Heat pump not working in winter?

Heat pumps work in a UK winter. The Met Office 1991 to 2020 UK January mean, the coldest month of the year, is 3.9°C. Modern air-source units are rated to keep heating down to between -20°C and -28°C outdoor air, well below the -2 to -4°C the English heating-design standard plans for, and they hold a coefficient of performance above 2.5 even at -7°C. The cold-weather myth survives anyway, and it is worth understanding why. Three things feed it.

1. Real physics, worded badly. A heat pump does get less efficient as the air outside cools, which is true of any machine moving heat against a bigger temperature gap. "Less efficient" becomes "doesn't work" in the retelling, and the two are not the same thing. A unit running at a COP of 2.6 on the coldest morning of the year is still delivering more than two and a half times the heat a direct electric heater would for the same electricity.

2. The early-adopter story. A wave of UK heat pumps went in during the 2010s, before the trade had settled on good design practice and before the training caught up. Some were undersized, some were bolted onto radiators far too small for a lower flow temperature, and a few households were left cold and angry. Those accounts travel widely. The hardware and the design standards have both moved on a long way since, but the memory of a neighbour's bad winter lingers.

3. Geography. A great deal of the English-language internet writing about heat pumps in the cold is written for North American or continental climates, where a -25°C night is a normal event. The top result for many UK searches about winter heat-pump trouble is an American troubleshooting page describing a problem most British homes will never see. The rest of this guide is about the UK specifically.

Yes, and with a wide margin. The first fear behind "not working in winter" is that the unit simply shuts down when it gets cold. Every major monobloc sold in the UK is rated to keep heating far below anything the British climate produces. The table below lists the rated minimum operating temperature for the five ranges that dominate UK installs.

Rated minimum operating temperature and maximum flow temperature for the five air-source heat-pump ranges that dominate UK installs.

Heat pump	Refrigerant	Rated to	Max flow
Daikin Altherma 3 H HT	R32	-28°C	70°C
Vaillant aroTHERM plus	R290	-25°C	75°C
Samsung EHS Mono R290	R290	-25°C	75°C
Bosch Compress 7400i AW	R290	-25°C*	75°C
Mitsubishi Ecodan PUZ-WM	R32	-20°C	55°C

Daikin Altherma 3 H HT

Refrigerant R32

Rated to -28°C

Max flow 70°C

Vaillant aroTHERM plus

Refrigerant R290

Rated to -25°C

Max flow 75°C

Samsung EHS Mono R290

Refrigerant R290

Rated to -25°C

Max flow 75°C

Bosch Compress 7400i AW

Refrigerant R290

Rated to -25°C*

Max flow 75°C

Mitsubishi Ecodan PUZ-WM

Refrigerant R32

Rated to -20°C

Max flow 55°C

Set those numbers against the British weather. The coldest single hour a UK heating system is designed to cope with, the CIBSE external design temperature the MCS heat-loss calculation uses, is around -2 to -4°C across most of England and -4 to -5°C in Scotland, with a small altitude correction. The rated floors above sit roughly 15 to 25°C below that. There is no realistic UK day on which a correctly chosen unit refuses to run. Whatever the worst morning of your worst winter looked like, it was comfortably inside the operating range.

The high-temperature R290 units add a second kind of headroom. A unit that can push a 70 to 75°C flow temperature can hold a comfortable house even on existing radiators that were sized for a gas boiler, which removes one of the classic reasons an older install struggled on a cold day. More on radiators in the Five things people actually hit in winter section below.

Running is one question. Running efficiently is the one that decides your bill. The number that matters is COP, the ratio of useful heat out to electricity in. The chart below plots COP against outdoor temperature for a modern monobloc at a 45°C flow. It slopes down as the air cools, which is the grain of truth inside the myth, but it stays well above the line where a heat pump would be no better than a plain electric heater.

The chart carries the real point of this guide. The curve is the heat pump's intrinsic property, and it is the same in every house. What differs from one home to the next is where the seasonal average lands inside the shaded band, and that is a question of commissioning and design rather than of the weather. The Electrification of Heat trial put a median of 2.80 across a broad mix of UK homes. The volunteer-monitored fleet on HeatpumpMonitor.org, full of carefully commissioned installs, averages 3.87 on the same kind of hardware. The gap between those two numbers is more than a degree, and none of it is weather. It is whether the system was sized, balanced and tuned properly.

That is also why a cold snap matters less to the annual bill than people fear. The hours below -2°C are a small slice of a UK heating season, and even at -7°C the unit stays above a COP of 2.5. The running-costs guide works through what a seasonal efficiency figure does to a year-one bill, and the heat-pumps-explained guide covers the physics of the curve in more detail.

Most winter complaints are real, but they are not the heat pump failing. They are five specific, fixable things, and knowing them in advance is the difference between a good install and a bad one.

1. The defrost cycle looks like a fault. In cold, damp air, frost forms on the outdoor unit's fins. Every so often the heat pump briefly reverses to melt it, which can mean a few minutes of steam coming off the unit, a puddle underneath, and the indoor heating pausing. This is designed behaviour, it lasts only a few minutes, and the system manages it automatically. Knowing it is coming is the whole fix. A unit that defrosts constantly in mild weather, on the other hand, is a sign of a sensor or refrigerant problem worth a service call.

2. Radiators sized for a gas boiler run cool. A heat pump runs a lower flow temperature than the 65 to 70°C a gas boiler used. The same radiator gives off less heat at 50°C than it did at 70°C, so a room with an undersized radiator can feel cool on the coldest days even though the heat pump itself is fine. The fix is a larger radiator in the rooms the heat-loss calculation flags, not a bigger heat pump. The property-suitability guide has the output-versus-flow-temperature chart.

3. Old gas-boiler habits. A gas boiler is happy to sit off all day and then blast the house warm in an hour. A heat pump is most efficient running low and steady, holding a gentle temperature around the clock. Households that carry over a deep overnight or all-day setback often find the system fighting to claw the house back up on a cold morning, drawing more electricity than a steady schedule would. The fix is a habit change, not a repair.

4. The flow temperature is set too high. A flow temperature set higher than the house needs sounds like it should help, and it does the opposite. It is one of the most common causes of a disappointing winter. A high flow temperature pushes the heat pump onto the steep part of the efficiency curve and inflates the bill without making the house any warmer. A weather-compensation curve, which the installer sets, lets the flow temperature track the outdoor air. Tuned well, it runs cool in mild weather and only climbs in the coldest hours.

5. Outdoor unit siting and drainage. A unit sitting on the ground with no drainage gap can ice up underneath in a hard frost, where the defrost meltwater refreezes. A unit boxed into a tight, unventilated enclosure recirculates its own cold exhaust and works harder than it should. Both are siting decisions made at install time: raise the unit on feet or a wall bracket with a clear drainage path, and leave the airflow clearances the manufacturer specifies.

Honesty cuts both ways. There are situations where a heat pump really does underperform in winter, and pretending otherwise is how the myth got its grip in the first place. The common thread is that every one of them is a design or specification error, not a limit of the technology.

An undersized unit from a rushed heat-loss calculation. If the heat-loss survey was guessed or skipped, the unit can be too small to hold the house on the coldest design days. It runs flat out, leans on its back-up immersion heater, and the bill climbs while the rooms stay cool. The MCS room-by-room calculation exists precisely to prevent this, which is why it is worth checking it was actually done.

An uninsulated house that needed fabric work first. A heat pump dropped into a solid-walled, draughty, loft-bare house without any fabric improvement has to meet a very high heat loss at a high flow temperature, which pushes it down the efficiency curve and up the running cost. The unit is not failing. The house was not ready. This is the case HeatPass most often returns a Not Yet verdict on, with an Improvement Plan naming the fabric work to do first.

The wrong tariff. A heat pump left on a standard single-rate electricity tariff can cost far more to run than the same unit on a heat-pump tariff that prices off-peak electricity cheaply. This is not a winter fault at all, but it shows up most painfully in winter when consumption is highest, and it gets misremembered as the heat pump being expensive to run in the cold.

In each case the answer is to get the specification right before committing, not to write the technology off. The point of a verdict before you buy is to catch exactly these. If the honest answer for your home today is Not Yet, that is worth knowing before the unit is on the wall, not after.

Because winter performance is set by the install, a handful of questions sort a careful installer from a quick one. None of them needs technical background to ask.

Did you do a room-by-room heat-loss calculation? A real MCS install includes one, and the design flow temperature and unit size both fall out of it. A quote produced from the floor area alone, or from the size of the old boiler, is a warning sign. Ask to see the calculation.

What design flow temperature are you sizing for? Lower is better for efficiency. A good UK design aims for a flow temperature around 45 to 50°C at the coldest design hour, which means the radiators and the unit are matched to run efficiently rather than hot. A design that assumes 60°C-plus is often a shortcut around upgrading a few radiators.

Will it run weather compensation, and who tunes it? Weather compensation is what keeps the flow temperature low in mild weather. Confirm the control supports it, that it will be enabled, and that someone will revisit the curve after the first cold spell rather than leaving the factory default.

How will the outdoor unit be sited and drained? Raised off the ground, a clear drainage path for defrost meltwater, the manufacturer's airflow clearances, and a noise position that respects the neighbours. These are the small decisions that quietly decide whether the unit copes with a hard frost.

A good installer welcomes these questions, because they are the same ones a good design already answers. HeatPass routes a qualified homeowner to a single local MCS-certified installer, and the verdict you get first is built on the same heat-loss logic the install should follow. If you want to see where your own home lands before you talk to anyone, the HeatPass check takes about two minutes from your postcode.