The Dynamic multistage multicommodity process selection - location - production - allocation (DMMSLPA) model for centralised.

Abstract

The concern of the present work is to determine the optimal or near—optimal growth strategy for a process industry from the point of view of the Central Planner of a country. This involves determining optimal locations of plants; selection of a process technology at each location; planning capacity build-up over time for each plant and allocation of raw materials from sources to plants and of product from plants to markets. The envisaged contribution of this work is in modeling this complex problem, as well as in solving it. Such a model would be useful to the planner in making appropriate recommendations to concerned decision making bodies, such as the licensing authorities (e.g. the concerned Ministries). This will enable the latter to come up with appropriate policy decisions. The problem is modeled in an Operational Research framework. While the framework presented here may be applied to any process industry, it is developed in the context of the Indian sponge iron industry, mainly because it is a relatively virgin industry with substantial growth potential, so that some interesting results for policy makers may be obtained. Given the reality of the particular problem, substantial modifications to the general formulation have been possible. A large sized mixed (0-1) integer linear program results. The modeling has been based on an extensive tour of several sponge iron related organizations. This has helped in establishing the relevance of the problem as well as in refining the model. Also presented here is an extensive survey of literature in the area of locational analysis. The problem addressed in the present work is substantially different from earlier works in that it addresses all the five issues of process selection, location, capacity planning allocation of raw materials and allocation of product simultaneously and in a dynamic frame- work. An algorithm has been developed to solve the problem. This makes use of the specific problem structure, and provides a good solution in reasonable computational time. _ The algorithm developed has been used to solve the problem for various Governmental policies such as regional dispersal of plants and disallowing certain already developed regions of the country from being chosen. In addition an extensive sensitivity analysis has been performed, varying the levels of demand for sponge iron, as well as changing the prices of various raw materials and freight rates. Such a sensitivity analysis provides an indication of the robustness of various locations. Some interesting conclusions emerge, especially in the light of the licenses issued by the Government for Sponge iron manufacture. While the primary purpose of such a model is to identify good locations and associated action plans for growth of the industry, it could also be used to rule out bad or unpromising locations from consideration. The latter is a very important aspect, especially when there are vested interests involved in plant location and related decisions. The example of the sponge iron industry to iron industry to illustrate the model has enabled incorporation of the specific features of the industry into the general formulation. However the general framework could be used for any industry where capacity augmentation is done by addition of modules of plant. This is especially true for process industries. The framework can be used both at the national level for industrial planning and by the private investor contemplating growth or expansion of the industry.