### Author Topic: scenario analysis  (Read 4607 times)

#### wimpi

• New Member
• Posts: 1
##### scenario analysis
« on: October 16, 2014, 17:37 »
hi,

i am using umberto nxt lca and would like to incorporate some kind of scenario analysis into my model:
let's say there is only one main process which yields one type of material. in scenario A, i would like to consider this material as a waste flow (leaving the system) and in scenario B, i would like to apply some further processing. Hence, the main process needs two exit flows: One flow in case the material leaves the system (scenario A) and one leading to another process  (scenario B). In scenario  A, all material leaves the system (99 kg in the following example). In scenario B, all material serves as an input for another process. my idea was to create a net parameter (e.g. "PARA", =0 for scenario A; =1 for scenario B) and to use  functions to define the respective flow leaving the main process:
Flow leaving the system: IF(=(PARA,0),99,0)  <<--scenario A
Flow leading towards further processing IF(=(PARA,0),0,99) <<--scenario B

However, I cannot use such kind of scenario modelling because umberto does not accept coefficients that are equal to 0, which is the case as in either scenario, one flow equals 0.

How can i avoid this or change my model to incorporate scenario modelling?
(I could use 0,000000001 instead of 0. however, i prefer a "clean" solution)

my model involves several scenarios and it is hence considerably inconvenient to creat a model for each scenario.

Thanks a lot

« Last Edit: October 16, 2014, 17:40 by wimpi »

#### uniquename

• New Member
• Posts: 25
##### Re: scenario analysis
« Reply #1 on: October 17, 2014, 11:10 »
Hi wimpi,

when setting up a model as you described, the impossibility of flows with the amount=0 is indeed a problem. I understand that you would like to do it and inserting a quantity of 1E-100 instead of 0 is a feasible workaround that will deliver results where the differences to a flow with the amount of 0 can be neclected.
You could of course also copy the model and just exchange the process. This would deliver exactly the results you are aiming for. Neither of these options is the (more elegant) solution that you were aiming for.

I know that the Umberto developers have already thought about these issue and possibilities to enable the calculation as you described it, but unfortunately no date is set for that task yet. So for now you will need to stick to the workarounds described.
Have a nice weekend!