The last two columns looked at how ‘best practice’ for thermal comfort in houses, typified by ‘passive solar’, has failed to gain more traction and reach its promise of better thermal comfort with lower energy. Now for the other end of the spectrum: how to prevent the all-too prevalent ‘worst practice’ that has given us the least climate-adapted houses in the Western world.

And prevention requires regulation.

Thermal Regulations

We have been late to the party to introduce standards for thermal efficiency for buildings, particularly houses. The reasons are numerous and political, too complex to deal with here, except to say that it has often been held back by the construction industry who maintain a primary interest in lower costs rather than better quality.

As part of the federal government’s greenhouse gas reduction strategy, the BCA (later NCC) flagged new standards in 2000, introduced thermal standards for houses in 2003, multi-residential in 2005, and for all other buildings in 2006 in a new ‘Section J’. Regulations are always designed to prevent worst practice rather than to promote best practice, and this is never truer than in this messy area.

Importantly for everything that follows, these standards are based on reducing carbon-based energy rather than thermal comfort. Unlike lighting, ventilation or acoustic standards which address human factors, Section J is primarily about energy efficiency, its first and fundamental failure.


When we did get around to addressing thermal standards in houses (and apartments) we adopted an unusual approach. Rather than go for stringent standards for building elements, as we did in Section J for non-residential, we adopted an interactive computer program to determine the optimal standards.

The government’s scheme was called NatHERS (National Housing Energy Rating Scheme) and it promoted the idea that a computer model could analyse the areas where undesirable heat loss and gain could be prevented. Designs could be adjusted against many variables to get the best ‘bang for the buck’ in ensuring thermal performance.

This variable interactive approach was driven by two interests; firstly, the building industry wanted to be able to make trade-offs to reduce the costs; and secondly, it was only possible because Australia, through the CSIRO, had developed one of the world's best computer programs to analyse thermal comfort. Both issues play a part in the failure of NatHERS.

A History of CSIRO thermal performance programs

In 1953, the CSIRO, one of our elite intellectual jewels, started an interest in computer models for thermal comfort in variable climatic conditions. Roy Muncey, and later John Spencer, developed a seminal computer simulation program called CARE. In the 60s the program changed into STEP (one zone) and then ZSTEP (10 zones).

In the 80s, with desktop computers, Angelo Delsante developed CHEETAH for user calculation of temperatures in buildings. In the 90’s Delsante and Steve Moller from the Victorian Gas and Fuel Corporation developed CHENATH as a basis for the program that was initially called NatHERS (where the S stood for System, not Scheme).

Delsante and Yuguo Li added a natural ventilation method to the model in the late 90s. All the software programs used variable climate conditions, and CHENATH used real weather data gathered by the Bureau of Meterology or BOM. So far, so good.

Changing from comfort to energy

The CSIRO programs were all primarily aimed at an analysis of thermal comfort in a house, not energy use. But the BCA experts realised that this did not give rise to easily digested standards, nor regulations. NatHERS needed to codify the output by expressing it as the energy required to maintain thermal comfort, rather than the thermal comfort itself.

Crucially, this led to several key assumptions. Firstly, they standardised the weather into 23 ‘climatic zones’, shown in a map, the only diagram in colour in the NCC. Secondly, they assumed a desirable internal temperature range in habitable rooms during the seasons (ignoring the PMV – see here), and that there would be an air conditioner in habitable rooms to achieve those desirable temperatures.

Thirdly they assumed a standardised occupancy and when those occupants would want the assumed levels of thermal comfort (or not), during the week / weekend. They also assumed that the occupants would be interactive with the house and would open and close windows for ventilation and curtains for insulation in accordance with passive solar design principles - see here).

In this way they had standardised the inputs so that the energy required to maintain that ‘idealised thermal comfort’ could be made, and the program’s output could be expressed in numerous ways: total kWh, kWh per square metre, or ‘stars’. And those levels could be adjusted over time to raise standards.

Accurate, BERS and FirstRate

In the 90s the government, through the BCA experts, opened a ‘market’ for software programs that would meet its desired regulatory approach it now called NatHERS. All would be required to use the CSIRO CHENATH ‘engine’.

In response, the CSIRO created a far better ‘front end’ that had a better user interface, and called it AccuRate, as it is to this day. When tested against 8 other computer software modelling programs in the mid 2000s it showed that AccuRate was amongst the best across various climate simulations.

Two further software programs were developed and are still in use. In Queensland Dr Holger Willrath, as SolarLogic, developed BERS, later BERS Pro, (Building Energy Rating Software) and Sustainability Victoria developed FirstRate, now in its fifth iteration (and no, v.2.0 was not called SecondRate).

In order to ensure that the NatHERS modelling was done accurately and consistently, a system for organisations to train and accredit assessors was established. The two remaining assessor organisations, ABSA and BDAV, are subject to government regulation to ensure standards of competency in its members, that their training and accreditation is assured, in one or more of the three software modules.

How do Accurate, BERS and FirstRate work?

An accredited assessor enters all the variables of a house design, originally done arithmetically in a spreadsheet, but more recently being done graphically. This establishes the base performance characteristics of the thermal envelope. This virtual house is then run through the CHENATH program, which uses real weather data to analyse its performance.

All three programs analyse what temperatures can be expected in a typical year of weather inside all the habitable and non-habitable rooms of that house, but not the semi-enclosed external areas. It runs a ‘virtual air conditioner’, with an assumed performance rating, and measures how much energy could be expected to be used to achieve the assumed thermal comfort. ‘Bands’ of anticipated energy use are established, and a design can be benchmarked against the bands to give ‘Star Ratings’.

NatHERS starts

The assumptions within the software were never explained well, indeed they were never explained at all. When NatHERS started in 2003 all hell broke loose. The software was clunky and not always rational, showing glitches in the system. The scientists had issues.

But that was as nothing compared to architects’ reactions. Angry and confused arguments broke out. For the first time many architects, particularly of lightweight houses, were being told decisively that their designs were not comfortable, but it was expressed in terms of the energy required to run an AC unit.

“We don’t have AC” they protested. Their uber-green credentials prohibited its use. And “our clients love our houses and are totally comfortable” they said without any more scientific POE (Post Occupancy Evaluation) than hearsay.

The two cultures clashed: the scientists could prove that the houses were not thermally comfortable; but in doing so they attacked the credibility of the architects, whose clients delighted in adjusting to a much wider range of ‘comfort’ temperatures, given that the houses were so open and close to nature.

In a crushingly inept bit of PR the scientists ran many of the best AIA ‘award winning’ houses through the various NatHERS programs, and all failed. In return, the architects failed the scientists.

Where to now?

The stand-off continues unabated. NatHERS is greatly improved, and simulates with thermal mass and ventilation much better, but the assumptions remain. A whole generation of architects have come to loathe NatHERS, or at best, tolerate it.

Having spent 20 years straddling the fence (as a practising architect and Board member of ABSA) I will attempt to heal the rift next week. If successful, the week after I become Secretary General of the UN.

Tone Wheeler is principal architect at Environa Studio, Adjunct Professor at UNSW and is President of the Australian Architecture Association. Please note that I do not read Instagram, Facebook, Twitter or Linked In. My sanity is preserved by replying only to comments addressed to [email protected]. The views expressed here are solely those of the author and are not held or endorsed by A+D, the AAA or UNSW.