← All insights Insight

Why domestic hot water is often the hardest part of heat pump design

Jonathan Aird · 2 June 2026

Space heating is the part of a heat pump retrofit everyone concentrates on. Hot water is the part that quietly determines whether the system hits its efficiency figure — and in leisure centres, healthcare buildings and anywhere with showers, it is frequently the larger problem.

The fundamental conflict

A heat pump becomes less efficient as the water temperature it produces rises. Space heating, on a well-considered retrofit, might run at 50 to 60 °C, and on a system with generous emitters lower still.

Hot water does not have that flexibility. Stored water generally needs to reach 60 °C for legionella control, and to get water in a cylinder to 60 °C the primary circuit has to run hotter than that. So the least efficient duty in the building is also the one with the least room for negotiation.

In a school, where hot water is a small fraction of total demand, this matters little. In a leisure centre with showers running all day, or a healthcare building, hot water can dominate the load — and if it is being served inefficiently, it drags the whole system's measured performance down with it.

The coil problem

This is the detail that catches out more retrofit schemes than any other, and it is entirely predictable from a survey.

An indirect cylinder transfers heat from the primary circuit to the stored water through a coil. That coil was sized for a particular primary temperature. Drop the primary temperature and the coil transfers less heat, because the driving temperature difference has shrunk. The cylinder still gets there eventually — but eventually may not be soon enough.

A school we surveyed had a 500 litre indirect cylinder with a 40 kW coil selected for approximately 80 °C primary flow. Serve that same cylinder from a heat pump at 55 °C and the coil's output falls dramatically. The cylinder will not recover between demand peaks, and at some point somebody runs out of hot water.

There are four ways out, and they need to be decided before plant selection rather than after:

• A larger coil or a plate heat exchanger. Increase the transfer area to compensate for the smaller temperature difference. Often means replacing the cylinder.
• More storage. Accept slower recovery and buffer the peaks with volume. Needs plant space, which is usually the thing in shortest supply.
• A separate high-temperature source. A dedicated high-temperature heat pump, or a retained boiler, serving hot water only while the main plant handles space heating at a lower, more efficient temperature.
• Keep hot water on the existing fuel for now. Unglamorous, sometimes the right first phase, and honest about what has and has not been decarbonised.

Legionella, and what it does and does not require

Legionella control is a legal duty and it is not a place for creativity. In practice it means stored water reaching 60 °C, distribution returning at 50 °C or above, and outlets reaching temperature within a defined period.

What is worth understanding is that these requirements apply to stored and distributed water, not to every part of the system at all times. Instantaneous and semi-instantaneous arrangements, which store little or no hot water, change the calculation. So do designs that use a heat pump to pre-heat and a smaller high-temperature source to finish — the heat pump does the bulk of the work in its efficient range and something else covers the last few degrees.

That last arrangement is often the most efficient answer available on a retrofit, and it is regularly overlooked because it appears to compromise the all-electric ambition.

Refrigerant choice matters more here than anywhere

Hot water is the one duty where the refrigerant genuinely changes what is possible.

CO₂ heat pumps are unusually well suited to it. They perform best with a large temperature lift and a cold return, which is exactly what hot water looks like: cold mains in, 60 °C-plus out. For a building with high, consistent hot water demand, a CO₂ unit dedicated to hot water alongside a conventional heat pump for space heating will frequently outperform a single system trying to do both.

Propane units are the other useful case. They reach higher flow temperatures than most conventional refrigerants — up to 80 °C on the units we have specified recently — which means existing cylinders and coils can often be retained. That has real value in a building where replacing the cylinder means getting it down a basement stair that no longer exists at the right size. Both refrigerants bring charge limits and plant location implications that have to be designed for.

What we look at

On any building where hot water is more than incidental:

• Actual demand profile, not the design figure — when the peaks occur and how long they last
• Existing cylinder volume, coil rating and the primary temperature it was selected for
• Recovery time required between peaks, which in a leisure centre may be very short
• Whether secondary circulation is running continuously and what it is costing
• Dead legs, and whether the distribution can hold its return temperature
• Whether hot water and space heating can sensibly be separated onto different plant
• Whether the heat pump can be metered separately for hot water, so its performance can be seen rather than inferred

That last point is worth dwelling on. Where hot water and space heating share plant and metering, a poor hot water strategy is invisible — it just shows up as a system-wide efficiency that is lower than expected, with no obvious cause. On the sites we monitor, heating and hot water are reported separately for exactly this reason. On one site a recent week showed heating running at 3.50 and hot water at 2.81. Both are respectable figures against their design targets. Averaged together they would have told you far less.

The summary

Hot water is not a detail to resolve after the heat pumps are chosen. It changes plant selection, refrigerant choice, cylinder and coil requirements, plant space and control strategy. Decided early it is a design problem with several good answers. Discovered late it is a cylinder that will not recover, in a building full of people who need a shower.