How BioPAT^® Supports Risk Mitigation

PAT tools generate process understanding and layout the design space around the safe production of efficacious products. This is achieved by consistent modelling and process execution by use of reliable online sensors and analytics, embedded in advanced automation routines. Therefore, knowing where the hazard lies, measures can be taken to avoid the risk before problems occur.

Every drawn sample that opens a closed vessel to the outside environment is a potential contamination risk. Even if that risk is minimized by novel technology and definitive protocols, a small chance still exists. Whilst data drives informed decisions and process control, the more sample points acquired, the higher the probability of a contamination failure. Thus, minimizing the sampling by using inline measurement techniques eliminates that risk.

Online monitoring of biomass data fed into a multivariate trajectory process model gives real-time plotting of the commercial batch where deviations from normal can be instantly reported. If this deviation is a contamination with no solution the quick decision can be made to stop bioreactor and start a new run. As the investigation starts, the combination of post contamination data from environmental swabs and the GMP compliant online data collection allows the rapid eliminating of possible route causes streamlining the investigation time. The investigation time as well as the down time of sequential equipment for reproducing a batch can be conducted faster in order to get back on schedule.

Starting the production of a bioprocess right-first-time is vital, missing components or incorrect concentrations can result in deviations and even failure. A system that scans the visible to infrared spectrum obtaining a finger print of the starting point can alert the user of minor deviations allowing time for corrective action and reducing the risk of losing a batch.

One of the key challenges to running processes today is managing the large amounts of data. Not only are people concerned with securing data from their bioreactor controllers but also all of the secondary and tertiary data sources such as offline metabolite analyzers, HPLC and manual tests.

Through the use of BioPAT^® MFCS as a central SCADA platform, users are able to automatically bring in live data from their control towers as well as 3rd party hardware using an OPC^® connection. This minimizes the number of places that critical batch data is stored. By having the data in a central location, users are able to easily and quickly compare batch data in both univariate and multivariate spaces with the use of the BioPAT^® SIMCA^® software. BioPAT^® MFCS also has a Replication Manager which automatically backs up all system configuration and batch data to a user defined network drive, providing robust data security and peace of mind.

BioPAT^® MFCS with S88 recipe control enables users to automate routine feeds such as media, glucose, ethanol or methanol. Through the use of online calculations and logic gates it is able to calculate the amount of feed to add based on an input and automatically add that volume via a substrate pump or gravimetric control. This reduces the number of manual steps required by the user.

Feeding the cells in a bioreactor is essential. Commonly, glucose additions supply cells with an adequate carbon energy source during growth and production phases. The process of sampling, calculating the feed and performing the manual addition is time consuming especially with multiple vessels running in parallel. Using MFCS-S88 Recipe control, the feed addition can be automated to the point where a user only needs to enter the off-line value and system performs the calculations and adds the correct fed amount. Thus, by using recipe control the number of manual steps can be reduced allowing the operators to perform other tasks. Further, by using automated sampling on online measurement, feeds can be started when users are not on-site reducing the need for after hours and weekend work.

For a typical E-coli process, the exponential phase begins after the lag phase when the starting nutrient component is depleted. Normally, using the DO signal as an indicator the operator manually takes samples to monitor when to start the exponential feed. The sample result is used to verify when this critical point occurs. This is tedious, and a very manual process. The classic DO sensor cannot be reliably used to automate this start due to the noisy control signal.

By combining the BioPAT^® Trace, along with the advanced control function of BioPAT^® MFCS, the real-time Glucose or Alcohol signal is fed to the controller. With this automated, timely and accurate signal, the feed start can be programmed into a recipe to automatically start without operator intervention.