Generalised cost benefit analysis for large infrastructure projects: the case of the Channel Tunnel rail link

Even in the absence of specific evidence it is surely unthinkable to imagine Napolean Bonaparte himself, when considering the potential benefits of invading England, failing to weigh against them the costs of building a tunnel under La Manche. In the light of the considerable literature of cost benefit analysis (CBA) and by gathering together several analyses that have been performed on the Channel Tunnel in more recent times, this thesis demonstrates that a comprehensive ‘big picture' CBA framework is missing, and then provides one: a new methodology, represented in Data Flow Diagram (DFD) form, for the generalised cost benefit analysis of large infrastructure projects. Now, almost 200 years after Napoleon’s death, the world spends nearly $40 trillion (almost £24 trillion) on constructing and upgrading roads, railways, power plants and other infrastructures so as to keep up with demand. Such huge investments require realistic plans, clear deadlines and tight budgets for the best allocation of resources but cost overruns and delays are all too frequent. What exactly causes the cost overruns or delay is difficult to identify, but better planning and focus on the accuracy of the appraisal is one way to reduce the problems. The new model presented herein should prove useful in assisting decision-makers for all large infrastructure projects, and as a foundation for developing an appropriate decision making tool for different categories of projects too, based upon the experience of the Channel Tunnel. This, after all, was a project that the Government had financed through the private sector with high investment risks and significant degrees of uncertainty and had then put out for construction and operation in two different tenders. In spite of never being viable for the private sector, the Eurotunnel Company looks in a better shape now and the tunnel is working properly, but the construction cost and deadlines were clearly exceeded.