I remember glancing at this a while back. I think I understand that you are intending to not write any flows for the single reach case and instead remove the flowjaface record entirely. I wonder if this is the best way to handle it. Perhaps it would be better to still include the header but just not write any records? I'm thinking that this may break other downstream processors, like zone budget. Do we have a similar problem for LAK/MAW/UZF? Can you remind me why this is a problem for SFT but not SFR (I think we discussed this once before).

I will need to refamiliarize myself with the issue. As I recall from reading my previous notes, I think the issue had something to do with the flowjaface budget term in single reach models (meaning its not needed). The transport model was either (1) not expecting this budget term and it was present, thereby throwing off the indexing or (2) was expecting this budget term and it was missing by virtue of there being no connected reaches.