The gap between design and built performance is seriously hampering our chances of meeting the 2016 target for zero carbon homes

What’s the biggest barrier to meeting the 2016 target for all new homes to be zero carbon? Well, there’s the continuing wait for a definition of what a zero-carbon home should be. Then there’s the fact that, as reported by the Zero Carbon Hub last month, the software used to calculate compliance with energy regulations isn’t up to the job and needs a complete rethink.

But maybe the most immediate concern is that the homes that have been built in recent years aren’t achieving anywhere near the performance targets they’ve been designed to. Studies carried out by Leeds Metropolitan University comparing the calculated heat loss through the fabric of 17 dwellings with the actual performance reveals that some homes are leaking 120% more heat than the design calculations suggest. Even the best performing ones are 10-15% above the target.

It was a similar story when Oxford Brookes University carried out tests on the Sigma eco-home at BRE’s Innovation Park - the first UK dwelling to hit level five of the Code for Sustainable Homes. Here they found that despite the building’s fabric being better than a standard home, its as-built heat loss was much worse than predicted.

Professor Malcolm Bell, head of the centre for the built environment at Leeds Met, says it is difficult to draw concrete proof from such a small sample. However, the results would suggest that there are significant performance gaps between what has been designed and what is being built. “From the tests we’ve done not one of them has produced the whole house heat loss co-efficient that you would get from calculations.”

Although airtightness testing for dwellings became compulsory in 2006, air leakage is only one of the factors determining the overall heat loss in a dwelling. Losses through the fabric from conduction, convection and radiation also need to be measured and the only readily available method for assessing this is by conducting a co-heating test (see box). This can be combined with other techniques such as infra-red thermography, which provides a more detailed picture of the heat loss mechanisms. All of which adds up to a relatively expensive test, which means few dwellings have been assessed to determine the as-built performance of the fabric.

However, from its studies, Leeds Met has pinpointed a number of areas where problems typically arise. These include thermal bypassing, particularly in the cavity of party walls, thermal bridging and airtightness around openings and service penetrations (see page 60). Despite tests being carried out on a number of different construction types, Bell says it’s difficult to say if one particular type of construction is more robust than another. What he does conclude is that it comes down to both design and workmanship on site. “Clearly you can design out some of the problems, like thermal bypass, but site operatives also need to understand what the issues are and be aware of things such as the continuity of thermal insulation.”

The seriousness of the gap between design and as-built performance hasn’t escaped the Good Homes Alliance (GHA), whose members include both large housebuilders and architects. Jon Bootland, director of the GHA, says there is a need for significant testing of fabric performance. “We need to get a better understanding of whether we are anywhere near achieving current standards and know what works and what doesn’t and what lessons can be learned.”

The GHA is now working on a three-year monitoring and feedback programme measuring the energy performance of four dwellings, each using different construction types. Fabric performance, occupant behaviour and services efficiency all have strong effects on energy consumption, says Bootland, but you need to be able to desegregate these to see where the problem areas are.

In a bid to pin down the fabric performance of these dwellings, co-heating tests have been conducted and the GHA will be drawing up guidelines for its members on testing.

Bootland suggests it would be especially valuable when construction teams are looking at new-build types or designs. “By building a prototype and testing it, problem areas can be identified before it is rolled out on a wider scale.”

For Bell, the biggest hurdle now is getting the industry to accept there is a problem. Only when they do, and more in-depth testing is conducted, can solutions start to be found.

What is a co-heating test?

If you’ve not heard of a co-heating test before, don’t worry, you’re not alone. It’s still a relatively new phenomenon in the UK and it’s the only way to get a definitive answer of what the as-built heat loss of a dwelling is, rather than what the design calculations predict.

It involves heating the inside of the dwelling using electric heaters to an average temperature, typically 25ºC. This temperature is then held for a period of between one and three weeks and the amount of energy needed to maintain the temperature is measured. The heat loss co-efficient for the dwelling can then be calculated by plotting the daily heat input against the daily difference in temperature between the inside and outside of the dwelling.

Equipment including sensors, circulation fans, data loggers and thermostats is required and, in an attempt to standardise the approach, Leeds Metropolitan University has drawn up a co-heating test protocol. This will allow like-for-like comparisons of different dwellings and build types.

Before the co-heating test is done, an air pressure test should also be undertaken. This will help determine the background ventilation rate that can be used to separate out the heat loss attributable to the fabric and the background ventilation.

The high temperatures used in a co-heating test also provide good conditions for carrying out a survey of the dwelling using a thermal imaging camera. This can reveal potential design and construction faults with the building’s envelope such as thermal bridges, which are not always easily identified under normal temperatures.

Where it can go wrong

Leeds Metropolitan University has carried out a number of in-situ tests to determine the as-built fabric performance of different dwellings, including a seven-year study on the Stamford Brooke development in Cheshire. Here are some of the common areas where problems can occur.

Thermal bypass

A large part of the discrepancy between the calculated and actual whole house heat loss co-efficient can be attributed to thermal bypass. This is most significant in dwellings with party wall cavities where upward air movement can occur as a result of thermal stack effects and pressure differences caused by wind. This effectively pulls heat away from the dwellings on either side.

Although it has traditionally been assumed that heat loss from party walls is insignificant, tests have demonstrated that it can in fact be very large, which is why the updated Part L of the Building Regulations is taking it into account.

Measures to reduce it include fully filling the party wall cavity with insulation, using solid party wall construction, horizontal cavity socks that close off the cavity and prevent air movement or sealing the top of the cavity with timber. However, these measures must ensure that the requirements for acoustic performance - the main reason for having a cavity - are not compromised.

Thermal bypassing can also occur wherever there is air movement around insulation, which reduces its effectiveness, potentially doubling heat loss. Ways of avoiding this include ensuring the air barrier and insulation are in contact with each other at all times, fitting insulation tightly to supporting structures and designing details so it is easy to accommodate tight-fitting insulation.


Uncontrolled air leakage can dramatically increase a dwelling’s heat loss. According to Leeds Met, it is important that it is addressed at the design stage and throughout construction. Designers need to identify a continuous line through the building envelope where the primary air barrier will be and care is needed at junctions and penetrations to ensure that integrity is maintained. Complexity should be avoided and the air barrier needs to be accessible for inspection, maintenance and repair and robust enough to last the lifetime of the building.

Secondary sealing using mastics, caulks and foam have been shown to shrink over time, resulting in air leakage. As a rule, they should not be relied on. Site operatives need to be aware of what the air barrier is and what they need to do to achieve it.

Thermal bridging

As higher levels of insulation are used in building envelopes, the relative impact of thermal bridges increases. Understanding and accurately modelling these is critical to predicting the building’s heat loss. Thermal bridges fall into three categories: repeating, such as metal wall ties used in cavity wall construction; non-repeating, such as door thresholds; and geometric, where complex junctions are formed and the thermal envelope can become compromised.

Designers should look to prevent thermal bridges where possible by not interrupting the thermal envelope or where it does occur making sure the thermal resistance of the materials are as high as possible - such as using polyester rather than steel wall ties. At junctions insulating materials should meet without any gaps.