Login
Purpose of the work
The airtightness test result is typicaly compared with a limiting value (compliance check), with the results of other tests of the same object by the same technician (when controling the evolution of airtightness during construction process) or with a test result of another technician (when verifying a suspicious result). These tasks need a reliable estimation of uncertainty, repeatability and reproducibility of test results in order to come to meaningful conclusions. In this article the results of repeatability and reproducibility experiments performend in the Czech Republic in the last years are presented, discussed and compared with the results of similar studies performed in Europe recently.
Method of approach
The reproducibility experiment (round-robin test), was repeated in 2010, 2012, 2013 and 2014 as a competence control of airtightness testers. The participants tested the same building (prepared before the test by the organisers) according to EN 13829 standard using their own equipment. The working procedure of participants was monitored in order to identify the deviations from the testing protocol.
In order to evaluate the repeatability, approximately 100 tests of a small testing facility (unheated box placed on the roof of a building) were performed during several days from April to December 2014.
Content of the presentation
The following topics are adressed in the contribution: motivations and benefits of round-robin tests, potential sources of measurement error – identified mistakes of technicians, round-robin test results and reproducibility of airtightness tests, analysis of the repeatability experiment results, comparison with results of other similar studies.
Results and assessment
In 2010, 2012 and 2014 the round-robin tests were held in the same single family building and under favourable climatic conditions. The average air flow rates at 50 Pa were 250, 240 and 280 m3/h respectively. The standard deviations ranged from 15 to 20 m3/h. In 2013 the round- robin test was held in a different building in strong wind which increased significantly the variability of test results. The average air flow rate at 50 Pa was 1440 m3/h with the standard deviation of 60 m3/h (4 % of the average value).
At the moment of abstract submission, only preliminary results of the repeatability experiment analysis were available. The average air flow rates at 50 Pa for each of testing days varied from 980 to 1030 m3/h, the corresponding standard deviations ranged from 3 to 12 m3/h (0,3 to 1,2 %) with 7 m3/h in average (0.7 %). The average air flow rate and standard deviation of all the tests performed are 1000 m3/h and 18 m3/h respectively. The test results seem to correlate with temperature.
Conclusions
The round-robin tests proved to be a suitable mean for a control of technicians competence allowing possible malpractice to be identified. Moreover, they bring an opportunity of meeting and discussion among the participants. Organisation of an international round-robin test might be beneficious.
Within a homogenous group of technicians the reproducibility seems to be relatively stabile in time. Reproducibility and repeatability standard deviations of air flow rate at 50 Pa (in m3/h) are very close to the results of a similar study performed recently by BBRI, although the air flow rates measured at 50 Pa are different in both studies. These results do not confirm the proportionality between measurement error and the air flow rate.
It seems that relatively good reproducibility can be achieved even under strong wind.
Note
For more information, please contact the reference author at: jiri.novak.4@fsv.cvut.cz