Carbon Monoxide: What Do You Know?

February 14th, 2017 | Posted by David Richardson in David Richardson | Home | Technical Blogs
David Richardson, Curriculum Developer & Instructor

David Richardson, NCI Curriculum Developer & Instructor

Many HVAC professionals assume the leading cause of carbon monoxide (CO) poisoning is a cracked heat exchanger. This assumption leads to inaccurate test methods that pass on from technician to technician. They try to assure customer safety, but unknowingly rely on misinformation.

No one wins in this scenario — both the contractor and customer have a false sense of security. They ignore other sources of CO after finding the heat exchanger is intact.

So, let’s look at one of the most inaccurate test methods commonly used to test for a cracked heat exchanger and why it could leave a potentially dangerous situation.

Testing at a Supply Register with a Store-Bought CO Alarm

One of the most common and deceptive test methods for determining heat exchanger integrity is placing a store bought carbon monoxide alarm on a supply register to see if it displays a CO reading.

The reasoning behind this: a cracked heat exchanger enables air blowing from the supply register to contain CO from the flue gases “leaking” into the air stream. If the dangerous gas is in the duct system, it will set off the alarm. Sounds good, right?

Store-bought CO alarms

Store-bought CO alarms used for testing simply aren’t sensitive enough for accurate measurements and diagnosis.

Let’s consider the limitations of a store-bought CO alarm. Carbon monoxide alarms are rated and certified by UL Standard 2034. The alarms are designed to tolerate CO as follows:

  • 70 PPM +/- 5PPM for 60 – 240 minutes
  • 150 PPM +/- 5 PPM for 10 – 50 minutes
  • 400 PPM +/- 10 PPM for 4 – 15 minutes

On the other hand, the false alarm (No alarm during these levels) is:

30 PPM +/- 3 PPM for 30 days and 70 PPM +/- 5 PPM for 60 minutes

These devices aren’t sensitive enough to alarm at low carbon monoxide levels and are useless for determining if a heat exchanger is cracked. They were never designed for this purpose.

Upgrading to a Low-Level CO Monitor or Combustion Analyzer

Those who understand a store-bought CO alarm isn’t sensitive enough for this test choose instead a low-level monitor or combustion analyzer. The reason? Those devices can read CO that is undetectable by traditional alarms.

Does a carbon monoxide reading from a supply register really point to a crack in the heat exchanger? No it doesn’t.

Unfortunately, many contractors make the same assumption with this test. The tester believes a cracked heat exchanger will cause flue gases to leak into the conditioned air stream. If CO is in the air stream, it will display on the low level monitor or combustion analyzer. They are designed to read low carbon monoxide levels. Sounds even better, right?

Perhaps. But does a CO reading from a supply register really point to a crack in the heat exchanger? No it doesn’t. This only shows CO is circulated by the furnace blower. It doesn’t identify the source. You still need to determine that. Remember, just because there isn’t any CO in the air stream doesn’t mean the equipment is safe.

Assumptions Made

This test usually leads  to two assumptions. The first assumption — cracks in a heat exchanger automatically allow flue gas to travel from the fireside to the conditioned air side of the heat exchanger. Sounds good in theory, but the opposite usually takes place.

Here’s why: The inside of a heat exchanger is under negative pressure. The outside of it is subject to the highest positive pressures in the HVAC system. Static pressure outside will be much higher than the pressure inside the heat exchanger.

CO - What Do You Know?

When it comes to CO, don’t make any assumptions. Knowledge and accurate measurements are your friends.

This is one reason to place roll outs along the burners of most modern furnaces. If a crack occurs, the additional heat will trip the roll out due to pressurization from the blower. Older natural draft furnaces didn’t have roll outs and you ended up with melted wiring.

So the second assumption made is that CO levels from flue gas will appear on the display of the test instrument. For a carbon monoxide reading to show up in a test like this, the levels in the flue gas would be extremely dangerous. When you mix two air streams, dilution occurs. The blower moves more air than the amount of flue gas produced by a typical residential furnace.

Some Other Assumptions

There are occasions when a cracked heat exchanger will produce excessive CO levels. This causes high carbon monoxide readings in the flue gas, NOT the conditioned air stream. The cause: blower air impinges the flame on surfaces it shouldn’t contact. You simply won’t pinpoint this issue testing for CO in the air stream.

So to discover this condition and verify the safety of a furnace or any fuel-fired equipment, measure the flue gases with proper test equipment. Improper testing leads to inconclusive results and potentially worse conditions.

Be sure you identify other potential sources of CO when hunting for cracks. By focusing on the furnace only, other items such as water heaters, gas ovens, and attached garages are easily overlooked.

Note: To clarify, cracked heat exchangers are never acceptable. They are a mechanical defect and need correcting. Don’t assume a cracked heat exchanger exists and replace it. Be a real troubleshooter and uncover the true cause of failure. This is what separates parts changers from technicians.

Finally, the answers are out there and together we will find them. If you need any additional information on this material or have any questions feel free to e-mail me at

You can follow any responses to this entry through the RSS 2.0 You can leave a response, or trackback.

Leave a Reply