An investigation of alternative methods for determining envelope airtightness from suite-based testing in multi-unit residential buildings

Abstract

Blower door testing is commonly used to characterize exterior façade air leakage in buildings. This characterization is needed to improve the accuracy of energy models and certify building performance. Performing these tests in large buildings requires air flow rates that are sometimes impractical to achieve using available equipment, which necessitates testing in individual zones within those buildings. However, when performing these tests in individual zones, the conventional whole-suite tests are unable to isolate the required envelope flow characteristics and pressure neutralization tests are, in most cases, too logistically difficult to perform, which necessitates an investigation into alternative methods. In this study, two methods developed for use in row housing were adapted to a high-rise multi-unit residential building (MURB) context in which the test suite is surrounded by adjacent suites: The Love and Passmore method, and the Lstiburek method. The efficacy of these adapted techniques was assessed in a single suite located in a high-rise MURB that followed typical North American construction habits. To carry out these assessments, a combination of field data and calibrated CONTAM simulations are used. For the case study building, the Love and Passmore method exhibits exterior boundary characterization errors of 108%, and is inappropriate for this characterization in high-rise MURBs. Alternatively, the Lstiburek method exhibits an exterior boundary characterization error of 0.2%, but additional analysis showed the Lstiburek method exhibits greater sensitivity to measurement noise than the pressure neutralization method. As a result of this increased noise sensitivity, and because of the similar labour and equipment requirements, it is therefore also not recommended as a replacement for the pressure neutralization method. It is important to note that buildings constructed using different details and techniques may exhibit different results, which is a limitation of this study. Details related to these tests, including a discussion of potential shortcomings and strengths, are presented in this paper.