Appendix XI: Additional Benefits Information

6.1 Developing an air quality estimate of full attainment with the
alternative ozone standards

As discussed in chapter 3, the modeled attainment scenarios were not
sufficient to simulate full attainment with each of the three
alternative ozone standards analyzed. To meet our analytical goal of
estimating the human health benefits of full simulated attainment with
each of these standard alternatives, it became necessary to derive an
estimate of the full attainment air quality increment through a simple
monitor rollback approach. 

We rolled back the values at each monitor such that no monitor in the
U.S. exceeded the alternative standard in question. This approach makes
the bounding assumption that ozone concentrations can be reduced only at
monitors projected to exceed the alternative standards. From a benefits
perspective, this approach leads to a downward bias in the estimates
because populations are assumed to be exposed at a distance weighted
average of surrounding monitors. Thus, any individual’s reduction in
exposure from a change at a given monitor will be weighted less if there
are other attaining monitors in close proximity. 

We determined projected attainment status of each monitor by calculating
design values (see chapter/appendix __ for a discussion of this
technique). However, to estimate changes in ozone-related health effects
resulting from improvement in air quality, the BenMAP model requires a
series of metrics. When performing a benefits assessment with air
quality modeling data, BenMAP calculates these metrics based on the
distribution of CMAQ-modeled hourly ozone concentrations for the ozone
season. However, because we were performing a benefits assessment based
on monitor values that have been rolled-back, it was necessary to derive
each of these metrics outside of the BenMAP model.  Thus, we first
developed a scaling ratio that related the calculated design value to
each of the ozone metrics.

A summary of this procedure is as follows:

Import partial attainment 0.080 calculated design values into the BenMAP
model

Perform a spatial interpolation of these design values using the Voronoi
Neighborhood Averaging algorithm. Design values are then interpolated to
the CMAQ grid cell.

Import distribution of air quality modeled daily and hourly ozone
concentrations into BenMAP. Create air quality grid in BenMAP using
spatial and temporal scaling technique. This procedure creates grid cell
level summer season ozone metrics (1 hour maximum, 5 hour average, 8
hour maximum, 8 hour average and 24 hour average).

Calculate grid cell-level ratio of each ozone metric to calculated
design value. The result of this calculation is a grid cell-level ratio
of metric to design value that can then be subsequently used to scale
the calculated design value and thus derive each of the metrics.

After having calculated these scaling ratios we then performed the
monitor rollback as follows:

Roll back the calculated 0.080 ppm partial attainment design value to
just equal the 0.080 ppm standard. This process creates a new baseline
design value grid.

Scale the design value grid cell values to ozone metric grid cell values
by using ratios described above. 

Create new 0.084 ppm baseline air quality grid from grid cell-level
ozone metrics.

Roll back the calculate calculated 0.070 ppm and 0.065 ppm partial
attainment design values at each monitor to just each the 0.070 ppm and
0.065 ppm standards, respectively.

Scale the calculated full attainment design value to grid cell-level
ozone metric using ratios described above.

Create new 0.070 ppm and 0.065 ppm air quality grids from grid
cell-level ozone metrics.

Perform benefits analysis with baseline and control grids.

To develop a 0.075 ppm full attainment air quality grid we performed an
interpolation of the 0.070 ppm full attainment air quality grid, rather
than a monitor rollback. This interpolation entailed the following
steps:

We identified any monitors that were projected to not attain 0.075 ppm
alternative in the 0.084 ppm base case air quality grid. 

For these monitors we calculated an adjustment factor that would scale
down the air quality improvement at that monitor. The purpose of this
adjustment was to ensure that the improvement in air quality at that
monitor reflected the attainment of the 0.075 ppm standard. This ratio
was calculated by dividing the improvement in the design value necessary
to attain 0.075 ppm by the improvement in the design value necessary to
attain 0.070 ppm. For example, a monitor whose baseline is 0.084 would
receive 2/3 of the air quality improvement from attaining 0.075 ppm than
they would from attaining 0.070 ppm.

We then interpolated these monitor-specific ratios to the grid
cell-level in BenMAP. 

Finally, we used these grid cell-level ratios as the basis for scaling
down the grid cell-level estimates of incidence and valuation from the
0.070 ppm analysis. 

6.2 Partial Attainment PM2.5 Incidence and Valuation Estimates

Tables __ through __ below summarize the estimates of PM2.5 incidence
and valuation resulting from the 0.070 ppm partial attainment scenario.
These estimates provided the basis for the full attainment PM2.5
co-benefit estimates found in Chapter 6 of this RIA. 

Table 6-15: Illustrative 0.070 ppm Partial Attainment Scenario:
Estimated Reductions in PM Premature Mortality associate with PM
co-benefit (95th percentile confidence intervals provided in
parentheses)

	Eastern U.S.	Western U.S. Excluding California	California	National PM
co-benefits

Mortality Impact Functions Derived from Epidemiology Literature

	ACS StudyA	510

(170--840)	0.17

(0.06--0.27)	47

(16—77)	550

(190--920)

Harvard Six-City StudyB 	1,100

(570--1,700)	0.4

(0.18--0.6)	110

(53--160)	1,300

(630--1,900)

Woodruff et al 1997 (infant mortality)	1.1

(0.5—1.7)	0.04

(0.02--0.06)	0.14

(0.07--0.2)	1.3

(0.6—2)

Mortality Impact Functions Derived from Expert Elicitation



Expert A	1,600

(170--3,000)	77

(8.2--150)	140

(15--270)	1,800

(190--3,400)

Expert B	1,200

(140--2,600)	56

(4--130)	110

(13--240)	1,400

(160--3,000)

Expert C	1,200

(140--2,700)	58

(6.6--130)	110

(12--250)	1,400

(150--3,000)

Expert D	820

(85--1,400)	40

(4.1—68)	76

(7.8--130)	940

(96—1,600)

Expert E	2,000

(890—3,000)	95

(44--150)	180

(82--280)	2,200

(1,000--3,500)

Expert F	1,100

(740--1,600)	51

(33—74)	99

(67--150)	1,200

(840—1,800)

Expert G	690

(0—1,300)	34

(0—65)	63

(0--120)	790

(0--1,500)

Expert H	880

(-46--2,100)	43

(-2.2—100)	80

(-4.2--200)	1,000

(-52—2,400)

Expert I	1,200

(60--2,200)	57

(3--110)	110

(5.5--200)	1,300

(69—2,500)

Expert J	950

(230--2,200)	46

(11—110)	87

(21--200)	1,100

(260--2,500)

Expert K	190

(0--970)	8.8

(0—47)	19

(0—95)	220

Expert L	860

(120--1,600)	33

(0.04--81)	79

(11--150)	970

(130—1,900)



Table 6-16: Illustrative 0.070 ppm Partial Attainment Scenario:
Estimated Reductions in Morbidity Associated with PM Co-benefit (95th
percentile confidence intervals provided in parentheses)

	Eastern U.S.	Western U.S. Excluding California	California	National PM
co-benefits

Morbidity Impact Functions Derived from Epidemiology Literature

	Chronic Bronchitis (age >25 and over)	380

(42-720)	12

(1.3—21)	43

(4.6—81)	440

(47—820)

Nonfatal myocardial infarction (age >17)	1,100

(560—1,700)	0.4

(0.2—0.6)	94

(47—140)	1,200

(610—1,800)

Hospital admissions--respiratory (all ages)	130

(59—200)	--	10

(4.4-15)	140

(63—220)

Hospital admissions-- cardiovascular 

(age >17)	270

(160—370)	--	20

(12—28)	290

(170—400)

Emergency room visits for asthma 

(age <19)	560

(310—820)	--	22

(12—32)	590

(320—850)

Acute bronchitis (age 8-12)	990

(-130—2,100)	32

(-4.1—67)	130

(-17—270)	1,200

(-150—2,400)

Lower respiratory symptoms (age 7-14)	8,400

(3,600—13,000)	3.6

(1.6—5.5)	1,200

(520—1,900)	9,600

(4,200—15,000)

Upper respiratory symptoms (asthmatic children age 9-18)	6,100

(1,500—11,000)	2.6

(0.7—4.6)	870

(220—1,500)	7,000

(1,800—12,000)

Asthma exacerbation (asthmatic children age 6--18)	7,700

(550—24,000)	3.4

(0.24—11)	1,100

(77—3,400)	8,700

(620—28,000)

Work loss days (age 18-65)	53,000

(46,000—61,000)	20

(17—22)	7,200

(6,100—8,200)	61,000

(52,000—69,000)

Minor restricted activity days (age 18-65)	320,000

(260,000—370,000)	120

(100—140)	42,000

(35,000—49,000)	360,000

(300,000—420,000)

 	 	 	 	 

A All estimates rounded to two significant figures. As such, confidence
intervals may not be symmetrical and totals will not sum across columns.
All estimates incremental to 2006 PM NAAQS RIATable 6-33: Illustrative
Strategy to Partially Attain 0.070 ppm: Estimated Partial Attainment
Value of Reductions in PM2.5-Related Premature Mortality Associated with
PM co-benefit (3 percent discount rate, in millions of 1999$) 95th
Percentile Confidence Intervals Provided in Parentheses

	Eastern U.S.	Western U.S. Excluding California	California	National PM
co-benefits

Mortality Impact Functions Derived from Epidemiology Literature

	ACS StudyA	$2,900

($410--$6,700)	$1

($0.14--$2.2)	$270

($38--$610)	$3,200

($450--$7,300)

Harvard Six-City StudyB	$6,600

($1,100--$14,000)	$2.1

($0.35--$4.6)	$610

($98--$1,300)	$7,200

($1,200--$15,000)

Woodruff et al 1997 (infant mortality)	$6.3

($1--$14)	$0.2

($0.03--$0.5)	$0.8

($0.12--$1.7)	$7.3

($1.1--$16)

Mortality Impact Functions Derived from Expert Elicitation



Expert A	$9,100

($810--$23,000)	$440

($51--$1,100)	$830

($75--$2,100)	$10,000

($930--$26,000)

Expert B	$6,900

($470--$21,000)	$320

($16--$1,000)	$630

($42--$2,000)	$7,900

($520--$24,000)

Expert C	$6,800

($620--$21,000)	$340

($38--$990)	$630

($57--$1,900)	$7,800

($710--$23,000)

Expert D	$4,800

($500--$11,000)	$230

($30--$550)	$440

($46--$1,000)	$5,400

($570--$13,000)

Expert E	$11,000

($1,800--$25,000)	$550

($120--$1,200)	$1,000

($160--$2,300)	$13,000

($2,000--$28,000)

Expert F	$6,200

($1,100--$14,000)	$290

($67--$630)	$570

(100--$1,200)	$7,100

($1,300--$15,000)

Expert G	$4,000

(0--$11,000)	$200

(0--$530)	$370

(0--$990)	$4,600

(0--$12,000)

Expert H	$5,100

($11--$16,000)	$250

(0.7--$770)	$470

($1--$1,500)	$5,800

($12--$18,000)

Expert I	$6,800

($570--$17,000)	$330

($35--$840)	$620

($53--$1,600)	$7,700

($650--$19,000)

Expert J	$5,500

($700--$17,000)	$270

($44--$810)	$500

($64--$1,500)	$6,200

($790--$19,000)

Expert K	$1,100

(0--$6,500)	$51

(0--$320)	$110

(0--$650)	$1,300

(0--$7,500)

Expert L	$5,000

($440--$13,000)	$190

($0.2--$640)	$460

($39--$1,200)	$5,700



Table 6-34: Illustrative Strategy to Partially Attain 0.070 ppm:
Estimated Partial Attainment Monetary Value of Reductions in
PM2.5-Related Premature Mortality Associated with PM co-benefit (7
percent discount rate, in millions of 1999$) 95th Percentile Confidence
Intervals Provided in Parentheses

	Eastern U.S.	Western U.S. Excluding California	California	National PM
co-benefits

Mortality Impact Functions Derived from Epidemiology Literature

	ACS StudyA	$2,500

($350--$5,600)	$0.8

(0.1--$1.9)	$230

($32--$520)	$2,700

($380--$6,100)

Harvard Six-City StudyB	$5,600

($890--$12,000)	$1.8

($0.3--$3.9)	$510

($82--$1,100)	$6,100

($980--$13,000)

Woodruff et al 1997 (infant mortality)	$5.3

($0.81--$12)	$0.2

($0.03--$0.38)	$0.7

($0.1--$1.4)	$6.2

($0.9--$14)

Mortality Impact Functions Derived from Expert Elicitation



Expert A	$7,700

($880--$19,000)	$370

($43--$940)	$700

($81--$1,800)	$8,700

($780--$22,000)

Expert B	$5,800

($510--$18,000)	$270

($13--$870)	$530

($46--$1,700)	$6,600

($440--$21,000)

Expert C	$5,800

($660--$17,000)	$280

($32--$830)	$530

($61--$1,600)	$6,600

($600--$19,400)

Expert D	$4,000

($520--$9,500)	$200 

($26--$470)	$370

($48--$880)	$4,600

($480--$11,000)

Expert E	$9,500

($2,000--$21,000)	$470

($98--$1,000)	$870

($180--$1,900)	$11,000

($1,700--$24,000)

Expert F	$5,300

($1,300--$11,000)	$250

($57--$530)	$480

($120--$1,000)	$6,000

($1,100--$13,000)

Expert G	$3,400

(0--$9,100)	$160

(0--$440)	$310

(0--$830)	$3,800

(0--$10,00)

Expert H	$4,300

($12--$13,000)	$210

($0.6--$650)	$390

($1.1--$1,200)	$4,900

($10--$15,000)

Expert I	$5,900

($600--$14,000)	$290

($30--$700)	$540

($55--$1,300)	$6,700

($550--$16,000)

Expert J	$4,600

($750--$14,000)	$230

($37--$680)	$420

($69--$1,300)	$5,300

($670--$16,000)

Expert K	$920

(0--$5,500)	$43

(0--$270)	$93

(0--$550)	$1,100

Expert L	$4,200

($460--$11,000)	$160

($0.16--$540)	$380

($42--$1,000)	$4,700

($400--$13,000)





Table 6-35: Illustrative Strategy to Partially Attain 0.070 ppm:
Estimated Partial Attainment Monetary Value of Reductions in Risk of
PM2.5-Related Morbidity Reductions Associated with PM co-benefit (in
millions of 1999$) 95th Percentile Confidence Intervals Provided in
Parentheses

	Eastern U.S.	Western U.S. Excluding California	California	National PM 

co-benefits

Morbidity Impact Functions Derived from Epidemiology Literature

	Chronic Bronchitis (age >25 and over)	$160

($8.7--$720)	$4.7

($0.3--$22)	$17

($1--$80)	$180

($10--$820)

Nonfatal myocardial infarction (age >17)

3% discount rate	$93

($25--$200)	$0.03

($0.01--$0.7)	$8

($2.1--$17)	$100

($27--$220)

Nonfatal myocardial infarction (age >17)

7% discount rate	$90

($23--$200)	$0.03

($0.01--$0.7)	$7.7

($2--$17)	$98

($25--$220)

Hospital admissions--respiratory (all ages)	$2.1

($1--$3.1)	---	$0.2

($0.08--$0.2)	$2.3

($1.1--$3.4)

Hospital admissions-- cardiovascular (age >17)	$5.5

($3.4--$7.5)	---	$0.4

($0.3--$0.57)	$5.9

($3.7--$8.1)

Emergency room visits for asthma (age <19)	$0.2

($0.04--$0.3)	---	---	$0.2

($0.09--$0.26)

Acute bronchitis (age 8-12)	$0.4

($-0.02--$1)	---	$0.05

($-0.002--$0.1)	$0.4

($-0.02--$1.2)

Lower respiratory symptoms (age 7-14)	$0.1

($0.04--$0.27)	---	$0.02

($0.006--$0.04)	$0.15

($0.05--$0.3)

Upper respiratory symptoms (asthmatic children age 9-18)	$0.2

($0.03—0.38)	---	$0.022

($0.005--$0.05)	$0.2

($0.04--$0.44)

Asthma exacerbation (asthmatic children age 6--18)	$0.34

($0.03--$1.3)	---	$0.05

($0.004--$0.18)	$0.4

($0.03--$1.5)

Work loss days (age 18-65)	$5.3

($4.6--$6)	---	$0.9

($0.7--$1)	$7.4

($6.4--$8.3)

Minor restricted activity days (age 18-65)	$7.9

($0.37--$15)	---
$1.1⠍〤〮ⴵ␭⤲␇ഹ␨⸰㈴ⴭㄤ⤷܇ޠޠޠޠޠഇ 敂䵮
偁吠捥湨捩污䄠灰湥楤散ⱳ䄠瑢䄠獳捯慩整㩳䴠祡㈠〰
⸵倠条⁥ⵃ㈱മ̍഍ഄ̍഍ഄግ䅐䕇†ക഍倓䝁⁅ᐠᔱ
഍

഍

഍഍⁁桔⁥獥楴慭整椠⁳慢敳⁤湯琠敨挠湯散瑮慲楴湯
爭獥潰獮⁥䌨刭 畦据楴湯搠癥汥灯摥映潲⁭桴⁥瑳摵
⁹景琠he American Cancer Society cohort reported in Pope et al
(2002), which has previously been reported as the primary estimate in
recent RIAs

B Based on Laden et al (2006) reporting of the extended Six-cities
study; to be reviewed by the EPA-SAB for advice on the appropriate
method for incorporating what has previously been a sensitivity
estimate.

C All estimates rounded to two significant figures. As such, confidence
intervals may not be symmetrical and totals will not sum across columns.
All estimates incremental to 2006 PM NAAQS RIA

A The estimate is based on the concentration-response (C-R) function
developed from the study of the American Cancer Society cohort reported
in Pope et al (2002), which has previously been reported as the primary
estimate in recent RIAs

B Based on Laden et al (2006) reporting of the extended Six-cities
study; to be reviewed by the EPA-SAB for advice on the appropriate
method for incorporating what has previously been a sensitivity
estimate.

C All estimates rounded to two significant figures. As such, confidence
intervals may not be symmetrical and totals will not sum across columns.
All estimates incremental to 2006 PM NAAQS RIA.

A The estimate is based on the concentration-response (C-R) function
developed from the study of the American Cancer Society cohort reported
in Pope et al (2002), which has previously been reported as the primary
estimate in recent RIAs

B Based on Laden et al (2006) reporting of the extended Six-cities
study; to be reviewed by the EPA-SAB for advice on the appropriate
method for incorporating what has previously been a sensitivity
estimate.

C All estimates rounded to two significant figures. As such, confidence
intervals may not be symmetrical and totals will not sum across columns.
All estimates incremental to 2006 PM NAAQS RIA

