UNIXRSTANDINC
THE
EFFECTIVENESS
OF
PRECURSOR
REDUCTIOHSIN
LOWERING
8­
HOUR
OZONE
CCNCENTRATlONS
IN
THE
EASTERY
UNITED
STATES
Steve
Reynolds
Charles
Blanchard
EnYair
I2Palm
Avenue
SanRafael,
CA
94901
Telephondfax:
(
415)
457­
6955
E­
mail:
steve@
srejnulds.
com
Jm9.7mI
I
Findings
for
Central
California
Various
VOC
and
NOx
emissions
reductions
effectlive
in
reducing
peak
1­
hour
ozone
levels
ViOC
reductions
have
modest
impact
on
peak
8hour
ozone
concentrations
Anthrapogenic
NOx
emissions
must
be
reduced
by
90
percent
to
reach
the
8­
hour
ozone
standard
Jurr9.2033
3
Purpose
of
the
Study
Develop
a
better.
understandingof
the
physical
and
chemical
mechanisms
underlying
observed
ozone
trends
Assess
the
technical
feasibility
of
attaining
the
8hour
ozone
standard
Jun:
9.
ZM3
Findings
for
Central
California
Anthropogenic
VOC
reductions
ineffective
because
biogenic
VOC,
less
reactive
VOC,
and
CO
continue
to
produce
ozone
Ozone
produced
per
NO
increases
as
NOx
emissions
decrease
NOx
emissions
reductions
may
increase
ozone
levels
in
some
areas
These
results
call
into
question
the
technical
of
attaining
the
8­
hour
ozone
standard
Jun
9.2M3
4
­­­
Objective
of
Current
Study
Extenld
basic
methodology
employed
for
centralI
California
to
eastern
U.
S.
and
assess
consistencv
I
of
findings
Photochemical
Modeling
Choice
of
study
area
based
on
recent
regulatory
modelling
for
the
eastern
U.
S.
and
availabilityof
code
with
process
analysis
facilities
EPA
encouraged
usage
of
and
assisted
in
providing
Clear
Skies
Initiative
modeling
files
CAMx
(
version
3.10)
with
process
analysis
I
Junr9.2033
7
Overview
of
Study
Approach
Combined
application
of
­
analyses
of
ambient
measured
ozone
and
precursor
data
­
photochemical
modeling
with
process
analysis
Employ
modelingto
provide
insights
into
physical
and
chemical
processes
associated
with
ozone
formation
Photochemical
Modeling
Some
field
measurementsduring
summer
of
1995
to
support
model
application
and
evaluation
e
5­
15
July
1995
ozone
episode
RAMS
used
to
develop
meteorological
inputs
Emissions
estimates
from
Clear
Skies
Initiative
Jurr
9.
m3
8
2
CAMx
Emissions
Sensitivity
Results
Subregions
Array
of
CAMx
VOC
and
NOx
sensitivity
runs
using
1996
base
case
emissions
AnthropogenicVOC
and
NOx
emissions
reduced
from
1996
base
case
levels
in
20
percent
increments
Results
used
to
produce
peak
1­
and
8­
hour
EKh4.
A­
typeisopleth
diagrams
Jlvr9.2CM
Subregions
9
CAMx
Runs
with
Process
Analysis
1996
and
2010
base
case
emissions
­
VOC
reductions
ranged
from
24
to
44
percent
­
NOx
reductions
ranged
from
12
to
40
percent
­
CO
reductions
ranged
from
9
to
29
percent
Additional
runs
with
1996
anthropogenicVOC
and
NOx
reduced
by
60
and
80
percent
3
.
Model
Predicted
Peak
Ozone
Levels
Peak
1­
and
8­
hour
ozone
isopleth
diagrams
based
on
19196
base
case
emissions
Various
combinations
of
VOC
and
NOx
emissions
reductions
initially
effective
in
reducing
peak
1hour
idzone
Jmr9,2m3
11
A
201OBorc
A
2010conrm~

V
2ozOBars
v
207J3cOnU"
l
%
bowOzonsCompsrablsu)
1999­
2001DIsign
Vduc
1­
9.
m1
I5
Model
Predicted
Peak
Ozone
Levels
VOC
emissions
reductions
have
modest
impact
on
peak
8hour
ozone
concentmions
Under
conditions
that
ylelded
peak
%
hour
ozone
levels
comparable
to
1999­
2001
design
values.
NOx
emissions
must
be
reduced
to
33­
75
percent
of
19%
base
case
values
to
reach
level
of
the
8­
hour
standard
Under
conditions
that
yielded
the
highest
8­
hour
ozone
concentrations,
anthropogenicNOx
emissions
must
be
reduced
to
75­
85percent
of
1996bise
case
values
to
reach
level
of
the
8­
hour
standard
Inn;
9.
I4
'
4'
ay
with
Maximum
8­
hourOzone
for
Epbodc
e
Juri;
9.
21133
17
18
urOzonc
Comparabls
lo
1999­
2001
Dorip
Value
Jvrr9.21133
I9
5
22
6
A
2010Bass
A
ZOlOCmvol
V
2028
Bmc
ZOZO
controlv
'
Day
wilh
MaximumShow
Ozonc
far
Epirodc
J­
9.2033
5
A
2010
Bmc
A
2010
Bmc
A
2010Convol
A
2010Cm~
1
V
2020
Bmc
V
2020
Bmc
2OZOConMl
v
2020Ca"
rn1
v
Day
wilh
8­
haur
Ozone
Compnrablc
LO
1999­
2001
Dsien
Vnlw
in
Mcvo
AM
1­
9.2003
27
J­
9.
ZW3
­
8
7
A
2010
Be
A
2010
Bate
A
2OIOconrol
A
2010
&"
lml
V
2020
Bar=
V
2020
Bass
v
2020
&
nu01
v
2020
conrol
'
DuywilhMoximurnS­
bourOzonc
for
Epiradc
Jlur
9,2003
29
A
2010
Bare
A
2010
Bnrt
A
2010
Conmnl
A
2010
Conm1
V
2020
Bssc
V
zou)
Bnrc
v
2020
Conlrol
v
2020conlrol
I
8­
hourOroncCampornblcu,
1999­
2031
Dssign
Vduc
Jun;
9.2003
31
8
Day
wih
Maximum
%
hourOmnc
far
Episode
Jm9.2003
33
A
2010
Bmc
A
2010
Bore
A
ZOIOConml
A
2010CO"
vol
V
ZMO
Bmc
V
2020
Bmo
'
I2020
conmi
'
I202Oco"
VOl
9
A
2010
Base
A
2010
Bare
A
20LOconvol
A
2010
canuol
V
2020
Bars
V
2020
Base
i1
v
2020
OInuOl
v
202oconvol
1
i
31
10
I
a
2010
Bnrc
A
2010
Bnss
A
2010
convol
A
2010
CaOrrUl
V
2020
Bore
V
2020
Blvlc
v
2020
conVal
v
2020
CUnVOl
I
Jw9.2003
42
Pmccss
onalysis
mults
for
July
14
­
NO
cycler
Process
Analysis
Results
Useful
dements
of
process
analysis
results
­
NO
cycles
(­
ozone
produced
per
NO)
­
mass
of
NO
reacted
­
mass
of
VOC
reacted
­
ozone
created
Ozone
production
involves
­
radical
initiation,
propagation,
and
termination
eated
by
photolysis
of
NO2
Jun;
9.2003
43
Jm9.2003
44
11
I
Process
Analysis
Results
VOC
emission
reductions
have
little
influence
on
NO
cycles
NO
cycles
increase
as
NOx
emissions
reduced
Increaseid
ozone
production
efficiency
at
lower
NOx
emissions
levels
Jw9.
Ua3
Summary
of
Modeling
Results
Ozonelproductionin
some
urban
core
areas,
such
as
New
York
City
and
Chicago,
are
VOC
limited
­
Moderate
NOx
emission
reductions
may
cause
increases
in
peak
8­
hour
ozone
levels
­
VOC
or
substantial
NOx
emission
reductions
needed
to
reach
8­
hour
ozone
standard
Jvlr
9.21303
41
Summary
of
Modeling
Results
As
NOx
emissions
decrease,
ozone
production
efficiency
increases
Substantial
NOx
emissions
reductions
required
to
reach
8­
hour
standard
in
several
areas
VOC
reductions
have
little
effect
on
NO
cycles
or
the
mass
of
VOC
reacted
1­
P.
ZM3
46
Analyses
of
Ambient
Ozone
and
Precursor
Data
Characterize
trends,
explain
observed
patterns,
assess
how
patterns
relate
to
modeling
results
Trends
determined
from
annual
statisticsusing
t­
tests
of
log­
transformed
data
(
sites
with
at
least
10
years
data)
­
4thhighest
&
haw
maximum
­
annual
I­
hour
maximum
Average
diurnal
concentration
profiles
for
21
high­
ozone
days
each
year
(­
10%
of
7
months)

Junc9.2W3
48
12
I
Observed
Ozone
Trends
Annual~
4M1ighestmaximum
daily
8­
hour
ozone
concentdons
declined
at
most
sites
in
New
York,
Philadt:
lpMa,
Chicago,
Indianapolis.
Atlanta
and
Baltimore.
Washington
areas(
1980­
2002)

Annual
I­
hour
maxima
also
declined,
by
greater
amounts
­
Much
iiinaller
trends
from
1992­
2002
49
Exampletrends
inthe
New
Yo&
BM
fw
the
mud
@­
highest
maximum
daily
8­
hour
ozonecnoecntrntion,
1980
throngb
2W2
J­
9.2M3
so
Sunumry
of
trends
io
mud
mdmm
I­
hour
ozone
and
amud
4 
highcsl
nnxirmin
dnily
&
hour
ozone
canccntmlion,
1980
hugh
ZOM
I­
hour
may
8­
hour
max
I
Juo~
9.2023
51
Jwr
9.
m3
52
13
I
Example
diurnal
pUem
of
ozo~
lcnt
Now
Yorksitu
I
~~~

Example
diurnnlpaUcm
ofozone
at
Baltimore­
Waslungton
situ
Exnmplc
diumol
pmcm
ofozone
at
Atlanta
sit1
9.2033
56
14
Comiparison
of
Modeling
and
Ambient
Data
Analysis
Results
Limilations
of
modeling
results
­
uncertainties
in
biogenic
emissions
estimates
­
unclear
whether
ozone
response
for
modeling
period
is
representative
of
other
days
when
moderate
to
relatively
high
8­
hour
ozone
concentrations
occur
Jun9.2003
57
Projectionot
historical
vends
10
2008­
2010
8­
hour
ozone
design
value
in
the
New
York
alnd
Philadelphia
anas
prrdicts
that
mny
riiw
will
fail
io
attain
I
Jw9.2033
59
Comparison
of
Modeling
and
Ambient
Data
Analysis
Results
Consistency
of
modeling
and
ambient
data
analysis
results
­
both
indicate
more
modest
changes
in
annual
4thhighest
maximum
8­
hour
ozone
levels
thanin
annual
maximum
I­
hour
ozone
levels
­
both
indicate
difficultyof
attainment
of
8­
hour
ozone
standardby
2010
­
both
indicate
increases
in
model
NO
cycles
and
ambient
03INOx
patios
lw9.
ZM3
58
Projcclion
of
historic4trends
a­
to
2008­
2010
8­
haur
ozone
$
designvalue
in
hke
Michigan
anapredictsthat
E
most
sites
arc
likely
to
attain
A5a
Jurr9.2033
60
15
Projection
of
lustorid
muds
to
2008­
2010
8­
hourozone
design
value
in
I'
the
Baltimore..
Washinglon
prrdins
that
mny
sites
will
fail
to
annin
Jlue9.2033
61
Ambient
Ratios
of
Mean
OzonefNOx
Increased
During
the
Late
1990s
63
~­

Ambient
Ratios
of
MeanOzone/
NOx
Increased
Duringthe
Late
1990s
FREQUENCY
OF
VOC­
LIMITED,
TRANSITIONAL,
AND
Nor­
LIMITED
IIOURS
AT
ALL
MONlTORS
CCnnlC.
li(
d
1591­
98
lo
all
arms.
frequency
of
NOx
7­
1593­
W
limited
how
is
greater
for
houn
S&
C.
lifd
IWI
with
ozonul20
ppbv
(
right)
All.­
1­
N**
.
mi
nid­
Nh&
IPPIP)
than
for
hours
with
ozone
5mIhmI.
k
ffi&,..
M
19Plw
>
80
ODbV
flCft).
..
.

Jw9.2033
 4
16
Conclusions
Various
'
VOC
and
NOx
emissions
reductions
may
be
efkctive
in
reducing
peak
I­
hour
ozone
levels
VOC
reductions
have
modest
impact
on
peak
8hour
ozone
concentrations
1­
9.
zw3
Conclusions
AnthropogenicVOC
reductions
ineffective
because
biogenic
VOC,
less
reactive
VOC,
and
CO
continue
to
produce
ozone
Ozone
production
efficiency
increases
as
NOx
emissildns
decrease
67
Conclusions
Under
conditions
that
yielded
peak
8­
hour
ozone
levels
comparable
to
1999­
2001
design
values,
NOx
emissions
must
be
reduced
to
50­
75
percent
of
19%
base
we
values
to
reach
level
of
the
8­
hour
standard
Under
conditions
that
yielded
the
highest
8­
hour
ozone
concentrations,
anthropogenicNOx
emissions
must
be
reducedto
75­
85
percent
of
1996basecase
values
to
reach
level
of
the
8­
hour
standard
66
Conclusions
Increases
in
ozone
production
efficiency
associatedwith
NOx
emissions
reductions
also
noted
in
central
California
modeling
study
Ozone
production
in
urban
core
areas,
such
as
New
York
City
and
Chicago,
may
be
VOC
liited.
­
Moderate
NOx
emissionreductions
may
cause
increases
in
peak
%
hour
ozone
levels
­
VOC
or
substantid
NOx
emission
reductions
needed
to
reach
8­
hour
ozone
standard
IwmS.
ZW3
68
17
­

Conclusions
Thesetresultscall
into
question
the
technical
feasibilif
y
of
attaining
the
8­
hour
ozone
standard
JmrY.
2m3
69
I
18
I
