Methodology for the survival analysis of urban building stocks

A new methodology is presented for estimating age distribution and survival functions of an urban building stock. This allows for a random sample of the undemolished stock together with a complete inventory count of the demolished part to create a Kaplan-Meier estimator for the survival function. This method can be applied to any building stock with difficult access to data. A demonstration of this method is applied to a sample of German urban stock and used to illustrate the data-collection process, the survival analysis and an estimate of the dynamics. Findings from this sample indicate a slower dynamic in the residential part in comparison with the non-residential part, which is related to seemingly different survival functions. Another finding is that the percentage of undemolished buildings seems much lower for older buildings than for the younger structures. Through a Weibull fit, little future change in this survival pattern is estimated. The estimates obtained in this preliminary study challenge many assumptions of service life, economic life and effective life time. It is found that the effective life time of building stocks is much higher than generally assumed, and possibly independent of their age. For recent buildings, the implication is whether the existing service life or economic life approaches are adequate to understand and plan for a potentially higher survival rate.