Life Cycle Comparison of Environmental Emissions from Natural Gas, ULS Heating Oil and Bio/ULS Heating Oil Blends in Vermont

Transcription

1 Life Cycle Comparison of Environmental Emissions from Natural Gas, ULS Heating Oil and Bio/ULS Heating Oil Blends in Vermont Submitted to: Vermont Fuel Dealers Association 250 Main St Montpelier, VT Prepared by: Exergy Partners Corp Meadowville CT Herndon, VA May 31, 2013

2 Analytical Summary While ultimate energy choices as to which fuel to consume are made by builders, remodelers and consumers, and these choices are most often based on economics, these choices are also influenced by perceptions of how efficiently, or inefficiently, our energy resources are being used and how such choices impact the environment, including the release of greenhouse gases into the atmosphere. Focusing on sustainability in the built environment requires life cycle assessments of building products and equipment. Sustainable energy production and consumption should also require life cycle, or fuel cycle, assessments from wellhead to burner tip. It is important, therefore, that consumers, builders and policy makers have the most accurate estimates of energy consumption, energy efficiency and environmental impacts when making energy choices for their homes. However, most efficiency standards and regulations that pertain to residential space heating and hot water appliances are site-based - that is, they only consider the impacts at the site where the energy is ultimately delivered. Given that the energy consumption and environmental impacts along the total energy production and supply chain are not included, reliance on site-based data can thus lead to inaccurate comparisons which may result in higher energy resource consumption, as well as higher levels of pollution. The Vermont Fuel Dealers Association requested an assessment of the claim made by Vermont Gas in its November 14, 2014 presentation titled Addison Natural Gas Project on slide 6 showing that natural gas is a cleaner fuel than oil, propane and wood (dry) that natural gas is cleaner fuel than heating oil. Figure 1 Page 6 of Vermont Gas in its November 14, 2014 Presentation Page 2 of 19

3 Figure 1 is scientifically inaccurate and misleading for the following reasons: 1. Figure 1 purports to present residential heating equipment commonly used in Vermont, but does not identify the heating equipment modeled and the source citations are inadequate. 2. The CO 2 graph is a fuel-based graph which does not account for upstream extraction, processing and delivery, nor does it include appliance efficiency in combustion. Therefore, the CO 2 graph is not related to any heating appliance. Total resource energy and fuel cycle emissions analysis are more comprehensive and accurate methods to assess the total energy and emissions impacts of fuel consumption at the point of use. These methods examine all energy consumption and emissions impacts associated with fuel use, including those from the extraction/production, processing, transmission, distribution, and ultimate energy consumption stages of the fuel cycle. Site energy analysis only takes into consideration the ultimate consumption stage. Significant energy is consumed, with resulting emissions of CO 2 and other greenhouse gases (GHG), during all stages of energy use. Greenhouse gas emissions are usually reported in pounds or metric tons of CO 2 equivalent (CO 2 e). CO 2 e is a measure used to compare different types of greenhouse gas emissions by converting all the various greenhouse gases to a carbon dioxide equivalent. This conversion to a common metric is accomplished using each gas s Global Warming Potential (GWP). GWP values have been established for the various GHGs by the Intergovernmental Panel on Climate Change (IPCC), the premier inter-governmental organization examining climate change and its impact on society. Therefore, it should be clear that CO 2 e emissions should drive policy and not CO 2 emissions. 3. The graph omits two other commonly used fuels ultra-low sulfur heating oil (ULS), which under Vermont law will be the basis for all heating oil sold in the state beginning in 2018; and biodiesel blends which are supplied in state in blends up to 20%. More accurate representation of a comparison of natural gas versus petroleum-based liquid fuels is shown in Figures 2-7. These figures focus on the near-term future where shale gas has become the marginal New England gaseous supply fuel, ULS heating oil required as the liquid base fuel and bio diesel blending will become the normative liquid fuel. Figure 2 and 6 clearly show that ULS bio-blends reach equivalence with natural gas GHG emissions at relatively low blending levels when assess using 20 year atmospheric time horizons 1. Increasing bio/uls blending merely further reduces GHG emissions beyond what could be achieved by neat natural gas. Figure 3 and 7 show that ULS heating oil with modern combustion technology and controls exhibits similar SO 2, NO x and PM, and CO emissions to natural gas 2. 1 See 100 Year versus 20 Year Time Horizon below 2 EPA AP 42-Compilation of Air Pollutant Emission Factors - Tables 1.3.1, and Page 3 of 19

4 Figure 2 Hydronic Non-Condensing Boilers Using New England Delivered Fuels 3 3 Basis for Figure 2 are ICF referenced reports (End Notes b and e modified as indicated in the body of this report to reflect changes for UN IPCC AR3 to AR4, updating biodiesel processing and shale gas information and using a 20 year time horizon for GHG atmospheric measurement. Page 4 of 19

5 Figure 3 Compilation of Air Pollutant Emission Factors (#2 Heating Oil Basis) 4 4 Basis for this graph: EPA AP 42-Compilation of Air Pollutant Emission Factors - Tables 1.3.1, and for small boilers (note EPA does not report on residential boilers, however consultation with Brookhaven National Laboratory confirmed the small boiler numbers are representative). Small particles (PM2.5) and SO 2 (1,500 ppm) values are from BNL report End Note j. Brookhaven National Laboratory testing confirms that B100 NOx emissions are at least 20% temperatures and less fuel bases less that ULS fuel based on lower combustions nitrogen. Page 5 of 19

6 Figure 4 Compilation of Air Pollutant Emission Factors (ULS Basis) Figure 5 Compilation of Air Pollutant Emission Factors Appliance Performance Adjusted Comparing natural gas, ULS and Bio/ULS blend on the basis of CO 2 e, SO 2, NO x and PM, and CO emissions, it is a fair conclusion to state that their emissions are relative comparable when Page 6 of 19

7 examining likely choices for boiler upgrades in Vermont with one caveat. Bio/ULS heating oil blends have the capability to further reduce emissions. (Figures 5 and 7) In summary, the accurate, traceable and credible comparison shown in Figure 6 and 7 comparable results between natural gas emissions and a 1.6% biodiesel and ULS blend. These two figures assume IGPCC AR4 20 year time horizon, shale gas at the margin and high efficiency non-condensing boiler end-use as the most likely existing residential upgrade. Figure 6 Comparing Natural Gas GHG Emissions versus #2 heating oil and 1.6% Bio/ULS Blend Page 7 of 19

8 Figure 7 Comparing Natural Gas Other Emissions with ULS and a 1.6% Bio/ULS Blend Therefore, public policies that encourage transition to modern ULS heating oil usage, moderate bio-blend mix usage and upgrading oil boilers and furnaces is equally as important and is encouraging natural gas upgrades. Liquid fuels like bio-blend ULS heating oil over time can significantly reduce GHG emissions when compared to natural gas by increasing the blending fraction. Page 8 of 19

9 Analysis Basis and Methodology Liquid Fuels Analysis The initial focus of this review will concentrate on Greenhouse Gas emissions for natural gas, number 2 heating oil (#2) and ultra-low sulfur (ULS) heating oil, biodiesel and biodiesel blends with #2 and ULS. When properly combined the three major emissions from natural gas, #2, ULS and biodiesel blends is expressed in CO 2 equivalents 5 (CO 2 e) which are CO 2, NO 2 and methane. In this regard natural, gas, #2, ULS and biodiesel and blend can be fairly compared. Sulfur dioxide and particulate matter will be addressed separately. Fuel Based Carbon Complete Combustion Comparison The CO 2 comparison presented (upper left quadrant) provides one view of four fuels, but it does not reflect residential heating equipment commonly used in Vermont as implied by the slide. The CO 2 data presented is merely the CO 2 emitted if one million Btus of the fuel is completely combusted or how much CO 2 is generated from the carbon content of the fuel. There are two documents i that form the basis for the GHG emissions used in this assessment. The complete combustion for natural gas uses in the first document is lbs/mmbtu for natural gas, lbs/mmbtu for #2 and ULS and 0 lbs/mmbtu for biodiesel. There would be no GHG emissions from the final use of this sustainable fuel. It is generally agreed that all of the carbon produced by combustion of biodiesel has been taken out of the air as a result of photosynthesis during the growing season; therefore, when a biofuel is combusted, an amount of CO 2 is added back to the air that is exactly equal producing a zero net impact. Figure 8 Assessment of Complete Combustion of Select Fuels 5 A CO 2 e converts NO 2 and methane into an equivalent amount of global warming potential in terms of CO 2 and adds the three components together. Page 9 of 19

10 Figure 8 represents the same 6 comparison for fuel based complete combustion adding the ULS, biodiesel (B100) and a blend of #2 or ULS with B100 that would exhibit equal CO 2 emissions per MMBtu to natural gas. # Equation 1 # # # # NG emissions #2 emissions B100 emissions #2 emissions F % where F equals the biodiesel blend fraction Figure 8 provides a comparison of the selected fuels that does not take into account fuel cycle efficiency 7, the actual appliance combustion efficiency, or the atmospheric lifetime analysis which are all required to reflect the complete impact of residential heating equipment commonly used in Vermont. Analytical Basis To achieve a rational understanding of the impact of residential heating equipment commonly used in Vermont today, a series of important variables must be considered as stated above. The foundation for this analysis is the comprehensive ICF study performed for the heating oil industry in The following is a series of analytical updates to this data, together with the means and methods used ICF Fuel Cycle Analysis Fuel Cycle GHG Emissions The fuel cycle GHG emissions comparison shows the amount of CO 2 equivalent emissions that is associated with delivering each MMBtu of the selected fuels to the burner-tip. These comparisons are presented for both 2006 and Changes in emissions intensity that occur over this time frame reflect changes in energy use and emissions for the various fuel cycle stages for each fuel, as well as changes in the supply base (e.g., changes to both domestic supply areas and LNG imports for natural gas) for each demand region. Table 1 ICF Base Report - Fuel Cycle GHG Emissions Findings 6 Complete combustion data for natural gas is from Table ES-2, heating oil and ULS is from ES-1 of Final Report Resource Analysis of Energy Use and Greenhouse Gas Emissions from Residential Boilers for Space Heating and Hot Water, Revised February 2009 by ICF International. 7 The fuel cycle energy efficiency, a measure of the resource energy required to extract, process and deliver the fuel from well to burner tip (before end-use equipment efficiencies) 8 Final Report Resource Analysis of Energy Use and Greenhouse Gas Emissions from Residential Boilers for Space Heating and Hot Water, Revised February 2009, ICF International Page 10 of 19

11 MMBtu MMBtu Natural Gas Natural Gas # ULS B B B % B % Table 1 shows the delivered burner-tip fuel-based CO 2 e emissions based on the information available at the time the report data was compiled. This data suggests that in the year 2020 a 24.2% biodiesel blend with ULS would be equivalent to natural gas (CO 2, CH 4 and N 2 O). However there remains several important updates and analysis that dramatically change this and substantially change the conclusion. Improved Biodiesel Fuel Cycle Efficiency The ICF study utilized a National Renewable Energy Laboratory (NREL) study to develop the biodiesel fuel cycle energy and emissions data. Concurrently, the Biodiesel Board studied the issue and published updated findings shown in Figure 9 Energy Balance: An Update 9 Updating the ICF base report biodiesel data yields Table 2 reducing the bio-blend fraction in 2020 to 22.6% for natural gas equivalence delivered to the burner tip. Table 2 ICF Base Report with Updated Biodiesel Efficiencies and Emissions - Fuel Cycle GHG Emissions Findings MMBtu MMBtu Natural Gas Natural Gas # ULS B B B % B % 9 Energy Balance: An Update, Dev Shrestha Co-Authors, A. Pradhan, D. S. Shrestha, A. McAloon, M. Haas, W. Yee, J. A. Duffield, and H. Shapouri, October 2008 Presentation Page 11 of 19

12 UN IPCC AR3 to AR4 Update The ICF study used the United Nations Intergovernmental Panel on Climate Change s Third Assessment Report (IPCC-TAR, 2001) on the effects of GHGs over a 100-year time horizon. This assessment weights the methane GHG impact at 23 times CO 2 over the 100 year timeframe. Table 3 IPCC Third Assessment Report (2001) TAR 20 year 100 year 500 year Carbon dioxide Methane Nitrous oxide The IPCC Working Group 1 presents GWP values based on the most up-to-date science, but does not recommend any rules on application of those values. Note that the latest science presented in the Fourth Assessment Report (AR4) released in 2007 rates the 100-year impact of methane at 25 times CO 2. Table 4 IPCC Fourth Assessment Report (2007) AR4 20 year 100 year 500 year Carbon dioxide Methane Nitrous oxide Updating the ICF report data to the AR4 findings further reduces the biodiesel blend for equivalence with natural gas GHG emissions to 22.2% in 2020 (Table 5). Table 5 ICF Base Report with Updated IPCC Findings from AR3 to AR4 100 Year Time Horizon MMBtu MMBtu Natural Gas Natural Gas # ULS B B B % B % Shale Gas Adjustments The 2009 ICF Base Report predated an understanding of the impact of Marcellus Shale gas on the region. A second report ICF International provided to the City of New York assessing shale gas forms the basis of this final delivered fuel GHG assessment based on the UN IPCC AR4 100 year time horizon. Page 12 of 19

13 Table 6 Delivered Natural Gas Mixture (NYC) with Shale Gas at the Margin (100 year Time Horizon) Units ICF NYC NG Report Mix with Shale Gas (P48) kg CO 2e /MMBtu ICF NYC NG Report Mix with Shale Gas (P48) lb CO 2e /MMBtu ICF Base Report NG Mix with LNG AR4 Adjust lb CO 2e /MMBtu Table 6 presents the natural gas mixture differences between the 2009 ICF report and the 2012 ICF report with respect to natural gas mixture GHG emissions characteristics as the industry moves from LNG to shale gas. Table 7 provides the 2020 delivered fuel bio/uls mixture is 24.5%. This blend number will be examined further by assessing end use efficiency and then by examining atmospheric time horizons. Table 7 Updated IPCC AR4 100 Year Time Horizon Shale Gas at the Margin MMBtu MMBtu Marginal Shale Gas Natural Gas # ULS B B B % B % End Use Efficiency To understand the impact of residential heating equipment commonly used in Vermont, it is essential to take into account the combustion efficiency of the heating appliance as well as the life cycle emissions of the delivered fuel. To this end, Brookhaven National Laboratory 10 (BNL) developed an accurate method to determine system efficiency for integrated heating and domestic hot water residential systems 11. The BNL model is more accurate in predicting actual building heating and DHW performance than the commonly used AFUE methodology. The comparison was performed on a 2,500 ft 2 ranch home with a basement and typical code construction. Table 8 provides the annual energy use for high efficiency non-condensing natural gas and oil-fired boilers sold in Vermont. Table 8 BNL Energy Findings for Northern New England Home Performance of Integrated Hydronic Systems, Project Report, May 1, 2007, Thomas A. Butcher, Brookhaven National Laboratory. AFUE leads to low estimates of the energy savings potential of modern, integrated systems, particularly where advanced controls are used. Page 13 of 19

14 Boiler & DHW MMBtu/year Description System Boston, MA 3 Current high efficiency oil boiler Current high efficiency gas boiler Table 9 applies the annual fuel use in Table 8 to the delivered fuel GHG emissions in Table 7 and yields a bio/uls blend ratio of 12.6% to achieve equivalent GHG emissions with natural gas based on a 100 year time horizon. Table 9 Updated IPCC AR4 100 Year Time Horizon Shale Gas at the Margin Including End-Use Efficiency NG Shale Gas at Margin lb CO Blend 2 / year MMBtu MMBtu ,601 NG Shale Gas at Margin year ,694 Blend # ,796 ULS ,980 B ,789 B ,106 B , % B , % 100 Year versus 20 Year Time Horizon The United Nations Intergovernmental Panel on Climate Change (IPCC) developed the concept of global warming potential (GWP) as an index to help policymakers evaluate the impacts of greenhouse gases with different atmospheric lifetimes and infrared absorption properties, relative to the chosen baseline of carbon dioxide (CO2). Scientific advancements have led to corrections in GWP values over the past decade, and now our policy decisions sorely need to reflect our new knowledge. In the mid-90s, policymakers for the Kyoto Protocol chose a 100-year time frame for comparing greenhouse gas impacts using GWPs 12. The choice of time horizon determines how policymakers weigh the short- and long-term costs and benefits of different strategies for tackling climate change. According to the IPCC, the decision to evaluate global warming impacts over a specific time frame is strictly a policy decision it is not a matter of science: the selection of a time horizon of a radiative forcing index is largely a user choice (i.e. a policy decision) [and] if the policy emphasis is to help guard against the possible occurrence of potentially abrupt, non-linear climate responses in the relatively near future, then a choice of a 20-year time horizon 12 U.N. Framework Convention on Climate Change, Report of the Conference of the Parties on its Third Session, Held at Kyoto from 1 December to 11 December 2007, Addendum, p. 31. Accessed at on March 12, 2008 Page 14 of 19

15 would yield an index that is relevant to making such decisions regarding appropriate greenhouse gas abatement strategies. 13 Short-lived pollutants that scientists are targeting today which actually warm the atmosphere are methane and hydrofluorocarbons (HFCs) which are greenhouse gases like CO 2, trapping radiation after it is reflected from the ground. Black carbon and tropospheric ozone, an element of smog, are not greenhouse gases, but they warm the air by directly absorbing solar radiation. Black carbon remains in the atmosphere for only two weeks and methane for no more than 15 years. Focusing on near term targets for GHG impacts is both an effective strategy and recommended policy as it can have a more dramatic effect in the short term than reductions in carbon dioxide, thus providing more time to develop appropriate carbon dioxide reduction strategies. This renewed focus on 20-year GHG targets stimulated a reassessment of the ICF life-cycle study using the AR4 20-year numbers for methane emissions in the production, transportation, delivery and combustion of heating oil, ultra-low sulfur diesel, bio-blends, natural gas and LNG. Table 10 shows that, based on a 20 year atmospheric time horizon, both #2 oil and ULS emit less CO 2 e emissions than the natural gas mixture modeled by ICF in the base study with LNG at the margin. Table 10 AR4 20 Year Time Horizon LNG at the Margin Including End-Use Efficiency MMBtu year Blend MMBtu year Blend NG LNG at NG LNG at ,557 Margin Margin ,557 # ,051 ULS ,326 B ,789 B ,106 BXX ,557 Not Required BXX ,557 Not Required Using NETL methane emissions from unconventional gas found in the Exhibit 5-3 of the ICF NYC Report 14 Table 11 shows that an AR4 20 year time horizon shale gas at the margin assessment requires a Bio/ULS blend of only 1.6% to be equivalent to natural gas CO 2 e emissions. Table 11 AR4 20 Year Time Horizon Shale Gas at the Margin Including End-Use Efficiency 13 Intergovernmental Panel on Climate Change, Radiative Forcing of Climate Change and an Evaluation of the IPCC IS92 Emission Scenarios, Cambridge University Press, p Assessment of New York City Natural Gas Market Fundamentals and Life Cycle Fuel Emissions, New York City Mayor s Office of Long-Term Planning and Sustainability, by ICF International 9300 Lee Highway Fairfax, VA 22031, August 2012 Page 15 of 19

16 NG Shale Gas at Margin lb CO Blend 2 / year MMBtu MMBtu ,116 NG Shale Gas at Margin year ,016 # ,051 ULS ,326 B ,789 B ,106 BXX ,537 Not Required Blend B , % CO2e Emissions Conclusions 1. The scientific community and U.S. energy policy leaders are now focusing on short-lived carbon forcers like methane as a means to quickly impact global warming issues. 2. The 20-year IPCC AR4 data shows that, with respect to a delivered fuels basis, a Bio/ULS blend of 14.4% would be sufficient to equal a marginal natural gas mix with shale gas mix CO 2 e emissions. But, this is not the whole story, appliance efficiency is not factored into this calculation. Since the bulk of New England homes using radiant heating boilers cannot cost effectively utilize condensing boilers, as the radiant heating surface areas are generally too small to allow low enough return water temperatures for condensing to occur for much of the heating season, new non-condensing boilers will form the bulk of future residential boiler upgrades. Therefore, a Bio/ULS blend of 1.6% would be sufficient to equal a marginal natural gas mix with shale gas mix CO 2 e emissions. 3. It is essential to understand the end-use application of heating appliances and the ability to apply them to existing homes. For example, condensing boilers, applied to existing homes with radiant-heat generally do not operate in condensing mode as the return temperatures are too high. Therefore, applying these appliances for policy purposes does not reflect scientific reality. 4. Using the 20-year IPCC AR4 data and comparing high efficiency non-condensing appliances shows ULS and bio-blends significantly lower in CO 2e emissions than shale gas or historically delivered natural gas mixtures to the region. 5. Thus, in the short term, less than 20 years, use of natural gas may have a stronger impact on global warming than heating oil applied in biodiesel blends. An evaluation over a hundred years, neat heating oil may have a stronger impact on global warming; however, this longterm effect is easily ameliorated by blending with biodiesel. NOx Emissions Nitrogen oxides (NOx) include both NO and NO 2 and are a pollutant emissions from all combustion sources. NOx is a concern nationally, mainly because it combines with hydrocarbons in the atmosphere and, under the influence of sunlight, forms ozone. With conventional yellow flame systems, the NOx emissions depend upon the firing rate and the combustion chamber. Page 16 of 19

17 Higher firing rates and increased refractory lining in the combustion chamber (hotter chamber) tend to produce higher flame temperatures and higher NOx. Current U.S. systems range from roughly 75 ppm to 180 ppm. Arguably, 110 ppm is about the average for oil combustion with yellow flame burners. Low-NOx residential oil burners based on high rates of recirculation of combustion products within the combustion chamber are available today. These burners have higher air velocity, and more of the air is introduced to the flame zone along the burner centerline, flame tubes to control recirculation, and flame tube slots or holes which control the amount and location of the recirculated flue gas. With these burners, achievable NOx emissions range from 40 to 65 ppm. It is technically feasible to achieve NOx emissions under 10 ppm with a nitrogen free fuel. Routes which have been developed towards achieving this goal include: 1) Increased recirculation rates with current low-nox burner designs with special provisions for startup; 2) new burner head designs; and 3) oil vaporization followed by combustion in radiant, porous media. SO2 Emissions The sulfur in any fuel results in sulfur oxides being released into the atmosphere when it is burned. During combustion in residential heating systems, roughly 98-99% of the sulfur in the fuel is oxidized to form sulfur dioxide (SO 2 ) and emitted from the stack. BNL testing shows that changing to lower sulfur content fuel (500 ppm) eliminates about percent of the sulfur dioxide emissions from residential oil heating systems. Ultra-Low Sulfur heating oil fuels (15ppm) produce immeasurable amounts of sulfur dioxides in the flue stack similar to natural gas. Particulate Emissions According to testing performed by Brookhaven National Laboratory, with respect to small particle emission for combustion of heating oil, they found a direct correlation between sulfur content and particulate emissions. Test data displayed in Figure 10 shows small particulate emissions for ULS to be in the immeasurable range below 0.1 mg/mj, which is essentially the same as natural gas. Page 17 of 19

18 Figure 10: Brookhaven National Laboratory Sulfur Test Results Brookhaven National Laboratory (BNL) tests show that PM 2.5 emissions for Ultra Low Sulfur heating oil are on the same order of magnitude as natural gas (Figure 11). Figure 11: Comparison of Average PM2.5 for Five Heating Fuel Types for Hydronic Boilers and Warm Air Furnaces Page 18 of 19

19 References a. Addison Natural Gas Project, Addison County Regional Planning Commission, Vermont Gas, November 14, 2014 Presentation. b. Final Report Resource Analysis of Energy Use and Greenhouse Gas Emissions from Residential Boilers for Space Heating and Hot Water, Revised February 2009 by ICF International. c. Energy Balance: An Update, Dev Shrestha Co-Authors, A. Pradhan, D. S. Shrestha, A. McAloon, M. Haas, W. Yee, J. A. Duffield, and H. Shapouri, October 2008 Presentation d. Performance of Integrated Hydronic Systems, Project Report, May 1, 2007, Thomas A. Butcher, Brookhaven National Laboratory. e. Assessment of New York City Natural Gas Market Fundamentals and Life Cycle Fuel Emissions, New York City Mayor s Office of Long-Term Planning and Sustainability, by ICF International 9300 Lee Highway Fairfax, VA 22031, August f. Life Cycle Greenhouse Gas Inventory of Natural Gas Extraction, Delivery and Electricity Production, October 24, 2011, DOE/NETL-2011/1522. g. U.N. Framework Convention on Climate Change, Report of the Conference of the Parties on its Third Session, Held at Kyoto from 1 December to 11 December 2007, Addendum, p. 31. Accessed at on March 12, h. Intergovernmental Panel on Climate Change, Radiative Forcing of Climate Change and an Evaluation of the IPCC IS92 Emission Scenarios, Cambridge University Press, p i. EPA AP 42-Compilation of Air Pollutant Emission Factors - Tables 1.3.1, and j. Evaluation of Gas, Oil and Wood Pellet Fueled Residential Heating System Emissions Characteristics, Brookhaven National Laboratory Report, Roger McDonald, December 2009 Page 19 of 19