Negative values in the Ecoinvent 3.1 database

Post Reply
s_mellino
Posts: 11
Joined: 25 Mar 2013 20:57

Negative values in the Ecoinvent 3.1 database

Post by s_mellino » 13 Apr 2016 13:59

Hello,

I am using the Ecoinvent Database 3.1 obtained from nexus-openlca website. I would like to know why some values in processes (for example the process "fuel cell production, stack polymer electrolyte membrane, 2kW electrical, future, alloc. default, U - CH" for the input waste plastic, industrial electronics) are negative.
I am running also MonteCarlo simulation and I obtain negative values for mean for many categories, how is it possible? I am supposing that It may be related to the negative value in the processes...

I hope that someone could help me.

Regards,

Salvatore

aciroth
Posts: 750
Joined: 09 May 2010 23:28

Re: Negative values in the Ecoinvent 3.1 database

Post by aciroth » 14 Apr 2016 12:32

Hi Salvatore,
ecoinvent processes have negative values for several reasons, the main being waste modeling, these can lead to negative results; this holds also for the example you mention. Basically waste flows are moved from output to the input side, and obtain a negative sign.
For the Monte Carlo simulation, the log normal distribution (which ecoinvent uses by default) as such does not have negative values, but "negative supply chains" (due to waste modeing) will lead to negative values always, and when these are combined with other supply chains thus the result can be negative overall.
Hope this helps?
Andreas

s_mellino
Posts: 11
Joined: 25 Mar 2013 20:57

Re: Negative values in the Ecoinvent 3.1 database

Post by s_mellino » 18 Apr 2016 13:21

Hi Andreas,
thank you for your reply. The problem is that it is difficult to interpreter and discuss the results of Monte Carlo Simulation when we have for a given category 0.55 as quick result and -16 with a standard deviation of 130 from the montecarlo. What do you think?
Salvatore

aciroth
Posts: 750
Joined: 09 May 2010 23:28

Re: Negative values in the Ecoinvent 3.1 database

Post by aciroth » 18 Apr 2016 13:55

Well this ("credibility" of the MC result) depends on the uncertainty specified (where in the system, how high and what kind of uncertainty distribution) and how many simulations you have performed. You can also test the uncertainty you have specified in openLCA, see attachments, from the inpout output sheet.
Best wishes,
Andreas
2016-04-18 13_46_08-Fotos.png
2016-04-18 13_46_08-Fotos.png (6.44 KiB) Viewed 6106 times
2016-04-18 13_46_22-Fotos.png
2016-04-18 13_46_22-Fotos.png (12.44 KiB) Viewed 6106 times

s_mellino
Posts: 11
Joined: 25 Mar 2013 20:57

Re: Negative values in the Ecoinvent 3.1 database

Post by s_mellino » 02 May 2016 16:21

Hello Andreas,
I understand what you said. The problem is when a result of MC is a negative value...how it could be possible? What is the meaning of a negative impact category?
Thank you for your time and your help.
Salvatore

Post Reply

Who is online

Users browsing this forum: No registered users and 12 guests