Economic pressures are forcing the chemical industry to decrease production of large amounts of commodity chemicals and to move to the production of smaller amounts of fine or specialty chemicals. These require different modes of operation, an important one being batch processing. Processes related to the production of specialty chemicals are characterized by significant price differences between the reactants and the products. Often a batch does not result in making a product of the desired quality and has to be discarded. Because the added value of the product in relationship to the reactants is so high, significant economic benefits can result from a general purpose modeling and optimization strategy for batch processes that decreases the number of batches that do not meet the product quality constraints. In contrast to continuous processes, batch processes are characterized by transient and nonlinear behavior. The linear approximate models, useful for continuous processes, can not be applied simply because there does not exist a steady state about which the process can be linearized. The use of batch reactor models that predict the time evolution of the concentration changes of the most important reactants and products offers the possibility for a much better understanding and a significant increase in the optimization of the process. The PI's plan to use "Tendency" or "Gray" models, because they do not depend upon the availability of a detailed and predictive model of the process ("White" model) as such models are very rarely available. In this technique an initial model is postulated (often a very crude one) and is successively updated and improved using experimental results. The equipment requested, a batch reactor, a gas chromotograph (to measure concentrations) and computer equipment for on line calculations and control of the process, are appropriate for the project. The PIs are respected researchers. Digital Equipment Corporation offers a 30% discount on equipment awarded under NSF grants. The PI has obtained an estimate of $40,330 for the computer equipment, therefore using the 30% discount and 1/3 cost share by Lehigh University an award of $45,941 out of FY87 funds is recommended.
