Modelling probability distributions of facilities management costs for an NHS acute care hospital building

Whole Life Cycle Costing (WLCC) provides a valuable insight into the economic efficiency of constructed facilities. It has however, been criticized by practitioners and academics alike for producing forecasts of a high-risk nature, a consequence of the large number of assumptions inherent in the modelling process. Most WLCC systems use deterministic assumptions but a stochastic approach to modelling the assumptions will achieve a better representation of the likely costs. Monte Carlo and Latin Hypercube simulation can be used to create forecasts of the whole life costs but the accuracy of these are strongly correlated to the quality of the input probability distributions, hence the purpose of this paper is to present a methodology for the definition of probability distributions that best represent the facilities management costs in an acute care NHS hospital building. The data used in this research were obtained from National Health Service (NHS) Estates on the Facilities Management (FM) costs of over 450 acute care NHS Trusts in England and Wales. The data was analysed to obtain the parameters of the theoretical distributions that best describe the FM costs for a local NHS Acute Care teaching hospital building. The distributions were then validated using various goodness-of-fit techniques. The result of this work might then be used as stochastic assumptions in the modelling of WLCC. The paper also discusses some issues of accuracy in distribution fitting, particularly the class interval rule and its effect on the quality of the results, and the selection of an appropriate goodness-of-fit test.