Cost management for underground infrastructure projects : a case study on cost increase and its causes

Abstract: Extensive investments in infrastructure for transportation are currently being carried out in Sweden. A substantial part consists of underground road- and railway projects. The aim of this thesis is to create a foundation upon which improvement can take place, with respect to cost management in the process for underground road- and railway projects. All projects are, to some extent, associated with uncertainties that fall into three groups: Risk, Inherent- and Inflicted uncertainties. The projects also are associated with two groups of unique features, namely geology and closed room. These features are associated with greater uncertainties, which justify that it is more difficult to estimate the cost in underground road- and railway projects than in those aboveground. The cost development has been studied for a number of Swedish road- and railway projects. All projects are associated with cost increase, and the largest occurs during planning. These results are in agreement with those from other international studies. Additional results from the Swedish projects are: (1) The unit price for tunnels and contracted prices for different tunneling works are associated with large variations; and (2) The overall process is stable, and that the outcome is predictable within wide statistically limits. These results provide a basis for improved cost estimates in the future. Currently, there is no systematic follow-up of the accuracy in early estimates, although, early cost estimates are based on cost from previous projects. Consequently, the quality and the uncertainties in the cost-estimates are unknown. A detailed case study reveals that the vast majority of cost increases occur elsewhere than within the unique feature for underground projects, namely as indirect and financial cost. The unique features have contributed to a minor, but yet substantial increase in the cost, of which the greatest increase originates from water treatment, reinforcement and tunnel safety. The mechanisms for the two former groups are evaluated, and mitigations are suggested. It is concluded that large cost increases are generated from optimistic prognoses, due to cognitive bias, and cautious choices in the mapping. The mechanisms behind water treatment cannot be investigated at this stage. Several of the causes for cost increase that have been identified in this thesis may be regarded as example of inflicted uncertainties. From the perspective of the client on road- and railway projects, activities within one project is very similar to those in others projects. Because a project organisation is temporary and unique, its abilities to improve are generally restricted to the individual project. To achieve a lasting improvement, experience from individual projects must be transferred to a more stable organisation, such as the parental organisations of the involved actors. There is also a need to improve cost management of indirect- and financial costs.