Using Ecoinvent in OpenLCA (for the beer bottle case)

Post Reply
KurtLeon
Posts: 12
Joined: 20 Sep 2013 09:21

Using Ecoinvent in OpenLCA (for the beer bottle case)

Post by KurtLeon » 03 Nov 2013 13:52

I tried to rebuild the Beer Bottle (Alu vs PET) with data from ecoinvent. The result was more or less zero.
So i tried to find out the reason for that. I compared the Sample data for electricity (electricity, at grid, US) with the ecoinvent data (electricity, medium voltage, at grid, US). I made 2 productssystems based on these processes, to compare it.
The result for the sample data for fossil (cumulative energy demand) was 2.18 MJ for 1 kWh - this seems ok to me.
The result for the data from ecoinvent for fossil was 2.174 E-11 for 1 kWh.
I don't know why I'm getting such an unbelievable result... I think there's an error with some unit or something else...
I really hope somebody can help me, with this problem!
Thanks!

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

Re: Using Ecoinvent in OpenLCA (for the beer bottle case)

Post by aciroth » 03 Nov 2013 14:18

There are indeed large differences between NREL and ecoinvent, one reason is that NREL (in the version we use for the case study) has several dummy processes that create a product "out of nothing". Just check out some of these processes.

KurtLeon
Posts: 12
Joined: 20 Sep 2013 09:21

Re: Using Ecoinvent in OpenLCA (for the beer bottle case)

Post by KurtLeon » 04 Nov 2013 17:02

Thanks for your prompt reply.
I'm aware of the fact that there are diffences, but not with a factor 1.000.000.000.000 - this seems to me more or less like a calculation error... maybe a problem with the unit... and the NREL data is at least plausible, but the ecoinvent data is wrong. I have the same problem with other processess as well, so it seems that not the data (nrel, ecoinvent) is the problem, but the calculation.

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

Re: Using Ecoinvent in OpenLCA (for the beer bottle case)

Post by aciroth » 06 Nov 2013 01:42

..as I tried to explain, the case is made for the NREL database; not sure how you made the ecoinvent system for the case but I would be very surprised if this was a calculation error since we really tested the recent openLCA versions (I assume you are using a recent version)?

KurtLeon
Posts: 12
Joined: 20 Sep 2013 09:21

Re: Using Ecoinvent in OpenLCA (for the beer bottle case)

Post by KurtLeon » 07 Nov 2013 12:31

Thanks for your reply. I'm not re-modelling the whole beer bottle example. I'm trying to compareonly one concrete process - from NREL the electricity, at grid, US and from ecoinvent the electricity, medium voltage, at grid, US. To compare it I made 2 productssystems based on these processes, to compare it.
The result for the sample data for fossil (cumulative energy demand) was 2.18 MJ for 1 kWh - this seems ok to me.
The result for the data from ecoinvent for fossil was 2.174 E-11 for 1 kWh - this seems wrong to me!

Do you know which problem I'm using to solve? I only compared to processes with an LCIA method - and I got very bad results for the econinvent data. I checked the units, but I've found no problem. I can send you my project database, so you can have a look on it - if this will help.

KurtLeon
Posts: 12
Joined: 20 Sep 2013 09:21

Re: Using Ecoinvent in OpenLCA (for the beer bottle case)

Post by KurtLeon » 10 Nov 2013 11:02

I just found out, that ecoinvent will work, if i'm importing the nexuspack-file into an empty projectdatabase. Importing it to the database from the beer bottle sample results in strange results. Do you have an explanation for this behaviour? It's the same with NEEDS...
ELCD still not works, no matter what I'm trying. The result of analysing productsystems with ELCD data is always 0.
Maybe somebody can help me in this case, thank you!

crodriguez
Posts: 7
Joined: 11 Dec 2013 19:25

Re: Using Ecoinvent in OpenLCA (for the beer bottle case)

Post by crodriguez » 12 Dec 2013 12:25

The problem using ELCD data and the ecoinvent method “cumulative energy demand” is that the characterization factors in the method refer to different flows that those in the inventory (e.g. "brown coal; 11.9MJ/kg" and "Coal, brown, in ground”). Thus, the result is 0. If another method is selected, e.g. Ecological Scarcity 2006, results are obtained for some categories.

The case study with NREL data was created in an “empty database”, that is, without reference data whereas the ecoinvent nexus pack contains already the reference elementary flows from openLCA. Although the name of the elementary flow might be equal, they are, in fact, different flows for the software (i.e. ecoinvent elementary flows are located for example in category Flows>Elementary flows>air while NREL are, for example, in Flows>air). Therefore, in the method the CFs are applied to the flows from NREL but not to those from ecoinvent.
The same reason is applied to NEEDs.

To avoid this and allow users to use the PETvsAluminium case study with other databases, a new database was uploaded in http://www.openlca.org/downloads.

We have imported the ecoinvent nexuspack into the new case study database and calculated the LCIA results for the electricity processes you wanted to compare. The results obtained are 2.18 MJ-eq for NREL (electricity, at grid, US) and 9.03 MJ-eq. for ecoinvent2 (electricity, medium voltage, at grid US). The difference in results is due to the different complexities of the product systems.

Post Reply

Who is online

Users browsing this forum: No registered users and 13 guests