Removing steel layer from insulated pipes increases pipe losses 10x

edit

I was trying to test well-insulated pipes but got CTF errors when I had high levels of insulation. The .err file suggested removing any highly conductive layers, so I removed the layer of steel from my pipe Construction. However, this caused pipe losses to increase tenfold, even for pipes with normal insulation levels. See graph below.

Am I missing something? Has anyone else encountered this?

I was able to recreate this with the example file PipeHeatTransfer_Schedule.idf, uploaded here: https://gist.github.com/ejhw/f3aeb5d4...