0 votes
60 views

I am currently conducting an LCA study for sugar production using openLCA, and I am experiencing a calculation problem where openLCA suddenly crashes/closes without showing an error message when I run a product system with Fine Sugar Milling or Fine Sugar Packing as the reference process/product.

My process structure is approximately:

Sugar Handling → White Crystal Sugar → Fine Sugar Milling → Icing Sugar → Fine Sugar Packing → Packed Icing Sugar

The input and output mass balances of these processes are consistent. The Icing Sugar flow from Fine Sugar Milling is also linked as the provider for the Icing Sugar input of Fine Sugar Packing.

However, when I create/run a product system using Fine Sugar Milling or Fine Sugar Packing as the reference, openLCA crashes and closes completely. I do not receive an error message.

Interestingly, when I use Molasses Storage/Molasses  or White Sugar Packing as the reference flow, the calculation can be completed successfully.

I have checked the process inputs/outputs, provider links, reference flows, and quantities. I also checked that there are no input, output, or quantitative reference values equal to zero in these processes.

Software and database information:

  • openLCA version: 2.6.2
  • Operating system: Windows
  • Database: ecoinvent 3.12
  • Calculation mode: Lazy / On-demand
  • No input/output or quantitative reference has a value of zero

Could you please advise how I can identify which process, flow, provider, or calculation dependency is causing openLCA to crash?

ago in openLCA by (180 points)

1 Answer

0 votes
ago by (148k points)
selected ago by
 
Best answer
Hi, thank you for sharing your detailed observations. openLCA runs really stable, these points can lead to crashes:

- not enough memory allocated (check - in windows - in the task manager during a calculation). You will typically see a "out of memory" note in the log file. Ideally allow 10GB RAM for ecoinvent calculations e.g., but the required amount depends a lot on the database and model. If you do not have sufficient physical RAM on your computer, use libraries.

- model is invalid (because you accidently deleted something e.g.) - validate your database for any issues

- some plain issues (full harddisc, databases on a remote server where access is slow..) can also lead to crashes.

That is about it. Ideally, check and report back?

Thank you,

Andreas
ago by (180 points)
Thank you for your assistance. I tried running the calculation again and found that my product system can be successfully calculated using the Eager/All calculation mode.

Is this considered normal? From several sources I have read, the Eager calculation mode generally requires more memory compared to Lazy. However, in my case, I am able to run the calculation successfully using Eager/All.

Could you please clarify whether this is expected behavior and whether there are any implications or differences I should be aware of when using Eager/All instead of Lazy?
ago by (148k points)
please follow the points I mentioned above, thank you. This is of course not expected behaviour.
...