A good star rating tells us something important about a home. But it does not tell us everything about what living there will be like. Understanding what a rating measures — and what it leaves unanswered — can lead to better decisions.

An early design for our own 39S House in Brisbane achieved a 9.6-star NatHERS rating.

The house we eventually built achieved 7.7 stars.

That sounds like a backwards step.

If the objective had been simply to achieve the highest possible NatHERS rating, it would have been. But the difference between those two numbers became useful because it forced us to ask a more important question:

What are we actually trying to optimise?

What is actually being rated?

NatHERS — the Nationwide House Energy Rating Scheme — models the thermal performance of a home.

It considers things including orientation, construction, insulation, windows and shading, and calculates the predicted annual energy required for heating and cooling under standardised assumptions. The result is expressed as a rating between zero and ten stars.

A higher rating generally means the building fabric should require less heating and cooling to maintain the modelled conditions.

That is valuable information.

NatHERS has made the thermal performance of Australian homes visible and comparable in a way that individual insulation values, glazing specifications and heating loads never could.

The National Construction Code has progressively increased the required thermal performance of new Australian homes, with NCC 2022 generally requiring the equivalent of 7-star NatHERS performance for new houses.

And NatHERS provides considerably more information than the headline star. Assessors can examine separate heating and cooling loads to understand what is driving the result.

The problem is not the measure.

The problem comes when we ask it to tell us more than it measures.

When 9.6 stars wasn't the answer

39S is an existing nineteenth-century worker's cottage in Petrie Terrace.

Its western side faces severe afternoon sun as well as traffic and stadium noise. The project needed to bring better daylight and ventilation into the existing house, manage those external conditions and make intelligent use of what was already there.

Thermal performance was an important part of that.

It wasn't the only part.

The early 9.6-star scheme showed us what could be achieved when the design was pushed particularly hard towards the NatHERS metric. It gave us a valuable benchmark.

But as the design developed, other considerations changed the architecture.

Existing fabric could be retained rather than replaced. Materials could be reused. Daylight needed to reach the old cottage. Natural ventilation needed to work when conditions allowed it. The western extension had to protect the house from afternoon sun while also helping shield it from traffic and stadium noise.

Those decisions had consequences.

One of them was that the NatHERS rating fell from 9.6 to 7.7 stars.

We did not discover that 7.7 stars was somehow better than 9.6.

Nor did thermal performance suddenly cease to matter.

Instead, the modelling allowed us to understand what we were gaining and giving up as the design developed.

That is an important distinction.

The highest possible score on one measure is not necessarily the best measure of the building as a whole.

And that raises the more useful question: what else do we need to know?

Performance is not one thing

Imagine two houses that both achieve seven stars.

One performs particularly well in winter but struggles during prolonged summer heat.

One has a western bedroom that becomes considerably hotter than the rest of the house.

One depends heavily on windows being opened at particular times.

One performs well thermally but has poor indoor air quality.

One is comfortable under the climate against which it was modelled but has little margin as summers become hotter.

Their headline rating might be the same.

Living in them may not be.

That is not a flaw in NatHERS. It is a consequence of asking a complex building to answer a particular question with a single number.

NatHERS estimates heating and cooling demand under standardised assumptions. It does not, on its own, establish long-term thermal safety, indoor air quality, moisture resilience or comfort under future climate conditions.

Those questions require different evidence.

People don't live in an annual average

Whole-house performance is useful because we need consistent ways to compare buildings.

But people occupy particular rooms, at particular times, during particular weather.

A bedroom at two in the morning matters differently from an unoccupied room in the afternoon.

A western room may behave very differently from one facing south.

And the duration of an uncomfortable condition matters.

A room becoming hot briefly is different from remaining hot throughout the night. A single hot afternoon is different from several hot days followed by nights that never properly cool.

Recent research in southeast Queensland makes this distinction particularly relevant.

Researchers monitored people aged 65 and over in their own bedrooms during summer, recording both room temperatures and physiological responses while they slept.

They found progressively stronger physiological responses as bedroom temperatures increased.

The research does not establish a universal temperature above which a bedroom becomes unsafe. Nor does it mean the results can simply be applied to every person or every house.

But it reinforces an important distinction.

Annual energy demand and what happens to someone in a particular room during a particular extreme event are different questions.

Both can matter.

Airtightness is not ventilation

Better-performing buildings also tend to place greater emphasis on controlling unwanted air leakage.

That is sensible.

Uncontrolled gaps around windows, doors and construction junctions allow conditioned air to escape and outside air to enter unpredictably.

But accidental air leakage has historically provided some of the fresh air in Australian homes.

As envelopes improve, we need to become more deliberate about what replaces it.

Air leakage and ventilation are not the same thing.

Leakage depends on wind, temperature and pressure. We cannot reliably control how much air enters, where it enters or whether it arrives when we want it.

Ventilation is intentional.

Sometimes the solution is straightforward: well-positioned openings and occupants who can use them when outdoor conditions are favourable.

But windows are not always a reliable ventilation strategy.

Outside air may be hot or humid. Bushfire smoke may make it undesirable. Traffic or aircraft noise may make open windows impractical. Security, insects and heavy rain can change how a house is actually operated.

In a more airtight building, deliberate mechanical ventilation may therefore become appropriate.

The important point is not that every house requires the same system.

It is that a good thermal rating does not, by itself, establish good indoor air quality.

Moisture is another question again

Temperature alone does not describe what is happening inside a building.

Humidity changes how heat feels to the body. It also affects how moisture behaves within the building fabric.

This becomes particularly important in warm, humid climates.

Cooling indoor air can create large temperature differences across walls, roofs and other parts of the envelope. Vapour, condensation, insulation, ventilation, material choice and construction detailing begin to interact.

Poorly understood, those interactions can create conditions that damage materials or support mould growth.

These are building-physics questions.

They matter to comfort and durability.

But they are not what a NatHERS star is designed to tell us.

Again, we have reached the edge of the question being measured rather than discovered a failure in the measure itself.

The climate will not stand still either

There is another difficulty when designing a building expected to last for many decades.

The climate during its later life will not be identical to the climate in which we design it.

NatHERS requires defined climate data so buildings can be modelled and compared consistently. Its climate files were substantially updated for NCC 2022 using more recent weather information.

That is important.

But a rating remains an assessment against a defined climate.

A house designed today may still be occupied in 2050, 2070 or beyond.

So there is another question worth asking alongside the rating:

What happens as the conditions change?

Future-climate modelling and testing against periods of sustained heat can reveal things an annual star rating was never intended to establish.

This is particularly important when a design performs close to a threshold.

Compliance tells us that the threshold has been crossed.

Resilience asks what happens when the conditions move.

So should we aim for more stars?

Yes — where doing so represents a sensible improvement to the building.

Better orientation, appropriate shading, insulation, glazing and a well-designed envelope can substantially reduce heating and cooling demand.

If the same house can achieve eight stars rather than seven through intelligent design, that is generally a worthwhile improvement.

But that is different from assuming the objective is to maximise the number regardless of the consequences.

Every intervention has consequences of its own.

More material has an embodied impact.

Higher-performing components cost money.

Additional layers and systems can add complexity and maintenance.

Some interventions produce substantial improvements. Others deliver increasingly marginal gains.

There comes a point where the next design decision may be more useful somewhere else.

Perhaps it is better external shading.

A ceiling fan.

Improved ventilation.

Better moisture management.

A smaller zone that can be efficiently conditioned during extreme weather.

Or simply retaining a useful part of an existing building rather than demolishing and replacing it.

There is no universal answer.

That is precisely why the rating is an input to a design decision rather than a substitute for one.

Measure what you can. Then understand what it means.

The lesson from 39S was not that star ratings don't matter.

Without the NatHERS modelling, we could not have understood the consequences of the design changes as clearly.

The 9.6-star scheme established what was possible.

The eventual 7.7-star result told us something about the choices we had made.

Neither number, on its own, could tell us whether we had designed the right house.

That required looking at the building more broadly: thermal performance, daylight, ventilation, noise, material use, existing fabric, future conditions and how the house would actually be lived in.

That is the useful role of performance modelling.

Measure what can be measured.

Understand what the measure means.

Then ask what remains.

Because a building is not a score.

It is somewhere people will sleep, breathe, cook, work, recover and grow older. It will experience weather and conditions that no model can predict perfectly, and it will consume resources in ways that no single metric can describe.

The aim is not to find one number capable of capturing all of that.

It is to use the information we have to make better decisions about what matters.