LARGER VIEW of heating furnace parts and air flow, showing the heat exchangerFurnace Heat Exchanger Leaks
Heat Exchanger Testing & Safety Inspection Procedures

  • HEAT EXCHANGER LEAK TEST - home - CONTENTS: how to inspect and test furnace heat exchangers for leaks & CO Carbon Monoxide Hazards; Photographs show clues indicating leaky, dangerous furnace heat exchangers. Furnace heat exchanger leak testing, A Complete List of Methods Used to Find or Test for Leaks in a Furnace Heat Exchange. 13 leak detection procedures for furnace heat exchangers. Heat Exchanger Testing and Test Devices: Who's Right? Recommendations for reliable furnace heat exchanger testing procedures
  • POST a QUESTION or READ FAQs about furnace heat exchanger leaks, cracks, damage, inspection & testing
InspectAPedia tolerates no conflicts of interest. We have no relationship with advertisers, products, or services discussed at this website.

Furnace heat exchanger inspection, troubleshooting, and leak testing guide: This heating system test article describes how to inspect furnace heat exchangers for leaks.

We compare and evaluate the reliability of all of the various furnace heat exchanger testing methods, we explain just how much leakage is "acceptable" by industry standards, and we conclude with recommendations for reliable heat exchanger testing and inspection.

Green links show where you are. © Copyright 2017, All Rights Reserved.

How to inspect a furnace heat exchanger for damage or leaks & carbon monoxide CO gas hazards

Visual Inspection of the Furnace Heat Exchanger can Detect Some Leaks but Not All

Photograph of rust on a furnace warning of unsafe heat exchanger. Photograph of furnace rust damage.

[Click to enlarge any image]

We recommend that you never rely on visual inspection alone to determine the safety of a furnace heat exchanger. Readers of this document should also

More about carbon monoxide - CO - is at CARBON MONOXIDE WARNING.

include the text from historical articles on methods used for testing furnace heat exchangers for leaks, and the allowable or standards for heat exchanger cracks, holes, leaks, or carbon monoxide hazards from such leaks.

Burns & soot in a heat exchanger (C) Daniel FriedmanWatch out: Dangerous carbon monoxide gas leaks, potentially fatal, can be present intermittently depending on variations in heating system operation and building conditions.

This article series answers most questions about central heating system troubleshooting, inspection, diagnosis, and repairs.

We describe how to inspect, troubleshoot and repair heating and air conditioning systems to inform home owners, buyers, and home inspectors of common heating system defects.

Readers of this article should also

Those considering using instruments to test heat exchangers for leaks should review RECOMMENDATIONS for GAS LEAK TEST INSTRUMENTS & gas detector tubes for indoor gas level tests.

In response to a reader who wrote that they have a G14Q3, installed in 1988, Lennox Pulse Furnace and who was wondering if there are any visual inspections for signs of a defective heat exchanger we provide the furnace heat exchanger inspection suggestions described in this article.

The photographs shown just above are two examples of rust and flame marks on a furnace that would be a basis for further inspection of the condition of the heat exchanger for cracks, rust perforation, or other unsafe conditions. But heat exchanger leaks can occur in a variety of locations and parts, including:

A Complete List of the Methods Used to Find or Test for Leaks in a Furnace Heat Exchanger

TIF 5000 refrigerant gas leak detector for halogensThe following is a catalog of all known methods for testing residential heating furnace heat exchangers for dangerous flue gas or carbon monoxide leaks. Contact us to add methods or critique. The list is arranged alphabetically rather than in any order of recommended procedure.

Watch out: SAFETY WARNING: Any evidence of furnace heat exchanger damage or of carbon monoxide or flue gas leaks should be taken seriously and those heating systems should be immediately checked by a professional.

Smoke test example (C) Daniel Friedman

Photograph of a rusted out furnace

Photograph of a rusted out furnace

Photograph of a rusted out furnace heat exchanger showing flame anomaly

Visual Inspection of the Furnace Heat Exchanger Alone is Unreliable for Detecting Furnace Heat Exchanger Leaks or Damage

Relying on visual inspection of heat exchangers: Given that a lot of the heat exchanger surface simply cannot be seen without completely disassembling the system, we would not rely on a visual inspection alone to decide if a system was damaged or not. There are other Tests using pressure testing or more commonly, tracer gas testing, that are more reliable.

Relying on gas detection instruments: Relying on gas detection instruments alone, without a visual inspection of the system is also dangerous and can falsely indicate that no problem in present when in fact the heating system is unsafe.

We discuss the reasons for this at SOME WARNINGS about relying on instruments for detection of hazardous gases in buildings.

Install CO detectors in buildings, as well as smoke detectors. Ultimately the combination of expert inspection, testing, and the use of carbon monoxide detectors and smoke detectors will make a significant improvement in the safety of any home heating system.

Smells and odors in the building could be an indicator of heat exchanger leaks -

A Review of 13 existing leak detection procedures

Gas utility servicemen use a variety of procedures in attempts to identify potentially dangerous leaks in furnace heat exchangers. The procedures would be used in response to service calls for

Thirteen different procedures [for testing furnace heat exchangers for leaks] were identified after surveying utility companies. Many of these methods do not give reliable results, thus often more than one method is used. Some methods are corrosive to the heat exchanger.

Others are performed under unrealistic conditions (such as increased pressure in the heat exchanger), or are so sensitive that they indicate leakage even in new heat exchangers.

Visual Inspection of the Furnace Heat Exchanger for Cracks, Burns, Leaks, Damage

One of the most common methods used is a visual inspection of the heat exchanger with a strong light. The serviceman must then make a judgment as to whether any cracks or holes he finds are large enough to warrant replacement of the furnace. Generally any fault is reason for recommending that the home owner obtain a second opinion of the need for replacement of a heat exchanger.

Smoke Bombs for Heat Exchanger Tests

Smoke test example (C) Daniel FriedmanThe use of smoke bombs is the second most common practice. With the flue outlet and burner access opening blocked a smoke bomb or candle is ignited in the heat exchanger.

Observations for smoke are then made in the circulating air side of the heat exchanger. When the bomb explodes a positive pressure is created inside the heat exchanger forcing the smoke out of very small cracks.

This makes the test very sensitive. Precautions are also needed with this method so the smoke does not get into the house where it may stain certain paints, fabrics, and tiles.

An air analysis test may be used in addition to the smoke test. This test detects combustion products in the circulating air stream as an indirect indication of heat exchanger cracks. If high readings are obtained there is no doubt that the heat exchanger has a crack or fault.

With slight increases a more thorough examination of the heat exchanger may be required. The drawbacks of this method are that it is complicated and time consuming to perform. Also the detection device is expensive.

Editor's note: See Matzen and other articles [below] for an evaluation and survey of this equipment. Significant to Matzen et al, DeWerth does not say these tests are necessarily invalid. Odorants, such as sulphur candles and oil of wintergreen have been used for detecting leaks.

A small quantity of the odorant is introduced into the combustion side of the heat exchanger; the hot air registers in the house are then checked for the characteristic odor. Any odor in the circulating air would indicate a heat exchanger leak. With the sulphur candle this test can be quite reliable, but very unpleasant smelling. The wintergreen is more pleasant, but the odor clings to the serviceman's hands and clothes, making the test unreliable.

Tracer Gas Tests for Furnace Heat Exchanger Leaks

Another test procedure is to release a small amount of carbon monoxide (CO) in the hot heat exchanger and then check for high concentrations of CO in the circulating air stream. The detector needed for this method is expensive and the use of CO is dangerous. Freon has been used as a tracer gas with a halogen leak detector.

Freon gas is released in the combustion side of the heat exchanger and if the halogen leak detector senses halogens in the circulating air stream a leak is present. A problem with this method is phosgene gas (a deadly poison) can be generated if the freon passes through a flame. Also, freon is heavier than air.

Spray Dyes for Heat Exchanger Tests

A spray can of fluorescent solution is marketed which is sprayed into the heat exchanger with a suspected leak and penetrates even the finest crack. Then the crack is detected using an ultraviolet light; however the heat exchanger must be accessible for close examination of all surfaces.

Chemical Smoke Tests of Furnace Heat Exchangers

The American National Standards for Gas-Fired Central Furnaces (Z21.47-1978) outlines a test method using a fuming or smoking material such as titanium tetrachloride. The material is introduced into the combustion chamber, and if combustion products are discharged through door cracks or other openings, their presence will be revealed by observing the smoke. A drawback to this method is that titanium tetrachloride is very corrosive.

Another widely-used procedure involves spraying a sodium salt solution into the burner flames and checking the circulating air stream with a propane torch for the presence of sodium ions. [Detection of Cracked Heat Exchangers in Warm Air Furnaces , J.F. Wunderlin, Wisconsin Gas Co., GV-6, June 1978.]

If the blue flame of the torch turns yellow, this indicates the presence of sodium ions and a leak. As the solution is sprayed into the burner flame, visual inspection of the heat exchanger surface below the burner port is needed.

If a solution of sodium chloride (table salt) is used, corrosion of the heat exchanger can be accelerated. Sodium bicarbonate salt solutions, on the other hand, are non-corrosive. The drawbacks to the sodium ion tracing method are that dust in the air can be mistaken for the sodium ion and acceptable leakages may be detected as the test is very sensitive.

A procedure similar to the sodium ion tracing is to trace the lithium ion, with the same drawbacks.

All of these methods have some undesirable drawbacks. The ideal method for detecting the leakage of flue gases into the circulating air stream would not be harmful to the furnace components of the home environment, would be performed under realistic operating conditions, and would not detect small amounts of acceptable leakage.

Furnace Heat Exchanger Pressure Measurements for Leak Detection

The pressure drop across the furnace heat exchanger surface has a significant effect on the tendency of flue gases to pass from one side of the exchanger to the other. Only if the flue gas side is more positive than the circulating air side will there be a tendency for flue gases to leak into the circulating air through any cracks or corrosion holes in the heat exchanger.

In order to determine the actual operating pressure inside and outside the heat exchanger, measurements were made with four furnaces with typical heat exchangers. ["A Simple Test That You Can Use to Check a Furnace for Leaks," American Artisan, September 1966.]

Two of the furnaces were equipped with atmospheric burners, one with an induced draft system, and the fourth with a power burner. Flue gas pressure taps were attached at two-inch intervals up the side of the heat exchanger from the burner port to the flue outlet. Circulating air taps were located adjacent to the flue gas taps.

Pressure measurements were taken with and without the circulating air blower in operation. The results showed that with atmospheric burners the average pressure on the flue gas side with the burners operating is about 0.02 inch water column (w.c.).

The average pressure on the circulating air side of the heat exchanger is 0.3 inches w.c. when the blower is operating. Thus the air side of the heat exchanger is more positive than the flue gas side when the circulating blower is on. If there were a hole leakage would be from the air side to the flue gas side.

When the blower is off the air side pressure is less positive than the flue gas side and leakage would occur from the flue gas side to the air side. This condition exists only during the short heat up period before the blower starts.

An induced draft system has a negative pressure of about 0.2 inch w.c. or less on the flue gas side of the heat exchanger at all times. The flue products are drawn through the combustion chamber causing the flue gas side to always be under a negative pressure. A positive pressure on the air side of the heat exchanger, caused by the circulating air blower, further insures that the flue gases stay in the combustion chamber.

With a power burner system the flue gas side of the heat exchanger must have a positive pressure of as much as 0.4 inch w.c. Thus, the pressure on the flue gas side is more positive than the air side, whether the circulating air blower is on or off. The only exception is the top of the heat exchanger where the circulating air blower may impinge directly onto the heat exchanger.

Air side pressure would be more positive than the flue gas side at these points. If there were a hole or crack, flue gases would leak into the circulating air stream at all times except if the hole or crack were at the top of the heat exchanger.

Table 1 summarizes the various systems and the leakage potential due to the pressure drop across the heat exchanger. In a furnace with atmospheric burners flue gases could only leak into the circulating air system if there were a hole or crack during the time the blower was not operating.

Induced draft furnaces would not allow any leakage through the heat exchanger as they operate under a negative pressure. A power burner system is more critical in that the flue gas side of the heat exchanger is always more positive than the circulating air side when the furnace is operating. Thus, if there were a hole or crack flue gases could leak into the circulating air stream.

This analysis of flue gas leakage based on the pressure difference across the heat exchanger is true for relatively small holes. For heat exchangers with large perforations flue gases may leak regardless of the pressure differential. Large perforations would be detected by visual observation of the heat exchanger or by the flame pattern. No further test procedure would then be needed.

Test methods should not subject the heat exchanger to abnormally high pressures, and should not detect insignificant leakage which may occur from pin holes in welded seams or other manufacturing imperfections.

Heat Exchanger Testing and Test Devices: Who's Right?

- Daniel Friedman

For an accurate understanding of how we should look for an unsafe heat exchanger the pass-fail criteria must first be defined. Matzen's heat exchanger test procedure looks for levels of CO as low as 2-3 ppm at the heat exchanger.

DeWerth's heat exchanger test procedure looks for levels of 200 ppm at the heat exchanger. Can we reconcile these views? Both articles are carefully reasoned; both articles cite authoritative sources for exposure levels. How are readers to interpret these differences?

The Matzen and DeWerth articles provide food for serious thought. Here Journal staff suggests some interim opinions and some questions. The reviewers end with a call for inspection priorities.

Matzen's Test Method for Heat Exchanger Leaks

Matzen calls for testing at very low levels of CO, in part because of the concern, beyond that of CO hazards, for the presence of formaldehyde (HCHO) which may be present in flue gases, and to which people have reactions at lower levels than those permitted for CO.

He describes using a calibrated CO monitor, testing for CO, with the presumption that if CO is present, HCHO may be as well.

Matzen recommends use of a CO monitor with digital readout to permit explicit level monitoring and reporting. He measures both at the heat exchanger, and in the house air - an important additional measurement which looks at actual levels rather than an inferred level based on measurement only at the heater.

These monitoring techniques will identify early failure where small quantities of 2-3 ppm are in evidence. Grossly damaged heat exchangers may admit 10-30 ppm in the room air in one cycle of the furnace, in his experience.

However this procedure does not deal with the question of actual CO exposure in the house at equilibrium. At such an inspection, data regarding actual house air changes per hour is generally not obtained.

Further, DeWerth argues that very low levels of CO found at the heat exchanger may be within acceptable limits and may not indicate equipment failure.

DeWerth's Test Method for Heat Exchanger Leaks

DeWerth's test for flue gas leaks at the heat exchanger of a gas-fired appliance looks for a level of 200 ppm flue product. Readers should pay particular attention to the careful sequence in which DeWerth develops his argument.

Is This Heat Exchanger Safe?

In testing any hypothesis, such as "This Heat Exchanger is Unsafe" there is more than one type of error possible.

Accepting for use a heat exchanger which is dangerous is not permissible since human life could be at risk. For this reason, DeWerth would be expected to be quite cautious in choosing the assumptions for his methodology.

Readers should look for, and will find, some examples of caution and significant margins for error (100% safety factor is commonly used) in DeWerth's assumptions throughout his analysis. DeWerth himself, has not presented the cumulative effect on risk of all of the conservative steps in his process.

Rejecting for use a heat exchanger which is safe, and which meets or exceeds industry manufacturing standards is also an error. While the consequences are less dire, the consumer's interests are not served by requiring replacement of an acceptable heat exchanger. Few heating companies will actually make such a replacement. Usually the entire furnace is replaced, for a typical cost around $2000.

Derivation of Test Criteria for Heat Exchanger Leaks

DeWerth's analysis does not test house air. Rather, his measurement standard, 200 ppm to be tested at the heat exchanger is derived from answering the following:

A Test Methodology for Furnace Heat Exchangers - Conclusions

DeWerth's description of Visual Inspection and Flame Pattern Inspection procedures should be studied by every home inspector, regardless of whether or not s/he uses other more sophisticated techniques. These first two, of DeWerth's Three Step Procedure, are visual-only, and should be followed by anyone inspecting a gas-fired heater in compliance with ASHI Standards of Practice.

DeWerth's third step involves purchase of special equipment, as does that of Matzen. However, in addition to a flue gas analyzer device, the DeWerth method requires purchase and use of a tracer gas and a gas regulating device.

The procedure is involved and time consuming. It's an appropriate procedure for a heating company expert who is called in to offer a second opinion regarding the condition of a heat exchanger.

Combining Both Methods of Testing for Heat Exchanger Gas Leaks

The choice to offer services which exceed the ASHI Standards of Practice is permitted by the Standards. Inspectors should beware of offering advanced services if they are not qualified. However this should not be taken as a suggestion by ASHI that an inspector may not make him/herself qualified in additional areas.

Every inspector should follow the first two (visual) steps of DeWerth's procedure

Qualified inspectors may follow a combination of the DeWerth Procedure with the Matzen suggestions for use of a CO monitor, substituting the CO Monitors for the more cumbersome tracer gas, as a step beyond visual inspection of heat exchangers.

CO Monitors were not dismissed as inaccurate. DeWerth raised questions of cost (now add the cost and time to use a gas cylinder/regulator) and a question of rejecting a good heat exchanger by a test which is too sensitive.>

Digital readout CO Monitors are now available. The inspector is free to elect the American Gas Association's standards of acceptability (or others if identified) while using this simpler and faster test instrument.

Inspectors who prefer Matzen's standards of acceptability will report any CO levels found in a building.

Inspectors who follow DeWerth's standard, itself a cautious one, would not call out a heat exchanger as "failed" if it leaked less than 15 ppm CO into the house living area at equilibrium conditions.

Based on DeWerth's explanation of the mechanism of failure and leak rates, a very low leakage detected

in the house air should report it as a concern.

A call for Priorities:

reviewers (A. Carson and D. Friedman) agreed that some perspective is needed in considering heat exchanger tests. Little data was provided, and only by Matzen, regarding deaths caused by faulty heat exchangers.

It is possible that some deaths and many complaints traced to high CO levels in homes are due to backdrafting and spillage from blocked or otherwise improper flues and vent connectors. Similar complaints may be attributed to inadequate supply of fresh makeup air in buildings, often seen as high CO 2 .

The concern for testing for leaks in heat exchangers and in some cases the methods used in the field are often poorly reasoned. If ASHI professionals are to expend training and inspecting time on life-safety issues perhaps we should first obtain loss data which would allow us to properly prioritize our efforts.

Virtually every home safety survey cites tripping hazards as the number-one cause of injuries in homes. Future research (and Journal articles) ought to address this question of priority. But if you test, do it correctly!

As a final note, report the authoritative sources used for your standards, identify your equipment, and state the procedures used and the quantitative measurement obtained. This may avoid embarrassment and will add credibility to your findings.

For inspectors who feel that documenting these data is too much work, we suggest skipping instrumented tests entirely.

It's little effort to pre-print a standard explanation identifying the test instrument, it's levels of performance, and the inspection standard used. The only "custom" writing necessary is the actual level measured in the building.


Continue reading at HEAT EXCHANGER LEAK ALLOWED or select a topic from closely-related articles below, or see our complete INDEX to RELATED ARTICLES below.


Or see

Furnace Heat Exchanger Articles

Suggested citation for this web page

HEAT EXCHANGER LEAK TEST at - online encyclopedia of building & environmental inspection, testing, diagnosis, repair, & problem prevention advice.



Or use the SEARCH BOX found below to Ask a Question or Search InspectApedia


Frequently Asked Questions (FAQs)

Click to Show or Hide FAQs

Ask a Question or Search InspectApedia

Use the "Click to Show or Hide FAQs" link just above to see recently-posted questions, comments, replies, try the search box just below, or if you prefer, post a question or comment in the Comments box below and we will respond promptly.

Search the InspectApedia website

Comment Box is loading comments...

Technical Reviewers & References

Click to Show or Hide Citations & References

Publisher's Google+ Page by Daniel Friedman