Corect procedure to accomplish fast Monte Carlo analysis

Question

Hello, i wanna know which is the correct procedure to accomplish a fast Monte Carlo analysis, since this process run "normally" take very longs time to finish. Which is an adequate cutoff to speed up enough the calculation? or, if i save the impact calculated for a process as LCI and then i run the Monte carlo on this, the results is correct? For last I have some dubt from theorethcal point of view.

Thanks

Answer 1

There is really not a straightforward answer on this. The goal and scope of your study defines what cutoff is appropriate. If your goal is to calculate a Monte Carlo Analysis as quick as possible to approximate results you may want to use a high cutoff value. If you want to be as accurate as possible you have to define a low cutoff and wait ...

Not sure if this is a very satisfying answer but I am afraid this is how it is. If you have not done this yet, you way want to assign more memory to your openLCA https://ask.openlca.org/594/how-to-assign-more-memory-to-openlca?show=594#q594

Comments

Thank you. Yes i have already assigned 14 Gb to OpenLCA, is it enough? The problem is that there is a very upstream process that doesn't allow me to reduce the cutoff, otherways a lot of error is generated. If I save the results of a calculation as a LCI and i make the monte carlo on that, can i speed up the process?
The last time, to accomplish 5000 itarations, it take almost 12 hour, with an i7-8700.