Effects of recycle ratios on process dynamics and operability of a whey ultrafiltration stage

There is limited information available to describe how the recycle ratios of the retentate and permeate streams can be used to counteract the effects of fluctuations in feed flowrate and composition on the flowrate and composition of the retentate stream from a whey ultrafiltration (UF) process. Based on dynamic modelling of a single membrane channel and a single-stage membrane module, it was found that the high retentate recycle ratio typically used in industrial UF of whey imposes slow dynamic responses and extensive oscillations on the flowrate and concentration of the retentate after a feedback controller is implemented. Even though the permeate recycle ratio can be used to control the specifications of the retentate, the high retentate recycle ratio used in typical whey UF processing places a limit on how fast the effects of fluctuations in feed flowrate and composition can be mitigated. Since a high retentate recycle ratio is required to improve the productivity of the whey UF process, a tradeoff must be obtained between the productivity and the time required to mitigate the effects of fluctuations in feed flowrate and composition.

This paper reports the effects of retentate and permeate recycle ratios on process dynamics for a single stage whey UF process. It provides important information about the operability of the stage, that is, how easily the flowrate and concentration of the retentate stream can be maintained within their desired specifications. With a high retentate reycle ratio of 0.96, there exists a trade-off between the productivity of the whey UF process and the speed with which the effects of the fluctuations in feed flowrate and composition can be mitigated by adjusting the permeate recycle ratio. For a given fluctuation of disturbance variables, an optimization based on the acceptable performance and the acceptable response time of the whey UF process is therefore required in order to determine the best nominal operating retentate and permeate recycle ratios. Although this study is focused on the operability of a single whey UF stage, the methodology presented in the study can be applied to multiple-stage industrial whey UF processes.