Process optimization of diafiltration with time-dependent water adding for albumin production

In this paper we investigate optimal time-dependent diluant adding in a diafiltration process that is designed to purify and to concentrate human albumin. Previous attempts considered arbitrarily constructed schemes, which may be, but need not to be, optimal. Employing methods of dynamic optimization we address two problems of optimal process operation: (1) the minimization of process time, and (2) the minimization of applied diluant volume. Control vector parameterization approach is applied considering different parameterized forms of the control function such as constant, linear, and N × piece-wise constant schemes. We found that traditionally used process operation, involving a pre-concentration and a constant-volume dilution mode step, is optimal in case of minimization of overall diluant consumption. Optimal control differs from the traditionally used operation types in case of process time minimization but the overall improvement is rather small for this specific application.

► Method for finding an optimal diafiltration process time is demonstrated.
► Method for finding an optimal diluant-saving strategy is demonstrated.
► Proposed optimization concept can be readily adopted for any diafiltration system.