All combustion generates by-products, and natural gas combustion, while cleaner than other fuels, is no exception. A number of natural gas-burning appliances are designed to be used indoors, without venting of the combustion products outdoors. Thus it is important to understand the nature and quantities of trace chemicals that are emitted by these appliances. Most of the attention in recent years has been focused on carbon monoxide and nitrogen dioxide. However, many other chemicals are formed as products of incomplete combustion, and are emitted into the indoor environment at trace levels.

Battelle’s Atmospheric Science and Applied Technology department recently completed an extensive study to explore the nature of trace organic and inorganic chemicals emitted by these appliances, including individual hydrocarbons, formaldehyde, other carbonyl compounds, nitrous acid, nitric acid, sulfur dioxide, nitrous oxide, nitric oxide, hydrogen cyanide, carbon dioxide, formic acid, acetic acid, polycyclic aromatic hydrocarbons, and ultrafine particles. Emissions were quantified both under normal and improper operating conditions. The data have been used to calculate emission factors for each trace contaminant. These emission factors can be used for indoor air quality modeling or empirical calculations to estimate the effect of the emissions on indoor air quality. These estimates can be used in turn for risk assessments. The emission factors from different appliances and different operating conditions can also be used by consumers to select appliances and conditions which minimize the air quality impact of natural gas appliance use.

The experimental portion of the project was designed to measure emissions of a large number of organic and inorganic chemicals from four different types of unvented natural gas appliances. These included a range burner, a convective space heater, a radiant space heater and a set of gas fireplace logs. All of the appliances are promoted for unvented use indoors. The approach used to measure the trace chemicals in the appliance emissions involved operating each appliance in a 17.3 m³ environmental chamber filled with clean humidified air. The emissions were captured and diluted in the chamber, and measured by a suite of instruments.

Pollutant Emissions Factors Study Highlights Nitrous acid was identified and quantified for the first time in natural gas appliance emissions. This species was emitted by all appliances studied, under all conditions. Formaldehyde was observed in the emissions under all conditions, and its emission factor was quantified. Hydrogen cyanide was measured at trace levels in several of the experiments. Nitrous oxide (N2O), a greenhouse gas, was observed in each of the tests for which its measurement was attempted. Formic and acetic acids were not observed in selected experiments designed to identify these species. Emissions of ultra fine particles and particle-bound polycyclic aromatic hydrocarbons were measured in selected experiments. The presence of hydrocarbons in the combustion air, for example from a gas leak in the appliance plumbing, has a noteworthy effect on the distribution of emitted oxidized nitrogen compounds. Hydrocarbon contamination of the secondary combustion air tends to increase the NO2/NO ratio in the emissions.

The measured contaminant concentrations and other necessary parameters were used with a mass balance equation to calculate emission factors in units of micrograms of contaminant per kilojoule of energy consumed (mg/kJ). Some experiments were performed with improper appliance operating conditions, to explore the effect on emissions. These experiments involved operating the appliance with obstructed air flow, incorrect physical configuration, or with vitiated air as the combustion air.

The primary results from this study are emission factors for a large number of pollutants. Emission factors for some of the contaminants apparently have not been measured before. The data can be used to examine the effect on emissions of factors such as firing rate, cooking load (for the range burner), convective vs. radiant design (for the space heaters), burner cap design, natural gas leakage, improper configuration (stacked logs), obstructed air flow, and vitiation of the combustion air.

This study greatly expands the data available on unvented natural gas appliance emissions. The results can be used to help identify conditions which lead to unacceptable emissions, identify practices that could be implemented to reduce occupant exposure to emissions, and can also be used to improve residential indoor air quality. The final report is available from the Gas Technology Institute as GRI-01/0097.

For further information, contact Chet Spicer at (614)424-5319, spicerc@battelle.org.