Quantifying Variations in Project‐Cost Estimates

Abstract

Traditional methods of estimating project costs do not attempt to assess the magnitude of the variation inherent in the estimate. As a result, there is a risk that decisions on strategy selection will be based on a high degree of uncertainty. Alternative approaches to cost estimating are available and this paper describes the method adopted for a major sewerage project in Kuala Lumpur, Malaysia. The component‐cost ranges and probability values are defined by a panel of estimators in order to reduce personal bias, and the alternative strategy costs are simulated on a random number basis, the values for each component being selected randomly within the specified ranges. The computed median costs are therefore defined within stated reliability ranges enabling the investment decisions to be taken with a higher degree of confidence.