Why your NatHERS or HERS rating skips the small stuff that bites you

By Ben Sorensen · September 26, 2026 · 7 min read
Modern energy-efficient house exterior
Photo by Naksha Banwao on Unsplash

I was standing in a client's roof cavity in Castlemaine a few months back, torch in hand, looking at a downlight cluster punched straight through an R4 batt like a cluster of little chimneys. The house had a 7.2-star NatHERS certificate stapled to the file in the kitchen drawer. Good rating, on paper. But that ceiling had eleven can lights, a bathroom exhaust fan with a leaky flap, and a cavity slider pocket that was basically an uninsulated void running the width of the living room. None of that shows up as a deduction anywhere on the certificate.

This is the bit nobody explains properly when they hand you a star rating: the software is very good at some things and structurally blind to others. Not because the assessors are sloppy — most aren't — but because the modelling tools were built to compare bulk fabric performance across a whole house, not to audit every joint, penetration and junction where heat actually escapes. If you want to know what a NatHERS rating actually leaves out, the honest answer is: the small, cumulative stuff. And the small stuff adds up to more than people expect.

What the software actually models

NatHERS (and the US HERS index it's often compared to, though they're not identical methodologies) both work from a simplified geometric model of the house: wall areas, window areas and orientations, roof and floor construction, insulation R-values by element, and a shading profile. The engine runs that model against a climate file for thousands of simulated hours and spits out a heating and cooling load, which becomes your star rating.

That's a genuinely useful piece of engineering. It tells a builder whether swapping from R2.5 to R4 wall batts, or upgrading from single to double glazing, moves the needle enough to justify the cost. Our piece on double vs triple glazing when the extra pane is worth it leans on exactly this kind of modelling logic. But the model treats each element — wall, roof, floor, window — as a clean, uniform plane. A wall is "R2.5" everywhere across its area. There's no line item for the timber stud every 450mm that bridges straight through the insulation, no line item for the corner where two walls meet and the batt gets compressed or skipped, no line item for the top plate where insulation typically thins out before it meets the ceiling.

Thermal bridges are the classic blind spot

Timber and steel framing conduct heat far faster than the insulation sitting between the studs. In a standard timber-framed wall, studs, plates, noggings and lintels can make up somewhere around 15 to 20 percent of the wall area, and every bit of that framing is a mini highway for heat compared to the batt either side of it. NatHERS software has historically applied a bridging-adjusted R-value at the assessment stage in the accredited software tools (the current generation does account for framing factor to a degree), but the granularity depends heavily on how the plans are drawn and what the assessor is given to work with. A lintel over a large window, a steel portal frame, a balcony slab that penetrates straight through the building line as a continuous thermal bridge — these get modelled at a very coarse level, if at all, unless the assessor specifically flags them.

I'd argue this is the single biggest gap between a star rating and how a house actually feels in July. You can have a 7-star result and still get a cold stripe down the wall where the studs are, because the model averaged that away into the wall's overall R-value assumption.

Penetrations rarely get their own line item

Downlights, exhaust fans, flues, service penetrations for wiring and plumbing, the gap around a bulkhead — these are treated by the software as negligible against the total insulated area, and in a sense that's mathematically fair. One downlight isn't going to move a star rating. But a house with a genuinely leaky building envelope, ten downlights, two unsealed exhaust fans and gaps around every service penetration is a fundamentally different thermal object to one with sealed IC-rated downlights and ducted exhaust with a backdraught damper, and the certificate won't tell you which one you're looking at.

This is really an air-sealing problem sitting inside an insulation problem, and I've written before about why the sequencing matters — see air sealing before insulation: the order of operations that actually saves energy. NatHERS doesn't model infiltration in fine detail either; it uses a default air-change assumption unless a blower door test result is actually input, which on volume-build housing almost never happens. If you want the real infiltration number, you need a blower door test done separately, not the star certificate. Our guide on vetting a blower door and thermal imaging auditor covers what that separate test should actually involve.

Junctions: where two good elements meet and become a bad one

A wall might be well insulated. A roof might be well insulated. The junction where the wall meets the roof, where insulation batts often get compressed, cut short, or simply not pushed all the way to the eave, is where a lot of real-world heat loss actually happens, and it's not something the software geometry captures at that resolution. Same story at the floor-to-wall junction on a suspended timber floor, and at window reveals where the frame meets the wall cavity.

This is one of the reasons two houses built to an identical plan set, by different builders, in the same climate zone, can genuinely perform differently in the field despite carrying the same star rating on paper. It's also part of why the same house can get two different NatHERS star ratings from two different assessors — differences in how conservatively junction detail gets assumed feed into that variance too, on top of the input differences that piece covers.

What an assessor can and can't verify on-site

Design-stage ratings are modelled off plans, which means the rating assumes the junction and penetration detail gets built as drawn. Our companion piece on what a NatHERS or HERS assessor actually checks on-site goes through the as-built verification process in more detail, but the short version is: even a site assessment is largely a visual and documentary check against the approved plans, not a thermal-bridging audit with an infrared camera. A thermal imaging survey after handover is a genuinely different, and in my view underused, way to catch what the rating process structurally can't.

The relevant Australian guidance sits under the NatHERS administrative framework, which is worth reading if you want the technical basis for what the software does and doesn't account for. In the US, the RESNET HERS standard documentation covers the equivalent methodology and its own known limitations around thermal bridging treatment.

What actually catches this stuff before it's a problem

None of this means the rating is worthless — it's a genuinely useful comparison tool, and I've made the "design tool, not build receipt" argument before in our piece on why your NatHERS or HERS rating is a design tool, not a build receipt. What it means practically is that a good star rating is necessary but not sufficient. If you're building or doing a deep retrofit, ask your builder specifically how they're detailing wall-roof junctions, whether downlights are IC-F rated and sealed, and whether penetrations get taped or sealed as a discrete trade task rather than an afterthought. Ask whether the insulation installer is being supervised for compression and gaps around noggings, not just whether batts are present — our piece on vetting an insulation installer's licence has the questions worth asking there.

A blower door test after first-fix, before the plaster goes up, catches most of this while it's still cheap to fix. After handover, a thermal camera on a cold morning will show you every thermal bridge and every leaky penetration in about twenty minutes, and it's genuinely worth the money on anything above a modest reno. The certificate in the drawer tells you the house was designed well. It doesn't tell you it was built that way down to the joint.

Tagged

Common questions

Does a higher NatHERS star rating guarantee fewer thermal bridges?
Not directly. A higher star rating usually correlates with better overall fabric performance, but the bridging at studs, lintels and junctions isn't itemised on the certificate, so two houses with the same star rating can have very different real-world bridging depending on framing detail and workmanship.
Can I ask my assessor to model thermal bridging in more detail?
You can ask, and some assessors will apply more conservative framing-factor assumptions or flag known bridging risks like balcony slabs and steel lintels, but it depends on what detail is on the plans and what the accredited software allows them to input at that resolution.
Is a blower door test part of a standard NatHERS assessment?
No. A standard rating uses a default air-change assumption unless an actual blower door result is supplied and entered, which is uncommon on volume-build housing. A separate blower door test is the only way to get a measured infiltration number.
Are HERS ratings in the US better at capturing thermal bridging than NatHERS?
Both systems have similar structural limitations around junction and penetration detail, since both rely on simplified geometric models rather than element-by-element site verification. Neither substitutes for a blower door test or thermal imaging survey.
About the author
BS
Ben Sorensen
Green building · Perth, AU

Ben covers insulation, glazing and home energy ratings — the building-envelope work that decides how hard every other system has to work.

Building-certification background (NatHERS-adjacent).

More from Ben Sorensen