| Abstract: | The Net Present Value (NPV) approach is often used as a tool to rank projects and it uses a total order approach. This paper applies a new method developed by De Baets and Van de Walle (1994, 1995) and which we will call ‘fuzzy multicriteria analysis’ (FMCA). The main objective of this new method is to weaken the total order into a fuzzy quasi-order in which incomparability between alternatives is possible. Although other methods such as Roy’s outranking method also deal with the issue of incomparability, there are important methodological differences between FMCA and outranking. We consider first an application of FMCA to a general NPV ranking problem. Some properties are derived which may be useful and we then describe an application to a practical example in which three projects yield the same NPV outcome. Different alpha-cuts imply different crisp quasi-orders. The use of FMCA shows that, in this sample case, though the projects have equal NPVs, a ranking may still be possible. Copyright © 1998 John Wiley & Sons, Ltd. |
