The following case studies detail duct leakage problems in homes with chimneys and tests that contractors can perform to detect and prevent unsafe system interactions between forced-air furnaces and fireplaces.

Case 1: Combustion gas release

Last fall, the Smiths purchased an $80,000 new, triple-wide manufactured home with a wood-burning fireplace in the family room, and installed it on their land in the country. It was constructed according to Housing and Urban Development (HUD) standards, like millions of other homes across the United States. The heating system was a forced-air electric furnace.

After the Smiths moved in, they woke up in the middle of the night smelling smoke throughout the home. They had burned a fire in the fireplace earlier, but they hadn’t noticed any odors. At some point after they went to bed, the fire died down and combustion gases began spilling into the home instead of going up the flue.

The incident was potentially serious, but everyone got out of the house safely.

The fireplace was approved for manufactured homes and included the required outdoor source of combustion air ducted directly to the combustion chamber. According to HUD codes, the home had been properly constructed and supplied with approved appliances.

On-site tests indicated that supply ducts were leaky. Using a digital manometer, technicians measured air pressure in the home and compared it with outside air pressure when the forced-air system ran. Forced-air furnace operation lowered indoor air pressure by 17 Pascals (1 in. wc = 249.1 Pascals) with reference to outdoor air.

High pressure flows to low pressure, so technicians reasoned that outside air (higher pressure) had flowed down the fireplace chimney in response to the home’s lower air pressure environment, spilling flue gases into the home in the process. An actual burning test confirmed that operating the forced-air system caused flames in the fireplace to deflect and flatten out as air came down the chimney.

Case 2: Fireplace fume release

Last summer, the Jones family moved into their $225,000 home built in a high-end subdivision on a new site. The home’s features included a wood-burning fireplace in the living room and a gas forced-air heating system.

The homeowners began using the fireplace as the weather grew cooler, and they noticed fireplace odors even when the fireplace was not in use.

One night they burned a fire, left it to burn down, and went to bed, closing the doors to their bedroom. Later family members awoke to the acrid smell of fireplace fumes.

The fumes were not visible, so the smoke alarm did not sound. However, the odors were carried all through the home by the forced-air system. One of the homeowners went to the hospital to be treated for smoke inhalation and the home retained smoke odors for months, in spite of clean-up efforts.

Unsafe house system interactions

Testing showed that both of these incidents were the result of what building scientists call “unsafe house system interactions” — in these cases, interactions between a forced-air heating system and a naturally-vented, wood-burning appliance.

Testing showed that, because of duct leakage, forced-air system operation lowered air pressure inside the home, with reference to outside air pressure. In the site-built home, the effect was minor (0 to -1 P); whereas in the manufactured home, the effect was much stronger (-17 P).

In the site-built home, when bedroom doors were closed, air pressure in the fireplace zone fell to -3 P with reference to outside air.

Check for duct leakage

Duct leakage was a component of both of these potentially fatal accidents.

When supply ducts leak, the furnace blower pumps air out of the home — through the leaks — faster than it can be replaced. Essentially the furnace blower functions like a large exhaust device, removing air from the home and pumping it into building cavities that contain ducts. When air is pumped out faster than it can be replaced, air pressure in the space drops with reference to outside air.

In response to lower air pressure inside a home, higher pressure outdoor air will come into the home through any opening it can find in the building envelope. In many homes, the largest hole and path of least resistance into a home is down a combustion appliance flue.

There are a number of tests contractors can do to detect and avoid the consequences of unsafe system interactions between forced-air furnaces and fireplaces. These tests should be employed when plans show forced-air systems and fireplaces will be included in the home.

Although duct tightness is perceived by many to be an efficiency issue, national training centers in Florida and North Carolina stress health/safety issues and air-quality impacts associated with duct leaks.

Forced-air on, all doors open

The simplest test is called “forced-air system on, all doors open.” Sometimes this is called the “dominant duct leak test.” The test requires a digital manometer and a length of plastic tubing.

The test can only be done once the home — particularly the air barrier — is complete. All exterior doors and windows are closed. All supply and return registers are open. All interior doors are opened, to make the house one interconnected zone. All appliances that take air from the home are turned off or prevented from operating during the test.

The plastic tubing is attached to the manometer reference tap and is run outside. The hose must not be pinched. A second tap on the manometer reads indoor air pressure. The manometer will tell the difference in air pressure between the house and outside air as various systems and appliances in the home are turned on.

Before turning anything on, a baseline air-pressure difference must be established — house air pressure with reference to outside, all systems off. Once the baseline is established, the forced-air system is turned on at the thermostat. The digital manometer measures the effect of forced-air system operation on house air pressure, compared to the baseline.

If a forced-air duct system is perfectly tight, all house air returned to the furnace will be resupplied to the home. Therefore, when ducts are tight, the effect of forced-air system operation should be balanced, or neutral.

National trainers point out that a zero pressure difference or neutral reading could mean that the system operation is balanced because:

1. It is perfectly tight;

2. Supply and return leaks are equal; or

3. The building envelope has so much leakage that outdoor air can move through the building envelope in sufficient quantity to balance any effect that duct leakage might have.

If a negative pressure is recorded (-5 P), air is leaving the home faster than it can be replaced, and house air pressure is lower than outdoor air pressure. Negative pressure indicates supply duct leakage. There may be leaks on the return side too, but a negative reading indicates that supply duct leaks are dominant.

If positive pressure is recorded (5 P), air is coming into the home faster than it can leave, and house air pressure is higher than outdoor air pressure. Leaks on the return side of the system draw air into the system from the surrounding building cavities, so positive pressure in the home indicates the presence of return duct leakage. There may be leakage on the supply side too, but a positive reading indicates return leaks are dominant.

Other tests are needed to clarify what pressure diagnostics are telling us about the forced-air system or the building envelope. However, “forced-air on, all doors open” gives a clear picture of the potential interaction between a furnace and fireplace.

Because natural draft in a fireplace is temperature driven, it is a relatively weak force at best. Negative pressure in a home will make it difficult for a fireplace to draft properly. At the startup or tail-down phase of a fire, draft is particularly weak, and air flow up a flue could be easily reversed if house air pressure is lower than outside air.

Canadian standards for depressurization in a fireplace zone prescribe no more than -5 P in zones containing naturally vented combustion devices. However, spillage has been shown to occur in -3 and -2 P environments. Currently no U.S. standards exist.

Forced-air on, all doors closed

After measuring the effect of forced-air operation with all interior doors open, the next step is to close all interior doors and check the pressure again — fireplace zone with reference to outdoors.

In the site-built home mentioned above, closing bedroom doors was the most likely cause of the spillage incident. When interior doors were closed, air supplied to the rooms could not get back to return grilles located in the central part of the home (the fireplace zone). Consequently, the return system was removing air from the fireplace zone faster than it could be replaced.

Air pressure in the zone dropped, compared to outside air. Air flowed down the chimney and into the home.

If “forced-air system on, all doors open” indicates a neutral or only slightly negative effect (0 to -1 P), it is not expected that the forced-air system operation would cause fireplace spillage. If negative pressure is stronger, unsafe interactions are possible.

If test results show positive air pressures, air may be entering the home from building cavities contaminated with insulation, mold, moisture, or dust, but would not have an immediate threat to health and safety caused by combustion gas.

Sealing supply duct leaks should reduce depressurization caused by forced-air system operation in the fireplace zone, making the home safer and more efficient.

If “forced-air system on, all doors closed” causes negative pressure in the fireplace zone, adding returns to the bedrooms or adding pressure relief grilles between bedrooms and central returns will allow air delivered behind closed doors to get back to the return side of the system.

Even though a more complete battery of tests always gives a better picture than one or two tests can provide, quick measurements of air pressure changes caused by house system operation allow contractors to detect and correct potentially serious safety problems in homes before anyone gets hurt.

Reprinted with permission from Technical Reference Bulletin 154, Summer 1999, from the Air Conditioning Contractors of America.