The process of building predates science by several thousand years and much of what we know about the subject has been learned by simple trial and error. It’s only in the past 80 or 90 years that scientists have turned their attention to construction and started to run the slide rule over it, so it’s hardly surprising that we are still learning things and that some of the earlier conclusions are now being challenged by more recent research.
Are U Values Accurate?
The idea that some materials hold on to heat better than others is intuitive and easy to grasp: it’s why we wear wool, not cotton, in winter. But the idea of quantifying the difference in performance is recent and very much the work of building scientists. Most of the original measuring was carried out in the 1950s using a laboratory hot box which gauges the temperature differences between inside and out and ascribes a U value to each material tested — good insulators, like wool, have low U values. Since then, U values have become the lynchpin of our understanding of energy efficiency.
When you design a wall, floor or roof, you have to calculate the U values of all the materials used, in the proportion and thickness specified, and combine all these figures into a global U value. The Building Regulations set a maximum level for U values and the task of the designer is to match or beat it. But in practice, the resulting performance of the building is often quite at odds with the theory. This may partly be due to poor building standards, but there are some who think that the hot box testing regime is flawed.
This debate all came to a head a few years back when multi-foil insulation started being marketed with claims that a 35mm-thick sheet was as good as 250mm blanket of mineral wool. Not in a hot box, it wasn’t. The conventional hot box test showed that, in general, multi-foils were equal to about 80mm of mineral wool, no more. But the multi-foil manufacturer’s claim was based on a very different test which involved building two comparison roofs and measuring the heat required to keep the enclosed space at an even temperature.
This caused a furore in the building physics community. A few respected researchers, and a solitary accreditation body, stood by the multi-foil tests; the majority suggested that it was all too easy to fudge the results and stood by the veracity of the hot box testing regime. Eventually, the powers-that-be came down against the multi-foil claim, but not before many had come to question just how accurate a predictor the hot box test actually is.
In theory, there’s nothing wrong with using heat-loss measurements on site as a method of establishing the energy performance of a dwelling. In fact, the practice, known as co-heating, is becoming widespread in the building physics community. The problem arises when you use co-heating as a tool for comparing one system with another because even minor variations in construction detail can have dramatic effects on the outcome. In other words, it’s very easy to manipulate the results to get a pleasing outcome and this highlights yet another problem for our understanding of building science and that is that much of the research is funded by commercial businesses wanting to establish their product’s credentials.
U Values and Old Homes
More recently, a new challenge has emerged for the prevailing orthodoxy of the hot box measuring of U values. It comes not from a manufacturer, but from heritage bodies — the Society for Protection of Ancient Buildings (SPAB) and Historic Scotland. They are concerned that old buildings will be harmed by the inappropriate use of insulation, put in place with the intended aim of improving energy efficiency. Their principal cause for concern is that introducing insulation products inside an existing building envelope may cause damp and mould problems, and they have become very active in promoting research into what actually happens when an older building gets retrofitted. As this is one of the main aims of the Government-sponsored Green Deal, which came into effect in January 2013, the need for such research is more pressing than ever.
SPAB now finds itself co-ordinating a number of research institutions including Glasgow Caledonian University, Cardiff University, as well as specialist investigators such as ArchiMetrics, who are undertaking a variety of projects across the country measuring the before and after effects of retrofitting insulation. Building scientists are starting to quantify the difference between the published U values of materials and their on-site performance by using heat flux plates attached to walls. One of the early conclusions is that many solid wall assemblies perform much better than the hot box tests of the constituent materials would suggest. In fact many unusual materials, like cob and straw have never been tested in a hot box, so their estimated U values are questionable to say the least.
Now once again, you might argue that SPAB has a vested interest in fixing the results of its research, as their remit is for the preservation, not the alteration, of ancient buildings. But SPAB doesn’t take a predetermined view here; they are not, per se, opposed to insulating old buildings, merely wanting to see that they are not damaged in the process.
SPAB’s work has led to the questioning of another fundamental tenet of our understanding of how buildings work, and that is how moisture behaves in walls and roofs. Again, our accepted standards are based on laboratory work undertaken in the 1950s, now referred to as the Glaser method. It starts from the assumption that water vapour pressure is higher inside a building than outside (it usually is) and hence water vapour will tend to be driven from inside to out. Thus, if you are constructing a wall using materials susceptible to water damage, you must create a vapour barrier on the inside to stop water condensing within the wall — a phenomenon known as interstitial condensation. The Glaser method is enshrined in the British Standards (BS 5250) and is the basis for the Condensation Risk Analysis which wall and roof assemblies have to comply with.
However SPAB’s work, which builds on a body of work carried out in North America – check out the website of the Building Science Corporation (very readable but utterly confusing!) – suggests that the Glaser method is horribly over-simplistic and that of all the problems likely to be faced by walls and roofs, vapour diffusion is very low down on the list. Worse still, vapour drive sometimes occurs in the opposite direction – i.e. out to in – and in this instance a traditional plastic vapour barrier might just do more harm than good. What actually goes on inside a wall assembly is complicated: there are air movements, convection currents and capillary actions as well as vapour diffusion, and some materials have the capacity to absorb moisture whilst others provide a condensation surface. Predicting where mould or condensation might occur is very difficult and there is currently no universally agreed model of how to do it, or how to improve on the Glaser method.
The main issue with walls seems to derive from rainwater getting in and not being able to dry out — this causes problems for insulation as well, because wet insulation is, by and large, pretty ineffective. Condensation issues are more prevalent with roofs, but more often than not, this is caused by relatively warm, moist air getting trapped in enclosed loft spaces where night-time temperatures can be very low.
All this has led to an interest in vapour-permeable construction systems which don’t attempt to build in any vapour barriers but allow the wall or roof assemblies to dry inwards and outwards. And behind this is a gaining interest in natural building materials which tend to work with moisture rather than trying to act as a barrier against it.
But is all this necessary? Proponents of the traditional view of timber frame with vapour barriers point out that, whilst the Glaser model may have flaws, it works pretty well in our climate and that we have nearly a million timber frame homes in this country built with plastic vapour barriers and that the failure rate is very low — lower, in fact, than masonry cavity wall work. If it ain’t broke, don’t fix it. While in hot, humid climates like Florida, the Glaser model may cause problems as the Americans have found, that’s just not so in the UK.
As the way we build homes is changing, the twin issues of insulation and moisture movements become more and more important and we can expect more research findings from building scientists over the coming years. To date, none of the recent research has overturned our previous theories, but it’s serving to refine them and is helping us understand that the way buildings work is often rather more complex than we first imagined. It’s also interesting to note that research into the way historic homes behave is now informing our understanding of how to design homes today, and to know what combinations of materials we should or should not be using.
The Controversies: Multi-Foil
Perhaps the major building controversy of recent years has been based around the merits (or otherwise) of multi-foil insulation, a composite insulative product which combines multiple layers of reflective films separated by wadding. The manufacturers claim that it is 3-5 times thinner than traditional insulation but performs to the same standard, however this is disputed by several building experts as well as the Building Regulations, which requires it to be used in tandem with another insulating material. The leading suppliers, however, claim that these concerns are based on misleading testing methods — multi-foils perform poorly in ‘hot box’ tests.
Insulating Old Buildings
Another key area of debate is around how effective U values are in measuring efficiency in old buildings. The argument made by historic home campaigners is that, if reducing U values were to be the sole priority of upgrading work, the risk is that old buildings could be harmed through the inappropriate installation of some insulation techniques, potentially causing damp and other deleterious effects. Put simply, the U values measure might not be as effective in old properties as it is in new.