The combustion-testing paradox in our industry
The recent literature review by Lawrence Berkeley Laboratory, titled: Assessment of Literature Related to Combustion Appliance Venting Systems, asserts that worst-case depressurization testing isn’t likely a significant risk reduction strategy for CO poisoning. The Building Performance Institute (BPI) and the Residential Energy Services Network (RESNET) require these worst-case tests but don’t require a complete combustion analysis measuring CO, O2, and temperature. The heating industry’s traditional leaders, Air Conditioning Contractors of America (ACCA) and National Fire Protection Association (NFPA) don’t require any kind of combustion testing in their standards. On the BPI-RESNET side you have an unnecessary and incomplete testing methodology, and on the ACCA-NFPA side you have a general neglect of combustion analysis.
Electronic combustion analysis is easy and quick. Combustion analysis measures carbon monoxide (CO) along with oxygen and flue-gas temperature. The test is important for both safety and energy efficiency. BPI and RESNET should promote electronic combustion analysis rather than worst-case testing for combustion appliances. The combustion-analysis equipment is widely available and not expensive, considering the valuable information it gives the energy auditor or heating technician.
I don’t think you can support mandatory worst-case testing based on risk reduction. The worst-case test is too complicated, inconsistently performed, and often not followed by effective combustion-safety improvement. Worst-case testing has dominated training and certification to the neglect of standards and best practices aimed at saving energy.
Here is a simpler procedure for combustion testing and hazard mitigation.
Make the following observations, and take appropriate action.
- Recognize soot anywhere near a combustion appliance as a sign that the appliance is producing CO at least some of the time.
- Recognize that rust in a chimney or vent connector may mean poor draft and spillage.
- Look for irregularities and flaws in the venting system.
- Find and seal all accessible return-duct leaks in forced-air heating systems.
- Verify that the home has a working CO alarm. If the home has no working smoke alarm in addition to no CO alarm, install a combination CO-Smoke alarm.
Then test, evaluate, and improve safety and efficiency
Test all combustion appliances with a modern flue-gas analyzer, which measures oxygen, temperature, and CO during weatherization, appliance replacement, or a heating service call.
- If measured CO is above 100 parts per million (as measured), take action to bring the reading to as low a value as possible.
- Notice the % of oxygen. For appliances with naturally drafting chimneys the value should be 5% or more to ensure adequate combustion air. If the oxygen is less than 5%, open a window and test again to see if the oxygen percent increases. If so, provide combustion air from outdoors. Avoid making a combustion air opening on a wall parallel to the winter prevailing wind. Don’t provide outdoor combustion air for appliances that operate above 5% oxygen.
- Notice the flue-gas temperature: For traditional open-combustion appliances the flue-gas temperature for gas should be above 325 degrees and for oil above 350 degrees to produce adequate natural draft.
- Operate the combustion appliances, including furnaces, at natural conditions and test for spillage. Furnaces, boilers and combustion water heaters shouldn’t spill for longer than one minute. Inspect the chimney and vent connector if the appliance spills for longer than a minute.
- Make appropriate repairs such as removing chimney obstructions, sealing leaks in the chimney and vent connector, sealing return ducts, and replacing horizontal runs of single-wall vent connector with double wall vent connector if the spillage isn’t related to depressurization.
- If you detect spillage or backdrafting and suspect that depressurization is causing it, then investigate depressurization further.
Other Blog Posts on Combustion Testing