TECHNICAL SUPPORT DOCUMENT
FOR THE
REGULATORY IMPACT ANALYSIS
FOR THE PARTICULATE MATTER
AND OZONE
NATIONAL AMBIENT AIR QUALITY
STANDARDS AND PROPOSED
REGIONAL HAZE RULE
IWfcfs TSD
Prepared by:
Innovative Strategies and Economics Group
Office of Air Quality Planning and Standards
U.S. Environmental Protection Agency
Research Triangle Park, N.C.
July 17, 1997
-------
This document was developed to provide technical support for the Ozone and PM
NAAQS and Regional Haze 1997 RIA as prepared by EPA professional staff. The analysis and
conclusions presented in this report are those of the authors and should not be interpreted as
necessarily reflecting the official policies of the US EPA. It should be noted that this document
was developed to derive national air quality, cost, benefit and economic impact estimates for these
rules. However, while useful to derive national estimates, information associated with any given
county or area is subject to significant uncertainties and should not be used to predict the
attainment status, cost or benefits which may result in a specific county or area
-------
The purpose of this technical support document is to provide additional background'
information on the methods and data used to calculate national benefits estimates for the
paniculate matter (PM) and Ozone National Ambient Air Quality Standards (NAAQS) and
Proposed Regional Haze Rule, dated July 16 1997. Background information is provided for a
number of components of benefits estimation. These components include:
(1) The results of a quantified uncertainty analysis using a Monte Carlo technique;
(2) The results of the Paniculate Matter benefits analysis, presented on a regional level of
aggregation;
(3) A memo documenting the methods and data used to estimate changes in health and
welfare incidences;
(4) A memo documenting the methods and data used to aggregate and monetize the
health and welfare incidence reductions;
(5) A discussion of the uncertainties associated with estimating benefits;
(6) A discussion of the methods and data used to estimate health benefits associated with
air toxics emission reductions.
11
-------
Quantified Uncertainty
Analysis
-------
1.0 Quantified Uncertainty Analysis
The monetized benefit results presented in chapter 12 of the regulatory impact
analysis (RIA) are point estimates. Point estimates are the most likely value, based on the
available information and is comparable to a mean or, or 50th percentile, estimate. There is
substantial uncertainty, however, associated with the monetized benefits for with each NAAQS
alternative.
The total monetized benefit associated with each health or welfare endpoint depends on
how much the incidence of the endpoint changes when a given NAAQS is implemented (e.g., how
many premature deaths will be avoided) and how much each unit of change is worth (e.g., how
much a premature death avoided is worth). The uncertainty associated with each of these
components is expressed as a distribution of values. For example, there is a distribution of
possible incidence values and a distribution of possible unit dollar values for each endpoint. The
uncertainty associated with the total monetized benefit for an endpoint combines the uncertainties
from these two sources
The distribution of the monetized benefit associated with each endpoint is estimated by
Monte Carlo methods. For each iteration of the model, a value is randomly drawn from the
incidence distribution for the endpoint and a value is randomly drawn from the unit dollar value
distribution for the endpoint. The monetized benefit for that iteration is the product of the two.
If this is repeated for many (e.g., thousands) of iterations, the distribution of the monetized benefit
associated with the endpoint is generated. The same Monte Carlo procedure is then used to
combine the monetized benefits from different (non-overlapping) endpoints to generate a
distribution of total monetized benefits. The mean of this distribution is used as the point estimate
of total monetized benefits. A Monte Carlo uncertainty analysis of the monetized benefits of
attaining the PM?., 15/65 standard, the PM,0 50/150 standard (99th percentile), and the ozone .08,
4th max. standard are presented in Tables 1 through 4.
-------
Table l:PMu: 15/65 standard
Quantified Uncertainty Range* for National Annual Monetized Health and Welfare Benefits'
' Estimates are incranental to the current PMNAAQS (50 |ig/m'annual; 150 ugAn'daily)
(billions of 1990 $; year = 2010)
ENDPOINT1
•Mortality'ishoit-tenn exposure
long-term exposure
•Chronic Bronchitis
Hospital Admissions:
•all respiratory (all ages)
all resp. (ages 65+)
pneumonia (ages 65+)
COPD(ages65+)
•congestive heart failure
•ischemic heart disease
•Acute Bronchitis
•Lower Respiratory Symptoms
•Upper Respiratory Symptoms
shortness of breath
•Work Loss Days
•Minor Restricted Activity Days
(MRADi)
Consumer Cleaning Cost Savings
Visibility
TOTAL MONETIZED BENEFITS
using short-term PM mortality
using long-term PM mortality
Partial Attainment Scenario
Sthpercentfle
$4.013
$10.970
$1.967
$0.021
$0.057
$0.029
$0.024
$0.016
$0.018
$0
$1
$0
$0
-$0.180
$0.227
$0.578
$0.197
$5.629
$22654
$33.515
Mean
$26305
$75.546
$21.281
$0073
$0.100
$0.046
$0038
$0035
$0049
$0001
$0.004
$0.001
$0.001
$0.014
$0.261
$1.001
$0.359
$7.754
$57670
$106.910
95th percentUe
$60.938
$186.893
$64.035
$0.130
$0.148
$0.065
$0.054
$0.056
$0.084
$0.002
$0.007
$0.002
$0.002
$0.210
$0.094
$1.451
$0.657
$9879
$113.257
$227.754
numbers may not completely agree dw? to rounding
2 only eodpotnts denoted with an • are aggregated into total benefits estimates
3 mortality estimates must be aggregated using either short-term exposure or long-term exposure but not both
due to double-counting issues
-------
Tabk2:PM,,: SO/ISO, 99th percentlle standard
Quantified Uncertainty Ranges for National Annual Monetized Health and Welfare Benefit*1
Estimates are incremental to the current PMNAAQS (SO ugfar1 annual; ISO ng/m* daily)
(billions of 1990 $; year - 2010)
ENDPOINT1
•Mortality'uhort-tenn exposure
long-term exposure
"Chronic Bronchitis
Hospital Admissions:
•all respiratory (all ages)
all resp. (ages 65+)
pneumonia (ages 65+)
COPD(ages6S+)
•congestive heart failure
•ischemic heart disease
•Acute Bronchitis
•Lower Respiratory Symptoms
•Upper Respiratory Symptoms
shortness of breath
asthma attacks
•Work Loss Days
•Minor Restricted Activity Days
(MRADs)
Consumer Cleaning Cost Savings
Visibility
TOTAL MONETIZED BENEFITS
using short-term PM mortality
using long-term PM mortality
Part
5th perceatfle
$0.267
$0.237
$0.178
$0.001
$0.003
$0.002
$0.001
$0.001
$0.001
SO
so
$0
$0
-$0.004
$0008
$0.019
$0018
$1.201
$2.666
$2609
Hal Attainment Scenarii
Mean
$1.728
$1.621
$1.937
$0.002
$0.006
$0.003
$0.002
$0002
$0.003
$0
$0
$0
$0
$0
$0.009
$0.034
$0032
$1.615
$5.402
$5.294
>
95th percentfle
$4.039
$4.008
$5.838
$0.004
$0.008
$0.004
$0.003
$0.003
$0.005
$0
$0
$0
$0
$0005
$0010
$0.049
$0.059
$2.022
$9.956
$9839
1 numbers may not completely agree due to rounding
2 only endpomts denoted with an * are aggregated into total benefits estimates
3 mortality estimates must be aggregated using either short-term exposure or long-term exposure but not both
due to double-counting issues
-------
Table 3: Oxone: .08 ppm, 4th max. standard
QuanHned Uncertainty Ranges for National Annual Mooeozed Estimates of Selected Benefits Categories'
Estimates are incremental to the 2010 baseline and includes benefits associated wim the ouient ozone NAAQS (.12 ppm, 1-hour)
ENDPOINT1
Oxone Health:
•L Mortality
Hospital Admhskms
•^ _n .MMMMtnnf /mil •a««\
•2. all rtapmimy (tii ages;
ad top. (ages 65+)
pneiunonia (ages 65+)
CQPD(ages65+)
emer. dept visits for asthma
•3. Acute Respiratory Symptoms
(any of 19)
asthma attacks
MRADs
•4. Worker Productivity
Ancillary PM Health:
*1. Mortality4: short-term eip.
long-term eiposure
•2. Chronk Bronchitis
Hospital Admissions:
•3. all respintory (all ages)
all resp. (ages 65+)
pneumonia (ages 65+)
COPD(ages65+)
•4. congestive heart failure
•5. ifchetnic heart disease
•6. Acute Bronchitis
•7. Lower Respiratory Symptoms
•8. Upper Respiratory Symptoms
shortness of breath
asthma attacks
•9. Work Loss Days
•10. Minor Restricted Activity Days (MRADs)
TOTAL MONETIZED BENEFITS
using short-term PM mortality
using long-term PM mortality
Partial Attainment Scenario
SthpereennV
41429
J0.004
JOD12
S0.004
-W.004
$0.002
SO
SO
SO
$0.024
$0.098
$0.279
S0.023
$0.001
$0.001
SO
SO
SO
SO
SO
SO
so
so
-S0.004
$0.006
$0015
$0.421
$0.650
Mean
$0345
$0.011
$0.036
$0.016
$0.005
$0.003
$0.002
$0
SO
$0.024
$0.643
S1.920
$0.246
$0.002
$0.001
$0.001
$0
$0
$0001
so
$0
$0
so
so
$0.007
$0027
$2.208
$3305
95th perotntik
$3.677
$0.019
$0.061
$0.031
$0.016
$0.004
$0004
$0
$0
$0.024
$1.490
$4.749
$0734
$0003
$0.002
$0.001
$0.001
$0.001
$0001
$0
$0
$0
$0
$0.005
$0.008
$0.039
$4935
$7.197
numbers may not completely agree due to rounding
only endpoints denoted with an • are aggregated into total benefits estimates
4 5th penentile benefits are aggregated without ozone mortality (i e, ozone mortality - zero)
PM mortality estimates must be aggregated using either short-term exposure or long-term exposure but not both due to
double-counting issues
-------
Tables 1 through 4 present the 5th percentile, mean, and 95th percentile estimates
associated with the selected PM and ozone standards. Tables 1 and 2 present uncertainty
estimates associated with all health and welfare benefits categories included in the total PM
benefits estimates. Table 1 shows that annual monetized benefits estimated for the selected PMi$
standard at the 5th percentile range between $23 billion and $34 billion. At the mean, benefits
range between $58 billion and $107 billion and at the 95th percentile, benefits range between
$113 billion and $228 billion.
Table 2 shows annual monetized benefits estimated for the selected PMto standard at the
5th percentile range between $2.6 billion and $2.7 billion. At the mean, benefits range between
$5.3 billion and $5.4 billion and at the 95th percentile, benefits range between $9.8 billion and $10
billion In all cases, ranges are driven by the difference between the long-term and short-term PM
mortality risk reduction estimates.
Table 3 presents the uncertainty estimates associated with a selected set of ozone benefits
categories. Total benefits estimates at are calculated by aggregating all categories denoted with
an asterisk. However, note that the aggregation scheme at the 5th percentile truncates ozone
mortality results at zero since mortality estimates below zero are considered to be biologically
implausible.
The ozone uncertainty analysis is conducted for only a selected set of benefits categories
for which uncertainty distributions could be estimated. For example, notice that benefits
associated with increased yields from commodity crops and fruits and vegetables are not included
in Table 3. Additionally, the ozone benefits in Table 3 are calculated from the 2010 baseline,
before attainment of the current ozone standard is modeled. This calculation prevents the results
in Table 3 from being directly comparable to all other ozone benefits estimates presented in this
RIA. All other estimates are presented incremental from the current ozone standard. The ozone
benefits uncertainty analysis is calculated from the 2010 baseline rather than incremental from
partial attainment of the current ozone standard because all ozone benefits analyses are calculated
in this manner. The majority of the benefits estimates are calculated using point estimates, which
allow subtraction of benefits attributable to the current standard.
However, this method of estimating benefits is problematic for the uncertainty analysis,
because the simple subtraction method cannot be applied to the 5th percentile and 95th percentile
results. To provide uncertainty benefits results from a comparable baseline (incremental from
partial attainment of the current ozone standard) and for a complete set of benefits categories, a
two-step procedure is presented below:
(1) The mean estimates presented in the ozone uncertainty analysis are comparable (within a
few percentage points) to benefits calculated in point estimate mode. A ratio can be
estimated which compares the point estimate results (which are calculated incremental to
partial attainment of the current ozone standard) to the mean estimates generated by the
uncertainty analysis (which are calculated prior to partial attainment of the current ozone
-------
standard). This ratio is calculated for the set of benefits categories that are common to
both analyses. This ratio is approximately equal to 0.51, meaning that for the 50th
percentile, benefits estimates calculated incremental from partial attainment of the current
ozone standard are approximately 51% of the benefits estimates calculated from the 2010
baseline, prior to partial attainment of the current ozone standard. This ratio of 0.51
provides a method for estimating potential 5th and 95th percentile benefits estimates
associated with partial attainment of the .08,4th max., standard, incremental from partial
attainment of the current ozone standard. Applying the 0.51 ratio to the 5th percentile
results in Table 3 provides a 5th percentile result of $0.251 billion to $0.332 billion. Using
the same procedure to estimate the 95th percentile result provides an estimate of $2.521
billion to $3.677 billion.
(2) The 5th and 95th percentile estimates calculated in step (1) above do not include benefits
associated with a number of other benefits categories, such as commodity crops and fruits
and vegetables. To include these categories in the 5th and 95th percentile estimates, it is
assumed that the distribution for the remaining benefits categories are similar to those
included in the uncertainty analysis. Using the point estimate results, it is estimated that
the remaining benefits categories (commodity crops, fruits and vegetables, nitrogen
deposition, and air toxics) contribute approximately 11 percent to 18 percent to the total
monetized benefits associated with partial attainment of the .08, 4th max. standard,
incremental from partial attainment of the current ozone standard. A midpoint of this
percentage range is 14.5 percent. This 14.5 percent factor is applied to the 5th and 95th
percentile estimates calculated earlier in step (1). The final results of these calculations are
presented in Table 4.
Table 4: Ozone: .08 ppm, 4th max. Standard
Ozone: Quantified Uncertainty Ranges for National Annual Monetized Benefits Estimates
Estimates are incremental from the current ozone NAAQS
(billions of 1990 $; year = 2010)
5th Percentile
$0.25 - $0.38
Mean
$1.3 -$2.1
95th Percentile
$2.9 -$4.2
-------
Participate Matter Regional
Benefits Results
-------
1.0 Regional Participate Matter Benefits
This section presents regional PM benefits results. The PM regions are identical to those
described in chapter 6 of the Regulatory Impact Analysis (RIA) for the Paniculate Matter (PM)
and Ozone National Ambient Air Quality Standards (NAAQS) and Proposed Regional Haze Rule.
Abbreviations used to identify each PM region are defined as follows:
WE = West
SE = Southeast
SC = South Central
RM = Rocky Mountain
NW = Northwest
ME = Midwest
Refer to chapter 6 of the RIA for a map detailing the borders of each region. In addition to the
six geographical regions, PM benefits are also provided at an aggregate national level, denoted as
Continental U.S.
Standards for which these PM benefits are provided include:
(1) the current PMW standard, 50 annual/150 daily ug/m3, 1 expected exceedance (pp. 1A-G);
(2) a 16 annual/65 daily ug/m3 PMM standard, 98th percentile (pp. 2A-G);
(3) a 15 annual/65 daily ug/m3 PM1$ standard, 98th percentile (pp. 3A-G);
(4) a 15 annual/50 daily ug/m3 PM15 standard, 98th percentile (pp. 4 A-G).
These regional benefits results are directly comparable to the PM cost results presented in chapter
6 of the RIA.
-------
»l 0 KKtau) PoH fiUnil* ftmlU - A*M«f IneWtne* ntir4,
At Qu»*y &c«ivto, PirttUAnirtrwt fj Y«ir
(JW S
PM
tlgrt
Uo(
Omk B
110
14
R
ilaitnmtttntto
Popfdli, IMt
l^ IIH
DIB
190S)
1701)
to
ID
KRAD
MO
Oiw,
0
CWKA f 1 «T
UWM
I4MU1
115
«7
«"
HouMtaM £olng Dinaoi
VldiBy-RtcniltaMJ
Ptdiftnl*»)Uihi, |HU
Irttt Kwxxnt vUnlfan
*«n tununla nkiiUm
Jr«t« »ea>e«»l» xifcnllnn
ue
>22S
PM tofrKKri (IncXidii len»4
-------
JOIONittovU P«M ttVnau tmHi • AttU*4tekfcAM vijK«Md»4B«mai
PU K/MQS <0tl
OrafcBnnMk
t. 10M
paaM uMlyih (10 UicUan)
45
ISO
J112
J7M
• Al R«tfJn1r, T»H MM (poeM «
IcfeMUR, TIKl,bl. tlH tpwMtiu^i
JcJunrnwdUorrtt, 1M*
Wwu« vid Uinh. 1««
(1
as
n
«
«o
97
0«diMV«tiL,tOtt
«jff«u Xittcii
202
37M
17W
1S7T
WoKUn Diy<
UFtAO
R40
HIM
HUM
777M
MM«y. Rxmfenl
tMH
r
wd Uiln. 19W
A*d tconcmlo vikallal
lit
111
118
n» KiMeW llncludn tomHtna mortdtyl
IUOO
19M
-------
JOIO Ntdml PoH EMnul* »«««• - AvolAid tncMmw wd UoMUttd ttnrt*
PM MAflS (OITM PMI4 •Undutf, SIMM, hcnrnMUHoOUMI O» itinjwO
AfrQnByfemfe. Ptrtbl AlUJxrurt, FiJ Y«nf
MocHlnl AO»S£JPU ftmtot etnflon)
EndiuM
Hwf
i 11 i»]
PM
Horti»y (Ionium tij).
Urtizy (ihott-twn op ) . P U to
Cn'iJi'lv BfWIChUl
pocM imVih (to iKCkm)
L1W»
(iLtBS
Ut
ti3
_122
to
to
so
to
HtK»
JO
JO
M
K
to
jo
Lmnr Roptnlory Sympliirnt
Upf>« Kllftrttoty Syn^toiM
Atthm/UtcU
U
1M
WMKlMiOtyl
1572
114 M
H
VouuMA lolna Otmigt
arm Hcnxrla vUnUcxi
JO
tl
10
PW SuMoUl (Kidu4H ton94«n) rnxMtyf
PM tuMotri (Inctudn PM10 il»rf4inn rat**)
PM Mddil ftxtudu PNP.J tbotMum mott»«Yl
tu
$31
-------
ZO10 Krtijnil Port Cttlml i Bmtti. AmUtd UcWiw* inj limited tmfU
PH HU01 (eumt PM 15 tintttt. 60/1M. tKnmtttll to CllTPt 09 MndViQ
Mt Qwky SctoiA) Pirtitl «nhr*», M Y«M
• SC PM R«d»i HJ irJail
Atoldt 4 hddMKi
(mlUlon
y eiuly (ihocUwm Of,} . PU 1 0
OnrteBnmMh
, 1994
VtkO
Sdwuti. iBMb
1(3
1M
170
1924
Hmp AiMtikxn - CwaMth* HMA Ftkn
How AiMiilom. ln*wrkH.irt ohill*
Aai>enneM»
Uwir RnpMocy Synpomi
Upp«r RnpkMoiy Syrrpiona
Startntuettntfi
4W
M71
im
6ZM
WcAlonteyi
URAD
RW
S2KH
«WI7
1MU3
t!7
HouMWd Silng Dump
VhMty> R«tf»M)eAil
VMAV-ftMklMtil
wKl MonoMO viiotion
»~*«tollan
in«l
-------
JOIO ttofenl PcW ErinuM emfli - *wWM IncUtm »d Umted BmtU
PM NMQS (onM PU10 lUirfiri. WIM, Nn
-------
2010 Nillaul PoH Eatniti 6«n«n i - AMkM takttnM ud MoMM bnrt*
PM tUAQS (arm* P«HO ttinchrt, SOflM. tmrwtil lo Cuwt O» lUfidM)
MrOodtyfewub, PiRtilMilmK
IdxfeM*rti NWPUflidon(13mBon)
Mtent 1MOt>
IIMUhyQaigJMfiBV. •»»•»»>
poaM uulyilt (10 Uwtknt)
J»
11
0
r
a
a
r;
_HS
10
H»if
Holt Adrrbilna.JtdwricKiirtOliitM
«ohwta, OH f IH (paoM iMiyi(«)
Wtw<^ fl««A«, n» oalK/iiHtf,ii)
Sttmvtt, 1IMt.b.n IIHtpeeMtnttfitil
,1M»
. UH
onp1
to
to
_JO
to
I«M> NiifMIMy Synpocn*
4?
•u
1M
»
J01
to
to
_JO
tl
12
_!?
ti
12
S2
WefkUMDayi
HHAO
^A
»1U3
Soft* Oimip
i ct tcononiic ntaton
• Ido^ctrm
tl»rl-»KM
PM iirtrtom jlntludM PM] i ihq«1-<»rTTi mofltKy)
tios
-------
2010 NMfaMl P«W Idhwtt B«n«a< -AwW.d hddmel ind UOMUad tartl
PM HMOS (wmrt PUIO Mitt* KVItO. hcmMttalM OUTMl W «indv<)
Air Oudy Switrto: PuttalMitnwt, MVtu
•MtPUfl»olon(«lnJ«in)
bxfe«M
InMti hcM*nc«
ImlMenf 111! II
Mortrty (VorMmn «(•) • PUIO
" tictnrxtDt
.Adrrhllm.
po«M CUV* (1 0 tacUcn)
Sdiwvtt. IWto
n
144
1H7
m
MM
Al Riifkil«y (tl ig
TMtmcnftli, 1M4
juMtikm . lictwtrfc Mem DbMM
ScAwO, 1t*O,AA MM (pooM
SwWUll Mo Honli 1 1 "99
srfiMiti md Kort. ;g«s
II
11
at
-s
»
to
10
I.MW HMplrtloty Symptonn
Uppv ntvmtery Synplomi
20J7
«44
1097
MIW)
AtO
11
sr
>D
Oimn«
VflOAv-
dn« fwcunfc viknUixi
dncr cMMnio ntatlon
dnnli
14
-uH
PM MMo
-------
JO Id Hufantl PoH Eitbnin lma< - A«W*d lmU*nM utd ItaMted tot*
Pit HMQS (1«IU. hwmwUJ 10 Cwrm PU1D riudinQ
MrOnOySeMuto: PirtWAIUlnwt, MYiu
:fJM5rr«or)
MortiKy (kx^tMm of • ig« HH)
llvbxy (iMd-Mnn «>p 1 - PH} !
I34M
i 1Q(Ot)
571U3
lusae
Pnuvnonl* (tfti U*)
How AricH«utDh.tii
IKO
1100
580
JM
AaltBtmcMlt
Vvdr+a t* tmVi
*PmtS*"1"
If lewnyi
moi
1U047
44830
ROM
»t
ti
n
JO
in
Work ti
UMkO
MO
10
PtcMM ff^ Uiftn. 1M4
dntt *wnocnte niotiwv
*»« KomxT^ nlmUM
I'M itaxoW (Wthj4»« tontKtmi mottiltr)
PMM ihorUlim nwrtiHtvl
J?l)5
UiiC
!OTrtgil«
-------
3010 Nuknil PcH EttnM* B*wl>» • /Wolitd Incttmt 1M Umtkwi <«•«•
PM NMQS (1 Mi, hown*rt»l to OmA PM 10 «udu«l
UritM Ant
CraJD«M
AvoldMl Incldtnct
(crntfti*)
M«i>t4uy B«rwflti
ImQlQPf 11101)
2U
IM
11217
(SIO
Ofrrta BrmcMh
(BOO
how
M RnpMgfy («l »««}
•nmUnttlL, 1M4
SdM« fill fIN (poeM
-------
1010 KiUentl POM EitlmMt 8«mfl >
Py WAQS (IWU, hcmiwiil 10 Cunri PU10 knw(. FJYii/
MoJdrt AIM' J6 PM Bttfoo l« T n*ai)
-§**-
[rrJOoni HI01I
405
124
I2UO
woa
fe^yt""
AMubw.
11'T
MhHtt. WM^hfl. TIM IfcdBftMljlU)
., IBM
109
It
Aali CrancttJt
OiAv •( H. 1IJT
If*
71T4
TtlO
Won ton
HMD
MD
O«re, 1W7
OltTO KH) R««**l, UBS
•1MO
17
J1S
HamhoM Sottig (KJMt*
, 1M4
driaKonxrk>
dnawtnenfe^
rM lUXMil (biduiHt loojH«m mwtaty)
PMliAlolJ flnclu*j> PM2.J «h
-------
JOIQ mtloni] PsW E«ttflj« t*n*fli.
I, hcnmotU l» Cun«M PM10 dindirt)
/* Only SttMifo' PirttaJWilrmtrt. FUIY»«l
SC PU RiUcn (7 1 ">•«>)
U«Uky (tag-twin w Witt*)
Pap* rflL. IBM
SdoMO i* it. t*Hi
199
J214
Orp>
-------
J01C Nrienil P»« EtUiuu grant. AwUid IncUMC* «nd UoMtod fl4««««
PU HMOS (11 CKlM
. IWi
IM
49
41
34
40
11
12
11
11
tl
Jl
AaliBrancMU
Ctoctny « iL. 1«9
Mtatittii
HLMitty*
m
JW7
ion
ITJ7
(0
to
to
HMO
MO
OT04J
til
>e
tn
Sdkig Dung*
dnM teonomfa nkoOon
*t«l tamnta itlutiai
PH fcttDiri (Mdu<)*« I
n* fcjUotri pncludM pm i thorn.
-------
JOtO Ntltonil PoW EMmli B.ix«i - ArcUtd Itttanc* ml Mwttted Bmftt
PM HMOS (1 tat. tostmmil \ t Cwrw* PM10 tUntwit)
AKJ»«ySc«n»rto Pinbl Muhnwt, Fmp^c| Monorrk nlatlcn
*«» *c«nxnla nlatlon
dnd tnmmlo nkaton
PM (vMoOl (tKludti (oofi-l^rn mwtdtty}
lixludu PM! J iiiort-iinn mortWXv)
J72
-------
2010 Ntifenil PoH £«*iul« e«n.n» . Awkfcd hddtnc* Md UoraawJ Bmrftt
PM KMQS <««M4. hcnfl>«
-------
»10 NKJOMl Pert Edtmti tmfti - AwkUd IncUm ud IbMfetd MnMf
PU fUAQS (IMS, hmmm] l« Cunrt PMtO *»*t*d)
PIRU MUtawt, M Viv
Knitpclnl
AyoJdid Incldwic*
1KCHI
. 1W9«
IS44!
UU
OrorioBorKMtt
Pnutmonli liftt (!•
COfO (ifu ««)
HMO
»T7
2U4
2MB
172
ttoo
r RitptUMy Synpl«m
20104
WM9
1I41U
JH7M
U7R7I
ji
(4
SI
It
113
HMD
MD
JH7R7I
MIJftMl
SJ01
11003
to
*»o« i BMornb ralBtlai
now
11044
PM «ub(o4il flnoudtt PMi i »horl<«ra mmliayl
155113
-------
. AmkM inddMn ti4 VowtM Btnfti
PU NAAOS (1 Mi, IntttRMfUJ lo Cum* PU10 MKtu4
A»QaltyScrafc. PanMAAihtMnt, MYw
UDdtM Ani- SE PM Ridan (» S mHoci>
EndpoM
Mortify MiOfMftinUB.t.PU2i
•hlrancMli
> AartHVm-
Hunum M iL, 1«M
Haw Mntulm.ltdwifcHMrlDhiiM
R»«f*no(y Symplono
AKhmi
>Aloi»C
AKidtt
WedcLoitOnyi
URAO
OKro, 1M7
O*» ml Roehtot*), 1«M
. WIT
A»>M«d Incliltnct
710
H317
JI1M
«''
tnt
11
it
tt
un
•an
tnu
U2<7
Ml
to
JO
w
to
dr»gt toamnte vUnOxi
dnd Ktmefnto nkjidon
im nxxtdty)
114
no
Hire
-------
3010 MtionaJ Port EKtnX» Omfli. AwhM IncMtncI tnd y«Mfe»4 ttntUt
fU WYWS ((3/U, tnwrartiTlt Cunrt PUtOflinjud)
SC PU R t^on (7 I moon] .
Eixtottt
WlMOCl
Uoo«ltiy GtnifM
UolUty (thai J«m up ) . PU3 6
ISM
HDii
flnwnanfe ;»p« »3«
• Conj«l>rt H«in Fitn
i AAnhtlam . lictofiic Hurt CNtMM
190
A3T
1M
1M
1IT
131
11H
10441
Ji
»
»3
II
-fi
VI
to
M
»
_!«
Work lw» Diyi
UAAO
111
to
HouHhoU Solng Oung<
-KtmiUofttl
- RiiW*r«lil
OMbi«,1MT
**« tcoiMcrlo ntaUo)
*«M KMWiMo nXnM liM'oiil {Indudn ton»<«fm nMrtt»y(
fM luMoOl QodudM ntit lho<^^^^m mort>avi
U
-------
2010HitloniJ P«M EotniM BwilHi. AwU«4 In44*c*indKnfertBtmttl
PM HMOS (IMS. JxrwiMil l« Cunrt PU10 (Undinl)
Ak Oaty Serartl PutttlAulvrwl, Ful VM/
Uodtktl Ani RM PM R«*B (3,« r*»)
[nOlonl lUOJt
21 r
M
JiM
Ti
101
M
How,
II
12
II
It
_H
to
to
w
w
__ffl
o
in
to
17
nw
»lt
lit)
1WS
om,
RAD
41H9
MSS»
115X0
HomtioM *<*i8 Oimij*
, IOM
*l A Honour ninUm
drict icuiMilo nkaHon
PM (ubtotil (IntJudtt teog-twm «V5rtJ«rt
PM ftJNfftil (Intlmlw PM2 1 thort-ltcin mort«flty)
Q
1
-------
JOIO tMbMl p«M.E«knili B«Mfti - AwWO Inddtnc* «fld UonMtaM hntfti
Ml HAAQS OW». iNnciMMllloCWnrt PUIO ««fird>
Mr Qut*y Smnufo piAlilAlutnwt.rJYtV
CntfixiM
Intnoni 1114)1
11 B
"
CHaic 6ro«i*k
• M RttftDoy (U ig*i)
« Kuplntity (tfu U*)
Htw. AAnlnkm.lwhKTik Hun Ohm*
i.AdrtHkif
• BnncMli
SchnrllMdUoirti, 1«M
toJMOttlVIt UMrli. IBM
DadwyidtriWt
to
to
M
to
to
to
to
to
M
-I
J1
JS
w
UN* HMplnlMV E»wplom»
n
292
M
WerkLoilDiyt
MRM>
22U
1«78T
8ttT
**4 400101*3 nkattn
dfMI ttooafflh vtlatfgn
PM tubtotJ^ (kxiudM toniMtnn
KludMPMl I iNofHtOT mortifltyl
_PM»yMo
-------
2010 KMtwJ Pott EitlffiMi B«ni1l«. Avt*Jrt ImktMC* ind MoMUud e*Ml»
t>U NAAOS (1VB!.
WE P U Ktokfl QtJ rrman)
CntoeM
ImWm 1110)1
UMUtyponirtwm wp • »o« ICKj
M«li»y ftJioil-C*™ «D 1 . PHI 6
its
170
HIT
Ctntte emtMh
HIM
M R«K*tt«y (U igt«)
, f lf« (px+3
li, 1MJ
Wwirt.indUofrtl.HM
212
4SO
1C!
1M
110
15)
JJ
lw*r R»K*»lwy Synptonw
0
-------
jOlfl tetfcnil Port Eittrwti B«wni - A«*M tndd«nc* Md UoMtnd MiwfM
PM N»AQ« (IBrt». hcnmMill«On«l f>U10 ftinhnQ
to! PwUilWiHmrt, FJY.W
•HEPKBtdenflt 11rtfall
CndMM
AnoMtd IncMmn
fndloni 1MOII
UwtilylthoiMifliitnl.PUlS
jH?T
Oirtrafl BfoncMii
SdiwifU. 1
"em
-Hi
Ho«».
CO/V <»fw «»;
K»tt
in
«47
929
tie
127
VI
StO
M
II)
AoltBnncMti
Imnr Rttplnloty Syir^Xof
70001
M
10
-&
WwKUuDtyt
UftAO
tai
to
us
HowJxDd Sofcg D4Jiuy»
VSit>«y.R«WMilaul
VMbMv-RnktortU
aria •menie ntaDai
(*r»d «conon* rakfftien
cfrad «conorNc vtioUon
PM tuMnttl pnAx}«« to«o-tnm morU«y|
PM tuMgtri (kvdudif 'MiS thwUtnn morttfM
5J14BC
J11S28
o WU irKKMUxl Nmltt
O
-------
JO 10 NrtMul PoM Efttmti l«n«ai - AwIM Inddax* ind MoMtiad B«*a<
Ml NAAQ5 (llflO, hcnmnUJ la Cum PM18 tfudirt)
M»d«l«d*m CofrtwtVU 8
^^
Rt/tnnc*
Atoldid inddwci
Motxt* (ttoM«mt no \ • PIH.6
<***ltto
1M>7
sni
J78XXJ
J37J11
•w
1427
110)
nw
JSS7
>?OS61
COK> (<««i 43*)
Hoip. AdmKilon . CongnlVi h4trt FUi*
Moio Mniiilax • !id»rafc Hitrt Dlitif t
tl09
J*B
Mt
1M
55J
lewv ftM^ntocy Stn^lomi
UpptrftMptiMxySvnVlotnt
SMmnettfHOl
HUM
MMM
419*11
WoA l«M Oty»
HRAO
AID
DIWJ1
1T4W7M
«KtJ8S(4
now
to
0,1«M
im
P»
-------
2016 MKfeotl PoH Ccthal* frtort. - Avolwd lndd«y» tnd MoMth»d Bmfli
PU NAAOS (1i«0, hcnmMiIyi
MAAD
I1MOW
Jll
HI
HaiHtaU Sc*ig otmigt
ChMtlU, 1MT
PliMordtndiltttv.
-------
3010 NtUonil Pdnt FKtruti KtntS i. Avdd*d lAMfanc* ind Vwtlttd frtmBt
PM K»AQS(IWW. hmmMUIlo Omit PU19*iM»nO
Uodritd Ann- SE PM lUdenft * nflcnl
fmlpohl
ynjittil IrxMtnci
(R>n«m mill
MM
211
Omite BrencMtH
tnt
Sttntuti. IIM^tkft T»S«(pooW.MPyitiJ
Haw
lit
2S7
«]
IB
M
78
J1
tl
It
1?
Oeolaiy « MU 1M8
Ufftr Knt/ndaet SyiWam
ShortnMjrftfMW
AjtVniAbctl
14H
1IU4
72S7
IW77
14319
10
w
to
M
WonlmOiyt
URAO
MO
CMn, 1
MMU
HI 743»
-iB
\Mbftv -RiTlihnbl
PtcMsrt in) Ilihi. ltr*l
drot «4«K>cpfe ninilon
«nM •eanrit tntoDcA
duct Konontg nlatlan
J115
PM luMolil [intkrfM too^OTi moruoy)
PM BiiHrtri ftntliHlH f Ml» »hcfM«m mmlilUvl
52201
-------
2010 Niltail PoM Edtntii Btntfll . Avridtd Indtoc* wd Ucntdod to
PU HMOS (1&SC. bcranw>M lo Cunrt PH10 lUmUrt)
tnfeoM
ffflllMn 1MOM
1MI>
128
12031
IOCM
11M8
HMO MM>iloRt-MR*if. 1M5
1M
4i4
1«4
IH
122
139
AEUiCnncMli
Pop* Mil, 1WI
O*o* ti, 1IH
lJ.. »H
5f
Wort L«i« D*r*
UFtAO
MO
M
IS)
tt42
HwwftoW Soloj Oung*
Ctwbll, I'dST
PtcMMnd Mitn. 1M4
PM (uMotit (IncludM tonftiaa wxt«*y)
PM M4«(ri (IndudM PM2.I «hod-««Tn nodtlM
o (cm meofiliid b»
Q
I
-------
J010 NiUnl PoM 6 ithuli tm«i • AnUtd MdwM Md limited B«nrt»
PU fWAQS HMO, hcnnwlil 10 Cui»rt FHID it«n*rt)
Mr OinPJty Semite PvtM MUhnw*, M Vur
UtxMtd AIM- PHI PU Rigtan {».»fT««v)
EndpoH
)»d hcldmei
MenMny B«fuin
211
11081
>m
SI
n
n
ii
it
11
Hew AiMtilau .
il ig«i)
Pnlwnoflli ( <*»)
Klip Adinhtbm . lldxrric Htllt
, 1BW
KM
1M
70
M
92
sa
AoJi aitnchtU
Icnrtr RiK*iloty Symptom
Upp*f RnpWwy Simpla™
^iAmJ /Utlcil
UM
17W
67M
to
M
W
M
_JO
U
fit
Tfi
MO
Went Loti Diyi
URAD
MO
69)09
4U7»
I9145J
PWlhrt >nd Ulhl 1 W*
**ct tccnoirig nLatlon
PM liiHotn (InekidM ions-Urm nortiKy)
PMtoUoti* (loduJ«tnM.i«ftcrt-ttnii moiliny)
11721
i Ho Irtu moottlrtd txmftt
-------
HJION«ion»IPeW EKtouit e«Mfl>- Anktei Jnddmu ud Mmttail Bidlntli]*nc(
(n>Honi
M*
UoUXy (feng-dnn up . iftl M*]
Popi * il, I«M
M
1140
CHOCK bnnatk
Ml«ilU.1M»
Knp AdVHtloni • Al R««f*»I«y(»l igt«)
COPO (•»" l»i)
Haip. AdnitiJan • CongM** Hun Ftkn
Haw
MwvtttndMorib. IN*
MMratniUaHi. 1
bodni/MiL,1M4
i:
II
II
SO
so
_S1
v>
LMrtf
UppK Rnpfrilory Spnptom
(Mottil, IIH
Mo4((l (Mf
1&3
n*3
TV)
71)
2311
10
M
WciklouOiyi
UMD
MO
(Mn nd RodwMd, 1M«
103310
63W
HouMtaM Soing Owug*
VWbBy-
drtrf Monomle viJmttor
mioUon
H* lutatold fkickujrt PM7 i lfx>rt-(«nii mort«»ty)
-------
2010 HubniJ Pot* Eitbul* Bimft*. Awtfej taddtno* ind Mcratod B4o.ni
P14 KMQ5 (IS«0. hffWTwOltO Cunn PU10 umtini)
(VtklMilnm^ Fitf Yur
••I'MEPMR.JanflB Jf7«onl
Endi»M
UotuiYf«hortJ«nn«i3).F'U3S
iurt
1ST7
I3SMI
»757«
flfcarandJJ. .
I MkTMltCM iaiidn Hurt run
Sdmuti, (IHOlt, MM^aeM«nUy>Ji)
Wnwifc, flHUk MM lpiMMinuyiJ)>
1481
21 U
7M
M7
110
tin
}13
to
Up9if flisjrintny Symptom
SIM4»uofi>*i0i
AitAnv /UQcti
10
JO
VVeiKtoiiOiyt
C
MO
tu
mi
jo
wo
J707
Dunig*
VhfcBy- Rtcnilferol
PJ^ifcrJ nd Uthi. 1»«4
) nioltn
*ict'i«norrto nknlfan
g nhmta
PM lUXotjl (hdtid t« fcKMH«m rtwUOy)
luUri PM 1 thoiUtfm nwrt>ot¥)
PM fuMotri nncli
IHIWS
-------
Abt Associates Inc.
Hampden Square-Suite 600 • 4800 Montgomery Lane • Bethesda, MD 20814-5341 • (301) 913-0500
MEMORANDUM
TO: Michele McKeever, EPA/OAQPS
FROM: John Voyzey, Ellen Post, and Erica Shingara, Abt Associates Inc.
DATE. July 14, 1997
SUBJECT: EPA Benefit Analyses for the Revised RIA for Ozone and Paniculate Matter (PM)
National Ambient Air Quality Standards (NAAQS): Estimating Changes in Health
and Welfare Incidence
This memorandum describes the method for estimating changes in the incidence of health
and welfare endpoints related to ozone, PM, and Regional Haze (RH) associated with attainment
of National Ambient Air Quality Standards (NAAQS) for ozone and/or PM and visibility targets
for RH.1 Since the pollutants contributing to RH are identical to those contributing to PM
formation, all concentration-response functions listed for PM are also applied in the RH benefits
analysis. While the methods discussed below apply to the estimation of both ozone-related and
PM-related changes in health and welfare endpoints, the discussion focuses on PM-related
changes for expository simplicity. "Ozone" can be substituted for 'TM," however, throughout
Similarly, while there are several health and welfare endpoints that have been associated with
ozone and PM, the discussion below refers only to a generic "health endpoint," denoted as y.
Finally, the discussion refers to estimation of changes in the incidence of the health endpoint in a
single county. National changes are estimated by summing the estimated changes in each county.
1.0 Estimating Incidence Change Based on a Single Concentration-Response Function
Attainment of a PM standard in a county implies a possible reduction in the concentrations
of PM in that county. (There may be zero reduction if the PM concentrations in the county
already satisfy the standard and if those concentrations are not further reduced by attainment in
adjacent counties.) Corresponding to a change in PM air quality, APM, there is a change in the
'The discussion here does not depend on standards being fully attained To a given level of attainment (either
full or partial) there corresponds a given change in ozone and/or PM concentrations. Associated with that change in
concentrations there are changes in the incidence of health effects. For simplicity, the discussion here refers to
attainment of standards It is to be understood that this may mean either full or partial attainment
Abt Associates Inc. I July 14. 1997
-------
health endpoint, Ay. Given a concentration-response function estimated by a study, and a
particular change in PM, APM, the corresponding change in the health endpoint, Ay,
corresponding to the particular APM can be calculated.
Different epidemiological studies may have estimated the relationship between PM and a
particular health endpoint in different locations. The concentration-response functions estimated
by these different studies may differ from each other in several ways. They may have different
functional forms; they may have measured PM concentrations in different ways; they may have
characterized the health endpoint, y, in slightly different ways; or they may have considered
different types of populations. For example, some studies of the relationship between ambient
PM concentrations and mortality have excluded accidental deaths from their mortality counts;
others have included all deaths. One study may have measured daily (24-hour) average PM
concentrations while another study may have used two-day averages. Some studies have assumed
that the relationship between y and PM is best described by a linear form (i.e., the relationship
between y and PM is estimated by a linear regression in which y is the dependent variable and PM
is one of several independent variables). Other studies have assumed that the relationship is best
described by a log-linear form (i.e., the relationship between the natural logarithm of y and PM is
estimated by a linear regression).2 Finally, one study may have considered changes in the health
endpoint only among members of a particular subgroup of the population (e.g , individuals 65 and
older), while other studies may have considered the entire population in the study location.
The estimated relationship between PM and a health endpoint in a study location is
specific to the type of population studied, the measure of PM used, and the characterization of the
health endpoint considered. For example, a study may have estimated the relationship between
daily average PM concentrations and daily hospital admissions for "respiratory illness," among
individuals age 65 and older, where "respiratory illness" includes International Classification of
Disease (ICD) codes A, B, and C. If any of the inputs had been different (for example, if the
entire population had been considered, or if "respiratory illness" had consisted of a different set of
ICD codes), the estimated concentration-response function would have been different. When
using a concentration-response function estimated in an epidemiological study to estimate changes
in the incidence of a health endpoint corresponding to a particular change in PM in a county, then,
it is important that the inputs be appropriate for the concentration-response function being used —
i.e., that the measure of PM, the type of population, and the characterization of the health
endpoint be the same as (or as close as possible to) those used in the study that estimated the
concentration-response function.
Estimating the relationship between PM and a health endpoint, y, consists of (1) choosing
a functional form of the relationship and (2) estimating the values of the parameters in the
function assumed. The two most common functional forms in the epidemiologjcal literature on
JThe log-linear form used in the epidemiological literature on ozone- and PM-related health effects is often
referred to as "Poisson regression" because the error term in the regression is assumed to have a Poisson distribution
rather than the usual normal distribution The form of the regression, however, is still log-linear
Abl Associates Inc. 2 Jufyl4,1997
-------
PM (and ozone) and health effects are the log-linear and the linear relationship. The log-linear
relationship is of the form
, (1)
or, equivalently,
= a + $PM , (2)
where the parameter B is the incidence of y when the concentration of PM is zero, the parameter
P is the coefficient of PM, ln(y) is the natural logarithm of y, and a = ln(B). If the functional form
of the concentration-response relationship is log-linear, the relationship between APM and Ay is
A, = y(e*'™ - 1] , (3)
where y is the baseline incidence of the health effect (i.e., the incidence before the change in PM)
For a log-linear concentration-response function, the relative risk (RR) associated with the change
APM is
(4)
Epidemiologjcal studies often report a relative risk for a given APM, rather than the coefficient, P
in the concentration-response function. The coefficient can be derived from the reported relative
risk and APM, however, by solving for P in equation (4):
P = \n(RK)/&PM (5)
The linear relationship is of the form
y = a * P/>A/ , (6)
where a incorporates all the other independent variables in the regression (evaluated at their mean
values, for example) times their respective coefficients. When the concentration-response
function is linear, the relationship between a relative risk and the coefficient, P, is not quite as
straightforward as it is when the function is log-linear. Studies using linear functions usually
report the coefficient directly.
If the functional form of the concentration-response relationship is linear, the relationship
between APM and Ay is simply
A.y = p*APA/ (7)
Abt Associates Inc. 3 July 14. 1997
-------
For a given PM change in a county, APM (measured appropriately), and a value for the
PM coefficient, P, the corresponding change in the health endpoint, Ay, is estimated - using
equation (3), if the concentration-response function is log-linear, or equation (7), if the
concentration-response function is linear. (If the concentration-response function is log-linear, the
baseline incidence is also necessary, as can be seen in equation (3))
A few epidemiologjcal studies, estimating the relationship between certain morbidity or
welfare endpoints and PM or ozone, have used functional forms other than linear or log-linear
forms. Of these, logistic regressions were the most common. The details of the models used in
these studies are given in the papers reporting the methods and results of the studies.
When only a single study has estimated the concentration-response relationship between a
pollutant and a given health endpoint, the estimation of a county-specific incidence change, Ay, is
straightforward, as noted above. The input components necessary to estimate incidence changes
using concentration-response functions from individual studies are shown in Exhibits 1, 2, and 3
for PM10-, PM2.5-, and ozone-related endpoints, respectively. In each exhibit, both the inputs
used in the study and the inputs used to estimate county-specific incidence changes in the benefit
analysis are shown.
2.0 Estimating Incidence Change Based on Multiple Concentration-Response
Functions: Pooling Study Results
When several studies have estimated concentration-response relationships between a
pollutant and a given health endpoint, the results of the studies were pooled to derive a single
estimate. If the input components discussed above (e.g., functional forms, pollutant averaging
times, study populations) are all the same (or very similar), a pooled, "central tendency"
concentration-response function can be derived from multiple study-specific concentration-
response functions. This is the case for all pollutant-endpoint combinations except ozone-
mortality, ozone-hospital admissions for COPD, and ozone-hospital admissions for pneumonia
Because some of these ozone studies measured daily 1-hour maximum ozone concentrations
while others measured daily (or some other) average ozone concentrations, it was not possible to
pool the concentration-response functions from same-endpoint studies to derive a central
tendency "pooled" concentration-response function for that endpoint. Instead, using the ozone
data appropriate to each study, national incidence distributions were derived corresponding to the
concentration-response function from each study, and these national incidence distributions were
pooled. That is, the pooling of results was done in "national incidence space" rather than in
"ozone coefficient space." The method is described in detail in a document discussing the
synthesis of results from multiple studies on ozone and mortality ("Assessment and Synthesis of
Abt Associates Inc. 4 Juty 14. 1997
-------
Exhibit 1; Summary of Inputs to Estimating PM10-Related Changes in Health and Welfare Effects
Emlpoint
Mortality
Post-neonatal infant
mortality (long-term
exposure)
Mortality (short-term
exposure)
Concentration-Response Function
Source
Woodruff etal, 1997
Ito A Thurston, 1996 (Chicago)
Kinney et al., 1995 (Los Angeles)
Pope etal., 1992 (Utah)
ocnwanz, 1 77ja (uirmmgnarn;
Schwartz ct al., 1996 (Boston)
Schwartz et al., 1 996 (Knoxville)
Schwartz et al., 1996 (Si. Louis)
Schwartz et al., 1 996 (Steubenville)
Schwartz et al., 1 996 (Portage)
Schwartz et al , 1996 (Topeka)
Functional
Form
PM-10 Averaging Time
Studied
Applied
log-linear
log-linear
log-linear
log-linear
log-linear
log-linear
log-linear
log-linear
log-linear
log-linear
log-linear
annual mean
2-day average
1 -day average
5-day average
3-day average
2-day average
2-day average
2-day average
2-day average
2-day average
2-day average
annual mean4
1 -day average
Population*
Annual Baseline
Incidence (per
100.000 of indicated
population)* •
Pollutant
Coefficient*
age 28-364
days
all
all
all
all
all
all
all
all
all
all
338
(infants < 1 year)
803
(nonaccidental
deaths in general
pop)
0.003922
0.000782
Hospital Admissions
All respiratory illnesses
Schwartz, 1995 (Tacoma)
Schwartz, 1995 (New Haven)
Schwartz, 1996 (Spokane)
log-linear
log-linear
log-linear
1 -day average
1 -day average
1 -day average
1 -day average
age 65+
age 65+
age 65+
504
(general pop.)
000170
Abi Associates Inc.
July 14. 1997
-------
Endpoint
COPD
Pneumonia
Congestive heart failure
Ischemia heart disease
Concentration-Rciponie Function
Source
Schwartz, 1994a (Birmingham)
Schwartz, 1994b (Detroit)
Schwartz, 1996 (Spokane)
Schwartz, 1994a (Birmingham)
Schwartz, 1994b (Detroit)
Schwartz, 1994c (Minneapolis)
Schwartz, 1996 (Spokane)
Schwartz and Morris, 1 995 (Detroit)
Schwartz and Moms. 1 99S (Detroit)
Functional
Form
log-linear
log-linear
log-linear
log-linear
log-linear
log-linear
log-linear
log-linear
log-linear
PM-10 Averaging Time
Studied
1 -day average
1 -day average
1 -day average
1 -day average
1 -day average
1 -day average
1 -day average
2-day average
1 -day average
Applied
1 -day average
1 -day average
1 -day average
1 -day average
Population*
age 65+
age 65+
age 65+
age 65+
age 65+
age 65+
age 65+
age 65+
age 65+
Annual Baseline
Incidence (per
100.000 of indicated
position)"
103
(general pop )
229
(general pop.)
231
(general pop )
450
(general pop.)
Pollutant
Coefficient'
0002533
0.0013345
000098
0.00056
Respiratory Symptoms/Dlnesiei not requiring hospitalization
Development of chronic
bronchitis
Acute respiratory
symptoms (any of 19)
Lower respiratory
symptoms (LRS)
Upper respiratory
symptoms (URS)
Schwartz, 1993b
Krupnick et al., 1990
Schwartz etal, 1994
Pope el al., 1991
logistic
logistic
log-linear
annual mean
1 -day average
COH
1 -day average
1 -day average
annual mean
1 -day average
I -day average
1 -day average
all
ages 18-65
(study
examined
"adults")
ages
8-12
asthmatics,
ages 9- II
n/a
n/a
n/a
38,187
(applied pop)
0.012
0.00046
00142
00036
J
Abt Associates Inc.
July 14. 1997
-------
Endpolnl
Shortness of breath (days)
Concent ration-Response Function
Source
Ostroetal, 1995
Functional
Form
logistic
PM-10 Averaging Time
Studied
1 -day average
Applied
J -day average
Population*
African-
American
asthmatics,
ages 7-12
Annual Baseline
Incidence (per
100.000 of indicated
BOBIllliilMri
r"i* "•"*•*••/
n/a
Pollutant
Coefficient c
0.00841
Welfare Endpointi
Consumer cleaning cost
savings
ESEERCO, 1994
linear
annual mean
annual mean
all
households
n/a
2.52
(dollars per
M8/m'l>MIO
per household)
NOTES:
1 The population examined in the study and to which this analysis applies the reported concentration-response relationship. In general, cpidemiological studies analyzed the
concentration-response relationship for a specific age group (e.g, ages 65+) in a specific geographical area This analysis applies the reported pollutant coefficient to all individuals in
the age group nationwide.
k annual baseline incidence in the applied population per 100,000 individuals in the indicated population. The rates given for mortality are national averages and were not actually used
in the analyses. County-specific mortality rates were used, as described in Section 3.0.
' a single pollutant coefficient reported for several studies indicates a pooled analysis; see text for discussion of pooling concentration-response relationships across studies
4 The following studies report a lowest observed pollution level:
Oslro et al., 1995 Shortness of Breath, days 19.63 ng/m' PMIO
Woodruff et al., 1997 Mortality (postneonatal infants) 11.9 fig/m1 PM10
Since these studies did not examine the concentration-response relationship for concentrations below the reported levels, this analysis does not estimate benefits for ambient
concentration reductions below these concentrations. The remaining studies did not report lowest observed concentrations
' All 1 -day averages are 24-hour averages. 2-day averages are 48-hour averages, etc
Abi Associates Inc.
July 14. 1997
-------
Exhibit 2: Summary of Inputs to Estimating PM2.5-Related Chan
Endpoint
Concentration-Response Function
Source
Mortality (Short-Term Exposure)
Schwartz el al, 1996
(Boston, Knoxville, St. Louis,
Steubenville, Portage & Topeka)
Functional
Form
»es in Health and Welfare Effects
PM2.5 Averaging Time
Studied
Applied
•
log-linear
2-day average
1 -day average
Population*
Annual Baseline
Incidence (per
100,000 of indkiled
population) b
Pollutant
Coefficient'
all
Mortality (Long-Tcrm Exposure)
Pope ctal.. 1995
log-linear
annual median
annual
median4
ages 30+
803
(nonaceidenuJ deaths
in genera) pop.)
0.001433
759
(nonaccidental deaths
in general pop.)
0.006408
Hoipital Admissions
All respiratory illnesses
Thurston et al., 1994 (Toronto)
linear
1 -day average
1 -day average
all
Respiratory Symptoms/Dhiesses not requiring hospitalization
Acute bronchitis
Lower respiratory
symptoms (LRS)
Asthma (moderate or
worse)
MRADs
RADs
Work loss days CWLDs)
Dockery ctal., 1989
Schwartz etal, 1994
Ostroetal, 1991
Ostro and Rothschild, 1 989
Ostro, 1987
Ostro, 1987
logistic
logistic
linear (with
log
pollutant)
log-linear
log-linear
log-linear
annual mean
1 -day average
daily 8-hour
average (9.00 am-
4-00 pm)
2-week average
2-week average
2-week average
annual mean4
1 -day average
1 -day average
1 -day average
1 -day average
1 -day average
ages 10-12
ages 8- 12
asthmatics,
ages 9- 11
ages 18-65
ages 18-6S
ages 18-65
n/a
3.45 XIO* f
n/a
n/a
n/a
780,000 days/year
(applied pop.)
400.531 days/year
(applied pop.)
150,750
days/year (applied
pop.)
00298
0.01823
0.0006
0.00741
0.00475
0.0046
J
NOlliS
Abl Associates Inc.
8
July 14. 1997
-------
' The population examined in the study and to which this analysis applies the reported concentration-response relationship. In general, epidemiological studies analyzed the
concentration-response relationship for a specific age group (e.g., ages 65+) in a specific geographical area. This analysis applies the reported pollutant coefficient to all individuals in
the age group nationwide.
* annual baseline incidence in the applied population per 100,000 individuals in the indicated population The rates given for mortality are national averages and were not actually used
in the analyses. County-specific mortality rates were used, as described in Section 3 0.
' a single pollutant coefficient reported for several studies indicates a pooled analysis; see text for discussion of pooling concentration-response relationships across studies.
' The following studies report a lowest observed pollution level.
Pope ct al., 1995 ' Mortality (long-term exposure) 9 ug/m* PM,,
Dockery et al., 1989 Acute Bronchitis 11.8 ug/m1 PM,, (20 1 ug/m' PM10)
Since these studies did not examine the concentration-response relationship for concentrations below the reported levels, this analysis does not estimate benefits for ambient
concentration reductions below these concentrations. The remaining studies did not report lowest observed concentrations.
' All 1-day averages are 24-hour averages, 2-day averages are 48-hour averages, etc.
r units on linear pollutant coefficient: hospital admissions per Mg/m' PM2.S per exposed individual
Abt Associate Inc. 9 July 14. 1997
-------
Exhibit 3; Summary of Inputs to Estimating Ozone-Related Changes in Health and Welfare Effects
Endpoint
Concentration-Response Function
Source
Functional
Form
Ozone Averaging Time
Studied
Applied
Population*
Annual Baseline
Incidence (per
100,000 ofindictted
poptililion) •
Pollutant
Coefficient1
Mortality (Short-Term Exposure)
Anderson el al.( 1996
Hoeketal., 1997 (in press)
Ito&Thurston, 1996
Kinneyelal.,1995
Loomisetal., 1996(1 1121)
Moolgavkarctal, 1995
Ostro etal, 1996
Samel etal, 1996, 1997 (HEI)
Verhoeff etal., 1996
log-linear
log-linear
log-linear
log-linear
log-linear
log-linear
log-linear
log-linear
log-linear
1 -day average *
1 -day average
1 -day average
daily 1 -hour max
daily 1 -hour max
1 -day average
daily 1 -hour max
1 -day average
daily 1 -hour max
1 -day average
1 -day average
1 -day average
daily 1-hour
max
daily 1-hour
max
1 -day average
daily 1-hour
max
1 -day average
doily 1-hour
max
all
all
all
all
all
all
all
all
all
803
(noruccidental
deaths in general
pop)
0.001126
0.001705
0.000677
000
0000182
0000611
0.00019
0000936
0000956
Hospital Admissions
All respiratory Illnesses
All respiratory Illnesses
All respiratory Illnesses
All respiratory Illnesses
Schwartz, 1996 (Spokane)
Schwartz, 1995 (New Haven)
Schwartz. 1995 (Tacoma)
Thurston et al , 1994 (Toronto)
log-linear
log-linear
log-linear
linear
daily 1 -hour max
1 -day average
1 -day average
daily 1 -hour max
doily 1-hour
max
1 -day average
1 -day average
daily 1 -hour
max.
age 65+
age 65+
age 65+
all
504
(general pop.)
504
(general pop.)
504
(general pop)
n/a
0.008562
00014
00036
j
1.62X10-' •
Abt Associates Inc.
July 14. 1997
-------
Endpnint
All respiratory Illnesses
COPD
COPD
COPD
Pneumonia
Pneumonia
Pneumonia
Pneumonia
Concentration-Response Function
Source
Thurston et al , 1992 (New York City)
Schwartz, 19 94 a
Schwartz, 1994b
Schwartz, 1996 (Spokane)
Schwartz, 1994 a
Schwartz, 1994b
Schwartz, 1994c
Schwartz, 1996 (Spokane)
Functional
Form
linear
log-linear
log-linear
log-linear
log-linear
log-linear
log-linear
log-linear
Ozone Averaging Time
Studied
daily 1 -hour max
1 -day average
1 -day average
daily 1 -hour max.
1 -day average
1 -day average
1 -day average
daily 1 -hour max
Applied
daily 1-hour
max
1 -day average
1 -day average
daily 1-hour
max
1 -day average
1 -day average
1 -day average
daily 1 -hour
max.
Population*
all
age 65+
age 65+
age 65+
age 65+
age 65+
age 65+
age 65+
Annual Baseline
Incidence (per
100.000 of indicated
poputaion)*
n/a
103
(general pop.)
103
(general pop.)
103
(general pop.)
229
(general pop.)
229
(general pop.)
229
(general pop.)
229
(general pop.)
Pollutant
Coefficient1
1.37X10* •
0.00314
0.00549
0.004619
0.00262
0.00521
0.002795
000965
Respiratory Symptoms not Requiring Hospitalization
Acute respiratory
symptoms
(any of 19)
Asthma attacks
MUADs
Krupnick et al , 1990
Whittemore and Kom, 1 980 and US
EPA, 1993
Ostro and Rothschild, 1 989
logistic
logistic
log-linear
daily 1 -hour max
daily 1 -hour max..
daily 1 -hr max
(avg over 2
weeks)
daily 1-hour
max
daily 1 -hour
max..
daily 1-hr
max (avg.
over 2 weeks)
ages
18-65
asthmatics
ages 18-65
n/a
n/a
780,000 days/year
(applied pop)
0.00014
00019
00022
Abl Associates Inc.
H
July 14. 1997
-------
Endpoint
RRADs
Concentration-Response Function
Source
Ostro and Rothschild, 1 989
Functional
Form
log-linear
Ozone Averaging Time
Studied
daily 1-hrmax.
(avg. over 2
weeks)
Applied
daily 1-hr
max (avg.
over 2 weeks)
Population*
ages 18-65
Annual Baseline
Incidence (per
100,000 of indicated
population)*
3 10,000 days/year
(applied pop.)
Pollutant
Coefficient"
0.0054
Welfare Endpolnti
Decreased worker
productivity
Crocker and Horst. 1 98 1 and US
EPA. 1994
percent
change
1 -day average
1 -day average
laborers
n/a
n/a
NOTES.
* The population examined in the study and to which this analysis applies the reported concentration-response relationship. In general, epidcmiological studies analyzed the
concentration-response relationship for a specific age group (e.g., ages 65+) in a specific geographical area This analysis applies the reported pollutant coefficient to all individuals in
(he age group nationwide.
' annual baseline incidence in the applied population per 100,000 individuals in the indicated population The rates given Tor mortality are national averages and were not actually used
in the analyses. County-specific mortality rates were used, as described in Section 3.0.
' a single pollutant coefficient reported for several studies indicates a pooled analysts; see text for discussion of pooling concentration-response relationships across studies.
* All I -day averages ore 24-hour averages, 2-day averages are 48-hour averages, etc.
• units on linear pollutant coefficient hospital admissions per ppb 0, per exposed individual
Abt Associates Inc.
12
July 14. 1997
-------
Available Epidemiological Evidence of Mortality Associated with Ambient Ozone from Daily
Time-Series Analyses")- The method of pooled analysis used for PM10 and PM2.S
concentration-response functions is described briefly below and in more detail in the attached
Appendix.
One method of pooled analysis is simply averaging the PM coefficients from all the
studies. This has the advantage of simplicity, but me disadvantage of not taking into account
the measured uncertainty of each of the estimates. Fytimarg!t with great uncertainty
surrounding them are given the same weight as estimates with very little uncertainty.
It seems reasonable that a "pooled estimate" which combines the estimates from
different studies should give more weight to estimates from studies with little reported
uncertainty than to estimates with a great deal of uncertainty. The exact way in which weights
are assigned to estimates of PM coefficients from different studies in a pooled analysis depends
on the underlying assumption about how the different estimates are related to each other.
Under the assumption that there is a distribution of PM coefficients, or p's (referred to
as the random effects model), the different coefficients reported by different studies may be
estimates of different underlying PM coefficients, rather than just different estimates of the
same PM coefficient. The random-effects model is preferred here to the fixed effects model
(which assumes that there is only one P everywhere), because it does not assume that all
studies are estimating the same parameter.3
Pooled analyses were carried out for the ten short-term exposure mortality studies, the
three "all respiratory illness" hospital admissions studies, the three COPD hospital admissions
studies, and the four pneumonia hospital admissions studies listed in Exhibit 1 (PM10); the six
short-term exposure mortality studies listed in Exhibit 2 (PM2.5); and the nine short-term
exposure mortality studies listed in Exhibit 3 (ozone)
Because only daily (24-hour) average PM concentrations were available, these were
used when applying short-term PM exposure pooled concentration-response functions in the
benefit analysts, even when one or more of the functions used in the pooling were based on
other averaging times. (The pooled analysis of PMlO-mortality, for example, includes studies
that used the average PM10 concentration on a single day as the pollution indicator as well as
studies that used the average PM10 concentration over a 2-, 3- or 5-day period. The studies
included in the pooled analysis of PM2.5-mortality all used 2-day PM2.5 averages.) Those
3 In studies of the effects of PM-10 on mortality, for example, if the composition of PM-10 varies among study
locations the underlying relationship between mortality and PM-10 may be different from one study location to another.
For example, fine particles make up a greater fraction of PM-10 in Philadelphia County than in Southeast Los Angeles
County If fine particles are disproportionately responsible for mortality relative to coarse particles, then one would
expect the true value of P for PM-10 in Philadelphia County to be greater than the true value of P for PM-10 m
Southeast Los Angeles County. This would violate the assumption of the fixed effects model
Abt Associates Inc 13 July 14. 1997
-------
studies which use multi-day averages are in effect using a smoothed data set, comparing each
day's mortality to recent average exposure rather than simply exposure on the same day.
The more nearly linear the concentration-response function, however, the less difference it
makes whether multi-day averaging functions are used with single-day PM data. (If the
functions were perfectly linear, it would make no difference at all.) The concentration-
response functions considered here are nearly linear. Because all of the studies in the pooled
analysis focus on the results of short-term exposure to pollution (as opposed to long-term
exposure, measured in years), it is appropriate to consider these studies together.
3.0 Baseline Incidences Used with Log-Linear Concentration-Response Functions
County-specific mortality rates were obtained for each county in the United States from
the National Center for Health Statistics (NCHS) Because most PM and ozone studies that
estimated concentration-response functions for mortality considered only non-accidental
mortality, county-specific baseline mortality rates used in the estimation of PM- and ozone-related
mortality were adjusted to reflect a better estimate of county-specific non-accidental mortality.
Each county-specific mortality rate was multiplied by the ratio of national non-accidental
mortality to national total mortality (0.93).
Although total mortality incidences (over all ages) were available for counties, age-specific
mortality incidences were not available at the county level. County-specific baseline mortality
incidences among individuals aged 30 and over (necessary for PM2.5-related long-term exposure
mortality, estimated by Pope et al., 1995) were therefore estimated by applying national age-
specific death rates to county-specific age distributions, and adjusting the resulting estimated age-
specific incidences so that the estimated total incidences (including all ages) equaled the actual
county-specific total incidences. For example, if the total of the estimated age-specific incidences
obtained in this way was 5% higher than the actual total incidence for a county, then each of the
estimated age-specific incidences was multiplied by (1/1.05)
Each county-specific hospital admissions baseline incidence rate was obtained by
multiplying the national hospital admissions rate for the relevant International Classification of
Diseases (ICD) code(s) per 100,000 exposed population by the county-specific population, and
then adjusting this incidence by the ratio of the county-specific proportion of the population aged
65 or older to the national proportion of the population aged 65 or older (Except for Thurston et
al., 1994, all hospital admissions studies used in the national benefit analysis apply only to
individuals 65 and older. The Thurston study used a linear concentration-response function,
which, unlike an exponential concentration-response function, does not require a baseline
incidence rate for calculation of PM-related incidence.)
Baseline incidence rates for all respiratory symptoms and illnesses included in the benefit
analysis and for restricted activity days were obtained from the studies reporting concentration-
response functions for those health endpoints No baseline incidence rates were available from
Abt Associates Inc. 14 July 14. 1997
-------
other sources for these endpoints. Finally, the consumer cleaning cost savings function is-a linear
function and therefore does not require a baseline incidence rate.
Abt Associates he 15 July 14. 1997
-------
4.0 References
Anderson et al. 1996. Air Pollution and Daily Mortality in London: 1987 - 92. BMJ 312: 665-
669.
Crocker T.D., Horst R.L., Jr. 1981. Hours of Work, Labor Productivity, and Environmental
Conditions: a Case Study. The Review of Economics and Statistics 63:361-368.
Dockery, D.W., RE. Speizer, D.O. Strain, J.H. Ware, J.D. Spengler, and B.G. Ferris, Jr. 1989.
Effects of Inhalable Particles on Respiratory Health of Children. Am. Rev. Respir. Dis.
139: 587-594.
Empire State Electric Energy Research Corporation (ESEERCO) 1994 New York State
Environmental Externalities Cost Study. Report 2: Methodology. Prepared by.
RCG/Hagler, Bailly, Inc., November.
Hoeketal. 1997. Effects of Ambient Paniculate Matter and Ozone on Daily Mortality in
Rotterdam, the Netherlands. Draft, submitted to Archives of Environmental Health
Ito, K. and Thurston, G.D. 1996. Daily PMlO/Mortality Associations: An Investigation of At-
Risk Subpopulations. Journal of Exposure Analysis and Environmental Epidemiology
6(1): 79-225.
Kinneyetal. 1995. A Sensitivity Analysis of Mortality/PM 10 Associations in Los Angeles
Inhalation Toxicology 7: 59-69.
Krupnick A.J., Harrington W., Ostro B. 1990. Ambient Ozone and Acute Health Effects*
Evidence from Daily Data Journal of Environmental Economics and Management 181-
18.
Loomis et al. 1996. Ozone Exposure and Daily Mortality in Mexico City: A Time-Series
Analysis. Health Effects Institute Research Report Number 75, October 1996
Moolgavkar et al. 1995. Air Pollution and Daily Mortality in Philadelphia. Epidemiology
6(5):476-484.
Ostro, B.D. 1987. Air Pollution and Morbidity Revisited: a Specification Test. J. Environ.
Econ. Manage. 14: 87-98.
Ostro B.D. and S. Rothschild. 1989. Air Pollution and Acute Respiratory Morbidity: An
Observational Study of Multiple Pollutants. Environmental Research 50:238-247.
Ostro, B.D., M.J. Lipsett, M.B. Wiener, and J C. Seiner. 1991. Asthmatic Responses to
Airborne Acid Aerosols. American Journal of Public Health 81 694-702
Abt Associates Inc. 16 July 14.1997
-------
Ostro, B.D., M.J. Lipsett, J.K. Mann, H. Braxton-Owens, and M.C. White. 1995. Air Pollution
and Asthma Exacerbations Among African American Children in Los Angeles. Inhalation
Toxicology.
Ostro et al. 1996. Air Pollution and Mortality: Results from a Study of Santiago, Chile. J. Of
Exposure Analysis and Environmental Epidemiology 6:97-114.
Pope, C.A., m, D.W. Dockery, J.D. Spengler, and ME. Raizenne. 1991. Respiratory Health
and PM10 Pollution: a Daily Time Series Analysis. Am. Rev. Respir. Dis. 144: 668-674.
Pope, C.A., m, Schwartz, J., and M.R. Ransom. 1992. Daily mortality and PM10 in Utah
valley. Arch. Environ. Health 47: 211-217.
Pope, C.A., III, M.J. Thun, M M Namboodiri, D.W. Dockery, J.S. Evans, F E. Speizer, and
C.W. Heath, Jr. 1995. Paniculate Air Pollution as a Predictor of Mortality in a
Prospective Study of U.S. Adults. Am. J. Respir. Cril. CareMed. 151 669-674.
Sametetal. 1996. Air Pollution and Mortality in Philadelphia, 1974-1988. Report to the
Health Effects Institute on Phase IB: Particle Epidemiology Evaluation Project, March 25,
1996 (draft, accepted for publication).
Samet et al. 1997. Paniculate Air Pollution and Daily Mortality: Analysis of the Effects of
Weather and Multiple Air Pollutants. The Phase IB Report of the Particle Epidemiology
Evaluation Project. Health Effects Institute, March 1997.
Schwartz, J. 1993a. Air pollution and daily mortality in Birmingham, Alabama. Am. J.
Epidemiol. 137: 1136-1147.
Schwartz,!. 1993b. Paniculate Air Pollution and Chronic Respiratory Disease. Environmental
Research 62: 7-13.
Schwartz, J. 1994a. Air Pollution and Hospital Admissions in Elderly Patients in Birmingham,
Alabama American Journal of Epidemiology 139:589-98
Schwartz, J. 1994b. Air Pollution and Hospital Admissions for the Elderly in Detroit, Michigan.
American Journal of Respiratory Care Med 150:648-55.
Schwartz,!. 1994c. PM10, Ozone and Hospital Admissions for the Elderly in MinneapoUs-St.
Paul, Minnesota. Archives of Environmental Health 49(5): 366-374.
Schwartz, J. 1995. Short Term Fluctuations in Air Pollution and Hospital Admissions of the
Elderly for Respiratory Disease. Thorax 50:531-538
Abt Associates Inc. 17 July J4, 1997
-------
Schwartz, J. 1996. Air Pollution and Hospital Admissions for Respiratory Disease
Epidemiology 1(\Y 1-9.
Schwartz, J. and R. Morris. 1995 Air Pollution and Cardiovascular Hospital Admissions. Am.
J. Epidemiol. 142:23-35.
Schwartz, J. D. W. Dockery, L.M Neas, D. Wypij, J.H. Ware, J.D. Spengler, P. Koutrakis,
F.E.Speizer, and E.G. Ferris, Jr. 1994. Acute Effects of Summer Air Pollution on
Respiratory Symptom Reporting in Children. Am.J. Respir. Crit. CareMed 150:1234-
1242.
Schwartz, J., Dockery, D., and L. Neas. 1996. Is Daily Mortality Specifically Associated With
Fine Particles? J. Air & Waste Man. Assoc. 46 927-939.
Thurston, G.D. K. Ito, P.L. Kinneym, and M. Lippman. 1992 A Multi-Year Study of Air
Pollution and Respiratory Hospital Admissions in Three New York State Metropolitan
Areas: Results for 1988 and 1989 Summers. Journal of Exposure Analysis and
Environmental Epidemiology. 2 (4):429-450
Thurston, G. K. Ito, C. Hayes, D. Bates, and M Lippmann. 1994. Respiratory Hospital
Admission and Summertime Haze Air Pollution in Toronto, Ontario: Consideration of the
Role of Acid Aerosols. Environmental Research 65: 271-290.
U.S. Environmental Protection Agency (U.S. EPA). 1993. External Draft, Air Quality Criteria
for Ozone and Related Photochemical Oxidants. Volume II. Office of Health and
Environmental Assessment, Environmental Criteria and Assessment Office, Research
Triangle Park, NC; EPA/600/AP-93/004b. 3v.
U S Environmental Protection Agency (U.S. EPA) 1994 Documentation for Oz-One
Computer Model (Version 2.0). Office of Air Quality Planning and Standards. Prepared
by: Mathtech, Inc., under EPA Contract No. 68D30030, WA 1-29. August.
Verhoeff, A.P. etal. 1996. Air Pollution and Daily Mortality in Amsterdam. Epidemiology
7(3). 225-230.
Whittemore AS, Korn EL. 1980. Asthma and Air Pollution in the Los Angeles Area. American
Journal of Public Health 70:687-696.
Woodruff, T.J., J.M. Grille, and K.C.Schoendorf. 1997. The Relationship Between Selected
Causes of Postneonatal Infant Mortality and Paniculate Air Pollution in the United States.
Environmental Health Perspectives. June (Forthcoming.)
Abt Associates Inc. IS Juty 14, 1997
-------
Appendix: Pooling the Results of Different Studies
Many studies have attempted to determine the influence of ozone pollution on human
health. Usually this involves estimation of a parameter p in a concentration-response function,
which may be linear or nonlinear, as discussed above. Each study provides an estimate of P, along
with a measure of the uncertainty of the estimate. Because uncertainty decreases as sample size
increases, combining data sets is expected to yield more reliable estimates of p. Combining data
from several comparable studies in order to analyze them together is often referred to as meta-
analysis.
For a number of reasons, including data confidentiality, it is often impractical or
impossible to combine the original data sets Combining the results of studies in order to produce
better estimates of P provides a second-best but still valuable way to synthesize information
(DerSimonian and Laird, 1986). This is referred to as "pooling results" in this report. Pooling
requires that all of the studies contributing estimates of p use the same functional form for the
concentration-response function. That is, the P's must be measuring the same thing.
One method of pooling study results is simply averaging all reported P's. This has the
advantage of simplicity, but the disadvantage of not taking into account the uncertainty of each of
the estimates. Estimates with great uncertainty are given the same weight as estimates with very
little uncertainty. For example, consider the three studies whose results are presented in Table Al.
Table Al. Three Sample Studies.
Study
Study 1
Study 2
Study 3
Estimate of P
0.75
1.25
1.00
Standard
Deviation
0.35
0.05
0.10
Variance
0.1225
0.0025
0.0100
The average of the three estimates is 1 0 However, the Study 2 estimate has much less
uncertainty associated with it (variance = 0.0025) than either the Study 1 or Study 3 estimates. It
seems reasonable that a pooled estimate that combines the estimates from all three studies should
therefore give more weight to the estimate from the second study than to the estimates from the
first and third studies. A common method for weighting estimates involves using their variances
Variance takes into account both the consistency of data and the sample size used to obtain the
estimate, two key factors that influence the reliability of results.
Abt Associates Inc.
July 14. 1997
-------
The exact way in which variances are used to weight the estimates from different~studies
in a pooled estimate depends on the underlying model assumed. The next section discusses the
two basic models that might underlie a pooling and the weighting scheme derived from each
Al The fixed effects model
The fixed effects model assumes that there is a single true concentration-response
relationship and therefore a single true value for the parameter p. Differences among p's reported
by different studies are therefore simply the result of sampling error. That is, each reported p is an
estimate of the same underlying parameter. The certainty of an estimate is reflected in its variance
(the larger the variance, the less certain the estimate). Pooling that assumes a fixed effects model
therefore weights each estimate under consideration in proportion to the inverse of its variance.
Suppose there are n studies, with the ith study providing an estimate P, with variance vs
denote the sum of the inverse variances. Then the weight, w,, given to the ith estimate, P,, is
1/v
This means that estimates with small variances (i.e., estimates with relatively little uncertainty
surrounding them) receive large weights, and those with large variances receive small weights.
The estimate produced by pooling based on a fixed effects model is just a weighted
average of the estimates from the studies being considered, with the weights as defined above
That is,
PA - 5>, • P,
The variance associated with this pooled estimate is the inverse of the sum of the inverse
variances:
1
S 1/v,
Abt Associates Inc. 20 July 14. 1997
-------
Table A2 shows the relevant calculations for this pooling for the three sample studies summarized
in Table Al.
Table A2. Fixed Effect Model Calculations.
Study
1
2
3
Sum
3,
0.75
1.25
1.00
Vi
0.1225
0.0025
0.0100
1/v,
8.16
400
100
£ = 508.16
Wj
0.016
0.787
0.197
1=1.000
Wi'0,
0.012
0.984
0.197
1=1.193
The sum of weighted contributions in the last column is the pooled estimate of P based on the
fixed effects model. This estimate (1.193) is considerably closer to the estimate from Study 2
(1.25) than is the estimate (1.0) that simply averages the study estimates. This reflects the fact that
the estimate from Study 2 has a much smaller variance than the estimates from the other two
studies and is therefore more heavily weighted in the pooling
The variance of the pooled estimate, vft, is the inverse of the sum of the inverse variances,
or 0.00197. (The sums of the Pj and Vjare not shown, since they are of no importance. The sum of
the Ify is S, used to calculate the weights. The sum of the weights, wp I = 1,.... n, is 1.0, as
expected.)
A2 The random effects model
An alternative to the fixed effects model is the random effects model, which allows the
possibility that the estimates P( from the different studies may in fact be estimates of different
parameters, rather than just different estimates of a single underlying parameter. In studies of the
effects of ozone on mortality, for example, if the behavior or susceptibility of populations varies
among study locations, the underlying relationship between mortality and ambient ozone
concentrations may be different from one study location to another. (Suppose, for example,
people in one location spend substantially more time outdoors than people in another location;
this would violate the assumption of the fixed effects model)
The following procedure can test whether it is appropriate to base the pooling on the
random effects model (versus the fixed effects model):
A test statistic, Q«, the weighted sum of squared differences of the separate study estimates from
the pooled estimate based on the fixed effects model, is calculated as:
1 „>
Abl Astociates Inc.
21
July 14. 1997
-------
Under the null hypothesis that there is a single underlying parameter, f), of which all the fts are
estimates, Qw has a chi-squared distribution with n-1 degrees of freedom. (Recall that n is the
number of studies in the meta-analysis.) If Qw is greater than the critical value corresponding to
the desired confidence level, the null hypothesis is rejected. That is, in this case the evidence does
not support the fixed effects model, and the random effects model is assumed, allowing the
possibility that each study is estimating a different p.
The weights used in a pooling based on the random effects model must take into account
not only the whhin-study variances (used in a meta-analysis based on the fixed effects model) but
the between-study variance as well. These weights are calculated as follows:
Using Q*. the between-study variance, T|2, is:
It can be shown that the denominator is always positive. Therefore, if the numerator is negative
(i.e., if Qw < n-1), then r\2 is a negative number, and it is not possible to calculate a random effects
estimate. In this case, however, the small value of Qw would presumably have led to accepting the
null hypothesis described above, and the meta-analysis would be based on the fixed effects model.
The remaining discussion therefore assumes that r|2 is positive.
Given a value for r\\ the random effects estimate is calculated in almost the same way as
the fixed effects estimate. However, the weights now incorporate both the within-study variance
(vj) and the between-study variance ( t|2). Whereas the weights implied by the fixed effects model
used only vit the within-study variance, the weights implied by the random effects model use v,
Let v{* = vj -Hi2. Then
v,
and
S'
Abt Associates Inc. 22 July 14, 1997
-------
The estimate produced by pooling based on the random effects model, then, is just a weighted
average of the estimates from the studies being considered, with the weights as defined above.
That is,
• E
P,
The variance associated with this random effects pooled estimate is, as it was for the fixed effects
pooled estimate, the inverse of the sum of the inverse variances:
1
rand
E 1/v,*
The weighting scheme used in a pooling based on the random effects model is basically the
same as that used if a fixed effects model is assumed, but the variances used in the calculations are
different. This is because a fixed effects model assumes that the variability among the estimates
from different studies is due only to sampling error (i.e., each study is thought of as representing
just another sample from the same underlying population), whereas the random effects model
assumes that there is not only sampling error associated with each study, but that there is also
between-study variability — each study is estimating a different underlying p. Therefore, the sum
of the within-study variance and the between-study variance yields an overall variance estimate
A3 An example
This section demonstrates the relevant calculations for pooling using the example in Table
Al above.
First calculate Q*. as shown in Table A3
Table A3: Calculation of Qw
Study
1
2
3
R
0.75
1.25
1.00
l/vs
8.16
400
100
1/v. * (R — pfc) 2
1601
1.300
3.725
£ = Q. = 6.626
In this example the test statistic Qw = 6.626. The example considers three studies, so Qw is
distributed as a chi-square on two degrees of freedom. The critical value for the 5% level (i.e.,
corresponding to a 95% level of confidence) for a chi-square random variable on 2 degrees of
Abl Associates Inc.
23
July 14. 1997
-------
freedom is 5.99. Because Qw = 6.626 > 5.99, hence the null hypothesis is rejected. That is-, the
evidence does not support the fixed effects model. Therefore assume the random effects model is
appropriate.
Then calculate the between-study variance:
, _ 6.626 - (3 - 1)
50fl H - 170066'6S
508.16
0.0267
From this and the within-study variances, calculate the pooled estimate based on the random
effects model, as shown in Table A4
Table A4. Random Effects Model Calculations.
Study
1
2
3
Sum
3,
075
1.25
1.00
v; + Ti2
0.1492
0.0292
0.0367
l/Cv.+n2)
6.70
34.25
27.25
1 = 6820
w;*
0098
0.502
0.400
1=1.000
w,» x 3,
00735
0.6275
0.400
1=1 101
The random effects pooled estimate, P,,,^ is 1.101. It's variance, v^,, is l/(68.2) = 0015.
Abi Associates Inc.
24
July K 1997
-------
Abt Associates Inc.
Hampden Square-Suite 600 • 4800 Montgomery Lane • Bethesda, MD 20814-5341 • (301) 913-0500
MEMORANDUM
TO: Michele McKeever, U.S. EPA/OAQPS
FROM: Ellen Post, John Voyzey, and Leland Deck, Abt Associates Inc.
DATE: July 16,1997
SUBJECT: Economic Valuations and Aggregation of Benefits of Health and Welfare Effects
in EPA Benefit Analyses for the Revised RIA for Ozone and Paniculate
Matter National Ambient Air Quality Standards (NAAQS)
This memorandum describes the methods used to estimate the monetized benefits of the
individual health and welfare endpoints associated with ozone and paniculate matter (PM), and
the methods used to aggregate these monetized benefits to estimate total monetized benefits, for
the revised Regulatory Impact Analysis (RIA) for Ozone, PM, and Regional Haze (referred to in
this memorandum as the revised RIA for ozone and PM NAAQS).
To provide an overview of how the health and welfare endpoints and their economic
valuations described in this memorandum compare to those used in similar analyses conducted by
EPA, Section 1 compares the resulting economic valuations to one other analysis conducted by
EPA and to the previous draft RIAs for the proposed ozone and PM National Ambient Air
Quality Standards (NAAQS). The three analyses with which comparisons are made are:
(1) §812 Retrospective Analysis of the Clean Air Act (U.S. EPA, 1997),
(2) Draft RIA for the Proposed PM NAAQS (U.S. EPA, 1996a), and
(3) Draft RIA for the Proposed Ozone NAAQS (U.S. EPA, 1996b).
The economic values presented in Section 1 are point estimates. There is uncertainty
surrounding any estimate of the monetized benefit associated with a unit change in a health or
welfare effect (e.g., an additional hospital admission avoided). Point estimates are often a central
tendency estimate taken from a distribution of possible values. Section 2 describes the derivations
of the distributions and point estimates of monetized values (unit dollar values) for those health
and welfare endpoints considered in this analysis.
Section 3 addresses the issues that arise when point estimates of dollar benefits from
different sources are aggregated into a point estimate of total dollar benefits. Finally, Section 4
discusses further aggregation issues that arise when the distribution of total dollar benefits is
Abt Associates Inc. I Juty 16.1997
-------
estimated, based on the distributions of the benefits predicted to result from separate, non-
overlapping sources.
1.0 A COMPARISON OF ECONOMIC VALUATIONS OF HEALTH AND
WELFARE ENDPOINTS IN SEVERAL EPA ANALYSES
Exhibit 1 shows the economic valuations used for each of the PM and ozone health and
welfare endpoints considered in the revised RIA for ozone and PM NAAQS as well as the three
analyses listed above. Each value presented in Exhibit 1 is the point estimate of the monetary
benefit associated with avoiding a unit of the given effect, and is therefore known as a unit dollar
value.
The analyses considered in Exhibit 1 address the uncertainty associated with the unit dollar
values in different ways. For example, the §812 Retrospective Analysis does not actually use the
unit dollar values reported in Exhibit 1. Instead, for each endpoint a Monte Carlo simulation
draws from a distribution of dollar values (the mean of which is the unit dollar value shown here).
In contrast, the 1996 Draft RIA for the proposed ozone NAAQS presents "Low", "Best", and
"High" estimates of unit dollar values, of which the "Best" estimates are the point estimates
shown in Exhibit 1. The 1996 Draft RIA for the proposed PM NAAQS presents only a point
estimate of the unit dollar value for each endpoint.
With the exception of the unit dollar value for a statistical life saved, the unit dollar values
shown for the revised RIA for ozone and PM NAAQS are rounded to three significant digits
(e.g., to the nearest $100 for hospital admissions, to the nearest SI for most respiratory
symptoms, and to the nearest SO.l for shortness of breath). The dollar values presented for the
three EPA analyses to which the current analyses are compared in Exhibit 1 are shown as they
were presented in each of the three reports; in some cases, numbers differ slightly due to rounding
conventions used in the different presentations.
Abt Associates Inc. 2 July J6,]997
-------
Exhibit 1. Economic Valuations for Health and Welfare Effects of Ozone and PM—Comparison of Three Benefit Analyses
Health or Welfare
Effect
Mortality
Statistical Lives
Saved
Life- Years Saved
Hospital Admissions
All Respiratory
Illnesses, all ages
PoDutant(s)'
Valuation
Measure'
Unit Values Reported in each Benefit Analysb
§812
Retrospective
1996 Draft
RIA-PM
NAAQS
1996 Draft
RJA-O,
NAAQS
Revised RIA
for O, and
PM
NAAQS'
PMn/PMu
o,
PM^/PMu
PMu/PMjj
o,
S per case
$ per case
$ per life-year
S per hospital
admission
$ per hospital
admission
$4.8 million
-.
$293,000
$6,100
$6,100
•
$4.8 million
__
—
$12,700
$4.8 million
...
—
$11,972
$4.8 million
$4.8 million
$293,000
$12,700*
$13,400*
Comments
The unit dollar value in the revised RIA for
ozone and PM NAAQS is different for PM
V£r4ttt ft hM^MiMi nnnArfimifv fv\*t t«
included for O, but not for PM, to avoid
double counting for PM (see Section
2.1.1), and because the study estimating a
concentration-response function for PM
defines "all respiratory illnesses" slightly
differently from the corresponding ozone
study. The unit dollar value for ozone
differs from that in the 1996 draft ozone
RIA because the adjustment of COI differs
(see comment below).
Abt Associates Inc.
July 16, 1997
-------
Health or Welfare
Effect
Pneumonia, age 2 65
COPD, age 2 65
Ischemic Heart
Disease, age 2 65
Congestive Heart
Failure, age 2 65
Emergency
Department visits for
Asthma
Pollutant(s)«
-------
Health or Welfare
Effect
Upper Resp.
Symptoms (URS)
Lower Resp.
Symptoms (LRS)
Acute Bronchitis
Acute KGspinuory
Symptoms-am; of 19
Asthma*
Shortness of breath
Sinusitis and hay
fever
Development of
Asthma
Restricted Activity
Work Loss Day
(WLD)
PolIuUnt(i)'
PMM
PMw/PMjj
PM18/PMu
(VPMU
Oj/PMjj
PM,.
o,
o,
PMw
Valuation
Measure*
$per
symptom-day
$per
^^•B 1 ii ftmmm
sympiom-oay
$ per case
Sper
symptom-day
Sper
symptom-day
Sper
symptom-day
...
Unit Values Reported In each Benefit Analysis
§812
Retrospective
$19
$12
$45
$18
$32
SS.30
quantified but
not monetized
...
1996 Draft
RIA-PM
NAAQS
$18.70
$11.82
$45
$18.31
$32.48
$5.29
—
1996 Draft
RIA-0,
NAAQS
—
—
...
$29.33
$32.48
...
quantified but
not monetized
quantified but
not monetized
Reviled RIA
for O, and
PM
NAAQSC
$19'
$12'
$45'
$18'
$32'
$5.30>
quantified but
not monetized
...
$ per day
$83
$83
...
$83
Comments
differences due to rounding.
The $18.31 vame (rounded to $18)
updates the $29.33 value. The new value
was derived from WTP estimates for
specific lower and upper respiratory
symptoms.
difference due to rounding
difference due to rounding
Abt Associates Inc.
July 16. 1997
-------
Health or Welfare
Effect
RAD
MRAD
RRAD
Welfare Effects
MTt^tfmf TV«»litMti»ittr
WOTKer rTOuUCUVliy
(resulting in enrages
in daily wages)
Visibility-
Residential
Visibility -
Recreational
PoDutant(f)'
PMu
(VPMjj
OJPMu
o,
deciview*
visual range*
Valuation
Measure*
$ per day
$ per day
—
change in daily
wages
Annual
household
WTP
Annual
household
WTP
Unit Values Reported In each Benefit Analysis
§812
Retrospective
quantified but
not monetized
$38
quantified but
not monetized
$1 per worker
per 10%
change in 0,
WTP per unit
decrease in
deciview =
S14
1996 Draft
RIA-PM
NAAQS
-
$38.37
—
—
WTP per unit
change in
visual range:
East -$149
West = $117
1996 Draft
RIA-O,
NAAQS
—
—
quantified but
not monetized
$1 per worker
per 10%
change in O/
Revised RIA
for O, and
PM
NAAQS'
...
S381
quantified but
not monetized
SI per worker
per 10%
change in 0)
WTP per unit
decrease in
deciview =
S14
see discussion
Comments
difference due to rounding
The WTP estimate reported in the §812
analysis updates the estimate used in the
PM NAAQS RIA, The new estimate is
given in terms of deciview and applies
nationwide (the WTP in the east is no
longer distinguished from that in the west).
The new nationwide WTP is very similar
to the old estimate for the eastern U.S.
(note that visibility units differ).
WTP for visibility improvements in Class I
areas are reported for three regions of the
U.S Separate WTP values are reported
for in-region and out-of region individuals
and for the "indicator" park in a region and
all other national parks in the region.
Abt Associates Inc.
Jufy 16,1997
-------
Health or Welfare
Effect
Consumer Cleaning
Cost Savings
PoDutant(s)'
TSP
Valuation
Measure*
$per
household per
(annual cost)
Unit Values Reported in each Benefit Analysis
§812
Retrospective
$2.50
1996 Draft
RIA-PM
NAAQS
$2.52
1996 Draft
RIA-0,
NAAQS
—
Revised RIA
for O, and
PM
NAAQS'
$250
Comments
difference due to rounding
NOTES:
' Pollutant(s) fcr which epidemiotogical evidence quantifies a concentration-response relationship for the given endpoint.
b most unit values quantify the willingness to pay (WTP) to avoid a case of the given effect However, for those effects measured in terms of symptom-days, the unit value
reflects the WTP to avoid one day of the given respiratory symptoms.
'Attainment of an ozone standard can result in anc^axy reductions mPM, by virtae of the control strategies used to achieve Similarly, attainment of a PM
standard could result in ancillary reductions in ozone. Because of a lack of air quality information on ancillary ozone reductions that would be likely to result from attainment of
a PM standard, the benefits of such ancillary ozone reductions cannot be estimated. The revised RIA for ozone and PM NAAQS does, however, estimate the benefits resulting
from ancillary reductions in PM associated with attainment of ozone standards. Health and/or welfare endpoints associated only with PM are therefore also relevant to the
analysis of the benefits of attaining an ozone standard.
* Incidence Hospital admissions for pneumonia and COPD not included in the RIA for PM NAAQS to avoid double-counting with the "all respiratory" admissions estimate.
The other two benefit analyses chose to avoid double-counting differently, see discussion below.
Asthma is cither self-reported asthma or moderate or worse asthma status
' Although worker productivity valuation is not reported as a unit value in the O3 NAAQS RIA, the implied value is as shown here
' deciview (dv) is a common visibility measure useful for characterizing visibility in terms of perceptible changes independent of baseline conditions. A decrease in deciview
corresponds to an increase in visibility. It is related to another common visibility measure, visual range (VR): dv - 10 ln[391 km/VR] where dv is unitless and VR is measured
in kilometers.
h rounded to the nearest $100.
1 Rounded to the nearest $1.
J Rounded to the nearest $0.1.
Abt Associates Inc.
Jufy 16. 1997
-------
2.0 DERIVATION OF DISTRIBUTIONS AND POINT ESTIMATES OF UNIT
DOLLAR VALUES
Willingness to pay (WTP) is the measure used for the value an individual places on
something, whether it is something that can be purchased in a market or not. WTP is the amount
of money such that the individual would be indifferent between having the good (or service) and
having the money.
For both market and non-market goods, WTP reflects individuals' preferences. Because
preferences are likely to vary from one individual to another, WTP for both market and non-
market goods (e.g., health-related improvements in environmental quality) is likely to vary from
one individual to another. In contrast to market goods, however, non-market goods such as
environmental quality improvements are public goods, whose benefits are shared by many
individuals. The individuals who benefit from the environmental quality improvement may have
different WTPs for this non-market good. The total social value of the good is the sum of the
WTPs of all individuals who "consume" (i.e., benefit from) the good.
In the case of health improvements related to pollution reduction, it is not certain
specifically who will receive particular benefits of reduced pollution. For example, the analysis
may predict 100 hospital admissions for respiratory illnesses avoided by attainment of a given
ozone NAAQS, but the analysis does not estimate which individuals will be spared those cases of
respiratory illness that would have required hospitalization. The health benefits conferred on
individuals by a reduction in pollution concentrations are, then, actually reductions in the
probabilities of having to endure certain health problems. These benefits (reductions in
probabilities) may not be the same for all individuals (and could be zero for some individuals).
Likewise, the WTP for a given benefit is likely to vary from one individual to another. In theory,
the total social value associated with the decrease in risk of a given health problem resulting from
a given reduction in pollution concentrations is
(1)
where Bi is the benefit (i.e., the reduction in probability of having to endure the health problem)
conferred on the ith individual (out of a total of N) by the reduction in pollution concentrations,
and WTPj(Bi) is the ith individual's WTP for that benefit. If a reduction in pollution
concentrations affects the risks of several health endpoints, the total health-related social value of
the reduction in pollution concentrations is
EE'^W (2)
M y-i
Abt Associates Inc. 8 July 16.1997
-------
where B0 is the benefit related to the jth health endpoint (i.e., the reduction in probability of
having to endure the jth health problem) conferred on the ith individual by the reduction in
pollution concentrations, and WTPfB£ is the ith individual's WTP for that benefit.
The reduction in probability of each health problem for each individual is not known, nor
is each individual's WTP for each possible benefit he or she might receive known. Therefore, in
practice, benefits analysis estimates the value of a statistical health problem avoided. For
example, although a reduction in pollutant concentrations may save actual lives (i.e., avoid
premature mortality), whose lives will be saved cannot be known ex ante. What is known is that
the reduction in air pollutant concentrations results in a reduction in mortality risk. It is this
reduction in mortality risk that is valued in a monetized benefit analysis. Individual WTPs for
small reductions in mortality risk are summed over enough individuals to infer the value of a
statistical life saved. This is different from the value of a particular, identified life saved. Rather
than "WTP to avoid a death," then, it is more accurate to use the term "WTP to avoid a
statistical death," or, equh/alentty, "the value of a statistical life."
Suppose, for example, that a given reduction in PM concentrations results in a decrease in
mortality risk of 1/10,000. Then for every 10,000 individuals, one individual would be expected
to die in the absence of the reduction in PM concentrations (who would not die in the presence of
the reduction in PM concentrations). If WTP for this 1/10,000 decrease in mortality risk is $500
(assuming, for now, that all individuals' WTPs are the same), then the value of a statistical life is
10,000 x SSOO, or $5 million.
A given reduction in PM concentrations is unlikely, however, to confer the same risk
reduction (e.g., mortality risk reduction) on all exposed individuals in the population. (In terms of
the expressions above, B, is not necessarily equal to Bj, for I *j). In addition, different individuals
may not be willing to pay the same amount for the same risk reduction. The above expression for
the total social value associated with the decrease in risk of a given health problem resulting from
a given reduction in pollution concentrations may be rewritten to more accurately convey this.
Using mortality risk as an example, for a given unit risk reduction (e.g., 1/1,000,000), the total
mortality-related benefit of a given pollution reduction can be written as
N
of units of risk reduction); * (WTP per unit risk reduction^
where (number of units of risk reduction^ is the number of units of risk reduction conferred on
the ith exposed individual as a result of the pollution reduction, (WTP per unit risk reduction), is
the hh individual's willingness to pay for a unit risk reduction, and N is the number of exposed
individuals.
If different subgroups of the population have substantially different WTPs for a unit risk
reduction and substantially different numbers of units of risk reduction conferred on them, then
estimating the total social benefit by multiplying the population mean WTP to save a statistical life
Abl Associates Inc. 9 July 16. 1997
-------
times the predicted number of statistical lives saved could yield a biased result. Suppose, for
example, that older individuals' WTP per unit risk reduction is less than that of younger
individuals (e.g., because they have fewer years of expected life to lose). Then the total benefit
will be less than it would be if everyone's WTP were the same In addition, if each older
individual has a larger number of units of risk reduction conferred on him (because a given
pollution reduction results in a greater absolute reduction in risk for older individuals than for
younger individuals), this, in combination with smaller WTPs of older individuals, would further
reduce the total benefit.
While the estimation of WTP for a market good (Le., the estimation of a demand
schedule) is not a simple matter, the estimation of WTP for a non-market good, such as a
decrease in the risk of having a particular health problem, is substantially more difficult.
Estimation of WTP for decreases in very specific health risks (e.g., WTP to decrease the risk of a
day of coughing or WTP to decrease the risk of admission to the hospital for respiratory illness) is
further limited by a paucity of information. Derivation of the dollar value estimates discussed
below was often limited by available information.
2.1 Premature Mortality
2.1.1 Statistical lives lost
The dollar value of avoiding one statistical death was estimated to be $4.8 million. This is
the mean of the estimates from 26 value-of-life studies identified by Industrial Economics, Inc.
(lEc, 1992) as "applicable to policy analysis." EC'S assessment mirrors that of Viscusi (1992)
and uses the same criteria used by Viscusi in his review of value-of-life studies. The $4.8 million
estimate is consistent with Viscusi's conclusion that "most of the reasonable estimates of the value
of life are clustered in the $3 to $7 million range."
Five of the 26 studies are contingent valuation (CV) studies, which directly solicit WTP
information from subjects; the rest are wage-risk studies, which base WTP estimates on estimates
of the additional compensation demanded in the labor market for riskier jobs. The 26 studies are
listed in Exhibit 2. The references for all but Gegax et al. (198S) and V.K. Smith (1983) may be
found in Viscusi (1992).
Although each of the studies estimated the average WTP for a given reduction in mortality
risk, the amounts of reduction in risk being valued were not necessarily the same across studies,
nor were they necessarily the same as the amounts of reduction in mortality risk that would
actually be conferred by a given reduction in PM concentrations. The transferabilhy of estimates
of the value of a statistical life from the 26 studies to the PM benefit analysis rests on the
assumption that, within a reasonable range, WTP for reductions in mortality risk is linear in
Abt Associates Inc. 10 July 16. 1997
-------
Exhibit 2. Summary of Mortality Valuation Eitimatei
(millkuM of 1990 dollat»)
Stedy
Valuation
(•fiHaaa
19901)
Kncaner and LaA (1991) (US)
Labor Market
0.6
Smith and Gilbert (1984)
Labor Market
07
09
Butkr(19S3)
1.1
Miller and Ourw (1991)
CooL Value
12
Mom aod VaHHi (19Wa)
Labor VOrkrt
. Magat, and Huber (1991b) I Cod. Value
Z7
GtgaxeiaL(19S5)
COOL Value
3.3
Maria and FMdtarapoukB (19«)
Labor Market
28
I Kneuper and Ledh (1991) (Aia«ralia) Labor Market
Abt Associates Inc.
n
July 16. 1997
-------
risk reduction. For example, suppose a study estimates that the avenge WTP for a reduction in
mortality risk of 1/100,000 is $30. Suppose, however, that the actual mortality risk reduction
resulting from a given PM reduction is 1/10,000. If WTP for reductions in mortality risk is linear
in risk reduction, then a WTP of $30 lor a reduction of 1/100,000 implies a WTP of $300 for a
risk reduction of 1/10,000 (which is ten times the risk reduction valued in the study). Under the
assumption of linearity, the estimate of the value of a statistical life does not depend on the
particular amount of risk reduction being valued.
Although the particular amount of mortality risk reduction being valued in a study may not
affect the transferabflity of the WTP estimate from the study to the PM benefit analysis, the
characteristics of the study subjects and the nature of the mortality risk being valued in the study
could be important. Certain characteristics of both the population affected and the mortality risk
facing that population are believed to affect the average WTP to reduce the risk. The
appropriateness of the mean of the WTP estimates from the 26 studies for valuing the mortality-
related benefits of reductions in PM concentrations therefore depends not only on the quality of
the studies (i.e., how well they measure what they are trying to measure), but also on (1) the
extent to which the subjects in the studies are similar to the population affected by changes in PM
concentrations and (2) the extent to which the risks being valued are similar.
Focusing on the wage-risk studies, which make up the substantial majority of the 26
studies relied upon, the likely differences between (1) the subjects in these studies and the
population affected by changes in PM concentrations and (2) the nature of the mortality risks
being valued in these studies and the nature of PM-related mortality risk are considered. The
direction of bias in which each difference is fikdy to result is also considered.
Compared with the subjects in wage-risk studies, the population believed to be most
affected by PM (Le., the population that would receive the greatest mortality risk reduction
associated with a given reduction in PM concentrations) is, on average, older and probably more
risk averse. Citing Schwartz and Dockery (1992) and Ostro et al (1996), Chestnut (1995)
estimates that approximately 85 percent of those who die prematurely from PM-related causes are
over 65. The average age of subjects in wage-risk studies, in contrast, would be well under 65.
There is also reason to believe that those over 65 are, in general, more risk averse than the
general population (see below) while workers in wage-risk studies are likely to be less risk averse
than the general population. Although Vacua's list of recommended studies excludes studies that
consider only much-higher-than-average occupational risks, there is nevertheless likely to be some
selection bias in the remaining studies—that is, these studies are tikety to be based on samples of
workers who are, on average, more risk-loving than the general population. In contrast, older
people as a group exhibit more risk averse behavior.
In addition, it might be argued that because the elderly have greater average wealth than
those younger, the affected population is also wealthier, on average, than wage-risk study
subjects, who tend to be blue collar workers. It is possible, however, that among the elderly h is
largely the poor elderly who are most vulnerable to PM-related mortality risk (e.g., because of
Abl Associates Inc 12 July 16. 1997
-------
generally poorer health care). If this is the case, the average wealth of those affected by a
reduction in PM concentrations relative to that of subjects in wage-risk studies is uncertain.
Finally, although there may be several ways in which job-related mortality risks differ from
PM-related mortality risks, the most important difference may be that job-related risks are
incurred voluntarily whereas PM-related risks are incurred involuntarily.
The direction of bias resulting from the age difference is unclear, particularly because age
is confounded by risk aversion (relative to the general population). It could be argued that,
because an older person has fewer expected years left to lose, his WTP to reduce mortality risk
would be less than that of a younger person. This hypothesis is supported by one empirical study,
Jones-Lee et aL (1985), that found the value of a statistical life at age 65 to be about 90 percent
of what it is at age 40. Citing the evidence provided by Jones-Lee et al. (1985), Chestnut (1995)
assumes that the value of a statistical life for those 65 and over is 75 percent of what it is for those
under 65.
The greater risk aversion of older people, however, implies just the opposite. Citing
Ehrlich and Chuma (1990), lEc (1992) notes that "older persons, who as a group tend to avoid
health risks associated with drinking, smoking, and reckless driving, reveal a greater demand for
reducing mortality risks and hence have a greater implicit value of a life year." That is, the more
risk averse behavior of older individuals suggests a greater WTP to reduce mortality risk.
There is substantial evidence that the income elasticity of WTP for health risk reductions is
positive (see, for example, Alberini et al., 1994; Mitchell and Carson, 1986; Loehman and Vo Hu
De, 1982; Gerkmg et aL, 1988; and Jones-Lee et aL, 1985), although there is uncertainty about
the exact value of tins elasticity). Individuals with higher incomes (or greater wealth) should,
then, be willing to pay more to reduce risk, all else equal, than individuals with lower incomes or
wealth. Whether the average income or level of wealth of the population affected by PM
reductions is likely to be significantly different from that of subjects in wage-risk studies,
however, is unclear, as discussed above.
Finally, there is some evidence (see, for example, Violette and Chestnut, 1983) that people
will pay more to reduce involuntarily incurred risks than risks incurred voluntarily. If this is the
case, WTP estimates based on wage-risk studies may be downward biased estimates of WTP to
reduce involuntarily incurred PM-related mortality risks.
The potential sources of bias in an estimate of WTP to reduce the risk of PM-related
mortality baaed on wage-risk studies are summarized in Exhibit 3 below.
Abt Associates Inc. 13 July 16. 1997
-------
Exhibit 3: Potential Sources of Bias in Estimates of WTP to Reduce the Risk of PM-
Rclated Mortality Based on Wage-Risk Studies
Factor
Age
Degree of Risk Aversion
Income
Risk Perception: Voluntary
vs. Involuntary risk
Likely Direction of Bias in MWTP Estimate
Upward (?)
Downward
Downward, if the elderiy affected are a random
sample of the elderly,
Unclear, if the elderly affected are the poor elderly.
Downward
The need to adjust wage-risk-based WTP estimates downward because of the likely
upward bias introduced by the age discrepancy has received significant attention (see Chestnut,
1995; lEc, 1992). In a similar vein, EPA's Science Advisory Board's Clean Air Act Compliance
Analysis Council highlighted the importance of life expectancy as an issue affecting mortality risk
valuation.1 If the age difference were the only difference between the population affected by PM
changes and the subjects in the wage-risk studies, there might be some justification for trying to
adjust the point estimate of $4.8 million downward. Even in this case, however, the degree-of the
adjustment would be unclear. There is good reason to suspect, however, that there are biases in
both directions, as shown in the table above. Because in each case the extent of the bias is
unknown, the overall direction of bias in the point estimate of $4.8 million is similarly unknown.
Adjusting the estimate upward or downward to compensate for any one source of bias could
therefore actually increase the degree of bias. The point estimate of $4.8 million was therefore
left unadjusted.
2.1.2 Life-years lost
Life-years lost is a possible alternative measure of the mortality-related effect of
pollution, as discussed in the memorandum on PM-related mortality. If life-years lost is the
measure used, then the value of a statistical life-year lost, rather than the value of a statistical life
lost would be needed. Moore and Viscusi (1988) suggest one approach for determining the value
of a statistical life-year lost They assume mat the wfflmgness to pay to save a statistical life is the
lCitod in EC (1993) as: Dr. Richard Schmalansee, Chair. Sckoce Advisory Board's Clean Air Act
Analysis Council, letter to EPA Administrator Carol M. Browna. Subject Scienoe Advisory Board's
review of (he Office of Policy. Flaming md Evaluation's (OPPE) and the Office of Air and Radiation's (OAR) progress
on the retrospective study of the impacts of the Clean Air Act, March 24.1993 (EPA-SAB-CAACAC-LTR-90-006)
Abl Associates Inc.
14
Jufy 16. 1997
-------
value of a single year of life times the expected number of yean of life remaining for an individual
They suggest that a typical respondent in a mortal risk study may have a life expectancy of an
additional 35 years. Using a mean estimate of $4.8 million to save a statistical life, their approach
would yield an estimate of $137,000 per life-year lost or saved. If an individual discounts future
additional years using a standard discounting procedure, the value of each life-year lost must be
greater than the value assuming no discounting. Using a 35 year life expectancy, a $4.8 million
value of a statistical life, and a 5 percent discount rate, the implied value of each life-year lost is
$293,000. The Moore and Viscusi procedure is identical to this approach, but uses a zero
discount rate.
2.2 Hospital Admissions
An individual's WTP to avoid a hospital admission will include, at a minimum, the
amount of money he pays for medical expenses (i.e., what he pays towards the hospital charge
and the associated physician charge) and the loss in earnings. In addition, however, an individual
is likely to be willing to pay some amount to avoid the pain and suffering associated with the
illness itself. That is, even if they incurred no medical expenses and no loss in earnings, most
individuals would still be willing to pay something to avoid the illness.
Because medical expenditures are to a significant extent shared by society, via medical
insurance, Medicare, etc., however, the medical expenditures actually incurred by the individual
are likely to be less than the total medical cost to society. The total value to society of an
individual's avoiding a hospital admission, then, might be thought of as having two components:
(1) the cost of illness (COT) to society, including the total medical costs plus the value of the lost
productivity, as well as (2) the individual's WTP to avoid the disutility of the illness itself (e.g.,
the pain and suffering associated with the illness).
It is useful to note the distinction which results from insurance between the total (social)
COI and the individual's COI. If the individual paid all the medical expenses associated with the
hospital admission, then the individual's WTP to avoid the hospital admission would, by
definition, be at least as great as, and probably greater than, the COL Because the individual
usually does not incur the total medical costs of a hospital admission, however, there is no
guarantee that the individual's WTP to avoid the hospital admission will be at least as great as the
total (social) COL The limited evidence comparing individual WTPs to social COI, however,
suggests that individual WTPs to avoid morbidity effects generally do in feet exceed the total
COIs associated with those effects (see ESEERCO, 1994).
In the absence of estimates of social WTP to avoid hospital admissions for specific
illnesses (components I plus 2 above), estimates of total COI (component 1) are typically used as
lower bound estimates Because these estimates do not include the value of avoiding the disutility
of the illness itself (component 2), they are biased downward. Some analyses adjust COI
estimates upward by multiplying by an estimate of the ratio of WTP to COI, to better approximate
total WTP. Other analyses have avoided making this adjustment because of the possibility of
overadjusting - that is, possibly replacing a known downward bias with an upward bias. The
Abl Associates Inc. 15 July 16. 1997
-------
previous RIAs for PM and ozone, as well as the revised RIA for ozone and PM NAAQS, do
adjust the COI estimate upward. This is the source of the major discrepancies in hospital
admission unit dollar values shown in Exhibit 1 for the §812 Retrospective Analysis versus those
shown for the PM and Ozone RIAs and the revised RIA for ozone and PM NAAQS.
2.2.1 Point estimates of COI
The COI estimates used in the revised RIA for ozone and PM NAAQS include the
estimated hospital and physician charges, based on the average length of a hospital stay for the
illness, and the estimated opportunity cost of time spent in the hospital.
Abt Associates Inc. (1992) estimated that physician charges for the first day of hospital
care for asthma Cm 1988) or COPD (in 1989) averaged S94 (in 1990 S); physician charges for
subsequent days of hospital care averaged $35. Average physician charges associated with
hospital care for asthma or COPD were assumed to provide reasonably good estimates of average
physician charges associated with hospital stays for the other illness categories considered here.
Estimated physician charges for a hospital stay of n days for any of the illness categories discussed
below, then, would be $94 + $35(n-l).
The opportunity cost of a day spent in the hospital can be estimated, for people in the
workforce, as the value of the lost daily wage. This is estimated at $83.00. If this value of a
work loss day (WLD) is elsewhere added into the total benefits analysis, however, including it as
a component of the WTP to avoid a hospital admission would be double counting. However, the
study on PM and work loss days (Ostro, 1987) considers only individuals 18 to 65 years old,
while all but one of die studies of PM and hospital admissions consider only individuals aged 65
and over. For all but one study of PM and hospital admissions (Thurston et aL, 1994, which
considered PM-2.S and hospital admissions for "respiratory illness"), then, the only possible
overlap is for individuals exactly 65 years old. This is not expected to be a significant amount of
overlap. One study of ozone and hospital admissions for "all respiratory illnesses" (Thurston et
al., 1992) also included individuals of all ages. Because the opportunity cost of time spent in the
hospital is not elsewhere included for ozone, as it is for PM, this component is included for all
individuals for this study.
To estimate the opportunity cost of a day spent in the hospital for an individual aged 65 or
older, rt is assumed that such an individual is not in the workforce. Although the value of a WLD
may be an inappropriate way to estimate the opportunity cost of a day spent in the hospital for
someone who is not employed (including the young and the elderly), this opportunity cost is
positive and should not be ignored. As a rough approximation, it was assumed that, for the
young, the elderly, and any other unemployed individuals the opportunity cost of a day spent in
the hospital is one-half what ft is for individuals in the workforce, or $41.50.
To derive estimates of the opportunity cost of a day spent in the hospital for respiratory
illness based on Thurston et al., 1994 or Thurston et aL, 1992, which considered individuals of all
ages, it is assumed that half of the PM- or ozone-related hospital admissions are among
Abl Associates Inc. 16 Jufy 16,1997
-------
individuals who are not employed, including the young and the elderly.2 The expected
opportunity cost of a day spent in the hospital for an individual randomly selected from among
those admitted to the hospital for PM- or ozone-related respiratory illnesses is therefore estimated
as (0.5)($83.00) -K0.5)($41.50) = $62.25.
However, because the value of work loss days for those in the labor force is added as a
separate component of the total benefit for PM, only the second component of opportunity cost
enters into the PM-related "all respiratory" hospital admissions benefit, which is, then,
(0.5)($41.50) = $20.75.
Finally, for all hospital admissions except for COPD, estimates of hospital charges were
based on discharge statistics provided by Elixhauser et al., 1993. For COPD, the estimate of
hospital charge and physician charge combined was based on data in Abt Associates, 1992.
2.2.2 Point estimate of total (social) WTP: Multiplying COI by the WTP/COI ratio
The revised RIA for ozone and PM NAAQS adjusts the total COI estimate for each
hospital admission endpoint by multiplying it by an estimate of the ratio of individual WTP to total
COI. It is argued below that, for two reasons, this adjustment is likely to be conservative for the
hospital admission endpoints considered and is unlikely to be an overadjustment.
The relationships between individual WTP and social (total) WTP may be clarified by the
following notation. Let
LP denote the value of lost productivity;
M denote the medical costs associated with the illness;
Mfafy denote the medical costs incurred by the individual (M.^, £ M);
COI denote the total cost of the illness;
CO!*,* denote the individual's COI;
D denote the value of avoiding the disutility of the illness;
WTPfaa, denote the individual's WTP to avoid the hospital admission; and
WTP denote the social WTP to avoid the hospital admission.
Then
COI=M+LP;
«^B^
);and
WTP-COI+D.
*This is approximately the same as the ratio of employed to total population in the United States lit 1994, for
example, this ratio was (123 mfllian)/(260 million), or 47 percent
Abt Associates Inc. 17 Julv 16.1997
-------
Finally, let
r = WTPindiv/COIand
a = COIMv/COI(0*o*l).
Then
_ ^
_ _ _ ^
COI COI COI
- D = (r - a)COI
The total (social) WTP to avoid a hospital admission can then be expressed as a function of r, a,
and COI:
WTP = COI + D = (1 + r - a)COI
Suppose, for example, that the avenge individual has a cost of illness (including both medical
expenses and lost earnings) that is half the total COL Suppose also that the ratio of individual
WTP to total COI is 2. The total (social) WTP to avoid the hospital admission would then be
WTP = (1 + 2 - 0.5)*COI = 2.5*COI.
Estimates of the ratio of individual WTP to total COI, r, are likely to vary substantially
from one health endpoint to another. Estimates of this ratio associated with asthma, for example,
ranged from 1.3 to 1.7 (Rowe and Chestnut, 1986; Rowe et al., 1984); an estimated ratio of 2.4
was reported for cataracts (Rowe and Neithercut, 1987).
The above repotted ratios provide a rough range of adjustment factors by which to
multiply estimates of COI to yield better approximations of the total social value of avoiding
hospital admissions. Based on the limited evidence cited above, multiplying COI estimates by 2 is
likely to improve the estimates of total WTP while minimizing the likelihood of overadjusting.
Two considerations suggest that 2 is likely to be a conservative adjust First, as the
disutility component of individual WTP, D, gets larger (with more serious illnesses), the ratio r is
likely to increase. A ratio associated whh asthma or cataracts may therefore be smaller than one
associated with a hospital admission for pneumonia or ischemic heart disease. Second, as can be
seen above, the total social value of avoiding a hospital admission (WTP) is a function not only of
COI and r, but of a as well. Multiplying COI by r to approximate WTP implicitly assumes that a
= 1 - that is, that the individual incurs the total medical costs To the extent that a < 1, however,
Abl Associates Inc. 18 July 16, 1997
-------
multiplying COI by r instead of by (1 + r - a) understates the total WTP. As a decreases, the
extent of the understatement increases.
In summary, WTP to avoid a hospital admission is estimated as
r+COI
= r+(hospital charge + physician charge + opportunity cost) ,
where
(1) the hospital charge is estimated by mean discharge costs provided by Elixhauser et al.,
1993, for all illnesses except COPD (for which hospital charge and physician charge
combined are estimated from information in Abt Associates, 1992);
(2) the physician charge is estimated as $94 + $3S(n-l), where n is the number of days
spent in the hospital;
(3) opportunity cost is the expected opportunity cost per day times n, and the expected
opportunity cost per day depends on the age composition of the hospitalized population
(e.g., all ages or only ages 65+) and whether opportunity cost for individuals in the
workforce is included elsewhere in the analysis, as described in Section 2.1.1; and
(4) the adjustment factor, r, is estimated to be 2.
The derivation of point estimates of total (social) WTP for all PM- and ozone-related
hospital admission endpoints is summarized in Exhibit 4.
Abl Associates Inc. 19 July 16. J 997
-------
Exhibit 4: Derivation of Cost of Dlncss (COD and Total WTP Estimates for Hospital Admissions Endpoints (1990$)
HfMpitftt Admissions Fort
Ischemic Heart Disease, ago
2 65 OCD codes 410-414)
Congestive Heart Future,
age 2 65 (ICD code 428)
COPD, age 2 65 (ICD codes
490-496)
Pneumonia, age 2 65 (ICD
codes 480-487)
illnesses." all ages (ICD
codes 466, 480-482, 485,
490-493)
Ozone-Related "all
respiratory illnesses," all
ages (ICD codes 466, 480-
486. 490-493)
Asthma (ICD code 493)*
"Resp Illness," age 2 65
(ICD codes 460-5 1 9)
Hospital
Charge
(1)
19.713
$7,532
$7,521
$7.181
$5,839
$5,973
$3,957
$5,682
Physician
Charge
(2)
$304
$374
-
$374
S339
$304
$234
$304
Opportunity
Cost per day
$41.50
$41.50
$41.50
$41.50
$20.75
$62.25
$62.25
$41.50
|vpui tuuii j ^
Arg Length
of Stay (days)
7
9
8
9
8
7
5
7
Cost
Total Opportunity
Cost
(3)
S291
$374
$332
$374
$166
$436
$311
$291
Total Cost of Illness (COT)
(l) + (2) + (J)
(4)
$10.308
$8,280
$7,853
$7,929
$6.344
$6.712
$4,502
$6,277
Estimate of Total
WTP: 2*(4)
$20,615
$16,559
$15.705
$15,857
$12.688
$13,425
$9,004
$12,553
* Emergency department visits for asthma are valued as hospital admissions for asthma, using the information in blixnauser ct ai., 1993 and the standard assumptions
used for all hospital admissions.
A hi Axioclates Inc.
20
Jufy 16, 1997
-------
2.2.3 Derivation of the distribution of WTP to avoid a hospital admission
There is substantial uncertainty associated with the adjustment factor, r. Acknowledging
that r is likely to vary from one endpoint to another and that the range of values cited above may
understate the true range of possible values appropriate to the particular endpoints considered
here, r is taken to have a continuous uniform distribution from 1.5 to 2.5, with a mean of 2. This
distribution is both simple and consistent with die point estimate of 2 and the range of estimates
cited above.
The hospital charge component of COI is generally an order of magnitude greater than the
other two components (in thousands versus in hundreds of dollars). Sample mean hospital
charges, as well as standard errors of the means, are provided by Elixhauser et al., 1993. Because
these sample means are generally based on very large samples, asymptotic normality of the sample
mean can be invoked.
The physician charge and opportunity cost are relatively small components of the COI
associated with a hospital admission. Including estimates of uncertainty surrounding these two
small components of WTP to avoid a hospital admission is therefore largely "fine tuning."
Because information concerning the distribution of these components of COI is lacking, these
components are omitted from the uncertainty analysis. The following distributional form was
used for the COI associated with each of the hospital admissions classifications: a normal
distribution with mean = the point estimate (Le., the mean hospital charge + physician charge +
opportunity cost) and standard deviation = die standard error of die mean hospital charge
reported in the Elixhauser et aL study. The distributions of WTP for all PM- and ozone-related
hospital admission endpoints are summarized in Exhibit S.
Abi Associates Inc. 21 Julv 16,1997
-------
Exhibit 5: Distributions of WIT to Avoid Hospital Admissions (1990S)
Hospital Admissions
For.
bcfaemic Heart Disease,
age a 65 (ICD codes 410-
414)
Congestive Heart Failure
age 2 65 (ICD code 428)
COPD,age265(ICD
codes 490496)
Pneumonia, age 2 65
(ICD codes 480-487)
PM-Rcltted"aU
ages (ICD codes 466,
480-482,485.490-493)
Ozone-Related "all
ages (ICD codes 466,
480-486, 490-493)
Asthma (ICD code 493)
•Reap. Illness," age 2 65
(ICD codes 460-519)
Distribution of WTP: WTP-r*OOI, where r- the ratio WIP/COI
Distribution of COL Normal:
Mean
$10.308
$8.280
$7.853
$7.929
S6.344
$6.712
$4,502
$6.277
Standard Deviation
$88
$117
$85
$112
$80
$65
$59
$55
Distribution of r
Continuous Uniform on [1 .5. 2.5]
Continuous Uniform on (1 .5. 2.5]
Continuous Uniform on (1.5. 2.5]
Continuous Uniform on [1 .5. 2 5]
Continuous Uniform on [1 .5, 2.5]
Continuous Uniform on [1.5, 2.5]
Continuous Uniform on (1 .5, 2.5]
Continuous Uniform on [1 .5, 2.5]
2.3 Respiratory Ailments Not Requiring a Hospital Admission
The distributions and point estimates of WTP to avoid the various respiratory ailments not
requiring hospitalization are summarized in Exhibit 6. The derivations of each of the point
estimates and estimated distributions are discussed below.
Abt Associates Inc.
22
Julvl6. 1997
-------
Exhibit 6: Point Estimates and Derived (Recommended) Distributions of the Estimates of
MW IT to Avoid Respirator
Endpolnt
Chronic bronchitis
URS (as defined by Pope et al
1991)
LRS (as defined by Schwartz et
al.. 1994)
"Presence of any of 19 acute
respiratory symptoms"
Acute Bronchitis
Shortness of Breath
Woric Loss Days
Restricted Activity Days
(MRADs)
lufrvlfwfa* f*r AI/rv^» Acfttmfl
Status
y Ailments Not Requiring H
Point Eatfanate of MWTP
$260.000
S19
S12
S18
$45
$5.30
$83
$38
519
capitalization
Derived Distribution of the Estimate of
MWTP (1990S)
A Monte Carlo-generated distribution,
based on three underlying distributions, as
JJJMLJ !• iTijui mot^ fiiftv ha*laTia/ in Section
2.3.1.
Cflntin\KWff wirfof™ distribution over the
interval [$7.00. $32.72]
Continuous unifonn distribution over the
interval [$5.27. $18.57]
Continuous unifonn distribution over the
interval [$0.00. $36.62]
Continuous uniform distribution over the
interval [$13.29. $76 74]
Continuous uniform distribution over the
interval ($0.00. $10.57]
N.A.'
triangular distribution centered at $38.37
on the interval ($15.72, S61.02]
_ . ., .. _ . .
interval f$l 1.81. $53.80]
*"N. A " indicates that a distribution is not available.
23.1 Chronic bronchitis
Chronic bronchitis is one of the only morbidity endpoints that may be expected to last
from the initial onset of the illness throughout the rest of the individual's life. WTP to avoid
chronic bronchitis would therefore be expected to incorporate the present discounted value of a
potentially long stream of costs (e.g., medical expenditures and lost earnings) associated with the
illness. Two studies, Viscusi et al (1991) and Krupnick and Cropper (1992) provide estimates of
WTP to avoid a case of chronic bronchitis. The study by Viscusi et aL, however, uses a sample
that is larger and more representative of the general population than the study by Krapnick and
Cropper (which selects people who have a relative with the disease). The valuation of chronic
bronchitis m this analysis is therefore based on the distribution of WTP responses from Viscusi et
aL(1991).
Both Viscusi et al. (1991) and Krupnick and Cropper (1992), however, defined a.case of
severe chronic bronchitis. It is unclear what proportion of the cases of chronic bronchitis
Abl Associates Inc.
23
July 16. 1997
-------
predicted to be associated with exposure to pollution would turn out to be severe cases. The
incidence of pollution-related chronic bronchitis was based on Abbey et al. (1993), which
considered only new cases of the illness.9 While a new case may not start out being severe,
chronic bronchitis is a chronic illness which may progress in severity from onset throughout the
rest of the individual's life. It is the chronic illness which is being valued, rather than the illness at
onset.
The WTP to avoid a case of pollution-related chronic bronchitis (CB) is derived by
starting with the WTP to avoid a severe case of chronic bronchitis, as described by Viscusi et al.
(1991), and adjusting rt downward to reflect (1) the decrease in severity of a case of pollution-
related CB relative to the severe case described in the Viscusi study, and (2) the elasticity of WTP
with respect to severity. Because elasticity is a marginal concept and because it is a function of
severity (as estimated from Krupnick and Cropper, 1992), WTP adjustments were made
incrementally, in one percent steps. At each step, given a WTP to avoid a case of CB of severity
level sev, the WTP to avoid a case of severity level Q.99*sev was derived. This procedure was
iterated until the desired severity level was reached and the corresponding WTP was derived.
Because the elasticity of WTP with respect to severity is a function of severity, this elasticity
changes at each iteration. I£ for example, it is believed that a pollution-related case of CB is of
average severity, that is, a 50 percent reduction in severity from the case described in the Viscusi
study, then the iterative procedure would proceed until the severity level was half of what it
started out to be.
The derivation of the WTP to avoid a case of pollution-related chronic bronchitis is based
on three components, each of which is uncertain: (1) the WTP to avoid a case of severe CB, as
described in the Viscusi study, (2) the severity level of an average pollution-related case of CB
(relative to that of the case described by Viscusi), and (3) the elasticity of WTP with respect to
severity of the illness. Because of these three sources of uncertainty, the WTP is uncertain.
Based on assumptions about the distributions of each of the three uncertain components, a
distribution of WTP to avoid a pollution-related case of CB was derived by Monte Carlo
methods. The meanjtfthis distribution, whicrxwas about $260,000, is taken as the central.
tendency estimate of WTP to avoid a pollution-related case of CB. Each of the three underlying
distributions is described briefly below.
The distribution of WTP to avoid a severe case of CB was based on the distribution of
WTP responses in the Viscusi study. Viscusi et aL derived respondents' implicit WTP to avoid a
3It is important that only new cases of chronic bronchitis be considered in this analysis
frec^'f e WTP firtfmttra r*R*rt Kfirtimft fflprndrhrrrs find/or knars iirff/KiatiH with this rhronif
illness, and incidences arc predicted separately for each year during the period 1970-1990. Ifthe
total prevalence of chronic bronchitis, rather than the incidence of only new chronic bronchitis
were predicted each year, valuation estimates reflecting lifetime expenditures could be repeatedly
applied to the same individual for many years, resulting in a severe overestimation of the value of
avoiding pollution-related chronic bronchitis.
Abt Associates Inc. 24 July 16.1997
-------
statistical case of chronic bronchitis from their WTP for a specified reduction in risk. The mean
response implied a WTP of about $1,000,000 (1990S)4; the median response implied a WTP of
about $530,000 (1990S). However, the extreme tails of distributions of WTP responses are
usually considered unreliable. Because the mean is much more sensitive to extreme values, the
median of WTP responses is often used rather than the mean. Viscusi et aL report not only the
mean and median of their distribution of WTP responses, however, but the decile points as well.
The distribution of reliable WTP responses from the Viscusi study could therefore be
approximated by a discrete uniform distribution giving a probability of 1/9 to each of the first nine
decile points. This omits the first five and the last five percent of the responses (the extreme tails,
considered unreliable). This trimmed distribution of WTP responses from the Viscusi study was
assumed to be the distribution of WTPs to avoid a severe case of CB. The mean of this
distribution is about $720,000 (1990S).
The distribution of the severity level of an average case of pollution-related CB was
modeled as a triangular distribution centered at 6.5, with endpoints at 1.0 and 12.0. These
severity levels are based on the severity levels used in Krupnick and Cropper, 1992, which
estimated with relationship between ln(WTP) and severity level, from which the elasticity is
derived. The most severe case of CB in that study is assigned a severity level of 13. The mean of
the triangular distribution is 6.5. This represents a SO percent reduction in severity from a severe
case.
The elasticity of WTP to avoid a case of CB with respect to the severity of that case of
CB is a constant times the severity level This constant was estimated by Krupnick and Cropper,
1992, in the regression of ln(WTP) on severity, discussed above. This estimated constant
(regression coefficient) is normally distributed with mean = 0.18 and standard deviation = 0.0669
(obtained from Krupnick and Cropper, 1992).
The distribution of WTP to avoid a case of pollution-related CB was generated by Monte
Carlo methods, drawing from the three distributions described above. On each of 16,000
iterations (1) a value was selected from each distribution, and (2).a value for WIT was generated
by the iterative procedure described above, in which the severity level was decreased by one
percent on each iteration, and the corresponding WTP was derived. The mean of the resulting
distribution of WTP to avoid a case of pollution-related CB was $260,000.
This WTP estimate is reasonably consistent with full COI estimates derived for chronic
bronchitis, using average annual lost earnings and average annual medical expenditures reported
by Cropper and Krupnick, 1990. Using a 5 percent discount rate and assuming that (1) lost
earnings continue Ullage 65, (2) medical expendh^
expectancy is unchanged by chronic bronchitis, the present discounted value of the stream of
medical expenditures and lost earnings associated with an average case of chronic bronchitis is
'There is an indication in the Viscusi paper that the dollar values in the paper are in 1987
dollars. Under this assumption, the dollar values were converted to 1990 dollars.
Abt Associates Inc. 25 July 16, 1997
-------
estimated to be about $77,000 for a 30 year old, about $58,000 for a 40 year old, about $60,000
for a 50 year old, and about $41,000 for a 60 year old. A WTP estimate would be expected to be
greater than a full COI estimate, reflecting the willingness to pay to avoid the pain and suffering
associated with the illness. The WTP estimate of $260,000 is from 3.4 times the full COI estimate
(for 30 year olds) to 6.3 times the full COI estimate (for 60 year olds).
13.2 Upper respiratory symptoms (UBS) and lower respiratory symptoms (LRS)
There are three sources of uncertainty in the valuation of upper or lower respiratory
symptoms: (I) an occurrence of URS or of LRS may be comprised of one or more of a variety of
symptoms (i.e., URS and LRS are each potentially a "complex of symptoms"), so that what is
being valued may vary from one occurrence to another; (2) for a given symptom, there is
uncertainty about the mean WTP to avoid the symptom; and (3) the WTP to avoid an occurrence
of multiple symptoms may be greater or less than the sum of the WTPs to avoid the individual
symptoms.
2.3.2.1 Upper respiratory symptoms
The concentration-response function for URS is taken from Pope et al. (1991). Pope et
al. describe URS as consisting of one or more of the following symptoms: runny or stuffy nose;
wet cough; and burning, aching, or red eyes. The children in the Pope study were asked to
record respiratory symptoms in a daily diary, and the daily occurrences of URS and LRS, as
defined above, were related to daily PM-10 concentrations. Estimates of WTP to avoid a day of
symptoms are therefore appropriate measures of benefit.
Willingness to pay to avoid a day of URS is based on symptom-specific WTPs to avoid
those symptoms identified by Pope et al. as part of the URS complex of symptoms. Three
contingent valuation (CV) studies have estimated WTP to avoid various morbidity symptoms that
are either within the URS symptom complex defined by Pope et al. or are similar to those
symptoms identified by Pope et al. In each CV study, participants were asked their WTP to
avoid a day of each of several symptoms. The median WTP from each of the studies, as well as
midrange estimates derived from them (TJEc, 1993) are shown in Exhibit 7. (Recall that, although
population mean WTPs are to be estimated, the median WTP reported by a study is generally
believed to be a better estimate of this than the mean WTP reported by the study because of the
undue influence of excessively large "protest" bids in CV studies.)
The three individual symptoms in Exhibit 7 that were identified as most closely matching
those listed by Pope etaLfor URS are cough, head/sinus congestion, and eye irritation. Adayof
URS could consist of any one of seven possible "symptom complexes" consisting of at least one
of these symptoms. Using the symptom symbols in Exhibit 7, the seven possible symptom
complexes that could constitute URS, as defined by Pope et aL are:
1. C
2. HC
Abl Associates Inc 26 Juh> 16. 1997
-------
3. E
4. C.HC
5. C,E
6. HC, E
7. C,HC,E.
Exhibit 7: Median WTP Estimate* and Derived Midrange Estimate! (in 1990 $)*
Symptom
Throat
congestion
Head/sinus
congestion
Coughing
Eye
irritation
Headache
Shortness
ofbreath
Pain upon
deep
inhalation
(PDI)
Wheeze
Coughing
up phlegm
Chest
tightness
Symptom
Symbol
TC
HC
C
E
H
S
P
W
CP
CT
Dickie etal.
(1987)
$3.77
$4.40
$1.26
$1.26
$0.00
$4.41
$2.52
$2.75**
$6.30
Tolley et al.
(1986)
$16.35
$17.61
$13.84
$15.72
$25.16
Loehmanetal.
(1979)
$8.20
$4.98
$10.57
Midrange
Estimate
$10.00
$10.00
$700
$15.72
$10.00
$5.00
$4.41
$2.52
$2.75
$6.30
estimates are WTP to cvoid one day of the symptom. Midnnge estimates were derived by Industrial Economic
Inc. (See ffic. 1993).
*»10K trimmed mem. (The median wtsO.)
Using the lEc midrange values for MWTP to avoid each of the separate symptoms, and
assuming that WTP to avoid symptoms is additive (an assumption that is discussed below), the
MWTP to avoid each of the seven types of URS is shown in Exhibit 8 below.
Abt Associates Inc.
27
July] 6,1997
-------
Exhibit 8; Estimates of MWTP to Avoid Upper Respiratory Symptoms
Symptom Combinations Identified
as URS by Pope et al. (1991)
C
HC
E
C,HC
C,E
HC,E
C,HC,E
MWTP to Avoid
$7.00
$10.00
$15.72
$17.00
$22.72
$25.72
$32.72
Average: $18.70
It is assumed that each of the seven types of URS is equally likely. The ex ante MWTP to
avoid a day of URS is therefore the average of the MWTPs to avoid each type of URS, or
$18.70. This is the point estimate for the dollar value for URS used in the benefit analysis. In the
absence of information on the uncertainty surrounding WTP to avoid each of the symptoms within
the URS symptom complex, an uncertainty analysis for WTP to avoid a day of URS would be
based on the distribution of MWTPs in the table above.
Finally, it is worth emphasizing that what is being valued here is URS as defined by Pope
et al, 1991. While other definitions of URS are certainly possible, this definition of URS is used
in this benefit analysis because it is the incidence of this specific definition of URS that has been
related to PM exposure by Pope et al., 1991.
2.3.2.2 Lower respiratory symptoms (LRS)
Schwartz et al. (1994) estimated the relationship between LRS and PM-10 concentrations.
The method for deriving a point estimate of MWTP to avoid a day of LRS is the same as for
URS. Schwartz et aL (1994) define LRS as at least two of the following symptoms: cough, chest
pain, phlegm, and wheeze. The symptoms for which WTP estimates are available that reasonably
match those listed by Schwartz et aL for LRS are cough (C), chest tightness (CT), coughing up
phlegm (CP), and wheeze (W). A day of LRS, as defined by Schwartz et al., could consist of any
one of the 11 combinations of at least two of these four symptoms. Using the lEc midrange
values for WTP to avoid each of these symptoms, and assuming that WTP to avoid symptoms is
additive, the WTP to avoid each of the 15 types of LRS is shown in Exhibit 9.
Abt Associates Inc.
28
July 16,1997
-------
Exhibit 9; Estimates of MWTP to Avoid Lower Respiratory Symptoms
Symptom Combinations Identified as LRS by
Schwartz ct al. (1994)
C.CT
C.CP
C,W
CT, CP
CT.W
CP,W
C, CT, CP
C, CT, W
C, CP, W
CT, CP, W
C,CT,CP,W
MWTP to Avoid
$13.30
$9.75
$9.52
$9.05
$8.82
$5.27
$16.05
$15.82
$12.27
$11.57
$18.57
Average: SI 1.82
It is assumed that each of the eleven types of LRS is equally likely. The ex ante MWTP to
avoid a day of LRS as defined by Schwartz is therefore the average of the MWTPs to avoid each
type of LRS, or $11.82. This is the point estimate used in the benefit analysis for the dollar value
for LRS as defined by Schwartz et al. The WTP estimates are based on studies which considered
the value of a day of avoided symptoms, whereas the Schwartz study used as its measure a case
of LRS. Because a case of LRS usually lasts at least one day, and often more, WTP to avoid a
day of LRS should be a conservative estimate of WTP to avoid a case of LRS.
In the absence of information on the uncertainty surrounding WTP to avoid each of the
symptoms within the LRS symptom complex, an uncertainty analysis for MWTP to avoid a day of
LRS as defined by Schwartz et al. would be based on the distribution of MWTPs in Exhibit 9.
Finally, as with URS, h is worth emphasizing that what is being valued here is LRS or
defined by Schwartz et aL, 1994. While other definitions of LRS are certainly possible, this
definition of LRS is used in this benefit analysis because h is the incidence of this specific
definition of LRS that has been related to PM exposure by Schwartz et al., 1994.
The point estimates derived for MWTP to avoid a day of URS and a case of LRS are
based on the assumption that WTPs are additive. For example, if WTP to avoid a day of cough is
Abl Associates Inc.
29
July 16.1997
-------
$7.00, and WTP to avoid a day of shortness of breath is $5.00, then WTP to avoid a day of both
cough and shortness of breath is $12.00. If there are no synergjstic effects among symptoms, then
it is likely that the marginal utility of avoiding symptoms decreases with the number of symptoms
being avoided. If this is the case, adding WTPs would tend to overestimate WTP for avoidance
of multiple symptoms. However, there may be synergistic effects — that is, the discomfort from
two or more simultaneous symptoms may exceed the sum of the discomforts associated with each
of the individual symptoms. If this is the case, adding WTPs would tend to underestimate WTP
for avoidance of multiple symptoms. It is also possible that people may experience additional
symptoms for which WTPs are not available, again leading to an underestimate of the correct
WTP. However, for small numbers of symptoms, the assumption of additivity of WTPs is
unlikely to result in substantive bias.
2.3.2.3 The distribution of WTP to avoid an occurrence of LRS or URS
There are three sources of uncertainty in the valuation of an of occurrence of upper
respiratory or lower respiratory symptoms: (1) URS and LRS may each be comprised of one or
more of a variety of symptoms (i.e., URS and LRS may each be described as a "complex of
symptoms"), so that what is being valued may vary from one occurrence to another; the
proportions of URI occurrences that are each different type is unknown; (2) for a given
symptom, there is uncertainty about the mean WTP to avoid the symptom; and (3) the WTP to
avoid an occurrence of multiple symptoms may be greater or less than the sum of the WTPs to
avoid the individual symptoms.
Information about the degree of uncertainty from either the second or the third source is
not available. There is, however, some information about the uncertainty associated with the fact
that an occurrence of URS or LRS may vary in symptoms. For example, seven different symptom
complexes that qualify as URS, as defined by Pope et al. (1991), were identified (see Section
4.2.3.1). The estimates of MWTP to avoid these seven different kinds of URS range from $7.00
(to avoid an occurrence of URS that consists of only coughing) to $32.72 (to avoid an occurrence
of URS that consists-ofcoughing plus head/sinus congestion plus eye irritation). There is no.
information, however, about the frequency of each of the seven types of URS among all
occurrences of URS.
To derive a point estimate of MWTP to avoid an occurrence of URS, it was assumed that
each of the seven different kinds of URS is equally likely. Denoting MWTP to avoid an
occurrence of the ith type of URS as MWTPb then MWTP to avoid an occurrence of URS is
7
MWTP - £ AflPTP, * (Probability that occurrence is of the ith type)
Abt Associates Inc. 30 July 16.1997
-------
= (\n)
The point estimate of MWTP to avoid an occurrence of URS, then, is just an average of the seven
estimates of MWTP for avoidance of the different types of URS:
= (1/7)
The same procedure was used for LRS.
The distribution of this estimator of the MWTP to avoid an occurrence of URS (or LRS)
depends on the distributions of the estimates of the individual MWTPj's. If these individual
estimates are normally distributed, then the estimator of MWTP will be normally distributed as
well. However, the variances of the individual estimates are unknown. In addition, there are
other sources of uncertainty, discussed above, that are not quantified but exist.
Because of insufficient information to adequately estimate the distributions of the
estimators of MWTP for URS and LRS, as a rough approximation, a continuous uniform
distribution over the interval from the smallest point estimate to the largest is used. For URS this
is the interval [$7.00, $32.72]. For LRS, it is the interval [$5.27, $18.57]
Alternatively, a discrete distribution of the seven unit dollar values associated with each of
the seven types of URS-identified could be used. This would provide a distribution whose -mean
is the same as the point estimate of MWTP. A continuous uniform distribution, however, is
probably more reasonable than a discrete uniform distribution. The differences between the
means of the discrete uniform distributions (the point estimates) and the means of the continuous
uniform distributions are relatively small, as shown in Exhibit 10.
Exhibit 10: A Comparison of the Means of Discrete and Continuous Uniform Distributions
of MWTP Associated with URS and LRS
Health Endpoint
URS (Pope etal. (1991))
LRS (Schwartz et al. (1994))
Mean of Discrete Uniform
Distribution (Point Est.)
18.70
11.82
Mean of Continuous Uniform
Distribution
19.86
11.92 .
Abt Associates Inc.
31
July 16.1997
-------
2.3.3 Presence of any of 19 acute respiratory symptoms
The sources of uncertainty in the valuation of LRS and URS (see Section 2.3.2) similarly
exist in the valuation of this health endpoint. In particular, (1) "the presence of any of 19 acute
respiratory symptoms" may be comprised of one or more of a variety of symptoms, so that what
is being valued may vary from one occurrence to another, (2) for a given symptom, there is
uncertainty about the mean WTP to avoid the symptom; and (3) the WIT to avoid an occurrence
of multiple symptoms may be greater or less than the sum of the WTPs to avoid the individual
symptoms. Applying the method described for URS and LRS to "presence of any of 19 acute
respiratory symptoms," however, is infeasible.
"Presence of any of 19 acute respiratory symptoms" is a somewhat arbitrary "health
endpoint" used by Krupnick et al. (1990). Moreover, not all 19 symptoms are listed in the
Krupnick study. It is therefore not clear exactly what symptoms were included in the study.
Even if all 19 symptoms were known, it is unlikely that WTP estimates could be obtained for all
of the symptoms. Finally, even if all 19 symptoms were known and WTP estimates could be
obtained for all 19 symptoms, the assumption of additivity of WTPs becomes tenuous with such a
large number of symptoms. The likelihood that all 19 symptoms would occur simultaneously,
moreover, is very small.
Acute respiratory symptoms must be either upper respiratory symptoms or lower
respiratory symptoms. In the absence of further knowledge about which of the two types of
symptoms is more likely to occur among the "any of 19 acute respiratory symptoms," it was
assumed that they occur with equal probability. Because this health endpoint may also consist of
combinations of symptoms, h was also assumed that there is some (smaller) probability that upper
and lower respiratory symptoms occur together.
To value avoidance of a day of "the presence of any of 19 acute respiratory symptoms" it
was therefore assumed that this health endpoint consists either of URS, or LRS, or both. It was
- also assumed that it is as likely to be URS as LRS and that H is half as likely to be both together.
That is, it was assumed that "the presence of any of 19 acute respiratory symptoms" is a day of
URS with 40% probability, a day of LRS with 40% probability, and a day of both URS and LRS
with 20% probability. Using the point estimates of WTP to avoid a day of URS and LRS derived
above, the point estimate of WTP to avoid a day of "the presence of any of 19 acute respiratory
symptoms" is
(0.40)($18.70) + (0.40)($11.82) + (0.20X$18.70 + $11.82) = $18.31 (12)
Because this health endpoint is only vaguely defined, and because of the lack in
information on the relative frequencies of the different combinations of acute respiratory
symptoms that might qualify as "any of 19 acute respiratory symptoms," the unit dollar value
derived for this health endpoint must be considered only a rough approximation.
Abl Associates Inc. 32 July 16,1997
-------
There is insufficient information to derive an accurate estimate of the distribution of the
mean WTP to avoid "the presence of any of 19 acute respiratory symptoms." A distribution was
therefore chosen to satisfy the basic criteria that it be reasonable (i.e., that it be consistent with the
reasoning used to derive a point estimate) and that h be mean preserving (i e, that the mean of the
distribution is the point estimate used). A continuous uniform distribution on the interval [SO.OO,
$36.62] was chosen as a rough approximation to the underlying (unknown) distribution because it
is simple and satisfies these two criteria. The mean of this distribution is $18.31. In addition, the
distribution covers the upper and lower bounds of the intervals on which the unit dollar values for
LRS and URS are distributed.
2.3.4 Acute bronchitis
Estimating WTP to avoid a case of acute bronchitis is difficult for several reasons. First,
WTP to avoid acute bronchitis itself has not been estimated. Estimation of WTP to avoid this
health endpoint therefore must be based on estimates of WTP to avoid symptoms that occur with
this illness. Second, a case of acute bronchitis may last more than one day, whereas it is a day of
avoided symptoms that is typically valued. Finally, the concentration-response function used in
the benefit analysis for acute bronchitis was estimated for children, whereas WTP estimates for
those symptoms associated with acute bronchitis were obtained from adults.
With these caveats in mind, a rough estimate of WTP to avoid a case of acute bronchitis
was derived as the midpoint of a low and a high estimate.9 The low estimate ($13.29) is the sum
of the midrange values recommended by lEc (lEc, 1994) for two symptoms believed to be
associated with acute bronchitis: coughing (S6.29) and chest tightness ($7.00). The high estimate
was taken to be twice the value of a minor respiratory restricted activity day ($38.37), or $76.74.
The midpoint between the low and high estimates is $45.00. This value was used as the point
estimate of MWTP to avoid a case of acute bronchitis in the benefit analysis.
In the absence of sufficient information to characterize the distribution of WTP to avoid a
case of acute bronchitis, this distribution is roughly approximated by a continuous distribution on
the interval from the low estimate to the high estimate, or [$13.29, $76.74].
2.3.5 Shortness of breath
A point estimate of MWTP to avoid a day of shortness of breath was derived as the mean
of the median estimates from two studies that evaluated this symptom. The median estimate from
Dickie et al., 1987, was $0.00; the median estimate from Loehman et aL, 1979, was $10.57. The
mean of these two medians is $5.29. Li the absence of sufficient information to characterize the
distribution of MWTP to avoid a day of shortness of breath, this distribution is roughly
*The derivation of the low and high ^frr**^ of WTP to avoid a case of acute bronchitis are explained in some
detail in EC (1994).
Abt Associates Inc. 33 Jvfy 16.1997
-------
approximated by a continuous distribution on the interval from the low estimate to the high
estimate, or [SO.OO, $10.57].
2.3.6 Moderate or worse asthma status
This health endpoint comes from Ostro et al. (1991), a study in which asthmatics were
asked to record in a daily diary a subjective rating of their overall asthma status each day (0=none,
1-mild, 2=moderate, 3=severe, and ^^incapacitating).
The unit dollar value used for this endpoint is based on a study which asked asthmatics to
estimate their WTP to prevent an increase in *1>ad asthma days" (Rowe and Chestnut, 1986).
Subjects were left to define for themselves what constitutes a bad asthma day. Rowe and
Chestnut found that WTP estimates depended in part on how the subjects defined a bad asthma
day. For example, the mean WTP among subjects defining a bad asthma day as one with any
symptoms was $11.81, whereas the mean WTP among subjects defining a bad asthma day as one
with more than moderate symptoms was $53.80. In general, WTP increased as the definition of a
bad asthma day increased in severity.
Although subjects' assessment of what constitutes a "bad asthma day" varied considerably
in the Rowe and Chestnut study, the subjective assessment of an asthma day being bad is very
similar to the subjective assessment of an asthma day being "of moderate or worse status" in the
Ostro study, in which subjects were also asked their subjective assessments. To estimate WTP to
avoid a day of asthma that is of moderate or worse status, the WTPs from the Rowe and Chestnut
study for all four severity categories in the study (with corresponding WTP estimates of $ 11.81,
$24.93, $39.37, and $53.80) were therefore averaged. The point estimate of WTP to avoid a day
of asthma that is of moderate or worse status is therefore $32.48.
As described above, the point estimate of MWTP to avoid a day of asthma that is of
moderate or worse status is the average of four sample average WTPs corresponding to four
categories of subjects* assessment of what constitutes a "bad asthma day" in a study by Rowe and
Chestnut (1986). The four sample average WTPs are $11.81, $24.93, $39.37, and $53.80. These
four averages may be considered four sample means drawn from four distinct (non-overlapping)
subpopulations, each of which is defined by the way it characterizes a "bad asthma day." The ith
sample mean is an estimate of the MWTP to avoid a day of asthma that is of moderate or worse
status among those individuals with the ith assessment of what constitutes a "bad asthma day."
There is, presumably, substantial variability surrounding each of these four estimates. The extent
of this uncertainty, however, is not known. Nor is the form of any of the four underlying
population distributions of WTP known.
The point estimate of MWTP to avoid a day of asthma that is of moderate or worse status
is thus the mean of four sample means, each of which is based on a sample of unknown size drawn
from an underlying population distribution of WTPs of unknown form. There is therefore
insufficient information to characterize the distribution of this estimate.
Abt Associates Inc. 34 July 16,1997
-------
As a rough approximation, the continuous uniform distribution over the interval [$11.81,
$53.80], defined by the lowest and highest of the four sample means, was used. This is actually a
supposition about the underlying population distribution of WTPs (recombining all four
subgroups), rather than the distribution of the estimate of the MWTP in the population. As such,
it would have greater variance than the distribution of the estimate of the MWTP and might thus
overstate the uncertainty surrounding this estimate. However, this interval is likely to understate
the true range of WTPs in the population distribution, because h is defined by the low and high
sample means rather than by the low and high WTPs in the population. As such, h would
understate the population variance. It would therefore understate the variance of the estimate of
the MWTP and thus understate the uncertainty surrounding the estimate.
Based on insufficient information, then, the interval [$11.81, $53.80] provides a rough
approximation to the distribution of the estimate of MWTP to avoid a day of asthma that is of
moderate or worse status. Because this health endpoint contributes only a small portion of the
total monetized benefits, and the uncertainty surrounding the estimate of MWTP to avoid this
health endpoint is likely to be only a very small portion of the total uncertainty surrounding total
monetized benefit, using a rough approximation rather than a good estimate of this distribution is
likely to have little impact on the assessment of overall uncertainty surrounding total monetized
benefit.
As with URS and LRS, the underlying population distribution of WTP could be a discrete
uniform distribution in which each of the four reported WTPs has equal probability. However, as
with URS and LRS, it is more reasonable to assume a continuous uniform distribution over the
range of these values — [$11.81, $53.80]. The point estimate (the mean of the discrete four-point
distribution) is $32.48. The mean of the continuous uniform distribution is ($11.81 + $53.80)72 =
$32.81. There is little difference in the means. The continuous uniform distribution over the
range [$11.81, $53.80] was therefore chosen.
2.3.7 Work loss days
Willingness to pay to avoid the loss of one day of work was estimated by dividing the
median weekly wage for 1990 (U.S. Department of Commerce, 1992) by 5 (to get the median
daily wage). This values the loss of a day of work at the median wage for the day lost Valuing
the loss of a day's work at the wages lost is consistent with economic theory, which assumes that
an individual is paid exactly the vahie of his labor.
The use of the median rather than the mean, however, requires some comment If all
individuals in society were equally likely to be affected by air pollution to the extent that they lose
a day of work because of h, then the appropriate measure of the value of a work loss day would
be the mean daily wage. It is highly likely, however, that the loss of work days due to pollution
exposure does not occur with equal probability among all individuals, but instead is more likely to
occur among lower income individuals than among high income individuals. It is probable, for
Abt Associates Inc. 35 July 16,1997
-------
example, that individuals who are vulnerable enough to the negative effects of air pollution to lose
a day of work as a result of exposure tend to be those with generally poorer health care.
Individuals with poorer health care have, on average, lower incomes. To estimate the average
lost wages of individuals who lose a day of work because of exposure to PM pollution, then,
would require a weighted average of all daily wages, with higher weights on the low end of the
wage scale and lower weights on the high end of the wage scale. Because the appropriate
weights are not known, however, the median wage was used rather than the mean wage. The
median is more likely to approximate the correct value than the mean because means are highly
susceptible to the influence of large values in the tail of a distribution (in this case, the small
percentage of very large incomes in the United States), whereas the median is not susceptible to
these large values. The median daily wage in 1990 was $83.00.
2.3.8 Minor restricted activity days
No studies are reported to have estimated WTP to avoid a minor restricted activity day
(MRAD). However, EC (1993) has derived an estimate of WTP to avoid a minor respiratory
restricted activity day (MRRAD), using WTP estimates from Tolley et al. (1986) for avoiding a
three symptom combination of coughing, throat congestion, and sinusitis. This estimate of WTP
to avoid a MRRAD, so defined, is $38.37. Although Ostro and Rothschild (1989) estimated the
relationship between PM-2.5 and MRADs, rather than MRRADs (a component of MRADs), it is
likely that most of the MRADs associated with exposure to PM-2.S are in fact MRRADs. For the
purpose of valuing this health endpoint, then, it is assumed that MRADs associated with PM
exposure may be more specifically defined as MRRADs, and the estimate of MWTP to avoid a
MRRAD is used.
Any estimate of MWTP to avoid a MRRAD (or any other type of restricted activity day
other than WLD) will be somewhat arbitrary because the endpoint itself is not precisely defined.
Many different combinations of symptoms could presumably result in some minor or less minor
restriction in activity. It has been argued (Krupnick and Kopp, 1988) that mild symptoms will not
be sufficient to result in a MRRAD, so that WTP to avoid a MRRAD should exceed WTP to
avoid any single mild symptom. A single severe symptom or a combination of symptoms could,
however, be sufficient to restrict activity. Therefore WTP to avoid a MRRAD should, these
authors argue, not necessarily exceed WTP to avoid a single severe symptom or a combination of
symptoms. The "severity" of a symptom, however, is similarly not precisely defined; moreover,
one level of severity of a symptom could induce restriction of activity for one individual while not
doing so for another. The same is true for any particular combination of symptoms.
Given that there is inherently a substantial degree of arbitrariness in any point estimate of
WTP to avoid a MRRAD (or other kinds of restricted activity days), the reasonable bounds on
such an estimate are considered. By definition, a MRRAD does not resuh in loss of work. WTP
to avoid a MRRAD should therefore be less than WTP to avoid a WLD. At the other extreme,
WTP to avoid a MRRAD should exceed WTP to avoid a single mild symptom. The highest EEc
midrange estimate of WTP to avoid a single symptom is $15.72, for eye irritation. The point
estimate of WTP to avoid a WLD in the benefit analysis is $83. If all the single symptoms
Abl Associates Inc. 36 Juty 16.1997
-------
evaluated by the studies listed in Exhibit 4.S are not severe, then the estimate of WTP to avoid a
MRRAD should be somewhere between $15.72 and $83.00. Because the EEc estimate of $38.37
falls within this range (and acknowledging the degree of arbitrariness associated with any estimate
within this range), the lEc estimate is used as the point estimate of MWTP to avoid a MRRAD.
2.4 Welfare Endpoints
The distributions and point estimates of WTP to avoid the welfare endpoints considered in
this analysis are summarized in Exhibit 11. The derivations of each of the point estimates and
estimated distributions are discussed below.
Exhibit 11: Point Estimates and Derived (Recommended) Distributions of the Estimates of
MWTP to Avoid Welfare Endpoints Considered in the Analysis
Endpoint
Worker Productivity
Visibility - Residential
Visibility • Recreational
Consumer Cleaning Cost Savings
Point Estimate of MWTP
$1 Per worker per 10% change in
o,
$14 per unit change in dv
Derived Distribution of the Estimate of
MWTP (1990S)
N.A.
Triangular distribution centered at $14 on
the interval [$8. $21]
see Section 2.4.3 and Exhibit 12
$2.52 per ugAn* change in PM10
per household
Beta distribution with mean=S2.S2,
standard deviation's 1.00 on the interval
[$1.26, $10.08]. The shape parameters of
this distribution are ec-1.2 and fl*7 J.
•T4.A." indicates that a distribution is not available.
2.4.1 Worker Productivity
The valuation used to monetize benefits associated with increased worker productivity
resulting from improved ozone air quality is based on information reported in Crocker and Horst,
1981 and summarized in U.S. EPA, 1994. Crocker and Horst (1981) examined the impacts of
ozone exposure on the productivity of outdoor citrus workers. Productivity impacts were
measured as the change in income associated with a change in ozone exposure, given as the
elasticity of income with respect to ozone concentration (-0.1427). The reported elasticity, which
is used as the central estimate in this analysis, translates a 10 percent reduction in ozone to a 1.4
percent increase in income. Given the average dairy income for outdoor workers engaged in
strenuous activity reported by the 1990 U.S. census, $73 per day, the 10 percent reduction in
ozone yields approximately $1 in increased daily wages. No information was available for
quantifying the uncertainty associated with the central valuation estimate.
Abl Associates Inc.
37
Jufy 16. J997
-------
2.4.2 Visibility - Residential
Based on predicted visibility improvements modeled for each county nationwide, national
monetary benefits are computed as follows:
Benefits - £, WTP * A*A>, * HHt
t
where:
Benefits = national benefits attributable to residential visibility improvements,
WTP = annual household willingness to pay per unit change in deciview (dv),
AdVj = change indvin county I between baseline and post-control scenarios
HH, = # of households in county I (calculated as county population / people per
household (2.68 according to 1990 Census)).
The household WTP value for improved residential visibility is based on a survey of
existing contingent valuation studies conducted by McClelland et al. (1991). Based on the results
of that survey, this analysis uses a central estimate of the annual household WTP for visibility
improvements of $14 per unit improvement in deciview.6 The uneven reliability of the published
studies examined by McClelland et al. (1991) complicated the characterization of the distribution
associated with the central estimate, leading to the selection of a hypothesized triangular
distribution of values to characterize the uncertainty associated with the visibility valuation The
upperand lower bound values of the triangular distribution were estimated using a consensus
function derived from a regression analysis of the results reported by the published studies
examined by McClelland et al. (1991).7 The consensus function approach, incorporating the part-
whole bias adjustment, yields estimated lower and upper bound values of $8 and $21,
respectively, for annual household WTP per unit improvement in deciview.
2.4.3 Visibility - Recreational
Section 169 A of the Clean Air Act established a goal of preventing and remediating
visibility impairments in national parks resulting from anthropogenic influences. Chestnut (1997)
has developed a methodology for estimating the value to the U.S. public of visibility
improvements in Class I visibility areas. Based on contingent valuation studies, Chestnut
l a txmmrm viability metric, eharariM-iw* visibility fa terras pf p-nxf/tiblf dlCTgfS in h
independent of baseline conditions.
The lower bound estimate included an adjustment to correct study •**""•*«*' for part-whole bias (failure to
differentiate values for visibility from those for other air quality amenities, such as reductions in adverse health effects).
For all studies which did not correct for part-whole bias, this analysis applied a correction of 0.25 (i.e., reported values
were multiplied by 0.2S). This represents the approximate midpoint of the range of applicable correction factors
reported by McClelland et al. (1991), Chestnut and Rowe (1989) and Irwin et al. (1990).
Abt Associates Inc. 38 Juty 16,1997
-------
calculates a household willingness to pay (HHWTP) for visibility improvements, capturing both
use and non-use recreational values, and attempts to account for geographic variations in
willingness to pay. The Chestnut methodology for determining HHWTP has been adopted for the
purposes of this benefits analysis.
In order to account for geographic variability in HHWTP, Chestnut (1997) divides the
recreational areas of the United States into three regions: California, Southwest, and Southeast.
The California region is comprised of the state of California; the Southwest is comprised of
Arizona, Nevada, Utah, Colorado, and New Mexico; the Southeast is comprised of Delaware,
Maryland, West Virginia, Virginia, Kentucky, Tennessee, North Carolina, South Carolina,
Georgia, Alabama, Florida, and Mississippi. The regions were developed to capture differences in
HHWTP values based on proximity to recreational areas. That is, in-region respondents typically
place higher value on visibility improvements at a local recreational area than out-of-region
respondents. Chestnut reports both in-region WTP and out-of-region WTP for each of the three
regions. The WTP in a North Carolina county for recreational visibility improvements in all three
regions, for example, is simply the sum of the out-of-region WTP for California and Southwest
visibility improvements and the in-region WTP for visibility improvements predicted for the
Southeast region.
Chestnut offers the following general functional form for WTP for visibility improvements:
WTP= HH * P *
where:
WTP = annual willingness-to-pay for visibility changes in a given year
HH = # of households: calculated as population / people per household (2.68
according to 1990 Census).
P = regression coefficient estimating the relationship between WTP and a
given change in visibility (as measured in Visual Range)
VR1 = annual average visual range (km) predicted in the baseline scenario
VR2 = annual average visual range (km) predicted in the post-control scenario
The estimated coefficient, P, described in the formula above, is derived from a regression
analysis of contingent valuation results, and quantifies the relationship between WTP and visibility
changes. Exhibit 12 documents the p coefficients reported by Chestnut (1997). As indicated,
Chestnut's coefficients vary by visibility region and by whether the value applies to residents
inside or outside of a given region. Exhibit 12 also indicates the WTP values in terms of a second
visibility metric, decrview.
Abt Associates Inc. 39 July ]6, 1997
-------
Exhibit 12: Visibility Valuation Estimates-Estimated p Coefficients from the National
Parks Visibility Valuation Study and Corresponding WTP per Deciview Improvement
Visibility
Region
California
Southwest
Southeast
Out-of-Region Household* | lo-Regfon Houieholdi
Estimated B
(95% CL)
73
(50-96)
110
(80-139)
40
(38-42)
CorrMpoodiBf WTP
per dcdview change
(95% CX)*
$7J
($5.0-9.6)
$11.0
(S8.0-S13.9)
S40
(S3.8-S4.2)
Estimated p (95%
CL)
105
(65-145)
137
(111-163)
65
(57-73)
Corresponding
VT ir per OCUVKW
change (95% CJ.)'
S10.S
(S6.5-S14.5)
$13.7
($11 1-S16.3)
$6.5
($57-$7.3)
' conversion from visual range to deciview based on relationship reported in Pitchford and Malm (1994)
Chestnut also concludes that, for a given region, a substantial proportion of the WTP is
attributable to one specific park within the region. This so called "indicator park" is the most
well-known and frequently visited park within a particular region. The indicator parks for the
three regions are Yosemite for California, the Grand Canyon for the Southwest, and Shenandoah
for the Southeast. The Chestnut methodology derives a weighting scheme which accounts for the
proportion of WTP attributable to the indicator parks versus that attributable to the other
recreational areas in a given region.
In accordance with the Chestnut (1997) methodology, this benefits analysis calculates out-
of-region and in-region benefits for a particular region for a gjven change in Class I area visibility
as follows:
where:
40%
60%
P
VR1
VR2
[40% * p,
60%
• total out-of-region benefits for a single recreational area region
! percentage of WTP attributable to indicator park
: percentage of WTP attributable to remaining parks in a region
: estimated out-of-region coefficient (capturing HHWTP)
: annual average visual range predicted in the baseline scenario
• annual average visual range predicted in the post-control scenario
•• total number of out-of-region households
and
Abi Associates Inc.
40
July 16.1997
-------
[50% 'Pfc^ *
50% * *
where:
Ben«wt** ~ total ui-region benefits for a single recreational area region
50% = percentage of WTP attributable to indicator park
p - estimated in-region coefficient (capturing HHWTP)
VR1 - annual average visual range predicted in the baseline scenario
VR2 - annual average visual range predicted in the post-control scenario
= total number of in-region households
In theory, summing Ben^^.^^ and Beiv,^, would yield the total monetary benefits
associated with a given visibility improvement in a particular recreational area region, which could
then be summed across regions to estimate national benefits. However, in addition to recreational
visibility benefits, this analysis estimates residential visibility benefits (discussed above). Chestnut
(1997) indicates that there is some uncertainty regarding the extent to which these benefit
categories may overlap (i.e., whether an individual's WTP for improvements in local visibility
(residential) is completely independent from that individual's WTP for visibility improvements in
recreational areas).
In an effort to account for the possibility of overlap between residential and recreational visibility
benefits, Chestnut (1997) proposes three alternative aggregation schemes:
• A low estimate valuing recreational visibility improvements to out-of-region residents
only, assuming that all of the recreational value to in-region residents is reflected in the
residential WTP.
• A central estimate valuing recreational visibility improvements by applying the out-of-
region WTP to all residents both in and out of the park region. This assumes that the
double counting of residential and park visibility values for in-region residents is reflected
by the differential between in-region and out-of-region WTP values for parks.
• A high estimate valuing recreational visibility improvements using the full WTP value
from the parks study added to the residential values with no adjustment for possible
double-counting.
This benefits analysis uses a Monte Carlo technique to combine the low, medium and high
estimates, with all three aggregation approaches treated as equally likely . Equal weights are
assumed due to a lack of information regarding the most appropriate of the three aggregation
assumptions. The alternative would involve assigning arbitrary weights to each of the three
approaches (for example, doubling the possibility of the central estimate compared to the low and
the high). However, in addition to being arbitrary, this would increase the mean benefits estimate
(compared to the equal weights approach) and decrease the overall uncertainty, two
insupportable outcomes. Therefore, by assuming equal weights for the three alternative
Abl Associates Inc. 41 July 16. 1997
-------
aggregation schemes, this analysis opts for the more conservative estimate of the benefits
associated with a given visibility change, and maintains the overall uncertainty in the calculation.
2.4.4 Consumer Cleaning Cost Savings
Several studies have provided estimates of the cost to households of PM soiling. The
study that is cited by ESEERCO (1994) as one of the most sophisticated and is relied upon by
EPA in its 1988 Regulatory Impact Analysis for SO2 is Manuel et aL (1982). Using a household
production function approach and household expenditure data from the 1972-73 Bureau of Labor
Statistics Consumer Expenditure Survey for over twenty cities in the United States, Manuel et al.
estimate the annual cost of cleaning per jig/m3 PM per household as SI.26 (S0.48 per person
times 2.63 persons per household). This estimate is low compared with others (e.g., estimates
provided by Cummings et al., 1981, and Watson and Jaksch, 1982, are about eight times and five
times greater, respectively). RCG/Hagler Bailly notes, however, that the Manuel estimate is
probably downward biased because it does not include the time cost of do-it-yourselfers.
Estimating that these costs may comprise at least half the cost of PM-related cleaning costs, they
double the Manuel estimate to obtain a point estimate of $2.52 (reported by RCG/Hagler Bailly in
1992 dollars as S2.70).
The Manuel study measured paniculate matter as TSP rather than PM-10 or PM-2.5. If a
one ug/m9 increase in TSP causes $1.26 worth of cleaning expenses per household, the same unit
dollar value can be used for PM-10 (or PM-2.5) only if particle size doesn't matter — i.e., only if
particles of all sizes are equally soiling. Suppose, for example, that PM-10 is 75%/>f TSP and
that all particles are equally soiling. Then 75% of the damage caused by a 1 ug/m3 increase in
TSP is due to PM-10. This is (0.75)($1.26) = $0.95. However, this corresponds to a 0.75 jig/m*
increase in PM-10. A one ug/m3 increase in PM-10 would therefore yield a dollar soiling damage
of $0.95/0.75 = $1.26.
Suppose, however, that only PM-10 matters. Then the $1.26 underestimates the impact
of a one ug/m3 increase in PM-10, because it corresponds to a less than one ug/m3 increase in
PM-10, e.g., a 0.75 ug/m3 increase in PM-10. In this case, the correct unit value per unit of PM-
10 would be ($1.26y0.75 = $1.68. If only PM-10 matters, then either (1) the dollar value can be
adjusted by dividing h by the percentage of TSP that is PM-10 and PM-10 can be used in the
soiling damage function, or (2) the dollar value can be left unadjusted and TSP, rather than PM-
10, can be used in the soiling damage function.
Finally, h is possible that, while both PM-10 and PM-2.5 are components of TSP that
cause consumer cleaning costs, the remaining portion of TSP has a greater soiling capability than
either the PM-10 or PM-2.5 component In this case, using either PM-10 or PM-2.5 air quality
data with a household soiling function based on TSP would yield overestimates of the PM-10- or
PM-2.5-related consumer cleaning costs avoided by attainment of a given set of alternative
standards.
Abt Associates Inc. 42 July 16, 1997
-------
There is, however, insufficient information on the relative soiling capabilities of the
different components of TSP. In the absence of such information, the conservative approach
would be to assume that only PM-10 matters and to make the appropriate adjustments. There is
also insufficient information, however, to make these adjustments in a national analysis. In
particular, the percentage of TSP that is PM-10 in each county in the United States is not known.
While it might be possible to substitute region-specific percentages for county-specific
percentages, doing this would simply introduce uncertainty from another source
2.5 Changes Over Time in WTP in Real Dollars
The WTP for health-related environmental improvements (in real dollars) could change
between now and the year 2010. If real income changes between now and the year 2010, for
example, it is reasonable to expect that WTP, in real dollars, would also change.
Based on historical trends, the Bureau of Economic Analysis (BEA) projects that, for the
United States as a whole as well as for regions and states within the U.S., mean per capita real
income will increase. For the U.S. as a whole, for example, mean per capita personal income is
projected to increase by about 16 percent from 1993 to 2005 (U.S. Dept. of Commerce, 1995).
It was argued above that the mean WTP in the population is the correct measure of the
value of a health problem avoided, and that WTP is a function of income. If the mean per capita
real income rises by the year 2010, rt might be inferred that the mean WTP would rise as well.
While this is most likely true, the degree to which mean WTP rises with a rise in mean per capita
income is unclear unless the elasticity of WTP with respect to real income is 1.0. The problem
derives from having means (mean income and mean WTP) instead of values for each individual, as
explained in the attached Appendix.
There is some evidence (Alberini et al., 1994; Mitchell and Carson, 1986; Loehman and
Vo Hu De, 1982) that the elasticity of WTP for health-related environmental improvements with
respect to real income is less than 1.0, possibly substantially so. If this is the case, then changes in
mean income cannot be readily translated into corresponding changes in mean WTP. Although an
increase in mean income is likely to imply an increase in mean WTP, the degree of the increase
cannot be ascertained from information only about the means.
Several factors, in addition to real income, that could affect the estimated benefit
associated with reductions in PM concentrations could also change by the year 2010.
Demographic characteristics of exposed populations could change. Technological advances could
change both the nature of PM emissions to the ambient air and the susceptibility of individuals to
those emissions. Any such changes would be reflected in concentration-response functions that
differ from those that describe current relationships between PM and various health endpoints.
While adjustments of WTP to reflect changes in real income are of interest, such adjustments
would by no means necessarily reflect all possible changes that could affect the benefits of
reduced PM concentrations in the year 2010.
Abt Associates Inc. 43 July 16. 1997
-------
Because there are likely to be changes in factors that cannot be captured that could change
benefits in the future, and because current mean WTPs cannot be adequately adjusted for changes
in mean real income because of the likely nonlinearity of the relationship between WTP and real
income, WTP estimates for the future year were left unadjusted.
3.0 AGGREGATION OF DOLLAR BENEFITS: POINT ESTIMATES
The dollar benefits from ozone reductions resulting from attainment of a proposed ozone
NAAQS is just the sum of dollar benefits from the reductions in incidence of all non-overlapping
health and welfare endpoints with which ozone is associated. Similarly, the dollar benefits from
PM reductions is the sum of dollar benefits from the reductions in incidence of all non-
overlapping health and welfare endpoints with which PM is associated. Finally, the total dollar
benefits accruing to attainment of a proposed ozone NAAQS is the sum of all ozone-related
benefits and all ancillary PM-related benefits.
If two endpoints are overlapping, then adding the benefits associated with each will result
in double counting of some benefits. Although study-specific point estimates of dollar benefits
associated with specific, possibly overlapping endpoints will be presented separately, estimation of
total benefits requires that the benefits from only non-overlapping endpoints be included in the
total. Four non-overlapping broad categories of health and welfare endpoints are included in the
estimation of total dollar benefits for the revised RIA for ozone and PM NAAQS: (1) mortality,
(2) hospital admissions, (3) respiratory symptoms/illnesses not requiring hospital admission, and
(4) welfare endpoints. The specific endpoints of interest, the studies that have estimated
concentration-response functions for these endpoints, and the populations considered in the
studies are listed in Exhibit 13 for ozone and Exhibit 14 for PM.
Exhibit 13; Studies of Ozone and Health and Welfare Endpoints
Endpotnt
(Short-term exposure) Mortality
Study
Population Coniidered
AnderaonetaL. 1996
Cifoeotes and Lave, 1996
Hoek et aL, 1997 Cm press)
Ito and TbiRton. 1996
Kimeyetal., 1995
Looms et al., 1996 (HET)
all
all
all
an
all
all
Abt Associates Inc.
44
July 16. 1997
-------
Moolgavkaretal.. 1995
Ostroetal., 1996
Samet et al.. 1996, 1997 (HEI)
Touloumi et aL, 1997
VerboeffetaL.1996
all
all
all
all
all
HotpKal Admbsiont
"all respiratory illnesses"
"til respiratcny illnesses"
"all respiratoiy illnesses"
"all respiratoiy illnesses"
"all respiratoiy JMry*g*'«"
COPD
COPD
COPD
Pneumonia
Pneumonia
P&CUDKXlift
EnicrggicY dcpt visits for tffthnui
m* + r
Schwartz. 1996 (Spokane)
Schwartz, 1995 (New Haven)
Schwartz. 1995 (Tacoma)
Thunton et al., 1994 (Toronto)
ThurstonetaL. 1992 (NYC)
Schwartz. 1996 (Spokane)
ocnwartz, iys»4a
Schwartz. 1994b
Schwartz, 1994a
Schwartz. 1994b
Srhwartr 1QQ4£
Wosdetal.. 1995
age 65+
age 65+
age 65+
all
all
age 65+
age 65+
age 65+
age 65+
age 65+
age 65+
all
Reipiratory Symptoms not Requiring HotptUllzation
Acute respiratoiy symptoms (any of 19)
A«il>TTi« ittacks
Chronic sinusitis and hay fever
Development of asthma
MRADs
RRADs
Welfare Endpolnti
Decreaaed worker productivity
Krupnicketal..l990
Whittemore and Kom, 1980 and
US EPA. 1993
Portney and Mullahy. 1 990
Abbey etal.. 1993. 1995
Ostro and Rothschild. 1989
Ostro and Rothschild. 1989
ages 18-65
asthmatics
adults
adult males, ages 25+
ages 18-65
ages 18-65
Based on Crocker and Hont.
1981 and US EPA, 1994
laborers
Exhibit 14: Studies of PM and Health and Welfare Endpoints
Abt Associates Inc. 45
July 16,1997
-------
Endpolnt
Study
Population Considered
Mortality
Short-term exposure mortality
Short-term exposure mortality
Long-term exposure mortality
Hof pK*l Admit ilont
•
*•«!! mnirafiii'if illnMBw"
ui icajjuuuijr HUH Am
"all respiratory illnesses"
COPD
Pneumonia
Congestive heart failure
Ischemic heart disease
Pooled analysis (10 functions)
Schwartz etal.. 1996
Pope etaL. 1995
all
all
ages 30+
Schwntr 10QS lOOfUnoolMiftnAtairi
5CQW4iU, lyyj, lyyo \poouit anaiysuj
Thurston et al.. 1994 (Toronto)
Schwartz. 1994 a, 1994b. 1996
Schwartz. 1994 a, 1994b. 1994c. 1996
Schwartz and Morris. 1 995
Schwartz and Morris. 1995
age 65+
all
age 65+
age 65+
age 65+
age 65+
Respiratory Symptomt/Dlnesie* not Requiring Hospttalization
. «_ i .,.
Acute respiratory symptoms
(any of 19)
w« < * •
Development of chronic
bronchitis
(LRS)
Tinner n*cnintnrv
-------
3.1 Aggregation of benefits from redactions in ozone
3.1.1 Mortality
Ten studies of the relationship between daily ozone concentrations and premature
mortality satisfy the study selection criteria used to identify acceptable studies to include in a
quantitative analysis of the relationship between daily ozone and mortality. This includes all the
mortality studies listed in Exhibit 13 except Touloumi et al., 1997, which is itself a meta-anarysis
of the results of studies in several locations in Europe. The study selection criteria and the
estimation of a distribution of the national incidence of ozone-related premature mortality, based
on the studies selected, are described in detail in ** Assessment and Synthesis of Available
Epidemiological Evidence of Mortality Associated with Ambient Ozone From Daily Time-Series
Analyses". The point estimate of the dollar benefit of reductions in ozone-related premature
mortality is the product of the mean of this distribution and the mean of the distribution of WTP
to save a statistical life ($4.8 million).
3.1.2 Hospital admissions
Two studies, Thurston et al., 1992 and Thurston et al., 1994, consider ozone-related
hospital admissions for "all respiratory illnesses" for all ages, a category which includes all other
hospital admission endpoint/population combinations for which ozone concentration-response
relationships have been estimated. To be consistent with the previous risk assessment for ozone,
the benefits of reductions of hospital admissions for "all respiratory illnesses" for all ages will be
based on the concentration-response function from Thurston et aL, 1992, and will be the only
hospital admissions endpoint included for ozone. The point estimate of the dollar benefit of
reductions in ozone-related hospital admissions is therefore the product of the mean of the
incidence distribution, based on Thurston et al., 1992, and the mean of the distribution of WTP to
avoid such a hospital admission ($13,425 — see Exhibits 4 and 5).
3.1.3 Non-hospital respiratory symptoms/illnesses
Among the ozone-related respiratory symptoms and illnesses not requiring hospitalization
listed in Exhibit 13, chronic sinusitis and hay fever, development of asthma, and RRADs are
quantified but not monetized (see Exhibit 1). Issues of aggregation of ozone-related respiratory
symptoms and/or illnesses therefore involve only acute respiratory symptoms (any of 19), asthma
attacks, and MRADs. There is possible overlap between all three of these endpoints. Acute
respiratory symptoms (any of 19) is more inclusive than asthma attacks (because h includes
asthma attacks as one of many possible symptoms) and applies to a more general population. It is
therefore preferred over asthma attacks as an endpoint to be included in the aggregation. It is
unclear, however, comparing "any of 19 acute respiratory symptoms" and MRADs, which is more
inclusive. Therefore, to be consistent with the 1996 draft RIA for ozone, MRADs are excluded
from the aggregation, to avoid possible double counting of benefits. The only endpoint among
the three that is included in the calculation of total ozone-related benefits, then, is "any of 19
acute respiratory symptoms."
Abt Associates Inc. 47 Jufy 16. 1997
-------
The health and welfare endpoints that are included in the estimation of total ozone-related
benefits are given in Exhibit 15.
Exhibit 15: Health and Welfare Endpoints Included in the Estimation of Total Ozone-
Related Benefits
Endpolnt
Mortality
Population to Which Applied
all ages
Study
pooled analysis of 10 studies
Hospital Admiiiioni
"all respiratory"
all ages
Thurstonetal, 1992
Respiratory Syraptoms/Dlnesses Not Requiring Hospttattzatkm
Acute respiratory symptoms (any of 19)
ages 18-65
Krupmcketal., 1990
Welfare Endpoints
Decreased worker productivity
laborers
Based on Crocker and Horst, 1981
and US EPA, 1994
3.2 Aggregation of benefits from reductions in PM
3.2.1 Mortality
Estimates of the benefits associated with reductions in short-term mortality related to
reductions in PM 10 cannot be added to estimates of the benefits associated with reductions in the
same endpoint related to reductions in PM2.S.
Because mortality is such an important endpoint, three separate estimates of total PM-
related benefits will be presented for each scenario, including the benefits of reductions in the
incidence of
(1) long-term exposure mortality among individuals age 30+ associated with reductions in
PM2.5;
(2) short-term exposure mortality among individuals of all ages associated with reductions
inPMlO.and
(3) short-term exposure mortality among individuals of all ages associated with reductions
in PM2.S.
3.2.2 Hospital admissions
Abl Associates Inc. 48 July J6, 1997
-------
The category of hospital admissions for "all respiratory illnesses" for all ages, for which a
PM2.5 concentration-response function was estimated by Thurston et al., 1994, clearly contains
the categories of hospital admissions for COPD and hospital admissions for pneumonia. It is
overlapping, however, with the other "all respiratory illness" hospital admissions studies
(Schwartz, 1995,1996) because, although those studies considered only ages 65 and up, the
definition of "all respiratory illnesses" was broader, including more ICD codes. It is likely,
however, that the ICD codes used by Thurston et al. include the most important PM-related
illnesses. Therefore only the Thurston study will be used for respiratory illnesses.
Ischemic heart disease and congestive heart failure are non-overlapping categories, and
each is non-overlapping with the category "respiratory illness." The point estimate of the total
dollar benefit of reductions in PM-related hospital admissions is therefore the sum of three means:
(1) the mean of the distribution of dollar benefits resulting from reductions in PM-related hospital
admissions for "all respiratory illnesses." This is the product of the mean of the
distribution of incidence, based on Thurston et al., 1994, and the mean of the distribution
of the corresponding unit dollar value ($12,688 - see Exhibits 4 and 5);
(2) the mean of the distribution of dollar benefits resulting from reductions in PM-related hospital
admissions for congestive heart failure. This is the product of the mean of the distribution
of incidence and the mean of the distribution of the corresponding unit dollar value
($16,599 - see Exhibits 4 and 5); and
(3) the mean of the distribution of dollar benefits resulting from reductions in PM-related hospital
admissions for ischemic heart disease. This is the product of the mean of the distribution
of incidence and the mean of the distribution of the corresponding unit dollar value
($20,615 - see Exhibits 4 and 5).
3.2.3 Non-hospital respiratory symptoms/illnesses
Among those PM-related respiratory symptoms or illnesses not requiring hospitatization,
acute bronchitis, development of chronic bronchitis, lower respiratory symptoms (LRS), and
upper respiratory symptoms (URS) are either non-overlapping symptom/illness categories or are
for non-overlapping populations. The benefits associated with these endpoints can therefore be
aggregated without any problem of double counting. Because shortness of breath among African-
American asthmatics, age 7-12, overlaps with either LRS or URS, and because this subpopulation
is likely to be much smaller than the subpopulations to which LRS and URS are applied, this
endpoint will not be included in the total. RADs do not have a unit dollar value and therefore
cannot be included in the aggregation.
Aggregation problems are presented, however, by the set of endpoints that are applied to
adults (either ages 18-65 or ages 18-70). These endpoints include acute respiratory symptoms
(any of 19), moderate or worse asthma, MRADs, and work loss days (WLDs). "Any of 19 acute
respiratory symptoms," MRADs, and moderate or worse asthma may all overlap with each other.
Abt Associates Inc. 49 July 16.1997
-------
MRADs and WLDs, however, are defined to be mutually exclusive (see Ostro and Rothschild,
1989). In addition, the study that estimated concentration-response functions for "any of 19 acute
respiratory symptoms" measured COH rather than PM10. For these reasons, the PM-related
benefits associated with reductions in "any of 19 acute respiratory symptoms," while shown
separately, will not be included in the calculation of total PM-related benefits. Because of
possible overlap between MRADs and moderate or worse asthma, and because the subpopulation
of asthmatics is much smaller than the population to which the MRAD endpoint was applied,
benefits associated with reductions in moderate or worse asthma are also omitted from the
calculation of total PM-related benefits.
Finally, one study reports concentration-response relationships between the same endpoint
and two different measures of PM. Schwartz et al., 1994, report estimated relationships between
lower respiratory symptoms and both PM10 and PM2.S concentrations for children ages 8-12.
Both estimated coefficients are statistically significant. Clearly, cases of LRS avoided from
reductions of PM10 and PM2.S cannot both be included in the estimation of total benefits. When
the benefits associated with attainment of PM10 standards are being estimated, it makes sense to
use the PM10 coefficient from the Schwartz et al. study, to avoid having to make assumptions
about how PM2.5 concentrations change in efforts to reduce PM10. Similarly, when the benefits
associated with attainment of PM2.5 standards are being estimated, h makes sense to use the
PM2.S coefficient from the Schwartz et al. study for the analogous reason. When the benefits
associated with ancillary reductions in PM resulting from attainment of an ozone standard are
being estimated, however, there is no clear choice of which coefficient to use. The numbers of
cases avoided are similar (e.g., about 9,000 cases of PM2.S-related LRS versus about 7,000 cases
of PMlO-related LRS from ancillary PM reductions resulting from the proposed ozone NAAQS
versus the 2010 baseline), and the dollar benefits are very small, in either case, compared to the
benefits contributed by other health endpoints. The choice of PM10 versus PM2.5 will therefore
have little impact. Given that, the (somewhat arbitrary) choice of PM2 5-related LRS will be
used.
The health and welfare endpoints that are included in the estimation of total PM-related
benefits are given in Exhibit 16.
Abt Associates Inc. SO Juty 16. 1997
-------
Exhibit 16: Health and Welfare Endpoints Included in the Estimation of Total PM-Related
Benefits
Endpoint
Population to Which Applied
Study
Mortality: Three Alternatives:
1. PM-2.5-related long-term exposure
mortality
2. PMlO-rdated short-torn
exposure mortality
3. PM2.S-reUted short-term
exposure mortality
ages 30+
allages
all ages
Pope etaL. 1995
Pooled analysis (10 functions)
Schwartz etal., 1996
Hospital Admissions
"all respiratory"
congestive heart failure
ischemic heart disease
all ages
age 65+
age 65+
Thurston etal. 1994
Schwartz and Morris, 199S
Schwartz and Morris. 1 995
Respiratory Symptoms/nineties Not Requiring HoipHalization
Acute bronchitis
Development of chronic bronchitis
riVLi.j-rejaieo tower rcspiraiofy
symptoms (LRS)
U ' atorv tonurURSt
U|ipci ••uppUBiuijT fjutfiUtua \\JKaJ
MRADs
Work loss days (WLDs)
Welfare Endpoints
Consumer cleaning cost savings
Visibility
ages 10-12
all
ages 8-12
. ^ 0 . .
ages 18-65
ages 18-65
all
all ages
Dockery etal.. 1989
Schwartz, 1993
Schwartz etal. 1994
Pope etal., 1991
Ostro and Rothschild, 1989
Ostro. 1987
ESEERCO. 1994
Pitchford and Malm. 1994
4.0 AGGREGATION OF DOLLAR BENEFITS: DISTRIBUTIONS
When considering only point estimates, aggregation of the benefits from different sources
is relatively straightforward. Once a set of non-overlapping categories has been determined, the
point estimate of the total benefits associated with the health and welfare endpoints in the set is
just the sum of the endpoint-specific point estimates. If each endpoint-specific point estimate is
the mean of a distribution of dollar benefits associated with that endpoint, then the point estimate
of total dollar benefits is just the sum of those means.
Abt Associates Inc.
51
July 16.1997
-------
The point estimate of total dollar benefits associated with attainment of a given PM
standard is just the sum of the point estimates of benefits from reductions in all non-overlapping
PM-related health and welfare endpoints. The point estimate of total dollar benefits associated
with attainment of a given ozone standard is the sum of the following two sums:
(1) the sum of estimates of benefits from reductions in all non-overlapping ozone-related
health and welfare endpoints; and
(2) the sum of estimates of benefits from ancillary reductions in all non-overlapping PM-
related health and welfare endpoints.
The estimate of total benefits may be thought of as the end result of a sequential process in
which, at each step, the estimate of benefits from an additional source is added. Each time an
estimate of dollar benefits from a new source (e.g., a new health endpoint) is added to the
previous estimate of total dollar benefits, the estimated total dollar benefits increases. The
uncertainty surrounding the estimate of total dollar benefits, however, also increases.
As an example, consider the benefits from reductions in ozone-related hospital admissions
for congestive heart failure. Because we don't know what this dollar value actually is, it may be
considered a random variable, with a distribution describing the possible values h might have. If
we denote this random variable as X, , then a good point estimate of the true value of benefits
from reductions in ozone-related hospital admissions for congestive heart Mure is the mean of
the distribution, E(X,). The Var(X,), and the 5th and 95th percentile points of the distribution
(related to the Var(X,)), are ways to describe the uncertainty surrounding the estimate.
Now suppose we add the benefits from reductions in ozone-related hospital admissions for
ischemic heart disease. Like the benefits from reductions in ozone-related hospital admissions for
congestive heart failure, this is a random variable, with a distribution. Denoting this random
variable as Xj, the benefits from reductions in the incidence of both types of hospital admissions is
X, + Xj . This random variable has a distribution with mean E(Xi + Xj) and, if Xt and X^ are
stochastically independent, a variance of VarQC, + Xj) = Varpt,) +
The benefits from reductions in all non-overlapping ozone-related health and welfare
endpoints (X,, X* ..., XJ and from the ancillary reductions in all non-overlapping PM-related
health and welfare endpoints (X^,, ..., XJ is X = X, + ... + XB The mean of the distribution of
total benefits, X, is
and the variance of the distribution of total benefits is
Abt Associates Inc. 52 Juty 16,1997
-------
If all the means are positive, then each additional source of benefits increases the point estimate of
total benefits. However, with the addition of each new source of benefits, the variance of the
estimate of total benefits also increases. That is,
but
...< Var(Xl - ... + JQ = Var(X)
This means that, as we add sources of benefits, we have larger and larger point estimates of total
benefits (as more and more sources of benefits are included in the total) about which we are less
and less certain. This phenomenon occurs whenever we add estimates of benefits — for example,
adding estimates of ozone benefits from different sources to get an estimate of total ozone-related
benefits, or adding an estimate of total ozone-related benefits to an estimate of total PM-related
benefits to get an estimate of total benefits.
Whether the 5th percentile point of the distribution of X = X, + ... + X, is less than or
greater than the 5th percentile point of the distribution of (X, + ... + X,,.,) will depend on how
much E0C.) adds to E(X) relative to how much VarpQ adds to Var(X). There is nothing
unusual about the 5th percentile point of X = Xj + ... + X, being lower than the 5th percentile
point of pC, + ... + X^i). There will always be more uncertainty about total benefits, X, than
about any subset of X, such as (X, + ... + X,,.,). It is therefore not unreasonable, for example,
that, while the point estimate of total benefits is higher than the point estimate of some subset of
total benefits (e.g., PM-related benefits), the 5th percentile point of the distribution of total
benefits is less than the 5th percentile point of the distribution of the subset of total benefits - if
the addition to the variance of the distribution of total benefits is large while the addition to the
mean of the distribution of total benefits is small.
A clear way to convey this pattern of an increasing point estimate of total benefits as well
as increasing variance of the distribution of total benefits, as the benefits from additional sources
are included in the total, would be to present the mean, 5th percentile point, and 95th percentile
point of successive subtotals of benefits until the final total is reached. For example, Exhibit 17
could be filled in, showing how the estimate of total benefits increases as new sources of benefits
are included, but how the uncertainty of the estimate of total benefits also increases.
Abi Associates Inc. 53 Juty 16,1997
-------
Exhibit 17: Estimates of Dollar Benefits From Successively Larger Subsets of Benefit Sources
Source
1
2
3
1+2+3
4
S
6
4+5+6
PM-related hospital admissions
PM-related respiratory symptoms
PM-related mortality
Total PMnrelated benefits:
Ozone-related hospital admissions
Ozone-related respiratory symptoms
Ozone-related mortality
Total Ozone-related benefits:
Total Benefits
Estimate of Benefit* from Source:
Mean (SthVoOe, 95th %0e)
1
2
3
1+2+3
4
5
6
4+5+6
Cumulative Eitimatet of Total Benefits from
Increasing Sublets of Sources:
Mean (5th%iK95th %0e)
1
1+2
1+2+3
1+2+3+4
1+2+3+4+5
1+2+3+4+5+6
1+2+3+4+5+6
Abt Associates Inc.
54
Jitfy 16,1997
-------
5.0 REFERENCES
Abbey, D.E, Petersen, F., Mills. PJC., and W.L. Beeson. 1993. Long-term Ambient
Concentrations of Total Suspended Particulates, Ozone, and Sulfur Dioxide and
Respiratory Symptoms in a Nonsmoking Population. Archives of Environmental Health
48(1): 33-46.
Abbey, D. B., Lebowitz, M.D., Mills, P.K., Petersen, F.F., Beeson, W.L., Buichette, R.J.
1995. Long-term ambient concentrations of paiticulates and oxidants and development
of chronic disease in a cohort of nonsmoking California residents. In: Phalen, R.F.,
Bates, D. V., eds. Proceedings of the colloquium on particulate air pollution and
human mortality and morbidity; January 1994; Irvine, CA. Inhalation ToxicoL 7: 19-
34.
Abt Associates, Inc. 1992. The Medical Costs of Five Illnesses Related to Exposure to
Pollutants. Prepared for U.S. EPA, Office of Pollution Prevention and Toxics,
Washington, DC.
Alberini, A., M.L. Cropper, T. Fu, AJ. Knipnick, J. Liu, D. Shaw, and W. Harrington. 1994.
Valuing Health Effects of Air Pollution in Developing Countries: The Case of Taiwan.
Discussion Paper 95-01, Resources for the Future, Washington, DC, October.
Anderson et al., 1996. Air Pollution and Daily Mortality in London: 1987 - 92. BMJ 312: 665-
669.
Chestnut, L.G. 1995. Dollars and Cents: The Economic and Health Benefits of 'Potential
Particulate Matter Reductions in the United States. Prepared for the American Lung
Association.
Chestnut, L. 1997. Draft Memorandum: Methodology far Estimating Values for Changes in
Visibility at National Parks. (April 15)
Chestnut, L. and R. Rowe. 1989. "Economic Valuation of Changes in Visibility: A State of the
Science Assessment for NAPAP," as cited in National Acid Precipitation Assessment
Program, Methods for Valuing Addic Deposition and Air Pollution Effects, NAPAP
State of Science and State of Technology Report No. 27, Part B. December.
Cifuentes, L. and L.B. Lave. 1996. Association of Daily Mortality and Air Pollution in
Philadelphia, 1983-1988. J. Air Waste Manage. Assoc.: in press.
Crocker T.D., Horst RI*, Jr. 1981. Hours of Work, Labor Productivity, and Environmental
Conditions: a Case Study. The Review of Economics and Statistics 63:361-368.
Abt Associates Inc. 55 July 16.1997
-------
Cropper, M.L. and A.J. Krupnidc 1990. The Social Cost* of Chronic Heart and Lung Disease.
Resources for the Future Discussion Paper QE 89-16-REV.
Cummings, R., H. Bumess and R. Norton. 1981. Methods Development for Environmental
Control Benefits Assessment, Volume V. Measuring Household Soiling Damages from
Suspended Air Particulates, A Methodological Inquiry. Report prepared for the U.S.
Environmental Protection Agency, Washington, D.C.
Dickie, M. et al. 1987. Reconciling Averting Behavior and Contingent Valuation Benefit
Estimates of Reducing Symptoms of Ozone Exposure (draft), as cited in Neumann, J.E.,
Dickie, M.T., and R.E. Unsworth. 1994. Industrial Economics, Incorporated.
Memorandum to Tun DeMocker, U.S. EPA, Office of Air and Radiation. March 31.
Dockery, D.W., F.E. Speizer, D.O. Stram, J.H. Ware, J.D. Spengjer, and B.G. Ferris, Jr. 1989.
Effects of Inhalable Particles on Respiratory Health of Children. Am. Rev. Respir. Dis.
139: 587-594.
Elixhauser, A., R.M. Andrews, and S. Fox. 1993. Clinical Classifications for Health Policy
Research: Discharge Statistics by Principal Diagnosis and Procedure. Agency for
Health Care Policy and Research (AHCPR), Center for General Health Services
Intramural Research, U.S. Department of Health and Human Services.
Empire State Electric Energy Research Corporation (ESEERCO). 1994. New York State
Environmental Externalities Cost Study. Report 2: Methodology. Prepared by:
RCG/Hagler, Bailly, Inc., November.
Ehrlich, I., and H. Chuma. 1990. A Model of the Demand for Longevity and the Value of Life
Extension. Journal of Political Economy 98(4): 761-782.
Gegax, D., S. Go-king and W. Shulze. 1985. Perceived Risk and the Marginal Value of Safety.
Working paper prepared for the U.S. Environmental Protection Agency.
Gerking, S., M. DeHaan, and W. Schulze. 1988. The Marginal Value of Job Safety: A
Contingent Valuation Study. Journal of 'Risk and Uncertainty 1: 185-199.
Hoek et al., 1997. Effects of Ambient Paniculate Matter and Ozone on Daily Mortality in
Rotterdam, the Netherlands. Draft, submitted to Archives of Environmental Health.
Industrial Economics, Incorporated (lEc). 1992. Memorandum to Tim DeMocker, Office of Air
and Radiation, Office of Policy Analysis and Review, U.S. Environmental Protection
Agency. November 6.
Abl Associates Inc. 56 July 16. J997
-------
Industrial Economics, Incorporated (ffic). 1993. Memorandum to Jim DeMocker, Office of Air
and Radiation, Office of Policy Analysis and Review, U.S. Environmental Protection
Agency, September 30,1993.
Industrial Economics, Incorporated (ffic). 1994. Memorandum to Tun DeMocker, Office of Air
and Radiation, Office of Policy Analysis and Review, U.S. Environmental Protection
Agency, March 31.
Irwin, J., W.Schulze, G. McClelland, D. Waldman, D. Schenk, T. Stewart, L. Deck, P. Slovic, S.
Lictenstein, and M. Thayer. 1990. Valuing Visibility: A Field Test of ike Contingent
Valuation Method Prepared for Office of Policy, Planning and Evaluation, U.S.
Environmental Protection Agency. March.
Ito, K. and Thurston, G.D., 1996. Daily PMlO/Mortality Associations: An Investigation of At-
Risk Subpopulations. Journal of Exposure Analysis and Environmental Epidemiology
6(1): 79-225.
Jones-Lee, M.W., et al. 198S. The Value of Safety: Result of a National Sample Survey.
Economic Journal 95 (March): 49-72.
Kinney et al., 1995. A Sensitivity Analysis of Mortal!ty/PM 10 Associations in Los Angeles.
Inhalation Toxicology 7: 59-69.
Krupnick, A. J. and M.L. Cropper. 1992. The Effect of Information on Health Risk Valuations.
Journal of Risk and Uncertainty 5(2): 29-48.
Krupnick, A. J. and R.J. Kopp. 1988. The Health and Agricultural Benefits of Reductions in
Ambient Ozone in the United States. Resources for the Future Discussion Paper QE88-
10, Washington, DC. August.
Krupnick A.J., Harrington W., OstroB. 1990. Ambient Ozone and Acute Health Effects:
Evidence from Daily Data. Journal of Environmental Economics and Management 18:1-
18.
Loehman, E.T., S.V. Berg, AA Arroyo, R.A. Hedinger, J.M. Schwartz, ME. Shaw, R.W.
Fahien, VH. De, R.P. Fishe, D.E. Rio, W.F. Rossley, and A£.S. Green. 1979.
Distributional Analysis of Regional Benefits and Cost of Air Quality Control Journal of
Environmental Economics and Management 6:222-243.
Loehman, E.T. and Vo Hu De. 1982. "Application of Stochastic Choice Modeling to Policy
Analysis of Public Goods: A Case Study of Air Quality Improvements. The Review of
Economics and Statistics 64(3): 474-480.
Abt Associates Inc. 57 July 16.1997
-------
Loomis et al., 1996. Ozone Exposure and Daily Mortality in Mexico City: A Time-Series
Analysis. Health Effects Institute Research Report Number 75, October 1996.
Manuel, E.H., RX. Horst, K.M Brennan, W.N. Lanen, M.C. Duff and JJC. Tapiero. 1982.
Benefits Analysis of Alternative Secondary National Ambient Air Quality Standards for
Sulfur Dioxide and Total Suspended Particulars, Volumes I-IV. Prepared for U.S.
Environmental Protection Agency, Office of Air Quality Planning and Standards, Research
Triangle Park, NC. [Crted in ESEERCO, 1994].
McClelland, G., W. Schulze, D. Waldman, J. Irwin, D. Schenk, T. Stewart, L. Deck, and M.
Thayer. 1991. Valuing Eastern Visibility: A Field Test of the Contingent Valuation
Method. Prepared for Office of Policy, Planning and Evaluation, U.S. Environmental
Protection Agency. June.
Mitchell, R.C., and R.T. Carson. 1986. The Use of Contingent Valuation Data for Benefit/Cost
Analysis in Water Pollution Control. Draft report submitted by Resources for the Future
to Office of Policy Analysis, U.S. Environmental Protection Agency. September
Moolgavkar et al., 1995. Air Pollution and Daily Mortality in Philadelphia. Epidemiology 6(5):
476-484.
Moore, M.J., and W.K. Viscusi. 1988. "The Quantity-Adjusted Value of Life". Economic
Inquiry 26(3): 369-388.
Ostro, B.D. 1987. Air Pollution and Morbidity Revisited: a Specification Test J. Environ.
Econ. Manage. 14: 87-98.
Ostro, 1995. Fine Particulate Air Pollution and Mortality in Two Southern California Counties.
Environmental Research 1Q\ 98-104.
Ostro B.D. and S. Rothschild. 1989. Air Pollution and Acute Respiratory Morbidity: An
Observational Study of Multiple Pollutants. Environmental Research 50:238-247.
Ostro, B.D., M.J. Lipsett, M.B. Wiener, and J.C. Seiner. 1991. Asthmatic Responses to
Airborne Acid Aerosols. American Journal of Public Health 81:694-702.
Ostro, B.D., M. J. Lipsett, JJC. Mann, H. Bnxton-Owens, and M.C. White. 1995. Air Pollution
and Asthma Exacerbations Among African American Children in Los Angeles. Inhalation
Toxicology,
Ostro et al., 1996. Air Pollution and Mortality: Results from a Study of Santiago, Chile. J. Of
Exposure Analysis and Environmental Epidemiology 6:97-114.
Abt Associates Inc. 58 Jufy 16,1997
-------
Pitchford, M.L. and W.C. Malm. 1994. Development and Applications of a Standard Visual
Index. Atmospheric Environment 29(5): 1049-1054.
Pope, C.A., m, D.W. Dockexy, J.D. Spengler, and ME. Raizenne. 1991. Respiratory Health
and PM|0 Pollution: a Daily Time Series Analysis. Am. Rev. Respir. Dis. 144:668-674.
Pope, C.A., m, M.J. Thun, MM. Namboodiri, D.W. Dockery, IS. Evans, F.E. Speizer, and
C.W. Heath, Jr. 199S. Particulate Air Pollution as a Predictor of Mortality in a
Prospective Study of U.S. Adults. Am. J. Respir. Crit CareMed 151: 669-674.
Portney, P.R. and J. Mullahy. 1990. Urban Air Quality and Chronic Respiratory Disease.
Regional Science and Urban Economics 20: 407-418.
Rowe, R.D. and L.G. Chestnut. 1986. Oxidants and Asthmatics in Los Angeles: A Benefits
Analysis—Executive Summary. Prepared by Energy and Resource Consultants, Inc.
Report to the U.S. EPA, Office of Policy Analysis. EPA-230-09-86-018. Washington,
D.C. March
Rowe, R.D. and T.N. Nehhercut. 1987. Economic Assessment of the Impacts of Cataracts.
Prepared for U.S. Environmental Protection Agency, Benefits Branch-OPPE by Energy
and resource Consultants, Inc., Boulder, Colorado.
Rowe, RJX, L.G. Chestnut, and W.D. Shaw. 1984. Oxidants and Asthmatics in Los Angeles: A
Benefits Analysis. Evaluation of the Ozone/Oxidants Standards. Si Duk Lee (ed.) Air
Pollution Control Association. Houston, pp. 366-392.
Samet et al., 1996. Air Pollution and Mortality in Philadelphia, 1974-1988. Report to the
Health Effects Institute on Phase IB: Particle Epidemiology Evaluation Project, March 25,
1996 (draft, accepted for publication).
Samet et al., 1997. Particulate Air Pollution and Daily Mortality: Analysis of the Effects of
Weather and Multiple Air Pollutants. The Phase IB Report of the Particle Epidemiology
Evaluation Project. Health Effects Institute, March 1997.
Schwartz, J. 1993. Particulate Air Pollution and Chronic Respiratory Disease. Environmental
Research 62: 7-13.
Schwartz,!. 1994a. Air Pollution and Hospital Admissions in Elderly Patients in Birmingham,
Alabama. American Journal of Epidemiology 139:589-98.
Schwartz,!. 1994b. Air Pollution and Hospital Admissions for the Elderly in Detroit, Michigan.
American Journal of Respiratory Care Mtd 150:648-55.
Abt Associates Inc. 59 July 16.1997
-------
Schwartz,! 1994c. PMia, Ozone and Hospital Admissions for the Elderly in Minneapolis-St
Paul, Minnesota. Archives of Environmental Health 49(5): 366-374.
Schwartz, J. 1995. Short Term Fluctuations in Air Pollution and Hospital Admissions of the
Elderly for Respiratory Disease. Thorax 50:531-538.
Schwartz,!. 1996. Air Pollution and Hospital Admissions for Respiratory Disease.
Epidemiology 7(1): 20-28.
Schwartz, J., and D.W. Dockery. 1992. Increased mortality in Philadelphia associated with daily
air pollution concentrations. Am. Rev. Respir. Dis. 145: 600-604.
Schwartz, J. and R. Morris. 1995. Air Pollution and Hospital Admissions for Cardiovascular
Disease in Detroit, Michigan. Am. J. EpidemioL 142:23-35.
Schwartz, J. D. W. Dockery, L.M. Neas, D. Wypij, J.H. Ware, J.D. Spengler, P. Koutrakis,
F.E.Speizer, and E.G. Ferris, Jr. 1994. Acute Effects of Summer Air Pollution on
Respiratory Symptom Reporting in Children. Am. J. Respir. Crit. CareMed. 150: 1234-
1242.
Schwartz, J., Dockery, D., and L. Neas. 1996. Is Daily Mortality Specifically Associated With
Fine Particles? J. Air & Waste Man. Assoc. 46:927-939.
Smith, K.V. 1983. The Role of She and Job Characteristics in Hedonic Wage Models. Journal
of Urban Economics 13:296.321.
Thurston, G.D. K. Ito, P.L. Kinneym, and M. Ltppman. 1992. A Multi-Year Study of Air
Pollution and Respiratory Hospital Admissions in Three New York State Metropolitan
Areas: Results for 1988 and 1989 Summers. Journal of Exposure Analysis and
Environmental Epidemiology. 2 (4):429-450.
Thurston, G. K. Ito, C. Hayes, D. Bates, and M. Lippmann. 1994. Respiratory Hospital
Admission and Summertime Haze Air Pollution in Toronto, Ontario: Consideration of the
Role of Acid Aerosols. Environmental Research 65:271-290.
Tolley, G.S. et al. 1986. Valuation of Reductions in Human Health Symptoms and Risks.
University of Chicago. Final Report for the U.S. Environmental Protection Agency.
January.
Touloumi et aL, 1997. Short Term Effects of Ambient Oxidants Exposure on Mortality: A
Combined Analysis Within the APHEA Project American Journal of Epidemiology (in
press).
Abt Associates Inc. 60 July 16. J997
-------
U.S. Department of Commerce, Economics and Statistics Administration. 1992. Statistical
Abstract of the United States, 1992: The National Data Book. 112th Edition,
Washington, D.C.
U.S. Department of Commerce, Bureau of Economic Analysis. BEA Regional Projections to
2045: Vol. 1. States. Washington, D.C: U.S. Govt. Printing Office, July 199S.
U.S. Environmental Protection Agency (U.S. EPA). 1993. External Draft, Air Quality Criteria
for Ozone and Related Photochemical Oxidants. Volume n. Office of Health and
Environmental Assessment, Environmental Criteria and Assessment Office, Research
Triangle Park, NC; EPA/600/AP-93/004b. 3v.
U.S. Environmental Protection Agency, 1994. Documentation for Oz-One Computer Model
(Version 2.0). Office of Air Quality Planning and Standards. Prepared by: Mathtech,
Inc., under EPA Contract No. 68D30030, WA1-29. August.
U.S. Environmental Protection Agency, 1996a, Regulatory Impact Analysis for Proposed
Particulate Matter National Ambient Air Quality Standard. Prepared by: Innovative
Strategies and Economics Group, Office of Air Quality Planning and Standards, U.S.
EPA. Research Triangle Park, N.C. December.
U.S. Environmental Protection Agency, 1996b. Regulatory Impact Analysis for Proposed Ozone
National Ambient Air Quality Standard. Prepared by: Innovative Strategies and
Economics Group, Office of Air Quality Planning and Standards, U.S. EPA. Research
Triangle Park, N.C. December.
U.S. Environmental Protection Agency, 1997. The Benefits and Costs of the Clean Air Act, 1970
to 1990. Prepared for U.S. Congress by U.S. EPA, Office of Air and Radiation/Office of
Policy Analysis and Review, Washington, D.C. (April, 1997 - Draft)
Verhoeff; A.P. et al., 1996. Air Pollution and Daily Mortality in Amsterdam. Epidemiology
7(3): 225-230.
Violette, D.M. and L.G. Chestnut. 1983. Valuing Reduction in Risks: A Review of the Empirical
Estimates. Report prepared for the U.S. Environmental Protection Agency, Washington,
D.C. EPA-230-05-83-002.
Viscusi,WJC. 1992. Fatal Tradeoffs: Public and Private Responsibilities for Risk. (New York:
Oxford University Press).
Viscusi, W.K., Magat, W.A., and Huber, J. 1991. Pricing Environmental Health Risks: Survey
Assessments of Risk-Risk and Risk-dollar Tradeoffs Journal of Environmental
Economics and Management 201:32-S7.
Abt Associates Inc. 61 July 16,1997
-------
Watson, W. and J. Jaksch. 1982. Air Pollution: Household Soiling and Consumer Welfare
Losses. Journal of Environmental Economics and Management. 9:248-262.
Weisel, CP., R.P. Cody, and P.J. Uoy. 199S. Relationship Between Summertime Ambient
Ozone Levels and emergency Department Visits for Asthma in Central New Jersey.
Environmental Health Perspectives. 103(Suppl. 7): 97-102.
Whittemore AS, Korn EL. 1980. Asthma and Air Pollution in the Los Angeles Area. American
Journal of Public Health 70:687-696.
Abt Associates Inc. 62 Juty 16,1997
-------
Appendix: The Relationship Between Mean WTP and Mean Income Venus the
Relationship Between Individual WTP and Individual Income
The elasticity between x and y can be shown to be the linear relationship between log x
and log y. That is, if b is the elasticity of y with respect to x, then
logy = a + blogx
for some a (assuming, for simplicity of discussion, that there is a constant elasticity). In
particular, if b is the elasticity of WTP for health-related environmental improvements with
respect to real income, then
log WTP = a + b log (income).
This implies that, unless b=l, the relationship between WTP and real income is nonlinear:
WTP = a(mcome)b , where a=ea
If 0=1, then the relationship between WTP and income is linear, and the relationship between
mean WTP, E[WTP], and mean (real) income, Efincome], is likewise linear:
E[WTP] = o Efincome].
Li this case, changes in mean income easily translate into changes in mean WTP. If mean income
in the year 2007 is 15% higher than in 1995, then
£[B7P]2007 = oE[iwcome]afl07 = al.!5£[i>icowe],W5 = l.lSE[W7P]l99i
If; however, b*l, then
E[WTP] = E[a(inoome)b] * a(E[income])b , for b\.
(Jensen's Inequality). Jensen's Inequality says that if Y=f(X) and f() is a concave function, then
E(Y] i f(E[X]). Iffit)isaconvexfiinction,thenE(Y]if(Epq). In this case, Y is WTP and X
is income, and the function relating WTP to income is concave ifb < 1 and convex if b>l. (Only
if b=l is the relationship between WTP and income linear.)
Abt Associates Inc. 63 July 16.1997
-------
Air Toxics Benefits of Options for Revision of the
National Ambient Air Quality Standard for Ozone
July 1997
Office of Policy, Planning and Evaluation
U.S. Environmental Protection Agency
-------
1. Introduction
Reductions in concentrations of ozone are achieved by reducing emissions of volatile
organic compounds (VOCs) and/or reducing emissions of oxides of nitrogen (NOx).
Many volatile organic compounds are substances which are listed as hazardous air
pollutants (HAPs) tinder section 112 of the Clean Air Act HAPs, also known as "air
toxics," are associated with a variety of adverse human health effects such as cancer,
reproductive and developmental effects, and neurological disorders, as well as adverse
ecological effects. This technical document presents the methodology and results of an
analysis of die air toxics benefits that would be expected from implementation of three
options for a revised National Ambient Air Quality Standard (NAAQS) for ozone:
• 8-hour average 0.08 parts per million (ppm), 5th maximum;
• 8-hour average 0.08 ppm, 4th maximum; and
• 8-hour average 0.08 ppm, 3rd maximum.
This analysis is focused entirely on estimation of benefits associated with avoided future
incidence of cancer due to reductions in emissions of carcinogenic VOC HAPs which are
adopted in response to a revised NAAQS. Only benefits from reductions which are
incremental to the current ozone NAAQS and other Clean Air Act requirements are
counted. The analysis addresses cancer benefits and does not address a variety of other
human health benefits and ecological benefits from reduced emissions of HAPs because
the available methods and data which support quantification of cancer benefits are not
generally available for the numerous other categories of benefits. In addition, the focus
of this analysis is on those VOC reductions associated with "partial attainment"
scenarios.4
This analysis of air toxics benefits makes use of HAP emissions estimates and HAP
concentration estimates developed as part of EPA's Cumulative Exposure Project1. This
modeling study estimates long-term average concentrations of 152 HAPs for each census
tract in the continental United States. HAP concentrations are estimated through
development of a comprehensive inventory of HAP emissions from stationary and mobile
sources, and simulation of dispersion of these emissions and other atmospheric processes
(e.g. reactivity, secondary formation, deposition) using an enhanced version of EPA's
Human Exposure Model (HEM).
1 "Partial attainment" scenarios include only those emissions reductions for which specific control
strategies could be identified. "Full attainment" scenarios represent all emissions reductions needed to
reach attainment with a particular standard, including all reductions represented by partial attainment plus
the additional reductions for which control strategies have not been identified.
1 The methodology for modeling HAP concentrations is presented in Rosenbaum et al. 1996. This
methodology has been reviewed by EPA's Science Advisory Board (Science Advisory Board 1996)
-------
2. EPA Methodology for Cancer Assessment
After considering the epidemiologic and experimental information pertinent to the question
of whether a pollutant poses a cancer hazard, EPA's methodology for quantifying estimated
carcinogenic risks from exposure to environmental pollutants is built on a quantitative dose-
response estimate known as a cancer slope factor (sometimes called a cancer potency
factor). The cancer slope factor is an extrapolation of the dose-response curve observed in
epidemiologic or experimental studies to the lower dose levels that are expected to be
typical of environmental exposures. Because mere are generally no toxicity data to define
dose-response curves in the range of environmental exposures, the extrapolation typically
assumes a linear shape below the observed range.
A central (or "best") estimate of cancer risk would be preferred, but in reality, this is not a
simple issue. "Best" estimates, from a statistical perspective, tend to be unstable; that is,
minor changes in the underlying data would translate into major changes in the estimated
risk. It is, therefore, highly misleading to characterize a single estimate as a "best" estimate
of risk, particularly when that estimate can be highly uncertain. In addition, a "best"
estimate would be a misleading estimate of the range of risks for individuals or populations.
For these reasons, the cancer slope factor is expressed as a plausible upper bound. That
means that these estimates should be interpreted as a plausible representation (reflecting
experimental uncertainty) of how high the cancer risk may be, but that the true risk may
also be lower. In this light, EPA's risk assessments are not meant to be estimates of true
risk, thus they are not comparable to actuarial risk statements such as annual highway
accident statistics.
Cancer slope factors can be multiplied by central estimates of anticipated human exposure
to give a plausible upper bound on an exposed individual's increased cancer risk from this
exposure. Matching the upper bound slope factor with a central estimate of exposure avoids
part of the problem of compounding conservatism that would result from multiplying upper
bounds by upper bounds. Units of exposure estimates must be matched to the units of the
cancer slope factor. For example, the cancer slope factor represents the incremental cancer
risk from a lifetime average exposure of 1 mg/kg-d, so the exposure estimate should
represent the total exposure averaged over a lifetime. Because ambient air exposures are
often expressed in units of micrograms per cubic meter (ug/m3), EPA's assessments provide
a quantity called an inhalation unit fjgfc which is an analogous quantity to the cancer slope
factor, except mat it is expressed in units of lifetime average ug/m3, assuming a set of
standard exposure assumptions. The lifetime cancer risk to a population exposed to a
carcinogenic air pollutant is therefore estimated as the product of 1) me iqhalation unit risk,
2) the lifetime average exposure to the population, and 3) the size of the population
exposed. This product yields an estimate of the number of lifetime cancer cases in the
population resulting from exposure to mat pollutant
-------
3. Selection of Pollutants for this Analysis
The analysis of cancer benefits focuses on three particular carcinogenic VOC HAPs:
benzene, 1,3-butadiene, and formaldehyde. These pollutants were selected through a
screening analysis which found that they accounted for approximately 90 percent of the
potential for ***'«*«"£ carcinogenic benefits from reducing emissions of VOC HAPs.
This screening analysis was conducted as follows:
• The set of pollutants which were candidates fbr.this benefits analysis was defined as:
organic HAPs classified by EPA as "known" or "probable" human carcinogens with
an inhalation unit risk (IUR) value reported in EPA's Integrated Risk Information
System (IRIS). Nineteen HAPs met these criteria and were included in the screening
analysis.
• A cancer potential index was constructed by multiplying the estimated national mean
outdoor concentration for each pollutant by its IUR. Multiplying by the IUR accounts
for the varying potency, or carcinogenic risk, of each of the pollutants considered.
Estimates of the national mean outdoor concentration for each pollutant were derived
from modeled air toxics concentrations developed as part of EPA's Cumulative
Exposure Project9
• The total cancer potential index of the candidate HAPs- was derived by summing the
lUR-weighted mean concentrations across all of the candidate HAPs. The
contribution of each individual HAP to the total was calculated by deriving the HAP's
lUR-weighted mean concentration by the total across all HAPs. As shown in Table 1,
1,3-butadiene was found to account for 55 percent of the total cancer potential index,
while formaldehyde accounted for 21 percent and benzene accounted for 17 percent.
None of the remaining candidate HAPs accounted for more than 2 percent of the total
cancer potential index.
Benzene is classified by EPA as a known human carcinogen; 1,3-butadiene and
formaldehyde are classified as probable human carcinogens.
1 For this analysis, only pollutant concentrations arising from current pollutant emissions are considered;
"background" concentrations, representing non-anthropogenic sources, persistence in the environment of
historical emissions, and other factors, are disregarded since they will not be affected substantially by
implementation of a revised ozone NAAQS.
-------
4. Methodology for Cancer Benefits Estimation
The method for estimating the partial attainment cancer benefits of the options for ozone
NAAQS revisions consists of three main steps:
• Estimate the changes in average outdoor concentrations of benzene, 1,3-butadiene and
formaldehyde for each state in the year 2010 due to partial attainment of the specified
option
• Estimate the changes in cancer risks in 201Q, using the projected changes in outdoor
concentrations for each pollutant in each state, projected 2010 population for each
state, and the IUR for each pollutant
•
• Multiply the avoided cases of cancer by the estimated value of reduced mortality
risks.
This method makes use of the fact mat the methodology for estimating cancer cases,
described above in Section 2, is based on a linear dose-response model which uses the
lifetime average daily dose for any specified population to estimate the total cancer risk
for that population. By this methodology, the change in the average concentration of a
carcinogen to which individuals in a state are exposed can be used to estimate the change
in total carcinogenic risk to the state's entire population. The actual changes in exposure
and risk may vary substantially across individuals within the state, but for this analysis of
national benefits these variations do not affect the aggregate estimate of benefits, and may
therefore be disregarded. Each of the three major steps in the methodology are discussed
in the following sections.
4. 1 Changes in HAP concentrations
To estimate changes in concentrations of the three target HAPs, three main data elements
are used: baseline population-weighted average concentrations of the HAPs for each
state; baseline emissions of HAPs for each state; and changes in emissions of HAPs, due
to partial attainment of a revised ozone NAAQS, for each state in the year 201.0.
The change in HAP concentrations in 2010 is calculated as:
where
= change in concentration (ug/mj) of HAP H in state S resulting from
partial attainment of a revised ozone NAAQS in year 2010
-------
= change in emissions (tons/day) of HAP H in state S resulting from partial
attainment of a revised ozone NAAQS
£5 H baseline emissions (tons/day) of HAP H in state S
GSJI - baseline concentration (ug/mj) of HAP H in stateS, net of background
concentration (population-weighted long-term average).
•
This approach assumes mat any reduction in emissions of a target HAP will translate into
a proportional reduction in the state's population-weighted average concentration of that
HAP. This assumption is discussed further in Section 5 below.
Baseline emissions (E) and concentrations (C) are derived from data developed for the
Cumulative Exposure Project E is equal to the total estimated emissions of each HAP in
each state in 1990, while C is a population-weighted average pollutant concentration for
each state, derived from dispersion modeling. Modeling was conducted at the census
tract level; statewide average HAP concentrations are calculated by weighting the
modeled concentration for each tract in a state with the number of residents in that tract,
as reported in the 1990 census4.
Change in HAP emissions (AE) are estimated by speciation of the VOC emissions
reductions developed for modeling of ozone attainment scenarios and cost estimates
presented in the NAAQS RIA. For each VOC control measure, estimates were made of
the percentage of the VOC reductions represented by the target HAPs. For example,
benzene is estimated to account for 2.66 percent of the projected VOC emissions
reductions from service station underground tanks. In each state where projected control
measures include VOC reductions from this source category, the amount of VOC
reduction is multiplied by 2.66 percent to estimate the benzene emissions reductions
achieved. Similar calculations are made for each control measure affecting emissions of
any of the target HAPs, and the HAP emissions reductions are summed across control
measures to estimate total statewide reductions in emissions of each target HAP.
Speciation factors, representing the percentage of the target HAPs in the VOC emissions
stream for each affected source category, were obtained from EPA's SPECIATE database
and from the technical literature (Wilson 1997).
A slightly modified approach is used for formaldehyde. Modeled pollutant
concentrations for this HAP consist of two components: primary formaldehyde, which
results from emissions of formaldehyde; and secondary formaldehyde, which results from
atmospheric reactions of several other emitted volatile organic compounds, referred to in
this analysis as "formaldehyde precursors.'* The approach for estimating changes in
concentrations of formaldehyde, using the approach described above, similarly consists of
two components: one which is based on data for primary formaldehyde emissions and
4 Data are available for each of the 48 continental states and the District of Columbia. Alaska and Hawaii
were not included in the toxics modeling study due to the unavailability of needed emissions data.
-------
primary formaldehyde concentrations; and a second which implements the same approach
using data on formaldehyde precursor emissions and resulting secondary formaldehyde
concentrations. In other words, AC^ is evaluated for both H = primary formaldehyde
and H = secondary formaldehyde; total results for formaldehyde are determined by
summing together the results for primary and secondary formaldehyde.
Only the VOC reductions estimated for partial attainment scenarios were included in this
analysis. Ah- toxics benefits cannot be directly estimated for the additional VOC
reductions included in full attainment scenarios; since specific measures to achieve these
additional VOC reductions can not be identified, it is difficult to directly estimate the
portion of these VOC reductions which will be accounted for by the three target HAPs.
Potential benefits of full attainment are further discussed in Appendix 4.
4.2 Estimating changes in cancer risks
Changes in incidence of cancer attributable to revisions in the ozone NAAQS are
estimated as: , •
POP
49 J
£ £
S-l H-l
IPA " £ £ IS.H
where
ISH = reduced incidence of cancer cases from HAP H in state S resulting from
partial attainment of a revised ozone NAAQS in year 2010
IUR,, = inhalation unit risk for HAP H
POPS = population in state Sin year 2010
IPA = total reduced incidence of cancer cases for all target HAPs in all states
resulting from partial attainment of revised ozone NAAQS in year 2010.
The IUR is an estimate of an individual's probability of cancer when exposed to a
pollutant concentration of one microgram per cubic meter of air (ug/m*) for a 70-year
lifetime. IUR values for the three target HAPs, obtained from EPA's Integrated Risk
Information System (IRIS), are:
benzene 8.3 x 10* per ug/m1
1,3-butadiene 2.8 x 10*4 per ug/m3
formaldehyde 1.3 x 10"5 per ug/m1.
-------
For this analysis, an adjustment to the IUR was necessary. This analysis addresses the
costs and benefits of a revised ozone NAAQS for the year 2010 only. As a result,
benefits should be calculated based on reduced pollutant exposures in only this one year.
The IUR, however, estimates the risk attributable to 70 years of exposure. Because of the
linear form of estimated dose-response relationship, the benefits of reduced exposure for
one year only can be estimated simply by dividing the IUR by 70.
Because the IUR is an estimate of risk to an individual, it must be multiplied by the size
of the population exposed (POP,) to estimate cancer incidence for that population.
Estimates of the population for each state in the year 2010 used to estimate air toxics
benefits are the same as those estimates used for ozone benefits modeling.
4.3 Valuation of avoided mortality
The value of a statistical life used to monetize avoided cancer cases is the same as
used elsewhere in the analysis of ozone NAAQS benefits, estimated as a Weibull
distribution with a mean of $4.8 million and standard deviation of $3.24 million. Three
values are used to illustrate the results produced by this distribution: the mean ($4.8
million), the 5th percentile ($0.75 million) and the 95th percentile ($11 million).
5. Key Assumptions in Exposure Estimates
This section discusses three key assumptions of the methodology for estimating reduced
exposure to the three target HAPs resulting from a revised ozone NAAQS.
5.7 Emissions-concentration relationship
Changes in 2010 HAP concentrations are derived from changes in 2010 HAP emissions
by using an estimated relationship between HAP emissions and HAP concentrations for
each state using data from the air toxics modeling study (Rosenbaum and Wei 1997).
The emissions-concentration relationship may understate the impact of emissions
reductions, due to limitations in the modeling methodology. Outputs from the modeling
exercise, representing estimated concentrations for 1990, were evaluated by comparing
modeled concentrations with available monitoring data. Measured air concentration data
for the three target HAPs were collected from a number of long-term monitoring
programs in operation in the early 1990s: 75 monitoring locations wereidentified for
benzene, 38 for 1,3-butadiene, and 34 for formaldehyde. For each monitor location, a
ratio was calculated of the modeled concentration divided by the measured concentration;
ratios less than 1.0 indicate that the modeled concentrations are underestimates, while
ratios greater than 1.0 indicate model overestimation. For each of the target pollutants,
70 percent or more of the monitor locations had ratios less than 1.0. For benzene, the
geometric mean of the ratios was 0.72, while for 1,3-butadiene it was 0.41 and for
formaldehyde it was 0.84.
-------
Although some of the underestimation of concentrations may be attributable to
underestimation of emissions, a substantial portion appears to be due to limitations of the
Gaussian dispersion modeling methodology. Among the model limitations likely to
contribute to underestimation are: inability to evaluate extreme meteorological events
such as stagnation (extended calm winds) which result in increased concentrations; poor
representation of stable atmospheric conditions that may occur at night; and limiting the
extent of dispersion modeling to SO kilometers (km) from the source, in accordance with
EPA recommendations for use of Gaussian models'.
If, as it appears, the tendency for underestimation of pollutant concentrations is at least
partly explained by dispersion modeling limitations, then the actual concentrations
resulting from given levels of emissions is understated. As a consequence, this method
will understate the impact of emissions reductions in reducing concentrations.
5.2 Aggregating the emissions-concentration relationship at the state level
This method assumes a fixed, linear relationship between emissions*and population-
weighted average concentrations at the state level. This fixed relationship is defined by
the baseline concentration and baseline emissions. The change in concentration is then
proportional to changes in HAP emissions; for example, if emissions of benzene in a
particular state decline by 2 percent, then the statewide population-weighted average
concentration is estimated to also decline by 2 percent6
The relationship between emissions and population-weighted outdoor concentrations is
influenced by a number of factors, including when the emissions occur, the weather at the
time the emissions occur, and where emissions occur. A given amount of emissions will
contribute more to the population-weighted average concentration if it occurs in a densely
populated area, or upwind of population, than if it occurs in a remote location, or
downwind. In addition, releases at ground level will generally result in a greater
population-weighted average concentration than elevated releases in the same location.
In the assessment of the impacts of any particular source, meteorological factors and
residential patterns could make a significant difference in whether there are large or small
changes in concentrations of pollutants to which individuals are exposed. For
implementation of a NAAQS, however, which is likely to require emissions reductions
from a large number of sources (including mobile sources) in a wide number of dispersed
locations, it is reasonable to assume that, in the aggregate, emissions reductions will
result in a proportional reduction in outdoor concentrations. While, in reality, the
1 The SO km distance limitation is likely to contribute to underestimation of concentrations all pollutants,
but is likely to be particularly significant for secondary formaldehyde, since formaldehyde precursors may
frequently be transported 50 km before transformation to formaldehyde occurs.
* For purposes of this analysis, background concentrations are excluded, since they will not be affected by
implementation of a revised NAAQS. This analysis is based on concentrations resulting from current
anthropogenic emissions only.
8
-------
relationship between emissions reductions and concentration reductions may not be
linear, this assumption is not likely to introduce any systematic bias to estimated
reductions in exposure concentrations.
5.3 Relationship between outdoor concentrations and exposures
This methodology also assumes that the population-weighted outdoor air concentration
for each HAP in each state is equal to the populatk)n-wd^ited exposure concentration.
While the calculation of cancer risks requires exposure concentrations, the modeling data
used in this analysis estimate outdoor concentrations. There are two ways in which the
assumed equivalence between outdoor concentrations and exposure concentrations might
be questioned: first, people generally spend the majority of their time indoors, and the
indoor concentration may not be equal to the outdoor concentration; and second, people
do not spend all of their time at home—much of their time is spent at work, at school, or
in other locations.
The relationship between indoor and outdoor concentrations of air pollutants has been the
subject of much research, including EPA's Total Exposure Assessment Methodology
(TEAM) studies, which have shown that for many individuals, indoor exposure to air
pollutants is much greater than outdoor exposure (Wallace 1987). Much of the difference
found between indoor and outdoor exposure is attributable to indoor sources, such as
cigarette smoking. Indoor sources of pollutants are not relevant to this analysis, however;
the key question for this analysis is the penetration of outdoor air pollutants into indoor
environments. If an outdoor air pollutant fully penetrates into indoor environments, then
the outdoor concentration can be assumed to equal the exposure concentration, bom
outdoors and indoors. If, however, there is less than complete penetration, assuming that
the exposure concentration is the same as the outdoor concentration would overstate
actual exposure.
For the three target HAPs in this analysis, an assumption of 100 percent penetration is
reasonable. A field sampling study of indoor and outdoor concentrations of volatile
organic compounds (VOCs) (Lewis 1991; Lewis and Zweidinger 1992) found that most
VOCs readily penetrate from outdoor to indoor air, even when air exchange rates are low
(i.e., 0.2 to 0.8 air changes per hour). Therefore, for three target HAPs, the long-term
average indoor concentration is expected to be very similar to the long-term average
outdoor concentration in the vicinity.7
Commuting patterns also affect the nri«tinti«iitp between the outdoor concentration and
actual exposure concentrations. As noted above, the outdoor concentration is calculated
by weighting the modeled census tract concentrations by the population residing in each
census tract However, people do not spend all of their time at home: substantial portions
7 Decay of 1,3-buttdiene on indoor surfaces may reduce indoor concentrations, but the reduction is
expected to be small at typical air exchange rates
-------
of time are spent at work or at school, and in transit to or from these locations. These -
commuting patterns are likely to cause increases in exposure. Most commuting patterns
take individuals from areas likely to have relatively low outdoor HAP concentrations (e.g.
suburbs) to areas likely to have higher outdoor HAP concentrations (e.g. downtowns and
industrialized areas). In addition, during time in transit for people in motorized vehicles,
exposures to pollutants emitted by these vehicles (including benzene, 1,3-butadiene and
formaldehyde) are substantially *nh*nff*^ By manning that all individuals stay at home
all of die time, therefore, the population-weighted outdoor concentration may understate
exposure concentrations, and understate the relationship between emissions and exposure
concentrations.
6. Results
Estimated reductions in 2010 cancer incidence for partial attainment of three ozone
NAAQS options are shown in Table 2. Estimated reductions are 0.7 annual cancer cases
for the 0.08 ppm, 5th maximum value option, 1.3 annual cancer cases for the 0.08 ppm,
4th maximum value option, and 2.2 annual cancer cases for the 0.08
-------
estimated national emissions reductions for VOC HAPs. Reductions of VOC HAPs other
than the three target HAPs of the cancer benefits analysis are an estimated 89 tons per day
(32,500 tons per year) under the partial attainment scenario fora 0.08 ppm, 4th maximum
value standard.
The benefits of reductions in emissions of HAPs are likely to be much greater than the
benefits which could be quantified in this analysis. There are several categories of
potential benefits associated with reductions in HAP emissions that would result from
adoption of a revised ozone NAAQS which have not been considered in this analysis.
Exclusion of these categories results in underestimation of the quantitative air toxics
benefits of a revised ozone NAAQS. The categories of benefits which have not been
quantified include:
• Other human carcinogens. VOCs which have been classified as "possible" human
carcinogens were not considered for this analysis. Emissions reductions for any of
these pollutants which actually are carcinogenic in humans would result in additional
benefits which have not been quantified.9 Also, motor-vehicle related control
measures will result in reductions in emissions of the class of HAPs referred to in the
CAA as polycyclic organic matter (POM). Many constituents of POM are classified
by EPA as probable human carcinogens, but different constituents have different IUR
values. POM was not included in this analysis because of the uncertainty in
estimating the proportions of specific POM.constituents in the overall POM
emissions reductions that would be achieved.
• Other human health effects. Many of the VOCs listed as HAPs in the Clean Air Act
are associated with a variety of other adverse effects, including reproductive and
developmental effects (e.g. decreased fertility, birth defects, diminished cognitive
function, impaired growth), central nervous system effects (e.g. impaired motor skills,
reductions in intelligence quotient), liver and kidney damage, and respiratory effects.
Benefits could not be quantified for these effects due to the lack of methods and data.
• Ecological effects. Hazardous air pollutants can also cause damage to a wide variety
of forms of plant and animal life. Methods are not available for quantification of
these effects.
* The same is true for any other VOCs which are human carcinogens but have not been classified as such
by EPA because of an absence of data.
11
-------
Table 1. Screening Analysis for Selection of HAPs for Cancer Benefits Analysis
•
HAP
1,3-butadiene
formaldehyde
benzene
acetaldehyde
ethylene dichloride
acrylonitrile
chloroform
ethylene dibromide
methylene chloride
carbon tetrachloride
hydrazine
hexachlorobenzene
acrylamide
bis(chloromethyl)
ether
propylene oxide
dichloroethyl ether
epichlorohydrin
2,4,6 trichlorophenoi
bromoform
TOTAL
Mean
Cone.
(ug/m3)
0.15
J.7
2.1
0.7
0.1
0.01
0.11
0.01
0.45
0.89
IE-OS
0.00013
IE-OS
2E-07
0.002
5E-06
O.OOOS
7E-08
0.02
Background
Cone.
(ug/m3)
0.49
0.48
0.06
0.08
0.008
O.IS
0.88
0.00009
0.02
Net
Cone.
(ug/m3)
O.IS
12
1.6
0.7
0.04
0.01
0.03
0.002
0.3
001
IE-OS
4E-05
IE-OS
2E-07
0.002
SE-06
O.OOOS
7E-08
0
Cancer
IUR
(ug/m3)-l
2.8E-04
13E-OS
8JE-06
2JE-06
2.6E-OS
6.8E-OS
2.3E-05
2.2E-04
4.7E-07
1.5E-05
4.9E-03
4.6E-04
1.3E-03
6.2E-02
3.7E-06
3JE-04
1.2E-06
3.1E-06
1.1E-06
Cancer
Incidence
Index
42.00
15.72
13.21
1.58
1.02
0.68
0.62
O.S1
'0.14
0.14
0.07
0.02
0.01
'0.01
0.01
0.00
0.00
0.00
0.00
76
Index %
55%
21%
17%
2%
1%
1%
1%
1%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
100%
Cumulative
Index %
55%
76%
94%
96%
97%
98%
99%
99%
100%
100%
100%
100%
100%
100%
100%
100%
100%
100%
100%
NOTES
1. Mean Concentration: arithmetic mean of modeled values for all census tracts
from Cumulative Exposure Project (December 1996 simulation).
2. Net Concentration = Mean Concentration - Background. Background is not expected to be
significantly affected by emissions reductions in near term. For pollutants with no value shown in
"Background" column, background has not been included in estimation of the mean concentration.
3. Only A and B VOC carcinogens with inhalation unit risk (IUR) value in IRIS are
included in this screening analysis.
4. Cancer Incidence Index = (Net Concentration)*(IUR)* 1,000,000
12
-------
Table 2. Estimated Reductions in Cancer Incidence in 2010 due to Partial Attainment of
a Revised Ozone NAAQS (annual cases avoided)'
Pollutant
Benzene
1,3 -butadiene
Formaldehyde (total)2
Total
0.08 ppm
5* Max
8-hour Avg.
0.1
0.5
0.1
0.7
0.08 ppm
4* Max
8-hour Avg.
0.3
0.9
0.1
1.3
0.08 ppm
S^Max
8-hour Avg.
0.4
1.6
0.2
2.2
'Incremental to current ozone NAAQS.
2Sum of primary formaldehyde and secondary formaldehyde.
13
-------
Table 3. Valuation of Estimated Cancer Risk Reductions in 2010 for Partial Attainment
of a Revised Ozone NAAQS1
Value of a Statistical Life
5* Percentile ($0.75 million/case)
Mean ($4.8 million/case)
95th Percentile ($1 1 million/case)
0.08 ppm
5* Max
8-hour Avg.
$0.5 million
$3.4 million
$7.7 million
0.08 ppm
4* Max
8-hour Avg.
$1.0 million
$6.2 million
$14.3 million
0.08 ppm
3" Max
8-hour Avg.
$1.7 million
$10.6 million
$24.2 million
'Incremental to current ozone NAAQS.
14
-------
Table 4. Estimated Net National HAP Emissions Reductions (tons per day) in 2010 for
Partial Attainment of a Revised Ozone NAAQS1
Pollutant
Benzene
1,3-butadiene
Formaldehyde (primary)
All other VOCHAPs
TOTAL
0.08 ppm
5* Max
8-hour Avg.
3.6
0.6
0.7
46.9
51.8
0.08 ppm
4* Max
8-hour Avg.
10.3
1.9
1.7.
88.8
102.7 •*
0.08 ppm
3* Max
8-hour Avg.
18.2
3.4
3.1
153.0
177.7
'Incremental to current ozone NAAQS.
15
-------
References
Lewis, C. (1991). "Sources of Air Pollutants Indoors: VOC and Fine Particulaie
Species.*1 Journal of Exposure Analysis and Environmental Epidemiology 1(1): 31-44.
Lewis, C. and R. Zwcidinger (1992). "Apportionment of residential indoor aerosol, VOC
and aldehyde species to indoor and outdoor sources, and their source strengths."
Atmospheric Environment 26Ad 2): 2179-2184.
Rosenbaum, A. and Y. Wei (1997). Memorandum to Jim Wilson, E.H. Pechan and
Associates. Baseline hazardous air pollutant (HAP) data. San Rafael. CA. ICF Kaiser
Consulting Group—Systems Applications International, Inc. March 14.
Rosenbaum, A. S., M. P. Ligocki, Y. H. Wei and J. P. Cohen (1996). Revised
Methodology For Modeling Cumulative Exposures To Air Toxics I: Outdoor
Concentrations; EPA Science Advisory Board Advisory Draft. Sacr-Rafael, CA.
Systems Applications International.
Science Advisory Board (1996). An SAB Report: The Cumulative Exposure Project.
Washington, DC. U.S. Environmental Protection Agency. EPA-SAB-IHEC-ADV-96-
004.
Wallace, L. (1987). Total Exposure Assessment Methodology (TEAM) Study:
Summary and Analysis, Volume I. U.S. Environmental Protection Agency. EPA-
600/6-87-002a.
Wilson, J. (1997). Task 4b Memorandum - HAP Emission Estimates. Springfield, VA.
E.H. Pechan & Associates. July 9.
16
-------
Appendix 1
Calculation of Estimated Reductions in Cancer Risk
-------
Benzene- 08-5thMAX
STATE
AL
AZ
AR
CA
CO
CT
OE
DC
FL
GA
ID
IL
IN
1A
KS
KY
LA
ME
MD
MA
Ml
MN
MS
MO
MT
NE
NV
NH
NJ
MM
NY
NC
NO
OH
OK
OR
PA
Rl
sc
SD
TN
TX
UT
VT
VA
WA
WV
Wl
WY
Benzene
Reductions
(tpd)
0.02
0.78
0.01
0.13
0.05
0.10
0.00
0.02
0.06
0.56
0.01
0.17
0.10
0.01
0.01
0.34
0.25
001
021
0.12
0.07
0.03
0.01
0.06
0.01
0.01
0.01
0.02
0.18
0.01
0.17
0.04
0.00
0.16
002
0.02
0.09
002
0.02
0.00
0.03
•1.06
0.01
0.00
0.46
0.03
012
0.06
0.00
Baseline
Benzene
Emissions
(tpd)
30
23
14
167
21
16
5
2
68
49
29
51
37
14
16
24
41
8
25
31
58
28
21
30
20
8
7
7
42
12
74
48
6
SB
19
30
67
5
24
9
37
170
12
4
37
55
17
27
6
Benzene
Reduction*
(%)
0.1%
3%
0.1%
0.1%
02%
1%
0%
1%
0.1%
1%
0.02%
03%
03%
0.1%
0.1%
1%
0.6%
0.1%
1%
0.4%
0.1%
0.1%
0.06%
02%
0.03%
0.1%
0.1%
0.3%
0.4%
0.1%
0.2%
0.1%
0%
0.3%
0.1%
0.1%
0.1%
0.4%
0.1%
0%
0.1%
-1%
0.1%
OH
1%
0.05%
1%
0.2%
0%
Baseline
Benzene
Concn
(ug/m3)
0.84
147
0.57
2.56
1.26
1.58
1.09
3.33
1.15
101
095
1.37
0.98
0.47
0.56
0.74
1.54
058
1.81
1.37
1.32
0.99
070
1.02
0.78
0.49
1.13
0.98
2.76
0.78
2.75
0.91
0.51
121
0.58
1.53
1.90
1.38
0.79
046
1.05
1.91
1.16
0.66
1.01
1.85
0.91
0.83
0.54
Benzene
Concn
Reduction
(ugftnS)
0.001
0.05
0.0004
0.002
0.003
0.01
0
0.04
0.001
0.01
0.0002
0.004
0.003
0.0005
00004
001
0.01
0.0005
0.02
0.01
0002
0.001
0.0004
0.002
00002
0.001
0.001
0.003
0.01
0001
0.01
0.001
0
0.003
0.001
0.001
0.003
0.01
0.001
0
0.001
-0.01
0.001
0
0.01
0.001
0.01
0.002
0
Benzene
Potency
. .Factor
(ugfrnSH
8JE-06
8.3E-06
8.3E-08
8.3E-06
&3B06
8.3E-08
8.3E-06
83E-06
8.3E-06
83E-06
83E-06
83E-06
8.3E-06
8.3E-08
8.3E-06
83E-06
83E-06
83E-06
83E-06
8.3E-06
8.3E-06
8.3E-06
8.3E-06
' 8.3E-06
83E-06
8.3E-06
8.3E-06
83E-06
83E-06
8.3E-06
8.3E-06
8.3E-06
8.3E-06
8.3E-06
8.3E-06
8.3E-06
8.3E-06
83E-06
83E-06
8.3E-06
8.3E-06
8.3E-06
8.3E-06
8.3E-06
&3E-08
8.3E-06
83E-06
83E-06
83E-06
2010
POP
(x 10*3)
4.798
5.522
2,840
37.644
4.658
3.400
817
560
17.363
8.824
1.557
12.515
6.318
2.988
2.849
4.170
4.683
1.323
5.657
6.431
9.636
5.147
2.974
5.864
1.040
1.806
2.131
1.329
8.638
2,155
18.530
8.552
690
11.505
3.639
3.803
12.352
1.038
4.205
826
6.180
,22457
. ' 2.551
651
7.627
6.658
1.851
5.590
607
Avoided
2010
Benzene
Cases
0.00
0.03
0.00
0.01
0.00
0.00
0.00
0.00
0.00
0.01
0.00
0.01
0.00
0.00
0.00
0.01
0.01
0.00
0.01
0.00
0.00
000
000
0.00
0.00
0.00
0.00
0.00
001
0.00
0.01
0.00
0.00
000
000
0.00
0.00
0.00
000
000
0.00
-0.03
0.00
0.00
0.01
0.00
0.00
0.00
'0.00
TOTAL CASES AVOIDED 0.1
i*v 6/17/97
-------
,3-butadiene-.08-5thMAX
STATE
AL
AZ
AR
CA
CO
CT
DE
DC
FL
GA
ID
IL
IN
IA
KS
KY
LA
ME
MD
MA
Ml
MN
MS
MO
MT
NE
NV
NH
NJ
NM
NY
NC
NO
OH
OK
OR
PA
Rl
SC
SO
TN
TX
UT
VT
VA
WA
WV
Wl
WY
1.3-tutadiene
Reductions
(tpd)
0.00
0.14
0.00
0.01
0.01
0.02
0.00
0.00
0.01
011
000
003
0.02
0.00
0.00
007
0.02
000
0.02
0.02
0.01
0.01
0.00
0.01
0.00
0.00
0.00
000
004
0.00
0.03
0.01
0.00
0.04
0.00
0.00
0.02
000
0.00
0.00
0.01
•020
0.00
0.09
0.08
001
002
0.01
000
Baseline
1.3-butadlene
Emissions
(tpd)
5
3
2
23
3
2
1
0
12
8
8
7
5
2
2
4
7
1
3
4
8
4
3
5
5
1
1
1
4
2
10
8
1
8
3
5
8
1
4
2
6
22
2
1
S
10
2
4
1
1.3-butadlene
Reductions
OK
4K
OK
0.03K
O3K
IK
OK
IK
O.IK
IK
OK
O.SK
0.4K
OK
OK
2K
0.3K
OK
1K
0.6K
02H
0.1K
OK
0.3K
OK
OK
OK
OK
1K
OK
0.4K
0.1K
OK
O.SK
OK
OK
0.2K
1K
OK
OK
OH
-1K
OK
OK
1K
OIK
1K
0.3K
OK
Baseline
l^butMflene
Concn
(ugftnS)
0.07
0.11
0.05
021
0.12
0.13
0.12
0.40
0.11
0.09
0.10
0.14
0.09
0.05
0.08
008
0.14
006
0.14
0.14
0.14
0.10
0.06
0.09
0.11
0.05
0.10
0.09
0.18
0.09
0.27
0.08
0.04
0.12
0.05
0.15
0.15
0.13
0.07
0.05
0.10
0.12
0.11
0.08
0.09
0.18
0.07
0.09
0.05
1>butadiene
Concn
Reduction
(ug/m3)
0
0005
0
0.0001
0.0004
0.001
0
0
00001
0.001
0
0001
00004
0
0
0.001
0.0004
0
0.001
0001
0.0002
0.0001
0
00002
0
0
0
0
0002
0
0001
0.0001
0
0.001
0
0
0.0003
0
0
0
0.0001
-0.001
0
0
0.001
0.0001
0.0006
0.0003
0
1>butadtene
Potency
Factor
(ug/m3M
2.8E-04
2.6E-04
2.8E-04
2.6E-04
2.8E-04
2.8&04
2.8S04
2.8E-04
2.8E-04
2.8E-04
28E-04
2.8E-04
2.8E-04
2.8E-04
2.8E-04
2.8E-04
2.8E^04
2.8E-04
2.6E-04
2.6E-04
2.8E-04
2.8E-04
2.8E-04
2.8E-04
2.8E-04
2.8E-04
2.8E-04
2.8E-04
2.8E-04
2.8E-04
2.8E-04
2.8E-04
2.8E-04
2.8E-04
2.8E-04
2.8E-04
2.8E-04
2.8E-04
2.8E-04
2.8E-04
2.8E-04
2.6E-04
2.8E-04
2.8&04
i8E-04
2.8E-04
28E-04
2.8E-04
2.8E-O4
2010
POP
(X10*3)
4.798
5.522
2.640
37.644
4.668
3.400
817
560
17.363
8.824
1.557
12.515
6.318
2.968
2.649
4.170
4.663
1.323
5.657
6.431
9.636
5.147
2.974
5.864
1.040
1.806
2.131
1.329
8.638
2.155
18.530
8.552
690
11.505
3.639
3.603
12.352
1.038
4205
826
6.160
, 22,657
2.551
651
7.627
6.658
1,651
5.590
607
Avoided
2010
Cases
000
0.10
0.00
0.01
0.01
0.01
0.00
0.01
0.01
0.04
0.00
0.03
001
0.00
0.00
0.02
0.01
0.00
0.02
0.02
0.01
000
0.00
0.01
0.00
0.00
0.00
0.00
0.05
0.00
0.07
0.00
0.00
0.03
0.00
0.00
0.02
000
0.00
0.00
0.00
-0.10
0.00
0.00
0.04
0.00
0.00
0.01
0.00
TOTAL CASES AVOIDED 0.5
-------
Primary Fomnald-.08-5thMAX
STATE
AL
AZ
AR
CA
CO
CT
DE
DC
FL
GA
ID
IL
IN
IA
KS
KY
LA
ME
MO
MA
Ml
MN
MS
MO
MT
NE
NV
NH
NJ
NM
NY
NC
NO
OH
OK
OR
PA
Rl
SC
SO
TN
TX
MT
VT
VA
WA
WV
Wl
WY
Form (prim)
.Reduction*
(tpd)
0.00
0.12
0.00
0.10
0.01
0.01
0.00
0.00
001
009
000
0.03
0.01
0.00
0.00
0.06
0.02
0.00
002
0.02
0.01
0.00
000
0.01
0.00
0.00
0.00
000
003
0.00
0.03
0.01
0.00
0.03
0.00
0.00
0.01
000
0.00
000
0.02
-0.17
0.00
. 0.00
0.08
0.00
0.02
0.01
0.00
Baseline
Form (prim]
Emission*
(tpd)
21
16
10
113
13
7
2
1
57
41
49
31
20
9
11
15
40
5
12
15
32
16
17
17
32
5
4
3
18
11
39
33
3
33
14
25
34
2
18
12
20
142
8
2
20
51
8
16
4
Form (prim]
Reduction*
(%)
0%
IK
0%
0.1%
0.1%
02%
0%
0%
0.02%
0.2%
0%
0.1%
0.07%
0%
0%
0.4%
0.04%
0%
0.2%
0.1%
0.04%
0%
0%
0.1%
0%
0%
0%
0%
0.2%
0%
0.1%
0.02%
0%
0.1%
0%
0%
0.04%
0%
0%
0%
0,1%
-0.1%
0%
0%
0.4%
0%
0.3%
0.1%
0%
Baseline
roriii (prim]
Concn
(ugftrd)
0.43
0.83
0.29
132
0.95
0.68
D.84
2.55
0.78
062
0.75
093
0.54
025
0.33
0.48
1.24
0.27
0.90
0.60
0.79
0.57
0.49
O.S6
0.75
0.31
0.45
0.41
1.29
0.56
1.80
0.52
024
0.75
0.40
0.80
0.96
0.58
0.49
0.38
0.51
133
0.79
O30
0.51
1.06
0.38
0.49
0.29
Form (prim)
Concn
Reduction
(ugftnS)
0
0006
0
0.001
0.001
0.001
0
0
0.0001
0001
0
0001
0.0004
0
0
0.002
00005
0
0002
0.001
00003
0
0
0.0004
0
0
0
0
0002
0
0001
0.0001
0
0.001
0
0
0.0004
0
0
0
0.001
-0.002
0
0
0.002
0
0.001
00003
0
Formald
D«^^hMK4M«
l'lH8IN»y
Factor
(ug/rn3)-1
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-OS
1.3E-05
1.3E-05
1.3E-05
13E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-OQ
1.3E-05
1.3E-05
1.3&05
1-3E-05
1.3E-05
1.3E-05
1.3E-05
1 .36-05
1.3E-05
1.3&05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
13E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1J&05
1.3&05
1J6-05
1.3E-05
1.3E-05
1.3&05
1.3E-05
1.3E-05
1.3E-05
2010
POP
(x 10*3)
4,798
5.522
2.840
37,644
4.668
3.400
817
560
17.363
8.824
1.557
12.515
6.318
2.968
2.849
4.170
4.683
1.323
5.657
6.431
9.836
5.147
2.974
5.864
1.040
1.806
2.131
1.329
8.638
2.155
18.530
8.552
690
11.505
3.639
3.803
12.352
1.038
4.205
626
6,180
22,857
2,551
651
7.627
6.658
1.651
5.590
607
Avoided
2010
Form (prim)
Case*
0.00
0.01
0.00
0.01
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
000
0.00
000
0.00
0.00
000
000
000
0.00
0.00
0.00
000
000
0.00
000
0.00
0.00
0.00
000
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
000
0.00
•0.01
0.00
0.00
0.00
0.00
0.00
0.00
0.00
TOTAL CASES AVOIDED 0.04
f»y«/17«7
-------
Secondary FormaW-.OB-SthMAX
STATE
AL
AZ
AR
CA
CO
CT
DE
DC
FL
6A
10
IL
IN
IA
KS
KY
LA
ME
MD
MA
Ml
MN
MS
MO
MT
ME
NV
NH
NJ
NM
NY
NC
NO
OH
OK
OR
PA
Rl
SC
SO
TN
TX
UT
VT
VA
WA
VW
Wl
WY
Precursor
Reductions
(u>d)
0.07
2.43
0.04
0.39
0.15
0.28
0.01
0.05
0.21
189
0.02
056
0.35
0.05
0.04
1 18
033
002
050
040
0.23
0.09
0.04
0.22
0.02
0.03
0.02
007
061
0.03
0.58
0.12
0.01
0.61
006
0.05
0.30
0.07
0.06
002
0.12
(3.41)
0.03
0.01
1.34
009
0.35
021
001
Baseline
Precursor
Emissions
(tpd)
63
46
31
327
43
28
9
5
163
111
82
108
73
33
37
51
96
17
50
59
124
60
45
73
55
20
14
12
69
29
137
103
10
119
42
60
121
9
52
23
75
412
24
9
73
110
28
60
10
Precursor
Emission
Reductions
(%)
0.1%
5%
0.1%
0.1%
0.4%
1%
0.1%
1%
0.1%
2%
0.02%
1%
0.5%
0.2%
0.1%
2%
0.3%
0.1%
1%
1%
0.2%
0.1%
0.1%
0.3%
0.04%
0.1%
0.1%
1%
1%
0.1%
0.4%
0.1%
0.1%
1%
0.1%
0.1%
0.2%
1%
0.1%
0.1%
02%
-1%
0.1%
0.1%
2%
0.1%
1%
0.3%
0.1%
Beseline
Form (See)
Concn
(ugmri)
0.09
0.14
0.05
0.27
0.13
0.20
0.17
035
0.13
0.16
007
017
0.10
0.04
0.05
0.08
0.14
0.05
0.23
0.14
0.18
0.12
0.06
0.11
0.06
0.05
0.06
0.12
0.36
0.06
037
0.11
0.01
0.14
0.05
018
0.20
0.12
0.10
0.02
0.12
028
0.10
0.04
0.12
0.25
0.08
0.07
001
Form (See)
Concn
RMfUCQOfl
(ugftnS)
0.0001
0.007
0.0001
0.0003
0.0006
0.002
• 0.0002
0004
0.0002
0003
0.00002
0001
0.0005
0.0001
0.0001
0.002
0.0005
00001
0.002
0.001
0.0003
0.0002
0.0001
0.0003
0.00002
0.0001
0.0001
0001
0.003
0.0001
0002
0.0001
0.00001
0.001
0.0001
0.0002
00005
0.001
0.0001
0.00002
0.0002
-0.002
0.0001
0,00005
0.002
0.0002
0001
0.0002
0.00001
TOTAL CASES AVOIDE
Formald
Potency
Factor
(ugftn3)-1
1.3E-05
1.3E-05
1.3E-05
13E-05
1.3E-05
1.3E-05
1.3E-05
13E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E45
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
13E-05
1.3E-05
13E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1JE-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-OS
1.3E-05
2010
POP
(x 10*3)
4.798
5.522
2.840
37.644
4.658
3.400
817
560
17.363
6.824
1.557
'12.515
6.318
2.968
2.849
4.170
,4.683
'1.323
5.657
6.431
9.836
5.147
2.974
5.864
1.040
1.806
2.131
1.329
8.638
2.155
18.530
8.552
690
11.505
3.639
3,803
12.352
1.038
4.205
826
6.180
22.857
2,551
651
7.627
6.658
1.851
5.590
607
Avoided
2010
Form (See]
Cesee
0.00
0.01
p.oo
0.00
0.00
0.00
0.00
0.00
000
000
000
0.00
000
000
000
000
000
000
0.00
000
000
000
000
0.00
0.00
0.00
000
000
001
0.00
0.01
0.00
0.00
0.00
0.00
0.00
000
• ooo
000
000
0.00
-0.01
0.00
0.00
0.00
000
000
000
0.00
> 0.03
r*v6T17<97
-------
Benzene-.08-4thMAX
STATE
AL
AZ
AR
CA
CO
CT
DE
DC
FL
GA
10
IL
IN
IA
KS
KY
LA
ME
MD
MA
Ml
MN
MS
MO
MT
NE
NV
NH
NJ
NM
NY
NC
NO
OH
OK
OR
PA
Rl
SC
SO
TN
TX
UT
vr
VA
WA
WV
Wl
WY
Benzene
Emissions
Reductions
(tpd)
1.44
0.49
0.12
0.13
0.05
0.10
0.00
0.02
0.06
056
0.01
0.17
0.11
0.01
0.01
0.35
0.37
001
0.22
0.12
0.07
0.03
015
0.06
0.01
0.01
0.01
0.02
018
0.01
0.17
1.33
0.00
0.19
002
0.02
009
0.02
0.20
0.00
147
0.19
0.01
0.00
0.47
0.03
0.12
0.06
0.00
Baseline
Benzene
Emissions
(«x>)
30
23
14
167
21
16
5
2
66
49
29
51
37
14
16
24
41
6
25
31
58
28
21
30
20
8
7
7
42
12
74
48
6
58
19
30
67
5
24
9
37
170
12
4
37
56
17
27
6
Benzene
Reductions
(%)
5%
2%
IS
0.1%
0.2%
1%
0%
IK
0.1%
1%
0.02%
03%
0.3%
0.1%
0.1%
1%
1%
0.1%
1%
0.4%
0.1%
0.1%
1%
0.2%
0.03%
0.1%
0.1%
0.3%
0.4%
0.1%
0.2%
3%
0%
0.3%
0.1%
0.1%
0.1%
0.4%
1%
0%
7%
0.1%
0.1%
0%
1%
0.05%
1%
0.2%
0%
Baseline
Btnxont
Condi
(ug/m3)
0.84
1.47
0.57
2.56
126
1.58
1.99
3.33
1.15
1.01
0.95
137
0.98
0.47
0.56
0.74
154
058
1.81
1.37
1.32
0.99
0.70
1.02
078
0.49
1.13
0.98
2.76
0.78
2.75
0.91
0.51
1.21
0.58
1.53
1.90
1.38
0.79
0.46
1.05
1.91
1.16
0.66
1.01
1.85
0.91
0.83
0.54
Benzene
Concn
Reduction
(ugftnS)
0.04
0.03
0.005
0.002
0.003
001
0
0.04
0.001
001
0.0002
0.004
0.003
0.0005
0.0004
0.01
0.01
00005
0.02
0.01
0.002
0.001
0.005
0.002
0.0002
0.001
0001
0.003
001
0.001
0.01
003
0
0.004
0001
0.001
0.003
0.01
0.01
0
0.1
0.002
0.001
0
0.01
0.001
0.01
0002
0
Benzene
Potency
Fm*^**f
. . rmGVBT
(ug/hrtM
6.3E-O6
8.3E-06
8.3E-06
8.3E-06
8.3E-06
83E-O6
8.3E-06
8.3E-06
8.3E-06
8.3E-06
83E-06
8.3E-06*
83E-06
8.3E-06
83E-06
83E-06
8.3E-06
83E-06
8.3E-06
8.3E-06
8.3E-06
8.3E-06
8.3E-06
83E-06
83E-06
8.3E-O6
8.3E-06
8.3E-06
8.3E-06
83E-06
83E-06
83E-06
8.3E-06
8.3E-06
83E-06
8.3E-06
8.3E-06
8.3E-06
8.3E-06
8.3E-06
8.3E-06
&3E-06
8.3E-06
S.3E46
8.3E-06
8.3E-06
8.3E-06
8.3E-06
83E-06
2010
POP
(x 10*3)
4.798
5.522
2.840
37.644
4.658
3.400
817
560
17.363
8.824
1.557
12.515
6.318
2.968
2.849
4.170
4.683
1.323
5.657
6.431
9.836
5.147
2.974
5.864
1.040
1.806
2.131
1.329
8.638
2.155
18.530
8.552
690
11.505
3.639
3.803
12.352
1.038
4.205
826
6,180
22,657
' 2,551
651
7.627
6.658
1.851
5.590
607
Avoided
2010
Benzene
Cases
0.02
0.02
0.00
0.01
0.00
0.00
0.00
0.00
0.00
001
000
0.01
000
0.00
0.00
0.01
0.01
000
001
000
000
000
0.00
0.00
0.00
0.00
0.00
000
001
000
0.01
0.03
0.00
0.01
000
0.00
0.00
000
0.00
0.00
0.05
0.01
0.00
0.00
0.01
0.00
0.00
0.00
000
TOTAL CASES AVOIDED 0.3
-------
1,3-butadiene- 08-4thMAX
STATE
AL
AZ
AR
CA
CO
CT
DE
DC
FL
GA
ID
IL
IN
IA
KS
KY
LA
ME
MO
MA
Ml
MN
MS
MO
MT
NE
NV
NH
NJ
NM
NY
NC
ND
OH
OK
OR
PA
Rl
SC
SO
TN
TX
UT
VT
VA
WA
WV
Wl
WY
1.3-butadiene
Emissions
Reductions
(tpd)
0.29
0.07
0.02
001
0.01
0.02
0.00
0.002
0.01
0.11
0.00
0.03
0.02
000
000
0.07
0.04
000
0.02
0.02
001
0.01
0.03
0.01
0.00
0.00
0.00
0.00
0.04
0.00
003
0.26
000
004
0.00
000
0.02
0.00
0.04
000
0.44
0.04
0.00
0.00.
0.08
0.01
0.02
0.01
0.00
Baseline
1,3-butadiene
Emissions
(tpd)
5
3
2
23
3
2
1
0.3
12
8
8
7
5
2
2
4
7
1
3
4
8
4
3
5
5
1
1
1
4
2
10
8
1
8
3
5
8
1
4
2
6
22
2
1
5
10
2
4
1
1,3-fcutadiene
Reductions
(%)
6%
2%
1%
0.03%
0.3%
1%
0%
1%
0.1%
1%
0%
05%
04%
0%
0%
2%
0.5%
0%
1%
06%
0.2%
0.1%
1%
0.3%
0%
0%
0%
0%
1%
0%
0.4%
4%
0%
0.4%
0%
0%
02%
1%
1%
0%
8%
0.2%
0%
0%
1%
0.1%
1%
0.3%
0%
Baseline
1.3-butadtone
Concn
(ug/hiS)
0.07
0.11
0.05
0.21
0.12
0.13
0.12
0.40
0.11
0.09
0.10
0.14
0.09
0.05
0.06
0.08
0.14
0.06
0.14
0.14
0.14
0.10
0.06
0.09
0.11
0.05
0.10
0.09
018
009
0.27
008
0.04
0.12
005
0.15
015
0.13
0.07
0.05
0.10
0.12
0.11
0.08
0.09
0.18
007
009
005
1.3-butadiefie
Concn
Reduction
(ug/mJ)
0004
0.002
0.0004
0.0001
0.0004
0.001
0
0003
00001
0.001
0
0.001
0.0004
0
0
0001
0001
0
0.001
0001
0.0002
0.0001
0.0005
0.0002
0
0
0
0
0002
0
0.001
0.003
0
0.001
0
0
0.0003
0
0.001
0
0.008
0.0002
0
0
0.001
0.0001
0.001
00003
0
1.3-butadlene
Potency
. . • Factor
(ugftn3)-1
2.8E-O4
2.6E-04
2.6E-04
2.8E-04
2.8E-04
2.8E-04
2.8E-04
2.6E-04
2.8E-04
2.8E-04
2.8E-04
2.8E-04
26E-04
2.8E-O4
2.8E-04
2.8E-04
2.8E£4
2.8E-04
2.8E-04
2-86-04
2.BE-04
2.8E-O4
2.8E-04
2.8E-04
2.8E-04
2.8E-04
2.86-04
Z8E-04
2.8E-04
2.8E-04
Z8E-04
2.8E-O4
2.8E-04
2.8E-O4
2.8E-04
2.8E-04
2.8E-04
2.8E-04
2.BE-04
28E-04
2.8&04
2.8E-O4
2.8&04
2.8E-04
2.8E-04
2.8E-04
2.8E-04
2.8E-04
286-04
2010
POP
(X 10A3)
4.798
5.522
2.840
37.644
4.658
3,400
817
560
17.363
8.824
1.557
12.515
6.318
2,968
2.849
4.170
4.683
1.323
5.657
6.431
9.836
5.147
2.974
5.864
1.040
1.806
2.131
1.329
8.638
2,155
18.530
8,552
€90
11.505
3.639
3.803
12.352
1.038
4.205
826
6.180
22.857
2.551
651
7.627
6.658
1.851
5.590
607
Avoided
2010
1.3-butadiene
Cases
0.08
0.05
0.01
0.01
0.01
0.01
0.00
001
0.01
0.04
0.00
0.03
0.01
0.00
000
002
0.01
0.00
002
0.02
0.01
0.00
0.01
0.01
0.00
0.00
0.00
0.00
0.05
0.00
0.07
0.10
0.00
0.02
0.00
0.00
0.02
0.00
0.01
0.00
0.19
0.02
0.00
0.00
0.04
0.00
000
0.01
000
TOTAL CASES AVOIDED 0.9
-------
Primary Formald-.08-4thMAX
STATE
AL
AZ
AR
CA
CO
CT
DE
DC
FL
GA
ID
IL
IN
IA
KS
KY
LA
ME
MD
MA
Ml
MN
MS
MO
MT
NE
NV
NH
NJ
NM
NY
NC
ND
OH
OK
OR
PA
Ri
SC
SD
TN
TX
UT
VT
VA
WA
WV
Wl
WY
Form (prim]
Emissions
deductions
(tpd)
0.24
0.07
002
0.11
0.01
001
0.00
0.00
0.01
0.09
000
0.03
002
0.00
0.00
006
003
000
0.02
0.02
0.01
0.00
0.03
001
0.00
0.00
0.00
.000
0.03
0.00
0.03
0.22
0.00
0.03
000
000
0.01
0.00
0.03
0.00
0.38
0.03
0.00
- 0.00
0.08
0.00
0.02
0.01
0.00
Baseline
Form (prim)
Emissions
(tpd)
21
16
10
113
13
7
2
1
57
41
49
31
20
9
11
15
40
5
12
15
32
16
17
17
32
5
4
3
18
11
39
33
3
33
14
25
34
2
18
12
20
142
8
2
20
51
8
16
4
Form (prim]
Reductions
1%
0.4%
02%
0.1%
0.1%
02%
0%
0%
0.02%
0.2%
0%
0.1%
0.1%
0%
0%
0.4%
01%
0%
02%
0.1%
0.04%
0%
02%
01%
0%
0%
0%
0%
0.2%
0%
0.1%
0.7%
0%
0.1%
0%
0%
0.04%
0%
0%
0%
2%
002%
0%
0%
04%
0%
0.3%
0.1%
0%
Baseline
Form (prim]
Concn
(ugftid)
O43
0.83
029
1.52
0.95
0.66
0.84
2.55
0.78
0.62
0.75
0.93
0.54
025
0.33
0.48
1.24
0.27
0.90
0.60
0.79
0.57
0.49
0.56
0.75
031
0.45
0.41
129
0.56
1.80
0.52
024
0.75
0.40
0.80
0.96
0.58
0.49
0.38
0.51
1.53
079
030
051
1.06
0.38
0.49
029
Form (prim]
Concn
(Uflftltf)
0
0.004
0
0.001
0.001
0001
0
0
00001
0001
0
0.001
0.0005
0
0
0.002
0001
0
0.002
0.001
0.0003
0
0
0.0004
0
0
0
0
0002
0
0001
0004
0
0.001
0
0
0.0004
0
0
0
0.01
0.0003
0
0
0.002
0
0001
0.0003
0
Formald
Potency
-. Factor
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
13E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
13E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
2010
POP
(Xl0*3)
4.798
5,522
2,840
37.644
4.658
3.400
817
560
17,363
8.824
1,557
12.515
6.318
2,968
2.849
4.170
4.683
1.323
5.657
6.431
9.836
5.147
2.974
5,864
1.040
1.806
2.131
1.329
8.638
2.155
18.530
8.552
690
11.505
3.639
3.603
12.352
1.038
4205
826
6.180
, 22.857
2,551
651
7.627
6.656
1,851
5.590
607
TOTAL CASES AVOIDED
Avoided
2010
Form (prim]
Cases
0.00
0.00
0.00
0.01
000
0.00
0.00
0.00
000
0.00
0.00
0.00
0.00
0.00
0.00
000
0.00
000
0.00
000
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
000
0.00
000
0.01
0.00
0.00
0.00
000
0.00
000
0.00
0.00
0.01
0.00
0.00
0.00
0.00
o.oo
0.00
000
0.00
0.07
tr* 719/97
-------
Secondary Formald-.08-4thMAX
STATE
AL
AZ
AR
CA
CO
CT
DE
DC
FL
GA
ID
IL
IN
IA
KS
KY
LA
ME
MD
MA
Ml
MN
MS
MO
MT
NE
NV
NH
NJ
NM
NY
NC
ND
OH
OK
OR
PA
Rl
SC
SO
TN
TX
UT
VT
VA
WA
WV
Wl
WY
Precursor
Emissions
Reductions
(tpd)
4.87
1.33
0.33
0.41
0.15
0.28
0.01
0.05
0.21
1.89
002
056
0.36
0.05
0.04
1.19
0.63
0.02
050
0.40
0.23
0.09
0.44
022
0.02
0.03
0.02
O.Q7
0.61
003
0.58
4.49
0.01
064
0.06
0.05
0.30
0.07
0.68
0.02
7.52
0.64
0.03
0.01
1.36
0.09
0.35
0.21
0.01
Baseline
Precursor
Emissions
63
46
31
327
43
28
9
5
163
111
82
108
73
33
37
51
96
17
50
59
124
60
45
73
55
20
14
12
69
29
137
103
10
119
42
60
121
9
52
23
75
412
24
9
73
110
28
60
10
Precursor
Emission
Reductions
8%
3%
1%
0.1%
0.4%
1%
0.1%
1%
01%
2%
002%
1%
0.5%
01%
0.1%
2%
07%
01%
1%
1%
0.2%
02%
1%
0.3%
003%
0.1%
0.2%
1%
1%
0.1%
0.4%
4%
0.1%
1%
0.1%
0.1%
02%
1%
1%
0.1%
10%
02%
0.1%
0.1%
2%
0.1%
1%
0.4%
0.1%
Baseline
Form (Sec)
Conen
(uo/m3)
0.09
0.14
0.05
027
0.13
020
0.17
0.35
0.13
0.16
007
0.17
0.10
0.04
0.05
008
0.14
005
023
0.14
0.18
0.12
0.06
0.11
0.06
0.05
0.06
0.12
036
006
037
0.11
0.01
0.14
005
018
0.20
0.12
0.10
0.02
0.12
028
0.10
Form (Sec]
Concn
Reduction
(ugftnS)
0.007
0.004
0.0005
0.0003
0.0005
0.002
0.0002
0.004
0.0002
0003
0.00002
0.001
0.0005
0.0001
0.0001
0002
0001
0.0001
0.002
0.001
0.0003
0.0002
0.001
0.0003
0.00002
0.0001
0.0001
0.001
0003
00001
0.002
0.005
0.00001
0001
0.0001
00002
0.0005
0.001
0001
0.00001
0.01
0.0004
0.0001
0.04 0.0001
0.12
025
0.08
0.07
0.01
0.002
0.0002
0001
0.0002
0.00001
Formald
Potency
Factor
(ug/m3M
1.3E-05
1.3E-05
1JE-05
1.3E-05
1JE-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1JE-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-O5
1.SE-05
1JB05
1.3E-05
1.3&05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
2010
POP
(x 10*3)
4.798
5.522
2.840
37.644
4.658
3.400
817
560
17.363
8.824
1.557
' 12.515
6.318
2.968
2.849
4.170
,%4.683
1.323
5.657
6.431
9.836
5.147
2.974
5.864
1«.040
1.806
2.131
1.329
8.638
2.155
18.530
8.552
690
11.505
3.639
3.803
12.352
1,038
4205
826
6.180
22.857
2,551
651
7.627
6.656
1.651
5.590
607
Avoided
2010
Form (Sec]
Cases
0.01
0.00
0.00
0.00
0.00
0.00
0.00
0.00
000
000
000
000
0.00
000
0.00
0.00
000
000
000
0.00
0.00
0.00
0.00
0.00
000
0.00
000
000
001
0.00
0.01
0.01
0.00
000
0.00
0.00
0.00
0.00
0.00
000
0.01
0.00
0.00
000
0.00
0.00
000
000
0.00
TOTAL CASES AVOIDED 0.07
re» 7/9/97
-------
Benzene-.08-3rdMAX
STATE
AL
AZ
AR
CA
CO
CT
DE
DC
FL
GA
ID
IL
IN
(A
KS
KY
LA
ME
MD
MA
Ml
MN
MS
MO
MT
NE
NV
NH
NJ
NM
NY
NC
NO
OH
OK
OR
PA
Rl
sc
SD
TN
TX
UT
VT
VA
WA
WV
Wl
WY
Benzene
Reduction*
(tpd)
1.44
0.78
0.12
0.14
0.05
0.10
0.00
0.02
0.06
0.56
0.01
027
0.12
0.01
0.01
0.36
0.36
0.21
0.22
0.14
0.07
0.03
0.15
0.06
0.01
0.01
0.01
0.18
0.18
001
0.17
1.34
0.00
0.20
1.18
0.02
0.09
002
0.20
0.00
3.77
1.36
0.01
0.00
3.91
0.03
0.12
007
000
Baseline
Benzene
•> • •
emissions
(tpd)
30
23
14
167
21
16
5
2
66
49
29
51
37
14
16
24
41
8
25
31
58
28
21
30
20
8
7
7
42
12
74
48
6
58
19
30
67
5
24
9
37
170
12
4
37
55
17
27
6
Benzene
Reductions
(%)
5%
3%
1%
0.1%
0.2*
1%
OK
1%
0.1%
1%
0.02%
1%
0.3%
0.1%
0.1%
1%
0.9%
2%
1%
0.4%
0.1%
0.1%
1%
0.2%
0.03%
0.1%
0.1%
3%
0.4%
0.1%
0.2%
3%
0%
0.3%
6%
0.1%
0.1%
0.4%
1%
0%
10%
1%
0.1%
0%
11%
0.05%
1%
0.2%
0%
Baseline
Benzene
Concn
(U0/m3)
0.64
1.47
0.57
2.56
1.26
1.58
1.99
3.33
1 15
101
095
1.37
0.98
0.47
0.56
0.74
1.54
058
1.81
1.37
1.32
0.99
0.70
1.02
0.78
0.49
1.13
0.98
2.76
0.78
2.75
0.91
0.51
1.21
0.58
1.53
1.90
1.38
0.79
0.46
1.05
1.91
1.16
0.66
1.01
1.85
0.91
0.83
0.54
Benzene
Conen
Reduction
(ugftn3)
0.04
0.05
0.005
0.002
0.003
0.01
0
0.04
0.001
0.01
00002
001
0.003
0.0005
0.0004
001
0.01
001
002
001
0.002
0.001
0.005
0.002
0.0002
0.001
0.001
0.03
0.01
0001
0.01
003
0
0.004
0.04
0.001
0.003
001
0.01
0
0.1
0.02
0.001
0
01
0.001
001
0.002
0
Benzene
•fc_ *— — -~- -
I'uinny
• Factor
(ugJMH
8.3E-06
8.3E-06
8.3E-06
8.3E-06
8.3E-08
8.3E-06
8.3&06
8.3E-06
8.3E-06
8.3E-06
8.3E-06
8.3E-06
8.3E-06
8.3E-06
83E-O6
83E-06
83E-06
83E-06
8.3E-06
8.3E-06
8.3E-O6
8.3E-06
8.3E-06
8.3E-06
8.3E-06
8.3E-06
8.3E-06
8.3E-06
8.3E-06
83E-06
83E-06
8.3E-06
8.3E-06
8.3E-06
8.3E-06
8.3E-06
8.3E-06
6.3E-06
8.3E-06
8.3E-06
8.3E-06
8.3E-06
8.3E-06
8.3E-06
8J&08
8.3E-06
83E-06
8.3E-06
8J&06
2010
POP
(x 10*3)
4.798
5.522
2.840
37.644
4,658
3.400
817
560
17.363
8.824
1.557
12.515
6.318
2.968
2.849
4.170
% 4.683
1.323
5.657
6.431
9.836
5.147
2.974
5.864
1.040
1.806
2.131
1.329
8.638
2.155
18.530
8.552
690
11.505
3.639
3.803
12.352
1.038
4.205
826
6.180
,22.857
2.551
651
7.627
6.658
1.851
5.590
607
Avoided
2010
Cases
0.02
003
000
0.01
0.00
0.00
0.00
0.00
0.00
0.01
0.00
001
000
0.00
0.00
001
0.01
000
0.01
0.00
0.00
0.00
000
0.00
0.00
0.00
0.00
0.00
0.01
000
001
0.03
0.00
0.01
0.02
0.00
0.00
0.00
000
0.00
0.08
0.04
0.00
0.00
0.10
0.00
0.00
.0.00
0.00
TOTAL CASES AVOIDED 0.4
rev 6/17/97
-------
,3-butadlene-.oe-3rdMAX
STATE
AL
AZ
AR
CA
CO
CT
OE
DC
FL
GA
ID
IL
IN
(A
KS
KY
LA
ME
MO
MA
Ml
MN
MS
MO
MT
NE
NV
NH
NJ
NM
NY
NC
ND
OH
OK
OR
PA
Rl
SC
SD
TN
TX
UT
VT
VA
WA
VW
Wl
WY
1,3-butadlene
Reductions
(tpd)
029
0.14
0.02
0.01
0.01
0.02
0.00
0.00
0.01
0.11
0.00
003
002
0.00
0.00
007
004
0.04
0.02
0.03
001
0.01
0.03
0.01
0.00
0.00
0.00
0.04
0.04
000
0.03
027
0.00
0.04
024
000
0.02
000
004
0.00
0.70
024
0.00
0.00
0.76
0.01
0.02
0.01
0.00
Baseline
1,3-butsdiene
Emissions
(tpd)
5
3
2
23
3
2
1
0
12
8
8
7
5
2
2
4
7
1
3
4
8
4
3
5
5
1
1
1
4
2
10
8
1
8
3
5
8
1
4
2
6
22
2
1
5
10
2
4
1
1.3-butadlene
Reductions
(%)
6%
4%
1%
0.03%
0.3*
1%
OH
1%
0.1%
1%
OK
0.5%
0.4%
0%
0%
2%
1%
3%
1%
1%
0.2%
0.1%
1%
0.3%
0%
0%
0%
4%
1%
0%
0.4%
4%
0%
0.5%
9%
0%
02%
1%
1%
0%
12%
1%
0%
0%
19%
0.1%
1%
0.3%
0%
Baseline
1 UmtediMM
Concn
(ugftnS)
0.07
0.11
0.05
021
0.12
0.13
0.'12
0.40
0.11
0.09
0.10
0.14
0.09
0.05
0.06
0.08
0.14
006
0.14
0.14
0.14
0.10
0.06
0.09
0.11
0.05
0.10
0.09
018
0.09
0.27
0.08
004
012
005
0.15
0.15
0.13
007
. 0.05
0.10
0.12
an
0.08
0.09
0.18
0.07
0.09
0.05
1>butadlene
Concn
Reduction
(ug/m3)
0.004
0.005
0.0004
0.0001
0.0004
0.001
0
0
0.0001
0.001
0
0001
00004
0
0
0.001
0.001
0002
0.001
0.001
0.0002
0.0001
0.0005
0.0002
0
0
0
0004
0002
0
0.001
0.003
0
0.001
0.004
0
0.0003
0
0001
0
0.01
0.001
0
0
0.01
0.0001
00007
00003
0
1,34MJtsdlono
Potency
Factor
(ugftidM
2.8E-04
2.6E-O4
£8E-04
2JE-04
2JE-04
2.8E-04
2.8E-04
2.8E-04
2JE-04
28E-04
2.8E-04
2.8E-04
2.8E-04
2.8E-04
2.8E-04
28E-04
2.6E-04
2.6E-04
2.8E-04
2.8E-04
2.8E-04
2.8E-O4
2.8E-04
2.8E-04
2.8E-O4
2.8E-04
2.8E-04
2.8E-04
28E-04
2.8E-04
2.8E-04
2.8E-04
2.8E-O4
2.8E-04
2.8E-04
2.8E-04
2.8E-04
Z8E-04
2.8E-04
2.8E-04
2.8E-04
2.8E-04
2.8&04
2.8E-04
2.8E-04
2.8&04
2.8E-04
2.8E-04
2.8E-04
2010
POP
(X 10*3)
4.798
5,522
2.840
37,644
4.658
3,400
817
560
17.363
8.824
1.557
12.515
6.318
2.968
2.849
4.170
4.683
1.323
5.657
6.431
9.836
5.147
2.974
5.864
1.040
1,806
2,131
1,329
8.638
2.155
16,530
8,552
690
11.505
3.639
3.803
12.352
1.038
4205
826
6.180
, 22.857
2,551
651
7.627
6.658
1.851
5.590
607
Avoided
2010
1.3-butadierw
Cases
0.08
0.10
0.01
0.01
0.01
0.01
000
001
0.01
004
000
003
0.01
000
0.00
002
0.01
001
002
002
0.01
0.00
0.01
0.01
0.00
0.00
0.00
0.02
005
0.00
0.07
010
0.00
003
006
000
0.02
0.00
0.01
0.00
029
0.12
0.00
0.00
0.40
0.00
0.00
0.01
0.00
TOTAL CASES AVOIDED 1 .6
r^ 6/17/9 7
-------
Primary Formald-.OS^rdMAX
STATE
AL
AZ
AR
CA
CO
CT
DE
DC
FL
GA
10
IL
IN
IA
KS
KY
LA
ME
MO
MA
Ml
MN
MS
MO
MT
NE
NV
NH
NJ
NM
NY
NC
NO
OH
OK
OR
PA
Rl
sc
so
TN
TX
UT
VT
VA
WA
WV
Wl
WY
Form (prim)
•Reduction*
(tpd)
0.24
0.12
0.02
0.11
0.01
0.01
0.00
0.00
0.01
009
0.00
0.04
003
0.00
0.00
006
0.03
003
0.02
0.02
0.01
0.00
0.03
0.01
0.00
0.00
0.00
. 0.03
V0.03
0.00
0.03
023
0.00
0.04
0.20
0.00
0.01
0.00
0.03
0.00
0.60
025
0.00
0.00
0.66
0.00
0.02
0.01
0.00
Baseline
Form (prtm]
Emissions
(tpd)
21
16
10
113
13
7
2
1
57
41
49
31
20
9
11
15
40
5
12
15
32
16
17
17
32
5
4
3
18
11
39
33
3
33
14
25
34
2
16
12
20
142
8
2
20
51
8
16
4
Form (prtm]
Reductions
(%)
1%
1%
03%
0.1%
0.1%
02%
0%
0%
0.02%
02%
0%
0.1%
0.1%
0%
0%
0.4%
0.1%
1%
02%
02%
0.04%
0%
02%
0.1%
0%
0%
0%
1%
0.2%
0%
0.1%
1%
0%
0.1%
1%
0%
004%
0%
0.2%
0%
3%
02%
0%
0%
3%
0%
0.3%
0.1%
0%
Baseline
Form (prtm]
Conen
(uo/mJ)
0.43
0.83
029
1.52
0.95
0.68
0.84
2.55
078
062
075
0.93
0.54
025
0.33
048
124
027
0.90
060
0.79
0.57
0.49
056
0.75
0.31
0.45
0.41
1.29
0.56
1.80
0.52
024
0.75
0.40
0.80
0.96
0.58
0.49
0.38
0.51
1.53
0,79
0.30
0.51
1.06
0.38
0.49
0.29
Form (prim)
Conen
Reduction
(ug/m3)
0.005
0.006
0.001
0.001
0.001
0.001
0
0
00001
0001
0
0001
0001
0
0
0.002
0.001
0002
0002
0.001
0.0003
0
0.001
0.0004
0
0
0
0.004
0.002
0
0001
0.004
0
0.001
0.006
0
0.0004
0
0.001
0
0.02
0.003
0
0
0.02
0
0.001
0.0003
0
Formsfd
Potency
. 'Factor
(ufl/mJM
13E-05
1.3E-05
13E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
13E-05
13E-05
13E-05
1.3E-05
13E-05
13E-05
13E-05
1.3E-05.
13E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
13E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3&06
1.3E-05
1.3E-05
1.3E-05
1 .IE-OS
1.3E-05
1.3E-05
1.3E-05
1.3E-05
2010
POP
(x 10*3)
4.798
5.522
2,840
37.644
4.658
3.400
817
560
17.363
8.824
1.557
12.515
6.318
2.968
2.849
4.170
4.683
1.323
5.657
6.431
9,836
5.147
2.974
5.864
1.040
1.806
2.131
1.329
8.638
2.155
18.530
8.552
690
11.505
3.639
3.803
12.352
1.038
4.205
626
6.160
22.657
2.551
651
7.627
6.658
1.851
5.590
607
Avoided
2010
Form (prim)
Cases
000
001
0.00
0.01
0.00
0.00
0.00
0.00
000
000
0.00
0.00
000
0.00
0.00
000
000
000
0.00
0.00
0.00
0.00
0.00
000
000
000
0.00
000
000
0.00
0.00
0.01
000
000
0.00
0.00
000
000
000
0.00
0.02
0.01
0.00
0.00
0.02
0.00
000
0.00
0.00
TOTAL CASES AVOIDED 0.1 1
r«v6/17f97
-------
Secondary FormaW-.08-3rdMAX
STATE
AL
AZ
AR
CA
CO
CT
DE
DC
FL
GA
ID
IL
IN
(A
KS
KY
LA
ME
MD
MA
Ml
MN
MS
MO
MT
NE
NV
NH
NJ
NM
NY
NC
NO
OH
OK
OR
PA
Rl
SC
SD
TN
TX
UT
VT
VA
WA
WV
W)
WY
Precursor
Reductions
(tpd)
487
2.43
0.33
0.42
0.15
0.28
0.01
0.05
0.21
1.90
0.02
0.70
0.39
0.05
004
1.20
0.63
069
0.50
0.47
0.23
0.09
0.44
0.22
0.02
0.03
0.02
062
061
0.03
058
4.54
0.01
0.67
401
0.05
0.30
0.07
0.68
002
11.89
4.36
0.03
0.01
13.00
0.09
035
022
001
Baseline
Precursor
Emissions
(tpd)
63
46
31
327
43
28
9
5
163
111
82
108
73
33
37
51
96
17
50
59
124
60
45
73
55
20
14
12
69
29
137
103
10
119
42
60
121
9
52
23
75
412
24
9
73
110
28
60
10
Precursor
Emission
Reductions
8%
5%
1%
0.1%
0.4%
1%
0.1%
1%
01%
2%
0.02%
1%
1%
0.2%
0.1%
2%
1%
4%
1%
1%
0.2%
0.1%
1%
0.3%
0.04%
0.1%
0.1%
5%
1%
0.1%
04%
4%
0.1%
1%
10%
0.1%
02%
1%
1%
0.1%
16%
1%
0.1%
0.1%
16%
0.1%
1%
0.4%
0.1%
Baseline
Form (Sec)
Concn
(ug/m3)
0.09
014
0.05
0.27
0.13
0.20
0.17
0.35
013
016
007
017
010
0.04
0.05
0.08
0.14
005
0.23
0.14
0.18
0.12
0.06
0.11
0.06
0.05
0.06
0.12
036
0.06
0.37
0.11
0.01
0.14
0.05
0.18
0.20
012
0.10
002
0.12
028
0.10
0.04
0.12
025
008
007
0.01
Form (Sec]
Concn
Reduction
(ug/m3)
0007
0.007
0.001
0.0003
0.0005
0.002
00002
0004
00002
0003
000002
0001
0001
0.0001
0.0001
0.002
0.001
0002
0.002
0.001
00003
00002
0.001
00003
000002
0.0001
00001
0.006
0003
00001
0.002
0.005
000001
0001
0.005
0.0002
00005
0001
0001
000002
0.02
0.003
0.0001
000005
0.021
00002
0001
0.0003
000001
Formald
Potency
Factor
(ug/m3H
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
13E-05
1.3E-05
1.3E-05
1.3E-05
13E-05
1.3E-05
1.3E-05
1.3E-05
13E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1 3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
13E-05
1.3E-O5
13E-05
2010
POP
(X 10*3)
4.798
5.522
2.840
37.644
4.658
3,400
817
560
17.363
8.824
1.557
' 12.515
6.318
2.968
2.849
4.170
%4.683
1 1.323
5.657
6.431
9.836
5,147
2.974
5.864
1.040
1,806
2.131
1.329
8.638
2.155
18.530
8,552
690
11.505
3.639
3.603
12.352
1.038
4.205
826
6,180
22.857
2.551
651
7.627
6.658
1.851
5.590
607
Avoided
2010
Form (Sec]
Cases
001
0.01
0.00
0.00
0.00
0.00
000
0.00
000
000
000
000
000
000
000
000
000
000
000
0.00
000
0.00
000
000
0.00
0.00
000
000
001
0.00
0.01
0.01
000
000
000
000
000
000
000
000
002
0.01
0.00
0.00
0.03
0.00
000
000
000
TOTAL CASES AVOIDED 0.1
r«« 6/17*97
-------
Appendii 2
HAP Emissions and Concentration Data
-------
ICF KAISER
Consulting Group
Systems Applications International, Inc.
MEMORANDUM
TO: Jim Wilson, E.H. Pechan and Associates
FROM: Arlene Rosenbaum and YiHua Wei
DATE: 14 March 1997
SUBJECT: EPA contract 68-W6-0028; WA 05; Task 2: Baseline hazardous air
pollutant (HAP) data
The emissions and concentration data provided here is derived from estimates for 1990
developed by SAI for the Cumulative Exposure (CE) Project for EPA OPPE. The data bases and
methods used to develop the CE data are described in the May 1996 draft report Revised
Methodology or modeling Cumulative Exposures to Air Toxics I: Outdoor Concentrations,
Systems Applications International, SYSAPP-96/33d
Tables 1 through 4 summarize estimates of emissions of benzene, 1,3-butadiene, formaldehyde
and formaldehyde precursors, respectively, at the state level of spatial aggregation, for the 48
contiguous states and the District of Columbia. The emissions of formaldehyde precursors are
reported as a composite, which is a weighted sum with weights based on molar yield and reaction
rates. The weighting procedure was developed for the Cumulative Exposure Project, and is
presented in Table 3-1 of the CE report noted above.
Estimates for specific source categories are also presented. The source categories are as follows:
• TRI non-metal: data from the 1990 Toxic Release Inventory for SICs 2000-3999, excluding
refineries (SIC 2911) and metal production (SICs 3300-3499)
• Other points: data from 1990 EPA National Interim Inventories and speciation profiles for
SICs 0100-1999,4000-9999, and combustion sources in SICs 2000-3999 (except refineries)
• Refineries: data from 1990 EPA National Interim Inventories and speciation profiles for SIC
2911
• Onroad mobile: data from 1990 EPA National Interim Inventories and speciation profiles
• Nonroad mobile: data from 1990 EPA National Interim Inventories and speciation profiles
101 Lucas Valley Road. San Rafael. CA 94903 * Telephone (415) 507-7100 • Fax 507-7177
-------
PAGE 2
• Manufacturing area: data from 1990 EPA National Interim Inventories and speciation profiles
for industrial activities
• Non-manufacturing area: data from 1990 EPA National Interim Inventories and speciation
profiles for non-industrial activities
• Municipal waste combustors (MWCs): data from a national inventory of MWCs, and
speciation profiles from EPA's MWC study
• Hazardous Waste TSDFs: -data from an EPA study of TSDFs
• TRI metals: data from the 1990 Toxic Release Inventory for metal production (SICs 3300-
3499)
The results show that the onroad mobile source category is the largest contributor of benzene (50
percent), 1,3-butadienc (41 percent), and formaldehyde precursors (48 percent). The
nonmanufacturing area source category is the largest contributor of formaldehyde (39 percent).
For all four HAPs the highest emissions totals are in Texas and California. Texas emissions
constitute approximately 11 percent of benzene, 9 percent of 1,3-butadiene, 13 percent of
formaldehyde, and 12 percent of formaldehyde precursors. The corresponding values for
California are approximately 10 percent for each HAP. In addition to mobile source
contributions, Texas shows particularly high emissions from refineries, while California shows
high emissions from nonmanufacturing area sources.
Table 5 presents estimates of population-weighted long-term average HAP concentrations of
benzene, 1,3-butadiene, primary formaldehyde, and secondary formaldehyde by state. The
averages are derived from census tract concentration estimates and 1990 population data form the
US Census. The highest concentrations of benzene, 1,3-butadiene and primary formaldehyde are
in D.C., while the highest concentrations of secondary formaldehyde are in New York.
Systems Applications International, Inc.
101 Lucas Valley Road. San Rafael. CA 94903 • Telephone (415)507-7100 • Fax 507-7177
-------
PAGE 3
Table I. Estimates of benzene emissions by state and source category (tpd).
State
AL
AZ
AR
CA
CO
CT
DE
DC
FL
GA
ID
IL
IN
IA
KS
KY
LA
ME
MD
MA
Ml
MN
MS
MO
MT
NE
NV
NH
NJ
NM
NY
NC
ND
OH
OK
OR
PA
RI
SC
SD
TN
TX
UT
VT
VA
WA
WV
Wl
WY
Total
FTPS TRI
Nonmet
I
4
5
6
8
9
10
11
12
13
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33.
34
35
36
37
38
39
40
41
42
44
45
46
47
48
49
50
51
53
54
55
56
0.22
0.00
0.04
0.20
0.00
0.01
0.06
0.00
0.00
1.05
0.00
0.22
0.04
0.03
0.00
0.11
1.06
0.00
0.10
0.00
0.12
0.00
0.04
0.03
0.00
0.00
0.00
0.00
0.03
0.00
0.0 1
0.27
0.00
0.19
0.00
0.00
0.15
0.00
0.95
0.00
024
2.84
0.00
0.00
0.68
0.29
0.12
0.01
0.00
9.15
Other
Point
2.21
0.47
1.01
1.33
0.50.
0.16
0.07
0.02
1.68
2.35
0.11
2.44
1.93
0.40
0.57
1.35
1.15
0.40
0.91
0.17
129
0.80
0.71
1.31
0.39
0.27
020
0.1 1
1 15
0.91
4.96
1.00
0.75
1.99
0.72
0.32
1.52
0.00
0.53
0.16
1.79
47.76
0.58
0.01
0.85
1.49
1.40
0.94
0.60
93.69
Ref Onroad Nonrd Manuf Nonmn MWC TSDF TRI
Mobile Mobile Area Area Metal
0.42
0.00
0.30
0.98
0.08
0.00
0.11
0.00
0.00
0.04
0.00
1.43
0.41
0.00
0.61
0.27
1.45
0.00
0.00
0.00
0.31
0.53
0.44
0.01
027
0.00
0.00
0.00
0.95
0.00
0.03
0.00
0.00
0.22
026
0.00
048
0.00
0.00
0.00
0.00
12.92
0.18
0.00
0.10
0.60
0.17
0.04
0.93
24.55
15.81
1220
6.96
91.03
12.13
8.10
2.17
1.61
40.95
27.40
4.89
28.46
18.94
8.33
9.65
12.43
1428
4.61
16.92
17.81
36.07
15.78
8.85
17.09
3.72
521
4.02
3.71
19.43
6.06
40.91
26.15
2.66
33.08
1123
11.62
32.30
2.63
11.79
2.86
20.73
60.69
5.68
2,76
20.02
18.82
5.58
15.73
2.23
802.12
2.93
5.81
127
35.39
3.88
3.86
.0.66
028
9.17
4.48
2.32
4.19
2.50
122
1.17
2.69
3.93
0.76
4.05
7.92
6.92
3.19
2.84
227
1.05
0.76
2.06
0.79
6.58
1.02
11.85
4.57
0.75
4.31
1.75
6.50
926
1.54
2.35
l.ll
3.47
14.11
2.07
034
4.96
11.15
1.42
3.14
0.57
210.19
0.94
0.02
0.93
8.40
023
0.94
1.08
0.00
0.66
0.82
0.02
4.08
2.27
0.22
0.94
1.12
9.72
0.00
0.33
0.64
3.94
0.68
1.52
1.73
0.36
0.12
0.03
OJO
8.36
029
2.57
1.85
026
3.40
I.S5
0.05
5.69
021
0.96
0.03
2.08
18.85
0.34
0.01
1.56
1.69
3.07
0.68
0.14
95.66
6.62
4.81
3.85
29.32
3.82
2.80
0.77
0.19
15.45
12.83
21.83
9.17
7.88
3.54
3.42
5.83
9.89
2.59
2.45
4.70
9.45
6.59
635
7.48
13.85
2.04
1.09
1.57
5.38
323
13.38
13.72
1.31
11.48
3.92
11.60
13.34
0.75
6.92
5.00
8J1
13.06
2.73
122
8.92
20.83
3.34
6.63
1.33
356.58
001
0.00
0.02
0.02
0.00
0.13
0.01
0.00
0.05
0.00
0.00
0.01
0.00
0.00
0.00
0.00
0,00
6.01
0.02
0.11
0.04
0.03
0.00
0.00
0.00
0.00
0.00
0.03
0.03
0.00
0.07
0.00
0.00
0.01
0.01
0.02
0.02
0.00
0.02
0.00
0.01
0.01 '
0.00
0.00
0.03
0.00
0.00
0.01
0.00
0.78
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
000
0.05
0.00
0.03
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.01
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.01.
0.00
0.00
0.00
0.00
0.01
0.00
0.00
0.00
0.00
000
0.00
000
0.12
0.93
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
1.03
3.40
0.00
0.00
020
0.00
0.00
0.10
0.00
0.59
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.51
0.00
0.00
2.91
0.00
0.00
3.89
0.00
0.00
0.00
0.00
0.00
0.12
0.00
0.00
0.00
2.12
000
0.00
15.80
Total
30.09
23.31
14.39
166.67
20.64
16.00
4.93
2.11
67.96
48.99
29.18
51.04
37.37
13.75
16.36
24.00
41.48
8.38
24.94
31.35
57.75
27.61
20.75
29.92
19.64
8.39
7.40
6.51
41.91
11.52
74.29
47.56
5.73
57.59
19.43
30.10
66.66
5.13
23.52
9.16
36.64
17026
11.70
4J4
37.12
54.87
1723
27.18
5.79
1608.64
Systems Applications International, Inc.
101 Lucas Valley Road. San Rafael. CA 94903 * Telephone (415) 507-7100 • Fax 507-7177
-------
PAGE 4
Table 2. Estimates of 1,3-butadiene emissions by state and source category (tpd).
State
AL
AZ
AR
CA
CO
CT
DE
DC
FL
GA
ID
IL
IN
IA
KS
KY
LA
ME
MD
MA
MI
MN
MS
MO
MT
NE
NV
NH
NJ
NM
NY
NC
ND
OH
OK
OR
PA
RI
SC
SD
TN
TX
UT
VT
VA
WA
WV
WT
WY
Total
FIPS TRI Other
Nonmet Point
1
4
5
6
8
9
10
11
12
13
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
44
45
46
47
48
49
50
51
53
54
55
56
0.03
0.00
0.00
0.00
0.00
0.08
0.04
0.00
0.00
0.07
0.00
0.05
0.00
0.24
0.00
0.43
0.42
0.00
0.00
000
0.04
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.10
0.01
0.00
0.35
0.00
0.00
0.09
0.00
0.01
0.00
031
3.98
0.00
0.00
0.00
0.00
0.32
000
0.00
6.58
0.33
0.03
0.08
021
0.03.
0.02
0.01
0.00
0.24
0.28
0.02
0.32
0.12
0.03
0.03
0.06
0.09
0.10
0.04
0.03
0.15
0.06
0.07
0.30
0.06
0.01
0.01
0.03
0.02
0.09
0.26
0.08
0.05
0.14
0.04
0.05
0.11
0.00
0.07
0.01
0.24
3.82
0.03
0.00
0.08
0.30
0.07
0.08
0.04
8.34
Ref Onroad Nonrd Manuf Nonmn MWC TSDF TRI
Mobile Mobile Area Area Metal
0.00
0.00
0.00
0.01
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.04
0.00
0.00
0.00
0.00
0.05
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.02
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.14
1.85
1.22
0.84
10.06
1.49
0.99
0.27
0.20
4.84
3.37
0.52
3.69
2.46
1.07
126
1.58
1.73
0.58
2.07
2.21
4.76
2.08
1.04
222
0.46
0.68
0.47
0.46
2.34
0.82
5.11
3.17
0.33
4.23
1.42
1.24
4.03
0.33
1.38
03*
2.51
735
0.69
0.35
2.44
2.03
0.68
2.08
0.28
97.62
0.90
0.91
0.40
525
0.71
0.40
0.11
'0.09
2.95
1.21
0.20
1.42
0.79
0.41
0.48
0.80
1.58
0.10
0.74
0.95
129
0.76
0.76
0.84
0.20
031
033
0.07
0.95
037
1.64
122
0.14
130
0.58
0.53
1.49
0.13
0.59
0.14
0.88
4.81
033
0.04
0.93
0.92
035
0.73
0.12
43.17
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
1.60
1.14
0.73
7.84
0.74
0.51
0.14
0.02
3.98
3.53
7.43
1.70
1.49
0.57
0.69
1.26
2.84
0.52
0.31
0.87
1.93
123
1.53
138
4.66
032
0.23
032
0.92
0.72
2.51
3.06
0.16
232
0.70
330
2.72
0.13
1.86
1.53
1.93
1.55
0.65
029
1.75
6.57
0.72
1.32
0.27
84.48
0.00
0.00
0.00
0.00
0.00
0.02
0.00
0.00
0.01
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0*00
0.00
0.02
0.01
0.01
0.00
0.00
0.00
0.00
0.00
0.01
0.00
0.00
0.01
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00'
0.00
0.00
0.01
0.00
0.00
0.00
0.00
0.11
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
o.od
0.00
0.00
0.00
0.00
0.00
0.00
0.00
000
0.00
0.00
0.00
000
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
000
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
000
Total
4.71
331
2.06
2338
2.97
2.02
0.57
031
12.02
8.45
8.18
7.21
4.86
232
2.46
4.13
6.71
1.31
3.16
4.08
8.18
4.13
3.40
4.74
538
132
1.04
0.90
424
2.00
9.62
7.54
0.67
8.34
2.75
5.13
8.44
0.58
3.91
2.03
5.87
21.54
1.70
0.68
520
9.83
2.14
421
0.71
24044
Systems Applications International, Inc.
101 Lucas Valley Road. San Rafael. CA 94903 * Telephone (415) 507-7100 • Fax 507-7177
-------
PAGE 5
Table 3. Estimates of formaldehyde emissions by state and source category (tpd).
State
AL
AZ
AR
CA
CO
CT
DE
DC
FL
GA
ID
IL
IN
IA
KS
KY
LA
ME
MD
MA
MI
MN
MS
MO
MT
NE
NV
NH
NJ
NM
NY
NC
ND
OH
OK
OR
PA
RI
SC
SD
TN
TX
UT
VT
VA
WA
WV
WI
WY
Total
FIPS TRI
Nonmet
I
4
5
6
8
9
10
11
12
13
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
44
45
46
47
48
49
50
SI
53
54
55
56
0.58
0.00
0.40
0.53
0.23
0.26
0.01
0.00
0.06
068
0.12
0.18
030
0.01
0.33
0.07
0.27
0.21
0.01
0.24
0.52
0.91
0.29
0.56
0.24
0.00
0.00
0.01
0.19
0.12
029
1.67
0.00
U9
0.01
1.09
0.32
0.00
0.70
0.12
0.48
1.64
0.00
0.00
0.82
0.15
0.26
0.55
001
16.60
Other
Point
1.84
0.15
0.78
1.55
0.15 .
O.IS
0.02
0.01
1.32
3.84
0.08
1.04
0.76
0.07
0.19
0.86
409
0.77
0.13
0.09
2.14
0.31
0.62
0.22
0.33
0.04
0.04
0.06
0.39
098
4.78
1.23
0.16
0.47
0.77
0.73
0.59
0.01
0.12
0.01
0.75
69.89
0.17
0.00
0.33
3.40
1.26
0.42
0.16
108.30
Ref Onroad Nonrd Manuf Nonmn MWC TSDF TRJ
Mobile Mobile Area Area. Metal
0.02
0.00
0.08
0.45
0.06
0.00
0.01
0.00
0.06
0.00
000
1.18
0.03
0.00
0.15
0.05
0.29
0.00
0.00
0.00
0.23
0.09
0.04
0.00
0.07
0.00
0.00
0.00
0.04
0.01
0.01
0.00
0.00
0.19
0.10
0.00
0.06
0.00
0.00
0.00
0.00
3.57
0.08
0.00
0.10
0.08
0.00
0.03
0.62
7.71
5.96
4.67
3.08
36.83
5.63
3.09
0.94
0.65
15.91
11.12
1.84
12.71
9.20
3.83
4.29
520
6.31
1.86
6.92
6.90
16.34
6.94
3.55
7.71
1.61
2.37
1.78
1.46
7.49
3.66
16.48
10.74
1.13
14.65
4.85
4.45
13.49
1.04
4.55
1.22
8.18
25.78
2.66
1.10
827
7.15
2.34
7.11
0.99
336.04
4.10
4.32
1.63
22.44
3.52
1.45
0.48
0.57
15.55
5.04
0.96
5.62
2.69
1.40
1.66
2.84
5.50
0.59
325
3.67
4.92
2.79
2.81
3.19
0.83
1.13
1.76
0.38
3.70
1.79
6.47
5.18
0.69
4.81
2.52
2.08
5.60
0.45
2JI
0.66
3.07
20.42
U6
022
3.45
3.96
1.03
2.53
0.42
177.79
0.12
0.01
0.05
6.35
0.26
0.05
0.37
0.00
0.02
0.05
0.01
2.88
1.04
0.04
0.58
0.65
6.17
000
0.03
0.09
0.37
0.65
1.75
0.06
0.42
0.02
0.03
0.00
1.50
0.39
0.01
0.08
0.30
1.84
1.84
0.01
2.45
0.02
0101
0.01
0.36
7.65
0.39
0.01
0.32
1.54
0.19
0.14
0.16
41.31
8.67
6.72
3.78
45.16
3.57
1.94
0.56
0.14
23.93
20.47
45.83
7.15
6.20
3.38
427
5.07
17.71
1.84
2.06
3.70
7.80
4.50
8.30
•5.14
28.47
1.86
0.86
1.19
3.95
3.59
10.76
13.75
0.99
9.58
3.39
1649
11.55
0.57
10.52
9.48
7.13
13.01
3.67
0.95
6.46
3456
2.95
5 16
1.66
440.47
0.00
0.00
0.01
0.01
0.00
0.05
0.00
0.00
0.02
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
o.bo
0.01
0.04
0.02
0.01
0.00
0.00
0.00
0.00
0.00
0.01
0.01
0.00
0.03
0.00
0.00
0.00
0.00
0.01
0.01
0.00
0.0 1
0.00
0.00
0.0 1 '
0.00
0.00
0.01
0.00
0.00
0.00
0.00
028
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.01
0.00
0.00
0.01
0.00
000
0.00
0.01
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.28
0.00
0.10
0.01
0.00
0.02
0.07
000
000
0.00
0.00
0.00
0.00
0.11
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.63
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.02
0.04
0.06
0.00
0.00
0.01
0.00
0.00
0.00
0.15
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
007
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.03
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.38
Total
21.29
15.87
9.82
113.33
13.42
6.99
2.39
1.36
56.86
41 20
48.85
30.79
20.27
8.79
11.47
14.77
40.35
5.27
12.41
14.75
32.49
16.19
17.37
16.88
31.97
5.42
4.46
3.12
17.55
10.54
38.92
32.66
3.26
32.82
13.56
24.87
34.08
2.09
18.21
11.51
19.98
142.10
8.33
2.29
19.77
50.84
8.03
1595
402
1129.51
Systems Applications International, Inc.
101 Lucas Valley Road, San Rafael. CA 94903 * Telephone (415) 507-7100 • Fax 507-7177
-------
PAGE 6
Table 4. Estimates of formaldehyde precursor emissions by state and source category (tpd).
State
AL
AZ
AR
CA
CO
CT
DE
DC
FL
GA
ID
IL
IN
IA
KS
KY
LA
ME
MD
MA
MI
MN
MS
MO
MT
NE
NV
NH
NJ
NM
NY
NC
ND
OH
OK
OR
PA
Rl
SC
SD
TO
TX
UT
VT
VA
WA
WV
WI
WY
Total
FIPS TRI
Nonmet
1
4
5
6
8
9
10
11
12
13
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
44
45
46
47
48
49
50
51
53
54
55
56
0.01
0.00
0.07
0.01
0.00
0.00
0.00
000
0.00
0.02
0.00
2.24
0.04
0.74
0.01
1.57
6.34
0.00
0.00
0.00
0.19
000
0.00
0.00
0.00
0.00
0.00
0.00
0.24
0.00
0.00
0.03
0.00
020
0.20
0.00
0.00
0.01
0.32
0.00
0.09
2733
0.00
0.00
0.07
0.00
0.66
000
000
40.40
Other
Point
2.64
0.14
0.99
2.67
0.15.
0.23
0.05
0.01
1.85
2.74
0.16
2.46
1.71
0.09
0.46
0.43
4.32
0.86
0.38
0.22
1.03
0.64
1. 11
7.82
0.40
0.09
0.05
0.23
0.69
222
2.88
0.78
025
2.09
1.03
0.53
0.93
0.01
0.43
0.11
1.50
145.80
0.53
0.02
0.68
2.66
2.06
0.52
0.28
19990
Ref Onroad Nonrd Manuf Nonmn MWC TSDF TRJ
Mobile Mobile Area Area Metal
0.28
0.00
0.09
1.32
0.02
0.00
0.06
0.00
000
000
000
i.04
0.13
0.00
0.41
0.10
1.30
0.00
0.00
0.00
0.10
0.17
0.20
0.01
0.06
0.00
0.00
0.00
0.28
0.00
0.03
0.00
0.00
0.16
0.17
0.00
0.20
0.00
0.00
0.00
0.01
1033
0.15
0.00
0.02
0.28
0.13
0.00
0.27
1733
30.97
21.60
14.55
173.24
25.19
16.06
4.50
331
80.94
55.68
8.94
62.36
^1.74
18.26
2131
2588
29.86
939
34.58
35.73
80.11
35.03
17.41
37.61
7.79
11.49
8.00
7.48
38.07
13.82
82.43
52.75
5.51
71.84
24.18
21.20
66.96
528
23.10
5.90
41.62
126.88
11.76
5.62
40.83
34.60
11.35
34.95
4.69
164237
11.47
11.82
5.63
66.52
9.23
5.81
1.57
1.09
37.57
16.12
2.73
1908
10.92
5.93
626
10.02
19.46
1.33
10.09
13.19
19.61
11.04
9.98
11.65
2.48
4.07
4.10
1.09
12.73
4.86
22.57
16.40
1.70
18.46
7.95
7.17
21.12
1.97
8.13
1.98
11.97
64.62
4.41
0.53
1234
12.42
4.10
10.72
1.43
577.43
0.76
0.01
0.77
4.18
0.09
0.78
0.78
0.00
0.55
068
001
2.55
1.60
0.17
0.62
0.73
6.40
0.00
0.27
0.51
323
036
0.71
1.45
0.14
0.09
0.01
025
6.47
0.13
2.18
1.54
0.10
230
0.76
0.03
3.88
0.17
0.82
0.02
1.63
13.59
0.13
0.01
121
0.58
2.53
0.53
0.05
66.36
16.47
1222
8.43
78.65
8.09
532
1.54
033
42.48
36.11
70.29
18.57
16.18
7.58
8.06
1221
27.95
5.18
4.82
933
19.67
13.15
15.79
14.67
44.17
432
221
321
1046
7.47
26.95
3128
2.42
24.11
7.79
30.61
27.51
1.44
18.89
15.11
18.05
23.65
6.79
2S2
18.05
5924
7.18
13.60
2.78
862.89
002
0.00
0.03
0.03
0.00
0.19
0.02
0.00
008
001
000
0.02
0.00
001
0.00
0.01
000
e!bi
0.03
0.16
0.06
0.05
0.00
0.00
0.00
0.00
0.00
0.05
0.05
000
0.11
000
000
0.02
0.01
0.03
0.03
000
003
0.00
0.02
0.02
0.00
0.01
005
000
000
002
000
1.19
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
000
0.00
0.00
c.oo
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
000
0.00
0.00
0.00
0.00
000
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
000
000
000
0.00
000
0.04
0.00
0.00
0.00
0.00
0.00
0.00
0.00
000
000
000
0.06
C.19
0.00
000
0.01
0.00
0.00
0.00
000
0.01
0.00
0.00
0.00
0.00
0.00
0.00
000
000
0.00
002
0.00
0.00
006
0.00
000
0.13
0.00
0.00
0.00
0.00
0.00
0.02
0.00
0.00
0.00
036
0.00
000
089
Total
62.66
45.79
30.58
326.61
42.78
2839
8.51
4.74
163.46
111 34
82 13
108.37
72.51
32.79
37.12
5095
95.63
1678
50.17
59.15
124.00
60.45
45.21
73.21
55.04
20.06
1437
1232
68.98
28.50
137.17
102.78
9.98
119 24
42.09
5958
120.77
8.87
51.72
23.12
74.89
41222
23.79
8.71
7325
10978
28.37
60.34
9.51
3408.77
Systems Applications International. Inc.
101 Lucas Valley Road. San Rafael. CA 94903 + Telephone (415) 507-7100
Fax 50T-7177
-------
PAGE 7
Table S. Estimates of population-weighted average concentrations of benzene. 1.3-
butadiene. primary formaldehyde, and secondary formaldehyde by state (ng/m').
State
AL
AZ
AR
CA
CO
CT
DE
DC
FL
GA
ID
IL
IN
IA
KS
KY
LA
ME
MD
MA
MI
MN
MS
MO
MT
NE
NV
NH
NJ
MM
NY
NC
ND
OH
OK
OR
PA
RI
SC
SD
TN .
TX
UT
VT
VA
WA
WV
WI
WY
FIPS
1
4
5
6
8
9
10
11
12
13
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
44
45
46
47
48
49
50
51
53
54
55
56
Benzene
0.84
1.47
0.57
2.56
126
1.58
1.99
3 33
1.15
1.01
095
1.37
0.98
0.47
0.56
074
1.54
0.58
1.81
1.37
1.32
0.99
0.70
1.02
0.78
0.49
1.13
098
2.76
0.78
2.75
0.91
0.51
1.21
058
1.53
1.90
U8
0.79
0.46
1.05
1.91
1.16
0.66
1.01
1.85
0.91
083
054
1,3 -butadiene
0.07
0.11
0.05
021
0.12
0.13
0.12
040
0.11
0.09
0.10
0.14
0.09
0.05
0.06
008
0.14
0.06
0.14
0.14
0.14
0.10
006
0.09
0.11
005
0.10
0.09
0.18
0.09
0.27
0.08
0.04
0.12
005
0.15
0.15
0.13
0.07
0.05
0.10
0.12
0.11
0.08
0.09
0.18
0.07
0.09
005
Primary
Formaldehyde
0.43
0.83
029
1.52
0.95
0.68
0.84
2.55
0.78
042
0.7:
0.93
0.54
025
0-V
0*48
124
027
0.90
0.60
0.79
0.57
0.49
0.56
0.75
0.31
0.45
041
129
056
1.80
0.52
0.24
0.75
040
0.80
0.96
0.58
0.49
038
0.51
1.53
0.79
0.30
0.51
106
038
049
029
Secondary
Formaldehyde
0.09
0.14
0.05
027
0.13
020
0.17
0.35
0.13
0.16
0.07
0.17
0.10
0.04
0.05
008
0.14
0.05
023
0.14
0.18
0.12
0.06
0.11
0.06
0.05
0.06
0.12
OJ6
0.06
0.37
0.11
0.01
0.14
O.C5
0.18
0.20
0.12
0.10
0.02
0.12
028
0.10
0.04
0.12
025
008
007
001
Systems Applications International, Inc.
101 Lucas Valley Road. San Rafael. CA 9-J903 * Telephone (415) 507-7100 • Fax 507-^17'
-------
July 9,1997
Mr. Daniel Axelrad (2126)
U.S. Environmental Protection Agency
401 M Street, S.W.
Washington, DC 20460
SUBJECT: EPA Contract No. 68-W6-0028, WA No. 5
Task 4b Memorandum - HAP Emission Estimates
Dear Dan:
Attached is the revised technical memorandum that explains how the HAP speciation was
performed for 2010 VOC emission changes. ,»
The incremental VOC emission reductions by control measure to meet the current
standard and each of the three ozone alternatives are speciated using the profiles in Table 1,
resulting in the benzene, 1,3-butadiene, formaldehyde, and "other HAP" emission reductions
shown in Tables 3 through 6.
Please call me at extension 102 with any questions about this information.
Sincerely,
Jim Wilson
JW/daw
-------
TECHNICAL MEMORANDUM TASK 4B:
SPECIATION OF YEAR 2010 VOC EMISSION REDUCTIONS
The weight fractions of benzene, 1,3-butadiene, formaldehyde, formaldehyde precursors,
and total other hazardous air pollutants (HAPs) for specific source category/control measure
pairings potentially implemented under a revised ozone national ambient air quality standard are
given in Table 1. These weight fractions are based on Pechan's best estimate of the volatile
organic compound (VOC) composition for these source categories in the year 2010. The
following sections describe the" methods used to develop HAP weight fractions, and to estimate
HAP emission reductions using VOC emission reductions in 2010 for each of the three ozone
alternatives. The general method for estimating 2010 HAP emission reductions is to multiply the
VOC emission reductions associated with each ozone NAAQS alternative by the HAP profiles
shown in Table 1.
A. DERIVATION OF SPECIATION PROFILES BY SOURCE CATEGORY
A description of the data sources used and assumptions made to derive the speciation
estimates that differ from the profiles obtained from SAI (and used in Task 2) is given below. It
should be noted that, for all of the sources described below, different speciation profiles are
recommended for use in modeling 2010 reductions, than were used to develop the ambient
concentrations under Task 2. • *
HAPs included in the other HAP category include any Title Hi-listed volatile organic
HAP. Common volatile organic HAPs include: toluene, xylenes, and glycol ethers from the
surface coating categories; acetaldehyde, ethylbenzene, n-hexane, xylenes, and toluene from
gasoline exhaust; and ethylbenzene, n-hexane, xylenes, toluene, and 2,2,4-trimethylpentane from
gasoline evaporation. Pechan/SAI reviewed the full speciation profiles for each source category
and totaled the weight fractions of these other HAPs in order to obtain the other HAP weight
fraction shown in Table 1. This memorandum (and the analysis) focuses on source categories
that are in the ozone NAAQS RIA control measure data file.
1. Web Offset Lithography
Pechan derived a new weight fraction for total other HAP (emissions of the other species
are not expected from this source category). Prior to issuance of the Control Technique
Guidelines (CTG), there were no significant HAP/precursor emissions for this source category
(STAPPA/ALAPCO, 1993). Following issuance of the CTG, reformulated fountain solutions
became commonly used hi an effort to reduce VOC emissions. These reformulated products
typically contain listed HAPs (e.g., glycols, glycol ethers) in place of the previously used
alcohols (EPA, 1995).
Pechan accounted for changes in the overall makeup of VOC emissions by assuming a 50
percent reduction in the VOC content of fountain solutions (STAPPA/ALAPCO, 1993),
rescaling the emissions for the source category (i.e., determining the fraction of VOC contributed
by fountain solutions), and a«"ming that all of me fountain solution VOC is made up of glycols
and glycol ethers. This approach lead to a total other HAP weight fraction of 0.615, which is
slightly less than the profile used by SAI to derive ambient concentrations under Task 2 (0.674).
2. Nonroad Engines
For nonroad gasoline engines, the profile is the EPA regional average for 4-stroke
gasoline engines. Although the reformulated gasoline control measure is likely to also result in
-------
changes to the final HAP speciation, data were not available to account for these changes.. See
the discussion under highway vehicles for additional information about the method used to
quantify speciated reductions.
*
3. Bulk Terminals and Service Stations
The SAI-derived regional profiles for summer blend gasoline were avenged to develop
the profile for this category (emissions from gasoline storage are assumed to represent the
majority of bulk terminal V0€ emissions). It should be noted that a different SAI profile
(#9968) was used to estimate ambient concentrations under Task 2. This profile was a petroleum
industry average containing formaldehyde, which is not a likely constituent of bulk terminal
emissions. The same profile is used for service station emissions speciation.
4. Surface Coating Sources
Estimating changes in speciation of future VOC emissions for most of these source
categories is not possible. Therefore, the SPECIATE profile has been used. Most of the VOC
should be made up of some petroleum distillate, such as mineral spirits, which contain only about
5% of other HAP. The reformulation profile shown in Table 1 for surface coating source
categories is also used for open top/conveyorized degreasing, since both are based on
manufacturers finding a non-HAP replacement for some, or all, of the petroleum distillate
solvent. Petroleum distillates (generic term for mineral spirits, stoddard solvent, and similar
solvents) are assumed to be by far the largest source of VOC reductions. They are currently used
both as a cleaning solvent and as a paint thinner. They will probably be replaced by water-borne
(aqueous) solvent reformulations, which still have some VOC in them. This remaining VOC
(e.g., 5-15% by weight versus 100% by weight for petroleum distillates) may still contain some
HAP. Since there is no way to predict what the reformulations will look like in 2010, we applied
the existing SAI profile for reformulated solvents. [NOTE: petroleum distillates are 100% VOC
and about 5% HAP, this is the baseline. In 2010, reformulated solvent with 58% HAP is
assumed to be in use.] Depending on the specifics of coating products and future available
reformulations for a given source category, the HAP content could increase or decrease. For
many of the categories, reformulation to water-based formulations may occur, and currently the
VOC in these formulations is made up of glycols and glycol ethers (the VOCs have 100% HAP
content). For other categories, such as marine surface coating, where water-based formulations
are not technically-feasible, the HAP content of the solvent used in the paint may be reduced
(e-g-» by switching from an aromatic to an aliphatic solvent).
For marine surface coating, information was found to adjust the total HAP fraction of the
VOC emissions as a result of MACT implementation (EPA, 1994). For the remaining source
categories, it is assumed that existing profiles will provide reasonable estimates of 2010 emission
reductions.
5. Highway Vehicles
HAP speciation for gasoline-powered motor vehicles is taken largely from information
obtained from EPA OMS based on complex model runs, and documented in a letter to the WAM
(Wilson, 1997). The OMS data contained estimated HAP speciation for exhaust and non-exhaust
VOC from the use of Federal RFG H, California RFC, and 2010 gasoline. Since the OMS data
contained only acetaldehyde as a formaldehyde precursor, additional precursor species were
added in to the profile using a profile from SPECIATE for a light duty catalyst-equipped engine,
and the SAI profile for summer blend gasoline.
-------
Since the SPECIATE and SAI data were used to fill in data gaps in the speciation for all
three gasolines, the weight fractions for total other HAPs and formaldehyde precursors are
similar for all three. In adding the SPECIATE/SAI data, Pechan used a non-exhaust fraction of
26 percent for Federal RFC II and 2010 baseline gasoline and 22 percent for California RFC
(EPA, 1993).
In the 2010 baseline, all non-California States will have either all 2010 baseline gasoline,
all Federal RFC, or a mix of both. In Table 1, highway vehicle profiles are listed for Phase II
Federal RFC, California RFC, and a 2010 baseline gasoline. Using reformulated gasoline VOC
reductions in a given State, the following steps are used to determine 2010 HAP emission
reduction estimates:
1 . Isolate the counties for which the Federal RFC program is selected as an ozone
control measure to meet a NAAQS alternative. Calculate baseline HAP emissions
using VOC emissions from the 2010 CAA Baseline inventory for each of these
counties (highway and nonroad gasoline vehicles). HAP emissions are calculated
by applying the weight fractions for 2010 baseline gasoline shown in Table I .
States and counties where RFC (either nonroad engine or highway vehicle) was
selected as a control measure under the 0.08 ppm. 5th maximum. 8-hour average
alternative, and the 0.08 ppm, 3rd maximum, 8-hour average alternative are listed
in Table 2. (No incremental VOC reductions occur as a result of reformulated
gasoline measures under the 0.09 ppm 3rd Max, 8-Hour'average alternative.)
2. Calculate "controlled VOC" emissions in each county by subtracting the
incremental VOC emission reduction for RFC from the baseline VOC emissions.
3. Apply the Federal RFG profile from Table 1 to. the county-level "controlled VOC"
emissions to calculate HAP emissions under each alternative.
4. Subtract controlled HAP emissions (step 3) from baseline HAP emissions (step 1)
to calculate benefits from Federal RFC.
The method used to calculate HAP emission reductions from RFG in Maryland - Washington
County under the 0.08 ppm, 5th maximum, 8-hour average alternative is shown below:
INCBEN = (VOC^ * BENZ*,*) - (VOCM * BENZFKFC)
where:
INCBEN incremental benzene reduction in 20 1 0;
VOC^e baseline VOC emissions from highway vehicles in 20 1 0 (2.63 tpd);
BENZ^e = weight fraction for benzene/2010 gasoline (Table 1); and
VOCttJ= VOC emissions from highway vehicles in 20 1 0 under 8-hour, 0.08 ppm,
5th maximum alternative;
weight fraction for benzene/Federal RFG (Table 1 ).
In the above formula, the variable VOC^, represents the baseline VOC emissions minus the
VOC reduction associated with reformulated gasoline. Using an incremental VOC reduction for
RFG of 0.3 1 tons per day (tpd), the incremental benzene reduction is:
INCBEN = (2.63 * 0.0396) - [(2.63 - 0.31) * 0.035*7 = 0.10 - 008 = 002 tpd
-------
Appendix 4
Benefits of Full Attainment
-------
Benefits of Full Attainment
The .analysis of the air toxics benefits of a revised ozone NAAQS presented in this
document focuses on the benefits of partial attainment scenarios, rather than full
attainment, because only the emissions reductions under partial attainment scenarios are
linked to identified control measures. These control measures provide the basis for
estimating the proportions of VOC emissions reductions which are accounted for by the
three target HAPs.
This appendix present an estimate of the air toxics benefits associated with full attainment
of the current standard, as well as the 5* maximum and 4* maximum options for a 0.08
parts per million (ppm) 8-hour average standard. Full attainment benefits are estimated
by extrapolation from partial attainment benefits using the national VOC emissions
reductions estimated for both scenarios:
IFA = (VOCFA/VOCPA) * IPA
%
4
where
IFA = total reduced incidence of cancer cases for all target HAPs in all states
resulting from full attainment of revised ozone NAAQS in year 2010
VOCpA= total national VOC emissions reductions (tons per day) necessary for
full attainment of revised ozone NAAQS in year 2010
VOCPA= total national VOC emissions reductions (tons per day) under partial
attainment of revised ozone NAAQS in year 2010
IPA = total reduced incidence of cancer cases for all target HAPs in all states
under partial attainment of revised ozone NAAQS in year 2010.
Table A-l shows the values for each of these variables, along with the estimated
monetized benefits using the mean estimated value for reductions in mortality of $4.8
million per statistical life.
-------
Table A-l. Estimated Cancer Risk Reduction Benefits in 2010 due to Full Attainment of
Current and Revised Ozone NAAQS
VOCFA
(tons per day, full attainment)
VOCPA
(tons per day, partial attainment)
IFA
(estimated 2010 cancer cases
avoided, partial attainment)
IFA
(estimated 2010 cancer cases
avoided, full attainment)'
Benefits
(monetized 2010 full attainment
benefits, at $4.8 MM per case
avoided)
Current
Standard
1616
467
0.6 cases
2.1 cases
$10.0 MM
0.08 ppm
5* Max
8-hour Avg.
896
475
0.7 cases
«*
1.3 cases
$6.2 MM
0.08 ppm
4* Max
8-hour Avg.
1611
736
1.3 cases
2.8 cases
$13.7 MM
'Estimated as:
IFA
* I
PA
-------
B. SPECIATION OF HAP EMISSIONS
The incremental State-level estimates of VOC emission reductions expected to occur to
meet the-current standard and the three ozone standard alternatives in 2010 are used along with
the profiles in Table 1 to derive estimates of benzene, 1,3-butadiene, formaldehyde, and other
HAP emission reductions. In addition to the current ozone standard, the three ozone N AAQS
alternatives for which VOC emission reductions are provided are:
Ozone Level Form Averaging Time
• 0.08 ppm 5th Maximum 8-hour average (Case Ic)
• 0.08 ppm 4th Maximum 8-hour average (Case li)
• 0.08 ppm 3rd Maximum 8-hour average (Case Id)
The control measures selected to meet the current ozone standard and each of the three ozone
alternatives are applied in the modeling to achieve VOC emission reductions beyond the control
levels applied in the 2010 Baseline Inventory. Control measures in the 2010 baseline that affect
VOC emissions include 2- and 4-year MACT standards, in addition to RACT controls and new
CTGs. National control measures incorporated into the 2010 baseline inventory include controls
for: consumer solvents, architectural coatings, TSDFs, municipal landfills, and onboard
refueling vapor recovery systems. CAA mobile and nonroad control measures reflected in the
baseline include: Federal nonroad engine standards, Phase IIRVP limits, I/M programs, and
Federal reformulated gasoline. The 49-State LEV program is also applied.
The annual VOC emission reductions by alternative, State, and control measure are
divided by 365 days to convert the values to an average daily basis. The average daily VOC
emission reductions by State and control measure are then speciated by multiplying the
incremental VOC emission reduction by the corresponding weight fraction in Table 1.
For the purposes of comparison of full attainment and partial attainment, VOC emission
reductions under the current standard, the 0.08,5th maximum, and the 0.08 ppm, 4th maximum
are shown in Table 7.
C. RESULTS
Observations about the emission results in Tables 3 through 6 are as follows:
• Alaska and Hawaii have no estimated VOC emission reductions because they are
excluded from the analysis.
• The VOC emission targets by nonattainment area do not always increase in
stringency between standards. * The target levels for the 8-hour 0.08 ppm 5th
maximum alternative may be more or less stringent than the target levels for the
current standard, or the target levels for the 8-hour 0.08ppm 3rd maximum
alternative. In Table 4, for example, negative incremental VOC reductions appear
for several States, including Indiana and Texas. Each of these States contains
counties mat comprise nonattainment areas for which VOC reductions are
expected to be needed to meet the current standard, but not to meet the 0.08 ppm
5th max, 8-hour average alternative, resulting in the negative incremental VOC
emission reductions shown in Table 4.
-------
• For Texas, negative VOC emission reductions are associated with the 0.08 ppm
5th maximum, 8-hour average alternative (Table 4). This is related to the
Houston area VOC emission reduction target, which is higher (less stringent)
under these two alternatives than for the current standard. In Table 6, incremental
VOC reductions are 115 tpd as a result of more stringent VOC targets for the
Houston-Galveston and Dallas-Fort Worth nonattainment areas under the 0.08
ppm 3rd maximum, 8-hour average standard.
• For certain States, such as Georgia and Michigan, the incremental VOC reduction
is constant across the three ozone alternatives.
• The Federal RFC control measure is selected in Arizona, Louisiana, Maryland,
Virginia, and West Virginia under the 0.08ppm 5th Max, 8-hour Average
alternative. In addition to these States, Federal RFC is also selected in Arkansas,
Mississippi, and Tennessee under the 0.08ppm 4th Max, 8-hour Average and 0.08
ppm 3rd Max, 8-hour Average alternative. The incremental HAP reductions
shown in Tables 4 through 6 for these States incorporate the HAP weight fractions
associated with both baseline gasoline and Federal RFC gasoline in 2010.
• Emission reductions in western, largely rural States like Wyoming and Montana
are attributable to the Tier 2 vehicle standards for light-duty vehicles and light-
duty trucks in the national ozone strategy.
• In California, the VOC target reductions for each nonattainment area increase in
stringency across the three alternatives. VOC measures to further control
emissions from degreasing operations, architectural coatings, and consumer
solvent use account for the majority of the incremental VOC reductions shown in
Tables 4 through 6.
• Benzene emission reductions range from 3.55 tpd under the 0.08ppm 5th Max, 8-
•> hour average alternative to 18.2 tpd under the 0.08 ppm 3rd Max, 8-hour average
alternative. Other HAP emission reductions range from 3 tpd to 153 tpd under
these same two alternatives.
• Under the 0.08 ppm 3rd Max, 8-hour average alternative, the highest other HAP
emission reductions are associated with the following control measures: high
enhanced I/M, Industrial Adhesives-content limits, and consumer solvent-content
limits.
References
EPA, 1992. Hazardous Air Pollutant Emissions from Process Units in the Synthetic Organic
Chemical Manufacturing Industry - Background Information for Proposed Standards, Volume
1C, November 1992.
EPA, 1993. Lifetime Emissions for Clean Fuel Fleet Vehicles, (EPA-AA-SRPB-93-01), Office
of Mobile Sources, Regulation Development and Support Division, October 1993.
EPA, 1994. Surface Coating Operations at Shipbuilding and Ship Repair Facilities -
Background Information for Proposed Standards, EPA 450-D-94-01 la, June 1994.
EPA, 1995. National Emission Standards for Hazardous Air Pollutants: Printing and
-------
Publishing Industry Background Information for Proposed Standards. EPA-453/R-95-002a,
February 1995.
Pechan/Mathtech, 1994. Regulatory Impact Analysis for the Petroleum Refinery NESHAP.
Revised Draft, prepared for U.S. EPA, OAQPS, prepared by E.H. Pechan & Associates and
Mathtech, Inc., March 15,1994.
STAPPA/ALAPCO. 1993. Meeting the 15-Percent Rate of Progress Requirement Under the
Clean Air Act: A Menu of Options, September, 1993.
Wilson, 1997. J. Wilson, E.H. Pechan & Associates, letter to D. Axelrad, U.S. EPA, "Task 4a:
HAP Speciation Methods", April 9,1997.
-------
Table 1
HAP Spcciations by Source Category
SOURCE
CONTROL
MEASURE
Wt. Fraction of HAP/Precursor in 2010 VOC
Benzene 1,3-Butadiene Formaldehyde Form.
Web Offset Lithography
Nonroad Gasoline Engines
Cutback Asphalt
Service Stations • Underground Tanks
Service Stations • Stage I
Oil and Natural Gas Production Field
Fugitives
Beverage Can Coating
Plastic Parts Coating
Wood Furniture Coating
Automobile Refinlshlng
Metal Coil/Can Coating
Metal Furniture Surface Coating
Machinery/Etec. Coating
Motor Vehicle Surface Coating
Surface Coating - Electronics
Aircraft Surface Coating
Marine Surface Coating
Aerosols
Adhesives • Industrial
Pesticide Application
Highway Vehicles: 2010 Gasoline
Highway Vehicles: Federal RFG
Highway Vehicles: CA RFG II
Architectural Coalings
Open Top/Convey. Degreaslng
New CTG (carbon
adsorber)
Reformulated gasoline
Switch to emulsified
asphalts
Vapor balance & P-V
valves
Vapor balance & P-V
valves
RACT
(equipment/maintenance)
Incineration
Incineration
Incineration
CARB BARCT limits
Incineration
VOC limits
VOC limits
Incineration
Add-on control levels/low
VOC
Incineration
Add-on control levels
CARB Tier 2 Standards •
Reformulation
Content limits
Reformulation - FIP rule
N/A
N/A
N/A
VOC content limits
Low VOC solvents
0
00401
0
0.0266
00266
0.0021
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.0396
00354
0.0320
0
0
0
00077
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0%
0
0
0
00079
00096
00061
0
0
0
00243
00202
0
0
0
0
0
0
0
0
0
0
0
0
0
t)
0
0
0
00067
0.0102
00112
0
0
Precursors Total
0
0.1302
00372
0.0288
0.0288
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.1340
0.1344
0.1399
0
0
Other HAP
0.626
0.2761
01820
0.1662
0.1662
0.3354
0.0523
0.0523
00523
0.0523
0.0523
0.0523
0.0523
0.0523
0.0523
0.0523
028
03899
0.3916
0.2374
0.1674
0.1677
0.1676
t
0.0523
0.0523
-------
Rubber and Plastics Mfg. Low VOC coatings 000 0 0.0523
Consumer Solvent CARB Limits 000 0 O.OS23
Highway Vehicle-LD Gasoline High Enhanced I/M 00396 00079 00067 0.134 0.1674
Highway Vehicle-Gasoline Transportation Control 0.0396 00079 0.0067 0.134 0.1674
Package
-------
Table 2
Counties for which Federal RFC* is Selected by Ozone
NAAQS Alternative
Baseline versus Ozone NAAQS Alternatives
Ozone NAAQS Altemafive:
State
Arizona
Arkansas
Louisiana
Maryland
Mississippi
Tennessee
Virginia
West Virginia
0.08 ppm 5th Max.
County 8-hour Average
PmalCo
Cnttendon Co
Ascension Parish
E Baton Rouge Parish
Ibervilte Parish
Livingston Pansh
Pointe Coupee Pansh
W. Baton Rouge Pansh
Washington Co
DeSotoCo
FayetteCo
Shelby Co
Tipton Co
Clarke Co
CulpeperCo
FauquierCo
King George Co
Madison Co
Spotsyivania Co
Warren Co
Berkeley Co
Jefferson Co
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
0.08 ppm 4th Max.
8-hour Average
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
0.06 ppm 3rd Max,
8-hour Average
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
NOTE 'Affects highway vehicle and nonroad engine emissions
-------
Table 3
State HAP Emission Reductions Incremental to 2010 CAA Baseline:
2010 Annual Average Emissions by HAP Category (tpd):
Current Ozone Standard
Annual HAP Emission Reductions by Pollutant (tpd):1
State Name
Cafifomia
Connecticut
Delaware
Indiana
Kentucky
Louisiana
Maryland
New Jersey
New York
Ohio
Pennsylvania
Texas
.
Incremental
FIPS State VOC Reduction
Code (tpd) Benzene 1.3 Butadiene
06
09
10
18
21
22
24
34
36
39
42
48
U.S. Total
111.23
21.37
1959
7.37
070
1553
055
77.10
94.24
1.57
4Z91
75.05
466.90
0.17
0.02
0.58
0.00
0.02
0.60
000
013
0.51
004
009
1.22
3.38
001
—
0.11
0.12
000
000
008
001
023
0.56
Formaldehyde
Formaldehyde Precursors
0.01
—
0.10
000 .
000
0.10
0.00
0.00
0.07 .
0.00
0.01
0.19
0.49
0.31
0.02
194
001
0.02
2.04
000
016
1.47
005
0.18
3.97
10.17
Other
HAP1
7.80
4.62
356
004
0.12
2.55
0.11
15.79
19.51
026
786
12.82
75.03
NOTES: 'Emission reductions are incremental to the 2010 CAA Baseline.
'Other HAPs include any Title Ill-toted volatile organic HAP, Common volatile organic HAPs include- toluene.
xytenes. and grycotethers. acetaldehyde, ethyfcenene. n-nexane, and 2,2,4-trimethylpentane.
-------
Table 4
State HAP Emission Reductions Incremental to Current Ozone Standard:
2010 Annual Average Emissions by HAP Category (tpd):
0.08 ppm Sth Max, 8-hour Average
Annual HAP Emission Reductions by Poflutant (tpd):*
State Name
Alabama
Arizona
Arkansas
California
Colorado
Connecticut
Delaware
District of Columbia
Florida
Georgia
Idaho
Illinois
Indiana
Iowa
Kansas
Kentucky
Louisiana
Maine
Maryland
Massachusetts
Michigan
Minnesota
Mississippi
Missouri
Montana
Nebraska
Nevada
New Hampshire
New Jersey
New Mexico
New York
North Carolina
North Dakota
Ohio
Oklahoma
Oregon
Pennsylvania
Rhode Island
Incremental
PIPS State VOC Reduction
Code (tpd) Benzene 1,3 Butadiene
01
04
05
06
08
09
10
11
12
13
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
44
0.53
55.97
0.27
168.28
1.15
27.92
2.23
5.00
1.59
1407
0.15
4.20
(4.72)
036
0.32
8.25
9.01
0.18
54.22
2.97
1.72
0.69
0.31
1.62
0.13
0.22
0.18
0.50
40.40
0.24
54.07
0.88
0.11
3.33
0.44
0.41
18.97
0.51
0.02
0.78
0.01
0.13
0.05
0.10
0.00
0.02
006
056
0.01
0.17
0.10
0.01
0.01
0.34
0.25
0.01
0.21
0.12
0.07
0.03
0.01
0.06
0.01
0.01
0.01
0.02
0.18
0.01
0.17
0.04
0.00
0.16
0.02
0.02
0.09
0.02
0.00
0.14
000
0.01
0.01
0.02
0.00
0.00
0.01
0.11
000
003
0.02
0.00
0.00
0.07
0.02
000
0.02
0.02
0.01
0.01
0.00
0.01
0.00
0.00
0.00
0.00
0.04
0.00
0.03
0.01
0.00
0.04
0.00
000
0.02
000
Formaldehyde
Formaldehyde Precursors Other HAP
0.00
0.12
0.00
0.10 .
001
0.01
0.00
0.00
0.01 v
009 *
0.00
0.03
0.01
000
0.00
0.06
0.02
0.00
002
0.02
0.01
0.00
0.00
001
0.00
0.00
0.00
0.00
0.03
0.00
0.03
0.01
0.00
0.03
000
0.00
0.01
000
0.07
2.43
0.04
039
0.15
0.28
0.01
0.05
021
189
0.02
0.56
0.35
005
0.04
1.18
0.32
0.02
050
0.40
0.23
0.09
0.04
0.22
0.02
0.03
0.02
0.07
0.61
0.03
0.58
0.12
0.01
0.61
0.06
005
0.30
007
0.09
8.96
0.04
1155
0.19
4.01
0.13
0.72
0.27
2.36
003
0.70
0.41
006
0.05
1.38
139
0.03
7.76
050
029
0.12
0.05
0.27
0.02
004
0.03
0.08
2.64
0.04
3.33
0.15
0.02
0.56
0.07
007
125
009
-------
Table 4 (continued)
Annual HAP Emission Reductions by Pollutant (tpd):*
State Name
South Carolina
South Dakota
Tennessee
Texas
Utah
Vermont
Virginia
Washington
West Virginia
Wisconsin
Wyoming
Incremental
FIPS Slate VOC Reduction
Code (tpd)
45
46
47
48
49
50
51
53
54
55
56
U.S. Total
0.43
0.11
1.79
(39.61)
073
0.10
2793
068
512
1.60
0.08
475.14
Benzene
0.02
0.00
0.03
(1.06)
0.01
0.00
0.46
003
0.12
0.06
0.00
3.55
1.3 Butadiene
0.00
0.00
0.01
(0.20)
0.00
0.00
0.08
0.01
0.02
0.01
0.00
0.62
Formaldehyde
Formaldehyde Precursors Other HAP
0.00
0.00
0.02
(0-17)
0.00
0.00
0.08
0.00
0.02
0.01
0.00
0.65
0.06
0.02
0.12
(3.41)
0.03
0.01
1.34
009
035
0.21
0.01
10.97
0.07
0.02
0.43
(8.75)
0.04
0.02
4.11
011
075
0.27
0.01
46.82
NOTE 'Emission reductions are incremental to the current 012 ppm hourly ozone standard.
-------
Tables
State HAP Emission Reductions Incremental to Current Ozone Standard:
2010 Annual Average Emissions by HAP Category (tpd):
0.08 ppm 4th Max, 8-hour Average
Annual HAP Emission Reductions by Pollutant (tpd):*
State Name
Alabama
Arizona
Arkansas
California
Colorado
Connecticut
Delaware
District of Columbia
Florida
Georgia
Idaho
Illinois
Indiana
Iowa
Kansas
Kentucky
Louisiana
Maine
Maryland
Massachusetts
Michigan
Minnesota
Mississippi
Missouri
Montana
Nebraska
Nevada
New Hampshire
New Jersey
New Mexico
New York
North Carolina
North Dakota
Ohio
Oklahoma
Oregon
Pennsylvania
Rhode Island
Incremental
FIPS State VOC Reduction
Code (tpd) Benzene. 1,3 Butadiene
01
04
05
06
08
09
10
11
12
13
16
17
18
19
20
21
22
23
24
25
26
27
26
29
30
31
32
33
34
35
36
37
38
39
40
41
42
44
36.36
52.55
3.85
17068
1.15
27.92
2.23
5.00
159
1407
0.15
4.20
3.30
0.36
0.32
14.39
1193
0.18
54.55
297
1.72
0.69
5.43
1.62
0.13
0.22
0.18
0.50
40.40
0.24
54.07
33.51
0.11
32.20
0.44
0.41
18.97
0.51
1.44
0.49
0.12
0.13
0.05
010
0.00
0.02
0.06
056
001
0.17
0.11
0.01
0.01
0.35
0.37
0.01
022
0.12
0.07
0.03
0.15
0.06
0.01
001
0.01
0.02
0.18
0.01
0.17
1.33
0.00
0.19
0.02
0.02
0.09
0.02
0.29
007
0.02
0.01
0.01
0.02
0.00
0.00
0.01
011
0.00
0.03
0.02
0.00
0.00
007
0.04
0.00
002
002
0.01
001
003
001
000
000
000
0.00
0.04
0.00
0.03
0.26
0.00
0.04
000
000
002
000
Formaldehyde
Formaldehyde Precursors Other HAP
0.24
0.07
002
0.11
0.01
0.01
0.00
0.00
0.01
009 '"
000
0.03
002
0.00
0.00
0.06
0.03
0.00
0.02
0.02
0.01
0.00
0.03
0.01
0.00
0.00
0.00
0.00
0.03
0.00
003
0.22
0.00
0.03
0.00
000
0.01
0.00
4.87
1.33
0.33
041
0.15
0.28
0.01
005
0.21
189
0.02
0.56
0.36
0.05
0.04
1.19
0.63
0.02
050
040
0.23
0.09
044
0.22
002
003
- 002
0.07
0.61
0.03
0.58
4.49
0.01
0.64
0.06
0.05 •
0.30
007
609
8.08
0.64
11.79
0.19
4.01
0.13
0.72
0.27
236
003
070
0.55
006
0.05
2.21
1.85
0.03
7.82
0.50
0.29
0.12
0.92
0.27
0.02
0.04
0.03
0.08
2.64
0.04
3.33
561
0.02
5.37
0.07
007
1.25
009
-------
Table 5 (continued)
Annual HAP Emission Reductions by Pollutant (tpd):*
State Name
South Carolina
South Dakota
Tennessee
Texas
Utah
Vermont
Virginia
Washington
West Virginia
Wisconsin
Wyoming
Incremental
PIPS State VOC Reduction
Code (tpd) Benzene 1.3 Butadiene
45
46
47
48
49
50
51
53
54
55
56
U.S. Total
509
0.11
75.04
20.83
0.23
0.10
28.24
068
5.12
1.60
0.08
736.22
0.20
0.00
2.47
0.19
0.01
0.00
0.47
0.03
0.12
0.06
0.00
10.27
0.04
0.00
0.44
0.04
0.00
0.00
0.08
001
002
0.01
0.00
1.86
Formaldehyde
Formaldehyde Precursors Other HAP
0.03
0.00
0.38
0.03
0.00
0.00
0.08
0.00
0.02
0.01
0.00
1.72
0.68
0.02
7.52
0.64
0.03
0.01
1.36
009
035
0.21
0.01
32.22
0.85
0.02
12.48
1.80
0.04
0.02
4.17
0.11
0.75
0.27
0.01
88.82
NOTE- 'Emission reductions are incremental to the current 0.12 ppm hourly ozone standard. • "
-------
Table 6
State HAP Emission Reductions Incremental to Current Ozone Standard:
2010 Annual Average Emissions by HAP Category (tpd):
0.08 ppra 3rd Max, 8-hour Average
Annual HAP Emission Reductions by
Incremental
FIPS State VOC Reduction
State Name Code (tpd) Benzene. 1,3 Butadiene
Alabama
Arizona
Arkansas
California
Colorado
Connecticut
Delaware
District of Columbia
Florida
Georgia
Idaho
Illinois
Indiana
Iowa
Kansas
Kentucky
Louisiana
Maine
Maryland
Massachusetts
Michigan
Minnesota
Mississippi
Missouri
Montana
Nebraska
Nevada
New Hampshire
New Jersey
New Mexico
New York
North Carolina
North Dakota
Ohio
Oklahoma
Oregon
Pennsylvania
Rhode Island
01
04
05
06
08
09
10
11
12
13
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
44
36.36
6076
385
174.94
1.15
27.92
2.23
500
1.59
14.07
0.15
104.81
15.02
0.36
0.32
14.46
11.88
5.18
54.55
3.49
1.72
069
5.43
1.62
0.13
0.22
0.18
4.62
40.40
0.24
54.07
33.87
0.11
32.42
29.89
0.41
18.97
051
1.44
0.78
0.12
0.14
005
0.10
0.00
0.02
0.06
056
0.01
0.27
0.12
0.01
0.01
0.36
0.36
0.21
0.22
0.14
0.07
0.03
0.15
006
001
001
001
0.18
0.18
0.01
0.17
1.34
0.00
0.20
118
002
009
002
0.29
0.14
0.02
0.01
0.01
0.02
0.00
000
0.01
0.11
0.00
0.03
0.02
0.00
000
0.07
0.04
0.04
0.02
0.03
0.01
001
0.03
0.01
0.00
0.00
0.00
0.04
0.04
0.00
0.03
0.27
0.00
0.04
024
0.00
002
000
Pollutant (tpd):*
Formaldehyde
Formaldehyde Precursors Other HAP
0.24
0.12
002
011
001 '
001
0.00
000
0.01
009 '"
000
004
0.03
0.00
0.00
006
003
0.03
002
0.02
001
0.00
0.03
0.01
0.00
0.00
0.00
0.03
003
0.00
0.03
0.23
0.00
0.04
0.20
000
0.01
000
4.87
2.43
0.33
0.42
0.15
0.28
0.01
0.05
0.21
189
0.02
0.70
0.39
0.05
0.04
1.20
0.62
0.69
0.50
0.47
0.23
0.09
0.44
022
0.02
0.03
0.02
0.62
0.61
0.03
0.58
4.54
0.01
067
4.01
0.05-
0.30
007
6.09
945
0.64
12.05
0.19
401
0.13
0.72
0.27
236
0.03
18.38
2.60
0.06
005
222
184
087
782
0.58
029
0.12
092
0.27
002
004
0.03
0.77
264
0.04
3.33
5.67
002
540
500
007
125
009
-------
Table 6 (continued)
Annual HAP Emission Reductions by Pollutant (tpd):*
State Name
South Carolina
South Dakota
Tennessee
Texas
Utah
Vermont
Virginia
Washington
West Virginia
Wisconsin
Wyoming
Incremental
PIPS State VCX) Reduction
Code (tpd) Benzene 1,3 Butadiene
45
46
47
48
49
50
51
53
54
55
56
U.S. Total
5.09
0.11
107.67
114.66
0.23
0.10
115.11
068
5.17
3.60
0.08
1.116.06
0.20
0.00
3.77
1.36
0.01
0.00
3.91
003
0.12
0.07
0.00
18.16
0.04
0.00
0.70
0.24
0.00
0.00
0.76
001
002
001
000
3.39
Formaldehyde
Formaldehyde Precursors Other HAP
0.03
0.00
0.60
0.25
0.00
0.00
0.66
0.00
002
0.01
000
3.08
0.68
0.02
11.89
4.36
0.03
001
1300
009
035
0.22
0.01
58.55
085
0.02
17.94
17.53
004
002
1871
Oil
076
065
0.01
15295
NOTE 'Emission reductions are incremental to the current 012 ppm hourly ozone standard'
-------
Table 7
National VOC Emission Reductions by Ozone Alternative:
Full Attainment Scenario
VCX: Emission Reductions in 2010 (tpd) under
Ozone Alternative
Current Standard1
0.08 ppm, 5th max2
008 ppm, 4th max2
Full Attainment Partial Attainment
1,615.9 466.9
895.9 475 1
1,6105 736.2
NOTES. VOC emission reductions are shown incremental to the 2010 baseline inventory
VOC emission reductions for the eight-hour standards are in addition to those for the current standard scenario.
-------
Appendix 3
Cancer Benefits of the Current Ozone Standard
-------
Cancer Benefits of the Current Ozone Standard
The existing National Ambient Air Quality Standard (NAAQS) for ozone is set at 0.12
parts per million, one hour average, with one exceedance allowed. Nine areas are
currently designated as being in non-attainment with the existing standard. Cancer
benefits of partial attainment of the existing standard in these areas in the year 2010 are
shown below; detailed calculations are shown in the attached tables. Estimates provided
here were developed using* the methods presented in the main body of this report for
estimation of benefits of a revised NAAQS. All estimates presented in the main body of
this report for revised NAAQS options represent benefits incremental to the benefits
shown below for the existing ozone NAAQS.
Estimated Cancer Cases Avoided in 2010 due to Partial Attainment of the
Existing Ozone NAAQS
Pollutant Cases Avoided
Benzene 0.1
1,3-Butadiene 0.4
Formaldehyde (total) 0.1
Total 0.6
Valued at $4.8 million per cancer case avoided, the annual monetized cancer benefits
from partial attainment of the existing standard are an estimated $2.9 million.
-------
Benzene-Current Std
STATE
CA
CT
DE
IN
KY
LA
MD
NJ
NY
OH
PA
TX
Benzene
Reductions
<«pd)
0.17
0.02
0.58
-
0.02
060
-
013
051
0.04
009
122
Baseline
Benzene
Emissions
(tpd)
167
16
5
37
24
41
25
42
74
58
67
170
Benzene
Reductions
(%)
01%
0.1%
11.8%
0%
0.1%
1.4%
0%
0.3%
07%
01%
01%
0.7%
Baseline
Benzene
Conen
(ug/m3)
256
1.58
199
0.98
0.74
1.54
1.81
276
275
121
190
191
Benzene
Concn
Reduction
(ug/m3)
0.003
0002
02
0
0.001
002
0
001
002
0001
0003
001
Benzene
Potency
Factor
(ug/m3>-1
83E-06
8.3E-06
8.3E-06
8.3E-06
8.3E-06
83E-06
8.3E-06
83E-06
83E-06
83E-06
83E-06
83E-06
2010
POP
(x 10*3)
37.644
3.400
817
6,318
4,170
4,683
5.657
8.638
18.530
11.505
12.352
22.857
Avoided
2010
Benzene
Cases
0.01
0.00
002
000
0.00
001
000
001
004
000
000
004
TOTAL CASES AVOIDED 0.1
-------
1,3-butadJene-Current Std
STATE
CA
CT
DE
IN
•or
LA
MO
NJ
NY
OH
PA
TX
1,3-butadiene
Reductions
(tpd)
001
.
011
.
-
0.12
-
-
008
-
001
0.23
Baseline
1,3-outadiene
Emissions
(tpd)
23
2
1
5
4
7
3
4
10
8
8
22
1.3-butadiene
Reductions
<%)
0.04%
0%
19%
0%
0%
2%
0%
0%
1%
0%
0.1%
1%
Baseline
4 * 'mlm.ifl.mji
1 v«"WMUlVHV
Concn
(ugftn3)
021
0.13
0.12
0.09
0.08
0.14
0.14
0.18
027
0.12
015
012
1.3-butadiene
Concn
Reduction
(uB/m3)
0.0001
0
0.02
0
0
0003
0
0
0.002
0
00002
0.001
1.34>utadlene
Potency
Factor
(ug/m3)-1
28E-04
28E-O4
28E-04
2.8E-04
2.86-04
28E-04
28E-04
28E-04
28E-04
28E-04
28E-04
2.8E-04
2010
POP
(x 10*3)
37.644
3.400
817
6.318
4.170
4.683
5.657
8.638
18.530
11.505
12.352
22.857
Avoided
2010
1.3-fcutadlene
Cases
0.01
000
0.08
000
0.00
005
000
0.00
017
000
001
012
TOTAL CASES AVOIDED 0.4
-------
Primary Formald-Current Std
STATE
CA
CT
DE
IN
KY
LA
MD
NJ
NY
OH
PA
TX
Form (prim)
. Reductions
(Hid)
001
.
0.10
.
.
010
•
.
007
.
001
019
Baseline
Form (prim)
Emissions
(tpd)
113
7
2
20
15
40
12
18
39
33
34
142
Form (prim)
Reductions
(%)
001S
0%
4%
0%
0%
0.2%
0%
0%
02%
0%
003%
0.1%
Baseline
Form (prim)
Concn
-------
Secondary Formald-Current Std
STATE
CA
CT
DE
IN
KY
LA
MD
NJ
NY
OH
PA
TX
Precursor
Reductions
(tpd)
031
0.02
1.94
0.01
0.02
2.04
-
0.16
1.47
0.05
018
3.97
Baseline
Precursor
Emissions
(tpd)
327
28
9
73
51
96
50
69
137
119
121
412
Precursor
Emission
Reductions
(%)
0.1 %
0.1%
23%
0.01%
'0.04%
2%
0%
0.2%
1%
0.04%
01%
1%
Baseline
Form (Sec)
Concn
(ugftnS)
0.27
0.20
0.17
0.10
0.08
014
0.23
0.36
0.37
014
020
028
Form (Sec)
Concn
Reduction
(ug/m3)
0.0003
0.0001
0.04
0.00001
0.00003
0.003
0
0001
0004
00001
00003
0.003
Formald
Potency
Factor;
(ug/m3M
1.3E-O5
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
1.3E-05
13E-05
13E-05
1.3E-05
2010
POP
(x 10A3)
37.644
3.400
817
6.318
4,170
4.683
5.657
8.638
18.530
11.505
12.352
22.857
Avoided
2010
Form (Sec)
Cases
000
000
001
0.00
0.00
000
000
000
0.01
000
000
001
TOTAL CASES AVOIDED 0.04
r«v 6/19*97
------- |