Hi
I`m relatively new to openLCA, but am currently modelling some systems for buildings, as a part of my masters thesis.
I`ve noticed, that while my model in general seems to be connected nicely, the output fails, when I disable one of the processes. It`s not a specific process, but accounts for them all.
The model has been constructed, dividing the building into structural building component (outer walls, inner walls, foundation, etc.). These all feed further into the model, e.g. transport, then use etc. Each building component contains various building material inputs (e.g. reinforcing steel, concrete, timber etc.) which are controlled by global parameters. I`ve chosen to use global parameters, as the input data are used several places.
I`ve controlled output from each process regularly, in order to assure that no process mistake is made. However, now, when assessing larger chunks of the model, e.g. transport of all building components, I`ve noticed some errors in the outputs, which I do not entirely understand. I hope some of you might be able to help me, or know this issue.
When I make changes in the input parameters, upon result extraction, I seem to get impacts out and a list of total requirements which makes sense to me, along with a sankey diagram that showns the modelled processes and the impact inheritance. However, if one of the building components does not occur in a scenario (e.g. one of the buildings does not contain inner walls at all, and hence those input parameters are all set to 0), the output is zero - I get neither impact results or total requirement values.
In my understanding, the act of disabling a building component, should just result in that one component not being accounted for, but everything else being transferred and evaluated.
I`ve checked this tendency for all my building component inputs, as I first thought I might have multiplied the disabled component somewhere by mistake, however the fault occurs throughout, regardless which component I `disable`, by setting all material input parameters to 0.
The erroneous output can be overcome/avoided, if I put a decimal numbers for just one of the components input materials/parameters, however, in theory, this would not be correct, as a materials is then accounted for which should not be in a specific scenario.
I`ve been trying to locate the reason my running through my model several times, but seems unable to locate the issue. Unfortunately I`m not able to read or analyse the structures behind, and hence I`m dependent on the user friendly interface.
I hope some of you may have experienced the same issue and are able to help me how to solve it, as I`m in slight time pressure as well, but would like to continue modelling in openLCA for my project.
(Not sure if it`s relevant, but I`m modelling in version 1.6.3, using ecoinvent 3.3 database - apos data, using ReCiPe Midpoint(H) as impact assessment method)
Cheers,
Sarah
