Quantification of loss and gain in performance using survey data: a study of earthquake-induced damage and restoration of residential buildings

This paper presents a method to assess and quantify changes in performance states of systems. The method is based on survey data containing performance states graded in a suitable scale. The changes in performance states are modelled using an association matrix obtained as a frequency table of measured performance states before and after an influential incident. Performance deviation, for both negative effects induced by a harmful incident and positive effects due to a constructive incident, is modelled. Relative change in performance is quantified by defining loss and gain matrices. Hadamard products of loss/gain matrices and the association matrix obtained from survey data are used to quantify performance deviation matrices. Finally, the overall change in performance is defined in terms of a performance deviation index, which is obtained by using the sum of the elements of performance deviation matrices. This index can be negative, zero or positive, and its value represents the degree of change in performance. A negative index indicates loss in performance; a zero index implies no measurable change, and a positive index means a gain in performance. An example application of the presented methodology for earthquake-induced damage and the subsequent restoration of residential buildings in South Iceland is presented. Copyright Springer Science+Business Media Dordrecht 2014