It is very important that as professionals we can agree on what we are testing.  Carbon Monoxide (CO) is one of the most confusing issues.

First, BPI's prime directive is health and safety first.  Monitoring the ambient CO during analysis is clear.  A personal CO monitor or a combustion analyzer can be used to monitor ambient CO.  Use the CO value on the analyzer for this.

Second, combustion analyzers can be used to measure the combustion byproduct CO in incomplete combustion.  To accomplish that, the probe of the analyzer must be sampling the undiluted gases in the flue.  Where you locate the probe depends on the appliance you are testing, and most of the combustion analyzers provide information on how to accomplish that.

Third, most combustion analyzers provide two CO values: CO and CO-AF or CO Air Free. CO-AF is the value that BPI is looking for for combustion testing.  (If in doubt, record the CO, the CO-AF, and the O2 levels.)  If the analyzer does not provide an O2 value or a CO-AF value, use the CO value and note that on the report.

CO-AF is a value calculated by the analyzer that subtracts the oxygen from the CO reading.  The analyzer should be initiated in clean air with a oxygen level close to 20.9%.  The formula is:

                                              CO ppm x ((20.9)/(20.9 – 02))

A furnace with a CO level of 2 and oxygen of 8.9% would have a calculated CO-AF reading of 3.5.

A furnace with a CO level of 46 and oxygen of 3.2% would have a calculated CO-AF reading of 54.3.

Fourth, since the CO-AF value will be higher than the CO value, it is important to get the right CO value to begin with and that means inserting a NOx scrubber in the path to the analyzer - either using an analyzer that has one built in or adding it to the hose.


If we can agree on the use of these values in this way, then I think we will be much further along in providing consistent, reliable results.

Views: 805

Replies to This Discussion

Paul,  Thank you for writing this clear description of the issue and the testing protocol.  Question:  What is the Action Level for CO-AF for natural draft appliances and gas ranges, and what is the recommended Action?  Thank you, Tamasin


Does this mean that if the air's O2 percentage is exactly 20.9 that the calculated CO-AF reading is infinite?                       Or that if the O2 percentage is greater than 20.9 that CO-AF is negative?  What do these numbers indicate in more practical/intuitive terms?  Thanks.

Curious where you found that BPI is looking for CO-AF.  I've always been told BPI is looking for CO not air free CO.  Would love to know where the clarification is as to which CO reading BPI's action levels are for.

In the BA Standards under review.  You'll find what is planned.  These standards are not what folks are expected to know, follow, and test under however.  We're just looking ahead.

Thanks, Tamasin.  Unfortunately the description of CO measurement in the present BPI BA Standard uses the terminology of "CO as measured" which is not a value that appears on a combustion analyzer screen.  It's CO or some form of CO Air Free like COAF.  So the question is which one are they referring to?   I vote for COAF because I KNOW that much of the time the analyzer probe is not in the right place, sampling CO that has not been diluted.  Also, the COAF value tells us not only how much CO is in the sample but also how much oxygen which tells us something about how well the appliance is functioning which is basically the point of the exercise.  Chris, if you get to a location with those oxygen levels, I would love to hear the results.  What can I tell you?  I'm not enough of a scientist to create that formula.

Thanks Paul.  I haven't seen >20.9% oxygen levels in a home, I'm just trying to understand the combustion chemistry that the formula is based on, and what exactly COAF is measuring that CO isn't, and why that is important.  

Okay, so based on this it looks like COAF is meant to correct for how much available O2 is available in the combustion air relative to how much the appliance requires for complete combustion.  The formula you gave essentially removes the excess O2 in the flue gasses, allowing you to measure the CO concentration as if there was zero excess O2 (that it, if it were undiluted).  And this allows an apples-to-apples comparison of combustion efficiency/CO production between appliances having different combustion air requirements?  Maybe this explanation could be useful for other people (like me) who are new to all this.


Home Energy Pros

Home Energy Pros was founded by the developers of Home Energy Saver Pro (sponsored by the U.S. Department of Energy,) and brought to you in partnership with Home Energy magazine.  Home Energy Pros is sponsored by the Better Buildings Residential Network.

Latest Activity

Dan McKee posted a blog post

9 Tips To Get Your Furnace Winter Ready

A well heated, cozy home is what makes the day better when the winter comes in full fledged. So, it…See More
5 minutes ago
Tauran Ivall replied to Tauran Ivall's discussion Energy Efficient Entry Door options out there?
"Thanks Graham, There is a distributor in Portland for them. Thanks for the link to the blog.  "
2 hours ago
Bob Blanchette replied to Rod Fox's discussion Utility Meter Accuracy
"The use of CT's screams commercial service, most likely over 200A. 200A and lower uses simply…"
2 hours ago
Dennis Heidner replied to Rod Fox's discussion Utility Meter Accuracy
"It's not just the CT's that you need to worry about.   The CT will pick up the…"
6 hours ago
Rod Fox replied to Rod Fox's discussion Utility Meter Accuracy
"Good point, Dennis. In the case of my original post, we're talking about a digital meter.…"
7 hours ago
Dennis Heidner replied to Rod Fox's discussion Utility Meter Accuracy
"I think you met the reply for Rod Frost."
7 hours ago
Kevin Wood replied to Tauran Ivall's discussion Energy Efficient Entry Door options out there?
" has training on the fiberglass doors. I wanted to transfer the link but I am not…"
11 hours ago
Vernon House posted photos
11 hours ago

© 2015   Created by Lawrence Berkeley National Laboratory.   Powered by

Badges  |  Report an Issue  |  Terms of Service