What does Monte Carlo analysis assess if enter uncertainty from Pedigree Matrix?

Question

Dear all,

please forgive me if my question was answered before but I did not find anything on ask.openLCA as well as the manual. My questions is the following:

Let's say I create a product system with three inputs: Polyethylene, Electricity and Heat. I select three providers from ecoinvent, which all have there own uncertainty. Then I create uncertainty data using the updated Pedigree appoach by you. for the product system.

When I then perform a Monte Carlo analysis are there both uncertainty distributions used? E.g. is the amount of electricity and heat varied according to the product system's uncertainty only or is the underlying uncertainty for the providers (from ecoinvent) analyzed as well?

Thank you very much in advance!

Answer 1

Whenever there is an uncertainty inside an exchange that is used in the product system, it will be varied as well and contribute to the overall uncertainty. This is usually the case for the unit process databases from ecoinvent, where nearly all exchanges have uncertainty data (some reasonable and some unreasonable) by ecoinvent.

I case you want to only vary the uncertainty of your own foreground processes, you can either use the ecoinvent system processes in the background (fully aggregated and without any uncertainty data), or you can delete all uncertainties in one of your copied ecoinvent unit processes databases with the following SQL command and afterwards insert your foreground uncertainties or import your foreground processes/models.

UPDATE 
 TBL_EXCHANGES 
SET
 DISTRIBUTION_TYPE = NULL,
 PARAMETER1_VALUE = NULL,
 PARAMETER2_VALUE = NULL, 
 PARAMETER3_VALUE = NULL;