How to Spot a Badly Insulated Home Before You Buy
Viewing times vary widely, but however long a viewing lasts, much of that time tends to go toward admiring finishes: double glazing that still looks new, a freshly painted hallway, no obvious damp staining anywhere. None of it says much about how much heat that house actually loses once the front door closes behind you. A thermal-imaging case study of a UK home fitted with uPVC double glazing less than ten years earlier found exactly that gap between appearance and performance: frames that looked flawless from inside the room turned out, once the plastic trim was removed, to be letting daylight straight through to the brickwork behind them. The lesson is not that glazing lies on purpose. It is that looking is not the same as checking — and a buyer who knows where to check gets a far more honest picture than one who simply glances at the windows.
Key Takeaways
- Recent or new-looking double glazing does not guarantee an airtight frame: a UK thermal-imaging case study found unsealed gaps behind trim on glazing installed under ten years earlier.
- A reading only means something with a large enough temperature gap — at least 10°C between inside and outside, ideally in winter, after the building has been heated for roughly 24 hours.
- Almost all the useful signal sits in six specific spots — doors, windows, the attic hatch, outlets, recessed lights, and wall plates — not in a general glance around the house.
- Even with a blower door running, draughts around frames, sills and skylights can often be felt directly with the back of a finger.
Why new glazing isn't proof of a tight home
An infrared camera does not see light; it reads surface heat. On most readings, darker blue or purple tones mean a cooler surface, while brighter yellow tones mean a warmer one, according to a US building-science guide on using thermal imaging to spot missing insulation (basc.pnnl.gov). That single rule is enough to interpret almost everything that follows, whether the reading comes from a rented camera or a professional survey.
Timing decides whether that reading is meaningful at all. The US guide recommends checking when the gap between indoor and outdoor temperature is as large as possible — early morning or evening in winter, or mid-to-late afternoon in summer. That guidance is often sharpened into something closer to a working rule, also reflected in the UK case study: a differential of at least 10°C, preferably in winter, with the building heated for about 24 hours beforehand so the walls and frames themselves have warmed through, not just the air inside them. A quick blast of heating an hour before a viewing will not do that; the building's fabric needs time to catch up with the air temperature.
That is roughly the setup under which the UK case study scanned a house with uPVC double glazing fitted within the previous ten years. The glazing itself looked sound. The scan told a different story: the frames were not properly sealed to the surrounding wall, and once a plastic trim panel was removed, daylight was visible straight through to the brick. Two of the opening panes also leaked along part of their seals, simply because the windows were not closing fully flush. As the case study put it, "the most obvious fault was draughts around the edges of the frames, from behind a plastic trim" — a defect invisible to anyone judging the window by its glass alone.
The same survey put numbers on what that kind of leak feels like in practice. Around the front door, cold single-glazed sections read close to 15°C against a hallway ambient of about 22°C, dropping to around 10°C at the bottom of the door — recorded in December with an outdoor temperature near 0°C. Draughts also showed up from under a window sill, something made more noticeable where the sill sits directly above a warm radiator, since the contrast between the warm air rising and the cold air leaking in is sharper there than elsewhere.
When and how to check
One caveat matters before trusting any exterior reading of windows: glass can reflect the cool night sky rather than show the true surface temperature behind it, which can make a perfectly sound window look like a cold spot on a thermal image. This is specific to exterior night-time scans and worth keeping in mind if a seller or agent shares outdoor thermal photos rather than an indoor check.The most rigorous version of this check goes further than pointing a camera at a wall. It depressurises the home using a calibrated fan mounted in a doorway, with exhaust fans, the clothes dryer and other equipment that could skew the reading typically switched off, while interior doors are opened. Incoming air then shows up as visible streaks on camera, because moving air carries a different thermal signature than still air. A twenty-minute viewing cannot replicate that setup, which is exactly why a quick scan during a viewing should be read as an approximation, not a verdict. The same case study offers a simpler fallback that still relies on the blower door running: "with the blower door running, it is also easy to feel these with the back of a finger" — a tactile check a buyer can try near a frame, sill or skylight, though without that depressurisation running, draughts may be harder to feel, especially when the temperature gap is small.
For those who want to try a scan themselves, thermal cameras can usually be rented from hardware shops, and smartphone-mounted thermal attachments are also sold commercially — a US guide puts the price as low as $250 to $500 (roughly €230–€470), though that figure is undated and should be treated as indicative rather than a current price to check before buying one.
The six-spot checklist: what's normal, what's a red flag
Across both the US guide and the UK case study, the same six locations keep producing almost all the useful information. In each case, some temperature variation is expected and harmless; the signal worth acting on is an extreme, irregular, or streaked pattern rather than any difference at all.
| Spot | What's normal | Red flag |
|---|---|---|
| Exterior doors | A little temperature variation at the edges or near the glass | Dark colouring or streaking during a depressurised check — poor fit, missing weatherstripping, or an unsealed opening |
| Windows | A small, even difference right at the frame edge | Extreme or irregular contrast, dark patches at the sill, and streaks across the meeting rail — on single-hung double-pane windows this kind of leak can often be corrected by adjusting the sash, replacing the weatherstripping, or caulking, though in some cases it may point toward replacement |
| Attic hatch | Even temperature across the hatch door itself | Dark areas and streaking around the opening, which signal heat loss and air leakage; dark colouring across the door means the hatch itself needs insulating |
| Outlets and switches | A slight difference is common, especially on an exterior wall | A large dark area around the fitting, pointing to significant air leakage — one documented example showed a clear cold spot at an outlet simply from missing insulation behind it |
| Recessed lights, vents, skylights | An even, lightly toned fixture or shaft with no obvious cold spotting | A very dark area with streaks radiating outward, signalling heat loss plus air leakage; a skylight shaft should read even, not show purple streaking down its length |
| Walls, including the top and base of the wall | An evenly warm surface across the whole wall | Dark streaking along the top of the wall or at its base, especially in corners where two exterior walls meet — usually missing insulation or an air path behind the plaster |
One wording note is worth flagging: US guidance often refers to the "top plate" and "bottom plate" of a wall, terms from American timber-frame construction where those are literal horizontal structural members. The table above uses "top and base of the wall" instead, since not every home is built that way. The underlying principle still holds regardless of construction type: the junction where a wall meets the ceiling, the junction where it meets the floor, and the inside corner where two exterior walls join are consistently the places where insulation gaps and air paths show up first.
Conclusion
A thermal scan, taken on one morning under one set of conditions, is a snapshot of how the building behaved that day — not a lifetime rating of its energy performance. What it is genuinely useful for is narrower and more practical: flagging likely air-leakage and insulation gaps under the right conditions, which is precisely the kind of information a viewing cannot otherwise provide. Whether that translates into a lower offer, a request for repairs, or simply a more realistic heating budget is a separate judgment call for the buyer to make. The more interesting question for anyone currently house-hunting may not be how to run the check, but whether to ask the seller if one has already been done — and if not, why not.
