Common Mistakes in Chemical Process Optimization

One of the main points that we have to be aware of during our daily work is to try to avoid mistakes during chemical process optimization. It is very typical to waste time for the company when we are undertaking and running processes in the lab that really will not be representative in the next step. One of the main points that we have to be aware of when we’re working in a lab environment is the small equipment. When we have very small equipment, the fluid rating regime that we use may be laminar because the dimension is very small; hence, the Reynolds number will be low (i.e., a low Reynolds number is a laminar regime). When we go to production, the equipment and geometry are much larger; hence, the Reynolds number will be higher than the critical number and we will be in a turbulent regime.
These two regimes, laminar and turbulent, are really different kinds of interactions among materials. In the laminar regime, the interaction between two materials that will give us a new product and a new process should be well controlled. If we are having very good success in the lab in the laminar regime where the interaction between materials is very low, then when we go to production equipment (if large) we will be in turbulent regime and the interactions between materials is a completely different mechanism, very high, and very random one. Each regime generates completely different results. In other words, what we did in the lab is not representative as the interaction of materials will be completely different in a different regime. Instead, tubulin will be in laminar or opposite, and it will create a very high confusion; moreover, we are wasting time and money for the company.

Another point that is critical is that we have a concept that we need only to change one parameter per experiment. So, if I want to find yield or high yield by purity, then we change the time and temperature. Once we have 5–6 temperature experiments, we can go to the concentration and find the best temperature to change the concentrations. When we do it this way, we are taking a very long route, which is not always the best route due to some synergism and interaction between temperature and concentration together. Therefore, changing one parameter per time does not provide the best a diverse point, the best regime, but only an interaction between them. I advise using a statistical tool to help conduct a full factorial design of experiments that takes into consideration more than one parameter at a time and includes it into these parameters; i.e., not only chemical or biological considerations but also physical considerations. The meaning of it is, when we are taking consideration which parameter will influence, normally in the chemistry, when the biological point will influence the type of material compared to the concentration. But this is only one point to consider. The second point is the physics. What will be the interaction between materials? What will be the flow we are a generating and how will it be representative in the next step? How will be the production equipment, when will be the conditions in the production equipment to generate some representative situations?
Consider what happens in the production equipment and implement it as part of the metrics of experiments that will be done; but also consider that they have some synergism and interaction between the parameters. In conclusion, take into account the regimes that you apply, the interaction between parameters, the chemical and the physical points, and the constraints or restriction you will get from the production step, and reflect these considerations in the lab. The only way to do this, of course, is to use theory and a simulation package to connect between your equipment and your process in a way that you will connect between the theory and the real-world situation.

VisiMix Video

The Influence of Mixing in the Process
New methodology using VisiMix software for the purpose of checking the influence of mixing in the processes.

VisiMix Demo Operation
Learning how to input data into the VisiMix software and get results the will help us understand the influence of mixing in our processes.

Lab Experiments
Learning how to set up the relevant experiments at the lab scale, to develop the processes from an engineering point of view.