Heat loss surveys: the complete guide for heating engineers

What this post covers: Everything you need to know about heat loss calculation as a heating engineer. The physics behind fabric and ventilation loss, the formulas, what happens when sizing goes wrong, how emitter sizing connects to flow temperature, and where modern software changes the workflow.

The short answer: A heat loss survey calculates exactly how much heat energy a building loses on the coldest expected day of the year. For heat pumps, this isn't optional and has to be precise. It determines system size, emitter sizing, flow temperature, and ultimately whether the homeowner's bills are reasonable. Software speeds up the process and pre-fills the paperwork, but the accuracy of the survey inputs remains the engineer's responsibility.

Sizing a heat pump isn't quite the same job as sizing a gas boiler. With a gas boiler, oversizing has historically been the safe move - the boiler modulates, the homeowner doesn't notice, and you avoid callbacks. However, the same logic doesn't apply to heat pumps. Oversize a heat pump, you get excessive cycling, reduced compressor lifespan, unnecessary capital cost, and sometimes a DNO upgrade that could kill the project entirely.

Accuracy is essential for heat pump sizing. In this guide we cover what's needed to do it right.

What is a heat loss survey?

It's probably worth starting with exactly what we mean by a heat loss survey.

A heat loss survey calculates how much heat energy a building loses on its “design day” - the coldest day statistically expected for that location. The heat loss itself isn’t a live measurement. It's a model of how the building will perform under worst-case conditions, which means the same methodology works for existing properties and new builds.

The design day matters because it sets the worst-case the system needs to handle. It's defined using weather data for the specific postcode - typically the 99th or 99.6th percentile cold weather temperature (i.e. the coldest it could get to on the coldest 1% or 0.4% of days).

From there, the goal of a heat loss survey is to correctly size the whole system - the heat pump itself, the emitters (radiators, underfloor heating), and the hot water cylinder. Each room's heat loss drives the emitter sizing for that room; the building's total heat loss drives the heat pump and cylinder sizing.

A complete survey requires:

• Measuring the dimensions of every room in the property (or tracing plans for new builds)

• Recording all window and external door dimensions

• Identifying construction materials for walls, floors and roofs to assign correct U-values

• Documenting ventilation characteristics, including draughts, chimneys, and extract systems

• Recording existing emitters (radiators or underfloor heating) and their output ratings

The result is the heat output the system needs to deliver to maintain a comfortable internal temperature when outside conditions are at their coldest.

In summary: A heat loss survey calculates the peak heating demand of a building by modelling fabric and ventilation losses under design-day conditions - typically the 99th percentile cold weather temperature (the coldest 1% of days) for that postcode. The calculation determines the minimum heat output required from the heat source and informs emitter sizing, flow temperature selection, and pipework design.

What are the two components of heat loss calculation?

Every heat loss survey involves two distinct calculations: fabric heat loss (how heat conducts through the building's solid elements) and ventilation heat loss (how warm air escapes and is replaced by cold outside air). Getting both right matters - and the most common errors in each are different.

We’ll explain both below, and briefly outline the formulas, but we’ll try to break it out in plain English too.

Fabric heat loss (conduction)

Fabric heat loss accounts for heat transferring through the solid materials of the building envelope: walls, windows, floors and roofs.

The formula:

Qfabric = U × A × ΔT

Where:

• U (U-value): How easily heat passes through the material, measured in W/m²K. Single-pane windows are typically around 5; well-insulated walls are around 0.2-0.3.

• A (Area): Surface area of the element in m².

• ΔT (Delta T): Temperature difference between inside design temperature and outside design temperature.

The most common error: getting the construction materials wrong. The U-value gap between a solid brick wall (typically 1.5-3.0 W/m²K depending on thickness) and a properly insulated cavity wall (around 0.3-0.5 W/m²K) is large enough that misidentifying one for the other can change the calculated fabric loss for that element by 4x or more.

Ventilation heat loss (infiltration)

Ventilation heat loss accounts for warm air physically leaving the property through draughts, chimneys and gaps in the building fabric, being replaced by cold external air that the heating system must then heat.

The formula:

Qventilation = V × ACH × 0.33 × ΔT

Where:

• V (Volume): Volume of the room in m³.

• ACH (Air Changes per Hour): How many times the room's air volume is fully replaced per hour. A modern airtight property might be around 0.5; an older draughty property could be 1.5 or higher.

• 0.33: A constant representing the volumetric heat capacity of air (Wh/m³K).

• ΔT: Temperature difference between inside and outside.

The most common error: misjudging the ACH. Getting the leakiness of a building wrong can double the ventilation heat loss calculation, sometimes more. A Victorian terrace with an open chimney and unsealed floorboards behaves completely differently from an airtight new build. The team at Open Energy Monitor has done useful work on how far CIBSE's standard ACH values diverge from blower door measurements; their analysis is worth reading.

The 2026 update to the CIBSE Domestic Heating Design Guide moved the recommended approach away from per-room ACH guesses toward whole-home air permeability. This new method calculates ventilation heat loss from the building's overall air permeability, then allocates it back to each room based on exposed surface area.

It's more accurate than picking an ACH value for each room from a table, and it handles awkward properties (uneven airtightness, open chimneys in one room and sealed in another) properly. Spruce calculates this automatically as you create the floorplan, so you're not stuck guessing ACH values when a blower door test isn't available. Our co-founder Steph has walked through the methodology in detail here, if you want the full picture.

In summary: Heat loss calculations combine two components: fabric loss (Qfabric = U × A × ΔT) and ventilation loss (Qventilation = V × ACH × 0.33 × ΔT). The most impactful errors are inputting wrong materials in fabric calculations and underestimating ACH in ventilation calculations.

Why does precision without accuracy make a heat loss calculation dangerous?

A heat loss calculation can look precise and still be wrong. It is just a formula, so if you put in the right numbers, you get the right answer. But the numbers depend entirely on what the engineer saw on site, what assumptions they made about the construction, and whether they correctly judged the ventilation.

Which is why speed gains from LiDAR scanning or plan tracing - hugely beneficial for scaling a business - need to be complemented by attention to detail. An engineer who scans a room in 10 seconds but misidentifies the wall construction still produces a wrong answer, only faster.

Experienced engineers will develop a feel for when a calculation looks wrong - when the total heat loss for a particular house is implausibly low or high given what they saw on site. Software supports that judgement and can help less experienced engineers develop it faster.

As a quick aside, historically the industry has veered towards oversizing. CIBSE defaults for ACH, party walls, solid brick U-values and window performance are all towards the conservative end when compared to actual measured data. We’ve written more about it here.

What happens when you get heat pump sizing wrong?

With gas boilers, sizing errors are largely self-correcting. A 30kW boiler installed in a home that needs 10kW will modulate, run inefficiently, but broadly work. The homeowner rarely notices. With heat pumps, the consequences of both undersizing and significant oversizing are more visible.

Undersizing

An undersized heat pump can't meet the design-day heat loss. On the coldest days of the year, the home won't reach the design temperature. The system runs continuously at full output and still falls short.

This results in an uncomfortable home on cold days, dissatisfied customer, and potential callbacks and complaints.

If you must err in one direction, a small amount of oversizing (10–15%) is safer than undersizing. The system can modulate down; it can't modulate up beyond its rated output. Szymon at Urban Plumbers has made the point that undersizing creates its own set of problems, particularly when manufacturer capacity data doesn't account for defrost cycles.

Significant oversizing

Oversizing by a large margin creates a different set of problems:

The radiator knock-on is worth understanding properly. Oversize the heat pump, then size the emitters to match that overstated heat demand, and you end up specifying triple-panel radiators in every room. They're expensive, they stick out from the wall, and the homeowner sees both. A lot of those projects die in the proposal stage - when an accurate heat loss calculation might have shown that standard double-panel radiators were sufficient all along.

In summary: Significant oversizing of heat pumps causes compressor cycling - the unit repeatedly switching on and off because it can't modulate below the building's actual heat demand on mild days. This reduces compressor lifespan, lowers efficiency (measured as Seasonal Coefficient of Performance or SCOP), and increases running costs.

How does emitter sizing connect to heat loss and flow temperature?

A heat loss survey is done room by room not just looking at the whole house. The main reason for this is to determine what emitters are required to meet the intended heat demand for the room. This calculation is directly linked to the flow temperature the system needs to run at.

Designing for low flow temperature

New heat pump installations should be designed for a low flow temperature - typically 35-55°C maximum. This has been built into MCS standards for a while, and building regulations are catching up - from March 2027 the Future Homes Standard makes low-carbon heating compatibility a requirement for new dwellings in England. A system designed this way:

• Keeps the heat pump operating in its efficient range

• Qualifies the property as "heat pump ready" for future upgrades

• Meets MCS compliance requirements for BUS grant eligibility

Our co-founder, Steph has written more on flow temperature and its relationship to retrofit decisions here, which is worth reading for the wider context.

Why room-by-room calculation is essential

Every room has a different heat loss - different wall areas, different window sizes, different exposure. The emitter in each room must be capable of delivering exactly that room's heat loss at the design flow temperature.

If radiators are undersized for the room's heat loss at the design flow temperature, the system has two options: run hotter (which reduces the heat pump's efficiency and raises running costs), or leave the room cold.

First, a quick note on ΔT50

Manufacturers publish radiator outputs at something called ΔT50 - meaning the average water temperature in the radiator is 50°C hotter than the room. This is the standard given gas boilers run at 70-80°C and rooms are around 20°C. This metric is useful for comparing radiators on a spec sheet, but it's less useful in a heat pump system where the water is much cooler. A radiator listed at 2000W at ΔT50 might only deliver around 800W in a real heat pump install.

So a correction needs to be applied - this is called the ΔT50 correction. Getting it right per room is fundamental to a system that runs at low flow temperature without leaving rooms cold.

Once the survey's done, the design comes down to two questions for each room: what radiator output is required, and does the existing emitter meet it at the system's design flow temperature? Doing that calculation manually for every room is where it becomes very time consuming on most spreadsheet workflows. Software handles it room by room, which both improves the accuracy and tells you exactly which radiators need replacing.

Where heat loss software changes the workflow

Engineers use heat loss software for the obvious reason that it takes the manual work out of the calculations. The more interesting question is what you do with the time that saves. Installers can take on more jobs in a week, or just get rid of their evening admin.

The alternative is running your own spreadsheets. Updating them every time industry standards change, filling in cells one by one, and hoping no broken references flow through into the rest of the survey. Software eliminates that risk, but there is still the question of what makes good software.

What good software actually changes:

U-value accuracy. Software that assigns U-values based on construction type, build period and wall material reduces risk of input mistakes, plus just saves installers time from having to look it up on every job. Good software also lets you add and save your own custom U-values for materials that don't fit standard libraries - useful for older properties and regional construction types. EPC data provides a starting point where one exists, though engineers should treat EPC U-values as indicative rather than definitive - EPCs are often based on assumptions rather than direct assessment.

Automatic ΔT50 conversion for emitter sizing. Radiator output data is published at ΔT50 as a standard reference condition. Adjusting that output to the system's actual design flow temperature requires a correction. Software that applies this automatically removes a step where manual calculation errors are common, and where getting it wrong results in undersized emitters.

Survey speed without sacrificing accuracy. LiDAR scanning on a mobile device, tracing over PDF floor plans - these reduce time spent on dimensions and inputs. LiDAR calculates room volume automatically, which feeds directly into the ventilation formula. Tracing off floorplans works for new builds and provides a good head start for retrofit surveys before going on site. The time saving creates more time on site for the observations that matter and building that trust with the homeowner.

In summary: Heat loss software standardises the calculation and design. Key workflow improvements include LiDAR room measurement (which calculates volume automatically for the ventilation formula), validated U-value libraries by construction type, automatic ΔT50 correction for emitter output, and automated generation of MCS-compliant documentation for BUS grant applications. Software accuracy is only as accurate as the quality of survey inputs - making sure these are right is something engineers should pay close attention to.

Get accurate heat loss calculations on every job

Get the heat loss right and everything downstream gets easier - the design, the proposal, the install, the running costs the homeowner sees.

Spruce is built to help engineers get the heat loss right, fast, and turn it into a system the homeowner trusts and you are proud of. LiDAR scanning, plan tracing, U-value libraries, automatic ΔT50 conversion and offline capability are all in one platform.

If you'd like to see how it fits your current survey workflow, book a demo or drop us a line at hello@spruce.eco.

Frequently asked questions about heat loss surveys and calculations

What's the difference between a heat loss survey and an EPC?

An EPC (Energy Performance Certificate) and a heat loss survey serve different purposes. An EPC assesses a building's energy efficiency rating across all energy uses - heating, hot water, lighting - and assigns a letter grade from A to G. A heat loss survey is a specific engineering calculation that determines the peak heating demand of the building in watts, used to size the heating system. EPCs use simplified assumptions and aren't accurate enough for heat pump sizing - using one as the basis for sizing is a common error. The correct approach is a dedicated room-by-room heat loss survey using the fabric and ventilation formulas above.

What is the design outside temperature, and where does it come from?

The design outside temperature is the coldest outside temperature the system needs to be designed to cope with - defined statistically as the 99th or 99.6th percentile of historical temperature data for that postcode. In most of the UK, design outside temperatures range from around -2°C to -5°C depending on location and altitude. Good software uses postcode-level weather data to apply the correct figure automatically. Using a single national default is a simplification that can over- or undersize systems in locations with significant climate variation.

What ACH value should I use for different property types?

Until recently, ACH (Air Changes per Hour) values were assigned per room based on construction type and condition, using default tables from the CIBSE Domestic Heating Design Guide. Research by Open Energy Monitor and others has shown these defaults significantly overestimate ventilation heat loss in most pre-2000 properties, which is around 93% of UK housing stock.

The 2026 update to the CIBSE domestic heating design guide moved the recommended approach away from these per-room defaults toward whole-home air permeability calculation, aligned with BS EN 12831-1:2017. This typically produces more accurate results, particularly for older properties. Spruce helps you calculate ventilation heat loss using this newer method.

If you're working from older CIBSE per-room values, treat them as a starting point rather than ground truth - and check the result against what you'd expect given what you observed on site.

What flow temperature should I design for?

New heat pump installations should be designed for a low flow temperature, typically 35–55°C maximum, in line with MCS requirements (with building regulations following from March 2027 under the Future Homes Standard). The system must be capable of maintaining the design internal temperature on the coldest day at this flow temperature - which means emitters must be correctly sized to deliver the required output at that temperature, not at the legacy 70°C or 80°C that older radiator sizing assumed. Designing for low flow temperature typically requires larger radiators than a traditional boiler system, which is why accurate room-by-room calculation matters. Undersized emitters mean the engineer has to either raise the flow temperature (which reduces efficiency) or accept that some rooms won't reach design temperature.

When is a heat loss survey not required?

A full room-by-room heat loss survey is required for any MCS-certified heat pump installation in the UK, and is a prerequisite for BUS grant applications. There is no compliant alternative. SAP assessments, EPCs and rule-of-thumb estimates based on floor area aren't acceptable substitutes under MCS standards. For new builds where construction isn't yet complete, a survey can be based on architectural plans and specified materials - the methodology is the same, the inputs come from plans rather than physical measurement.

What does a heat loss survey cost, and who pays for it?

Survey costs vary by installer and property complexity, but a thorough room-by-room survey for a typical residential property typically takes one to three hours on site, plus software input time. Some installers include the survey cost in their overall installation quote; others charge separately. For the homeowner, the survey is the foundation of a correctly sized, efficient system - the cost of a wrong survey is a heat pump that cycles, underperforms, or requires expensive post-installation correction. For the engineer, time invested in a thorough survey reduces callbacks and warranty disputes considerably.