Optimization of Electrolytic Cells Having Large Numbers of Variables

A generalized reduced gradient method was used to optimize profit for a model of a chlor-alkali cell which included 42 variables, 37 equality constraints, and 2 inequality constraints. Many of the equations were nonlinear. The technique optimized all variables simultaneously. In addition, Lagrangian multipliers were used to determine sensitivity of the objective function with respect to cell constants in the constraints. In this manner, sensitivity of the optimal solution to design and operating variables was explored. The methodology developed in this study may be applied to other electrolytic cells for which algebraic mathematical models exist.