Soft Machines

I’m off to the seaside. Normal service will resume in a couple of weeks.

A surprisingly large fraction of the energy used in developed countries is used heating and lighting buildings – in the European Union 40% of energy used is in buildings. This is an obvious place to look for savings if one is trying to reduce energy consumption without compromising economic activity. A few weeks ago, I reported a talk by Colin Humphreys explaining how much energy could be saved by replacing conventional lighting by light emitting diodes. A recent report commissioned by the UK Government’s Department for Environment, Food and Rural Affairs, Environmentally beneficial nanotechnology – Barriers and Opportunities (PDF file) ranks building insulation as one of the areas in which nanotechnology could make a substantial and immediate contribution to saving energy.

The problem doesn’t arise so much from new buildings; current building regulations in the UK and the EU are quite strict, and the technologies for making very heat efficient buildings are fairly well understood, even if they aren’t always used to the full. It is the existing building stock that is the problem. My own house illustrates this very well; its 3 foot thick solid limestone walls look as handsome and sturdy as when they were built 150 years ago, but the absence of a cavity makes them very poor insulators. To bring them up to modern insulating standards I’d need to dryline the walls with plasterboard with a foam-filled cavity, at a thickness that would lose a significant amount of the interior volume of the rooms. Is their some magic nanotechnology enabled solution that would allow us to retrofit proper insulation to the existing housing stock in an acceptable way?

The claims made by manufacturers of various products in this area are not always crystal clear, so its worth reminding ourself of the basic physics. Heat is transferred by convection, conduction and radiation. Stopping convection is essentially a matter of controlling the drafts. The amount of heat transmitted by conduction is proportional to the difference of temperature, the thickness of the material, and a material constant called the thermal conductivity. For solids like brick, concrete and glass thermal conductivities are around 0.6 – 0.8 W/m.K. As everyone knows, still air is a very good thermal insulator, with a thermal conductivity of 0.024 W/m.K, and the goal of traditional insulation materials, from sheeps’ wool to plastic foam, is to trap air to exploit its insulating properties. Standard building insulation is made from materials like polyurethane foam, are actually pretty good. A typical commercial product has a value of thermal conductivity of 0.021 W/m.K; it manages to do a bit better than pure air because the holes in the foam are actually filled with a gas that is heavier than air.

The best known thermal insulators are the fascinating materials known as aerogels. These are incredibly diffuse foams – their densities can be as low as 2 mg/cm3, not much more than air – that resemble nothing as much as solidified smoke. One makes an aerogel by making a cross-linked gel (typically from water soluble polymers of silica) and then drying it above the critical point of the solvent, preserving the structure of the gel in which the strands are essentially single molecules. An aerogel can have a thermal conductivity around 0.008 W/m.K. This is substantially less than the conductivity of the air it traps, essentially because the nanscale strands of material disrupt the transport of the gas molecules.

Aerogels have been known for a long time, mostly as a laboratory curiousity, with some applications in space where their outstanding properties have justified their very high expense. But it seems that there have been some significant process improvements that have brought the price down to a point where one could envisage using them in the building trade. One of the companies active in this area is the US-based Aspen Aerogels, which markets sheets of aerogel made, for strength, in a fabric matrix. These have a thermal conductivity in the range 0.012 – 0.015 W/m.K. This represents a worthwhile improvement on the standard PU foams. However, one shouldn’t overstate its impact; this means to achieve a given level of thermal insulation one needs an insulating sheet a bit more than half the thickness of a standard material.

Another product, from a company called Industrial Nanotech Inc, looks more radical in its impact. This is essentially an insulating paint; the makers claim that three layers of this material – Nansulate will provide significant insulation. If true, this would be very important, as it would easily and cheaply solve the problem of retrofitting insulation to the existing housing stock. So, is the claim plausible?

The company’s website gives little in the way of detail, either of the composition of the product or, in quantitative terms, its effectiveness as an insulator. The active ingredient is referred to as “hydro-NM-Oxide”, a term not well known in science. However, a recent patent filed by the inventor gives us some clues. US patent 7,144,522 discloses an insulating coating consisting of aerogel particles in a paint matrix. This has a thermal conductivity of 0.104 W/m.K. This is probably pretty good for a paint, but it is quite a lot worse than typical insulating foams. What, of course, makes matters much worse is that as a paint it will be applied as a very thin film (the recommended procedure is to use three coats, giving a dry thickness of 7.5 mils, a little less than 0.2 millimeters. Since one needs a thickness of at least 70 millimeters of polyurethane foam to achieve an acceptable value of thermal insulation (U value of 0.35 W/m2.K) it’s difficult to see how a layer that is both 350 times thinner than this, and with a significantly higher value of thermal conductivity, could make a significant contribution to the thermal insulation of a building.