Building Heat Loss Calculator

Step 1 - Temperature Difference (ΔT)

Everything is driven by the difference between the room’s target temperature and the outside design temperature.

Target: 21°C  −  Outside: −3°C  =  ΔT of 24°C

This represents the worst-case conditions the system must be able to cope with.

Step 2 - Fabric Loss

For each surface (walls, windows, floor, ceiling, doors) we apply:

Q_fabric = U × A × ΔT

U is the U-value (W/m²K - how easily heat passes through the material), A is the surface area in m². A lower U-value means a better-insulated surface.

Example: a 10 m² cavity-insulated wall (U = 0.50) with a ΔT of 24 loses 0.50 × 10 × 24 = 120 W. The same wall uninsulated (U = 1.00) would lose 240 W.

Windows and doors are taken out of the walls they sit in - their area is deducted from the walls facing outside first, then from any facing an unheated space, before the wall’s U-value is applied. Each opening is therefore counted once, at its own U-value, rather than twice. Every opening is costed at the full outdoor ΔT, whichever wall it sits in, and if the openings you enter come to more area than those walls can hold the room says so.

Ground floors lose heat to the ground, not the outside air, so they use a smaller ΔT based on the ground temperature (around 10°C) rather than the outdoor design temperature. Upper floors sit on the heated room below, so their floor isn’t counted (unless you set an exposed-floor type on that room).

Ceilings work the same way in reverse. Only rooms on the top storey sit under the roof, so only they are given the roof’s U-value; a room with a heated storey above loses next to nothing upwards. Where part of a lower storey sits under the roof, set the ceiling type on that room and it is counted.

Walls are counted by what they face. A wall to a heated room loses almost nothing. A wall to an unheated space (garage, store, unheated stairwell) loses at about half the outdoor ΔT, since that space sits roughly midway between inside and outside.

Step 3 - Ventilation Loss

Warm air leaking out of the room carries heat with it:

Q_vent = 0.33 × Volume × ACH × ΔT

0.33 is the volumetric heat capacity of air (W/m³/K), ACH is the number of air changes per hour. The rate is the property default unless you set one on the room. Wet rooms such as bathrooms and kitchens need more (2-3 ACH) and circulation spaces around 2 ACH, and a room whose name suggests one of those is flagged if it is left on a lower rate - but nothing is raised for you. The rate used is always the one showing on the room.

Example: a 48 m³ room at 1.0 ACH and ΔT of 24 loses 0.33 × 48 × 1.0 × 24 = 380 W through ventilation alone.

Open fires and flues pull warm air out of the room continuously. Each open flue you record adds a fixed allowance of roughly 40 m³/h of cold-air infiltration on top of the room’s air changes.

Step 4 - Total Room Heat Loss

Q_total = Q_fabric + Q_vent

Each room’s total is summed to give the whole-property heat loss.

Step 5 - System Sizing

A heat source has two duties, and we work out each one:

• Space heating - the total heat loss, plus a 20% uplift for boilers (pipework losses and warming a cold building). Heat pumps take no uplift; they are sized to run continuously at design conditions.
• Hot water - a combi has no store and heats mains water as it flows, so its figure follows the Hot Water Flow Rate you record: flow × 4.18 × a 35 K rise ÷ 60. That 35 K rise is the condition a combi's output is declared at, so the answer maps onto how combis are sold. Anything with a cylinder has to reheat that cylinder instead, so its figure follows the Hot Water Volume.

DHW kW = Volume × 4.18 × (store − 10°C mains) ÷ (3,600 × reheat hours)

A boiler reheats a cylinder to 60°C in an hour; a heat pump reheats a cooler store to 50°C over three, which is why the same cylinder asks far less of it. A cylinder system left with no volume recorded carries no hot water allowance at all.

Example: 8,000 W total heat loss × 1.20 = 9.6 kW of space heating. On a combi at 12 L/min, hot water asks 29.3 kW - the larger duty - so the suggestion is a 30 kW combi. On a system boiler with a 210 litre cylinder, hot water asks 12.2 kW, still the larger, so 13 kW. On a heat pump with the same cylinder, hot water asks only 3.3 kW, so space heating governs and the suggestion is 8 kW.

The heat pump figures assume the unit does what heat pumps do: switch to a separate hot water setpoint for the length of the charge, reaching a 50°C store, then drop back to space heating afterwards. That setpoint belongs to the unit and is not the Flow Temp you set above - running underfloor at 35°C does not mean the cylinder is charged at 35°C. Reaching a 50°C store needs roughly 55°C on the primaries, which most current units manage; an older one capped lower would reach a few degrees less.

A 50°C store sits below the 60°C pasteurisation temperature. That is normally a weekly immersion boost - an electric element, and no load on the heat pump - so it is not in this figure.

The suggested size is the larger of the two duties, not their sum. They never run at once: every domestic wet system gives hot water priority, so space heating pauses for the length of a cylinder charge or a running tap. Adding them would size for a condition that does not occur - on a small property the hot water figure alone would roughly double the answer. The tool tells you which duty governs and what the other came to.

That is also why hot water usually governs a combi: a combi is named by its hot water output, which is why a small flat still gets a 24-30 kW boiler while its heating demand is a fraction of that. It is a starting point for a conversation with a heating engineer, not a specification.

Step 6 - Emitter Sizing (ΔT50)

A room’s heat loss is the heat its emitter has to deliver. It is not the number printed on a radiator, because radiators are sold by their output under a fixed test condition - ΔT50, meaning the mean water temperature sits 50°C above the room (75°C flow / 65°C return into a 20°C room).

Output = Rating × (ΔT / 50)ⁿ    ΔT = mean water temp − room temp

Run an emitter cooler than that test condition and it delivers less, so the rating you need to order rises. Run it hotter - a boiler at 75/65 into an 18 or 19°C room, where the water averages more than 50°C above it - and it delivers a little more, so the rating needed comes out slightly below the room’s heat loss. Both are ordinary; the tool states which way round it is for your figures. The exponent depends on the emitter: n = 1.3 for radiators (and for a mixed system, which is sized on its radiators), 1.1 for underfloor, whose large low-temperature surface falls away more gently, and 1.0 for fan coils, which are close to linear.

Example: a room losing 1,600 W at 21°C needs a 1,650 W ΔT50 radiator on a 75/65 boiler, but roughly 4,800 W ΔT50 at a heat pump’s 45/40 - about three times the rating for the same duty.

Radiators are often sold in BTU/h rather than watts, so the results table quotes the rating that way too. It is the ΔT50 rating converted at 3.412 BTU/h per watt - not the room’s heat loss converted - so on a heat pump it sits far above what the room actually loses, for exactly the reason above.

Underfloor and fan coils are not sold on the ΔT50 basis radiators are. The rating shown for such a room converts its demand on the same curve so every room can be read down one column; it is not a figure to order underfloor by. A room on underfloor is judged against its own output in W/m² in Step 7.

If the water would average no warmer than the room - a flow temperature set too low for the target temperature - then no emitter of any size can heat it. Rather than quote a rating, the tool reports the room as Cannot be sized and flags it.

Flow and return temperatures default to values typical of the heat source and emitter you pick; you can override them, and once you do the tool keeps your figures.

Step 7 - Underfloor Heating

Underfloor output varies far too much with screed, pipe spacing and floor covering for us to derive it, so you enter the output figure in W/m² from the underfloor system’s own design data. We multiply it by the room’s floor area to get what the floor can cover, and anything left over is reported as a shortfall with the ΔT50 rating a supplementary emitter would need.

Your figure is used exactly as entered. We compare it against the floor surface temperature limits in BS EN 1264 - 29°C in occupied rooms, 33°C in wet rooms - and flag it if it looks higher than a floor at that limit could deliver. That is a prompt to check the figure, not a correction of it.

Step 8 - Radiator Schedule

The schedule records the radiators you intend to fit against the rating each room needs. You enter the catalogue output at ΔT50 and the quantity, one line per radiator size, and the line totals are added up and compared with the rating from Step 6. A room with underfloor only has to make up what the floor cannot cover, so that shortfall is what it is judged against - whether the room is underfloor throughout or a mix of the two.

The smaller figure beneath each line total is what those radiators put out at your design flow and return, using the same curve as Step 6. On a heat pump it sits far below the catalogue figure - which is exactly why the rating needed is so much larger than the room’s heat loss.

Outputs are used exactly as entered. We hold no product catalogue, so we cannot confirm that a product code and an output belong together, or that the radiator will fit the wall intended for it.

Defaults and how they cascade

A room uses its own setting where it has one, otherwise its storey’s, otherwise the property’s. Storeys carry their own construction because buildings do - underfloor downstairs with radiators up, a solid ground floor under a timber first floor, a roof only over the top.

To save entering the same thing on every room, the first room surveyed on a storey sets that storey’s emitter type, floor type, ceiling type, room height and underfloor output; the room then goes back to inheriting, and later rooms on that storey start from the same place. Any room can still override any of them, and a room that has been overridden is outlined and offers the storey’s value back.

Target temperature never cascades at all. A hall, a bedroom and a bathroom on one landing are heated to three different figures, so there is no storey-level temperature: each room takes its own from its Room Type, and you can change it on the room.

Heat Loss per m²

Dividing a room’s heat loss by its floor area gives a figure that can be compared across rooms and properties. The bands we describe it with follow the guidance quoted in MCS 031: under 30 W/m² is very low heat loss and 120-150 W/m² is very high. We use it as a sanity check and warn on rooms that fall far outside the range real properties occupy.

Where these figures come from

Every constant the tool applies, and its source:

Referencing a standard describes where a method or value originates. It does not mean this tool is certified against it, or that its output constitutes a design to that standard.

Method & Standards

This calculator uses the simplified steady-state method for domestic heat loss, calculating peak heat demand at the worst-case outdoor design temperature. It is not an MCS-certified design tool, and its output is not an MCS or BS EN 12831 design.

To keep data entry quick it makes simplifications a full survey would not. It does not model:

• Heat moving between two heated rooms at different temperatures - a warm bathroom beside a cooler landing nets to zero here.
• Thermal bridging, thermal mass, solar gain, or heat given off by people and appliances.
• Measured airtightness - ventilation comes from the air change rate you select, not a pressure test.
• Exposure, orientation, shelter or altitude corrections to the outdoor design temperature.
• Distribution losses, pipe sizing, flow rates, hydraulic balancing or system volume.
• Whether the radiators on the schedule are real. The outputs are the ones you type in; we hold no product catalogue and do not check a code against its published data.
• Heat pump capacity derating at low outdoor temperatures, defrost cycles, or hot water reheat scheduling.
• Room shape. Every room is taken as a rectangle of four walls, with its ceiling the same area as its floor - an L-shaped room has to be entered as more than one room, and a sloping or vaulted ceiling is understated.
• Below-ground walls. Only a ground or basement floor uses the ground temperature; a basement’s walls are costed at the full outdoor ΔT, which is cautious rather than accurate.
• Downward heat loss from a heated floor into what is beneath it.

Nothing is taken from what a room is called. Its Room Type sets its design temperature and air change rate, and decides whether the EN 1264 wet-room floor limit applies (Step 7) - so renaming a room never moves a figure.

A room without a complete length, width and height cannot be costed, so it is left out of the property total, the suggested heat source and the report. Such rooms are marked “Not counted” on the survey and named in both the report and the spreadsheet, but it is worth checking every room is measured before relying on a total.

Results are only ever as good as what is entered. U-values, dimensions, air change rates, underfloor outputs and radiator outputs are all your inputs, and a wrong U-value or a mis-set wall facing will produce a confident-looking figure that is badly wrong. The warnings shown against rooms are prompts to check, not a validation that the survey is correct.

Treat the results as indicative sizing guidance, not a design certificate.