Hi, are you using the EN15804 add-on? We also observe that the amount of waste is higher than in other sources, but believe this is the correct result from ecoinvent. Waste needs to be specifically addressed to show up in the calculation result, and if you do not do this (or do this incompletely), you get 0 to lower result. You of course need to pay attention to not double count waste in waste treatment chains, but I believe we do this. This is an explanation for lower results in other systems. Another explanation for especially GaBi can be that GaBi seems to model incomplete life cycle systems only, following only the "assumed relevant" parts of a supply chain.
Hope this helps, best wishes,