Current building pressurization test methods considerably reduce the scope of validity of the test results when stack effect is significant without clear justification. This work attempts to fill this gap based on physical modelling and numerical simulations. In steady and no-wind conditions, it gives the error in the test result with leaks discretely distributed over the building height assuming the flow exponent is the same for all leaks. It also gives an analytical expression of the error when the building can be modelled with two leaks, one at the bottom and one at the top. The results show that the maximum absolute value of the error in the two-leak case has a very high probability to be the upper limit of the absolute value of the error obtained with any leakage distribution of more than two leaks. Consequently, one can use the two-leak model to obtain the minimum induced pressure to generate to contain the error below a given limit. For example, using an induced pressure greater than twice the stack pressure should be sufficient to contain the error in absolute value below about 5%; the error should be below 14% if the induced pressure is greater than the stack pressure. In turn, this paper gives recommendations for testing together with tables and charts to be used by field technicians.

当前的建筑物加压测试方法在烟囱效应显着而没有明确理由的情况下，大大降低了测试结果的有效性范围。这项工作试图基于物理建模和数值模拟来填补这一空白。在稳定和无风的情况下，如果所有泄漏的流量指数均相同，则泄漏结果会在整个建筑物高度上离散分布，从而给测试结果带来误差。当可以对建筑物进行两次泄漏建模时，它也会给出错误的解析表达式，一次泄漏在底部，一次泄漏在顶部。结果表明，在两次泄漏的情况下，最大误差绝对值具有很高的概率，成为任何大于两个泄漏的泄漏分布所获得的误差绝对值的上限。所以，可以使用两泄漏模型来获得最小诱导压力，以产生以将误差限制在给定极限以下。例如，使用大于两倍烟囱压力的诱导压力，应足以将绝对值的误差控制在约5％以下；如果诱导压力大于烟囱压力，则误差应低于14％。反过来，本文提供了测试建议以及供现场技术人员使用的表格和图表。