Technical Support Document: Calculating Benefit Per-Ton estimates

Overview

The full attainment PM2.5 benefits in this RIA are a composite of
benefits from the modeled partial attainment scenario and extrapolated
benefits based on benefit per-ton metrics. This technical support
document (TSD) documents the steps we took to calculate the benefit
per-ton metrics that were used as the basis for the extrapolated
benefits. This document focuses specifically on the steps followed to
calculate the metrics and makes extensive reference to previous
documents that describe the air quality modeling data that provided the
inputs to these benefit estimates. 

Summary of Calculation Steps

The basic calculation steps were as follows:

Model the changes in ambient PM2.5 associated with 25%, 50% and 75%
reductions in PM2.5 precursors from various classes of industrial
sources 

Input the changes in air quality to the Environmental Benefits Mapping
and Analysis Program (BenMAP) and estimate monetized benefit estimates
of these air quality improvements

Divide the total monetized benefit by the total emission reduction

Below we provide an expanded discussion of each of these three steps.

Generating Modeled PM2.5 Concentrations

EPA employed a Response Surface Model (RSM) to generate estimated PM2.5
concentrations. The RSM uses advanced statistical techniques to
characterize the relationship between a very complex air quality model
such as CMAQ and input parameters in a highly economical manner.  The
RSM is simply a reduced-form prediction model that uses statistical
correlation structures to approximate model functions through the design
of complex multi-dimensional experiments. Additional technical detail
regarding the model may be found in the technical support document for
the PM2.5 RIA. 

In this version of the PM2.5 RSM, EPA selected 12 emission control
factors based on precursor emission type and source category.  The
selection of factors was based on three fundamental areas:

type of precursor emission (NOx, SOx, NH3, POC, PEC, or VOC), 

emission source category (EGU point sources, NonEGU point sources, area
sources, agricultural sources, nonroad sources, and mobile sources)

location of urban areas contributing to residual PM2.5 after
implementation of the Clean Air Interstate Rule in 2015 and
geographically distinct in contribution to sources of PM2.5
concentrations.

This version of the PM2.5 RSM uses the 2015 CAIR inventory and included
the following 12 source pollutant factors:

NOx EGU = NOx IPM EGU point source emissions

NOx NonEGU Point and Area = NOx IPM Non-EGU point source, area source,
and agricultural source emissions 

NOx Mobile = NOx nonroad source and mobile source emissions

SOx EGU = SOx IPM EGU point source emissions

SOx NonEGU Point = SOx IPM Non-EGU point source emissions

SOx Area = SOx area source and agricultural source emissions 

NH3 Area = Ammonia area source and agricultural source emissions  

NH3 Mobile = Ammonia non-road source and mobile source sources

POC/PEC Point (EGU and NonEGU) = Elemental carbon and organic carbon IPM
EGU point source and IPM Non-EGU point source emissions

POC/PEC Mobile = Elemental carbon and organic carbon nonroad source and
mobile source emissions

POC/PEC Area = Elemental carbon and organic carbon area source and
agricultural source emissions

VOC All = Volatile organic carbon IPM EGU point source, IPM Non-EGU
point source, area source, agricultural source, nonroad source, and
mobile source emissions

One of the chief analytical benefits of the RSM is that it can model the
change in air quality resulting from a change in emissions from any
given source/pollutant factor while holding all other factors constant.
For, example, the RSM can estimate the improvement in PM2.5 air quality
resulting from a 50% reduction in area source NH3 while holding other
emissions constant. This ability to hold other emissions constant is
essential to the process of creating benefit per ton estimates that are
specific to each source/pollutant combination. 

To generate estimates of changes in air quality associated with emission
reductions in each source/pollutant combination, we simulated a 25%, 50%
and 75% reduction in each of the 12 factors, each time holding all other
factors constant; this step generated 36 different air quality modeling
runs. For each modeling run, the RSM produces 36 km CMAQ air quality
grids suitable for input to the BenMAP model for subsequent generation
of benefits estimates.

Estimating Monetary Value of Human Health Benefits

We used the BenMAP model to estimate changes in human health resulting
from each of these modeling scenarios. The benefits assessment used
techniques, health impact functions and valuation functions that were
consistent with the Regulatory Impact Analysis that supported the PM2.5
National Ambient Air Quality Standards. Readers interested in a
comprehensive discussion of these methods and functions may refer to
chapter 5 of this document. 

The BenMAP model produced estimates of the change in the incidence of
health effects and estimated a monetary value for the modeled air
quality change in each of the 12 source/pollutant combinations.
Consistent with the PM2.5 NAAQS RIA, we estimated total benefits for
each modeling scenario using two different PM2.5 mortality functions:
Pope et. al (2002) and Laden et. al (2006). For analytical simplicity we
estimated benefits using a 3% discount rate only. 

Calculating the benefit per ton estimates

Having modeled the PM2.5 air quality change and estimated the total
valuation for each scenario, the final step was to estimate the benefit
per ton of PM2.5 precursor abated. This calculation step involved simply
dividing each valuation estimate by its corresponding reduction in PM2.5
precursor tons abated. For example, in a given scenario we reduced
emissions by 1,000 tons and benefits were $2,000,000 then the benefit
per ton would be $2,000. 

Results

Table 1 below summarizes the benefit per ton estimates for each of the
36 scenarios.

PM2.5 Mortality Estimate	Percent Emission Reduction	Source/Pollutant
Combination



CarbonArea	CarbonMobile	CarbonEGU/

Non-EGU	NH3Area	NH3Mobile	NOxEGU	NOxMobile	NOxNEGU	SOxArea	SOxEGU
SOxNEGU	VOC

Pope et al. (2002)	25%	$240,548	$174,152	$147,210	$12,276	$27,895	$2,570
$4,176	$2,491	$12,651	$24,166	$15,203	$325

	50%	$234,903	$172,105	$147,298	$12,364	$30,713	$3,106	$4,556	$2,833
$13,163	$22,344	$15,020	$408

	75%	$229,860	$169,523	$146,567	$12,421	$33,529	$3,644	$4,919	$3,173
$13,581	$20,226	$14,858	$491















	Laden et al. (2006)	25%	$514,082	$372,797	$316,484	$26,293	$59,780
$5,494	$8,961	$5,346	$27,088	$51,740	$32,532	$695

	50%	$500,528	$368,015	$316,292	$26,431	$65,775	$6,643	$9,766	$6,078
$28,183	$47,609	$32,084	$874

	75%	$488,309	$362,045	$314,310	$26,500	$71,757	$7,796	$10,531	$6,804
$29,065	$42,856	$31,681	$1,051

 	 	 	 	 	 	 	 	 	 	 	 	 	 

Discussion

There is some variability among the benefit per ton estimates. The first
source of variability is attributable to the PM2.5 mortality estimate
used to calculate total benefits. Consistent with the results of the
PM2.5 RIA, benefit estimates derived using the Laden et al. (2006)
mortality estimate are roughly double those estimates based on Pope et
al. (2002). 

The second source of variability among the estimates is due to the
different potential for a given ton of each of the 12 source/sector
combinations to form PM2.5 in the atmosphere. For example, carbonaceous
particles emitted directly from mobile sources tend to have a great
potential to form PM2.5 in the atmosphere than do volatile organic
compounds. As such, the monetized benefits of abating a ton of mobile
source carbonaceous particles is significantly greater than that of
reducing a ton of volatile organic compounds. 

 Abt Associates, Incorporated. Environmental Benefits Mapping and
Analysis Program (Version 2.4). Prepared for Environmental Protection
Agency, Office of Air Quality Planning and Standards, Air Benefits and
Cost group. Research Triangle Park, NC.  2005. 

  "U.S. Environmental Protection Agency, February 2006.  Technical
Support Document for the Proposed PM NAAQS Rule, Response Surface
Modeling, Office of Air Quality Planning and Standards, Research
Triangle Park, NC (Docket # OAR_2001_0017_0770).

 U.S. EPA, Regulatory Impact Analysis for the PM2.5 NAAQS. 2006. 

 Benefits estimated using a 7% discount rate would be somewhat smaller.
The partial attainment valuation estimates in the Ozone RIA derived
using a 7% discount rate are approximately 15% lower than those using a
3% discount rate.

 PAGE   

 PAGE   4 

