Open cmutel opened 5 years ago
This is tough - there are a lot of problems with the ecoinvent parameters, including zero division errors, undefined functions (e.g. UnitConversion
despite the spec saying this should be UC
and "This formula has not yet been implemented in the ecoinvent context and should not be
used"), references to parentvalue
, references to external spreadsheet values (from PSI!), etc. 257 datasets can't be successfully recalculated.
This functionality was added in 5f75c2f690307ff82f42fe360dc8ff769a817ff5, but isn't recommended for now. Overall model accuracy actually decreases by about 1%, because even though the particulates problem mentioned above is solved, more new problem are introduced.
See market for hard coal in ZA - the particulates are determined by the variable
transportation_losses
, which is defined elsewhere.