Hello,
I am currently working on a two-step fermentation process. How do I use Response Surface Methodology or similar techniques to optimise the final output (from the second step) for a two-step process? The output of the first step does affect the second step but optimising the first step does not necessarily optimises the output of the second step. So the only objective is to maximise the output of the second step.
Can anyone point me to the right direction (method, keywords, papers etc.)?
Thank you in advance!
Hello, Eqwan! Hope you're doing well. I thought about your question for some time, and came up with this idea: why don't you try to get the optimum conditions for the second phase, and then try to get the same outputs from your first one?
It all depends on what are you trying to optimize, of course. Is it some compound? Or is it the growth of the bacteria/yeast? I believe that if you're trying to get maximum yield, then you should go try optimizing the first step and get the most of that last step as well.
For all these, I would recommend reading the almighty RSM book, "Response Surface Methodology: Process and Product Optimization Using Designed Experiments" by Montgomery and Myers. It's a piece of art.
For analizing the data and getting the best out of it, I would recommend Design Expert software (it's not free, but it is the best of the best). I was using MiniTab but the first is way much better!
Hope to have thrown some light in the subject. Cheers from Argentina!
Hello Pablo Gustavo Kildegaard !
Thank you very much for your answer. Hope you are doing well as well.
I am rethinking my question again. Maybe I didn't ask it properly. I'm trying to maximise biogas yield on the second step but the second step is in standardised conditions. The only parameters that are manipulated (temperature etc.) is in the first step. So I guess your suggestion is true in my case.
Since the second step is in standardised conditions, although the yield comes from the second step, I assume in my case the second step should just be ignored as if it doesn't exist?
Thank you for your suggestion on the book and the software as well. Will look into it!
Cheers from France!
Hi Eqwan Roslan ! Yeah, I believe you could just optimize your first-step yield, and then consider the best condition to get your biogas without having to reach that last step at intermediates states, hence saving some time and money :)
I assume that for your standardized biogas synthesis, you would need maximum yield in the first step. From this, I would suggest doing a Central Composite design (for 2 variables) or a Box-Behnken design (if you have 3 or more variables) using your final yield as a response, and then find that maximum response.
If you're not sure if your variables are at optimum conditions, then a more exploratory analysis, prior to these RSM, would came very in handy! Usually, you tend to "go" in the direction of the gradient which suggests the location of that maximum/minimum. All these stuff is very well explained in that Montgomery book I recommended before.
Take care!
- Badges
- Science topic