Design temperatures being read incorrectly from newest TMY3 weather files

[ekpresent]

@carlobianchi89 pushed some new TMY3 weather files recently to fix some bad sensor issues in a few specific cases. They also used 2021 ASHRAE design conditions instead of 2009, which didn't seem like it would be a problem. However, we are seeing some issues. The one we are aware of currently is that in the LA weather file, the 0.4% MCWB (2021 value: 63.5) seems to be being used as the cooling design setpoint where the 1% DB was being used before (2009 value: 82.1). Or at least that seems to be the most likely explanation so far. This is causing some pretty extreme problems with HVAC sizing. I don't know where the issue is - whether it's improperly formatted in the weather file, improperly read by OS-HPXML, or somehow a ResStock issue.

@joseph-robertson @afontani

[shorowit]

Can someone provide the before/after LA weather file?

[joseph-robertson]

I'm looking at this with @whiphi92 right now. It looks like these new TMY weather files with "2021 ASHRAE design conditions" have a new element "WSF" in their design conditions header row. Thus, offsetting by an index of 1. The OpenStudio SDK thus parses, for example, coolingDryBulb1 incorrectly as coolingMeanCoincidentWetBulb1...

[whiphi92]

GitHub doesn't support epw files. I will send these to you on teams @shorowit

[ekpresent]

Sounds like we should just ask @carlobianchi89 to put the 2009 header on all weather files for now?

[afontani]

@carlobianchi89 and I talked and he is going to use the 2009 header to keep things consistent.

[shorowit]

I'm looking at this with @whiphi92 right now. It looks like these new TMY weather files with "2021 ASHRAE design conditions" have a new element "WSF" in their design conditions header row. Thus, offsetting by an index of 1. The OpenStudio SDK thus parses, for example, coolingDryBulb1 incorrectly as coolingMeanCoincidentWetBulb1...

The design conditions in an EPW format should reflect this format. So it sounds like the 2021 design conditions were not properly generated.

[afontani]

This was an off-by-one error that has been fixed in the processing of the weather files.