The house and the constraint that came first
The case-study property was a three-bedroom 1930s semi in Leeds with 118 m² of floor area, solid walls, and a gas boiler that had used 16,400 kWh over the previous year. A room-by-room heat-loss survey put design demand at 6.1 kW at −3°C. The owner wanted an air-source heat pump, but the existing 180-litre cylinder and a narrow airing cupboard made hot-water storage the first constraint, not heat-pump output.
Why storage, not boiler size, drove the design
Heat pumps deliver lower flow temperatures than gas boilers, so domestic hot water often needs a larger cylinder and a buffer to avoid short cycling. The measured morning draw was 140 litres: two showers and one bath between 06:40 and 07:25. A standard 180-litre cylinder would recover too slowly at the 45°C design flow temperature. The design team used bananzia to compare a 9 kWh and a 12 kWh thermal store against that draw, rather than estimating from cylinder volume alone.
What the bananzia comparison changed
The comparison showed that the 9 kWh store would keep the heat pump at full output for 52 minutes after the morning peak, while the 12 kWh store cut recovery to 31 minutes. The cause was not the heat pump’s nominal 7 kW rating but the interaction between tank standing losses, reheating time, and the 45°C flow temperature. The solution was a 12 kWh store, a 210-litre hot-water cylinder, and a 7 kW monobloc heat pump set to weather compensation. The extra store occupied a 600 mm by 700 mm footprint in the former airing cupboard.
Commissioning checks that made the store work
The installer used 28 mm primary flow and return pipework between the heat pump and the store, with a 22 mm cold feed and a 22 mm hot-water distribution manifold. The store was charged to 2.5 bar, the expansion vessel was sized for 12 kWh at a 55°C maximum, and the cylinder thermostat was set to 48°C with a 5°C hysteresis. A low-loss header kept the primary and secondary circuits hydraulically separate, which mattered because the underfloor heating loop and the cylinder had different pressure losses.
Result after one heating season
Monitoring from October to April recorded a seasonal performance factor of 3.0 for space heating and hot water, with average flow temperature at 43°C. The owner reported no morning hot-water shortfall, and the heat pump cycled less than once per hour during mild weather. Electricity use was 4,850 kWh over the period, against 16,400 kWh of gas previously. The useful takeaway is that bananzia should be used to test storage against measured draw patterns and heat-loss data; a cylinder label alone will not show whether a low-carbon system recovers fast enough.