What Makes Scale-Up of Industrial Biotechnology so Difficult?

by Mark Warner (Warner Advisors LLC) Two questions have dominated the responses received after my Lessons Learned series on commercializing industrial biotechnology (here) and the deep-dive into the industrial biotechnology commercialization process (here), the questions are what makes scale-up of industrial biotechnology so difficult and of course how can risk be reduced in the process? Looking back at the materials, there are a few areas that need to be expanded in greater detail to answer those questions.

From the Lessons Learned series, the third lesson “there is no substitute for a fully integrated pilot process” generated a lot of discussion on why that is so. Many in our advanced biotechnology industry come from the chemical or petrochemical industries, which are rooted in process modeling as their primary scale-up tool. This is significantly different than advanced biotechnology, where scale-up is based on extended pilot operations. The cost and timeline of building an integrated pilot, or demonstration scale plant, challenges ventures attempting to bypass these steps, which can end badly. Let me focus on what makes biotechnology processes unique and why pilot testing is so critical to a successful scale-up.

First, let’s understand how traditional chemical processes are scaled-up by modeling as a comparison.

Modeling did not completely replace the need for piloting, but often limited the scope of pilot testing to verification of key parameters.

Now, let’s compare that to industrial fermentation based processes. Industrial fermentation typically starts with feedstocks that are less pure and more complicated from a reaction standpoint than a traditional chemical reaction. Anyone who has seen the massive wall posters of metabolic pathways in very small font knows what I am talking about. It is generally not practical to model the entirety of the individual reactions (and competing side reactions), but rather only practical to generate an average rate equation for the overall process. While this can be used to represent the process from a “macro” perspective, it will not accurately predict the minor constituents in the fermentation broth that can impact both the fermentation and recovery productivity. This example is specific to fermentation, but the principle equally applies to other bio-based processes.

Given this inability to accurately model biotechnology processes, pilot and demonstration plant operation is the only reliable method to generate the information needed to scale and design equipment. This is why integrated pilot operation is so critical to project success. Here are a handful of my lesson’s learned specific to scale up of biologic processes:

The liberal media – A reference to the fermentation media, of course. ...

The “other” problem – as discussed above, it is not practical in a commercial biotechnology process predict all compounds generated during the fermentation, or that come along with the feedstock. Typical chemical analysis used in engineering scale-up will identify key compounds, but then everything else that cannot be identified gets lumped into a category of “other”, often referred to during the design process as “OS” or “other stuff”. ...

Don’t push the rope – hopefully we all learned at an early age that you cannot push a rope, you need to pull with it. The same principle applies to process scale-up. You need to first determine what your commercial scale facility will look like conceptually and use the pilot operation to prove out key parameters needed to build the process (i.e., “pull” the information needed from the pilot). ...