Natural experiment to measure change in energy use and indoor environment in dwellings with smart heat pump retrofits

Authors: Sahar Zahiri, Rajat Gupta, and Sam Hampton

Published in:  Proceedings of ECEEE 2021 summer study (virtual), 6-11 June 2021 Link

Date Published: June 2021


The UK Government announced in 2020 its ‘Ten Point Plan’ for a green industrial revolution, which includes a challenging target to install 600,000 heat pumps per year by 2028. Balancing electricity supply and demand locally is key to the success of achieving this target.

This paper uses a natural experiment approach to gather early insights into the change in energy use, indoor temperature and relative humidity profiles before and after installation of smart ground source heat pumps (GSHPs) and smart controls in nine social housing dwellings (5 bungalows, 4 flats) located in a socially-deprived area of Oxford (UK). The GSHPs replaced night-storage heaters and integrate smart controls to optimise heat production in line with outside weather and resident preferences. Indoor and outdoor temperature and relative humidity were continuously monitored at 15’ intervals using blue-tooth enabled data loggers across the nine dwellings.

In a subset of two dwellings, electricity use was monitored remotely using CT loggers from October to December 2020. Household surveys were conducted to establish the household characteristics, socio-demographics and the way residents heat their home. No correlation was observed between actual annual energy costs (self-reported) and EPC ratings. Indoor temperatures were found to be more stable across the nine dwellings after installation of heat pumps. Despite having similar size, number of occupants and occupancy patterns, there was wide variation in the range of indoor temperatures measured across the sample.

Post-heat pump installation bungalows experienced higher increase in indoor temperature as compared to flats, with mean indoor temperatures of over 25°C observed in mid-terraced bungalows, due to improved air-tightness as a result of cavity wall insulation, constant heating and limited window opening.

Smart controls were found to be regularly used by residents to easily increase the heating set point temperature to overcome the low output temperatures of the heat pumps. Following heat pump installation, daily electricity use increased to 14.3 kWh/day (against 7.8 kWh/day) in the bungalows, and to 9.2 kWh/day (against 6 kWh/day) in the flats, however, when normalised for weather, daily electricity use was found to be reduced by 49%. Electricity use for heating increased during the evening peak period, making a strong case for connecting to time-of-use tariffs to change the timing of electricity use so that heating is run in periods when electricity tariff is cheap and heating is avoided during the expensive peak periods.

Insights for EnergyREV:

After heat pump installation, indoor temperatures were found to be more stable across the case study social housing dwellings. However, this form of low carbon heating requires continuous engagement with residents encouraging them to move to an ‘always on’ heating system. This will mean re-training of tenant liaison officers in housing associations. Electricity use for heating also increased during the peak evening period making a case for exploring demand side response through smart controls.