Prioritization of Korean Rail Investments Utilizing Extended Capital Budgeting Problem

Constructing a railroad network includes a process of selecting projects that will form the railroad network based on efficiency measures. Such a selection process has many difficulties, since it is usually made under budget constraints and those projects are not mutually independent. In many applications, however, Capital Budgeting Problem (CBP) models have been used under an assumption that alternative projects are mutually independent to maximize NPV. Accordingly, the configuration of long-term railroad networks in application of these models will likely to produce differences according to the interdependency between projects. In order to overcome this shortcoming of CBP, extended forms of CBP have been utilized. This paper analyzes the difference in long-term railroad network configurations from applications of CBP and some Extended CBP under budget constraints for Korea’s actual case. The resulting long-term network configurations from CBP and Extended CBP are compared that from the network design problem which explicitly considers projects interdependency. Configurations from CBP, Extended CBP and NDP application show noticeable differences. The desirability of these configurations is compared with additional measure of effectiveness. CPU times and computing resources relating to the creation of models and the solution of problems are also discussed.