How reliable are blower door measurements?

Summary of a paper presented at the joint 45th AIVC conference and ASHRAE 2025 IEQ conference “IEQ 2025: “Rising to New Challenges: Connecting IEQ to a Sustainable Future” will be held on September 24-26, 2025, in Montreal, Quebec together with the 13th TightVent and the 11th venticool conferences.

Key words: Airtightness testing; Blower door test; Measurement uncertainty; Repeatability; Reproducibility.

Blower door results increasingly determine compliance, acceptance, and liability in construction projects. But how reliable are these measurements in practice? This paper examines the uncertainty of airtightness testing, based on published evidence and large-scale reproducibility studies.

Introduction

Airtightness is a key element of building performance, influencing energy efficiency, indoor air quality, and moisture control. Blower door measurements are widely used worldwide to demonstrate compliance with energy performance regulations and to support quality assurance in construction. As a result, test outcomes can have significant technical, contractual, and financial implications.

Despite this widespread application, the reliability of airtightness testing is not always well understood. Yet, uncertainty is an essential component of any reported measurement result. As stated in the Guide to the Expression of Uncertainty in Measurement, “without such an indication, measurement results cannot be compared, either among themselves or with reference values given in a specification or standard” (JCGM 2008).

In this paper, the uncertainty associated with airtightness measurements is examined through a review and analysis of repeatability and reproducibility studies, drawing on both published literature and new datasets.

Repeatability of the measurement

Repeatability describes the precision obtained when airtightness measurements are performed on the same building, by the same operator, using the same equipment, and within short time intervals. Table 1 summarizes results from nine published studies assessing the repeatability of blower door tests.

Table 1. Main Results from the Repeatability Studies Reported in Literature.

With the exception of (Persily 1982), (Bracke et al. 2016) and (Kölsch et al. 2025), all studies report a consistent standard deviation between 1% and 2%. In (Persily 1982) and (Kölsch et al. 2025), high wind speeds were recorded during testing, rendering standard deviation an unreliable indicator of repeatability. When only calm-day measurements are considered, (Persily 1982) reports a standard deviation of 1.1%, in line with the other studies. The higher deviations observed by (Bracke et al. 2016) may be explained by very low measured airflow rates or by elevated wind speeds, which were not documented.

Repeatability studies provide valuable insight into measurement uncertainty and its sources, and they are useful for evaluating the intrinsic performance of test equipment and protocols. However, three important limitations must be noted:

·         Each study applies to a single building. Although the consistency across studies suggests that a repeatability of 1–2% can often be expected, this remains an empirical observation rather than a universal rule.

·         Repeatability does not capture systematic errors, as repeated biases are not reflected in the standard deviation.

·         Human factors and protocol variations (e.g. differences in building preparation or pressure sequences) are excluded, since all tests are conducted by the same operator using the same procedure.

Reproducibility of the measurement

Values from existing literature

Compared to repeatability, far fewer studies have investigated reproducibility. Reproducibility refers to tests conducted on the same building by different operators using different equipment. Table 2summarizes results from four published studies.

Table 2. Main Results from the Reproducibility Studies Reported in Literature.

As expected, standard deviations under reproducibility conditions are significantly higher than those observed for repeatability. In addition, the consistency seen across repeatability studies is not observed for reproducibility. In (Murphy et al. 1991), the elevated standard deviation can be attributed to the limited experience of some operators, a factor not present in the other studies. Another key difference between studies lies in building preparation practices. In (Delmotte and Laverge 2011) and (Novak 2015), the building preparation was performed by a single person, whereas in (Murphy et al. 1991) and (Rolfsmeier et al. 2010), each operator prepared the building independently.

(Rolfsmeier et al. 2010) further quantified this effect. When operators prepared the building themselves (17 tests), deviations in q₅₀ ranged from −4% to +55% relative to the reference value. When the building was prepared in advance by a single person (12 tests), the deviation range narrowed substantially, to −6% to +7%.

15 studies in Czech Republic

Since 2010, round-robin tests have been organized almost annually by Asociace Blowerdoor CZ (A.BD.CZ). The four tests reported in (Novak 2015) are part of this broader dataset. Figure 1illustrates the evolution of the observed standard deviation over time.

Figure 1. Observed standard deviation over time, for yearly reproducibility studies conducted by Asociace Blowerdoor CZ between 2010 and 2024.

The clear reduction in standard deviation can be attributed to increasing familiarity with airtightness testing procedures as the method has become more widely adopted. In addition, some participants have taken part in multiple round-robin sessions and benefited from detailed feedback, leading to improved performance. Since 2020, the standard deviation has stabilized below 3%.

Large study in Belgium

Since 2015, Flemish regulations have required airtightness testing to comply with the STS-P 71-3 standard and to be performed by qualified testers (De Strycker et al. 2018). In this framework, the Belgian Construction Certification Association (BCCA) organized practical examinations in late 2014 and early 2015. Building preparation was carried out by the operators, but compliance with the standard was verified by an examiner prior to testing.

From 99 test results, the average airflow rate at 50 Pa was 1024 m³/h, with a standard deviation of 4.4%. When pressurization and depressurization results are considered separately – yielding 198 individual measurements – the standard deviation increases to 7%. This demonstrates that averaging both modes, as required by the STS standard, significantly reduces measurement uncertainty.

Conclusion

This paper evaluates the reliability of blower door testing by reviewing published repeatability and reproducibility studies and by introducing new reproducibility datasets from the Czech Republic (A.BD.CZ) and Belgium (BCCA). The main findings are as follows:

·         Under non-windy, repeatability conditions, the intrinsic measurement standard deviation lies between 1% and 2%. These values were obtained with experienced testers.

·         Under reproducibility conditions, when experienced testers follow verified procedures, the standard deviation generally remains below 4%. This uncertainty increases when the procedure does not require averaging pressurization and depressurization results.

·         Under reproducibility conditions involving inexperienced testers or inconsistent building preparation, deviations can exceed 20%.

In practice, measurement uncertainty is usually expressed as the range within which the “true” airtightness value is expected to lie with high confidence (95%). For the results discussed in this paper, this range can be approximated as ±2 times the standard deviation. For example, a standard deviation of 2% implies that the “true” airtightness value is expected to lie within ±4% of the reported result.

These findings highlight the importance of clear and enforceable guidelines for building preparation, as well as robust qualification schemes for testers supported by high-quality training.

References

Bracke, Wolf, Jelle Laverge, Nathan Van Den Bossche, and Arnold Janssens. 2016. ‘Durability and Measurement Uncertainty of Airtightness in Extremely Airtight Dwellings’. International Journal of Ventilation 14 (4): 383–94.

Brennan, Terry, Gary Nelson, and Collin Olson. 2013. ‘Repeatability of Whole-Building Airtightness Measurements: Midrise Residential Case Study’. Paper presented at Workshop on Building and Ductwork Airtightness Design, Implementation, Control and Durability: Feedback from Practice and Perspectives.

De Strycker, Maarten, Liesje Van Gelder, and Valérie Leprince. 2018. ‘Quality Framework for Airtightness Testing in the Flemish Region of Belgium – Feedback after Three Years of Experience’. 39th AIVC Conference (Antibes Juan-Les-Pins, France).

Delmotte, Christophe, and Jelle Laverge. 2011. ‘Interlaboratory Tests for the Determination of Repeatability and Reproducibility of Buildings Airtightness Measurements’. 32nd AIVC Conference.

JCGM. 2008. JCGM 100, Evaluation of Measurement Data - Guide to the Expression of Uncertainty in Measurement.

Kim, Andrew K, and Chia Yu Shaw. 1986. ‘Seasonal Variation in Airtightness of Two Detached Houses’. ASTM Symposium on Measured Air Leakage of Buildings (Philadelphia, USA).

Kölsch, Benedikt, Valérie Leprince, Joachim Zeller, and Iain S. Walker. 2025. ‘Uncertainty in Building Airtightness Tests: Comparison of Regression Techniques Using a Comprehensive Dataset of 6,000 Tests’. Energy and Buildings 330 (March): 115328. https://doi.org/10.1016/j.enbuild.2025.115328.

Murphy, W.E., D.G. Colliver, and L.R. Piercy. 1991. ‘Repeatability and Reproducibility of Fan Pressurization Devices in Measuring Building Air Leakage.’ ASHRAE Transactions 97 (2).

Novak, Jiri. 2015. ‘Repeatability and Reproducibility of Blower Door Tests‒five Years’ Experience of Round-Robin Tests in the Czech Republic’. 9th Buildair Symposium (Germany).

Persily, A. 1982. ‘Repeatability and Accuracy of Pressurization Testing’. DOE/ASHRAE Conference ‘Thermal Performance of the Exterior Envelopes of Buildings II’ (Las Vegas, USA).

Prignon, Martin, Arnaud Dawans, and Geoffrey Van Moeseke. 2018. ‘Uncertainties in Airtightness Measurements: Regression Methods and Pressure Sequences’. 39th AIVC Conference (France).

Roberts, Ben, David Allinson, and Kevin Lomas. 2023. ‘Evaluating Methods for Estimating Whole House Air Infiltration Rates in Summer: Implications for Overheating and Indoor Air Quality’. International Journal of Building Pathology and Adaptation 41 (1): 45–72. https://doi.org/10.1108/IJBPA-06-2021-0085.

Rolfsmeier, S, K Vogel, and T Bolender. 2010. ‘Ringversuche Zu Luftdurchlässigkeitsmessungen Vom Fachverband Luftdichtheit Im Bauwesen e.V.’ 5th International BUILDAIR (Copenhagen, Denmark).