CSIRO Report On Reflective Paint
Building Research Establishment Information Paper
P Whiteley, BSc, MRSC, FTSC
And D Gardiner, BSc, PhD
Solar Reflective Paints
The theoretical advantages of solar reflective paints for roofs are not always realised in practise, and the performance of several examples tested has not been good enough to demonstrate wholly satisfactory products of long life at acceptable cost. Some paints lost reflective efficiency rather quickly, others tended to crack and could cause cracking of bituminous and asphalt surfaces. Nevertheless, benefits can be expected from the correct use of some paints and improvements are expected in the near future. This paper is intended to provide guidance on the relative merits of paints and other treatments and on the range of results likely to be achieved.
Many roof surfacing materials absorb a large part of the solar radiation which falls on them, this is mainly converted to heat which can affect their behaviour and can be transmitted (depending on the conductivity of the roof) to other components and the internal air. A small amount (in the ultra-violet) is absorbed and causes degradation of organic materials, e.g. roofing products and paints.
The theoretical advantages of roof surfaces which reflect solar energy are considerable and are generally accepted. The benefits include:
- A reduction in the rate of deterioration caused by thermal chemical reactions
- If the surface reflects the ultra-violet part of the spectrum (or absorbs it harmlessly), a reduction in photochemical indeed degradation.
- A reduction of mechanical stresses because the maximum thermal expansion of the roofing material and structure is requested.
- Where desirable (e.g. in hot climates or in temperate climate not summers), a possible reduction of temperature in the roof and the interior of the building.
There are possible disadvantages as well. Cooler surfaces remain damper longer, which may increase those forms of degradation (such as hydrolytic mechanisms) which depend on moisture. Also if the cooling effect keeps the temperature below the softening point of bitumen; or the glass transition temperature of polymer for increased periods, there is more risk of brittle fracture and less chance of recovery of strains through creep mechanisms. However, the theoretical disadvantages seem to be less important in practice and the net result of using reflective surfaces is usually a gain in durability, although less than might be expected, unless unsuitable paints area used.
The energy of the solar spectrum is distributed between ultra-violet (4 to 5 per cent), visible about 40 per cent and infra-red (about 55 per cent). White surfaces are highly reflective in the visible range, but absorb UC and some IR. Darker colours absorb progressively more visible light and most black surfaces absorb over 95 percent of the total. However some dark and black pigments have good IR reflectance and can produce a lower solar heat gain than their appearance suggests. Aluminium metal and paints when new have good specula (mirror) reflection from UV to IR, about 55 per cent, with a similar figure from reflectance meters, but their overall efficiency is reduced by their low emissivity but not enough to compensate for their increased absorption. Dirt deposited on reflective surfaces greatly reduces their efficiency.
The reduction of temperatures at any level of particular roof structure, or of the interior air, depends on the thermal conductance of the roof and the only figures that can be usefully quoted are those immediately on or below the reflective surface.
There have been many measurements published from various sources which indicate that a black surface on a poor conductor at peak periods of radiation e.g. mid-day with bright sun and little wind and an air temperature of 25°C, may reach 80°C, whereas a white surface with a solar reflectance of 80 per cent will only reach 40°C. The increasing levels of roof insulation now being adopted and the use of wans deck designs especially those with the insulation immediately below the waterproof layer, may well produce greater fluctuations of surface temperature and demand efficient solar reflection.
It is sometimes believed that a white roof will be colder at night; except for the period around sunset when the roof is already cooler this will not be true because the emissivity which causes loss of heat by radiation is the same for most non-metallic surfaces regardless of colour. A bright aluminium metallic surface, on the other hand, with a somewhat lower reflectance but a much lower emissivity, may be warmer than a white one during the day but also0 warmer at night.
Methods of Improving Solar Reflectance of Asphalt and Bituminous Roofing
- The normal sanded finish which is marginally better than a black surface (and is protective by absorbing UV radiation before it reaches the sensitive bituminous material).
- A layer of white or light-coloured mineral chippings (e.g. white spar).
- Asbestos cement, concrete or grc tiles or slabs.
- The application of suitable paints or light-coloured liquid waterproofing coatings.
Note: Various white-pigmented polymer-based sheets are available as alternative to bitumen-based sheet roofing, and bitumen felt can be obtained with metal foil facing.
If paints are used they need to be specifically formulated for compatibility with roofing materials, particularly for asphalt and bituminous felt. The use of any other paints can be disastrous; bituminous surfaces may bleed and turn the paint brown, or, worse still, may be cracked to a considerable depth by the shrinkage of the paint film and subsequent weathering.
A flexible binder is an obvious requirement to accommodate the movements of the roofing material, particularly where bitumen is present as it may soften when hot. Paints may be based on water-thinned emulsions or solvent-thinned resins or polymers. Aluminium pigmented paints are most frequently based on bituminous binders. Emulsion paints are generally more convenient in use but sometimes have inferior adhesion. Some may require a specific primer.
The relative merits of suitable paints versus chippings:
- Negligible extra loading on the roof (compared with up to 16kg/m2).
- Slightly higher reflectance when new, with possibly greater retention of reflectance on; ageing, although not invariably.
- Paints do not conceal any points of failure or leakage.
- Chippings may block rainwater outlets or be blown about in strong winds. Paints permit light. Occasional maintenance foot traffic may press chippings into the surface and damage it.
- Paints are less permanent and they need renewal every few years Chippings may become dirty and lose reflectance but have some capacity to absorb heat.
- Paints will not perform satisfactorily where ponds form because of deposition of dirt or algae, or because the alternate swelling and shrinkage increases their rate of breakdown. Chippings may not be completely submerged by shallow pools but algae can grow on them and they will also restrict the drying out.
Chippings and painting may be complementary e.g. when used on flat and pitched areas of one roof respectively; paints may also be used to restore the reflectance of chippings (by spray application) or of tiles and slabs, and relatively cheap emulsion paint can be used if it does not come into contact with asphalt or bitumen.
Solar reflective paints have been marketed for many years with varying results. Bad application or poor surface conditions always affect the performance of coatings, and failures can often be attributed to these factors. Even so, a comparison of most of the available paints, stated at the Building Research Station in 1977, illustrated the wide diversity of behaviour and the general failure to perform well for an adequate time. The test surfaces included new bitumen felt sheets on a rigid polyurethane foam base at 3° pitch and 30° pitch with a ponded area, old bitumen felt sheet on a wooden base, and a re-asphalted roof. (Earlier work on various roof substrates had included many temperature measurements which were not repeated systematically in later tests, adequate correlation with reflectance (weighted according to spectral energy distribution) being known.
The various coatings deteriorated in different ways, the least serious being solely the loss of reflectance, which could fall from 70 to 40 per cent in one year on pitched areas and to 20 per cent in ponded areas. But other coatings showed cracking of the felt, or of hot-applied bitumen adhesive exuded from the edges of the joining strips, or of asphalt, which could danger their waterproofing ability. This was caused by the shrinkage properties of the paint on ageing. Once a crack is produced it will absorb radiation, become hotter than the adjacent surface and penetrate further into the substrate. During the night the substrate hardens and prevents a return to the original state, a ratchet effect. One white coating showed fairly severe failure of this type, a second moderately bad behaviour and the other remained intact up to 2 years although losing reflectance.
The aluminium-pigmented bituminous paint lost reflectance but remained in good condition and such paints can be regarded as particularly protective of the bituminous felts, emulsions and asphalts, aluminium paint tended to produce blistering. Although undoubtedly moisture must have been present to cause the blisters, it is assumed that they increased because of the low vapour permeability of aluminium paint.
The best result both for temperature reduction and for durability was produced by a liquid waterproof coating, applied at the manufacturer's specified thickness which is far in excess of that for normal paint and hence costs considerably more. The temperature reduction was achieved even though the colour was distinctly yellowier than the other paints, and was maintained as a result of chalking; the coating also showed a marked surface crazing but this did not result in breakdown of the film in 3 years. (Other thick, liquid, roof coatings in different tests did not show similar crazing but serious soiling reduced their reflective efficiency.)
The behaviour of the paints on an asphalt roof was generally worse than on the bitumen felt. Trouble had been experienced when the paint had been applied during the autumn; dew formation interfered with recoating next morning, and persisted longer over the colder, more reflective sections. The liquid waterproofing composition failed in the absence of a satisfactory primer, which has since been rectified. The second most durable coating on felts was also durable on asphalt but was too severely darkened after 2 years to be an effective solar reflector. The best overall results were produced by an emulsion-based paint which has been sold for many years for this use, giving better results on asphalt than on bitumen felt.