Monte Carlo simulation of construction costs using subjective data

Abstract

This paper critically examines the problems arising from the assumptions of independence and triangular distribution in the risk analysis of construction costs. These two assumptions have been widely adopted by researchers and practitioners alike. However, they lead to bias in the analysis. With the aid of a spreadsheet and a risk analysis add-on program, the nature of the resulting bias can be illustrated using both empirical and simulated data. Various possible solutions which seek to reduce the bias are explored. Their practicality under real life constraints is examined. A tentative practical solution is proposed and evaluated against the existing approach. The method and logic of the solution are also discussed. The approach is applied to two sets of data to illustrate its use and to test the reasonableness of the approach.

Keywords:

• Costs
• dependence
• spreadsheet
• Monte Carlo simulation
• probabilistic cost estimate