Toward more reliable cost estimations for asset replacement projects

Abstract

The majority of existing infrastructure assets are ageing, and sooner or later, they will have reached the end of their service life. As a result, many asset owners are confronted with large-scale replacement projects in the coming decades. Annually, they allocate a significant number of resources, but these budgets are never adequate for executing all maintenance and replacement activities for their respective assets. It can be said that the same pattern of consequent cost overruns and budget exceedance that is observed in new infrastructure projects over the years also exists for life-cycle budgets. Based on the literature, some scholars investigate the various factors that can lead to budget overruns in new infrastructure projects. Accordingly, others are trying to deal with this issue for new infrastructure projects by proposing different models. However, little information about this issue exists for replacement projects. This research attempts to fill this gap by identifying the factors contributing to the cost overrun problem in replacement projects and providing a model to deal with them. The first part of the research contains a theoretical and practical analysis. In the first, a literature review is conducted to determine the cost overrun factors other scholars have spotted for new infrastructure projects and examine their applicability to the replacements. Once these factors are identified, then a practical examination of these factors is conducted by examining multiple case studies concerning past replacement projects. In that part, all the available cost assessments between different cost estimation phases are revied, and valuable findings are extracted, leading to the development of a conceptual model capable of explaining the cost overrun issue in reconstruction projects. Based on that, two essential characteristics of replacement projects influence the accuracy of their cost assessments. The first is associated with the fact that their budgets are usually structured based on preliminary cost estimations with a high error range and low reliability. The second is related to the fact that usually, replacement budgets are structured many years before the actual initiation of works. During this long-time frame, many changes can happen, such as price (de) escalation, which can significantly influence the initially estimated cost. All this information extracted from this twofold analysis is used in the second part of this research, which includes the process of developing and validating a model capable of dealing with these identified factors. The proposed model consists of two building blocks, each intended to target the factors that originate from the two specific characteristics of replacement budgets. More particularly, the first building block suggests a probabilistic model that can improve the current ways of preliminary estimating costs. The second intends to incorporate the price escalation effect into the estimation by using historical prices and a stochastic model. Based on the validation process, the proposed model could provide more accurate results than the current practices but also a probable range of costs and a projection of them in the future. Consequently, its use can contribute to reducing the cost overruns in reconstruction projects.