We've been investigating differences in emissions of conventional pollutants from the EPS compared to the EPA's National Emissions Inventory and have discovered that our estimates are very, very off compared to what EPA has. EPA's inventory is much more robust than ours.
As one example, EPA's PM2.5 emissions in buildings are more than 10x what we have in the model, driven primarily by the fact that we are using an industrial boiler biomass emissions factor for buildings, but uncontrolled wood burning in fireplaces is the main culprit. Using a controlled rate for industrial boilers is wrong. This is one example, but across the board we are off, usually significantly. Even where we have the correct mapping, the numbers are off: PM2.5 from heavy duty trucks are 20% of what EPA has in the EPS. Ships are almost 6x higher in the EPS than in EPA's inventory (and we've confirmed EPA is including international shipping).
An alternative approach is to combine the historical data from EPA with the historical data from EIA's AEO to calculate emissions factors on a per sector and fuel basis. This would ensure we at least come close to aligning with EPA and is a significant improvement relative to to what we have in there now. GREET has given us trouble in the past, so I think this is all around an improvement. This is easier than trying to use the emissions factors EPA uses, because it already would convert to BTUs and EPA's emissions factors are scattered throughout many different AP-42 documents, with different units.
We've been investigating differences in emissions of conventional pollutants from the EPS compared to the EPA's National Emissions Inventory and have discovered that our estimates are very, very off compared to what EPA has. EPA's inventory is much more robust than ours.
As one example, EPA's PM2.5 emissions in buildings are more than 10x what we have in the model, driven primarily by the fact that we are using an industrial boiler biomass emissions factor for buildings, but uncontrolled wood burning in fireplaces is the main culprit. Using a controlled rate for industrial boilers is wrong. This is one example, but across the board we are off, usually significantly. Even where we have the correct mapping, the numbers are off: PM2.5 from heavy duty trucks are 20% of what EPA has in the EPS. Ships are almost 6x higher in the EPS than in EPA's inventory (and we've confirmed EPA is including international shipping).
An alternative approach is to combine the historical data from EPA with the historical data from EIA's AEO to calculate emissions factors on a per sector and fuel basis. This would ensure we at least come close to aligning with EPA and is a significant improvement relative to to what we have in there now. GREET has given us trouble in the past, so I think this is all around an improvement. This is easier than trying to use the emissions factors EPA uses, because it already would convert to BTUs and EPA's emissions factors are scattered throughout many different AP-42 documents, with different units.