On most commercial and industrial sites with substantial heat demand, heat is the biggest single line on the energy bill. And yet the first thing most of us do is buy new plant.
Nobody asks the awkward question first: are we already paying to make heat that we then throw away?
Usually the answer is yes. Compressors, chillers, process cooling and boiler flues are all pushing useful heat out of the building while a boiler fires to make more.
So heat recovery ought to be one of the most straightforward wins on site. It rarely is. The kit is the easy part. Get the timing, the temperatures, the controls or the evidence wrong and a promising project becomes an expensive addition to the plant room that nobody can defend.
And the building won't tell you. Buildings lie, cheerfully, and the bill is the only witness.
I've watched these seven mistakes play out across sites. Here they are, with the check that stops each one.
1. Plant sizing: treating nameplate capacity as real demand
A heat source or demand rated at a certain capacity does not run at that capacity all day. Nameplate data tells you what equipment can do under specified conditions. It tells you nothing about the site's average load, its operating hours, its peaks or how those demands overlap.
Size a recovery system around the wrong number and it will be too large, too small, or useful for far fewer hours than the business case claims. That is how a good idea gets a reputation for not working.
Do this instead: build a time-based load profile. Record source and demand temperatures, flow rates and operating hours across representative shifts and seasons. Distinguish peak demand from average demand, and account for diversity: different processes do not necessarily reach peak load together. CIBSE’s AM17 guidance for large non-domestic heat pumps covers demand characteristics, sizing and part-load performance for a reason.
If you'd rather not build that profile yourself, our Projects team designs against a measured load profile instead of a datasheet.
2. Heat pumps: buying one before cascading available heat
A heat pump can raise low-grade heat to a useful temperature. That doesn't mean it should be the first thing you buy.
Most sites I walk round already have heat available from compressors, process cooling, boiler flue gases or chiller heat rejection. Some of it can be used directly, with no heat pump at all. So the right question is not "which heat pump should we install?" It is "what heat do we actually have, and where can it do useful work?"
Buy the compressor first and you have paid to upgrade heat you could have used for free.
Do this instead: list the sources and demands, then match them in stages. Look first for direct heat exchange. Consider a heat pump where the recovered heat is too cool for the demand and needs a temperature lift. The Carbon Trust heat pumps guide also recommends understanding the waste heat and the site’s thermal demands before selecting a system.
3. Temperature matching: using valuable heat for the wrong job
Heat has a temperature grade, and it matters more than most people expect.
Put your high-temperature source onto a low-temperature job like pre-heating or space heating, and you have left a valuable demand to be served by a boiler. Try to serve a high-temperature process with low-grade heat and you will spend a fortune lifting it.
Either way, the plant room gets busier and the expected benefit does not appear.
Do this instead: map source and demand temperatures, including the required flow and return temperatures. Match the highest-grade heat to the demands that need it. Then cascade what remains to lower-temperature duties. Keep the proposed arrangement grounded in actual operating conditions, not just a single temperature on a plant datasheet. That matching work is where independent design and delivery earns its keep, before anything gets bought.
4. Simultaneity: missing heat and cooling that run together
One system can reject heat while another in the same building fires a boiler to make it. Heating and cooling often run together too, across different zones, processes or schedules.
That overlap can be the opportunity. Assume it without checking the timings and the project looks better on paper than it ever will in operation.
Do this instead: put source and demand on the same timeline, across shifts and seasons. Where timings don't line up, assess controls, scheduling or thermal storage. Don't assume storage solves it: include its losses, controls and cost.
5. Metering: having no way to prove the saving
This is the one that costs you the most credibility, and the most money. If you can't measure the useful heat recovered and the energy used to recover it, you can't show what the system delivered.
A lower bill after installation is not proof of savings. Production may have fallen. The weather may have changed. The site may be running fewer hours. Buildings lie, and they will lie cheerfully to a finance director.
That makes reporting hard for everyone: the facilities team managing performance, the sustainability lead reporting progress and the finance team testing the investment case.
Do this instead: plan metering before procurement. Define the measurement boundary and meter the heat source, useful heat delivered and electricity used by compressors, pumps, fans and controls. Record relevant operating conditions too. Establish a baseline, then compare post-installation performance against an adjusted baseline that accounts for meaningful changes such as production, occupancy, weather and operating hours.
This is measurement and verification thinking: evidence that compares like with like. CIBSE AM17 includes metering and monitoring, commissioning and operation because a system’s performance has to be checked after handover, not just promised before it.
This is exactly what Enerlyse exists to do: half-hourly data, an adjusted baseline, and a report that still stands up when finance asks how you know.
6. System design: forgetting the loads around the heat exchanger
Recovered heat is not free if the recovery system needs substantial energy to capture and move it. Pumps and fans use electricity. Added pressure drop can make existing equipment work harder. Poor access can complicate maintenance. Fouling can reduce heat transfer if cleaning is not planned.
Ignore those practical details and they can eat into the benefit, while a system that is awkward to service may gradually perform worse.
Do this instead: include the whole system in the design review. Check pump and fan energy, pressure drop, controls, isolation, access for maintenance, expected fouling and a realistic cleaning regime. Ask who will maintain it, how often, and how performance will be checked. Compare net benefit, not gross recovered heat. The parasitic load is part of the capital case, not an afterthought in commissioning.
7. Investment case: filed as maintenance, so it is never funded and never owned
Heat recovery gets treated as a plant-room improvement rather than an investment that needs an owner, a staged plan and evidence.
So it competes for maintenance budget against a leaking roof, has no agreed verification plan, and never gets signed off. Or it does get installed, and nobody ever checks whether the benefit lasted past the first cold snap.
Do this instead: answer the three questions a finance director will actually ask. What does it cost? What does it save, net of the parasitic load? And how will you prove both? If you can't answer the third, you don't have a case, you have a plant-room wish.
Then make it investable: opportunity, cost, operating assumptions, risks, measurement plan and route to funding, in sensible stages if that helps a decision get made. Agree in advance who owns delivery, reporting and follow-up. If capital is the blocker, PowerPacks explores flexible financing so a good scheme doesn't die at the funding stage.
Heat recovery: sequence the decision before buying kit
The practical order is simple:
- Measure actual energy use, heat sources and demands.
- Model the demand across operating hours, shifts and seasons.
- Cascade available heat to the demands it can serve, checking temperature and timing.
- Size equipment against the real load profile, not nameplate capacity.
- Meter the recovered heat, useful output and parasitic energy.
- Verify performance against a properly adjusted baseline and report what changed.
Data comes before kit. That is how you make a decision the facilities team can operate, the sustainability team can report and the finance team can scrutinise.
Most sites are being reactive: paying to make heat, throwing it away, then paying again to make more. The cheapest source of capital is the energy you are wasting.
Not sure where you sit? Start with the free Energy Health Check and find out what your building is actually doing.
If you want to talk through a heat recovery opportunity, email andrew@hawleyenergy.co.uk or call 01484 958761 to arrange a no-pressure 30-minute call. No sales pitch, no slide deck. Bring your plant room and I will bring the questions.
if you want to read more, click here: https://www.hawleyenergy.co.uk/wyms