The first question householders ask when looking into upgrading their heating system with a heat pump, is how much it will cost. This is no surprise. Even with government subsidies such as the current Boiler Upgrade Scheme, an air source heat pump is an expensive purchase. For information providers like heat pump installers, tradespeople hubs, or decarbonisation advocates, providing personalised cost information is all but impossible. It’s important though. Our consumer research showed that, without cost information up front, many householders are discouraged and stop exploring heat pump adoption.

A large number of factors influence cost, which until now made it challenging to estimate without detailed information about a building or even a site visit. However, a rich dataset exists that contains information about past heat pump installations including the hardware installed, the kind of home it’s installed into and the amount paid. This is held on the MCS database, which has information on every installation that has received a government grant.

We’ve created a tool that uses data from the MCS database to give householders an estimation of the price of a heat pump installation in their own home. We’re making this tool open source and we want to partner with anyone who’s interested in using it. For installers, the tool will help householders understand the costs before they even pick up the phone. And for information websites, this will add extra richness and detail to your offer. If you’re interested, get in touch with us and we’d be pleased to work with you.

The tool is based on a machine learning model. The model explores connections between the data collected by the MCS, the features of a property from its energy performance certificate (EPC) and the overall cost of the heat pump installation at that property. The machine learning process identified the features that were most useful for predicting costs, such as property type and age, and used them to build a model that can calculate costs based on these features. We then created a tool that speaks to the model. Consumers interested in getting a heat pump can use the tool to enter data about their own property and it can return an estimated cost for a heat pump installation based on the model’s calculations.

It’s a really exciting tool which we think offers something very new and useful to householders. It’s really important, though, to note its limitations. The model provides the most accurate prediction possible using the limited available data. However, many other factors influence the cost and it isn't designed as a substitute for a quote from an experienced installer. Moreover, because it looks at past installations to produce its findings, it can only produce an output for homes it has data to look at. This means that it can’t give a figure for flats, for homes in Northern Ireland, or for very large or very small homes. When output to the tool, the model’s results are rounded to the nearest hundred pounds and shown as a range. We would expect at least 50% of heat pump installations for such a property to fall within this range.

The tool is hosted in a demo form on the Nesta website. The price estimations it makes are correct, but we’ll work with interested partners to improve the design of the user interface.

We think the model offers two key benefits to householders.

While we are hosting a working demo of the tool on the Nesta website, we’re keen that it gets seen and used as widely as possible. We know that Nesta isn’t the go-to place for householders seeking information about heat pumps – others do this really well and we’re happy for that to continue. However, as an organisation, we work in the open and share what we make with others who share our aims. We foresee the following ways partners might use the tool.

1: Integrate it directly into your web page

The tool can be integrated into a partner’s web page allowing them to display the tool and its findings but with their own information and materials surrounding it. We can collaborate on the design to make it more appealing and use resources such as our recent report called How to reduce the cost of heat pumps to enrich the information offer.

2: Use the calculations in the model with your own interface

Partners can create their own interface that accesses the model and presents its outputs in the way they like.

3: Integrate the model into existing or new models or calculations

If partners already have their own methods of cost estimation, or are developing them, this model can be integrated into those calculations or provide a comparison or sense check for them.

Potential users are diverse. For example, they could be heat pump installation firms who want to provide initial information to potential customers before their first contact; they could be information providers who want to enrich their information offer; or they could be creators of complex tools that provide cost estimation for whole house retrofit.

If you could use the model or the tool, get in touch with us and we can work with you to help you to do so.