I opened a discussion with one of the NDIR sensor manufacturers, explaining these requirements to them. In short, they said the high temperature is problematic.

Your idea with the buffer tube is good, and we did something similar in the past, when several A3s were installed on buses and wanted to have some protection from the high pressure jets used to wash them.

But on PM2.5/PM10 readings there were two issues: 1) a decrease in amplitude, compared to units running at the outside. 2) a delay in registering the elevated readings, as the air had to slowly travel the tube.

2) is not a problem for you. With enough time, 1) should reach the real value too. So if you can have this run for let’s say, half an hour, you should get close to the real value found in the hot area. You’ll need a tube + a fan, to force the air from the sauna all the way through.

Planning on doing this anywhere soon? Let me know if we can help.