Metrics for Homes: Why Measurement Could Transform Housing Performance

Professor Allinson's presentation at the Winter Ideas Exchange event opened with a striking comparison. When buying a smartphone, consumers can access dozens of performance metrics—display quality, battery life, camera specifications. Twenty years ago, few people knew what a megapixel was; now we instinctively understand that 20 is better than 10.

Cars have had meaningful metrics for decades. Miles per gallon influences both driving behaviour and purchasing decisions. Modern vehicles provide real-time dashboards showing performance across multiple measures.

Yet homes—the most expensive purchase most people make—offer almost nothing comparable. The measurement device in most homes might tell you how many cubic feet of gas have been consumed since the metre was installed. That's about it.

The Energy Performance Certificate represents our best attempt at housing metrics, but Professor Allinson highlighted a fundamental problem: of all the numbers on an EPC, only the first two are actually measured. Everything else is based on assumptions and models. We invest in collecting data, then don't trust the information it produces.

Learning from Test Houses

Loughborough University operates a pair of 1930s semi-detached houses near campus that serve as living laboratories for building physics research. The houses have had identical maintenance since construction and were renovated identically in 2016. They face south, experience identical weather, and can be operated to simulate normal household behaviour—opening windows, drawing curtains, generating the heat that comes from people watching television or sleeping.

This matched-pair approach enables true scientific experimentation. Make a change in one house, compare it with the other, and observe the actual difference.

One study examined zonal space heating controls—motorised TRVs that allow different rooms to be heated to different schedules. The test house results showed energy savings of around 12%. Promising, but what would happen in real homes with real people?

A field trial across 68 gas-heated, occupied semi-detached homes in the Midlands provided the answer. Of these, 37 received the zonal controls while the rest continued as a monitoring control group. The results were illuminating: some households increased their gas consumption significantly, using the technology to improve comfort rather than save energy. Others achieved substantial reductions. The average saving across the

intervention group was just 3.5%—real, but far from the 12% test house result or the 50% some manufacturers had claimed.

The lesson is clear: test house experiments show what's possible, but field trials reveal what actually happens when technology meets human behaviour.

Busting the Insulation Myth

With overheating increasingly on the agenda, Professor Allinson addressed a persistent misconception: that insulation causes homes to overheat.

"All the work we've done shows that's not true," he stated firmly. "Insulation stops heat coming in—that's a good thing."

Research has demonstrated that energy efficiency measures significantly reduce cold discomfort without increasing overheating risk. Climate change mitigation and adaptation are entirely compatible, which matters enormously for retrofit strategy.

The real problems with overheating lie elsewhere. Ventilation provision in new homes tends to be poor. Through the heat of the day, the goal should be to seal the building, close shutters or blinds, and prevent heat entering. Then, when temperatures drop at night, ventilation should expel accumulated heat as quickly as possible.

External shading and effective ventilation are the genuine solutions—but both require occupants to understand when and how to use them. The information gap, not the insulation, is the problem.

The Heat Transfer Coefficient: A Metric That Matters

At the heart of Professor Allinson's vision is a single metric: the Heat Transfer Coefficient, or HTC. This describes the average rate of heat loss from a building per degree of temperature difference between inside and outside. If a home has an HTC of 100 watts per Kelvin and there's a 10-degree difference between indoor and outdoor temperatures, the heat loss is 1,000 watts.

The HTC is already calculated for every EPC—but never reported. The information is generated then lost, never recorded anywhere accessible.

Historically, measuring HTC required emptying a house for two to three weeks, filling it with electrical heaters, and holding indoor temperatures at 25°C. Expensive and impractical—only around 350 such tests have been conducted worldwide.

The breakthrough has been developing "in-use" measurement methods. With smart meter data, indoor temperature sensors, and weather information, the HTC can be measured while people continue living normally in their homes. In the past five years, over 10,000 such measurements have been taken. The UK leads the world in this approach, thanks largely to work being driven by the Department for Energy Security and Net Zero.

Closing the Performance Gap

The "performance gap"—the difference between how buildings are expected to perform and how they actually perform—has been discussed for thirty years. Professor Allinson presented new analysis suggesting we can now understand why it exists.

Using data from the zonal heating controller field trial, the team compared predicted gas consumption (from SAP models) with actual measured consumption. The correlation was poor, with an R-squared of just 0.3.

They then examined whether the problem was the HTC. Comparing SAP-predicted HTC values with in-use measured values showed similarly poor correlation (R-squared 0.27).

But here's the crucial finding: when measured HTC values were fed back into the prediction models, the correlation between predicted and actual gas consumption jumped dramatically—from 0.3 to 0.6. Using all in-use measurements pushed the R-squared to around 0.9.

The models aren't fundamentally wrong. They're being given wrong inputs about thermal performance. Measure the building properly, and predictions improve dramatically.

What Measurement Could Enable

Professor Allinson outlined the practical applications of widespread HTC measurement:

Targeted retrofit:

Rather than relying on EPC ratings that may not reflect reality, identify which homes are genuinely performing poorly. Some homes with C-rated EPCs might actually perform like an A or B; some rated B might perform like a D or E. Measurement reveals which properties truly need intervention.

Heat pump sizing:

Heat pumps tend to be massively oversized—more expensive, larger, and noisier than necessary. Accurate HTC data enables right-sizing.

Quality assurance:

If you pay for loft insulation, you should be able to verify it was installed properly and delivers the expected improvement. Measurement provides accountability.

Preventative maintenance:

With sensors already in place for HTC measurement, the same infrastructure can identify damp and mould risk, diagnose whether problems stem from insulation, heating, or ventilation inadequacy, and provide real-time health warnings during extreme weather.

New business models:

Warm rents and heat-as-a-service become viable when performance can be measured and guaranteed.

The Revolution Ahead

Smart meter are now in over 50% of UK homes. Internet of Things devices increasingly include environmental sensors, often without owners realising. The computing power in our pockets can transform raw data into meaningful metrics.

Professor Allinson believes we're on the cusp of a data-driven revolution in home performance measurement—one that could happen quickly over the next five years. The UK is genuinely world-leading in this space, with DESNZ actively exploring how measurement might become commonplace in policy.

The paradigm shift is from estimation to measurement. From models based on assumptions to insights based on reality. From talking about the performance gap to actually closing it.

"Why are there so few metrics for homes?" Professor Allinson asked the audience. "Is it status quo bias? We've got used to EPCs and any change feels like a step too far?"

Perhaps. But as he noted, Loughborough knows something about challenging the status quo—the Luddites famously smashed knitting machines nearby. Whether this digital revolution proves as significant as the industrial one remains to be seen, but the direction of travel seems clear.

Measurement changes everything.