Blog post by Bill Lemmon, Automation Engineer at Superior Controls
With nearly a decade in the automation industry but no biotech experience, I joined Superior Controls’ Albany, NY automation team with more than a little trepidation. GAMP, GMP, and WFI, were just jumbled letters to me. Bob Patrick, VP of Engineering at Superior, however, informed me that my first 3 weeks would be spent at the company’s Seabrook, NH headquarters where I would participate in intense training on detailed validation, project management, and the proper implementation of biotech automation projects.
I spent 3 weeks reading through the “Superior Navigator,” an in-house, web-based, self-paced tutorial containing thousands of pages, pictures, videos, and quizzes to ensure that I understood the basics of Good Automated Manufacturing Practice (GAMP) required to manufacture FDA-regulated therapeutics. As part of this training, I developed a PLC program and operator interface to monitor and control a batch reactor. Prior to programming the system, I had to develop the design documents in accordance with GAMP 5.0 all before writing any code. Later, I executed a Software Factory Acceptance Test using Good Documentation Practice (GDP).
Next, Bob Patrick asked me to attend a 4-day intense Fundamentals of Bioprocessing Engineering class at Worcester Polytechnic Institute. The course, run by WPI’s Biomanufacturing Education and Training Center, was half lecture and half hands-on lab work using equipment that one would expect to see in current manufacturing facilities. What follows is a summary of what I learned in this fascinating course:
The first day, after a quick introduction of the instructor and participants, we dove right into the process that would take a simple 1 milliliter of frozen seed and grow it to manufacturing volumes. Growth of the desired product could happen in roller bottles, spinner flasks or wave bioreactors. This lead to the introduction to the phases of the cell growth process: the lag phase, log (exponential) phase, stationary phase and the death phase. We began with small vials then transferred the material to larger containers because cell growth would lead to production volumes of 5000 liters and more.
On a quick walk to the lab, we saw WPI’s equipment for growing the 1 ml seed in 250 ml, 500 ml and 1 L flasks with larger structures that could grow in 5 liter, 10 liter, 20 liter, all the way to 100 liter reactors. While in the lab, two ways were demonstrated for counting the number of cells grown to ensure a transfer happened at the correct time. The first method consisted of diluting a sample from any of the containers and looking at it under a microscope and manually counting the cells in a few grids. This is something that could be done in 15 minutes if you have years of training and experience. The second method was much easier and used an automated system that after a minute of processing would give the cell count with live and dead cells as well as produce images of all 31 sample counts.
Back in the classroom, we reviewed the equipment that was used to build a bioreactor, from the pneumatic and solenoid valves to pH and conductivity probes. Then once more in the lab, it was shown how the devices connect to the reactor for both stainless steel reactors and disposable reactors.
The next day the topic was how to remove cells and particulates from the media of the final stage. The discussion began with disk centrifuges and led to the lab where we put one together. I had worked with disk centrifuges while serving on a submarine that was built in the 1960s and I could see that technology hasn’t changed too much, of course back then I used it to separate oil and water, but the theory was still the same. Because centrifuges are not 100% effective, a series of 10 μm to 0.1 μm filters are next in series to ensure everything has been removed. Further discussion continued with PLC operations specifically directed towards proportional integral derivative (PID) operations and then we had instruction on factors to consider when working with the various types of cells utilized in biopharmaceutical processing.
Day three was about amino acids, the proteins they produce, and how to remove the desired or undesired component with chromatography in the downstream process. In the lab, we utilized proteins from jellyfish to view the bioluminescence under a black light as the proteins traveled through a column filter. When performing the load and wash steps, the group was able to see the protein enter and bind to the filter. We were then able to perform the elute process and see the protein pass through the filter and fill up the test tubes. This lab was also performed the previous week by WPI’s biomanufacturing master’s degree class.
Our last day was focused on critical utilities for bio-manufacturing facilities or more specifically, water. The water coming into the facility could have suspended solids, dissolved solids (ionized, non-ionized and organic), colloidal materials, dissolved gases, and bacteria. To remove these unwanted contaminants particle filters, membrane filters, ultrafilters, reverse osmosis, carbon filtration, ultraviolet disinfection, and deionization are used in combination with lots of sampling and documentation to ensure it meets the gold standard of distilled water. This process allows purified water and WFI (Water for Injection) to be used in the facility.
To conclude, I came out of the class with a much better understanding and appreciation for the FDA regulations required in biotech and pharmaceutical manufacturing. I feel I’m a better project engineer and understand the steps required to properly implement a custom biotech automation project. Bob Patrick tells me that Superior Controls currently has 36 fixed-price automation projects in-house and is anticipating more. He recently assigned me to a biotech project here in Albany that I now feel very comfortableworking on.