W2UP2 – Simulating heat pump performance in dwellings
23rd October 2015 Alison Parker

Simulating heat pump performance in dwellings

David Samuel, UCL MRes EDS



In the transition to a low carbon economy, air source heat pumps will be a critical demand side technology. Heat pumps function on electricity and ambient heat alone, therefore cutting the need for gas consumption and potentially exploiting the use of renewable power. The added benefit of air to water heat pumps is their ease-of-retrofit to domestic buildings: they can often replace gas boilers on existing water distribution systems, reducing installation costs.

The government has stated its intention to support air to water heat pumps under future Renewable Heat Incentives, subject to successful testing and affordability [1].


Fig 1. Dwelling heating system with air source heat pump [2]

Research Gap

The Energy Savings Trust conducted a recent field trial of both ground and air source heat pumps across various UK sites. The results for the air source units broadly showed a performance gap, between their design efficiency and delivered energy efficiency post-installation, as well as reduced performance compared to other European trials [3,4]. Consequently, research is needed to understand this performance gap, to improve installation practice and maximise energy and carbon savings.

Objectives and Method

Use a theoretical model of an air to water heat pump in a dwelling to simulate performance (in MatLab).
Compare the model to measured field data, and evaluate its limitations.
Perform a sensitivity analysis on the model to test the influence of key characteristics on unit performance. The varied model parameters include: heat pump sizing, control time, building type (i.e. thermal mass and heat loss coefficient) and radiator design.
Use the results to recommend further research and improved installation practice; for example, targeted building types for air source heat pumps.

Fig 2. Method summary.

Model-data comparison results

Fig 3. Temperature profile comparison between model and measured.

The calibrated model was validated on an average scale, matching measurements at hour time-resolution.
Key strengths of the model were its theoretical (thermodynamic) transparency; and its flexibility, as many scenarios could be analysed.
Key limitations were its representation of modulating controls, dynamic heating response, and building design.
The model could be improved by incorporating variable speed controls, and possibly by integrating the MatLab component model into a building simulation package such as EnergyPlus. This would allow more accurate representation of building scale thermodynamics and design.

Sensitivity Analysis

Building parameters varied were:

Heat loss coefficient (W/K)
Building thermal mass (J/K)
Radiator heat transfer (W/K)
Heat pump parameters varied:

Power sizing (kW)
Operating period (hours per day)
A large number of different cases were tested. A key resulting trend was that there is a trade-off between:

(i) oversizing the heat pump size (power capacity) to deliver peak heat loads

(ii) undersizing and operating more constantly to minimise energy consumption.

The oversized case was found to offer heating system flexibility for dwellings with high heat loss coefficient, high thermal mass, and when quick warm-up periods are required, e.g. when occupants return to a cold house. In the model, given the heat pump was fixed speed (on or off), it was found that temperature controls were critical in minimising energy consumption; and limiting the opportunity for overheating. The effect of the temperature controls is to cycle the heating system on and off – in order to regulate heat load and room temperature. The undersized/constant operation case should be targeted to lower heat loss and thermal mass buildings. However, this places more importance on correct sizing of the heat pump capacity pre-installation; and this can result in reduced heating system flexibility, if building/occupancy conditions change after installation.

Further simulation work is required to test modulating control strategies, and to compare heat pumps with and without thermal stores (hot water tanks).


[1] DECC (2011) Renewable Heat Incentive

[2] ICS Heat Pump technology (2010) website

[3] DECC (2009) Microgeneration Certification Scheme: MCS 007 Issue 2.0

[4] Delta Energy and Environment (2011)


Primary: Prof. Rob Lowe Secondary: Dr. Catalina Spataru

Project Team

David Samuel

Phill Biddulph
Robert Lowe
Catalina Spataru