Procedures/Methods used during Heat Exchanger Inspections

Heat exchanger inspections have almost become a science unto themselves. In many organizations, finding a cracked heat exchanger constitutes a really good day, and sometimes a boost to the wallet.

The following are various procedures and test methods to tell how to identify a cracked furnace heat exchanger. Note that NONE of the methods are 100% reliable. That’s why there are multiple procedures listed within each method.

Ht_xchgr_viewKeep in mind that the whole point of a heat exchanger check is safety. However, the current mind set in the industry is that “every heat exchanger crack is dangerous”. AGA and GAMA insist that even a hairline crack in a heat exchanger constitutes a defect and requires replacement.

The prevailing viewpoint is that any hole or or heat exchanger crack can and will get larger and when it does, it is an immediate hazard. Taken as a “scientific” observation, that is true. However, as a practical matter, few cracks have proven to be real hazards that cause or pass carbon monoxide or create a fire or explosion.

This ContractorTalk forum entry is a compilation of a number of heat exchanger tests that can be used to identify a cracked fire box or if a heat exchanger has holes. It describes most of the available tests and has images of some of the testing tools and methods of how to check for a cracked heat exchanger.

ACHR News – Heat Exchanger Test

Heat Exchanger Testing ACHRNewsHere’s a 2006 article from that describes how to test a heat exchanger and describes other air requirements that should also be checked. The author does a good job covering all the basics, however, he comes to the same faulty conclusion about heat exchanger holes and cracks in residential furnaces that befall most mechanics.

The author describes lab tests where 1/8″ holes were drilled into various locations in a heat exchanger. He then describes the changes in O2 readings seen on the combustion analyser. He readily admits that the rise in O2 is due to air entering the heat exchanger through the hole. He even goes on to suggest that placing an obstruction (blocking plate) in the discharge air stream will increase static pressure in the furnace and increase the amount of air passing INTO the heat exchanger.

His conclusion is that any change in O2 readings when the blower comes on is an indication of a defective heat exchanger. At the very end of the article he explains that this information should be noted on a service ticket and signed by the customer as proof that the customer was notified of the defect.

The very last bullet point at the bottom of the article directs the mechanic to explain the health risks to the customer. Sub-point “a.” says “A defective heat exchanger allows flue gasses to enter the building.”

Since the author stated that the static pressure in the furnace pushed air INTO the drilled holes of the heat exchanger, how are flue gasses supposed to enter the building?  The CO laden products of combustion that accumulate in the front of the furnace and may pool in the utility room, but it is not being picked up and delivered to the indoor air stream unless it is pulled through return-air openings around the furnace.

This is the conflict in thinking that most folks ignore.


The first thing you’ll look for when inspecting a heat exchanger is whether there is a change in the flame and combustion gasses when the burner is firing and the indoor blower comes on. That implies that air is passing from the indoor air stream to the “fire-side” of the heat exchanger, which is true.

And this is my point – in furnaces built since the 1970’s air ALWAYS passes from the indoor-air side of a heat exchanger crack TO the fire-side. 

It’s impossible for it to happen the other way around when duct work and an air-conditioning coil is connected to the furnace. Short of a concussive ignition, there’s no way that a burner flame will ever create enough pressure inside a heat exchanger to overcome the static pressure on the outside of the heat exchanger that is created by the indoor blower.

On natural-draft furnaces, the flue is always “negative” with respect to the inside of the building. If it’s not always negative, it’s not a flue, it’s a hole in the roof. (If the flue doesn’t always draw, it has been mis-designed, incorrectly installed, or the building structure itself has been changed. A properly installed flue always drafts up and out through the roof.) So, since the flue pulls the burned gasses from the furnace up and out of the structure, how is a hole or crack or big split going to allow flue gasses to pass from the fire side to the indoor air side of the heat exchanger?

At best, the burners will generate .02″ to .04″ water column presure within the heat exchanger which will be hot gasses that will quickly rise and exit via the draft diverter and into the flue. This path will offer less resistance than the pressure needed to push through a split in the heat exchanger. When the indoor blower starts, it creates static pressure inside the furnace cabinet and outside the heat exchanger that pushes large amounts of air INTO the heat exchanger cracks. However, when the indoor blower first starts, there’s a brief moment when air is moving quickly through the heat exchanger, before static pressure has built up, when flue gasses can be pulled through a crack. Most of the time it is imperceptible, it happens so quickly. But, there are circumstances where long duct runs, belt-drive blowers, soft-start or electronic blower motors, and blowers with failing capacitors can take a second or two to get up to speed and cause a delay in establishing static pressure in the furnace.

Furnaces with draft inducers are even less likely to ever allow flue gasses to pass to the indoor air stream. The pressure inside the heat exchanger is ALWAYS negative. Flue gasses are pushed by the draft inducer up the flue. That’s why big gapping holes in tubular heat exchangers cause flame rollout but don’t contribute to flue gasses getting into the indoor air stream.

Power draft furnaces (in shot power burners) are a different story. They positively pressurize the inside of a heat exchanger and can definitely push flue gasses (and potenially CO) into the indoor air stream. Mechanics have to recognize the kind of equipment they’re servicing and adjust their procedures accordingly.

Note that the tests were meant for “standard” furnaces, they don’t apply to Lennox Pulse Furnaces, commercial furnaces, duct heaters, etc.

Pay Attention to the Installation

Pay attention to how the furnace is installed and used. If a furnace is simply placed in an area with NO duct work or air-conditioning coil on it, then this entire discussion is moot. Since there’s little to no air restriction to the discharge air, there will be little to no static pressure inside the furnace. In this case, the indoor blower air stream can actually suck combustion products through the crack and into the inside of the building.

This same caveat applies to furnaces connected to oversized duct work. If static pressure is not created inside the furnace cabinet, then heat exchanger cracks can allow flue gasses into the indoor air. If the furnace has an air-conditioning coil or properly sized duct work that creates static pressure in the furnace, heat exchanger cracks may upset combustion a little, but they are not the eminent danger everyone wets their pants about.

Carbon monoxide definitely can be pulled from a heat exchanger crack in unit heaters, horizontal duct furnaces and direct-fired make up air heaters.

Walking Past Potentially Fatal Problems

The interesting part is to note how many times mechanics zero in on heat exchanger cracks, spend inordinate amounts of time looking for them while ignoring the fact that they are servicing a 100,000 btu furnace connected to a flue with a 35,000 btu water heater in a utlity room with a 30,000 btu gas dryer that measures 8ft x 12ft with an 8ft drywalled ceiling and a solid door attached to the opening of the room. When their customers drop dead on a frigid cold night because someone closed the utility room door, they can at least report that the heat exchanger was intact.

Side note – it takes more fresh air than you think to support gas appliances.

An induced-draft furnace requires 15 cubic feet of free air for every 1 cubic foot of gas that it burns. A cubic foot of gas is around 1,000 btuhs, which means a 100,000 btuh furnace needs 100 x 15 or 1500 cubic feet of clean air available for every hour of operation.

If the furnace is tucked in a utility room that’s 8 feet wide and has an 8 foot high ceiling, then the room better be 23 feet long, or the furnace will run out of combustion air when it operates continuously for a full hour. Add a water heater connected to the same flue, in the same room, and now that utility room has to be even bigger.

We’ve weatherized and built homes so tight that there is very little infiltration air. During extremely cold weather occupants avoid opening outside doors and certainly keep their windows shut. This means that there are fewer air changes in the home and less chance for oxygen to be replenished. Because homes have less infiltration air, issues with a shortage of combustion air and problems with flues have become a bigger problem.

Even the AGA doesn’t recognize the “conflict” between their directions to watch the burner flames on blower start up and the conclusion that a crack in a heat exchanger will somehow pass flue gasses and potentially CO to the indoor air stream. Here’s a link to AGA’s test procedures. They’re suggesting the use of a tracer gas of 14.3% non-odorized methane in nitrogen and a 200ppm calibrated combustible gas leak detector.

Their “scientific results” are that the procedure was field tested by 7 major gas utlities during the 1982-83 heating season and was reported by them to be a major improvement over other methods. (I wonder what that means?)

AHRI – Air-Conditioning, Heating and Refrigeration Institute

AHRI also published a test sheet. It uses a CO tester on the indoor air stream as a primary test method. Here’s a link to the AHRI web site with a link on it to the “fact sheet”. I wonder how they conclude carbon monoxide gets into the indoor air stream?

HARDI – Heating, Air-Conditioning, Refrigeration Distributors International

HARDI gives a good explanation of the tracer gas method, and admits that other gasses can interfere with the test and that the test itself is not conclusive.

Here’s why you’re checking for heat exchanger cracks, big cracks. Someone bypassed a roll-out switch to keep a furnace running. The new homeowner found the roll-out switch and cracked heat exchanger himself when the furnace quit working.

My concern is NOT that you have a heat exchanger crack, and I don’t mean to imply that heat exchanger cracks are OK to live with.

My issue is that you’re being lied to, and at the same time, most mechanics walk past the things that might kill you and your family.

If you have a cracked heat exchanger, you need to replace it. But, you also need to make sure (or have your HVAC mechanic insure) that your home’s appliances and construction and flue and combustion air will safely support your gas appliances.

In almost every carbon monoxide poisoning case I’ve looked at, the problem was declared to be the furnace or boiler or water heater. But on further investigation, the actual cause of the poisoning turned out to be blocked flues or limited combustion air or improper use. And, in each of these cases, the HVAC contractors that serviced the equipment said they had checked the flues and looked at the duct work and considered the HVAC systems and boilers safe.

Here’s a 2011 article “12 Must-Do’s On a Furnace Clean and Check” in a national trade magazine that lists the checks that service technicians should be performing on a furnace “clean and check”. The checks, as listed, are necessary and should be done. However, the article also illustrates what I mean about equipment-focused service versus application awareness.

  • Nothing in the article mentions the need to confirm that there is adequate combustion air available to support all the gas appliances in the area.
  • Nothing is stated about confirming flue draw, size or correct installation for appliances using a standard Class B metal flue.
  • Nothing says to inspect the flue for damage – broken joints, incorrect pitch, blocked or crushed flue cap, excessive rust or mineral deposits (high condensation of flue gasses.)
  • Nothing is stated about checking the indoor blower wheel for excessive dirt stuck on the blades (which reduces overal air flow).

If technicians follow the 12 points and only check the equipment, they could be ignoring potentially dangerous problems with the building that may have an adverse affect on gas appliance operation.

Although this is an article about a boiler, it shows what can happen when the entire combustion air zone is not evaluated. In this case, a spark-ignition boiler on a snow-melt system kept locking out and almost killed the homeowner. In fact, it could have killed the representative looking at the job.