A multi-cell model for predicting ventilating air-flows is described in detail. Comparisons between prediction and measurements indicate that the method is capable of giving relatively high accuracy for a wide range of ventilation conditions. This applies not only to the prediction of whole-house air change rates, but also to the much more difficult problem of room ventilation rates.

The advantages and disadvantages of multi-cell and single-cell methods are discussed. It is argued that the multi-cell approach is potentially more accurate and more useful. Future development of the British Gas model is outlined.