TECHNOLOGY WAYS FOR CHEMICAL INDUSTRY PRODUCTION MIGUAL
and OMAR Departamento
(Received for publication
1900 La Plata, Argentina
4 April 1985)
Abstract-A method is presented for selecting optimal technology ways for the production of a given amount of chemicals. We assume that a set of feedstocks is locally available and may be in limited quantity. Also, several alternative process technologies are available for transforming the feedstocks into final products. These technologies are characterized by technical coefficients of consumption of raw materials, chemicals, utilities, labor, etc., and by products produced, investment costs for different plant sizes, and other operation and maintenance costs. Also introduced are intermediate chemicals, which are produced and consumed within the system. The approach generates a network of chemical transformations, implying numerous alternatives. Additional alternatives are generated by allowing the possibility to import some feedstocks and intermediate or final products. We want to select the optimal one from among them. The objective function is to minimize the total cost of supplying the chemicals demanded by the market through local production or imports. This optimization problem could be classified as a nonlinear convex objective function constrained by linear inequalities. A theorem proves that the optimal solution must be in a vertex of a feasible solution polygon. This theorem permits construction of an algorithm that gives fast convergence to the optimal solution. It includes an appropriate linearization strategy for the objective function. By using a technology-economic database, software has been developed to state automatically the objective function and the constraints. Few specific data are needed, such as market demand to be satisfied, availability of raw materials and list of processes and chemicals not already included in the system. An analysis of possible uses of the method is presented, giving advantages and limitations. Cases included are optimal selection of alternative industrial investment projects, optimal exploitation of natural resources, comparison of alternatives of geographic localization of industrial projects, definition of industrial promotion regulations matching public and private objectives, etc. Use of shadow prices makes it possible to give appropriate weights to such political objectives as energy savings or efficient utilization of fossil feedstocks.
DISTILLATION AND DYNAMIC
GEORGE J. PROKOPAKIS of Chemical Engineering, University
TOWERS: STEADY-STATE SIMULATION and WARREN of Pennsylvania,
D. SEIDER Philadelphia, PA 19104, U.S.A.
(Received for publication 4 April 1985) Abstract-Azeotropic distillation towers have been in operation for many years, often for the dehydration of alcohol. However, it is only recently that mathematical models have been sufficient to accurately trace their steep fronts in concentration and temperature. These fronts are extremely sensitive to small differences in the concentrations of entrainer and water in the nearly pure bottoms product and to the boil-up rate. Furthermore, these variables must be carefully adjusted to prevent the models from predicting a vapor overhead stream that cannot be condensed to form two liquid phases, giving a high concentration of alcohol in the stripper bottoms and low recovery of alcohol in the azeotropic tower. In fact, it is so difficult to avoid operating conditions with low recovery of alcohol in the azeotropic tower that we found it necessary to develop a strategy to avoid this problem. This is the subject of our first paper and was accomplished by formulating a nonlinear programming problem with an appropriately selected objective function and inequality constraints. The results of this algorithm showed three regimes of operation, which were simultaneously reported by Magnussen et al. (1979), who found an instance of instability for one of the regimes. These observations suggested the dynamic simulation studies in our second paper, which reveal unusual responses to disturbances and support the observations of Magnussen and co-workers. For simulation of azeotropic towers, it is necessary to include the MESH (material balance, equilibrium, summation of mole fractions, heat balance) equations for each tray. These comprise a large set of stiff ordinary differential equations, which are integrated with semiimplicit or implicit integration formulas as described in the paper. (For further information, the reader may refer to AIChE J. 29, 49 (1983) and AIChE J. 29, 1017 (19831.)