So I looked through the past 2-3 shifts' 100k analysis for after a hwp change. They appeared to only take short runs to get the charge asymmetry down. I took a different approach: we marked the runlet as suspect but let it go a full run. Perhaps this isn't the smoothest way to run?
Anyway, when run 14109 is finished analysing the QwAC should probably spend an extra minute comparing the first few runlets. I suspect it will go away, especially because everything else looks fine.