It seems like a no-brainer: cooling of buildings and neighbourhoods can be helped by having light coloured roofs. But the question has caused no end of controversy. Let’s settle it once and for all.

What does the science say?

The technical name for a roof that reflects heat is that it has a high albedo.

High-albedo roofing has two scientifically proven benefits: it helps to mitigate the urban heat island effect and reduce cooling energy demand and costs. It can also increase thermal comfort in non-conditioned buildings.

Benefits are both direct and indirect. 

Direct benefits to individual buildings occur by reducing absorbed shortwave radiation through the roof. Neighbourhood-scale indirect benefits result from reduced ambient air temperatures, particularly when high-albedo surfaces are deployed on a large scale. 

These conclusions come from several sources, including one, (published in the journal Energy and Buildings) which comprised whole-building energy simulations of a set of archetypical single family residential buildings in three locations with distinct characteristics in the Los Angeles area (one coastal and two inland). This location is about the same degrees latitude north of the equator as many Australian cities are south of it. 

The simulations show that benefits from the indirect effect can be the same order of magnitude as the direct effects. The systematic replacement of dark surfaces with white could lower heat wave maximum temperatures by 2 degrees Celsius or more, according to other research from Yale University. 

A recent study published in the Proceedings of the National Academy of Sciences found that in the absence of adaptive urban design, and separate from climate change, urban expansion increases average temperatures by 1-2°C. therefore if those roofs were painted a light colour this effect would be cancelled out.

But these benefits depend on the climate and building characteristics. For heating-dominated climates, white roofs may not be appropriate.

The highest energy and thermal comfort benefits were observed in the Los Angeles study in a low-performance building (defined by airtightness and ceiling insulation levels). There, simulations indicated an energy savings of 41 per cent and thermal comfort improvement of 23 per cent due to a combination of direct and indirect effects.

Incidentally, Los Angeles uses the albedo effect to cool itself also by painting several streets in a light grey paint to reduce road-top temperatures by as much as 5.5°C. 

Roof on a building in Los Angeles, California, given a highly reflective surface to keep out unwanted solar heat. Source: Koskim, on behalf of The Vinyl Roofing Division of the Chemical Fabrics and Film Association, CC BY-SA 3.0.

What roofs for which parts of Australia?

Research commissioned by the RTAA, performed by the University of Newcastle, found that light coloured tiles yielded “energy savings between 25-36 per cent compared to dark coloured tiles”. 

Lower cooling energy demand was found even with insulation or sarking, although it is smaller. 

The sustainability guide Your Home 5th edition states that light coloured roofs are estimated to reflect up to 70 per cent of summer heat gain – around 50 per cent more than a dark roof. Dr Chris Reardon, principal author, said: “In cooling-dominated climates, you need more than one layer of sarking – two layers separated by 25 mm gap plus bulk insulation”. 

A dark roof could be beneficial in colder, heating-dominated climates like Canberra and possibly even Melbourne, but it shouldn’t be considered north of Brisbane.

Many thermal rating tools, such as those used to assess NatHERS ratings, work off historical climate data. Houses built now might not be prepared for the climate change affected temperatures of the future. The publication says that it would be prudent to look at projected temperatures over the 50-80 year lifetime of a house, with CSIRO research predicting temperatures up to 4°C warmer on average by 2100. 

Therefore even in southe000rn parts of Australia, homes should be properly insulated to protect both heat and cold extremes, and light roofs should be installed.

Other types of roofing

Of course other environmentally sound types of roof are possible. Roofs may be “green” – a layer of sedum vegetation –  or covered in photovoltaic panels. 

A study in Adelaide found that besides delivering cooling in summer, green roofs also act as an insulating layer to keep buildings warmer in winter. It found that covering 30 per cent of the roof area in vegetation would reduce the electricity consumption by 2.56 (W/m2/day).

Building control

As we reported before, some councils are intrinsically opposed to light-coloured roofs and turn down planning applications.

“White roofs would generally be incongruous with heritage items and are likely to be incongruous with the character of council’s heritage conservation areas because they would be in stark contrast to the overall character of those areas,” one council spokesman has said. 

“Development controls in such areas require council to have regard for the colour and finishes of materials, assessing each case on its merits. In some cases a white roof may be acceptable as part of a contemporary in-fill design. 

“White or light coloured roofs are also very reflective in nature, and given the undulating landscape in the Woollahra area it is necessary to consider the impact light reflected from a white roof would have on neighbouring properties overlooking the roof when considering any application.” 

But although a surface may reflect more heat it does not necessarily produce more glare or light reflection: pale colours with a matt finish, for example.

Why do light roofs cool and dark roofs heat?

All materials have the capacity to absorb and emit the energy they receive from the sun. This can be used to maximise the proportion of incident solar energy absorbed and control how much is emitted. 

Absorption is a measure of how much incident radiation is received by a material. The emissivity of the surface of a material is a measure of its effectiveness in emitting energy.

Absorption and emittance of incident solar energy in a material.

The emitted energy is in the form of long wave or thermal radiation. In scientific terms:

  • Absorption refers to the ability of a material to absorb solar radiation (approximately the wavelengths 400 nanometre – 10 micrometre (UV-A, visible and near infrared).
  • Emissivity or emittance refers to the ability of a material to emit infra-red radiation (from materials with temperatures between -40°C to 100°C, of approximately the wavelengths 10 micrometre – 100 micrometre).

Since energy cannot be lost or destroyed, any radiation not absorbed into a material will be reflected. 

The radiation that is absorbed will have the effect of heating the material, which then itself will emit heat (infrared radiation) in order to reach the same temperature as its surroundings. 

If a material absorbs a great deal of solar radiation it is said to have a high absorption coefficient. If it emits a lot of radiation it is said to have a high emissivity coefficient. 

The converse is true for low absorption or emissivity rates. 

  • An absorption coefficient of 1 would mean that all of the solar spectrum radiation was absorbed, and of 0 would mean that none of it was absorbed and all of it was reflected. By definition a true “black body” surface has an absorption coefficient of 1
  • An emissivity coefficient of 1 would mean that all of the incident radiation was re-emitted as infrared radiation, and of zero would mean that none of it was

(In practice these extremes are not reached.)

Selective surfaces

The need for spectral selectivity is dependent on the intensity of the incoming solar radiation and the temperature at which the surface is to operate. 

A totally black absorber and a totally white reflector will absorb and reflect respectively all solar energy arriving on them, regardless of wavelength. 

Selective absorbers will absorb only in the spectral region of the solar radiation and be transparent in the thermal infrared. Selective reflectors will reflect only in the thermal infrared and be transparent in the spectral range. This is used in the design of photovoltaic solar panels.

In general the solar energy absorbed by a wall or roof can be approximated according to the surface colour (see Table 1 below). The amount of solar energy absorbed also depends upon the angle at which it arrives (See Table 2). 

The emissivity and absorption coefficients for some common roofing and building materials can be found in Table 3 below. For the purpose of maximising the cooling effect, choose a low ratio between absorption and emissivity. (The coefficients for some products varies with the temperature. As a guideline, the figures are based on a temperature of 300oK. )

The amount of solar energy absorbed also depends upon the angle at which it arrives, as shown here.

[table id=28 /]

[table id=29 /]

[table id=32 /]

David Thorpe is the author of Passive Solar Architecture Pocket Reference, Solar Technology and Sustainable Home Refurbishment

Join the Conversation


Your email address will not be published. Required fields are marked *

  1. I’ve come to a very strange conclusion: that if a material absorbs visible light well but cannot absorb IR well… this meaning it will be unable to emit IR equally, then it would heat up until it starts to emit in the visible range;: the implication being that you could end up with the paradoxical situation of a surface becoming warmer from a cooler source, thereby breaking a fundamental law: that heat travels from hotter to cooler.

  2. This is a great article and your research is well done.
    The theory around foil is rubbish (if you live outside a vacuum), the sooner it leaves the BCA the better. The whole theory is based on a still air gap, which loses its effectiveness with any ventilation.
    The new ICANZ manual mentions condensation 96 times because in any part of Australia that falls below 14 deg C the risk needs to be assessed, foil sarking is a cheap air barrier but it can cause condensation issues with leads to mould.
    The foil association guys need to see the warning expressed in the original paper, by Robinson and Powlitch p23. Foil like Vapour Permeable Membranes’s can help improve airtightness, foil can be used somewhere north of Rockhampton as overnight temperature don’t fall below 14 deg C.
    Light or dark roofs do influence the heat island effect, yet reflecting the heat back has a warming issues for the atmosphere. Using solar panels could be an effective solution as it converts the energy to electricity rather than heat.

    Looking forward to the next article.

  3. “aluminium foil insulations under any roof”
    There are two issues here:
    1. aluminium foil is effective against “radiant” heat, so great installed as the outer surface, but is very poor insulator of conductive heat (which is what happens once the heat gets through the roofing material).
    2. it does not take long for a layer of dust to deposit on the upper surface of the foil. This layer of dust almost completely changes the properties of the surface.

    Foil facing downwards in the roof space (even reflective sarking) is of benefit in winter, as it does not get any dust on it.

  4. I have been amazed and horrified driving around new build Western Sydney suburbs to see whole tracks of new houses with BLACK tile roofs.

    So I read this article hoping to read that they were high ‘albedo’ and actually really cool in summer. Didn’t see it.

    In which case – is this a stitch-up – by the building supply industry ?

    Or is black considered more overall-energy-efficient because it absorbs more heat in winter ?

  5. It’s astonishing this the highly important topic for the FIFTH ESTATE to reveal, has not incorporated the tremendous thermal power of durable aluminium foil insulations under any roof.

    Once a downward facing pure aluminium (99% purity) surface exists, facing a 50-100mm airspace clearance, the astonishing scientific properties of low emissivity of any downward radiation is replused to the tune of approx 97%, whereby only 3% (the reciproxal of 97% reflectivity) is emitted downward. NB mist upward facing foil mebranes are coloured plastic and hence have zero thermal contribution against radiation – ie they are highly absorptive with close to zero reflectivity.

    The point of the story?

    The colour of a roof can largley be a defused thermal issue, when foil exists beneath. The watts of energy per sqm will be drastically reduced by foil. Of course, a light coloured roof must mathematically make an additional contribution.

    The power of a single aluminium surface will offer up more freedom for people to choose the colour of their roof. This leads to substantial lowering of energy demand to run cooling appliances.

    Numerous examples exist where bulk insulations cannot resist hugh intensity radiant energy loads upon roofs. Well how can they when the recently revised insulation standard AS/NZS 4859.1 & .2 (2018), slated for enforcement into the NCC May 2019, has wilfully refused to account for radiant energy flows upon all and every insulation material.

    Its incredible that the radiation on buildings subject rarely gets oxygen for vigorous public discussion.

    The ABCB and Standards Australia are both guilty on this point. Why? Because both of them will not urgently push the case for mandating durable radiant heat controlling mechanisms such as foil beneath roofs. An attempt was made in 2013 but it quickly fizzled out.

    The reason is the bulk insulation industry continue to have the whip hand in national thermal insulation regulations, since about 1986. They chair all the Standards committees.
    Its a stitch up. They write the standards and any pro-foil voices on standards have been crushed. To make matters worse, the the Consumers Federation of Australia on the relevant standards committee will not and never has demanded that there be radiant heat control of residential roofs. I believe that person has been on the committee for 20-25 years.

    I am the sole person who has repeatedly tried to open this discussion and have engaged numerous politicians over the 8 years to try to see sense and reason. To date only one federal politician has risen to the occasion – just one – Tim Wilson MP (Vic) who spoke on 16 October 2017 in parliament about the power of aluminium foil and it’s ability to contribute to ‘demand management’ of energy use.

    This subject needs rapid acceleration for further FIFTH ESTATE discussion, who seem to be the only media mechanism brave enough to raise tough and complex subjects, that mainsteam media refuse to touch.

    In the Public and National Interest.