Fermentation Simulation Model
Manufacturing an antibiotic in a production scale fermenter. A big production fed batch fermenter was making an antibiotic. The problem was to understand what controlled the process so that it could be made more predictable and increase productivity. Working with a microbiologist, who was the expert on the biology of the particular organism, I developed a dynamic simulation model which represented the stoichiometry of the metabolic pathways in a simplified form of reaction kinetics, but always recognising the need to preserve mass and energy balances. The model generated time-series for variables which could be compared with those gathered from the operation of the plant. Furthermore, of course, the model exposed what were the inaccessible internal conditions underlying those accessible measurements.
Right from the beginning, the microbiologist said it behaved like a fungal organism, but not initially the one we were dealing with. As the model was refined, it became more and more accurate in its predictions of what was observed. It also became more complex, close to the boggle limit. The work also caused a rethink of the methods of assay, since the means now existed to make an accounting of all the materials on a conservative basis.
One day, the plant operators made a mistake and started a batch with a hundred times smaller initial inoculum of live organism than they should have done. They asked the microbiologist for advice; stop and waste all the materials or what? With great faith in the model the microbiologist ran the simulation with the starting conditions far different from any used to develop it. Having seen that the model predicted a normal outcome but four days late, he told them to hang on. They did. The model was right and it was paid for! (A description of this work was published. See list of publications).

If you have a problem in the behaviour of a fermentation brewing or similar process and would like to discuss a fresh look at the process, bringing quantitative methods to bear on it, then contact me. I am always ready to give a little time to discuss a new puzzle, in confidence, of course. We'll only worry about fees when we have some defined work. I can be flexible about how I work with you.