MARAMA 2007/2017/2020 Modeling Emissions Inventory Version 2 Preliminary Trends Analysis October 31, 2011 Prepared by: Abigail Vickers, Julie McDill and Pat Davis Mid-Atlantic Regional Air Management Association (MARAMA) For: Mid-Atlantic/Northeast Visibility Union (MANE-VU)
Preparation of this report was funded by a subgrant to MARAMA from the Ozone Transport Commission (OTC) under EPA Grant XA-973181. Preparation of the emissions inventory was funded in part by MARAMA s EPA Grant XA-973511 with supplemental funding for additional elements from the above OTC subgrant.
TABLE OF CONTENTS Data Sources and Assumptions 1 Annual VOC Emissions 9 Annual NOx Emissions 21 Annual SO 2 Emissions 31 Annual PM2.5 Emissions 41 Mobile Analysis 51 Projected Emission Analysis 59
DATA SOURCES AND ASSUMPTIONS a. Introduction b. Inventory Data Source and File Locations c. Next Steps d. Screening Modeling Inventories e. Methodology Used to Summarize Emissions In this Report October 31, 2011 Page 1
October 31, 2011 Page 2
INTRODUCTION The MARAMA Version 2 regional inventory will be used to concurrently address national ambient air quality standard (NAAQS) for the new ozone and fine particle and to evaluate progress towards longterm regional haze goals. The emissions inventory is a component of the regional integrated, oneatmosphere air quality modeling platform used to support state air quality attainment demonstrations. The MANE-VU+VA regional inventory addresses emissions in the region that includes Connecticut, Delaware, the District of Columbia, Maine, Maryland, Massachusetts, New Hampshire, New Jersey, New York, Pennsylvania, Rhode Island, Vermont, and Virginia. Local air planning agencies include Philadelphia and Allegheny County, Pennsylvania. For Version 2, which is summarized in this document, not all parts of the inventory are complete. For example, the nonroad sector in Virginia is not included in our inventory. Virginia is participating in the SESARM inventory process and developed their nonroad emissions with SESARM. They have provided MARAMA with 2007 emissions. However, SESARM has not yet produced future year inventories, so no emissions for 2017 or 2020 are available for this document. Table 1 provides a break-down of each sector of the inventory by state. Yes means the sector is included in the inventory and is complete. Table 1 Inventory sector status by state Nonroad Mobile MAR 2007 2017 2020 2007 2017 2020 2007 2017 2020 CT Yes Yes Yes Draft No No Yes Yes Yes DE Yes Yes Yes Draft No No Yes Yes Yes DC Yes Yes Yes Draft No No Yes Yes Yes ME Yes Yes Yes Draft No No Yes Yes Yes MD Yes Yes Yes Draft No No Yes Yes Yes MA Yes Yes Yes Draft No No Yes Yes Yes NH Yes Yes Yes Draft No No Yes Yes Yes NJ Yes Yes Yes Draft No No Yes Yes Yes NY Yes Yes Yes Draft No No Yes Yes Yes PA Yes Yes Yes Draft No No Yes Yes Yes RI Yes Yes Yes Draft No No Yes Yes Yes VT Yes Yes Yes Draft No No Yes Yes Yes VA Yes* No No Draft No No Yes Yes Yes Area NonEGU Point EGU 2007 2017 2020 2007 2017 2020 2007 2017 2020 CT Yes Yes Yes Yes Yes Yes Yes No No DE Yes Yes Yes Yes Yes Yes Yes No No DC Yes Yes Yes Yes Yes Yes Yes No No ME Yes Yes Yes Yes Yes Yes Yes No No MD Yes Yes Yes Yes Yes Yes Yes No No MA Yes Yes Yes Yes Yes Yes Yes No No NH Yes Yes Yes Yes Yes Yes Yes No No NJ Yes Yes Yes Yes Yes Yes Yes No No NY Yes Yes Yes Yes Yes Yes Yes No No PA Yes Yes Yes Yes Yes Yes Yes No No RI Yes Yes Yes Yes Yes Yes Yes No No VT Yes Yes Yes Yes Yes Yes Yes No No VA Yes Yes Yes Yes Yes Yes Yes No No * Provided by SESARM October 31, 2011 Page 3
INVENTORY DATA SOURCE AND FILE LOCATIONS The MARAMA Version 2 NIF and ORL inventory files used to prepare the charts in this document are posted on the MARAMA FTP site. The following FTP log-in information can be used to access the files and documentation: MARAMA 2007 and 2020 Inventory are Version 2 Onroad Mobile is MOVES Version 1 2007 Inventory File Location ftp.marama.org Username: regionalei Password: marama2007 NIF Location: /2007 Emission Inventory Version 2 Final/NIF_Files ORL Location: /2007 Emission Inventory Version 2 Final/SMOKE_Files Documentation: /2007 Emission Inventory Version 2 Final/Documentation Projection Inventory Files Location ftp.marama.org Username: regionalei Password: marama2007 NIF Location: / 2017 2020 Projection Version 2 Final/NIF_Files ORL Location: / 2017 2020 Projection Version 2 Final/SMOKE_Files Documentation: / 2017 2020 Projection Version 2 Final/Documentation File Names on the FTP: Point Sources 2007: EGU and Non-EGU: MANEVU+VA_Point_2007_NOF.zip 2017: MANEVU+VA_Point_2017_NOF.zip 2020: Non-EGU: MANEVU+VA_Point_2020_NOF.zip 2020 EGU Emissions have not been prepared yet Area Sources 2007: MANEVU+VA_Area_2007_NOF.mdb 2017: MANEVU+VA_Area_2017_NOF.zip 2020: MANEVU+VA_Area_2020_NOF.zip Nonroad Mobile- NMIM 2007: 2007MARAMANRNMIMv1_1.zip 2017: 2017MARAMANMIMv1_1.zip 2020: 2020MARAMANMIMv1_1.zip Nonroad Mobile- Marine/Aircraft/Rail 2007: Marine: MANEVU+VA_MAR_2007_CMV_v1_4.mdb Aircraft: MANEVU+VA_MAR_2007_Aircraft_ v1_2.mdb Rail: MANEVU+VA_MAR_2007_Rail_v1_3.mdb 2017: MANE-VU+VA_MAR_2017.mdb 2020: MANE-VU+VA_MAR_2020.mdb Onroad Mobile 2007: SMOKE-MOVES Version 1 2017: MOVES Runs have not been completed 2020: MOVES Runs have not been completed. October 31, 2011 Page 4
NEXT STEPS FOR THE MARAMA MODELING EMISSIONS INVENTORY This summary of the MARAMA Version 2 modeling inventory is preliminary in that emissions from the Electric Generating Unit (EGU) and on-road sectors are not complete for future years 2017 and 2020. In addition, the 2007 on-road mobile emissions were prepared using a preliminary version of the SMOKE- MOVES model and is considered a draft. Finally, emissions from other regions are not currently available. This section provides an overview of future plans to update and extend this preliminary inventory for use in SIP modeling. ELECTRIC GENERATING UNITS (EGU) Electric Generating Units (EGU) emissions for future years 2017 and 2020 have not yet been estimated. MANE VU states are participating in an inter-regional process under the Eastern Regional Technical Advisory Committee (ERTAC) that is working to produce a new model capable of estimating future EGU emissions. The model will use base year dispatching patterns, EIA fuel, region specific growth factors, and state supplied unit modification information to estimate future emissions. We expect the ERTAC EGU emission projection model to be complete in Mid-2012, at which time it will be used to generate projected emissions for 2017 and 2020. ON-ROAD MANE VU states are working with OAQPS and OTAQ staff to implement the SMOKE-MOVES model. This model uses emission factors calculated by the MOVES Look Up Table Model to calculate county level hourly emissions for use in air quality modeling exercises. The implementation process has been protracted. As a result, the 2007 on-road mobile emissions reported in this document were prepared using a preliminary version of the SMOKE-MOVES model. In addition, on-road mobile emissions for future years 2017 and 2020 have not yet been estimated. To distinguish this preliminary 2007 on-road inventory from the final version, it has been identified as MOVES 2007 Version 1. UPDATES OF BASELINE INVENTORIES MARAMA Version 2, as is summarized in this document, was used in the OTC 2011 level 2 screening modeling exercise. However, a new MARAMA Version 3 of the inventory is under development. There are only expected to be minor differences between Versions 2 and 3. DEVELOPMENT OF A CONTROLLED INVENTORY A controlled inventory for 2017 and 2020 will be built based on Version 3. The controlled inventory will account for control measures to reduce future emissions. October 31, 2011 Page 5
METHODOLGY USED TO SUMMARIZE EMISSIONS IN THIS REPORT The annual emissions inventory consists of individual records of emissions by year, pollutant, Standard Classification Code (SCC), county and state. This section describes our use of the SCC codes to graphically summarize the inventory in this report. The U.S. Environmental Protection Agency (USEPA) developed the Source Classification Code (SCC) system. Standard Classification Codes are a way of organizing emissions by source type. The SCC is a ten-digit code. The codes are organized in a hierarchical fashion, with the first two digits indicating a very broad classification of sources, and each subsequent pair of digits indicating a more specific source classification. The last three digits are the most specific. For example: Tier 1 is the first two digits in the SCC. It is the highest level of the hierarchy and is an overall description of a broad source category. Tier 2 is the second and third digits found in the SCC and are more specific than Tier 1. As the viewer moves from left to right, the groups of digits, or Tiers, allow for more specific information. For the example above, Tier 1 (21) indicates that this information pertains to all stationary source fuel combustion categories. Tier 2 (02) links the data to the industrial category, which is more specific than Tier 1 but less specific than Tier 3. Tier 3 (006) shows that the data is relevant to natural gas usage only. Finally Tier 4 (000) is the most specific Tier, dividing the information into formal groups. This particular number indicates all boiler types. Though occasionally the Tier 4 level contains 4 numbers instead of 3, the above example is the general trend. For the analysis in this document, emissions for all pollutants were totaled at either the SCC tier 1 or SCC tier 2 levels. For presentation in the tables the SCC written definitions were used rather than the SCC for easy comprehension. October 31, 2011 Page 6
SCREENING MODELING INVENTORIES The MARAMA inventories have been used by the OTC in screening modeling runs. The screening modeling platform consists of operational portions of the SIP quality platform as they become available and uses proxy data to complete the platform. EMISSIONS INVENTORIES USED IN SCREENING MODELING WITHIN THE OZONE TRANSPORT REGION (OTR) Table 2 compares the inventories used in Level 1 and Level 2 Screening Modeling runs. As can be seen in the table, the completed sectors of the MARAMA Version 1 inventory were used for 2007 emissions in Level 1 screening, while for Level 2 screening modeling, both 2007 and 2020 used the MARAMA version 2 inventory. When the inventory is not complete estimations for 2020 were made based on percentage reductions from the 2007 base year emissions, as shown in Table 2. The percentage reduced was estimated based on anticipated implementation of federal control measures. An additional 5% reduction was taken for NOX proxy future emissions. This is to represent additional control measures taken within the OTR beyond federal measures. For example, EGU NOx and emissions in the OTR are estimated to drop a total of 70% and 30% respectively by 2020. Table 2 Comparison of Level 1 and Level 2 Screening Modeling Mobile EGU Other Point Nonroad Area Level 1 - Complete in 2010 Level 2 - Complete in 2011 2007 Scenario 3 (2020) 2007 Scenario 4 (2020) OTR Non-OTR OTR Non-OTR OTR Non-OTR OTR Non-OTR NOX & VOC NOX & VOC NOX VOC NOX VOC NOX & VOC NOX & VOC NOX VOC NOX VOC Mobile 6 Proxy EPA/Chief 2005 & Proxy from 2007 Proxy from 2007 MOVES NEI2008 Proxy from 2007 Proxy from (2002-2009 Proxy (2002-2009 2007 Ver. 1 NEI2008 interpolation) interpolation) 75+5=80% 30% 75% 30% 70+5=75% 30% 70% 30% MARAMA 2007 Ver. 1 MARAMA 2007 Ver. 1 MARAMA 2007 Ver. 1 Proxy (2002-2009 interpolation) EPA/Chief 2005 & Proxy (2002-2009 interpolation) EPA/Chief 2005 & Proxy (2002-2009 interpolation) EPA/Chief 2005 & Proxy (2002-2009 interpolation) EPA/Chief 2005 & Proxy (2002-2009 interpolation) *An additional 5% reduction of NOx emissions was applied to the MARAMA version 2 inventory to account for new control measures that will be adopted in the region. OTHER REGIONAL INVENTORIES Proxy from 2007 Proxy from 2007 MARAMA 2007 Ver. 2 NEI2008 Proxy from 2007 0+5=5% 30% 0% 30% Add. 5%* 0% 30% The modeling domain includes states in the midwestern and southeastern United States. Inventories for the Midwest are prepared by Lake Michigan Air Directors Consortium (LADCO). Inventories for the Southeast are prepared by Southeastern States Air Resource Managers (SESARM). Neither entity had a complete 2007 inventory at the time of modeling. In Level 1 screening modeling the USEPA 2005 modeling platform emissions were used to represent emissions outside the OTR. Estimations for 2020 were made based on a percentage reduction from the 2007 base year emissions. The percentage reduced was estimated based on anticipated implementation of federal control measures. Proxy from NEI2008 65+5=70% 30% 65% 30% 65+5=70% 30% 65% 30% Proxy from 2007 Proxy from 2007 MARAMA NEI2008 MARAMA 2020 Proxy from 2007 Ver. 2 Ver. 2 NEI2008 65+5=70% 30% 65% 30% Add. 5%* 65% 30% Proxy from 2007 Proxy from 2007 MARAMA NEI2008 MARAMA 2020 Proxy from 2007 Ver. 2 Ver. 2 NEI2008 35+5=40% 30% 35% 30% Add. 5%* 35% 30% Proxy from 2007 Proxy from 2007 MARAMA NEI2008 MARAMA 2020 Proxy from 2007 Ver. 2 Ver. 2 NEI2008 October 31, 2011 Page 7
October 31, 2011 Page 8
ANNUAL VOC EMISSIONS CONTENTS a. Regional Emissions Pie Chart b. Regional Top VOC Source Categories c. VOC Biogenic VOC Emissions by State d. Top VOC Source Categories by State SUMMARY The charts in this section summarize VOC emissions in the MANE-VU and VA region. The regional 2007 pie chart shows that 2007 VOC emissions are predominantly biogenic. Regional anthropogenic VOC emissions in 2007 are split more or less equally between mobile, area and nonroad sources. The bar chart of biogenic emissions per unit area show that biogenic VOC emissions tend to be greater as you move south due to higher temperatures. Bar charts present the Source Classification Codes (SCC) with the maximum anthropogenic emissions in 2007 first for the entire region and then for each state. Projected emissions in 2020 are provided for comparison, where available. These charts are summarized at the SCC Tier 2 level. These charts show that onroad mobile, various nonroad categories; solvent usage and residential wood combustion are the SCC Tier 2 sources with the largest VOC emissions in 2007. Because onroad mobile have not yet been calculated for 2020, no conclusion can be drawn as to what SCCs will predominate in 2020. October 31, 2011 Page 9
October 31, 2011 Page 10
Total: 7,953,639 TPY October 31, 2011 Page 11
Note: States are listed north to south October 31, 2011 Page 12
* Predominantly Wood * Predominantly Wood October 31, 2011 Page 13
* Predominantly Wood October 31, 2011 Page 14
* Predominantly Wood * Predominantly Wood October 31, 2011 Page 15
* Predominantly Wood * Predominantly Wood October 31, 2011 Page 16
* Predominantly Wood * Predominantly Wood October 31, 2011 Page 17
* Predominantly Wood * Predominantly Wood October 31, 2011 Page 18
* Predominantly Wood October 31, 2011 Page 19
October 31, 2011 Page 20
Annual NOx Emissions CONTENTS a. Regional Emissions Pie Chart b. Regional Top NOx Source Categories c. Top NOx Source Categories by State SUMMARY The charts in this section summarize NOx emissions in the MANE-VU and VA region. The regional pie chart shows that 2007 NOx emissions are predominantly from mobile sources. The bar charts present the Source Classification Codes (SCC) with the maximum emissions in 2007 first for the entire region, and then for each state. Projected emissions in 2020 are provided for comparison, where available. These charts are summarized at the SCC Tier 2 level. These charts show that onroad mobile, EGUs, and nonroad SCCs are the Tier 2 sources with the largest NOx emissions in 2007. Because onroad mobile and EGU sources have not yet been calculated for 2020, no conclusion can be drawn as to what SCCs will predominate in 2020. October 31, 2011 Page 21
October 31, 2011 Page 22
Total: 2,862,968 TPY October 31, 2011 Page 23
October 31, 2011 Page 24
October 31, 2011 Page 25
October 31, 2011 Page 26
October 31, 2011 Page 27
October 31, 2011 Page 28
October 31, 2011 Page 29
October 31, 2011 Page 30
ANNUAL SO 2 EMISSIONS CONTENTS a. Regional Emissions Pie Chart b. Regional Top SO 2 Source Categories c. Top SO 2 Source Categories by State SUMMARY The charts in this section summarize SO 2 emissions in the MANE-VU and VA region. The regional pie chart shows that 2007 SO 2 emissions are predominantly from point sources. The bar charts present the Source Classification Codes (SCC) with the maximum emissions in 2007 first for the entire region and then for each state. Projected emissions in 2020 are provided for comparison, where available. These charts are summarized at the SCC Tier 2 level. These charts show that regionally, electric generation is the category with the largest 2007 emissions, representing about 75% of all emissions. For many individual states residential distillate oil combustion is a significant contributor to SO 2 emissions. Further, for most OTC states, SO 2 emissions from residential fuel combustion exceed all other categories except EGUs. While 2020 electric generation emissions have not yet been quantified, they are expected to drop dramatically. However, emissions from residential distillate oil combustion are expected to drop by only 15 to 20% by 2020. These estimates reflect only emissions controls already adopted. October 31, 2011 Page 31
October 31, 2011 Page 32
Total: 2,298,676 TPY * Predominantly Distillate Oil October 31, 2011 Page 33
* Predominantly Distillate Oil * Predominantly Distillate Oil October 31, 2011 Page 34
* Predominantly Distillate Oil * Predominantly Distillate Oil October 31, 2011 Page 35
* Predominantly Distillate Oil * Predominantly Distillate Oil October 31, 2011 Page 36
* Predominantly Distillate Oil * Predominantly Distillate Oil October 31, 2011 Page 37
* Multiple Fuels * Predominantly Distillate Oil October 31, 2011 Page 38
* Predominantly Distillate Oil * Predominantly Distillate Oil October 31, 2011 Page 39
* Predominantly Distillate Oil October 31, 2011 Page 40
ANNUAL PM2.5 EMISSIONS CONTENTS a. Regional Emissions Pie Chart b. Regional Top PM2.5 Source Categories c. Top PM2.5 Source Categories by State SUMMARY The charts in this section summarize PM2.5 emissions in the MANE-VU and VA region. The regional emissions pie chart shows that 2007 PM2.5 emissions are predominantly from area sources. The bar charts present the Source Classification Codes (SCC) with the maximum emissions in 2007, first for the entire region, and then for each state. Projected emissions in 2020 are provided for comparison, where available. These charts are summarized at the SCC Tier 2 level. These charts show that residential heating, predominantly by wood, is a significant source in the region. These emissions are expected to drop approximately 10-15% by 2020. October 31, 2011 Page 41
October 31, 2011 Page 42
Total: 465,280 TPY * Predominantly Wood October 31, 2011 Page 43
* Predominantly Wood * Predominantly Wood October 31, 2011 Page 44
* Predominantly Wood October 31, 2011 Page 45
* Predominantly Wood * Predominantly Distillate Oil October 31, 2011 Page 46
* Predominantly Wood * Predominantly Wood October 31, 2011 Page 47
* Multiple Fuels * Predominantly Wood October 31, 2011 Page 48
* Predominantly Wood * Predominantly Wood October 31, 2011 Page 49
* Predominantly Wood October 31, 2011 Page 50
MOBILE ANALYSIS CONTENTS a. Spatial Map of 2007 Onroad Emissions b. Regional MOVES Analysis by state c. Monthly mobile emissions by state SUMMARY Mobile emissions are an important component of the inventory, representing 51% of the NOx emissions in the region. This section further analyzes the mobile inventory. The spatial maps show that mobile emissions are concentrated on the eastern seaboard along the I-95 corridor. Other pockets of high emissions occur in the urban areas. Following the maps, a set of four bar graphs compare on-network versus off-network emissions for each pollutant. On-network emissions are emitted while the vehicle is underway and are directly related to the number of miles driven. Off-network emissions are either evaporative and occur continuously regardless of whether the vehicle is driven, or occur at start-up before the emission control equipment is warmed up. Different control strategies may work for on-network versus off-network emissions. NOx, SO2, and PM2.5 are predominately on-network and are directly related to number of miles traveled. VOC emissions are predominantly off-network and result from evaporation, which occurs whether or not the vehicle is driven. Most PM2.5 emissions occur during engine start-up before the control equipment is heated up and effective. The final set of graphs show monthly traces of 2007 emissions for each state. In general, PM2.5 emissions are highest in the cold, winter months. NOx emissions are relatively constant all year round. VOC emissions are somewhat lower during the summer months as a result of the use of summer fuel. There are VOC spikes in spring and fall when the fuel transition occurs, particularly in warmer, southern states like Virginia. This is because of high emissions on warm days in the spring and fall where winter fuel is still in use. October 31, 2011 Page 51
October 31, 2011 Page 52
October 31, 2011 Page 53
October 31, 2011 Page 54
October 31, 2011 Page 55
October 31, 2011 Page 56
October 31, 2011 Page 57
October 31, 2011 Page 58
FUTURE YEAR PROJECTION ANALYSIS CONTENTS a. Selection of Future Modeling Years b. Nonroad c. Area d. MAR e. NonEGU Point October 31, 2011 Page 59
October 31, 2011 Page 60
SELECTION OF FUTURE MODELING YEARS Future emissions inventories are used to predict what additional controls will be needed to achieve air quality standards and visibility goals. In general, years are selected based on the standards for which attainment must be demonstrated. Based on advice from MANE VU member states MARAMA has completed significant portions of a regional emissions inventory for the years 2017 and 2020 projected from a 2007 base year. The projected change in emissions from 2007 to 2017 and 2020 for the sectors that are complete are summarized in this section. Anthropogenic VOC emissions are projected to drop sharply for Nonroad sources, while for area and nonegu point VOC emissions are largely unchanged. VOC emissions for Marine/Air/Rail sources drop sharply between 2007 and 2017 and are largely unchanged between 2017 and 2020. NOx emissions are projected to drop sharply between 2007 and 2020 for Nonroad, area, and Marine/Air/Rail source groups and generally hold steady for nonegu point sources. SO 2 emissions are projected to drop sharply between 2007 and 2020 for Nonroad (except for NY). Emissions from Marine/Air/Rail source groups also drop sharply. SO 2 emissions generally hold steady for area and nonegu point sources, but drop in some states. PM2.5 emissions are projected to drop sharply between 2007 and 2020 for Nonroad, Marine/Air/Rail source groups. Trend vary by state for area and nonegu point sources. October 31, 2011 Page 61
Note: VA 2007 nonroad files provided by SESARM. 2017 and 2020 have not yet been prepared. October 31, 2011 Page 62
October 31, 2011 Page 63
Note: VA 2007 nonroad files provided by SESARM. 2017 and 2020 have not yet been prepared. October 31, 2011 Page 64
October 31, 2011 Page 65
Note: VA 2007 nonroad files provided by SESARM. 2017 and 2020 have not yet been prepared. October 31, 2011 Page 66
October 31, 2011 Page 67
Note: VA 2007 nonroad files provided by SESARM. 2017 and 2020 have not yet been prepared. October 31, 2011 Page 68
October 31, 2011 Page 69