CO2 calculated incorrectly, zone air contaminant balance

edit

I am running into a wall and can not find a solution to my problem.

I have a single zone building with a constant air volume mechanical system. The ventilation rate is fixed on a schedule at 0.19 m3/s. 27 people come in and out of the room on a schedule. The CO2 should rise to about 1200 ppm and then reach steady state. This is a school so there are 180 days of class. The contaminant balance works correctly most of the time. But a couple of days for only some of my weather files, it does the calculation incorrectly and the CO2 shoots up. I've included an example where ventilation is blue line and CO2 is red line. Occupancy is the same for each day.

https://drive.google.com/file/d/1JJ80...

I am running 70 files from historical weather and not every simulation has this problem. Only a handful do, but I don't see any pattern when it happens. My timestep is 5 minutes.

I could obviously post-process the answer, but I need to make sure that the contaminant balance is working because I'm going to be running some DCV control schemes.

More information , 12/21/2023 8:12 AM PST

It looks like the problem is in my weather files. They were purchased from whitebox technologies so I shouldn't upload the whole file here. But for one file, the three days for which the CO2 calculation has a problem is October, 12 and 13 and Dec 13. If I take the weather for Oct 11 and substitute it for October 12 and 13 and the weather for Dec 12 and sub for Dec 13, the problem goes away. So bizarre. I see nothing wrong with the problematic days of weather. I've posted a chunk of the weather file here:

https://drive.google.com/file/d/1tpVS...

October 11 weather produces correct results for CO2 calc. October 12 and 13 product incorrect result, matching the image in the first link.

Of course I could do a bunch of manipulation of my weather files to get rid of the problematic days, but I really don't like that I don't understand the root cause. Anyone have any clues?