Modelling Approaches for Bio-Manufacturing Operations - Measurement, Monitoring, Modelling, and Control of Bioprocesses - page 99

The approach was exemplified by using it to evaluate the impact of manufacturing
decisions upon technical and business outcomes in a case study focussing upon a
mammalian cell culture process delivering clinical trial material. The model was
used to track resource utilisation profiles and running costs incurred by different
manufacturing options, thus enabling the provision of adequate resources, material
balancing and cost calculation. Cost models were based upon conventional chem-
ical engineering-style calculations, but they also accounted for the additional
expenditure incurred by implementing stringent cGMP rules. The manufacturing
process was represented as a set of unit operation and ancillary tasks in a dynamic,
discrete event simulation environment, for which resource pools were available for
entities such as capital equipment, membranes, chromatography columns, bags,
buffers and labour. Values for these were initialised at the start of the simulation and
updated as appropriate during the run whenever new stocks were purchased or
prepared. Hence the financial and time-related impacts of operating in a resource-
constrained environment could be evaluated. This meant that batches were pro-
cessed only if adequate resources were available for that to occur. Otherwise, a
batch was forced to wait until resources become available in sufficient quantities.
This represents the same situation that would occur in a real plant and can thus
enable identification of bottlenecks. The simulation contained pre-programmed
blocks that represented specific tasks such as a culture or purification operation. The
blocks were cloned to a workspace and then connected together to simulate the
sequence of steps in an entire process flowsheet, with the generic blocks customised
for specific requirements of the relevant step for its location in the process. A batch
of process material is represented by an 'item' i.e. a computer-generated entity that
is 'loaded' with stream properties and which is modified as it passes from one
simulated unit operation to the next. Items were passed through the model, updating
cost, resource and material balance property data such as volume, titre, impurity
level etc. over the course of the simulation.
Example 2
Models are also useful when deciding whether or not to conduct a plant retrofit,
where typical commercial questions might focus upon the timing of the retrofit and
also for determining whether the altered final process merits the production
downtime and retrofit cost. Mustafa et al. [ 30 ] describe a model which evaluated
the economic impacts of an expanded bed retrofit in place of a centrifuge and
packed bed-based protein separation. The study sought to simulate and trade off a
higher product yield on one hand with the costs of plant shutdown (loss of pro-
duction), capital investment and process re-validation on the other. The method
enabled strategic, process and economic evaluation of the two options. In this
specific case, the results indicated that the re-validation cost and the timing of the
retrofit were important, as was the effective downtime cost associated with lost
production. In a related study, Mustafa et al. [ 29 ] developed a software method to
evaluate the business and process aspects of two different manufacturing flow-
sheets employing packed bed chromatography in one case and expanded bed
adsorption in the other. The method assessed the advantages of expanded bed in
terms of capital cost reduction and higher product yield compared with the higher