Hi @Peter Svensson ,
When you say that you can't get it to work, are you seeing any error?
With regards to limitation with temp storage, as explained here, the only limitation is you can't use more than 1 GB of disk space, which is the quota for each sandbox.
As explained here, Azure Automation assigns a worker to run each job during runbook execution in the sandbox. While workers are shared by many Automation accounts, jobs from different Automation accounts are isolated from one another so, I believe few of your runbook execution's isolation might be the reason for you to not able to append content to your intended csv file.
So, to try resolving the issue in this case, write your runbooks in such a way that you append the content from each runbook execution to the csv file that's stored in blob storage.
Example:
Runbook 1 -> $env:TEMP\xxx.csv -> cost.csv in blob storage
Runbook 2 -> $env:TEMP\yyy.csv -> cost.csv in blob storage
Runbook 3 -> $env:TEMP\zzz.csv -> cost.csv in blob storage