Heat Pump First Winter: What's Normal and What Isn't?
Heat pump first winter in the UK: what's normal, what isn't, how to judge your first bill fairly, a worked weekly log, and when to escalate.

A heat pump's first winter is when most households find out what their system actually does. It runs differently from a gas boiler, so plenty of normal behaviour looks like a fault, and a few real faults hide behind behaviour that looks normal.
This guide treats the heat pump first winter as an evidence-gathering exercise rather than a wait-and-see one. It covers what is normal, a symptom-by-symptom triage list, how to judge your first bills fairly against the installer's estimate, a worked example of a six-week log, and the formal route if your installer stops answering. The benchmark figures come from the Electrification of Heat (EoH) demonstration - a government-funded field trial run by Energy Systems Catapult that monitored 742 heat pumps in British homes of every age and type.
What is normal in a heat pump's first winter?
Five behaviours that look concerning but are the system working as designed.
- The outdoor unit runs most of the day. A heat pump delivers heat continuously at low output instead of in short boiler bursts, and in a cold snap it can run almost non-stop. Long run-times are efficient, not a warning sign. They only matter if the rooms still fail to reach target. Our cold weather operation guide explains why.
- Defrost cycles, with steam and dripping water. In damp weather around freezing, frost builds on the outdoor coil and the unit periodically reverses for a few minutes to melt it. A cloud of vapour and a puddle under the unit are expected; make sure the condensate has somewhere to drain so it does not freeze into a slip hazard.
- Lower efficiency on the coldest days. In the EoH trial the median air source system delivered 2.80 units of heat per unit of electricity across the year, and 2.44 on the coldest days, when daily mean temperatures fell as low as -6°C. A dip in a freeze is physics, not a fault.
- The backup heater firing briefly in hard frosts. Many systems carry an electric immersion or inline heater for very cold days and for weekly legionella cycles on the hot water cylinder. A few hours in a cold week is normal; running every day is not (see the immersion backup guide).
- Rooms 1-2°C under target in a cold snap. A heat pump sized to the design outdoor temperature can lag slightly on the very coldest days, and heavy masonry houses take time to respond. A sustained gap of 3°C or more is a different matter.
What is not normal, and what should you check first?
Symptom-by-symptom triage: the check you can make yourself, then when it becomes the installer's job.
- Rooms 3°C+ below target for several days. Check first: thermostatic radiator valves fully open in the cold rooms; no furniture or curtains over radiators; outdoor temperature in your log. Call the installer when: it happens in mild weather too, or the whole house falls short in a cold snap.
- A fault or error code on the controller. Check first: look the code up in the manufacturer's manual or app and note the date, time and outdoor temperature. Call the installer when: the code repeats, or the manual says it needs an engineer.
- Knocking, grinding or rapid clicking from the outdoor unit. Check first: leaves, ice or debris around the fan; whether it only happens during defrost. Call the installer when: the noise is mechanical and persists outside defrost - stop using it if it is severe.
- Bills far above the estimate. Check first: weather-adjust the comparison (next section) and strip out non-heating electricity. Call the installer when: the gap survives the adjustment, or the backup heater shows long daily hours.
- Hot water runs out in normal use. Check first: cylinder schedule and setpoint against what the installer configured. Call the installer when: settings match the handover pack and it still runs short - see cylinder coil sizing.
- System pressure falling. Check first: pressure gauge reading against the figure on your handover paperwork; any damp under radiators or pipe joints. Call the installer when: pressure keeps dropping after one top-up, or you cannot find the leak.
- The unit cycles on and off every few minutes. Check first: whether many radiator valves are shut, cutting the water volume. Call the installer when: it continues with valves open - see our short-cycling guide.
Faults that do appear in the early years tend to cluster around sensors, controllers and small refrigerant leaks; our common failure modes guide covers what each one costs and what the warranty covers.
How do you judge your first bill against the installer's estimate?
An unadjusted monthly bill is the most common source of false alarms.
The installer's running-cost estimate is an annual figure built on average weather. Comparing a January bill with one twelfth of it will always look alarming, because winter carries roughly half of a year's heating electricity - our seasonal usage guide shows the monthly profile. Three adjustments make the comparison fair.
- Strip out everything that is not heating. Your bill includes the standing charge, cooking, lighting, appliances and any EV charging. Take your pre-heat-pump summer usage as a baseline and subtract it, or use the heat pump's own energy reading if the controller or app reports one.
- Adjust for the weather. Heat demand rises with the gap between your indoor target and the outdoor temperature. With a 20°C target, a week averaging 0°C needs about 20 degrees of lift while a week averaging 7°C needs 13 - so the cold week needs roughly half as much heat again before efficiency falls at all. Note the average outdoor temperature for every week you log.
- Calculate your own efficiency. If your system reports heat delivered as well as electricity used, divide one by the other. That figure is your seasonal performance factor (SPF, the real-world measured version of the SCOP on the datasheet). Weekly figures between the mid-2s and around 3 in a British winter sit in line with the EoH results; persistent figures near 2 in mild weather are worth raising.
If, after all three, consumption is still well above the estimate, something is wrong: the weather compensation curve, the flow temperature, the backup heater, or the sizing. Our SCOP optimisation guide walks through each lever.
What does a first-winter log look like?
An illustrative six weeks for a three-bed semi with a 20°C target - invented numbers, real method.
The figures below are illustrative, not measurements from a real home. They show how a simple weekly log separates weather from faults.
- Week 1 (7°C average outside): 105 kWh used, 315 kWh of heat delivered, SPF 3.0. Rooms: on target.
- Week 2 (5°C average outside): 125 kWh used, 363 kWh of heat delivered, SPF 2.9. Rooms: on target.
- Week 3 (0°C average outside): 188 kWh used, 470 kWh of heat delivered, SPF 2.5. Rooms: lounge 1°C under.
- Week 4 (-2°C average outside): 222 kWh used, 532 kWh of heat delivered, SPF 2.4. Rooms: lounge 1°C under; backup heater 3 h.
- Week 5 (6°C average outside): 121 kWh used, 339 kWh of heat delivered, SPF 2.8. Rooms: back bedroom 2°C under.
- Week 6 (4°C average outside): 138 kWh used, 387 kWh of heat delivered, SPF 2.8. Rooms: back bedroom 2°C under.
Reading it. Week 4 used more than twice the electricity of week 1. Taken alone, that looks like a fault. Against the outdoor temperatures it is exactly what the weather predicts: the heat delivered tracks the temperature gap, and efficiency dipping to 2.4 in a -2°C week matches the field trial's coldest-day median. The lounge running 1°C under in the freeze is the normal lag described above.
The real finding is weeks 5 and 6. One room falling short in mild weather, while the rest of the house is fine, points to that room - a radiator too small for a low flow temperature, an unbalanced circuit or a valve partly shut - not to the heat pump. That is a specific, fixable question to put to the installer, backed by two weeks of data, and it is very different from ringing to say the system "feels expensive".
Month 1: why should you leave the settings alone?
Let the house and the new system settle before tuning.
Week 1-4 after commissioning: hands off the controller settings. Let the property's thermal mass + the heat pump's weather compensation find their natural operating points. Specifically:
- Don't adjust the weather compensation curve. Installer defaults are reasonable + your observations need a full month's data to identify whether they need tuning.
- Don't tweak temperature setpoints daily. Pick a target temperature (typically 19-21°C) + leave it; the heat pump optimises around a stable target far better than a moving one.
- Don't compare bills to gas-boiler-era bills. The first month may include partial-month effects + transition costs that aren't representative. Wait for the full month's electricity bill.
- Do note unusual events: any fault codes, persistent cold rooms, hot water shortages, very loud noise. Photograph or screenshot if possible.
Month 2: what should you start logging?
Build the data that month-3 decisions rest on.
Weeks 5-8: start the weekly log above without changing settings yet.
- Room temperatures: two or three main rooms, morning and evening, alongside the outdoor temperature.
- Weekly electricity: the heat pump's own reading if available, otherwise the whole-house meter minus your baseline.
- Manufacturer app data: efficiency, run-time, defrost and fault history - screenshot it weekly, because some apps only keep a short history.
- Comfort notes: rooms that run cool or warm at particular times of day.
By the end of month 2 you have eight weeks across a range of outdoor temperatures, which is enough to see whether a problem follows the weather or follows a room.
Month 3: should you tune or escalate?
Act on the pattern the log shows.
Weeks 9-12: act on what you have learned.
- Everything on target and efficiency in line: no action beyond booking the annual service (our owner maintenance checklist covers the rest).
- Whole house too cool in cold weather, fine when mild: raise the weather compensation curve slightly - see our curve setup guide and the weather compensation explainer. Small changes (±0.1 on the slope, ±1°C on the offset) make a noticeable difference.
- Whole house too warm, or efficiency low in mild weather: lower the curve in the same small steps.
- One room short in all weathers: a radiator or balancing question for the installer, as in the worked example.
- Short-cycling: apply the fixes in our short-cycling guide.
- Any 'not normal' symptom that persists: call the installer with the log. Dated readings get a faster, better diagnosis than a general complaint.
What if your installer does not respond?
The formal route for an MCS-certified installation.
Most installations funded by the Boiler Upgrade Scheme are carried out by an installer certified under MCS (the Microgeneration Certification Scheme, the UK standards body for small-scale heat pump installations). MCS sets out a three-step route when a problem is not fixed:
- Complain to the installer in writing. Explain the problem and the outcome you want. MCS treats 14 days as a reasonable maximum for a response, and by then you should at least have an agreed plan.
- Escalate to MCS. Send what went wrong, photos, the contract and your log. MCS acknowledges within 5 working days, updates you at least every 10 working days and aims to resolve most complaints within 8 weeks.
- Take it to dispute resolution. If you disagree with the outcome or it is unresolved after 12 weeks, eligible disputes go to the Furniture & Home Improvement Ombudsman, free of charge, within 12 months of MCS concluding. For contracts signed before 1 January 2025, the route can instead run through the installer's consumer code (RECC or HIES).
The first-winter log is what makes each of these steps work: it turns "it doesn't seem right" into dated evidence. Manufacturer warranty claims follow a separate path, covered in our warranty claims guide.
For the questions new owners ask most in the first months, see our heat pump FAQs for new owners.
So when should you actually worry?
The short version.
Worry about patterns, not moments. Long run-times, defrost steam, a dip in efficiency during a freeze and a bigger January bill are the system behaving as designed. A room that stays cold when it is mild outside, a code that keeps returning, mechanical noise, falling pressure, or consumption that stays high after weather adjustment are the patterns that justify a call - and a few weeks of logged readings is what gets them fixed first time.
Frequently asked questions
Is it normal for my heat pump to run all day?
Why is my first winter heat pump bill higher than expected?
What efficiency should I expect in my first winter?
When should I call the installer in my first winter?
How long does a heat pump take to settle in?
What can I do if my installer will not fix a problem?

Heat pump FAQs from new owners

Heat pump cold weather operation

Common heat pump failures in the first five years
