In a magical setting on the shores of Loch Lomond, Cameron House Hotel and Spa is a luxurious, up-market resort. Its centrepiece is an imposing Scottish baronial style house, built in the 19th century, close to the water’s edge.
But tragically, the most historic part of the building suffered a devastating fire in December 2017. Since then, much of the structure has stood open to the elements.
Now the forensic examinations are complete, the owners are setting about restoring the building. Because of the extent of the damage, the plan is to completely rebuild the internal structure. But with the hotel forming a key part of the business, understandably the hope is to progress the work as swiftly as they can.
The main architects employed by the client are responsible for the new internal layout and construction, and conservation architects Simpson & Brown were employed alongside, specifically to advise on the historic structure and its conservation.
There was general agreement that the energy performance should be improved if possible, Simpson & Brown said.
“Although historic buildings do not have to meet the same energy standards as new buildings, there is an expectation in the building regulations that the insulation levels are improved wherever possible. Insulating the building will help to reduce the energy required to heat the building, so our client was keen to improve the thermal performance.”
The combination of the damage caused by the fire and the need to proceed quickly has led to most of the building’s internal structure having to be removed and rebuilt using modern techniques and materials such as concrete and steel, Simpson & Brown added, but the original traditional appearance is being retained. In order to preserve the iconic appearance, as so often with historic buildings, the only available option is to insulate internally.
Internal insulation always needs to be installed with careful consideration of the moisture performance of the completed build-up. There are a number of risks if too much moisture accumulates within the structure, including condensation, mould, and damage such as corrosion and rot, in particular where the back of the insulation meets the (cold) masonry.
But there was a considerable extra challenge here, as while the building had been without its roof, the 800mm thick sandstone walls would have absorbed extra water – and wet stone can take many years to dry out. This could affect the conditions in the insulation and even in the fit-out as the rooms were finished and decorated.
Simpson & Brown wanted to get a better understanding of how the wet stone and insulation would interact with each other, with the Scottish weather outside, and the warm conditions and occupant activities inside. They explained:
“In conservation circles there is a lot of concern about how to insulate traditional buildings and protect the fabric; we know that internal insulation has particular risks, which is why we called in Greengauge to get their technical input to back up our advice.”
The first task was to establish what kind of stone the walls were built from, what its moisture properties were, and just how wet it actually was, so Toby Cambray and Jack Preece from Greengauge travelled to Loch Lomond to carry out site investigations.
The “wettability” of the wall from outside (i.e. the likely rate of water uptake from wind-driven rain) was tested using Karsten tubes. These are small pipes stuck against the wall, and filled with water – the faster the water soaks into the wall, the more water the stone is likely to take up in wet weather. Samples of stone were also taken and tested back in the Greengauge lab, to assess how quickly moisture is taken up through the thickness of the stone. Lastly a dozen small holes were drilled deep into the wall, and the existing moisture content of the stone was analysed every 50mm in, until the drill reached the loose rubble core.
The average moisture content ranged from very dry (less than 0.5% by mass) to moderately wet (around 7% average at one test hole) – though Greengauge warn that there were parts of the building they could not access, so they cannot be certain that their samples were fully representative.
Samples from locations facing in different directions showed that south- and west-facing walls were wettest, confirming the moisture had come from wind-driven rain (rather than simply absorbed from the damp Scottish air). In many cases the walls were wetter away from the surface, showing that although the drier June weather may have been drying out the surface, the water deeper in was taking longer to evaporate.
The next step was to investigate the way the moisture content of the walls might change once the building was re-roofed, and rain was no longer falling on the inside and tops of the walls – taking into account the properties of the stone and the starting conditions found during the site investigations.
Moisture can enter, move through, and leave the elements of a building is several ways: rain can seep or be blown in to the masonry; water vapour can condense when meeting a cold surface, or be absorbed from the air into a hygroscopic (water-attracting) material. Movement can be driven by wind or gravity, by vapour diffusion and by capillary transport.
How damp or dry any element of construction is at any point in space and time depends on environmental conditions (temperature and moisture) indoors and outdoors, what the element and all the other components are made of – and what has been happening previously.
WUFI is a software tool that has been devised to predict the behaviour of moisture in a construction build-up over time, taking these variable factors – including the properties of the particular building materials – into account. Crucially, the calculations carried out in WUFI are cumulative over time, with the analysis building on the changes that have already taken place. This so-called “dynamic” modelling makes it possible to predict how conditions within a construction will progress – over 20 years or more if required, through the seasonal cycles of weather for the locality. Thus WUFI modelling can help to predict how quickly a wet construction might be expected to dry out and stabilise; also whether any components are at risk of mould in the meantime.
The somewhat simpler Glaser analysis (as set out in BS 13788) does not take all of these factors into account.
The WUFI modelling indicated that once the building was roofed, any part of the wall that was saturated would dry out very quickly to begin with, dropping below what is regarded as the ‘level of concern’ of 5% moisture by mass within a year. (The limitations of one-dimensional modelling may overestimate this in some cases however, so the prediction must be treated with caution.)
Moisture levels would then be expected to continue falling, but more slowly, edging down to around 1.8% at 20 years and stabilising at some point beyond that.
The team then needed to explore how this process might be affected by – and affect – the proposed internal insulation – as well as how the insulation would deal with the ‘regular’ moisture from both inside and outside in the long term.
There are a number of possible approaches used for internal insulation, each with their pros and cons. Given the programme and other constraints, the contractor was keen to use a system with straightforward installation, preferably avoiding build-ups that involve several layers or several trades.
One practically straightforward approach is to build an internal frame and fill this with insulation batts –a system using lightweight steel framing and mineral wool insulation had been proposed. According to Simpson & Brown, this is a common technique for internal insulation, but the team was concerned about the possible moisture implications of this approach at Cameron House.
The proposal included a vapour control layer (VCL), as is usual, to keep warm moist air and water vapour from the interior away from the cold surface of the stone. However it can never be guaranteed that this layer won’t be breached one day: by wear and tear – or by someone wielding a drill. And in the case of Cameron House, the stone was already wet – and the VCL would also prevent moisture from leaving the masonry and passing into the interior.
Modelling (including an allowance for some warm air from inside to reach the back of the insulation, as is realistic) showed the team was right to be concerned, and that the relative humidity behind the insulation would remain very high – close to or even at 100% – for over five years, leading to a significant risk of mould growth, water damage and corrosion.
The proposed approach had included an approximate 50mm ventilated gap between the insulation and stone. In order to remove sufficient water to prevent the dangerous build-up of moisture, Greengauge warned that there would need to be a considerable flow of air through this gap. Although it was difficult to calculate with certainty, they suggested there would probably need to be the equivalent of a 50mm diameter hole every 1m along the top and bottom of each storey – unlikely to be acceptable from some points of view!
This meant a different approach had to be found. An alternative to using a cavity and vapour control layer to limit wetness at the back of the insulation is a ‘breathable’ internal insulation system that does not block moisture, but rather allows it to move, both from the room towards the surface of the masonry but also, crucially, back out to the building’s interior, where it can be removed by ventilation.
If this is to work safely, the material also needs to be able to absorb and redistribute the moisture effectively enough, without too much building up at any one point: in technical terms, it needs to be sufficiently hygroscopic and sufficiently capillary active to cope with the particular moisture loads it will face.
A number of hygroscopic and capillary active insulation systems are available: some made with natural materials such as hemp or timber, some with a mineral base.
Given the preference from the contractor for simple installation, the first breathable option investigated was a ‘one-step’ system: a rigid wood fibre board, backed with a spongier layer of wood wool which moulds to accommodate irregularities in the wall surface, meaning no base layer is required to level the wall.
Modelling (again including an allowance for some air to penetrate from the room) showed that as expected that the combination of the cold wet Scottish weather, high moisture burden in the walls, plus moisture from the interior, would raise the humidity in the cooler zone at the back of the insulation, leading to moisture absorption into the insulation. However because the back of insulation in this zone is relatively spongy and light, modelling predicted that it would not have the capacity to carry this moisture fast enough to the dryer layer on the room side. For at least 5 years the back of the insulation would be expected to remain well above 20 % moisture content by mass – considered the “danger level” for natural materials like wood fibre, and likely to lead to deterioration of the wood fibre itself.
Denser materials may do better in these conditions. However using a denser wood fibre board would require preliminary applications of lime levelling and bonding coats, which it was not felt there was time for in the compressed programme required. One dense hygroscopic insulation material is diathonite plaster — a hygroscopic lime-, clay- and cork-based insulation that is mixed with water then trowelled or sprayed directly onto masonry.
To model the drying out of the wall with this option, the team had to include the water included with the diathonite, which is applied as a wet mix. That meant that the structure would have a higher total moisture content at the 1-year point. However, the calculations predicted that the moisture levels would fall below 20% after a couple of years, confirming that the denser diathonite is better able to distribute moisture away from the humid, cold areas close to the cold stonework than the wood fibre alternative. Diathonite is also less susceptible to decay.
Greengauge also investigated the performance of a thinner layer of 80mm, rather than 120mm, of diathonite. Although the thermal performance would be reduced, it would mean that there would be less diathonite-mixing water to shed initially, and the diffusion of moisture from the wet masonry out to the interior might be able to take place more quickly.
Modelling confirmed that this would be the case, and in fact, because diathonite is more hygroscopic than the masonry behind it, it would actively attract moisture out of the wall and allow it to be carried away by the building’s ventilation, meaning the wall would dry out more quickly than it would if left uninsulated.
Nevertheless, with either thickness of diathonite, the moisture levels in and adjacent to the insulation would be high for a couple of years – depending on how wet the different areas started off. While this is not a problem for the stone or the diathonite, it is also proposed to build a plasterboard inner wall with a service void behind as part of the final high-spec interior – and this section could be more vulnerable.
Modelling showed that the air within this proposed void would indeed be quite humid for the first couple of years, posing a risk of mould, and possibly cause corrosion of the steel within it. Repeating the modelling with a later starting point, after the diathonite and masonry had been allowed to dry for six months, showed dramatically reduced risk.
Greengauge therefore recommended that if they are to use a diathonite insulation, the construction team should allow as much time as possible, and certainly several months, for the wall to dry out before completing the internal fit-out.
The knowledge about which elements of the walls were wettest also gives the team the option to leave the wettest areas till last. Similarly, a few areas of the exterior had a painted cement render. Modelling confirmed that these areas would be slowest of all to dry out, as the paint was not vapour permeable. These areas too should be given the maximum possible drying time, Greengauge advised – or even left uncovered altogether. Greengauge’s advice was to allow the wall to dry as much as possible, ideally to below about 5-6% moisture content, before fitting the internal plasterboard.
Greengauge also recommended the team give themselves a head start by getting the drying-out process under way before any construction began at all, by temporarily covering the masonry in a way that would keep off the rain but allow drying by wind and sunshine.
Toby Cambray says: “The team have taken on board my recommendations to promote drying between now and the time the plasterboard goes on. They will cover the exposed walls to keep rain off as much as possible, the diathonite will go on as soon as possible firstly so it has a chance to dry itself, but it will then promote the drying of the wall by coupling to the mortar and creating a bigger surface area from which water can evaporate.”
Given the fact that the building will need to go on drying after the envelope is once again complete, the ventilation system, always critical (particularly given the density of baths and showers in luxury hotels) has another, vital role – one which Toby Cambray says he has emphasised at every opportunity .
To further improve the thermal performance and comfort, Greengauge also advised on ways to reduce thermal bridging in places where the new internal structure met the external fabric – thermal bridging is often overlooked, yet can contribute significant heat loss – and sometimes introduces a surface mould risk too.
The conservation architects were pleased to get the additional certainty they needed to go forward and insulate such a sensitive building.
“There is a lot of discussion and hearsay about internal insulation,” Simpson & Brown say, “but there are not many answers. Greengauge’s investigations and input gave us the confidence we needed to back our advice.”