What size heat pump do you need? A guide to sizing for UK homes

Last reviewed: 14 May 2026

In short

Domestic air source heat pumps installed in UK homes come in five common output bands — 5, 7, 11, 14, and 16 kW — with model-specific variants filling in between. The size your home needs is decided by a room-by-room heat-loss survey carried out to BS EN 12831-1:2017 under the MCS heat pump design standard (MIS 3005-D V3.0, mandatory since 5 December 2025). The survey produces a calculated heat load in kilowatts at the local design temperature — for Reading, that’s −3°C. The installer then specifies the smallest unit whose output at design conditions meets the calculated load. Most UK homes end up in the 5–12 kW range; only large or poorly-insulated properties need 14 kW+.

Contents

• What “size” means in heat pump terms
• The five typical UK output bands
• How the survey decides the size
• Why oversizing has been the historic problem
• Why undersizing produces different — but visible — problems
• Cascaded systems — when one unit isn’t enough
• Hot water heating — a separate calculation
• Three things to check on your quote
• What this means for homes in Reading

What “size” means in heat pump terms

A heat pump’s “size” is its rated heat output in kilowatts at a defined air and flow temperature. The figure you’ll see on the model name — for example, a Mitsubishi Ecodan PUZ-WM85VAA is 8.5 kW — is the unit’s output at a standard test point called A7/W35 (7°C outdoor air, 35°C flow temperature). This is the headline number, but it’s not the number that decides whether the unit fits your home.

The number that matters is the output at design conditions — what the unit actually delivers on the coldest day your installer designs for. Heat pump capacity declines as outdoor temperature falls; a 7 kW unit (A7/W35) might only deliver 5.5–6 kW on a −3°C day. Your installer’s sizing decision uses the design-day output, not the headline figure.

There are three test points to know about:

• A7/W35 — the headline rating, used in most marketing material
• A2/W35 — closer to a UK design day; output is typically 80–90% of A7/W35
• A-7/W35 — a colder-climate reference point; informational for most UK locations

For Reading-area properties, the design temperature is −3.0°C (we’ll come back to where that figure comes from). The A-3°C to A-2°C zone is what your sizing has to work in.

The five typical UK output bands

UK domestic heat pumps cluster in five common output bands. Manufacturers fill in the gaps with intermediate models (4, 6, 8, 9, 10, 12, 17 kW), but the five “round” bands cover most installs. Here’s how they map to typical UK properties:

4–5 kW

Typical fit: modern 2–3 bed flats, well-insulated terraces, end-of-terrace bungalows, small detached new-builds.

Examples: Daikin Altherma 3 M 04, Mitsubishi Ecodan PUHZ-W50, Vaillant aroTHERM plus VWL 35/5, NIBE F2040-5.

The 4–5 kW band has expanded since 2023 as modern monobloc units have improved part-load efficiency at lower capacities. Below 4 kW remains rare in UK domestic stock.

6–7 kW

Typical fit: typical 3-bed semi with cavity wall and modern insulation, 2-bed terrace with some fabric upgrades, small 3-bed detached.

Examples: Daikin Altherma 3 M 06, Mitsubishi Ecodan PUZ-WM60, Vaillant aroTHERM plus VWL 75/5, Samsung EHS 6 kW.

The 7 kW unit is the single most common heat pump size installed in UK homes. It fits the modal property — 3-bed semi, cavity-walled, modern insulation — which calculates around 5–6 kW design-day load. Across the BUS-administered install base, the median capacity is around 8 kW and the modal install is 7 kW.

8–9 kW

Typical fit: 3–4 bed semi or detached with mixed-period fabric, 3-bed Edwardian terrace with retrofit insulation.

Examples: Daikin Altherma 3 M 08/09, Mitsubishi Ecodan PUZ-WM85, Vaillant aroTHERM plus VWL 105/5, Worcester Bosch Compress 7400i AW 9.

10–12 kW

Typical fit: 4–5 bed detached, larger Victorian/Edwardian terrace, 4-bed semi with solid walls, 1930s detached without fabric upgrades.

Examples: Daikin Altherma 3 M 11/14, Mitsubishi Ecodan PUZ-WM112, Vaillant aroTHERM plus VWL 125/5, Worcester Bosch CS 7000iAW 13.

The 11–12 kW band is the upper limit for most BUS-eligible installs on a standard single-phase 100A supply. Properties needing 14 kW+ may require a three-phase supply (or a DNO upgrade), or a cascaded two-unit install instead.

14–16 kW

Typical fit: large 5+ bed detached, period properties with no fabric upgrades, large rural homes off the gas grid.

Examples: Daikin Altherma 3 H HT 14/16, Mitsubishi Ecodan PUZ-HWM140, Vaillant aroTHERM plus VWL 155/5, NIBE F2120-16.

Single-unit installs above 14 kW are uncommon in standard UK domestic retrofit — they push against single-phase supply limits and most properties at this load are better served by a cascaded pair of smaller units.

How the survey decides the size

The size of your heat pump is not chosen by eye, by floor area, or by “what we usually fit in a 3-bed”. It is chosen by a room-by-room heat-loss survey carried out to BS EN 12831-1:2017 under the MCS heat pump design standard. The current version is MIS 3005-D V3.0, which became mandatory for MCS-certified contractors on 5 December 2025.

The survey produces three outputs that drive the sizing decision:

1. Total dwelling heat loss in kilowatts at the design external temperature
2. Room-by-room heat loss in watts — used for radiator sizing
3. A recommended design flow temperature — typically 45–55°C for retrofit, 35–45°C for new-build or substantially refurbished properties

The design external temperature for a Reading-area property is −3.0°C (99.6% percentile, Heathrow weather reference) from CIBSE Guide A Table 2.5. This is the temperature your heat pump has to work in for the coldest 0.4% of hours per year — a conservative reference point.

Your installer then reads the manufacturer’s performance table for the proposed unit at A-3°C and the specified flow temperature, and checks that the unit’s output at that point is at least equal to your calculated load. That’s the sizing match.

There’s an important nuance under V3.0: the “next size up” convention has weakened. Older guidance encouraged a 20% safety margin — specify a unit 1.2× the calculated load to be safe. V3.0 doesn’t require this. The modern view is that BS EN 12831-1:2017 already builds in a diversity factor at calculation, and additional oversizing risks short-cycling losses (more on this below). Where a calculated load of 6.8 kW would historically have been bumped to 8 kW, V3.0 sizing says specify the 7 kW unit and let modulation handle the small under-shoot.

If your installer can show you the survey output and the manufacturer’s performance table side by side, you can verify the sizing yourself. If they can’t, the sizing is not on documented evidence.

Read our deep-dive on heat loss surveys for what the survey checks and how the calculation works.

Why oversizing has been the historic problem

UK heat pump installs through the early BUS-grant period (2022–2023) showed a systematic oversizing bias. A 2023 Energy Saving Trust review estimated that 30–40% of installations were oversized by at least one model band. There were two drivers:

Sizing software defaults. Pre-2017 calculation tools (including older versions of the MCS Heat Load Calculator) summed infiltration as if every wall faced the wind simultaneously. This over-estimated total heat loss by 15–25% on typical UK properties. The BS EN 12831-1:2017 update introduced the diversity factor — recognising that wind doesn’t blow on all sides at once — and brought calculated losses down to where they belonged.

Installer “safety margin” practice. Oversizing was reputationally safer than undersizing. A heat pump that fires its immersion backup on a cold day generates complaints; a heat pump that short-cycles in spring doesn’t, because the homeowner doesn’t notice. So installers had a one-way incentive to specify the next size up. The result was thousands of UK homes running heat pumps designed for properties one band larger than theirs.

The consequences of an oversized unit:

• Lower SCOP. An oversized heat pump spends more time at very low part-load, where inverter efficiency degrades. Real-world SCOP can be 0.3–0.5 below the design figure, which compounds over the 15-year life of the unit.
• Short-cycling wear. Frequent start-stop cycles wear the compressor faster and reduce equipment life.
• Higher upfront cost. A 12 kW unit costs around £1,500–£2,500 more than an 8 kW unit before the BUS grant.

The corrections in MIS 3005-D V3.0 + BS EN 12831-1:2017 are explicitly aimed at this problem. (Our sizing-matters explainer covers the consequences in more depth.)

Why undersizing produces different — but visible — problems

Where oversizing reduces SCOP quietly, undersizing produces problems you’ll notice. An undersized heat pump:

• Runs its electric immersion backup on cold days. Below the unit’s “balance point” — the outdoor temperature at which heat pump output equals heat demand — the immersion fires to make up the gap. The immersion runs at COP = 1 (straight electric resistance), at typically 1–3 kW for several hours on a cold UK day. The bill impact is immediate and obvious.
• Fails to reach setpoint. Homeowners report rooms running 1–3°C below thermostat target on the coldest days.
• Generates complaint volume. Cold rooms and high bills together are the most common heat-pump-failure profile, and the one that historically drove installer over-cautious sizing.

The V3.0 calculation framework is designed to specify the unit accurately rather than over-cautiously. The trade-off is that small calculation errors no longer have the “next size up” safety net — which makes survey rigour the dominant factor, not installer caution.

Cascaded systems — when one unit isn’t enough

For properties with a calculated load above ~12–14 kW, or where single-phase electrical capacity is a constraint, an MCS-certified installer may specify two heat pumps in cascade — two smaller units running in parallel, staging on as load demands. MIS 3005-D V3.0 explicitly permits cascaded systems, and they are recorded separately on the MCS install certificate.

The advantages:

• Single-phase compatibility. Two 8 kW units typically run within a 100A single-phase supply where one 16 kW unit may not.
• Better part-load efficiency. At low load (spring/autumn), only one unit runs, often at higher SCOP than a single 16 kW unit modulating down hard.
• Redundancy. If one unit fails, the other can carry partial load until repair.

The cost: two units typically run £4,000–£8,000 more than a single equivalent unit in equipment + labour. Cascaded installs are normally specified only where a single-unit install would either fail the electrical supply check or compromise part-load efficiency to an unworkable degree.

Hot water heating — a separate calculation

The headline sizing decision is for space heating. Hot water cylinder reheat is a separate demand the same heat pump also serves. For a typical 3-bed household:

• Cylinder size: 180–250 L unvented cylinder, sized at roughly 50 L per occupant plus a 50 L margin
• Reheat cycle: heat pumps run at a higher flow temperature (~50–55°C) for hot water, which reduces COP for that cycle. A full reheat from cold typically takes 1.5–3 hours and uses 4–8 kWh of electricity.
• Sizing implication: where the heat pump is sized close to space-heating design load, the hot water cycle runs in series with space heating and may cause brief shoulder-season output strain. Where the unit has modest space-heating headroom, hot water runs are absorbed without issue.

For most installations, the sizing question is dominated by space heating; hot water reheat is a secondary check. Our hot water cylinder guide covers cylinder sizing in detail (link will be added when A6.3 is published).

Three things to check on your quote

The single most useful sizing question you can ask before signing a quote: “What is my design-day heat load in kilowatts, and what is the specified unit’s output at −3°C at the design flow temperature?” If the installer can’t answer both halves with a paper trail from the survey calculation, the sizing isn’t on documented evidence.

Three specific checks on the quote:

1. The design external temperature is named. A compliant quote should reference the CIBSE Guide A weather station (Heathrow, for the Reading area) and the design temperature used (−3.0°C, 99.6% percentile). If only “design conditions” appears with no number, the calculation may not have used the right temperature.

2. The unit’s output at design conditions is stated. The figure should appear in the design notes — for example, “Specified unit: Daikin Altherma 3 M 08; rated output at A-2°C / W45 = 6.8 kW; calculated design load = 6.6 kW.” If only the nominal A7/W35 figure appears, you can’t tell whether the unit is sized to design conditions or just to the marketing rating.

3. Any oversizing is justified in the design notes. Where the specified unit exceeds the calculated load by more than ~15%, MIS 3005-D V3.0 expects the design notes to explain why — typically because the next size down doesn’t quite meet design load, or because cascaded part-load operation has been considered. Unexplained 20%+ oversizing is a flag.

These are checks any homeowner can run; they don’t require technical training, only the survey report itself.

What this means for homes in Reading

Reading sits in the Thames Valley CIBSE weather region, with the design external temperature −3.0°C (99.6% percentile, Heathrow reference) applying uniformly across the area. What varies is the calculated load, driven by housing-stock fabric. Typical sizing outcomes by neighbourhood character:

• Central Reading and lower Caversham — Victorian and Edwardian terrace stock with solid 9” brick walls. Calculated heat loss commonly 7–12 kW before fabric upgrades, with sized heat pumps typically 11–14 kW. Where internal wall insulation has been added, the load drops materially — a fully internally-insulated Victorian terrace can return 6–8 kW.

• Tilehurst, Earley, Whitley, eastern Reading — inter-war and post-war semi stock with cavity walls. Where the cavity has been filled (most properties since the 1990s), calculated load typically falls in 5–8 kW, with sized heat pumps 6–9 kW.

• Lower Earley, Woodley, western and southern modern estates — 1980s+ insulated cavity construction. Calculated load typically 4–7 kW, with sized heat pumps 5–8 kW.

• Caversham Heights, Caversham Park, central conservation areas — period properties where fabric upgrades face planning constraints. Sizing typically lands at the upper end of the property’s pre-upgrade calculated load, with the design flow temperature often raised to 50–55°C to keep emitter sizing manageable — at the cost of a lower SCOP for the life of the unit.

The Reading housing distribution means most installations will fall in the 6–11 kW range, with the modal install around 7 kW. The 14–16 kW band tends to apply to larger period properties without retrofit insulation; the 4–5 kW band tends to apply to modern flats and the newer end of Lower Earley / Woodley stock.

Patterns are directional, not deterministic. Two terraced properties on the same Reading street can calculate differently because one has internal wall insulation and the other doesn’t. The room-by-room survey is what distinguishes them — not the postcode and not the property type.

Related guides

• Heat loss surveys for heat pumps — the room-by-room survey that produces the kW load figure used to size your heat pump
• Why heat pump sizing matters — the consequences of oversizing and undersizing in depth (link will be added when A4.2 publishes)
• Heat pump flow temperature — the design decision that compounds with sizing: flow temperature affects unit output and SCOP for the life of the install
• Do you need bigger radiators for a heat pump? — the emitter-sizing consequence of the design flow temperature chosen at sizing

---

Get a quote for an MCS-compliant heat pump install in Reading — survey, design, install, BUS grant administration handled by our team. | Call us on 0118 [number] | Email [address]

---