There is a stack of factors that make it practically impossible for installers to right-size heat pumps. The standards encourage you to layer error upon error, meaning you end up with a calculated heat loss result that significantly overestimates the real value. Oversizing inflates costs so may reduce your conversion rate, and also reduces performance, leading to higher bills for the homeowner. Here are some things to watch out for to avoid oversizing in your jobs.
Why oversizing heat pumps matters
Oversizing heat pumps creates two significant problems:
Cost and feasibility barriers
At a basic level, specifying larger heat pumps than necessary means homeowners pay more than they need to for their units. But the real damage happens for bigger properties, where inflated numbers create major barriers:
At best, you install a cascade system, adding cost and complexity to the job, or failing that, homeowners are forced into expensive three-phase power upgrades.
At worst, the home is deemed "unsuitable" for heat pumps entirely.
When our calculations are systematically high, we're telling people they can't have heat pumps in properties that could actually be perfectly suitable. This isn't just about cost anymore - it's actively blocking adoption in homes that could successfully transition away from fossil fuels.
The cycling problem
Heat loss calculations are done for the coldest expected temperature - but most of the time, it's much warmer. This means the house needs far less heat than the calculation suggests. While heat pumps can modulate their output, they can only turn down so far (some brands are better than others). When even the minimum output is too high for what the house needs, the heat pump has to cycle on and off. This both reduces efficiency (compressor starts are inefficient) and shortens the lifespan of expensive equipment.
Factors pushing oversizing
Here's what we're seeing in terms of factors pushing oversizing:
Magic Heat Creation
When calculating room-by-room heat loss, many tools assume adjacent rooms are at 18°C. They do this so they don’t need to deal with the complexity of what room is touching what. But it results in a situation where you are saying the kitchen is losing heat to the living room and the living room is losing heat to the kitchen 🤨. Add all the rooms up and you've got this mysterious extra heat loss that can't exist in reality.
Air Changes Per Hour (ACH)
The CIBSE domestic heating guide recommends values by room type and building age. That premise seems very reasonable, but the values simply don't match reality. Real blower door test data typically shows infiltration below 10 m3/m2/hr (~0.9 ACH at 4Pa), yet CIBSE values are above 1 for all pre-2006 homes. The team at Open Energy Monitor dug into this - looks like CIBSE are probably missing a factor of two in the calcs 😬. And if there's a fireplace? CIBSE says bump the ACH to 4 - a value that is apparently equivalent to having a whole wall open!
Solid Brick Wall U-values
Measured data suggests solid brick walls perform much better than CIBSE values indicate. There is a fair bit of variation but the average measured value is 1.59 vs. CIBSE's recommended value of 2.11. If the average value is right for your property, you've overestimated external wall heat loss by 30%.
Party Wall Assumptions
CIBSE guidance says to "assume the adjoining property is unheated, even when it is known that a heating system is installed". In reality, your neighbour’s house is likely at a very similar temperature to yours. While there's some heat loss through thermal bridging, assuming the whole wall loses heat to 10°C significantly overestimates losses. If neighbouring properties really are unoccupied for long periods, then maybe that’s a conversation worth having with the homeowner, but in most cases, this assumption just doesn’t make sense.
Window U-values
The default U-values in CIBSE for double glazing are stuck in the past. Standard "Double glazed window with UPVC or wood frame" is listed at 2.8 W/m2K, yet building regs required 2.2 or lower back in 2003! Even 20-year-old windows perform better than the default. Current regulations set a limiting value of 1.6 (notional 1.2). And these calculations ignore curtains, which can dramatically reduce heat loss.
Internal Gains
Unlike SAP calcs, Heat loss calculations do not include the contribution of internal gains. A typical home has mid-hundreds of watts coming from appliances, cooking, and the people inside. Solar gains can also be really substantial. Now it might not be sunny on your coldest day, so there is some logic to leaving some of these out, but they do very much still contribute to cycling in general operation.
Layer all these conservative assumptions together and you can see why heat pumps keep ending up oversized.
Ok, not everyone
It's worth noting that there's a growing community of engineers actively using measured data and real-world monitoring to right-size systems. But those installers are swimming against the tide of industry defaults and standard practices, and are left trying to convince customers that their design is right compared to others.
When industry defaults push everyone toward oversizing, properly sized systems can look suspicious. And for newer engineers just entering the space it’s very counter-intuitive that you should need to push back on the standards.
Of course, there is a balance here - Szymon at Urban Plumbers, has some excellent content showing how undersizing (particularly when manufacturer’s capacity data doesn’t account for defrost) creates its own problems. The goal isn't to totally eliminate any safety margin, but to stop layering multiple 30% oversize factors one on top of the other.
Getting this right
CIBSE itself acknowledges these limitations in its preface, advising professionals to use their judgment. Working on heat loss tools at Spruce has shown me how critical it is to make accurate sizing easier for engineers. As our understanding improves through real-world data and monitoring, hopefully we'll see standards evolve to help installers size heat pumps with appropriate margins, not layers of worst-case assumptions.
How Spruce can help
Spruce helps you deliver high quality installs and streamline your workflow so you can save days every month. All the tools you need to attract more leads, save time on-site, produce high quality system designs, generate paperwork and win more jobs!
If you’re interested in discussing any of this or learning more about Spruce, we’d love to hear from you – please reach out to hello@spruce.eco or book a call with us here.
This post was originally shared by Spruce co-founder Steph Willis on LinkedIn.