Efficiency in Process Development

TEA-guided Process Development

In the early days of Amyris, we had an advisor named Hans van Dijken. If you haven’t had the pleasure of meeting him, it’s definitely your loss - not only did he have an encyclopedic memory of all things yeast physiology, he was a very generous and entertaining person. We learned a LOT from him. Anyhow, as Amyris was considering how to expand its -omics capabilities, Hans was fond of saying something like “The only -omics that matter in fermentation is Econ-omics!” I think Hans may have coined that expression 20+ years ago …

That lesson continues to be a critical one for our industry. As such, I’d share some examples from my past that can help drive the theoretical point home more clearly, and perhaps give you an idea or two about what you could do differently.

So obviously, technoeconomic analysis can help you understand where the big levers are on cost. Having a model helps you translate technology performance into economic terms and informs your business plan. And if you are a leader at your company, you want to have a firm grip on where you should be investing your resources for the maximum gain - even if it’s a simplified model: For example,

Let’s assume you have confidence in and can reduce your complex TEA model into a simple equation[1] like this:

Now, this is too simple - other things are in play. For example, can the technology improvement be implemented without CapEx investment? Is one tech development group “less overbooked” than another so the opportunity cost is lower?, etc. This decision can’t be made in a vacuum. But by simply keeping these ideas front of mind, you can make smarter decisions about where you are investing your process development dollar.



For example, I once consulted with a partner on a process for a novel product. A quick assessment suggested that things looked good – the process fit in the plant well and it was in their implementation wheelhouse; the production costs seemed reasonable and there was enough profit to go around. But then we really interrogated the costs – really pushed on the assumptions in the more important unit operations, and we learned that the costs in a few areas were actually significantly higher than was originally projected. The reasons for that are many, but the key is, because we had invested in and interrogated the TEA, we learned BEFORE the deal was signed that this wasn’t a good move for the group. So while I had to share the bad news to an unhappy client, the good news was that they avoided making a big mistake, and that was in the long run much better for them.

I worked with another client that had an existing TEA done for their early-stage technology. It was a complicated, excel-based monstrosity – hard to get into. I was asked to build upon it, which, while difficult, ended up being a good move by the company – not only did I uncover some optimistic original assumptions which threw the accuracy of the original TEA into question, but I also was able to correct some mistakes in how certain processes were modelled AND identify a few big opportunities for process improvement. The combination had the potential to result in more than a 50% cost reduction if the ideas stemming from the modeling worked in the lab. And while I’ve done this kind of thing before, and certainly experience with different types of processes helps, probably the biggest lesson there was that simply by getting a second set of eyes on the model, the company was able to improve the quality and accuracy of their projections.

If you don’t have a process flow diagram and a high-level TEA, I’d be happy to help you build one. If you do have a TEA and are thinking you’d like a second set of eyes on it and/or some thoughts about what it is telling you, let me know – I’ve been fortunate to have built/interrogated many cost models in the past and would be happy to work with you on yours.

[1] By the way this can be a very effective exercise – see here more:   Benjamin KR, Hill PW, Meadows AL, Singh AH, Cherry JR, “Use Cost Models to Guide R&D”, Chemical Engineering Progress, June 2016, pp. 44-50