Error & Uncertainty in whole house heat loss (Co-Heating) measurements

Samuel Stamp, UCL

Project Abstract

The so-called performance gap between designed and as-built building performance threatens to undermine carbon reduction strategies in the built environment. Space heating, as the largest domestic energy use, can principally be tackled through behavioral change and via heating system or building fabric efficiencies. Work to date has shown that the as-built fabric heat loss of tested buildings is on average 60% higher than design. This type of work is however still relatively in its infancy and is largely based on the use of co-heating tests. Importantly, the errors associated with co-heating are not well understood or typically addressed. The test’s reliability and accuracy are embedded in the available mix of weather and the building characteristics. Furthermore, the test procedure requires a building to be unoccupied for 2-3 weeks and is therefore often cited as unsuitable for developers. Questions over where the test sits in assessing thermal performance and providing feedback therefore still remain.

It is therefore crucial to firstly understand the errors and uncertainty in the co-heating test. The sensitivities to types of weather and building characteristics need to be clear. Questions on test protocol, particularly in new low energy buildings, may then be answered. It is hoped by understanding the test in more detail (heat loss mechanisms and building parameters) the uncertainty in the result can be reduced and the more detail on thermal performance could be provided.

Research Aims

Quantify and characterise types or error for various building types
Identify potential improvements to methodology; pratical protocol and data analysis methods
Develop more useful and detailed test outputs to provide greater feedback

Research Method

Sensitivities of various types of error on various types of dwellings will be mapped through the use of synthetic data and building simulation.
Experimental protocol or analysis methods can then be developed to handle identified errors.
Field Tests will help understand the praticalities of these issues and be used to support simulation work.
Through working with industry partners (architects, designers, builders & developers) field Tests will also be useful to determine the wider context around the role co-heating tests can forfill.

Involvement in Wider Projects

The NHBC Co-heating Steering Project will ‘investigate the sensitivity of variations in methodology’. Six organisations, including UCL, have carried out co-heating tests on a test dwelling at BRE over January – April 2012. An identical control house was also monitored over the same period. It is hoped finding from this series of tests and their subsequent findings will help develop guidance on an ‘optimal test method suitable for use in a range of buildings.’
IEA Annex 58 will look in ‘Reliable Building Energy Performance Characteristaion Based On Full Scale Dynamic Measurement.’ My involvement in this project will hopefully provide insight from participant based around Europe and the rest of the World and will introduce dynamic analysis methods to potentially apply to co-heating data.