IMPROVEMENT OF LIPASE PRODUCTIVITY IN BIOPROCESSES USING A STRUCTUREDMATHEMATICAL-MODEL

A simple structured mathematical model coupled with a methodology to e stimate state variables and parameters was developed and applied to Ca ndida rugosa batch and continuous lipase production. Process modelling was carried out using ACSL (Advanced Continuous Simulation Language). Once model parameters were determined, the whole model was validated, showing satisfactory results. In order to estimate the best strategy to improve lipase productivity, different simulations of batch, fed-ba tch and continuous cultures were made. The maximum enzyme productivity was predicted in continuous culture al moderate dilution rates (0.09 h(-1)) and substrate feeding concentration of 8 g l(-1). However, the highest predicted lipase activity was reached in fed-batch cultures wi th a prefixed substrate-feeding in order to maintain constant-at their optimal values-the relation substrate/biomass or the specific growth rate. The fed-batch process mode matched the simulation results. (C) 1 997 Elsevier Science B.V.