1. A Collaborative Process to Develop a 2016 Emissions Modeling Platform
11/9/2017

2. Origins of the Collaborative Effort
Versions of the 2011 Emissions Modeling Platform have been used since 2014
2014 National Emissions Inventory version 2 is coming in December, 2017
It’s now time to develop an updated modeling platform
Regional modeling organizations and states asked to be more involved in the development of the next modeling platform
States need a new platform to develop State Implementation Plans for the National Ambient Air Quality Standards for Ozone, and also for Regional Haze
States often have better access to local information
States and regions would like to have more input into the methods used, particularly those used to develop “projections” of emissions to future years
A process to select a post-2011 base year finished in early 2017: 2016 was selected
EPA is supporting a collaborative process to develop a 2016 emissions modeling platform based on 2014NEIv2

3. Improvements in the 2014 NEI Version 2
Updates to PM emissions from ag. tilling, ag. dust, construction dust, mining and quarrying, road dust
Corrections to spatial distribution of livestock NH3 and the addition of VOC for agricultural livestock
Residential fuel combustion updates
Updates to oil and gas emissions
Locomotive emissions and offshore oil platforms updated to year 2014
Refinements to point sources (e.g., stack parameters, locations, hazardous air pollutants)
Updates to onroad input data including age distributions, speeds, fuels, and activity data (e.g., vehicle miles traveled and populations)
Updates to nonroad emissions in some states
Limited updates to fire and biogenic emissions

4. Emissions Modeling Process
Steps needed to convert emissions inventories into the resolution and formats needed by air quality models:
Temporal Allocation
Temporal Profiles
Emission Inventories (annual/ monthly/daily)
Speciation Profiles
Speciation
Spatial Allocation
Hourly, speciated, gridded emissions for air quality modeling
Spatial Surrogates

5. Emissions Modeling Platforms
Contents:
Starting point is a version of the National Emissions Inventory (e.g., 2011 NEI version 2)
Related data sets needed for processing are also included
Software and scripts to process inventories into air quality model inputs
Typically include a historic or “base” year and one or more future years
Historic platform development process:
Base and future year inventories developed by EPA
Ancillary data and software needed to process the inventories for AQM developed by EPA
Platform released for public comment
EPA incorporates comments into a new version of the platform
REPEAT

6. Development Plan for the Collaborative Effort
In October, regional modeling leaders and EPA worked together to prepare an 18 page development plan for the effort
Defines organization, principles and procedures for collaboration, process management, high-level schedule and milestones, expected data products
For each emissions sector specifies timelines for input data availability, known issues, and leaders who have volunteered to- date
Google Drive worked well to facilitate this collaboration WfJwZ2P5VSegHU52iP_6nU/edit?ts=59ce5f35

7. Organizational Structure of the Effort
Overall co-leads: Alison Eyth of EPA and Zac Adelman of the Lake Michigan Air Directors Consortium (LADCO)
Coordinate communication, develop processes to be followed, help resolve issues, specify documentation requirements, facilitate distribution of data to stakeholders
Coordination committee: regional, state, EPA leaders that help finalize the process and resolve issues that come up
Sector-specific Workgroups: co-led by one regional/state and one EPA staff (where possible) with participants from states/locals/regions
Focus on developing 2016, 2023 (for ozone), and 2028 (for haze) emissions estimates and improving modeling of specific emissions sectors

8. Sector Workgroups
The sector-specific workgroups are where the real work will happen
Planned workgroups:
Biogenic sources, fires
Electric generating units, oil and gas sources, other non-EGU point sources including airports
On-roadway mobile sources, off-roadway mobile sources (e.g., tractors, construction, recreational boats), commercial marine vessels, locomotives
Nonpoint sources (e.g., dust, residential pollution sources, gas stations)
Julie McDill (MARAMA) developed a simple survey for MJOs to use to gather state/local input on the effort
Included questions about regulatory drivers and timeframe by which data is needed
Outlined a draft schedule
Provided a section in which people could sign up for workgroups
Responses due by November 10
Let Julie know if you want to participate but haven’t received a survey

9. Next Steps Survey results will be collated by MARAMA
Workgroup leads will receive lists of interested members
Overall co-leads will try to fill remaining workgroup co-lead spots
Coordination committee will meet at the end of November
Review survey responses
Continue specifying processes for documentation, approval and sign off, success metrics, plans for data sharing and add these to the Development Plan
Sector workgroup meetings will be scheduled
Start by reviewing available 2016 data, data from 2014 NEI version 2 to be released in December, data / methods for projections, and emissions modeling input data
Identify issues the workgroup needs to address
Discuss sector-specific schedules and work items
2016 Type A point source submissions are due to EPA by January, 2018

10. Three Phases of Development
Alpha: a first attempt at 2016 modeling inventories using readily available 2016 and 2014NEIv2 data (February, 2018)
Beta: an improved 2016 modeling platform with preliminary projections to 2023 and (Summer, 2018)
Version 1.0: usable modeling platform for 2016, 2023, and 2028 that can be used as a building block for State Implementation Plans and for other modeling studies (Winter, 2019)
End goal: an emissions modeling platform based on best currently available data and methods that states, EPA, and others can use for modeling assessments
Any questions on the overall effort?

11. Origin of data used for oil and gas projections
EIA’s Natural Gas summary (million cubic feet) and EIA’s Crude Oil Production summary (thousands of barrels) provided annual production data for desired years for most states
Natural Gas Production: htm
Oil Production:
Exceptions included being able to obtain data from STATE websites for Indiana and South Dakota
In cases where the EIA tables didn’t provide historic state data, production was assumed to be constant over the 2014 to timeframe unless EPA could identify historic data on state websites

12. Preliminary Projections of 2014 to 2016
A preliminary version of 2016 was developed based on 2014NEIv1
Due to the changes coming in 2014NEIv2, the preliminary version will only be used as a first pass at modeling the year 2016
Projection factors were applied to 2014NEIv1 for non-point oil and gas sector (np_oilgas) and point oil and gas sector (pt_oilgas)
Oil production projection factors applied to “Oil production and exploration” SCCs
Natural Gas production projection factors applied to “Gas production and exploration” SCCs
Any other non-point SCCs not in any of these two categories gets an average of the oil and gas projection factors
NSPS controls applied (if any) from 2014 to 2016 are not included

13. 2014 and 2016 non-point oil and gas emissions based on 2014NEIv1 : NOX

14. 2014 and 2016 non-point oil and gas emissions based on 2014NEIv1: VOC

15. 2014 and 2016 point oil and gas emissions based on 2014NEIv1: NOX

16. 2014 and 2016 point oil and gas emissions based on 2014NEIv1 : VOC

17. Longer-term oil and gas projections: Historical state data bridge to Supply Region-level forecast method
Historical state production data publicly available on EIA website used to project sources from base year to a recent year (e.g or 2016)
Use AEO Supply Region production forecasts to project from 2015 to desired future year (e.g., 2023, 2028)
Some adjustments based on historical trends for “low production” states may be needed
Combine factors to get a state-specific projection factor from base year to desired future year
Use SCC-oil/gas/both cross-reference to apply state-specific factors
US EPA OAQPS, Emission Inventory and Analysis Group

18. Oil and Gas Projections: NEMS (or Supply) Regions
US EPA OAQPS, Emission Inventory and Analysis Group

19. Oil and Gas Projections: Historical state data + AEO Supply Region
AEO Supply Region change
Historical state change X
Combined Factors representing both historical state data and AEO supply region change for each state
US EPA OAQPS, Emission Inventory and Analysis Group

20. 2023en Oil and Gas Projections: NSPS Controls
For future-year NSPS controls (oil and gas, RICE, Natural Gas Turbines, and Process Heaters), we attempted to control only new sources/equipment using the following equation to account for growth and retirement of existing sources and the differences between the new and existing source emission rates.
Control_Efficiency (%) = 100 * (1- [(Pf-1)*Fn + (1-Ri)t + (1-(1-Ri)t)*Fn]/ Pf)
Pf = growth rate expressed as ratio (e.g., 1.5=50% cumulative growth)
t = number of years between base and future years
Fn = emission factor ratio for new sources
Ri = retirement rate, expressed as decimal (e.g., 3.3%=0.033)
US EPA OAQPS, Emission Inventory and Analysis Group