Buildings with natural draft combustion appliances should be routinely tested to ensure that the spillage of combustion products into the building is unlikely. Combustion safety testing is critical because of the potential for severe health effects from improperly venting appliances, including carbon monoxide poisoning.

Spillage of combustion products into the building can be caused by a variety of conditions including:

• Blocked or partially blocked chimneys, vents, or vent connectors.
• Improper equipment installation.
• Cracked heat exchangers.
• Leaks in the venting system (disconnected flue pipes, open cleanest door etc.).
• Low vent temperatures.
• Combustion appliance zone depressurization. As buildings are made tighter, it becomes easier for exhaust fans and forced air system imbalances to create potentially hazardous depressurization conditions.  

Many cases of improperly venting combustion appliances have been related to depressurization (or negative pressures) in the room that contains the combustion appliance. Depressurization can be caused by exhaust fans, dryers, imbalanced forced air distribution systems, and forced air system duct leakage. As buildings (or combustion appliance rooms) are made tighter, these problems can be made worse, although very leaky buildings can also have venting problems related to depressurization. Figure 5 below estimates the amount of depressurization that can be caused by various exhaust fan flows. For example, from Figure 5 we can see that a 400 cfm exhaust fan will depressurize a 2,500 CFM50 building (or room) to approximately 3 Pascals. That same 400 cfm fan would produce over 10 Pascals of depressurization in a 1,000 CFM50 building.

The presence of code approved combustion air intakes does not ensure that venting problems will not occur.  Significant combustion room depressurization is frequently found even after code approved combustion air intakes have been installed. Passive combustion room air intakes typically do not provide sufficient airflow to relieve negative pressures caused by distribution imbalances, duct leakage, or large exhaust appliances. For example, a typical 6″ passive inlet can at best supply only about 50 cfm at a 5 Pa negative building pressure. And because passive air intakes are often poorly installed (i.e. many sharp bends, long runs), they typically provide much lower flows than designed. Building codes typically give little or no guidance on how one would design a combustion air opening when competing exhaust appliances are present (the 2000 Minnesota Energy Code is the only code we are aware of to give such guidance).

The only way to be reasonably sure that venting problems will not occur in a building is to perform combustion safety tests. Described below is a test procedure designed to locate existing or potential combustion safety problems in buildings. These procedures are offered only as an example of what other organizations in North America typically recommend for testing. The Energy Conservatory assumes no liability for their use, and contractors should have a working knowledge of local codes and practices before attempting to use the procedures outlined below.

If combustion safety problems are found, tenants and building owners should be notified immediately and steps taken to correct the problem including notifying a professional heating contractor if basic remedial actions are not available. Remember, the presence of elevated levels of carbon monoxide in ambient building air or in combustion products is a potentially life threatening situation. Building or duct sealing work should not be undertaken until existing combustion safety problems are resolved, or unless air sealing is itself being used as a remedial action.

This chart can be used to estimate the amount of house depressurization caused by operating exhaust fans.  To use the chart, find the intersection between the airtightness (CFM50) of the house and the cfm capacity of the exhaust fans in question.  The amount of depressurization caused by the fan(s) is read off the diagonal house depressurzation lines.  For example, a 400 cfm kitchen range hood operating in a house with an airtightness level of 2,500 CFM50 would depressurize the house by approximately 3 Pa relative to the outside.  This same fan operating in a 1,000 CFM50 house would produce over 10Pa of depressurization.  NOTE: This chart was generated by assuming that all houses have a “House Leakage Curve” with an exponent (or slope) of n = 0.65.

Test Procedures

This procedure is not intended to cover all circumstances you will find in the field. A basic understanding of the dynamic interactions between building pressures, air flow and mechanical system operation is required to fully utilize the procedures presented below. 

• Measure Ambient CO Level in Building: 
• Zero your digital CO tester outside before entering the building. CO tester should have 1 PPM resolution.
• Measure the ambient CO level in all occupied areas of the building. Be sure to measure ambient CO levels in kitchens and in combustion appliance rooms. 
• Investigate any ambient CO levels above 2 ppm.   Note:  Areas close to very busy streets may have ambient CO levels above 2 ppm.

Survey of Combustion Appliances:

Walk through the building and survey all combustion appliances including furnaces, water heaters, fireplaces, woodstove and auxiliary heating units, dryers and cooking stoves.

Write down the following information on a survey form:

• Location, type and input of combustion appliances.
• Signs of visible deterioration and leaks in flue pipes and connections.
• Presence of gas leaks, signs of spillage or flame roll-out.
• Location, size and operable condition of combustion air supply(s).
• Evidence of rusted interior surfaces of heat exchangers.

Gas or fuel leaks are a very serious safety problem requiring immediate remedial action.

Survey of Exhaust Fans:

Walk through building and note the location and rated capacity (or estimated capacity) of  all exhaust fans including kitchen and stove fans, bath fans, dryers, whole house vacuum systems, attic vent fans (not including whole house ventilation fans) etc.

Measure Worst Case Fan Depressurization: 

With this test procedure, the goal is to measure worst case depressurization in all combustion rooms with natural draft appliances and fireplaces. This measurement gives us an indication of the likelihood of exhaust and air handler fans causing the combustion appliances to backdraft and spill. The procedures below measure worst case depressurization under 3 separate operating conditions; running exhaust fans only, running exhaust and air handler fans, and running the air handler fan only. These tests are very sensitive to wind effects, and on windy days it can be very difficult to get accurate results.

Initial Preparation 

Close all exterior windows and doors and be sure furnace, water heater and other vented combustion appliances are off. Close all interior doors. Set up the digital gauge to measure the pressure difference of the combustion appliance zone (CAZ) with reference to (WRT) outside. Record the existing baseline building pressure. Note: The DG-700 gauge has a built-in “Baseline” feature which makes it easy to zero out the existing building baseline pressure and display the actual change in building pressure caused by fan operation. See the DG-700 manual for specific operating instructions.

1.  Exhaust Fans Only

Turn on all exhaust fans found in the survey above (for dryer, clean out lint filter before turning on).  Now determine the worst case position of interior doors with the smoke test below:

Smoke Test:  While standing in the main body of the building, squirt smoke under each door containing an exhaust fan (except the CAZ currently being tested).  If the smoke goes into the room, open the door.  If the smoke comes back into the main body of the building, keep the door closed.  Now squirt smoke under the CAZ door (while continuing to stand in the main body).  If smoke goes into the CAZ, leave the CAZ door shut.  If smoke comes back into the main body of the building, open the door.  

Measure the depressurization of the CAZ WRT outside caused by turning on the exhaust fans (i.e. the change in building pressure from the baseline condition). Depressurization should not exceed the appropriate House Depressurization Limits (HDL) listed in the table below. If it is windy, it is often necessary to turn fans off and on several times to obtain good pressure readings.   

Fireplace Zones:  For Fireplace Zones, repeat the same procedure and measure and record depressurization of fireplace zone WRT outside from exhaust fan operation. Depressurization should not exceed the appropriate HDL listed below.

2.  Air Handler and Exhaust Fans 

With exhaust fans continuing to run, turn on the air handler fan (note: air handler fan only, do not turn on burner) and close any supply registers in combustion appliance room. For both CAZ and Fireplace Zone tests, re-determine worst case position of all interior doors with the smoke test described above. If cooling is available, be sure air handler fan is running at high speed. Repeat worst case depressurization measurements. 

3.  Air Handler Fan Only

Turn off all exhaust fans and leave air handler operating (if cooling is available, be sure air handler is running at high speed). For both CAZ and Fireplace Zone tests, re-determine worst case position of all interior doors with the smoke test described above. Repeat worst case depressurization measurements.          

If the HDL are exceeded for any of the worst case depressurization tests above, pressure relief is needed. Pressure relief could include duct system repair, undercutting of doors, installation of transfer grills, eliminating or reducing exhaust fan capacity, or instructing homeowner on safe exhaust fan operation.  If negative pressures in the combustion appliance zone (or basement) are a function of return leaks in that area, check for leaks in the return ductwork, plenum, filter access door and air handler cabinet.  Pay particular attention to panned under floor joists (used as returns) as they typically have many leaks.

House Depressurization Limits (HDL) Table

Spillage Test (natural draft and induced draft appliances):

This test identifies actual spillage of combustion byproducts into the living space under worst case depressurization conditions. 

• With building set up in worst case depressurization mode (as specified above), fire up each combustion appliance.
• If appliances are common vented, conduct test on smallest input appliance first, then test with both appliances running.
• When burner lights, check for flame rollout (stand away from burner).
• Check for spillage (using chemical smoke) at the end of the spillage test period (see Table 10 below). For natural draft appliances, spillage is tested at the draft diverter. When an induced draft heating system is vented in common with a natural draft water heater, spillage is checked at the water heater draft diverter. For a single induced draft appliance, spillage is checked at the base of the chimney liner or flue, typically using the drip tee at the bottom of the liner.

Spillage Test Period Table

If spillage continues beyond the spillage test period, remove the negative pressure in combustion room by turning off fans and/or opening an exterior window or door.

Re-check for spillage. If spillage stops, there is a pressure induced spillage problem. If spillage continues, check flue and chimney for obstructions, and check compatibility of appliance BTU input with chimney size.

Spillage of combustion products beyond the spillage test period is a health and safety concern. If the problem is a blocked flue or chimney, or inadequately sized flue or chimney, consult a professional heating contractor. If the problem is pressure induced, provide pressure relief. Re-check for spillage following attempt to provide pressure relief. If spillage continues, contact a professional heating contractor to investigate the problem.

Draft Test (natural draft appliances):

This test measures flue draft pressure in the venting systems of all natural draft combustion appliances under worst case depressurization (not to be done for sealed combustion or induced draft appliances).

• Drill a small hole in the vent pipe approx. 2 feet downstream of the draft diverter or barometric damper. Insert a static pressure probe.
• Measure draft pressure (vent WRT combustion room) with Magnehelic or digital pressure gauge after 5 minutes of operation.
• Compare measured draft with minimum draft pressures below:

Minimum Draft Pressures Table

If measured draft is below the minimum draft pressure above, check for flue or chimney obstructions, disconnected vents, open chimney cleanout doors etc.. Also remove sources depressurization (e.g. turn off exhaust fans) and test again to determine if CAZ depressurization is contributing to poor draft.

Carbon Monoxide Test:

This test measures carbon monoxide levels in all operating combustion appliances.

After 5 minutes of appliance operation, measure the CO level in the flue products of all combustion appliances.

• CO should be measured before appliance draft diverter, or barometric damper.
• CO levels should be below 100 ppm in all flues.
• For gas stoves, measure CO from oven exhaust port and 3 feet above burners with all burners running. CO level should be below 50 ppm.
• If CO found in gas stove, re-measure ambient kitchen CO after 10 minutes of stove operation.

The presence of CO and spillage requires immediate remedial action.

Heat Exchanger Integrity Test (Forced Air Only):

This test is used to determine if a crack or hole is present in the furnace heat exchanger. A crack or hole could allow products of combustion into the building, and/or promote carbon monoxide production through flame distortion and impingement. There are 3 main types of tests which can be performed:

1.  Flame Distortion Test 

This test involves watching the furnace flame when the furnace air handler first turns on. Any distortion of the flame indicates a hole or crack in the heat exchanger. This test can be done in conjunction with the flame rollout component of the spillage test. Another method for conducting a flame distortion test is to slowly extend a match up and down into each combustion chamber with the burner off and the air handler fan on, and watch for movement of the flame head.

2.  Blocked Flue Test

With the furnace off, block the flue ports leading from the combustion chamber to the draft diverter or barometric damper.  Squirt smoke into the combustion chamber. Turn on the furnace fan and watch to see if the smoke is disturbed when the fan comes on. Smoke movement indicates a hole or crack in the heat exchanger.

3.  Tracer Gas Test

A number of testing procedures exist for injecting a tracer gas into the combustion chamber (usually with the furnace fan off) and then measuring or detecting the tracer gas on the warm air side of the heat exchanger.

If any of the above heat exchanger tests provides a positive indication for a cracked heat exchanger, immediate action should be taken to notify the residents of the potential danger, and a professional heating contractor should be contacted to investigate the problem.

Turn off fans and return appliance controls to their original settings once the test procedures have been completed.

Special thanks to Advanced Energy, Sun Power and the Center for Energy and Environment (CEE) for their work in developing and refining the combustion safety test procedures above.