan check should have been carried out. As I have already explained, such plans as were provided were seldom more than planning application drawings that gave no relevant construction details. So far as I can tell, requests for information either were not dealt with promptly or went unheeded altogether. I am satisfied that there was very little that Campagna could do about this, and it is clear that both Liberty and MDIS accepted the situation.

But in any event the plan check - whenever it was carried out - should have recorded the numbers of the drawings inspected, together with any other relevant information provided, and should have contained a record of any queries that were raised together with the response received. To the extent that it was possible to do so, Campagna should have checked the items listed on the plan check report form.

Where possible, Campagna should have attempted to obtain any relevant ground reports and structural calculations (including timber frame details where relevant). To the extent that Campagna was not able to check these because they lacked the necessary expertise, I consider that they should have referred the matter to MDIS.

In the very many cases where the plans provided gave no, or no sufficient, information and there was little prospect of obtaining any more detailed information, I consider that Campagna should have reported that fact to MDIS. Having said that, I have already indicated that Liberty and MDIS were generally aware of the problem of lack of the necessary drawings and specifications and accepted the realities of the situation.

In the absence of any construction details being provided at the plan check stage, it was for Campagna to check during site inspections that appropriate construction techniques were being adopted. I consider that it was Campagna's duty when on site to find out what the builder was proposing to do in relation to critical details such as damp proof courses and cavity trays. If a technical auditor was not familiar with a particular type of detail, I would expect him to consult a standard reference work, such as the HomeBond manual, in order to satisfy himself that what was being done was appropriate. The evidence indicated that Campagna's technical auditors usually carried a copy of the HomeBond manual when carrying out inspections, and this is what I would expect.

I consider that the principal purposes of this inspection were, first, to check the depth of the foundations to ensure that they reflected good practice and/or the details on any drawings (if available) and, second, to check the nature of the ground. For example, the technical auditor should look for any signs of fill. The inspection should take place before any concrete is poured. Further, I consider that it would be the duty of the technical auditor to find out whether or not there was in existence a report on the ground conditions and, if there was, to obtain a copy of it before approving the foundations. If there was a report and it was not provided in response to a (suitably timed) request, the technical auditor would have to notify the developer that to proceed with the construction without the technical auditor having seen the report on the ground conditions could well jeopardise the issue of a Certificate of Approval.

I consider that the first foundation inspection in respect of each house type should take about 10 minutes. Thereafter I would expect foundation inspections of similar house types to take 5-10 minutes. In the case of more complicated structures, such as apartment blocks, I would expect these times to be considerably longer.

This inspection should take place before any plasterboard is put up or plaster applied to blockwork. The technical auditor should ensure that this is clearly understood by the builder at the outset.

This inspection also provides the opportunity for the technical auditor to check details such as the installation of cavity trays and damp proof courses. It may be that some of these will already have been covered up by the time of the pre-plaster inspection. In that case, I consider that the technical auditor should make arrangements with the developer to inspect one or two sample properties at an earlier stage in the construction sequence so that these details can be seen and approved. Whilst in the case of some of the first houses built this may require a special visit to the development, it should thereafter be something that can be done whilst at the site to carry out other inspections.

An important part of this inspection is to check the underside of the roof, in particular to check that the sarking felt has been properly installed and that difficult details, such as chimneys and valley gutters, have been properly completed. It is essential that the technical auditor is provided with a ladder so that the roof space can be inspected at close quarters. To ensure this, the technical auditor should liaise with the builder at an early stage and make arrangements for the provision of an appropriate ladder. If the builder is unwilling or difficult, it can be made clear that no Certificate of Approval will be issued if the technical auditor has not been able to carry out a proper examination of the roof space and any other areas that require access from a ladder.

I would expect a pre-plaster inspection properly carried out to last some 20-30 minutes in the case of the first house of a particular type. Thereafter, for houses of the same type, I would not expect the inspection to take longer than 15-20 minutes.

At this inspection there will be relatively little to see inside the house and I consider that the inspection should concentrate on the roof and external envelope. I would expect the technical auditor to have both a torch and a pair of binoculars for this inspection. I discuss the use of binoculars in more detail later in this judgment, particularly in relation to Mountgorry Wood.

The technical auditor should inspect the roof space by looking through the access hatch from a ladder. If there is an important detail, such as a chimney, that is not clearly visible from the hatch, I would expect the technical auditor to ask for a board or plank so that he can go into the roof space.

If there is no scaffold or other means of high-level access, the technical auditor should examine the roof closely through binoculars. Should he see something that concerns him which requires closer access, then that concern should be raised with the developer and, if necessary, MDIS before the Certificate of Approval is issued.

I would expect a final inspection to take not less than 15-20 minutes per house. I would not expect the final inspection of houses of a similar type to one already inspected to take very much less. In my view a similar level of thoroughness will be required for almost every house.

At an early stage during the inspections at a particular development, the technical auditor will need to form a view about the general quality of workmanship on that site. If he has reason for concern at the quality of the workmanship, extra vigilance will be called for. If he considers that the builder is deliberately attempting to cover up work before it can be inspected, he should make it clear that he will not approve the issue of a Certificate of Approval unless he is given an opportunity to see the work that has been covered up.

However, in the absence of some reasonable cause for concern, I would not expect a technical auditor to call for the opening up of work that has been executed.

If, whilst a technical auditor is on site, there is access to any building under construction from a scaffold, I would expect the technical auditor to go up the scaffold and examine the work, particularly such matters as the construction of cavity walls. This applies even if the property with the scaffold is not the unit or units that he has come to inspect.

I would not expect a technical auditor to call for a mobile scaffold (assuming that there is one on site) in order to inspect part of a building under construction unless he has a good reason for doing so. This is for the reason that releasing or moving a scaffold is likely to cause disruption to the builder which a technical auditor should seek to avoid.

I do not regard it as possible to give any indication of the frequency at which a particular site should be visited because this will depend on the construction programme. The important point is that the inspections should be sufficiently frequent to enable the technical auditor to see every important stage of the construction of a particular house type. To visit at monthly intervals may be a useful rule of thumb, but it is no more than that.

I consider that the list of equipment that a technical auditor should have with him when carrying out inspections that is set out at paragraph 5.11 of Mr Scott’s report is appropriate (of course, the need for safety equipment such as hard hat and boots will depend on the conditions at and rules of the particular site). To these I would add the HomeBond manual, or some similar reference work. In the case of final inspections, I consider that the technical auditor should also carry a pair of binoculars.

There were various inspection report forms that were in use from early 2003 onwards, which listed the items that should be inspected and, in some cases, contained “prompts” in the form of notes indicating particular details that should be checked. It is obviously good practice for technical auditors to use such forms, but it has to be said that Campagna’s technical auditors probably did not make as much use of such forms as they might have done. Their comments tended to be sparse and sometimes uninformative. Whilst brevity in such matters should not of itself be the subject of criticism, Campagna’s case might well have been assisted if these forms had contained a little more information. It was not really satisfactory for Campagna’s technical auditors to communicate information to each other orally, as the witnesses said they did, whilst not at the same time recording that information in the relevant reports. MDIS did not consider the reports as much of a working tool from its point of view, since for a significant period they were simply filed unread.

It may be helpful if at this point I give my brief views of the witnesses who gave evidence at this hearing.

Mr Michael Evans is a Senior Claims Adjuster at Liberty, whom he joined in June 2005. He is a recently qualified solicitor, having done his training with CMS Cameron McKenna who act for Liberty in this action. His training was not continuous: it consisted of 6 months with Cameron McKenna and then 6 months at Liberty.

Much of his evidence was uncontentious, but he was cross-examined at some length about his visit to the site at Tayleur’s Point and his observation about the different heights of the chimneys along a road of identical houses facing the sea. The purpose of the cross-examination was to establish the basis on which Liberty had decided that the chimneys presented an imminent danger to the health and safety of the occupants of the houses (the state of affairs that would trigger liability under the scheme). I found that his evidence was of limited value: the assessment, such as it was, of the condition of the chimneys had been carried out by engineers and Mr Evans was doing little more than summarising the gist of the views that had been given to Liberty by some of those engineers.

Mr Dan Gray is the Class Underwriter for the Latent Defects business at Liberty. He joined Liberty in June 2006. He was brought in to take over the account from Mr Lynch, who retired in June 2009. Again, his evidence was largely uncontentious and his cross-examination was in large measure an exercise directed at putting some of the relevant documents before the court. I found him to be a truthful and straightforward witness.

Mr Ian Turnick is a Partnership Director of MDIS. His evidence was largely concerned with the procedure for vetting potential developers, both in the United Kingdom and in Ireland. He was another straightforward witness who was prepared to make concessions when they were justified and, so far as his evidence concerned matters that were within his direct knowledge, I accept it.

Mr Gary Devaney is the Managing Director and Chairman of MDIS, the company which he established in 1997 for the purpose of specialising in the provision of long-term insurance warranties for the construction industry. Whilst he was not involved in the day to day running of the scheme, he clearly took a very close interest in it. Although his recollection of the detail of events that occurred several years ago was not perfect, as one might expect, I found him to be a truthful and straightforward witness.

Mr Devaney’s background is in insurance, not construction, and so I find that his evidence on matters related to construction, on which his cross-examination touched at points, was of limited value (but through no fault of his).

Mr Devaney explained that MDIS has been put on notice by Liberty of a potential claim against it in respect of the losses that have arisen on the scheme and was advised to notify its professional indemnity insurers, which it did. Apparently no steps have been taken in relation to that potential claim, but its existence suggests that it may not necessarily be in MDIS's interest for Campagna to be held liable to Liberty. I can therefore see no reason why witnesses called on behalf of MDIS should be predisposed either to defend Campagna or to blame it.

Mr Paul Byrne is currently the Business Development Director for MD Professional and Support Services Ltd, a subsidiary of MDIS. Between 2003 and 2009 he was the Technical Services Manager for MDIS. His responsibilities as Technical Services Manager involved agreeing operational procedures with the site audit surveyors and ensuring that there was appropriate and effective communication between those surveyors and MDIS.

I found Mr Byrne to be an articulate, careful and truthful witness. He, like some of the Campagna witnesses, had formerly worked for Carillion Specialist Services Ltd - in his case as Deputy Area Manager (Northwest). For example, he knew Mr Williams and Mr Hilton from that period. Given his knowledge of both parties, I thought that he gave his evidence very fairly.

Mr Andrew Gee is an Architectural Technician employed by Aidan O'Connell and Associates Ltd, who are based at Portlaoise in Ireland. In that capacity he has been involved in investigating some of the claims made by homeowners that have given rise to this litigation. In particular, he was involved in the investigation of certain defects at the Mountgorry Wood development.

His evidence was limited to describing the investigations that he had carried out and a report that he had written about them. I found him to be a truthful and straightforward witness.

Mr Aidan O'Connell is the Managing Director at Aidan O'Connell & Associates Ltd. His firm was instructed as the engineers to investigate some of the claims. Although many of the in-depth investigations were done by others in his firm, he said that he probably saw a sample of every defect at Mountgorry Wood. He was a clear and articulate witness who struck me as someone who generally knew what he was talking about.

Mr Conor Holland joined the loss adjusters, OGS, on graduating as a Building Surveyor. He was involved in the investigation of some of the claims relating to the failure of the render at Tayleur's Point. He was asked a number of questions about his understanding of the policy wording, but his evidence was not really controversial. He struck me as a perfectly straightforward witness.

Mr Michael Gillett is the Head of Claims for Liberty. He joined Liberty in 1995 and has over 25 years experience in the insurance industry, mainly on the claims side. He was cross examined about the various changes to the policy wordings between 2003 and 2007 and as to the similarity of the scheme in the UK and the scheme in Ireland and his expectation of the likely level of claims in Ireland. I regard his evidence as fairly peripheral to the issues that I have to decide, but I have no doubt that he was a truthful witness.

Mr Chris Kerrigan was the Regional Manager for the loss adjusters, OGS. His evidence mainly described visits that he made to Derrycorris Drive and to Tayleur's Point to inspect properties whose owners had made claims under the scheme, and his dealings with some of the individual owners.

I found Mr Kerrigan to be an honest witness, but I felt that on occasions he was prepared to answer a question without having thought about it properly. For example, when asked about a passage in his witness statement where he dealt with the non-compliance of the chimneys at Tayleur's Point with the Irish Technical Guidance note, he said that this was a matter for others to assess. However, it was quite clear from the relevant passage in his witness statement, and he confirmed later in his evidence, that he was expressing a view that he himself had formed at the time based on his own knowledge of the Technical Guidance note.

This was not a difficult matter because it concerned the ratio of the height of the chimney to its maximum width: Tayleur's Point was in an area where this had to be less than 4.5. It was not very difficult to make a crude assessment on site as to whether or not this ratio had been exceeded, which is what Mr Kerrigan obviously did.

Mr David Hilton was the first witness called on behalf or of Campagna. He is an Associate Director of Campagna, and he joined the company in 2003, by which time he had had about 11 years experience as a technical auditor.

Mr Hilton had met Mr Bob Williams when he worked with the Property Services Agency (later to become Carillion Specialist Services). Mr Hilton was one of the technical auditors who carried out inspections on the three developments which are the subject of this hearing.

I found Mr Hilton to be a confident and assertive witness. However, he had fairly little recollection of the actual inspections that he carried out. This is not surprising given the lapse of time. He was reluctant to accept that any of the defects that have since come to light were there to be seen during his inspections. He accepts that he would not have seen everything and that, where the standard of work in a particular area had been shown to be satisfactory, he would sometimes take on trust the fact that the work had been carried out to a similar standard in places that he had not been able to inspect.

Mr Bob Williams was the founder of Campagna Ltd (the English company), which was incorporated on 17 June 2002, and is its Managing Director. He has 15 years experience of technical auditing for latent defects.

Mr Williams was, perhaps unsurprisingly, a rather defensive witness. He was reluctant to accept that any of the defects of which Liberty complains might have been present at the time when either he or any of his colleagues carried out their inspections. In one or two instances he expressed opinions about compliance with regulations that were simply wrong. For example, in relation to the technical guidance in relation to chimneys that the height should not exceed 4.5 times the width of the chimney, he said that it was "common for professionals to assume that the 4.5 x dimension is "gospel" and cannot be altered. This is not the case. These properties are outside the zone which is covered by the exposure guidance.".

When it was pointed out to him that the relevant houses were within the zone to which the 4.5 height to width ratio applied, he said

“. . . what I meant was outside the 4.5, the chimney height 4.5. I think it's a slight error there."

He was then compelled to accept that, since he had referred to the "properties"being outside the zone, his explanation that he meant that the chimney was outside the height to width ratio was unsustainable. He then said that it must have been something that he missed when he proof read the statement.

I do not find this to be satisfactory evidence. This is an example of Mr Williams seeking to justify what he did at the time by relying on a misplaced interpretation of the technical guidance. However, I do not think that Mr Williams was trying to mislead the court: this is one of those cases - not uncommon - where the exigencies and pressures of the litigation have persuaded a witness to adopt an unrealistic view of the facts.

Mr Stuart Gough is a director of Campagna. Unlike his colleagues, Mr Williams and Mr Hilton, Mr Gough is an engineering graduate. After graduation he worked for 12 years as a Building Control Surveyor. He is a member of both the Association of Building Engineers and the RICS. In 1997 he moved to Tarmac Professional Services where he worked as a technical auditor. He joined Campagna when it was founded in June 2002.

Mr Gough was a confident and articulate witness and was a man of firm views. He emphasised, not unreasonably, the periodic nature of the inspections by Campagna and the fact that they could not be expected to find every defect that existed. However, he was adamant that the inspections that he and his colleagues carried out were sufficient and that, for example, Campagna never issued a Certificate of Approval for a property without having done a final inspection of that property. When it was suggested to him that, where there was no evidence of a final inspection having been carried out for a property in respect of which a Certificate of Approval had been issued, an alternative possibility was that there had been no final inspection of that property, his short response was that "or" (ie. the alternative) was not an option.

Mr Stephen Scott was the expert called on behalf of Liberty. He is a Chartered Building Surveyor. He has been in practice for 21 years and is the Managing Director of Watts Consultancy Ltd, one of the UK and Ireland's largest Building Surveying practices. However, as Ms Franklin pointed out, he has never acted as a technical auditor for an insurance company.

Mr Scott prepared a detailed and comprehensive report running to 170 pages, supported by three files of appendices. I found Mr Scott to be a knowledgeable and thorough expert and, on the whole, I accept his evidence. However, in a few areas I was not satisfied that Mr Scott's conclusions were justified on the material that he cited. Where I have not accepted his views I have given my reasons for not doing so in the relevant part of this judgment.

Mr Nevan Mulrooney was the expert called on behalf of Campagna. The presentation of the expert evidence by Campagna was not entirely satisfactory. Originally, Campagna had retained two experts. The first was a Mr Howard Jenkins, who was retained to deal with the defects. The second was Mr Mulrooney, who was retained initially to deal with Irish construction standards. However, a few weeks before the trial, Campagna decided to dispense with the services of Mr Jenkins, who by that time had been involved in preparing the experts joint statement with Mr Scott, and to instruct Mr Mulrooney to deal with the defects as well as construction standards in Ireland. Mr Mulrooney produced an initial report running to about 100 pages and, shortly after the start of the trial, a supplementary report which responded to the report of Mr Scott.

Mr Quiney was very critical of the reports of Mr Mulrooney. He submitted that in his reports Mr Mulrooney often did not distinguish between a recital of fact and an expression of opinion. In the light of this, Mr Quiney submitted that much of Mr Mulrooney's evidence should either be ignored or rejected. In addition, he submitted that there were several questions put to Mr Mulrooney in his instructions which he did not answer in his reports. He submitted also, correctly, that Mr Mulrooney had had only a limited opportunity to inspect the sample properties: in particular, he had not seen the inside of any of the properties at Derrycorris Drive. I accept that Mr Quiney's criticisms have some force. To a large extent I consider that his first criticisms largely arise out of loose drafting by Mr Mulrooney, rather than any reluctance to engage with the issues. The second criticism makes a fair point and Mr Mulrooney's evidence must be assessed in the light of it.

Mr Quiney also submitted that Mr Mulrooney was a partisan witness who did not give his evidence objectively. I think that there is some force in this, but I formed the view that when Mr Mulrooney was pressed in cross examination he generally gave his evidence fairly - although I think it is right to say that his evidence was at times coloured by a degree of sympathy for his clients. I am certainly not prepared to reject Mr Mulrooney's evidence out of hand: on the contrary, I thought that parts of his evidence were perfectly sound and in some cases I accept it, as will appear in subsequent parts of this judgment. I take into account also that Mr Mulrooney's role was changed at the eleventh hour and that is not something for which he is to blame.

I now propose to deal with each of the defects in the sample properties that are the subject of this hearing. I will take each development in turn in the order in which it was built, which was not the order adopted in the reports.

The single storey houses, which have dormer bedrooms, have steep roofs with a pitch of 45°. As a result they are sometimes referred to as the "chalet bungalows". As I shall explain, it is clear from the evidence that on a tiled roof with a pitch of 45° every tile had to be nailed. This concerns Nos 5, 13, 25, 25a and 91.

The tiles that were used, the Roadstone Stonewold Flat Concrete Tile, were specially constructed with punch outs through which the fixing nails could be driven without breaking the tile. I assume that the tiles had a lip on the inside face from which they could be hung from a batten without being nailed. In addition, the manufacturer’s information shows that the two vertical edges of the tiles were moulded so that each tile could interlock with the next one. This moulding did not, so far as I can tell, contain any form of mechanical lock that would prevent the tile from moving against the neighbouring tile, but the overlapping mouldings would, I suspect, have provided significant friction. Each tile weighed just under 5 kg.

The manufacturer’s technical data stated that on all roofs "at pitches 45° and over" each tile should be nailed or mechanically fixed. In its section on the roof tiling/slating, the HomeBond manual divided roofs into those with a pitch less than 45° and those with a roof pitch of 45° to 55°. In terms of situation, buildings were classed as normal, exposed or severely exposed, according to location. For normal buildings with a roof pitch of between 45° and 55° the HomeBond manual advised that every tile should be head nailed or tail clipped over the entire roof area. At one point Ms Franklin submitted that it was not necessary to nail the tiles at every course on these 45° roofs at Tayleur's Point. Whilst I accept that a 45° roof is right on the borderline, the recommendation in the technical documentation is clear: on a roof of 45° or more every tile has to be nailed.

During 2005 and 2006 several tiles were blown off some of the roofs of these houses during high winds. There is also evidence, and I find, that these tiles had not been nailed when they were fixed, as they should have been: on many of the tiles which had been blown off the roof the manufacturer's punch outs for the nails were still intact, and the battens in the area from which the tiles had come had no nail holes. The evidence, both of Mr Kerrigan and in the contemporaneous documents, shows that in the case of some houses there were large parts of the roofs where the tiles were not nailed.

It was the evidence of Mr Williams that during one of his early visits to Tayleur's Point he asked how the tiles were to be fixed. He says that he was told by the builder that they would be nailing every tile. Mr Hilton said that he remembered being told this by Mr Williams before he went to the site for the first time. In his evidence Mr Williams explained that the developer at Tayleur's Point, Logan Court Properties, had been in business for 15 years and that their representative on site, Alan Drumm, was the brother of the managing director. He said that it was a family business with a reputation to maintain. Mr Williams said that it was Alan Drumm who told him that they were going to nail every tile on the “chalet bungalows” (Day 8/68-70). Mr Williams said that, since he regarded Mr Drumm's firm as reputable builders, he had no reason to doubt what he was told. I accept this evidence.

In Mr Hilton’s oral evidence, which was given by video link (for reasons relating to Mr Hilton's domestic circumstances), he said that he checked some of the tiles by pushing them with a batten and they didn't move. He said that he did this from a scaffold. He said that he saw the roofer with a hammer in his hand and a bag of nails and he assumed that he was doing his job properly. However, Mr Hilton did not mention any of this in either of his two witness statements. I have to confess that I have some reservations about this evidence, because I think that if Mr Hilton could really remember having tested tiles with a batten during his visits to site, this is something that he would have mentioned in his witness statements because its importance was self evident.

It was not suggested by Mr Scott, the expert instructed on behalf of Liberty, in his report that Campagna's technical auditors ought to have checked the fixing of the tiles by pushing them with a batten in the manner suggested by Mr Hilton; this suggestion arose for the first time in the course of Mr Hilton’s cross-examination and was then to some extent adopted by Liberty. Until then Mr Scott’s position was that the technical auditors had the opportunity to see the roofing in progress, because they were on the site at the appropriate times and so they had an opportunity to look and check the fixing method whilst the tiles were being laid.

In cross examination, after agreeing (in answer to a question from the court) that the technical auditor would want to satisfy himself that the nailing regime was appropriate to the location, Mr Williams said (at Day 8/70):

“Yes, and you would see that as you were going round, and you could climb on them and test some of the tiles on some of the roofs. You wouldn't test every house.”

Mr Williams did not say how he would test the tiles but I assume that he meant that one could attempt to dislodge a tile with one's hands and would then be able to tell whether or not it had been nailed.

It seems to me, as I indicated during the evidence, that the difficulty with this type of test carried out whilst the roofer is actually working on the roof, is that any sensible roofer who has been told to nail every tile is likely to do so if he thinks he is being watched. The problem is what happens when the technical auditor’s back is turned. Since I accept the evidence of Mr Williams that he was told by Alan Drumm that every tile would be nailed on the 45° roofs, I consider it to be more probable than not that this is what the roofers were told to do. If they had not been told to nail every tile but told instead, say, to nail every second row, this is what they would have done and I would have expected Mr Hilton to have noticed it.

In a report dated 23 June 2006 (which concerned No 13) Mr Kerrigan, of OSG, said that OSG’s enquiries suggested that the tiles were nailed every fourth row only. However, information received by Mr Kerrigan from Mr Hamm (of M T Hamm Ltd, referred to below) after many of the roof coverings had been stripped indicated that there was no regular pattern to the nailing of the tiles.

I accept that if a house was being tiled whilst Campagna's technical auditors were on site, they should have climbed onto the scaffold to check that the roofer was nailing every tile as Mr Drumm had instructed. However, I consider that as soon as the roofers became aware that technical auditors were on site they would have taken care to ensure that they nailed every tile if that is what they had been told to do. If the technical auditor checked the tiles that had already been fixed, assuming that there was a ready means of doing so, limitations of access may well have restricted his check to the tiles in the area in which the roofer was currently working.

In the absence of more cogent evidence as to precisely how the tiles could or would have been checked by a technical auditor, I suspect that a simple manual check of a few tiles carried out in the way I have described would have been frustrated because the roofer would have ensured that, whilst the technical auditors were on site, he fixed the tiles in accordance with his instructions. It was probably when the technical auditor’s back was turned that the poor fixing would occur, as Mr Scott accepted (Day 10/171-2).

In this context I was referred to the decision of HH Judge Hicks QC in George Fischer Holdings v Multi Design Consultants (1998)61 Con LR 85, in which he firmly rejected an argument along these lines. However, the facts of that case were very different from the facts here. That case concerned the construction of a warehouse roof which was covered with large panels. The panels did not extend over the full distance from the apex to the eaves and so there were horizontal joints or laps between the panels. The roof leaked and a central issue was whether the construction of these laps in the panels was properly checked during construction by the quantity surveyors who were retained for that purpose. In fact the quantity surveyors never went on to the roof during construction because they said that access was not safe. That defence was firmly rejected. The other line of defence was that the defective formation of the lap joints would not have been detected even if the surveyors had carried out visits during their construction because the work in question was probably not carried out during the visits; alternatively, if it had been, the workmen would have taken particular care whilst under the eye of the inspecting surveyor. Judge Hicks rejected this defence at every level. As to the suggestion that the workmen would “put on a show", the judge observed that either they would not be able to do that all the time or, if they did, since the surveyor should have been present when most of these joints were being constructed, that would still have achieved much the same end. Whilst I do not wish to be taken as in any way disagreeing with what Judge Hicks said in that case, it seems to me that the position is altogether different in this case where there are many thousands of roof tiles and no one could realistically suggest that the technical auditors could have watched anything more than a tiny fraction of them actually being fixed.

I consider that the roofers at Tayleur's Point were probably instructed by Mr Drumm to nail every tile, but that they did not always do so. I think that it is likely that when Campagna's technical auditors were on site they would have nailed every tile in accordance with the instructions. Even if the technical auditor had checked the tiles that were being fixed, for the reasons that I have already given, I do not consider it more probable than not that he would have found that some of the tiles had not been nailed. He may have done: but I do not feel able to conclude that this would have happened as a matter of probability.

I accept that it might have been possible, in principle, for Campagna's technical auditors to check the tiles on a roof that had been completed by attempting to move individual tiles by pushing them with a batten. However, I have reservations as to whether these particular tiles, where not nailed, would have been easily moved if pushed in this way. As I have said, the tiles are heavy, they have a moulded overlap at each side and they have the tiles from the layer above resting on them. It might take quite a lot of force to move a tile in these circumstances and, since no one has actually tried it, I am not in a position to conclude that it would have been reasonably possible to test the tiles in this way. Alternatively, if I was to accept Mr Hilton's evidence that he did try pushing tiles with a batten, then the fact that they did not move would be consistent either with the fact that they had been nailed or that it was simply not possible to dislodge them by pushing them with a batten.

Whilst I do not find this question easy, I consider that in the light of the evidence as a whole it has not been demonstrated that Campagna’s technical auditors should have discovered that many of the tiles on the steep pitched roofs of the dormer houses had not been nailed. I am satisfied that they were told that every tile would be nailed and there is no evidence that they saw anything that indicated the contrary. In the absence of such indication, they trusted the builder. I am not persuaded that they were wrong to do so. On this allegation, therefore, I find that Liberty has failed to prove its case.

The gable ends of the dormer houses are stepped and there is a balcony at first floor level on the recessed part of the elevation. Part of the roof which overhangs the balcony projects to the same line as the edge of the roof on the opposite side of the gable. The overhang is then stepped back for the lower part of the roof. The part of the roof that overhangs the balcony extends out about 1 m from the face of the gable wall, and the length of the overhang (as measured in the plane of the roof) is about 3.5 m. It is alleged that the section of the overhang above the balcony is not properly supported with the result that it is prone to excessive movement which can result in tiles becoming dislodged. Whether or not this section of the overhang is in any danger of collapse, I consider that the present risk of a tile that weighs nearly 5 kg falling with potentially fatal results is quite serious enough to amount to imminent instability that requires rectification within the meaning of the policy.

In the plan check report dated 21 May 2003 Mr Hilton made the following observation:

“Trusses OK, but additional info requested for cut roofs."

It was accepted that this reference referred to the overhang above the balcony (although Mr Hilton suggested at one point that he thought that there were other cut roofs on the development to which it would have applied also) and that it reflected a concern by Mr Hilton as to how it was designed. In cross examination Mr Hilton said that his recollection was that there was a cantilevered beam at the bottom end of the overhang, and that he must have checked that this was so in order to issue the Certificate of Approval. He said that he was unable to remember now what calculations or checking he carried out in order to satisfy himself that the design was adequate. I do not find this evidence surprising: it would perhaps have been more surprising if Mr Hilton had been in a position to remember such details of events that occurred over seven years ago.

Mr Scott relied on the detail that he said was shown in drawing number 99.195.3.1.102. This was a drawing prepared for the purpose of submitting the planning application. It showed the ground and first floor plans (the latter being the dormer bedrooms), front and rear elevations, side elevations and one section. One of the side elevations showed what appeared to be a vertical support from the balcony to the underside of the roof overhang. The first floor plan showed what appeared to be the same support (in the form of a small circle on the balcony). However, the front elevation, whilst showing a support to the balcony from the ground (which was also indicated on the drawing), did not show any support of the roof overhang from the balcony.

I am prepared to accept that these details, albeit shown on the planning drawing and inconsistent though they were, should have put a technical auditor on enquiry as to the precise means of support of the overhang and, in particular, whether it was to be by way of a vertical strut from the balcony or by way of some form of cantilevered beam within the construction of the roof itself. The note on the plan check report to which I have referred above indicates that Campagna did query, or at least did intend to query, the means of support of the roof overhang above the balcony. Drawing number 99.195.3.1.102 was one of those listed in the plan check report as having been received and checked by Campagna.

Mr Scott said, at paragraph 8.14.3 of his report, that "In the event that Campagna cannot prove that the requested information was issued, reviewed and approved by the TA I am of the opinion that the TA should not have issued the Certificate of Approval". Insofar as this is a matter for Mr Scott at all, it seems to me that it simply reverses the burden of proof.

There is no documentation that indicates whether or not Campagna ever received any response to the request referred to in the plan check. If this was how matters stood then, in the light of Mr Hilton's evidence, I would have been reluctant to find that Campagna issued a Certificate of Approval for any of the properties with a cut roof without having received some information about the construction of the roof overhang that had satisfied them that it was appropriate.

However, it appears to be the case, and Mr Hilton accepted this in cross examination, that the Certificate of Approval for No 13 Tayleur's Point followed a final inspection of that house on 14 May 2003 and was issued on 16 May 2003, a week before the plan check. It is clear, therefore, that Mr Hilton could not have had any information about the construction of the roof overhang at the time when this Certificate of Approval was issued, otherwise he would not have made the comment about the overhang that he did on the plan check on 23 May 2003 (because No 13 was one of the houses to which it would have applied).

On 4 March 2004 there was a further plan check for the houses at Tayleur's Point, which listed the same drawings as the earlier plan check report. Mr Hilton was unable to say what the reason may have been for this second plan check. He said that he may have received additional information, but that he did not know. However, and in the meantime, Certificates of Approval for Nos 7, 8, 9, 10, 11 and 12 had been issued between 6 and 12 February 2004.

Since it appears that Mr Hilton must have issued the Certificate of Approval for No 13 before carrying out the first plan check, and the Certificates of Approval for Nos 7-12 before the second plan check, I am not prepared to accept as a matter of inference that Mr Hilton would probably have had a response to his query about the construction of the roof overhang before issuing the Certificate of Approval for any of the dormer houses. Ms Franklin submitted that the absence of any comment about the overhangs on the second plan check indicates that Campagna must have received satisfactory information about the design of the overhang. However, if this was the case, I would have expected that information to be noted on the plan check as part of the information received. It was not. Accordingly, I reject this submission.

In these circumstances I am compelled to conclude that Campagna did not satisfy themselves about the adequacy of the design or construction of the roof overhang before issuing Certificates of Approval for any of the dormer houses. In this they were in breach of their duty to Liberty in that they fell below the standard of a reasonably competent latent defects auditor.

However, I am not in a position to determine whether or not the roof overhang of the dormer houses was in fact unsafe as a structure (as distinct from the tiles on the overhangs being unstable: see paragraph 147). The evidence is inconclusive. In a report dated 23 June 2006 Mr Kerrigan noted that bricks immediately under the soffit boarding of the overhang had become displaced but he could not say that the roof was imminently unstable. Mr Scott accepted that the repairs may have been a “belt and braces” job (Day 11/72).

A report by M T Hamm Ltd, Consulting Engineers, dated 18 July 2006, noted whilst that there was no excessive deflection of the overhang, its fixing was "suspect" and may have allowed the overhang vibrate in high winds. This may have resulted in tiles being blown that were not properly fixed. But I consider that the principal cause of this state of affairs was the poor fixing of the roof tiles, rather than the design of the roof overhang. However, the consequences (in terms of remedial work) of the potentially unsatisfactory nature of the design or construction of the roof overhang is not a matter that I have to decide at this hearing.

The dormer houses had fireplaces in the ground floor living rooms against the side wall and the chimney projected through the lower part of the pitched roof. The effect of the position of the fireplace and the configuration of the roof meant that the chimney had to be fairly high in order to carry away the smoke without causing a nuisance or irritant to the occupiers of the houses.

The chimneys were about 3.7 m high (measured from the top of the chimney to the highest point at which the roof met the chimney) and about 580 mm x 600 mm in external cross section. The Irish Building Regulations 1997 Technical Guidance Document provides that the height of an unsupported chimney should not be more than 4.5 times its lowest horizontal dimensional in wind zone A. Wind zone A is the least exposed of the three zones identified in the document and includes Tayleur’s Point. The document goes on to state:

“The proportions given in this paragraph are intended for general application. More slender chimneys may be built if they can be shown by calculation to be stable in the particular wind environment of the building.”

When Mr Williams was asked in evidence on what basis he asserted that the chimneys at Tayleur's Point complied with the technical guidance he said (Day 8/05):

“It was my experience. I'd seen the chimneys of that slenderness ratio, you know, higher than 4½, up to 6½ times the least width In my working life, which had been subject to building control in the UK. There are some chimneys around Taylor's Point which are still there, have been there for many years. . . . I was using my experience from what I'd seen in the UK. And I am convinced that what was there was satisfactory and if I'd asked for calculations to come in, they would have proven exactly what I was prepared to accept."

Mr Williams was not prepared to be shifted from his position. He was adamant that the chimneys at Tayleur's Point were safe and represented an acceptable risk for the PGS.

The report of M T Hamm Ltd, to which I have referred in relation to the previous defect, concluded that the chimneys were unstable in high wind conditions as a result of their excessive height to width ratio. The report said that they would have to be increased in cross section to 800 mm x 800 mm or, alternatively, reinforced with steel supports. Unfortunately, no calculations were attached to this report. However, the writer stated that the height of the chimney was 3.6 m which, if divided by 4.5, produces a width of 800 mm. This suggests that the writer of the report did not carry out any calculations but simply worked back from the deemed to satisfy provisions in the Technical Guidance Document. In the course of cross-examination Mr Kerrigan said that he thought some calculations had been done by an engineer employed by M T Hamm, a Mr Dolphin, although he had not seen these calculations. I reject the suggestion (made in Liberty’s closing submissions) that M T Hamm’s recommended cross section of 800 mm x 800 mm was the product of any calculation: in my view, it would be a very unlikely coincidence for a calculation to arrive at precisely the same dimensions as given by the formula in the Technical Guidance Document.

RA Duggan & Associates, consulting engineers instructed by one of the residents (Mr Devine, of No 5), wrote in a letter dated 9 February 2007 that the minimum plan dimension of the chimneys should be 825 mm. Again, although this time stated explicitly, this dimension was arrived at by dividing the height of the chimney (3.715 m, according to the writer) by 4.5.

There is no evidence of any failure of any of the chimneys at Tayleur's Point, but on the other hand I have not been told of any contemporaneous calculations that support the dimensions as built. The definition of Major Damage in the policy that was in force at the end of 2003 was as follows:

“Any defect in the design, workmanship, materials or components of the Structure that is first discovered during either the Defects Insurance Period or the Structural Insurance Period and which causes physical loss, destruction or damage or causes imminent instability to a New Housing Unit that requires complete or partial rebuilding or rectification work to the New Housing Unit."

The question here is whether or not the under design of the chimneys has resulted in "imminent instability" to the relevant houses. The writer of the M T Hamm report seems to have thought so (because the chimneys were "unstable in high wind conditions"), but it is quite unclear whether this conclusion was justified. As Mr Mulrooney observed, it is a little curious that these chimneys suddenly became imminently unstable a year or two after they were built.

It is not for Mr Williams, as a technical auditor, to rewrite the advice in the Technical Guidance Document, however much he considers the chimneys to be satisfactory. It seems to me that, in the absence of any evidence to the contrary, a chimney that does not comply with the height to width ratio stated in the Technical Guidance Document should be assumed to present at least a risk of instability. I consider that Mr Williams should have recognised this.

It may be that Mr Williams was right and that, if appropriate calculations had been done at the time, the dimensions of these chimneys could have been justified (I refused a late attempt to introduce recent calculations). There is no evidence that any chimney has failed, or has even shown signs of possible failure (although there is evidence of cracking on some chimney breasts, I do not see how this could be caused by the stack of the chimney being lighter than it should have been). But the fact is that there is no evidence that a contemporaneous check was carried out at the time when the chimneys were being built: although Ms Franklin submitted that the certifying architects must have had calculations done, there is no evidence of it. Mr Mulrooney accepted that a design check on the adequacy of the chimneys should have been carried out. I agree. In my judgment Mr Williams had one of two choices if he was to discharge his duty to Liberty. Either he could arrange for the dimensions of the chimneys to be verified by calculation (Campagna had its own in-house engineer who could probably have done this), or he should have communicated his views to MDIS and left it to them to decide whether or not they wish to seek advice from a consulting engineer.

Campagna did neither of these things because Mr Williams was so confident that the chimneys were adequate. As I have said, it was not for Mr Williams to substitute his own judgement for the guidance given in the Technical Guidance Document. Accordingly I find that Campagna was in breach of its duty to Liberty in failing to take any steps to have the structural stability of the chimneys checked by calculation or to report to MDIS the fact that the chimneys did not comply with the required height to width ratio. In my judgment in this respect Campagna fell below the standard of a reasonably competent latent defects auditor.

This defect concerns Nos 100, 101 and 106, which are terraced houses. They are built of cavity blockwork walls which are finished with an external render coat. The defect - and it is not disputed that it is a defect - consists of cracking of the render, particularly at corners. The blockwork at the corners of the building was finished with a steel angle bead and the cause of the cracking is corrosion of the steel of the angle bead.

This corrosion was the result of water penetration behind the render. In Mr Scott's view, the water penetration occurred because the render coat was applied after the fascia boards and gutters had been installed instead of the other way round. The effect of this sequence of work is that it creates a joint between the render and the fascia board behind the gutter through which water can penetrate. Water could then run down the face of the blockwork behind the render and into contact with the steel angle bead. As is well known, when iron corrodes it expands. The expansion of the angle bead caused the render to crack. This then permitted the ingress of further water which then aggravated the problem.

This was not a topic that was explored at great length during the trial. Mr Scott's analysis of the cause of the cracking of the render was not challenged and I am satisfied that it is correct. I reject Mr Mulrooney’s assertion that the incorrect sequence of work has not resulted in any damage. I am satisfied also that to fix fascia boards and gutters before applying the render coat is poor practice and should not have been done. I consider that any reasonable technical auditor ought to have appreciated that. In his two reports Mr Mulrooney did not express an opinion on this point. In cross examination Mr Quiney took the position that since Mr Mulrooney had not expressed any opinion as to what Campagna should or should not have done, he was not required to cross-examine Mr Mulrooney on the subject. I consider that Mr Quiney was entitled to take this position: it was not for him to elicit Campagna's case for them.

The real issue in relation to this defect is whether or not Campagna should have picked up during the visits to the site that an incorrect sequence of work in relation to the fitting of the fascia boards and gutters and the render coat was being adopted in the case of these houses.

There is no evidence that Campagna carried out a plan check for these houses. But there is also no evidence that there was any written specification in relation to either the choice of the steel angle bead or the sequence of work. Whilst angle beads are made in other materials I do not understand it to be alleged that it was negligent of Campagna's technical auditors to permit the use of steel angle beads as opposed to angle beads made of some other material. The case is really put on the basis that the render coat was self-evidently a critical element of the weatherproof envelope and so its application should have been the subject of specific attention by the technical auditors.

Nos 100 and 101 Tayleur's Point were inspected (after the foundation stage) during site visits on the following dates:

No 106 Tayleur's Point was similarly inspected during site visits on the following dates:

None of the above inspections was described as a pre-plaster inspection, and to this extent it seems that Campagna was in breach of its obligations under the 2004 Retainer. However, I doubt if that fact has any relevance to Campagna's failure to discover or address the wrong sequence of work in relation to the render coat. I consider that it was very likely that during the inspections of Nos 100 and 101 on 3 March 2005 it would have been apparent that the render coat had been or was being applied after the fascia boards and gutters had been installed. This is because it is likely that the fascia boards and gutters were fixed as soon as the roofing was complete and whilst the scaffold was still in place, so Campagna's technical auditor would have seen the render coat being applied at a time when the fascia boards and gutters were already in place.

In the case of No 101, there was a particular defect in the render at the end of the roof gutter over the front bedroom. The render was not continued behind the stop end of the gutter at an internal corner in the exterior of the building, because the latter had been installed first and the render completed around it.

In cross examination Mr Gough was very reluctant to accept that applying the render after the fascia board had been installed would give rise to any serious risk of leakage (because he said that the joint was not exposed, being in his view protected by the roof overhang and the gutter), but he did agree that it was not good practice. I consider that Campagna's technical auditors should have appreciated that the fascia boards and gutters should not have been installed before the render coat and that this should have become apparent during the inspection of No 100 on 3 March 2005. In addition, there are photographs which clearly show the render overlapping the end of the fascia board which indicates that it was applied after the fascia board had been installed. In my view a careful examination on a final inspection, if not earlier, should have picked this up.

Campagna's technical auditors should have appreciated that applying the render after the installation of the fascia boards and gutters so as to leave small areas of brickwork not protected by the render would present weaknesses in the external weatherproofing. If the problem had been noticed then, Campagna should either have made it clear to the builder that the Certificate of Approval would be withheld unless the fascia boards were removed so that the render coat was applied to the brickwork underneath it, or have reported the matter to MDIS for its view as to what should be done. At the very least, if the matter had been drawn to the attention of the builder the incorrect sequence of work would not have been repeated on the houses to which the render coat had not been applied at the time of the inspection on 3 March 2005.

For these reasons I consider that in failing to appreciate and remedy the consequences of applying the render after the installation of the fascia boards and gutters Campagna fell below the standard of a reasonably competent latent defects auditor.

This defect arises because of the way in which the render and coping stones have been installed on top of the parapet wall of No 106. In effect, the render has been applied after the coping stone, and therefore the top of the render forms a ledge. Water has been able to penetrate behind the render and compromise the waterproof envelope.

I am not satisfied that this is a detail that a reasonably competent technical auditor would have picked up. It is not an inevitable consequence of the sequence of the work; rather it is a function of the dimensions of the coping stones. Mr Scott did not address this defect in isolation and I do not consider that Campagna were in breach of duty in failing to notice it.

In October 2007, following reports of water ingress to some top floor apartments at Mountgorry Wood, the roofs were inspected. It was found that there was extensive damage to the roof covering. The roof of Block C was particularly bad. It is common ground that this damage was caused by people walking on the roof after it had been laid: the issue is how and when this happened.

The roofs have a very low pitch: according to the plans, it is about 10°. It has been measured on site as between 10°-12°. The lowest two or three courses of the roofing material have an even lower pitch, of 5°-7°.

There is some doubt about the type of roof covering that was originally specified and shown on the drawings for these blocks. There was a discrepancy in the specification because it appears to have provided for two forms of roof covering. The written specification provided for slates/tiles on 50 x 50 mm battens and sarking felt fixed to pre-fabricated roof trusses, whereas the relevant elevation drawing indicated that the roof covering was to be “Selected sheet-metal deck roof to manufacturers detail and specification”.

In the event the product that was eventually selected was known as Decra Parc Tile. According to the BBA Detail Sheet (No 95/3122) these tiles may be installed on conventional steel or timber structures with a minimum pitch of 12°. In fact they are not tiles in the ordinary sense of the word, but steel sheets moulded to a shape simulating 8 conventional tiles and measuring about 1.3 m x 415 mm. These sheets are laid on battens and are interleaved along the horizontal edges, in that the higher sheet has a downstand that overlaps an upstand in the sheet below it. However, I consider that the Decra Parc Tile corresponds with the description of a “sheet-metal deck roof” indicated on the drawing.

It is not clear why Decra Parc Tile was chosen for a roof with a 10° pitch when there were other roofing products manufactured by Decra that would have been suitable. It is quite clear that Campagna were aware of the fact that the roofs of the blocks were designed with a 10° pitch, because that is recorded in their documents. It is well known that the pitch of a roof will determine the choice of material with which it must be covered (or the method of fixing). Roofs with very low pitches are prone to wind damage so that, for example, ordinary tiles that are simply held in position by gravity are unlikely to be suitable unless the building is in a very sheltered location. I consider that competent technical auditors in the position of Campagna should have satisfied themselves that the selected roof covering was suitable for these particular roofs. It should have required no more than a telephone call to Decra, or a visit to Decra’s website, to establish that the Parc Tile was not recommended for a roof with a pitch below 12°. Campagna’s failure to do this was a breach of its obligations to Liberty in that it fell below the standard of a reasonably competent latent defects auditor.

However, it is not clear whether or not Decra Parc Tile roofing is in fact unsuitable for these buildings. When the repairs were carried out to rectify the damage the replacement was not total and the damaged parts of the roof were replaced with the same sheeting. There is no evidence before the court as to the reason for this decision, but it is a reasonable inference that it was considered by those acting for Liberty that to replace the damaged sheets by new sheets of the same type would not only be cost-effective but also be suitable for the conditions. In addition, there is no evidence that the choice of roofing material has permitted any water penetration. The water penetration that occurred is entirely explicable by the physical damage to the roof sheeting and poor detailing.

A refinement to the problem about the choice of roof covering is presented by the fact that the lowest layers of the sheeting are at an even lower pitch, about 5-7°. This section of the roof consists largely of the overhang and so, from a weatherproofing point of view, is probably less critical than that the remainder of the roof. The evidence suggests that this section of the roof does not share the main roof framework but rests on a specially constructed extension attached to the main roof trusses. Once again, when the repairs were carried out to these lower sheets the detail was not changed and the damaged sheets were simply replaced by new sheets to the same specification. I do not consider that the presence of this feature adds anything of significance.

If Campagna had made the enquiries of Decra that I consider that they should have made, it is likely that one of the other Decra products would have been chosen instead of the Parc Tile. Whether or not this would have made any difference to the resultant position is a matter of speculation. I am inclined to doubt it. This is because I assume that the alternative products would have been just as vulnerable to damage by foot traffic as the Parc Tile, with the result that - if the damage was caused by foot traffic - it would have been likely that alternative roof decking would have been damaged to much the same extent as the damage to the Parc Tile sheeting which actually occurred.

The roof should be laid from top to bottom (unlike some conventional tiles) and the sheets are nailed - roughly horizontally - to the battens where they overlap. Once laid the Decra roofing can be walked on, but only if care is used to walk along the line of the battens where the sheeting is supported. If a person walks on the sheeting where it is unsupported it will be damaged and, very likely, the underlying felt also.

There were complaints of water ingress to second floor rooms in blocks B and C at Mountgorry Wood. Subsequent investigation revealed that the Decra roofing was very badly damaged, probably by foot traffic. The sample properties with which this hearing is concerned and which suffered from water penetration through the roof are apartments 37 and 39, both of which are in Block C. Accordingly in this judgment I will concentrate mainly on the roof of Block C.

Only three potential causes of this damage to the roof of Block C have been identified. The first is storm damage. The second is damage during construction. The third is damage post-construction caused, it has been suggested, by maintenance men carrying out repairs. The evidence of Mr Aidan O'Connell, an engineer, was that storm damage could be eliminated. That has not been challenged. Accordingly, I eliminate it also as a potential cause. This leaves damage by foot traffic: the question is whether it occurred during construction or at some later stage.

As to damage during construction, it is on the face of it unlikely that any competent roofing contractors could cause such extensive damage and then leave the roof in that damaged condition. If the damage was caused during construction, then it would be likely to have occurred only if the Decra sheeting was laid from the bottom up, because then the lower sheets might be walked on whilst the upper sheets were being laid. If the Decra sheets were laid from the top down, as the manufacturer recommends (according to Mr Mulrooney, and I have no reason to doubt his evidence on this), then extensive damage by foot traffic during construction would seem to be unlikely.

A further difficulty with the hypothesis that damage was caused during construction is that the first sign of water ingress appears to have been in about the spring of 2006. This is on the basis of a statement by the owner of No 37, a Miss McCann, who was said to have moved into the property in December 2004 and told a member of Mr O'Connell's firm in November 2007 that the water penetration first occurred about 18 months earlier. Mr O'Connell’s evidence as to how long it would have taken before the effects of water penetration would have been seen was as follows:

“There would certainly have been an ingress of water, in my opinion, yes. Water, if you have significant rainfall, and we do, certainly in that area of Ireland, is quite close to the sea, there would have been, and I would have expected that there would have been an ingress of water, most definitely. Whether it would have manifested immediately internally within the apartment, that would definitely not be the case, but very shortly thereafter, yes.”

If Miss McCann moved into the property in December 2004 but did not see any signs of water penetration until the spring of 2006, it seems that two winters must have gone by before any signs of water penetration appeared. This assumes, of course, that she is correct in her recollection as to when signs of water first appeared. However, she said that the first leaks had been attended to and repaired by the developer. The next signs of water penetration appeared in October 2007, as a result of which Mr O'Connell's firm was called in. Interestingly, this was also about 18 months after repairs had apparently been carried out to deal with the first incident of water ingress.

The roof space above Miss McCann's apartment was inspected in December 2007 by an engineer from Mr O'Connell's firm. It was recorded that there was a noticeable leak on the felt in the attic above the living room, and a lack of fibreglass insulation above the ceiling in some parts of the roof space. It is possible that the latter was the result of the removal during the previous repairs of insulation that had become damp and the failure to replace it.

On the basis of this evidence it seems to me that it is at least possible that the leaks through the roof took rather longer to manifest themselves in the apartments below than Mr O'Connell thought likely.

If, on the other hand, the damage was caused at a later stage by workmen carrying out repairs to the roof, then it would seem likely that these repairs were carried out during the winter of 2005/2006, if Miss McCann's dates are correct. It is not clear what repairs might have been carried out or whether they involved anyone going up onto the roof itself. One of Liberty’s witnesses was a Mr Cleary, who worked for OSG, the loss adjusters. He was not called but certain parts of his first witness statement were agreed. These included a paragraph in which he said that in August 2008 he contacted the management company responsible for the development asking for details of any post-completion maintenance that had been undertaken on the roof of Block C. He was told by the management company that they had no records of any post completion maintenance on that roof. This, of course, is not conclusive, but it does not assist the damage by maintenance post-construction hypothesis.

A further difficulty with the hypothesis that the damage was caused during the carrying out of repairs is that the point at which it is suggested that repairs may have been carried out on Block C is close to the parapet at the end of the building. I would have thought that the natural method of access would be by way of a ladder positioned against the wall close to the point where the repairs were thought to be required, in which case workmen would have no need to walk on any area of the roof apart from that in the immediate region of the repair. This is not consistent with the observed damage.

Neither of these hypotheses is a clear favourite. Each faces the difficulties that I have indicated. It seems to me that the best clue lies in the nature of the damage itself. Taking the roof of Block C, it is in effect fairly random. It is not consistent with men gaining access to the roof at point A and walking across it in order to carry out repairs at point B. Further, the most extensive area of damage is along a section of the roof bordering the gutter extending up the roof for about four sheets of the Decra material. As Mr O'Connell observed, no one in their right mind would walk along the overhanging edge of this roof if there was no scaffold below. If this damage was caused post-construction, it seems to me that it would have had to have been the result of drunken vandals rather than of traffic by men carrying out repairs. But vandals would have had no means of access to the roof (which is doubtless why they have not been put forward as a possible hypothesis).

It seems to me that the observed damage would only be consistent with careless walking on the roof during construction if the sheets were laid from the bottom up which, to someone not familiar with the product and the techniques of laying it, might seem a natural way of doing it (as Mr O'Connell thought was the case, albeit wrongly as it happens). A contractor who was capable of allowing his operatives to damage the roof whilst working on it is, in my view, all too likely to have been capable also of adopting the wrong method of laying it. In other words, to have laid it from the bottom up instead of from the top down, as Decra recommends.

The roof of Block C was not the only one that was damaged in this way. Similar damage was seen on Blocks A and B, but it was not as severe as that on Block C. There is no evidence as to whether or not there was any similar damage to Block D, but it appeared to be accepted on both sides that if there was any similar damage to Block D it was fairly slight. If the damage was caused during the carrying out of repairs, then it would follow that each of the three roofs of Blocks A, B and C must have independently required repair within two or three years of construction. However, if this happened it could have been purely coincidental.

The sequence in which the roofing was carried out to the various blocks may be of some relevance. As can be seen from the table below, the roofing to Block D was carried out in September 2003. Block C was probably roofed in early 2004 (this can be deduced from the facts that by the end of September 2003 construction was only up to first floor level and that "plastering out" took place in February/March 2004). The roof trusses for Block B were installed in March/April 2004 and by August 2004 the second fix was in progress. This suggests that the roofing to Block B probably took place during April/May 2004. The roofing to block A took place in August/September 2004. From this timetable it seems to me at least possible, but I can put it no higher than that, that the roofing of Block D was carried out by a different team to those that roofed Blocks A, B and C. This could account for the apparent lack of damage observed to the roof of Block D.

As a matter of probability, therefore, I conclude that all, or at least the major part, of the damage to the roof of Block C was done during construction by a careless and inexperienced contractor.

The next question is whether the damage should have been apparent to a reasonably competent technical auditor. The relevant extracts from the site inspection reports are set out below (omitting those stages before roofing started):

In relation to Block A, at the time of the visit on 2 September 2004 the report described "roofing in progress", which must refer to the laying of the Decra sheeting because the rafters were already in place. Mr Mulrooney accepted that if there had been an inspection from a scaffold by a technical auditor, he should have seen the damage if it had occurred by the time of the inspection.

Unfortunately, the site inspection report gives no indication as to how much of the Decra sheeting had been laid at the time of the inspection. It could have been just the lowest two layers, or it could have been most of the roof. There is no way of knowing. If, at the time of the inspection, the roofers had only just started to lay the Decra sheeting, then it is unlikely that there would have been any appreciable damage. By contrast, if the work was nearly complete then the damage would have been there to be seen. Accordingly, so far as Block A is concerned, it is impossible to say whether or not there would have been any damage to the roof that a technical auditor could have seen at the time of the visit in September 2004.

In relation to Block B, if the roof trusses were being installed on 2 April 2004, then the laying of the Decra sheeting could not have begun. On the next visit, 3 August 2004, the second fix was in progress. I would not expect this to require scaffolding, so I think it likely that the scaffolding to Block B would have been struck by the time of this inspection.

Turning to Block C, on 3 March 2004 the Block was being plastered out, so the roof must have been weathertight. One month later, on 2 April 2004, the work was described as nearing completion. On 3 August 2004 the work was described as being almost at completion, with making good to external brickwork still required. Mr Mulrooney said that he inferred from this that they might be striking the scaffold at that point because the making good of the brickwork is the last thing to be done. Access to the external brickwork at first floor level would require the presence of the scaffold at that level, and so the making good of the brickwork would be done whilst the scaffold was being taken down. Mr Quiney suggested to Mr Mulrooney that this would indicate that the scaffold was very probably still in place at the time of the previous inspection on 2 April 2004, some four months earlier. Mr Mulrooney accepted this as a possibility.

On the basis of these reports, it is not possible to say whether or not there would have been any access to the roofs at Blocks A or B that would have enabled a technical auditor to inspect the roof at a time when most of the Decra sheeting had been laid. So far as Block C is concerned, taking this evidence as a whole, I find that the scaffold probably was still in place at second floor level to at least a part of the perimeter of the building at the time of the inspection on 2 April 2004. By that time the roof of Block C would have been complete and the damage would have occurred. Of those three blocks, the damage to Block C was the most severe and is readily visible in the photographs. Accordingly, a reasonably competent technical auditor who went up a scaffold should have seen it.

If I am wrong about this, the question remains as to whether or not the damage to the roof of Block C should have been noticed at the time of the final inspection on 2 September 2004. This inspection would have been at a time when there was scaffolding to Block A only, and it is not suggested that it would be reasonable to expect a technical auditor to attempt to inspect the other roofs from a long ladder (even assuming there was one available). I would reject also any suggestion that such an inspection should have been carried out from a cherry picker. There is no evidence that there would have been a cherry picker on-site and this is not a cost that Campagna could be expected to incur.

In my view, the real question is whether at the final inspections Campagna’s auditors should have been equipped with binoculars with which to inspect the roof of the building being inspected. The reality is that there was no other method of inspecting the roofing if there was no means of access to it. Mr Scott said, and I accept, that he would not expect a technical auditor to carry a pair of binoculars as part of his usual equipment when visiting a site. But this still leaves the question of whether binoculars should be taken when carrying out a final inspection of any property, if that includes the roof, in circumstances where there is unlikely to be any means of access. On this, Mr Scott said in his report that where it would not be possible to see the roofs from ground level except at a distance, the technical auditor would require the use of binoculars to assess the roof covering and the component parts of the roof (paragraphs 5.9.7 and 5.9.8). In his oral evidence, at Day 10/146, Mr Scott was rather less emphatic. He said:

“Well, I think if I knew that I couldn't see the roof, having been to site on numerous occasions previously, then it may be something that might be worthwhile bringing with him, but I wouldn't necessarily say it formed part of their daily kit."

The documents make it quite clear that one of the purposes of the technical auditor's role is to take reasonable care to check that the external envelope of the building is weathertight because damage by water penetration is an insured risk. In order to achieve this I consider that the technical auditor must carry out some inspection of the roof of the property at the time of the final inspection, both external and internal (so far as access permits).

I would expect a technical auditor going to Mountgorry Wood (or any other site where roofs might have to be inspected from the ground) on a final inspection to carry a pair of binoculars because he could not rely on the availability of any means of access to roof level. An inspection using a pair of binoculars would be far from perfect, but I am quite satisfied that any serious defects, such as the damage to Block C, in particular, would have been seen through binoculars. Mr Mulrooney accepted, perhaps inevitably in the circumstances, that if a technical auditor had seen the state of the roof of Block C as it is shown in the photographs, he should not have issued a Certificate of Approval.

In addition, I consider that a careful examination of the roof of Block C with a pair of binoculars would have revealed also the loose felt which was flapping free at the base of the glazed roof light. Again, Mr Mulrooney accepted that if a technical auditor had seen the state of that felt on Block C as it is shown in the photographs, he should not have issued a Certificate of Approval.

For these reasons I consider that in failing to identify the damage to and defects in the roof of Block C, Campagna fell below the standard of a reasonably competent latent defects auditor and was in breach of its duties to Liberty.

At each end of each block there was a low parapet along the side of the roof which was capped with coping stones. The Decra sheeting was turned up where it abutted the parapet. Lead flashing was tucked horizontally under the coping stones and folded down over the turned up ends of the Decra sheeting to provide a waterproof barrier. The end wall of each block consisted of two leaves of brickwork separated by a cavity. At the top of the brickwork, bridging the cavity, there was to be a horizontal damp proof course (“DPC”) laid immediately under the coping stones. Two or three courses below that there should have been a cavity tray built in to the outer leaf of brickwork and then folded up at an angle and built in to the brickwork of the inner leaf one course higher up. No such cavity trays were installed and the DPC was badly laid and, as a result, there was water penetration into the interior of the building through the upper part of the gable wall. This affects apartment No 24 (Block B).

It follows that the cavity tray had to be installed during the construction of the top few courses of brickwork of the parapet wall. The horizontal DPC at the top of the parapet wall did not have to be installed until immediately prior to the installation of the coping stones. During the period following the construction of the cavity walls to their full height (including the element that constituted the parapet) there would be a period before the final DPC was laid at the top of the wall during which the cavity tray would be clearly visible to anyone who chose to inspect the wall.

For ease of reference I will set out again the relevant parts of the table summarising the inspections carried out at Mountgorry Wood.

Mr Mulrooney accepted that when the roofing was in progress at Block A the gable end parapet walls must have been completed to their full height (Day 12/178), so that anyone who inspected them from a scaffold would have been able to see whether or not the cavity tray at the top of the parapet wall was in place. By the same reasoning, it would have been possible to inspect the parapet walls of Block B at the time of the inspection on 2 April 2004. In each case, it would have been possible to see that the cavity trays that should have been installed just below the top of the cavity walls were not there. This did not require any opening up beyond the removal of whatever temporary cover had been placed over the top of the cavity. In my judgment, Campagna's technical auditor should have noticed the absence of the cavity trays in the parapet walls at the gable ends of Blocks A and B.

In the case of Block C, the scaffold would have been in place at the time of the inspections on 3 March and 2 April 2004, as I have already found, but it is not possible to say whether or not the coping stones would have been laid so as to close the gable end parapet walls and therefore prevent inspection. However, since I consider that the technical auditor should have noticed the absence of the cavity trays to the parapet walls on Block B on to April 2004, he should have been put on enquiry about the possible absence of the cavity tray in the corresponding parapet walls of Block C also. If, by the time of the inspection, the coping stones had already been laid, he should have told the developer to open up the cavity for inspection (by removing a couple of coping stones and the horizontal DPC), failing which the Certificate of Approval would be withheld.

By contrast, it would not be possible to tell whether or not the DPC that bridged the cavity at the top of the parapet walls had been properly laid unless the technical auditor was physically there at the time when it was being laid, because it would have been covered up almost immediately as the coping stones were laid on top of it. The DPC could only be inspected thereafter by removing one of the coping stones. I assume that the DPC would have been laid directly onto the top of the brickwork and the lead flashing folded down on top of it, rather than the other way round - but, whether or not this happened, I doubt very much if the DPC would have been visible once the coping stones had been laid, however carefully and closely the wall was examined.

There is no evidence that any technical auditor was on site at the time when the parapet coping stones were being installed. In any event, the odds would have been against it. Taken by itself, the absence of a proper DPC at the top of the parapet wall was not something that a reasonably careful technical auditor could have been expected to discover, unless the DPC had been laid and there were grounds for requiring the cavity to be opened up in the circumstances that I have just described.

I have already held that I do not consider that the technical auditor was under a general duty to require the opening up of any work that he had not had an opportunity to inspect. However, if, in one or other of the circumstances that I have described above the technical auditor had noticed the absence of the cavity trays and had told the developer that there would be no Certificate of Approval unless the upper three or four courses of the brickwork of the parapet walls were dismantled and proper cavity trays installed, the developer would almost certainly have complied. It may be that the contractor would have been sufficiently jolted by this that he would have taken steps to ensure that the rest of the details at the top of the parapet walls (such as the flashings) were properly installed. But, leaving this aside, I am satisfied that if there had been properly installed cavity trays at the top of the parapet walls, the water ingress that occurred in that area would have been substantially, if not wholly, prevented. Accordingly, in failing to notice the absence of these cavity trays Campagna fell below the standard of a reasonably competent latent defects auditor. Accordingly, the consequent claims for water penetration to the relevant apartments were the result of Campagna’s breach of its obligations to Liberty.

For part of the front elevation of each of the blocks the wall was constructed of exposed brickwork. That wall was carried up above roof level (by about a metre) to form a parapet. It should have had a cavity tray built in just above roof level. This affects apartment No 40 (Block C).

The evidence, which is not conclusive on this point, suggests that this cavity tray may not have been installed. One of the parapets was inspected and two bricks were removed from a course just above roof level and no cavity tray or other form of damp proof course was seen. For the purpose of this judgment, I will assume without deciding that the required cavity tray was not installed.

It is evident from the nature of this parapet wall, and this was confirmed in evidence, that its construction from roof level to full height would only take a matter of days. It would be pure chance if one of Campagna’s technical auditors happened to be on site whilst it was being constructed. Unlike the small parapets at each end of each block, this parapet wall did not have to be left open after it had been built up to its full height. This is because it was not capped with coping stones which could not be laid until the roof covering and associated flashing was complete, but was instead finished at the top by a course of bricks laid at right angles to the line of the wall - rather like a typical brick garden wall. Accordingly, once the brickwork had been completed, any cavity tray or damp proof course would in all probability no longer be visible. As I have already found, in such circumstances I do not consider that a technical auditor could be expected to require the builder to open up parts of the completed parapet wall so that he could check for himself that the cavity tray had been installed. Unless he had a good reason for suspecting that the cavity tray might have been omitted, to require opening up would almost certainly engender hostility from the builder (who would probably have no contractual remedy for the costs or delay incurred as a result, since it was not being required by his employer) and thereby damage the goodwill of the PGS.

There is absolutely no evidence that any of Campagna’s technical auditors was on site at the time when these parapet walls were in the course of construction. In these circumstances there is accordingly no basis upon which Campagna can be criticised for not having seen the absence of the cavity tray in these parapet walls during the site inspections.

A more difficult question is whether Campagna should have had any reason to suspect that the cavity trays had not been installed in these parapet walls. If Campagna had noticed, as I have found they should, that the cavity trays had not been installed in the low parapet walls at the ends of the blocks, then it might be said that they should have considered the possibility that the cavity trays might have been omitted in these parapet walls also.

This was not an aspect that was dealt with in the report of Liberty’s expert, Mr Scott. His report addressed only the absence of the cavity tray and the horizontal damp proof course in the low parapet wall at the gable end of Blocks B and C. He did not consider the brick parapet wall on the front elevation of the blocks. He therefore did not consider the question of whether or not a competent technical auditor should have been put on enquiry as to whether or not the cavity tray had been installed in the front parapet walls if he had discovered that they had not been installed in the gable end parapet walls.

In the absence of expert evidence I do not consider that this is an issue that I should resolve against Campagna. It is not obvious that the absence of a cavity tray in the gable end parapet walls made it likely that no cavity tray had been installed in the front parapet walls. It is entirely possible that, if the relevant enquiries had been made at the time by the technical auditor, it would have emerged that the contractor had not been told to install a cavity tray in the gable end walls. In this situation there would be no reason to assume that cavity trays had been omitted in any other location where they had been specified. Their omission in the gable end parapet walls would thus have been the result of an omission in the design or specification for which the builder was not responsible. There is no evidence as to what the contractor was or was not told to install by way of cavity trays in the gable end wall and consequently there is no basis for any finding that Campagna’s technical auditors should have been put on enquiry by the absence of cavity trays in the gable end walls as to the possibility of a similar absence of cavity trays in the front parapet walls.

Two principal defects are alleged, one of which was a direct consequence of the damage to the roofing by foot traffic which I have already discussed. That is the damage to the sarking felt below the Decra Park Tiles, which had sagged in various places. This was said to have been the result of the damage to the roofing sheets and I did not understand this to be challenged (but it was certainly Mr Hilton's view - see Day 7/143-144).

The second defect related to apartment No 24 and consisted of a failure to connect the ventilation extract from the bathroom to the ventilation outlet roof tile in the roof.

I consider that it was the duty of a technical auditor carrying out a final inspection to go into the roof space of the property concerned. In the case of all the sample properties there was some form of access to the roof space usually by means of a hatch in the ceiling of a room below (although there was evidence that once the properties were inhabited and furnished in some cases these hatches were not readily accessible). I understood Campagna to accept that a technical auditor should have looked in the roof space on a final inspection if there was a ladder available. In my judgment, this concession does not go far enough. As I have already said, I consider that Campagna's technical auditors should have taken steps to ensure that a suitable ladder was made available at final inspections. It only required a telephone call to the developer to make it clear that if there was no ladder available, there would be no Certificate of Approval: in that event, either ladders would then have been provided or the property would not have been insured under the second stage of the PGS.

In the case of Block C, I consider that if a technical auditor had done nothing more than put his head into the roof space he should have noticed that the sarking felt was hanging down in places. I accept the evidence of Mr Mulrooney that it would not necessarily have been evident whether or not the felt was properly taped at the joins or overlaps (Day 12/189), but this does not really affect the position. Once a technical auditor had noticed sagging felt, he should have drawn the position to the attention of the developer so that he could put it right. In addition, if a technical auditor had noticed the sagging felt he ought to have been put on inquiry that this might have been the result of some damage to the roof in the relevant place and that this possibility required investigation. This is a further reason why I consider that a competent technical auditor should have discovered the damage to the roof of Block C before he issued a Certificate of Approval.

I am not satisfied that the failure to connect the soil pipe extract ducting to the ventilation tile would necessarily carry a risk of water ingress: that implies that the vent tile itself was inappropriately designed. However, if the vent tile itself was not properly installed, that would involve a risk of water ingress.

From the evidence in the reports of OSG and Aidan O’Connell and Associates I am unable to reach a conclusion as to whether or not both of these defects existed. If it was only the former, I am not satisfied that it carried a risk of a claim under the policy. Accordingly, Liberty has not satisfied me that the existence of this defect involved any breach of duty by Campagna.

The two sample properties that are concerned with this defect are apartment 10 in Block A and apartment 24 in Block B. The balconies partly project beyond the face of the building and are partly recessed, so that access can be obtained to the balcony from both of the two bedrooms and from the living room. The balconies are formed by grooved timber decking supported in part by the extension of the floor slab to the face of the building and thereafter by an external steel framework.

The timber decking is fixed so that it goes right up to the frames of the balcony doors or French windows. There was water ingress into the apartments because water collecting on the slab could penetrate under the frames of the doors and windows and thus under the internal flooring that was supported off the same slab. There were several unsatisfactory features about the balconies as actually constructed, but the essential design fault was said to be the lack of any suitable upstand from the concrete slab below the timber frames of the doors and windows. The problem was compounded by the fact that the gutters above the balconies were inadequately sized to discharge the water that collected on the roofs, so that at times of heavy rain water cascaded down from the gutter onto the balcony or balconies below.

Following a report in about July 2007 of dampness in apartment 10, which was sold in 2005, it was inspected by Mr Cleary of OSG in October 2007. The damage was reported as water damage to the laminate floors in the living room and the two bedrooms which gave onto the balcony. A further inspection was carried out by Mr O'Reilly of Aidan O'Connell and Associates on 31 October 2007. It was noted that the finished floor level of the external timber decking was about 20 mm higher than the finished floor level of the internal floors. There were unacceptably high levels of moisture in the internal floors and walls in the parts closest to the doors to the balcony. Aidan O'Connell and Associates were unable on a visual inspection to say precisely how the moisture was getting in to the property.

Aidan O'Connell and Associates carried out a further inspection of apartment 10 in December 2007 during which they took up some of the timber decking where it abutted the doors and windows. This revealed that the timber decking was screwed directly into the slab below through the damp proof membrane which had been laid between the decking and the surface of the slab. This membrane was turned up approximately 10 mm at the door threshold. Another defect was that the door to the living room was not mounted on the outside edge of the internal wall as shown on the plans, but on the inside edge. This left part of the end of the internal wall exposed to the weather.

From a visual inspection of the balcony floor as constructed all that could be seen was a sealant joint between the decking and the door frame. There was about 30 mm of the bottom section of the door frame projecting above the level of the decking. The flooring on the inside was laid flush to the insides of the door frames.

The principal complaint in relation to these arrangements in apartment 10 is that there was no upstand on the floor slab at the line of the door and window frames. This would have provided a physical barrier to the migration of water along the upper surface of the slab from the outside of the building to the interior. Instead, the barrier as installed was some form of mastic joint.

Apartment No 24 was the subject of a complaint about the ingress of moisture into the living room through the gable end wall. This was a result of the failure to insert a cavity tray near the top of the gable wall and a properly laid horizontal DPC to seal the cavity under the coping. I have already discussed these defects. However, the investigation of those defects seems to have provided the opportunity to enable inspection by the experts of the construction of the balcony and the waterproofing details at the point where the balcony decking meets the lower sections of the door and window frames. Apartment No 24 is a first floor apartment in Block B, which is a two storey building. The balcony to the apartment is therefore immediately below the gutter.

The experts were able to lift sections of the timber decking, which revealed that it had been fixed directly onto the structural slab with long screws screwed into a long “rawlplug” in the slab. There was no form of DPC or other waterproof membrane in spite of the fact that the slab was also the ceiling of the apartment below, although there was some form of plastic or bitumen damp proof membrane applied to the top surface of the slab in the region of the door threshold and a stepped lead flashing at the threshold. The joint between the slab and the underside of the door and window frames was again formed by mastic. There was no form of upstand. Rather surprisingly, and unlike Apartment No 10, there were no signs of dampness to the interior of apartment No 24.

Two things are evident from this inspection. First, there was no consistency of detail in the construction of the balconies in the different blocks, and in each case the workmanship was not of good quality. In Block A there appeared to have been some effort to introduce a membrane between the decking and the slab, whereas there was no such attempt in Block B. Second, the detail in the apartment No 24 appears to have worked as a waterproof barrier.

I do not find it entirely surprising that the detail in apartment No 24 seems to have been effective. In his report, Mr Mulrooney included an example of a standard detail for the junction between the slab and a door frame at the edge of a balcony. This was not the same type of threshold detail as that involved at Mountgorry Wood, but it is an example of a design where the barrier to prevent the ingress of water between the underside of the door frame and the slab consists solely of a silicone seal: on my reading of this drawing there is no form of upstand in the slab (I do not agree with Liberty’s suggestion that the slab is stepped - I take the higher level shown inside the threshold to be the applied flooring).

In his evidence Mr O'Connell agreed that there was nothing about the detail of the balcony as constructed that would have provided any indication that there was not a proper detail underneath the decking or at the thresholds. He said that the only features that were unusual were the fact that the internal finished floor level was slightly below the level of the decking on the balcony and that there were no weep holes at the bottom of the door and window frames. Nevertheless, Mr O'Connell said that what was there to be seen would not necessarily have rung any warning bells. As Mr O'Connell pointed out, they had to open up the detail in order to see what was there. There is no evidence that at the time of any of the inspections at Mountgorry Wood there was a balcony in the course of construction. In his report (at paragraph 7.11.18), Mr Scott said that "one can only make a considered guess that the construction of the balconies was available to inspect prior to final inspection at some stage to any one of the 4 Blocks". The result of a “considered guess" is not a basis on which I can find as a matter of probability that a balcony was in the course of construction at one of Campagna's visits.

It is, I think, common ground that the drawings that were made available to Campagna did not provide any information about the balcony threshold detail. If Campagna's technical auditor had asked the developer what form of barrier against water ingress was to be provided between the slab and the underneath of the door and window frames and had been told that it was to be by way of a silicone joint without any form of upstand, he would in my judgment have been entitled to treat that as a satisfactory detail.

If he had gone further and had asked, at the relevant stage during the construction of Block D (which was the first to be built) whether there was to be a waterproof membrane between the balcony decking and the floor slab, I consider it likely that he would have been told that a membrane was to be installed because that is what good practice would have required (as well as being what was subsequently done in Block A, the balconies of which were probably still to be constructed at the time of the plan check). I would not have expected the builder or developer to tell the technical auditor that it was not proposed to install any form of waterproof membrane between the slab and the decking on the balconies.

This leaves the question of whether the technical auditor should have required the builder to open up the floor of a sample balcony so that he could inspect the construction detail at the thresholds of the doors and windows. For the reasons that I have already given, I do not consider that a technical auditor was under any obligation to require the opening up of work that he had not had an opportunity to inspect unless he had some reason to think that the work covered up might not have been done properly or that it involved the use of an inappropriate detail. In the case of the balconies, I do not consider that either of these features was present. Whilst a diligent technical auditor might have felt that the balcony details were sufficiently important to have one example opened up for his personal inspection, I do not consider that it was negligent or a breach of Campagna's duty of care not to take this step.

For these reasons, I consider that Campagna was not in breach of its retainer in relation to the balconies.

As I have already mentioned a further defect in the construction was the fact that the doors to the living room were hung from the inside edge of the interior wall, and not its outside edge as shown on at least one set of plans (although there was a suggestion that these particular plans were in fact prepared for fire safety purposes). This is not an allegation that has been made by Liberty, and, even if it had been, I doubt whether Campagna's failure to pick it up could be classed as negligent.

This defect relates to apartment No 24. The ducting from the roof ventilation tile was not properly connected to the flexible duct from the internal soil stack. This caused damp staining to the ceiling in the corridor outside apartment No 24. In his witness statement, Mr O'Connell said that when this area was opened up it was apparent that the soil vent stack in the corridor outside apartment No 24 was not connected to the vent pipe so that the moisture from the stack could remain in the attic area above the corridor resulting in condensation and subsequent dampness to the corridor ceiling.

It is not apparent from any of the reports, witness statements or photographs whether or not this particular section of ducting could have been checked from any accessible part of the roof space. In the light of this I do not consider that there is any material upon which I can properly find that this was a defect which Campagna's technical auditors ought to have noticed.

I have already discussed the problems with the canopies in some detail. A further point made by Mr Scott, and amply demonstrated by the photographs, is that the lead flashings at the junction between the canopy and the wall are inconsistently dressed throughout the development. For example, in some cases the flashings are inserted at every course on each side of the canopy, whereas in others they are inserted at every other course (so that there are fewer pieces of flashing). In terms of weatherproofing, this may not matter but it demonstrates an inconsistency of workmanship that should have served as a warning bell to a technical auditor that increased vigilance might be required.

For the reasons that I have already given, I consider that Campagna should have investigated the method by which the canopies were to be fixed to the building. They did not do so. It seems that no canopy was inspected in the course of construction and neither were there any inspections of the inside of the canopy before the ceiling below it was installed. There was a conflict in the evidence as to whether the canopies were completed by the time of the pre-plaster inspections or whether the openings were sheeted up and the canopies installed later. Either way, Campagna's evidence was that no canopies were in the course of construction at the time of any site visits. In his report, Mr Scott suggested that this was unlikely. I have to confess that, whilst I share Mr Scott's scepticism, I do not consider that that is a sufficient ground for rejecting the evidence of Mr Williams and Mr Hilton on this point.

However, taking the evidence as a whole it is clear that Campagna’s technical auditors did not know how the canopies had been or were being fixed to the building and took no steps to find out. In these circumstances, I consider Campagna was in breach of its duty to Liberty in relation to the canopies.

If Campagna had taken proper steps to find out how the canopies were to be fixed to the building and had found, as I find that they should have done, that simply nailing the framework to the building was unsatisfactory, they could have taken steps to ensure that a proper fixing detail was adopted. It is of course possible, as Ms Franklin suggests, that the builder would not have told the technical auditors that it proposed to nail the frame of the canopy to the wall of the building but would instead have said that some form of suitable anchor was going to be used. However, I reject this as the likely outcome. I consider that, if asked, the builder would be more likely to tell the truth about the proposed method of fixing (because, if he lied, he would be aware that he might well get caught out) or, possibly, to say that he had not given the matter any consideration and was going to leave it to the workmen on site. In either case, the technical auditor would then have been able to discuss the method of fixing and to tell the builder that if he did not adopt a suitable method he could not expect to obtain a Certificate of Approval.

Ms Franklin further submitted that the fact, which appears to be the case, that the canopies on some houses have not collapsed indicates that the fixing details must have been different on those houses. She submits that if the technical auditors had inspected one or two canopies during the course of construction, they might well have seen ones that were being properly constructed using correct fixings rather than the ones that were just nailed to the wall of the building In my view, the first limb of this submission is not necessarily correct. However inadequate the fixings to the wall may have been the canopies were not wholly unsupported because some of the load may have been taken to a greater or lesser extent by the bay window below. These windows were not intended to be load bearing but as a matter of fact they would have been capable of taking some load. Further, I suspect that in some cases the fitting between the frame of the window and the canopy was tight, with the result that the window frame was able to take enough of the load to prevent the canopy from dropping. Whilst there is not sufficient evidence for me to make a finding on this point, and so I do not do so, it does provide one explanation as to why the canopies may have failed in some cases and not in others. A further reason for the absence of universal failure is that in some cases the nailed fixings may have been better than in others and just sufficient to keep the canopy supported. I therefore do not accept Ms Franklin’s submissions on this point.

The defects alleged (as described in the report of Mr Scott) are incomplete and poor installation of the sarking felt and the lead flashing around the chimneys. In particular, the sarking felt was not adequately lapped under the flashings to the chimney to form a proper barrier. Liberty alleges that this subsequently allowed water to penetrate the waterproof envelope of the property, so that there was water penetration around the chimney stack which then damaged the ceiling of the front bedroom. These defects concern Nos 6, 11 and 12 Derrycorris Drive.

Although the defect as described by Mr Scott also refers to poor installation of the lead flashing to the chimney, Mr Scott does not discuss this in his report. There was evidence from Mr Kerrigan that, when he visited the site in 2008, he could see the untidy flashing round the chimneys from the road, but on the basis of this evidence alone I do not feel able to reach a conclusion as to whether or not this represented the state of affairs at the time when Campagna’s technical auditors carried out their final inspections. I shall therefore say no more about the flashing to the chimneys.

It is common ground that the remaining defects are not ones that have anything to do with the plan check. These defects, if they exist, would be the result of poor workmanship. It is alleged that they should have been noticed at the pre-plaster or final inspections. In his report, Mr Scott said this (at paragraph 6.20.3):

“I am of the opinion based on the evidence following the joint site inspection that the installation of the sarking felt to this property (and other properties inspected by the TA at the same date namely no 6, 5, 11) all revealed:

- Omission of any timber trimmer arrangement to the chimney stack necessary to support the lead flashing/sarking felt interface

- Gaps to the party wall/chimney abutment

- Sagging and untidy installation of sarking felt

- Rips and tears to areas of sarking felt"

Mr Scott does not contend that Campagna should have seen anything wrong with the lead flashing around the chimneys. This would not have been visible from the inside at a pre-plaster inspection. What I understood Mr Scott to be saying was that the poor detailing of the sarking felt around the chimney and the absence of the small timber trimmer (which would have acted as a support for the lead flashing) should have been seen when looking at the roof from the inside, whether at a pre-plaster or at a final inspection.

Mr Scott considered that the presence of these defects indicated a generally poor standard of workmanship. However, by the time that Mr Scott had carried out his inspection remedial works had already been carried out to the interior of the roof at No 12. The photographs appended to his report showed the roof space after the remedial works had been carried out and so any defects in the original work shown in those photographs or that were present at the time of the inspection had been left unremedied when the remedial works were carried out.

In view of this Ms Franklin submitted that the defects in the original work that Mr Scott has identified cannot have presented any risk under the policy if no one thought it necessary to repair them. More generally, she submitted that photographs taken some 6 years after the event and after some remedial works had been carried out did not provide reliable evidence as to the situation at the time of Campagna's inspections. She submitted that the evidence of Mr Hilton, who was actually there, should be preferred to what she described as the reconstruction exercise carried out by Mr Scott.

Although the photographs showed examples of untidy sarking felt which Mr Scott accepted was part of the remedial works, Mr Scott said that he had also seen untidy and poorly installed sarking felt that he considered was part of the original construction. This was the sarking felt around the chimney which, Mr Scott said, had not been tucked up against the sides of the chimney underneath the lead flashing: instead, it was left hanging down (and, in the case of No 14, of which there was a photograph of the felt apparently as originally installed, it was not cut to size). He said that the primary barrier against ingress of water around the chimney was the external lead flashing and that the sarking felt was only a secondary barrier. It follows therefore that if the lead flashing was sound there should be no leaks around the chimney.

The PGS covered the costs incurred in rectifying any part of the waterproof envelope of the house as a result of the ingress of water caused by a defect in the design or workmanship of the waterproofing elements of the waterproof envelope of the house. The waterproof envelope was defined to include the roofs. If a technical auditor noticed a defect in the construction of the roof which gave rise to a risk that water penetration may occur, he should not issue a Certificate of Approval unless and until that defect was put right.

Although, as I have said, the primary barrier against water penetration at the junction between the chimney and the roof was the lead flashing around the chimney, and the sarking felt below was a secondary barrier, the object of this method of construction is to ensure that any water that penetrated past the flashing was caught by the felt and then carried harmlessly away to the gutter or roof overhang. If the sarking felt is laid in such a way that it cannot or might not perform this function, then I consider that the property would not represent a standard risk for the purposes of the PGS. Accordingly, it would be a matter that would have to be addressed by the technical auditor.

I find it a little surprising that the defects in the installation of the sarking felt around the chimneys were not rectified when the remedial work was carried out to some of the houses. However, I suppose that it may have been thought by those responsible for specifying the remedial works that the difficulty and expense of relaying the sarking felt and dressing it properly up against the sides of the chimneys was not justified and that adequate protection against water ingress for five years or so could be achieved by rectifying the defects in the lead flashing and ensuring that it provided a weathertight seal. I therefore do not accept Ms Franklin's submission that the defects in the installation of the sarking felt around the chimneys did not represent a risk in the context of the PGS just because it was not rectified. As I have indicated, I consider that the failure to rectify it was probably driven by considerations of cost and pragmatism.

The photographs show that the sarking felt that was left hanging down beside the chimneys was not difficult to see. I am satisfied that in some cases, including the relevant sample properties, it was not properly installed. It is true that a technical auditor would have to make a point of looking at the chimney when he was inspecting the roof space, but that is something that I would expect a competent technical auditor to do. It is well known that the junction between a chimney and the roof is a weak point in the waterproof envelope of a building and I would expect a technical auditor to give it special attention, both at the pre-plaster inspection and at the final inspection. For the reasons that I have given elsewhere in this judgment, it is not an adequate excuse for a technical auditor to say that there was not a suitable ladder available with which to gain access to the roof space. It was the technical auditor's responsibility to make sure that a suitable ladder was available.

In my judgment, therefore, Campagna's technical auditors should have noticed the defects in the sarking felt around the chimneys and their failure to do so was a breach of their obligations under the 2004 Retainer to Liberty. This applies to Nos 6, 11 and 12. Whether or not this breach of duty has resulted in any loss to Liberty is open to question if the defect was never rectified when the subsequent remedial works were carried out. However, that is not a question that I have to answer in this judgment.

There was a further problem at Nos 11 and 12. At these properties there was a leak through the lead flashing lining the valley gutter. The evidence about this defect is unsatisfactory and confused. Mr Scott refers to a report from a consulting engineer, Mr Larry Holton, dated September 2008, which is supposedly appended to his report as Appendix 2.7. In fact appendix 2.7 is a report by Mr Holton dated 29 August 2008 which relates to No 36.

However, in the loss adjusters’ correspondence file for No 11 there is a report by Mr Holton dated 16 September 2008, which I take to be the report that Mr Scott intended to append as his Appendix 2.7. That report concerned Nos 11, 12, 53 and 54 Derrycorris Drive. In each case there had been a leak that had caused damage to the ceiling of the main bedroom, causing its collapse in the case of No 11, that was in a different place to the leaks that came through the junction between the roof and the chimney. Mr Holton concluded that the cause of the leak was that either the felt in the valley gutter was not overlapped sufficiently or that it was jointed in the centre of the valley. According to Mr Holton, the lead flashing should be laid over the felt in the valley gutter and should then be covered by the roof tiles at each side. This I accept. However, this had not been done in these properties because the underside of the lead flashing that formed the valley gutter could be seen from inside the attic.

In my judgment, this is also a defect which should have been noticed during both the pre-plaster inspection and the final inspection. Exposed lead is not difficult to see and a competent technical auditor should have known that the lead was supposed to have been laid with sarking felt underneath it. In any event, the correct detail would have been evident at other houses (apart from of course, Nos 12, 53 and 54). Accordingly, I find that the failure to observe this defect was a breach by Campagna of its obligations under the 2004 Retainer.

This defect relates to Nos 6, 11, and 12. The evidence about the nature of this defect also is not very satisfactory. Mr Scott refers to a letter dated 16 September 2008 from Mr Holton and to a report from OSG dated 8 May 2008. I have not been able to find either of these documents.

There are various references in the documents to these roof vents from the bathrooms not being properly flashed or to a poor arrangement of the extract ducting from the bathroom (for example, a report by Mr Holton in respect of No 12 dated 16 October 2008), but I have been unable to find any clear description of either aspect of the defect. There is one photograph, taken in the attic of No 24, which is said to show the installation of the vent tile and the absence of a secondary layer of felt to the opening around the vent tile.

Mr Scott said that the poor installation and connection of the flexible duct between the bathroom ceiling fan and the proprietary ventilation tile “should would [sic] have been evident” to the technical auditor when carrying out an inspection of the properties. In relation to the ventilation tile, Mr Scott said that the fact that the sarking felt had not been adequately cut or prepared to receive the tile and that a secondary “undercloak” had not been installed around the tile ventilator would have been evident to a technical auditor when looking up at the underside of the roof during a pre-plaster inspection.

Mr Hilton, when pressed in cross-examination by Mr Quiney, said that if he had gone into the attic of No 24 and had seen the vent tile arrangement shown in the photograph, he would not have passed it. For the reasons that I have already given, I consider that it was the duty of the technical auditor to examine the underside of the roof during a pre-plaster inspection and to enter (at least to the extent of putting his head and shoulders into it) and inspect the roof space during a final inspection.

The vent tile arrangement shown in the photograph is, in my opinion, something that was fairly easy to see by anyone who was looking for it. It was a type of feature which technical auditors should check because it represented a potential weak point in the waterproof envelope.

Mr Mulrooney expressed the view that this was not a defect that was covered under the terms of the policy. In so far as the arrangement of the sarking felt around the vent tile was not weatherproof, then in my judgment it presented a risk covered by the policy. It was a defect that a competent technical auditor should have noticed and I consider that Campagna was in breach of its obligations under the 2004 Retainer by failing to do so.

However, I am not persuaded that the other aspect of the arrangement described by Mr Holton, namely the fact that the installation of the vent and ducting was such that there could be a back flow of condensed steam to the extractor fan either represented a risk covered by the policy or was something that a competent technical auditor should have noticed. There is no evidence before me as to what it was about the arrangement that should have indicated to a technical auditor that something might be amiss.

Accordingly, I am not persuaded that this aspect of the defect is one that Campagna's technical auditors should have spotted. I therefore reject Liberty's claim in respect of this aspect of the defect.

Twelve of the houses at Derrycorris Drive had a lean-to extension at the rear which contained the kitchen. This had a sloping tiled roof which joined the rear wall of the house just below the first floor bathroom window. The problem with this arrangement as it was built was that water was able to penetrate through the joint where the roof of the lean-to met the wall of the house resulting in a leak into the kitchen. This defect affected Nos 12, 24, 25 and 26.

Liberty's principal criticism of this joint is that there was no sufficient upstand between the topmost layer of tiles on the roof of the lean-to and the underside of the bathroom window. Where the tiles met the wall there was a line of lead flashing which disappeared from view under the bathroom windowsill and under the pebbledash render on the walls on either side of it. There was a horizontal mastic joint all the way along the top of the flashing which appeared to be the only barrier against water entry. It was difficult to tell by a visual examination of the photographs whether the lead flashing stopped underneath the render or whether it was turned up the brickwork and then subsequently covered by the render as it should have been.

What is quite clear from the photographs is that there was no upstand of any sort under the windowsill. The lead was just dressed right up to the wall immediately below the sill and then sealed with mastic. It was Mr Scott's view that good practice for a detail such as this is that where the flashing meets the wall it is turned up to form an upstand of at least 75 mm. This was impossible under the window because the underside of the windowsill was only about 20 mm above than the top of the upper layer of the roof tiles.

It is not clear why the builders allowed this to happen. According to Mr Scott the timber frame drawing showed that there was a distance of 411 mm between the underside of the windowsill and the highest point of the truss of the timber frame of the lean-to roof. Allowing for a 75 mm upstand, this left over 330 mm to accommodate the roof covering including the battens. This should have been more than sufficient.

There can be no doubt that this detail was clearly to be seen by any technical auditor who looked at the rear of a house after the lean-to had been constructed, or whilst it was in the course of construction (if the framework was in place). Campagna's case is that there was nothing wrong with this detail in principle, because it should have been perfectly possible to construct a weatherproof detail given adequate workmanship.

The HomeBond manual recommended a 150 mm upstand at the junction between a lean-to roof and the vertical wall against which it abutted. The detail shown in the manual also provided for a further piece of lead flashing dressed into the brickwork above and turned down over the upstand. Mr Scott’s opinion was that any competent technical auditor should be familiar with this guidance. Mr Scott said that since the detail as constructed did not meet accepted standards Campagna's technical auditors should have paid closer attention to the way in which the detail was constructed. He said that having seen that there was only about 20 mm between the underside of the windowsill and at the top of the roof, a competent technical auditor should have concluded that this did not represent a standard risk in terms of the PGS.

Mr Hilton disagreed. When it was put to him that the lean-to as constructed was clearly inconsistent with the recommendation in the HomeBond manual and that this should have rung an alarm bell in his mind he said (at Day 7/125-130):

When Mr Hilton was asked whether he checked whether the flashing went up behind the render he said that since he signed it off, he must have been satisfied that it did. Mr Williams was also adamant that the detail shown in the photographs was a simple design solution that “gave me no cause for concern that it wouldn't work because I've seen it lots and lots of times before" (Day 8/35). Mr Williams had taken a photograph of a pitched roof with a dormer window set into it so that the roof tiles went right up to the underside of the window sill. He relied on this as being a typical example of the type of detail used at Derrycorris Drive. The roof in question was of some considerable age and it turned out that it was at Queens’ College, Cambridge. It was, of course, not possible to tell from the photograph precisely how the joint between the underside and the top of the tiles had been formed. Further, as Mr Scott pointed out, the roof of the Derrycorris lean-to has a pitch of about 22°, which is much shallower than the pitch of the roofs shown in the photographs upon which Mr Williams relied.

Mr Scott said in evidence that he had looked closely at the construction of one of the lean-to roofs (at No 36) and found that the lead flashing terminated horizontally against the wall and was not dressed up underneath the render. I consider that it is likely that this is how they were all built.

I do not accept the submission of Ms Franklin that Campagna’s technical auditors were justified in assuming, if indeed they did assume this, that the lead flashing where the lean-to roof met the wall of the house had been dressed up vertically underneath the render. However, I agree with her that this was the issue. This was a detail that could have been checked on site fairly quickly and without any difficulty and I consider that the technical auditor should have done this. It did not have to be taken on trust.

In my view, Campagna's technical auditors should have appreciated that the detail was unusual, and not the detail recommended, and looked into it when the first lean-to roof was being built (or had been built). They should have checked the plans or, if they did not have any plans at that stage, they should have asked to see the plans to which the builders were working. If those plans showed the details described by Mr Scott, then they should have asked why they were not being followed. If, on the other hand, they showed a different detail, then the technical auditor should have discussed the detail of junction between the roof of the lean-to and wall of the house and satisfied himself that whatever the builders were proposing to do either involved an adequate upstand or was a detail that they were confident would work.

Whilst I consider that it might not have been unreasonable to accept the as-built configuration of the roof and the window sill, if there was a good reason for it and provided that particular care was taken with the mastic seal, Campagna should have insisted that on either side of the window the lead flashing was dressed up the wall under the render (as Mr Hilton says he assumed was the case).

In my view Campagna did not investigate the detail of the construction of the lean-to as they should have done and, in permitting the construction of the lean-to to be carried out as it was, they were in breach of their obligations to Liberty under the 2004 Retainer. Proper steps, if taken, should have ensured that the leaks did not occur or, at least, not within the 5 years of the relevant part of the cover.

At Appendix B attached to this judgment I set out, by reference to each of the sample properties, the answer to issue 4.3, namely whether or not Campagna incorrectly issued the Certificate of Approval. On the assumption that certificates were issued for every property, this shows that in the case of every one of the sample properties, save for apartment No 10 at Mountgorry Wood, the certificate should not have been issued owing to the presence of defect(s) that should have been noticed by Campagna.

Whilst I have found that in certain respects Campagna did not discharge its duties to Liberty, I have asked myself several times why this should have been so in Ireland but not, apparently, in the United Kingdom where many of the same technical auditors were also inspecting houses under construction for Liberty.

In the course of considering the evidence I formed the clear impression that the quality of the building at the First Developments was in some areas, particularly the roofing, very poor. It seems that the degree of specification by the designers was limited and that the builders may often have been left to work out construction details for themselves. Any supervision that there may have been must have been cursory. I consider that Campagna should have become aware of this and been more inquisitive and pro-active about what was being done on site.

To the extent that either Liberty (or MDIS) or Campagna placed any reliance on the certificates that would be signed by the architects or engineers responsible for the design to confirm that the properties substantially complied with the relevant regulations, I consider that they may have been unwise. It is possible that these architects or engineers carried out inspections during construction, but I have seen no evidence of it. My overall impression is that many shortcuts were taken in the construction of the First Developments and no-one put a stop to it. Perhaps the building boom in Ireland was accompanied by a fall in the standard of work: however, on the limited evidence of three developments it would not be safe to form a conclusion about this.

I consider that some of the issues do not lend themselves to an intelligible answer that can be condensed into a few sentences, and so I have dealt with them rather more discursively. However, I hope that my conclusions in relation to the issues are now reasonably plain from this judgment. I will, of course, hear the parties if there are any further questions that either party feels have not been sufficiently answered.