Duration:
2017-2021
Status:
Completed

Building airtightness tests have become very common in several countries, either to comply with the minimum requirements of regulations or programmes, or to justify input values in calculation methods. This raises increasing concerns about the reliability of those tests.

There are four key sources of uncertainty in airtightness testing:

  • Measurement devices (accuracy and precision);
  • Calculation assumptions (e.g. reference pressure, regression analysis method);
  • External conditions (wind and stack effect impact); and
  • Tester behaviour.

While competent tester schemes and independent checking procedures show potential to reduce errors due to tester behaviour, there have been extensive yet sterile debates about how the building pressurisation test standard ISO 9972 should address other sources of uncertainty. As a result, no changes were made to these aspects in the new version of the standard published in September 2015.

Under the current standard, the zero-flow pressure shall not exceed 5 Pa for the test to be valid. Consequently, in moderately windy conditions, it may be impossible to perform a pressurisation test in accordance with the standard, even using precautions with a careful uncertainty analysis.

The goal of this project was to provide guidance on estimating uncertainties due to wind, which may be used to extend the applicability of standard ISO 9972. This should allow testers to provide conservative estimates of a building's permeability in windy conditions and confirm compliance to a given requirement.

Publications

Events

Conferences:
  • 38th AIVC Conference, September 2017, Nottingham:
    • Topical session "Integrating uncertainties due to wind and stack in declared airtightness results"
  • 39th AIVC Conference, September 2018, Juan-les-Pins
    • Topical session "Integrating uncertainties due to wind and stack effect in declared airtightness results"
  • 40th AIVC Conference, October 2019, Ghent:
    • Topical session "Integrating uncertainties due to wind and stack effect in declared airtightness results"