A monoclonal antibody process had to move from a 0.1-liter lab vessel to a 20,000-liter production tank, but media and experiments were far too expensive to test the way to an answer. We built a mechanistic model of what physically happens inside the tank, validated it against real lab data, then ran hundreds of production-scale scenarios virtually, without consuming media, losing batches, or spending time in the plant.
Impact for the client:
- A defensible design space, documented for the CMC filing
- The best operating design identified, plus viable backups for when the ideal isn’t achievable
- Infeasible designs ruled out on screen instead of on the floor
- A validated model that carries forward to the next scale-up, site, and product
