Clarkson, Andrew Irvin; (1995) A Study of the Use of Process Simulation and Pilot-Scale Verification Trials for the Design of Bioprocesses. Doctoral thesis (Ph.D), UCL (University College London).

Preview

Text 10017237.pdf Download (9MB) | Preview

Abstract

This thesis examines the use of process simulation tools and pilot-scale verification trials for the design of efficient bioprocesses. The use of process simulation tools requires the development of predictive, robust unit operation models were the models are used for the calculation of mass and energy balances, and ultimately economic analysis and optimisation. Verification trials are employed to assess how the model compares to reality. Models describing key unit operations such as protein precipitation and centrifugation are often very simplistic and do not take into account the added complications that biological materials present, such as hindered settling at high solids concentrations in centrifuges and susceptibility to shear forces. The generation of useful engineering models, testing by comparison with real process data and their use in design are covered in this work. Two models have been developed in this thesis; batch protein precipitation and disc-stack centrifugation. The batch protein precipitation model calculates the enzyme and total protein solubilities upon precipitant addition, together with the precipitate phase particle size distribution and enables the effects of precipitant concentration and batch ageing conditions to be predicted. A mass and activity balance is then completed around the unit operation. The disc-stack centrifuge model is capable of predicting the separation of a range of biological materials including whole yeast cell, cell debris and shear-sensitive precipitate particle suspensions. A centrifuge feedzone breakage model has also been developed, which accounts for the shear breakage of precipitate particle suspensions that occurs in the feedzone of the centrifuge. The capacity to predict the much finer particle size distribution which enters the active disc stack where particle separation occurs enables accurate predictions of separation performance to be made. The centrifuge model also enables mass and activity balances to be completed around the unit operation. The models have been linked together so that they predict mass balances around a complete process sequence for the isolation of an intracellular yeast enzyme. Pilot-scale process verification trials have been conducted for the process sequence. The simulations and experimental verification trials provide total protein concentration and ADH activity data for all streams throughout the process. In the small scale trials DNA and cell debris concentration were also measured and simulated. Results show that the simulated results follow the trends of the experimental data extremely well. The utility of verification trials in indicating where further modelling is required, such as the centrifuge feedzone is demonstrated. Process perturbation trials have been used to show that the models can be used outside their normal operating conditions. The models developed for the yeast ADH test bed have also been tested for a process for the isolation of β-galactosidase from Escherichia coli. Results have shown that only limited experimental data is required to calculate the parameters used in the models to effect an accurate simulation. This thesis demonstrates the use of a combination of modelling and experimental verification trials for the design of bioprocesses. Recommendations for future work in the areas of further model improvement and development of other unit operation models, investigation of simulation frameworks, incorporation of on-line control techniques and the use of "what-if" studies are made.

Download activity - last month

Export as JSONXMLCSV

Download activity - last 12 months

Export as JSONCSVXML

Downloads by country - last 12 months

Export as XMLCSVJSON

Archive Staff Only

View Item