Improving cost contingency estimation for infrastructure projects based on the impact of cost overrun causes of similar projects – illustration using Saudi Arabia survey data

Abstract

Cost overrun on infrastructure projects is widespread and represents significant financial risks to stakeholders. Despite the progress made, there is a lack of research on the estimation of project cost contingencies, thus the main objectives is by taking into consideration the actual impact of the cost risk of similar projects, this thesis developed and validated a cost contingency estimation model. A survey of 160 project managers of infrastructure projects in Saudi Arabia was conducted to elicit the cost overrun causes. After cluster analysis, the causes were reduced to four dimensions: scope changes, market and regulatory, inadequate planning and control, and unforeseen circumstances. These four dimensions were then used to develop a risk-based cost contingency estimation model (RBCCEM) to improve the accuracy of cost forecasting and then validated using a bootstrapping approach. The accuracy of cost estimation measures were used to compare RBCCEM with fixed cost contingency (10%), reference class forecasting (RCF P50 & P90), and hybrid (combination of RBCCEM & RCF P50). The contributions of this study are; first, it contributes to the research on the cost overrun causes by updating the existing knowledge of cost overrun causes to better understand the complexity of these causes and to manage risk effectively. Second, the results show that RBCCEM produces contingency estimates that are more consistent and brings project costs closer to actual costs than alternative approaches, such as RCF and hybrid. Moreover, this study contributes to knowledge by introducing a new way of modelling the cost contingency of projects using the cost overrun causes of similar infrastructure projects and providing empirical comparisons of cost overrun outcomes with those using alternative estimation methods. Together, this study provides important theoretical and managerial insights and opens the way for more research within the cost overrun context, particularly with regard to cost contingency estimation in large projects.