Heat Pump Sizing: Room-by-Room Heat Loss Calculation and Flow Temperature Design

Quick Answer: A heat pump must be sized to the building's heat loss at the design external temperature (typically −2°C to −4°C in the UK), not to the existing boiler rating. MCS 3005 requires a room-by-room heat loss calculation per BS EN 12831, with flow temperatures selected to suit the emitter (radiator or underfloor) capacity. Typical UK domestic heat losses range from 4–12 kW. Oversizing causes short-cycling and poor SCOP; undersizing causes cold rooms.

Summary

Heat pump sizing is the single most important decision in any air source or ground source installation. Unlike a gas boiler — which can be oversized with little efficiency penalty because it modulates and only runs when called — a heat pump's seasonal efficiency (SCOP) depends on running steady-state at low output for long hours. An oversized heat pump short-cycles, wears out compressors prematurely, and delivers poor SCOPs (typically 2.5 or less instead of the 3.5–4.0 a correctly sized unit can achieve).

The MCS (Microgeneration Certification Scheme) installation standard MCS 3005 is the binding requirement for any heat pump installation that the homeowner wants to use for grant funding (BUS — Boiler Upgrade Scheme, £7,500) or for renewable energy certification. MCS 3005 mandates a room-by-room heat loss calculation using BS EN 12831 methodology, an emitter survey, and a flow temperature design that the existing or upgraded emitters can deliver.

Sizing a heat pump from the customer's gas bill, the existing boiler kW, or rules of thumb (e.g. "1 kW per 10m²") is non-compliant and unreliable. The customer's gas bill includes hot water, cooking, and standing losses; the boiler is almost always oversized; and rules of thumb do not account for insulation, glazing, or air permeability. See heat pumps for the overview and bus grant guide for grant funding.

Key Facts

Quick Reference Table

Quoting a heating job? squote turns a 2-minute voice recording into a professional quote.

Try squote free →
Building Type Typical Heat Loss (kW) Recommended HP Size Flow Temp Design
Pre-1920 solid wall, 100m² 9–14 9 kW + back-up 50–55°C
1930s cavity, uninsulated, 100m² 7–10 8 kW 50°C
1980s cavity insulated, 100m² 5–7 6 kW 45°C
2000s well insulated, 100m² 4–6 5 kW 40–45°C
Modern Part L2 new build, 100m² 3–5 4–5 kW 35–40°C
Passive house standard, 100m² 1–2 3 kW (smallest) 35°C
2-bed flat, 60m², top floor 3–4 4 kW 40–45°C
4-bed detached, 180m², 1990s 9–13 9 kW + DHW priority 45°C

Detailed Guidance

Step 1: Building survey

Before any calculation, measure the dwelling. The MCS heat loss calculation requires accurate floor areas, wall areas (subtracting windows), ceiling/roof areas, ground floor type, and glazing details. Where existing drawings are not available, measure on site and produce a sketch plan.

For each room, record:

Step 2: U-value determination

Where measured U-values are not available, use assumed values from age and construction:

Element Pre-1920 1920–1980 1980–2000 2000+
Solid brick external wall 2.1 — — —
Uninsulated cavity wall — 1.6 — —
Insulated cavity wall (60mm) — — 0.45 —
Modern cavity wall (100mm PIR) — — — 0.25
Single glazing 5.7 5.7 — —
Double glazing (1980s) — — 3.0 —
Modern A-rated glazing — — — 1.4
Uninsulated suspended timber floor 1.5 1.0 — —
Insulated floor (100mm PIR) — — 0.45 0.22
Roof — uninsulated 2.3 — — —
Loft — 270mm mineral wool — — — 0.16

These are indicative. PAS 2035 retrofit surveys require condition-based assumptions or in-situ measurement.

Step 3: Heat loss calculation per room

BS EN 12831 formula for each room:

Q_room = Σ(U × A × ΔT)_fabric + (n × V × 0.33 × ΔT)_ventilation

Where:
  Q_room = heat loss in watts
  U      = U-value W/m²K
  A      = area m²
  ΔT     = design temperature difference (Tint − Text)
  n      = air change rate per hour (typically 0.5–1.5)
  V      = room volume m³
  0.33   = specific heat capacity of air, Wh/m³K

Worked example — 4 × 4m living room, 2.4m high, two external walls (insulated cavity, 1980s), one window 2m², floor solid uninsulated:

External walls: 2 × (4 × 2.4) − 2 = 17.2 m² × 0.45 × 23 = 178 W
Window:         2 m² × 3.0 × 23 = 138 W
External floor: 16 m² × 1.0 × 18 (reduced ΔT) = 288 W
Roof (mid-floor, no loss): 0
Infiltration:   16 × 2.4 × 1.0 × 0.33 × 23 = 292 W
Total:          896 W living room

Sum all rooms for whole-dwelling heat loss. Add 5–10% for DHW priority and standing losses.

Step 4: Emitter capacity check

Once heat loss is known per room, check whether existing radiators can deliver that output at the design flow temperature. Radiator outputs in catalogues are quoted at Δ50 (ie. mean water temperature 50°C above room). At lower flow temps the output reduces by the ratio:

Output_new = Output_Δ50 × (ΔT_new / 50)^1.3

At 45°C flow / 35°C return / 21°C room:
  Mean water = 40°C, ΔT = 19
  Factor = (19/50)^1.3 = 0.282
  → Radiator delivers 28% of catalogue output

This is why heat pump conversions often need radiator upgrades. K2 (double convector) or K3 (triple) panels at the same dimensions deliver more output. Underfloor heating bypasses the problem entirely at 50–70 W/m².

Step 5: Flow temperature selection

The lowest flow temperature that the emitters can support at the design condition gives the best SCOP. Iterate:

Step 6: Domestic hot water sizing

DHW requires a cylinder sized for the heat pump's output (typically 200–300L for a 4-bed) and a heating coil rated for low-temperature operation (>3.0 m² coil area for a 9 kW heat pump). Schedule DHW reheat overnight to avoid clashing with peak space heating demand.

Step 7: Document the design

Final heat loss schedule includes:

Frequently Asked Questions

Can I size a heat pump from the customer's gas bill?

No. Gas bills include hot water, cooking, standing losses, and behavioural variation. The boiler's modulating range means actual heating energy delivered is a poor proxy for design heat loss. MCS 3005 requires a room-by-room calculation. Heat-loss-from-gas-bill as a sanity check is fine, but it cannot be the sizing basis.

Why is oversizing worse than oversizing a boiler?

Heat pumps achieve high SCOP by running steady-state at part load. Oversizing forces short-cycling: the compressor reaches setpoint quickly, shuts off, the system loses heat through the buffer/cylinder, restarts. Each start has compressor wear, refrigerant pressure spikes, and an efficiency penalty. A correctly sized heat pump runs almost continuously in cold weather at low modulation — exactly the opposite of how a boiler operates.

Do I need to replace all the radiators?

Not necessarily. Check each radiator's required output against its existing output derated to the design flow temperature. Often only 2–4 radiators in the coldest rooms need upgrading (typically bedrooms with old single panels). Use thermographic survey or radiator size calculator. Upgrade to K2 or K3 at the same length to preserve room layout.

What's the difference between MCS 3005 and BS EN 12831?

BS EN 12831 is the heat loss calculation method (the maths). MCS 3005 is the heat pump installation performance standard (the rules of installation) which references BS EN 12831 as the sizing methodology. For grant funding, MCS-certified installers must produce a calculation following MCS 3005, which uses BS EN 12831 internally.

How much does a sizing survey cost?

A standalone MCS heat loss survey typically costs £350–£600 for a 3-bed house, taking 3–4 hours on site plus 4–6 hours desk work. The cost is usually absorbed into a full quotation by MCS installers, but RICS-registered consultants and independent assessors offer survey-only services for homeowners wanting an unbiased size before tender.

Regulations & Standards