Draft MOVES2009 Highway Vehicle
Temperature, Humidity, Air Conditioning,
and Inspection and Maintenance
Adjustments
United States
Environmental Protection
Agency
-------
Draft MOVES2009 Highway Vehicle
Temperature, Humidity, Air Conditioning,
and Inspection and Maintenance
Adjustments
Assessment and Standards Division
Office of Transportation and Air Quality
U.S. Environmental Protection Agency
v>EPA
NOTICE
This technical report does not necessarily represent final EPA decisions or
positions. It is intended to present technical analysis of issues using data
that are currently available. The purpose in the release of such reports is to
facilitate the exchange of technical information and to inform the public of
technical developments.
United States EPA-420-P-09-003
Environmental Protection . ^ _
Agency August 2009
-------
Table of Contents
Table of Contents i
1. Introduction 1
2. Temperature Adjustments 1
2.1 Data Sources for Temperature Effects 1
2.2 Temperature Adjustment Methodology 2
2.3.1 Temperature Effects on Gasoline Start Emissions 3
2.3.2 Temperature Effects on Gasoline Running Emissions 7
2.4 Effects of Temperature on Diesel Fueled Vehicles 9
2.5 Cold Weather Effects 11
2.5.1 Cold Weather CO Requirement 12
2.5.2 Cold Weather HC Requirement 12
2.5.3 Cold Weather PM Effects 13
3. Humidity Adjustments 16
4. Air Conditioning Adjustments 17
4.1 Air Conditioning Effects Data 18
4.2 Method for Calculating Air Conditioning Effects 20
4.3 Air Conditioning Effects on Emissions 22
4.3.1 A/C Correction Factors for HC, CO and NOx Emissions 22
4.3.2 Full A/C Correction Factors for Energy Emissions 23
4.3.3 Uncertainty Analysis 23
4.4 Adjustments to Air Conditioning Effects 26
5. Inspection and Maintenance Programs 28
5.1 Inspection & Maintenance inMOBILE6 28
5.2 Inspection & Maintenance in MOVES 28
5.3 Development of MOVES IMF actors 29
5.4 Development of MOVES IM Compliance Inputs 32
6. References 34
Appendix A: Mean Start Emission by Temperature 36
Appendix B - Calculation of Specific Humidity 38
-------
1. Introduction
The emission rates in the MOVES model database represent a single (base) scenario of
conditions of temperature, humidity, air conditioning load and fuel properties. MOVES is
designed to adjust these base emission rates to reflect the conditions for the location and time
specified by the user. MOVES also includes a methodology for adjusting the base emission rates
to reflect the effects of local-run Inspection and Maintenance (I/M) programs. This report
describes how these adjustments for temperature, humidity, I/M and air conditioning were
derived. Adjustments for fuel properties are being addressed in a separate report. The crankcase
emission processes are chained to running exhaust, engine start and extended idling emissions
and are thus similarly affected by the temperature adjustments describe in this report.
2. Temperature Adjustments
In EPA's previous emissions model (MOBILE6), passenger car and light-duty truck tailpipe
emissions were adjusted relative to its base emission rates at 75 degrees Fahrenheit based on:
1. ambient temperature [1], and
2. for start emissions, an adjustment factor based on the length of the soak time. [2]
MOVES will take a similar approach, but we will substantially alter the nature of the
temperature adjustment factors.
2.1 Data Sources for Temperature Effects
For this analysis, we used almost entirely "Bagged" tests. Those data set consisted of
Federal Test Procedure (FTP) and LA-92 tests for start emissions. For the temperature effects on
running emissions we used the Bag-2 emissions of those FTPs as well as US06 tests (without
engine starts). Some second-by-second test data were used (but only) to validate the effects of
temperature on running emissions (HC, CO, and NOx). The data used in these analyses come
from the following four sources:
1. EPA's Mobile Source Observation Database (MSOD) as of April 27, 2005. Over the
past decades, EPA has performed emission tests (usually the FTP) on tens of thousands
of vehicles under various conditions. EPA has stored those test results in its Mobile
Source Observational Database (MSOD). (EPA has supplemented those tests with the
results from many non-EPA testing programs.)
For the MSOD data, we limited our analysis to only tests from the vehicles that were
tested at two or more temperatures. In this analyses, those paired (MSOD) tests covered
the temperature range from 15 to 110 degrees Fahrenheit. Many (most) of those bagged
tests (FTPs) were also used in our earlier MOBILE6 analyses.
Information on EPA's MSOD is available on EPA's website:
-------
http://www.epa.gov/otaq/models.htm
2. A testing program in Kansas City also yielded pairs of test (using LA92s tests rather than
FTPs) from the vehicles that were tested at two or more temperatures.
3. EPA's Office of Research and Development (ORD) contracted (through the Clean Air
Vehicle Technology Center, Inc.) the testing of five cars (model years 1987 through
2001). Those vehicles were tested using both the FTP and the EVI240 cycles at
temperatures of: 75, 40, 20, 0 and -20 °F. These five vehicles supplemented the vehicles
from the MSOD and Kansas City . [3]
4. Under a contract with EPA, the Southwest Research Institute (SwRI) tested four Tier-2
vehicles (2005 model year car and light-duty trucks) over the FTP at temperatures of: 75,
20, and 0 °F. These four vehicles also supplemented the vehicles from the MSOD and
Kansas City.
2.2 Temperature Adjustment Methodology
For our analyses, we stratified the paired-test data by the same parameters that MOVES uses
to define the Source Bins, namely: fuel type, regulatory class, and model year groups (listed on
the next slide).
For this analysis, we started with the model year groups used in MOVES for start emission
rates. By combining several model years into single groups, we consolidated those (MOVES)
model year groups into these six model year groups:
- 1960 to 1980
-- 1981 to 1982
-- 1983 to 1985
-- 1986 to 1989
-- 1990 to 2004
- 2005 and later
A preliminary analysis of the test data indicated that the Tier-0, Tier-1, and LEVs all
exhibited similar increases in emissions by the time the ambient temperature drops from 75° F to
20° F. A single additive adjustment factor (for each of HC, CO, NOx) can represent this.
Both the Federal FTP and California's Unified Cycle are 3-mode (or 3-bag) tests in which
the first and third modes are identical driving cycles, but the first mode begins with a cold-start
and the third mode begins with a hot-start start. We used the adjusted difference of Bag-1 minus
Bag-3 emissions to estimate the cold-start emissions (in grams) for each test.
Similarly, we used the emissions from the FTP Bag-2, EVI240, and US06 tests to estimate the
ratios (i.e., multiplicative changes) in the hot-running emission rates.
-------
We combined the test data from the passenger cars and the light-duty trucks. Therefore, the
only stratifying parameter in this analysis (of gasoline-fueled vehicles) was the model year
grouping. (Analyses on the heavy-duty vehicles and diesel-fueled vehicles will be presented at a
later meeting.)
Then, within each model year group, we used regression analysis (of cold-start and hot-
running emissions versus temperature) to find a polynomial fit to describe the change in
emissions as a function to temperature.
We limited those polynomials to a multiple of "temperature minus 75° F" to either the first,
second, or third degree. This produced (additive) adjustment factors that exhibit zero change at
75 degrees Fahrenheit.
2.3 Effects of Temperature on Gasoline Fueled Vehicles
Based on earlier analyses, EPA decided to model, in MOVES, the effects of ambient
temperature on HC, CO, and NOx emissions:
1. Using additive (rather than multiplicative) adjustment factors.
2. Using multiples of one of the following:
the temperature minus 75° F, or
the square of the difference of the temperature minus 75° F, or
the cube of the difference of the temperature minus 75° F.
This approach guarantees a value of zero (change) for the additive adjustment factor at
75° F (i.e., the nominal temperature of EPA's FTP test). Those multipliers/coefficients
are stored in the MOVES database table named StartTempAdjustment.
Since the logarithms of the emissions (rather than to the emissions themselves) tend to be
normally distributed (i.e., a log-normal distribution), it is often useful to apply regression
analysis to the logarithms of the emissions. However, restricting the adjustment factors
to one of these three forms made it impractical to use regressions of the logarithms of the
emissions.
3. Setting the value of the adjustment factors equal to zero for temperatures higher than 75°
Fahrenheit.
2.3.1 Temperature Effects on Gasoline Start Emissions
2.3.1.1 HC and CO Start Emissions for Gasoline-Fueled Vehicles:
As described in an earlier analysis, we used the difference in the Bag-1 emissions minus the
corresponding Bag-3 emissions to estimate the cold-start emissions (in grams per start) for each
test. For the gasoline-fueled vehicles, those cold-start emissions were then stratified by model
year group. The mean emissions at 75 °F were subtracted from each of the means to determine
-------
the change in emissions as functions of ambient temperature. (See Appendix A for the resulting
average changes.)
As noted at the beginning of this section, EPA had decided to model the changes in cold-start
emissions as a polynomial (linear, or a quadratic, or a cubic) of the temperature minus 75° F.
Thus, the shape of each adjustment curve at temperature below 75° F would determine the shape
of that curve at temperatures above 75° F. However, the predetermined shape of the curve at
temperatures above 75° F was not always in agreement (directionally) with the test data above
75° F. Therefore, EPA decided to set the value of those additive adjustment factors equal to zero
for temperatures higher than 75° F. We did not use the changes in emissions from temperature
above the FTP temperature range (68° to 86° F); however, those values are included (if available)
in Appendix A.
We performed a linear, quadratic, and cubic regressions on the data in Appendix A and then
selected the best fit from among those three. The following equations were, thus, chosen as the
"best fit" predictors of the change in cold-start emissions (in grams) as functions of the ambient
temperature:
For the Pre-198Is:
HC temperature Adjustment = tempAdjustTermA * (Temp. - 75)
where: tempAdjustTermA =-0.630705748 R-sqr = 0.99271
CO temperature Adjustment = tempAdjustTermA * (Temp. - 75)
where: tempAdjustTermA =-4.677330289 R-sqr = 0.98973
Each of those linear coefficients is stored in table StartTemp Adjustment, (for the cold-start,
i.e., opModelDof 108)
For the 1981-1982s:
HC temperature Adjustment = tempAdjustTermA * (Temp. - 75)
where: tempAdjustTermA =-0.413584322 R-sqr = 0.98368
CO temperature Adjustment = tempAdjustTermA * (Temp. - 75)
where: tempAdjustTermA =-4.630546442 R-sqr = 0.97761
For the 1983-1985s:
HC temperature Adjustment = tempAdjustTermA * (Temp. - 75)
where: tempAdjustTermA =-0.360706640 R-sqr = 0.88660
CO temperature Adjustment = tempAdjustTermA * (Temp. - 75)
where: tempAdjustTermA =-4.244442967 R-sqr = 0.96367
For the 1986-1989s:
HC* temperature Adjustment = tempAdjustTermB * Sqr_of (Temp. - 75)
where: tempAdjustTermB = 0.002413998 R-sqr = 0.98895
CO temperature Adjustment = tempAdjustTermA * (Temp. - 75)
where: tempAdjustTermA =-1.089740827 R-sqr = 0.99401
-------
HC test data for this model year range were available down to an ambient
temperature of-20° F. However, the "best fit" HC regression curves (linear,
quadratic, and cubic) all exhibited less than ideal fits to those data at temperatures
from zero through 20° F. Deleting the test data at -20° F and rerunning the
regressions produced an improved estimate of the cold-start HC emissions in that
critical temperature range. Therefore, this proposed quadratic regression is based on
the changes in cold-start emissions at only temperatures from zero through 75° F.
For the 1990-2005s:
HC* temperature Adjustment = tempAdjustTermB * Sqr_of (Temp. - 75)
where: tempAdjustTermB = 0.002924240 R-sqr = 0.99409
CO* temperature Adjustment = tempAdjustTermA * (Temp. - 75) (Eqn x.x)
where: tempAdjustTermA =-1.141434345 R-sqr = 0.99017
* As with the regressions performed on the test data from the 1986 through 1989 model
years, both the HC and CO regressions produced superior estimators of both HC and
CO cold-start emissions (at temperatures above zero degrees F) when the test data at -
20° F was omitted. Therefore, both of these regressions were based on the changes in
cold-start emissions only at temperatures from zero through 75° F.
2.3.1.2 Temperature Effects on Gasoline NOx Emissions
For the effects on cold-start NOx emissions associated changes in ambient temperature, we
attempted the same model year stratification that we used for the HC and CO emissions.
However, as is illustrated in the following graph (Figure 1), the "by model year" temperature
effects on cold-start NOx emissions did not lend themselves to linear, quadratic, or cubic
regressions (possibly due to insufficient sample size). Also, not unexpectedly, most of the
coefficients produced by those regression analyses were not statistically significant.
-------
Figure 2-1 Effects of Ambient Temperature on Changes in Cold-Start NOx
9.0
6.0
3.0
0.0
-3.0
-20
20 40 60 80 100
Temperature (degrees Fahrenheit)
120
A visual inspection of Figure 1 suggests that only three model year groups (1990-1993, 2001,
and 2005) exhibited patterns that would result in meaningful regression analyses. We attempted
to group the data into various other model year groups. The only grouping that produced useful
regression analyses was the one in which we average together all of the NOx results (from
Appendix A) to obtain the following table:
Delta NOx
Temp (grams)
-20.0 1.201
0.0 1.227
19.4 0.202
20.7 0.089
22.4 -0.155
Delta NOx
Temp (grams)
31.0 -0.007
40.0 0.876
48.8 0.127
49.8 0.333
5i:0 0.325
Delta NOx
Temp (grams)
54.2 0.438
76.3 0.000
95.3 0.225
97.1 0.370
105.8 0:543
-------
Performing regression analyses on these data (again, using only the changes in the NOx cold-
start emissions for temperatures below 86° F as explained in Section 3.2), we found the "best fit"
equation to be:
NOx temperature Adjustment = tempAdjustTermA * (Temp. - 75)
where: tempAdjustTermA =-0.009431682 R-sqr = 0.611349
Although the value of R-squared is not as high as for the HC and CO regression equations,
the coefficient is statistically significant. If we were to evaluate that equation for temperatures
higher than 75° F, it would predict a negative change (i.e., a decrease) in the cold-start NOx
emissions (i.e., a decrease in cold-start NOx emissions), but the actual data indicate that the cold-
start NOx emissions increase as the ambient temperature rises above 90° F. Therefore (as with
the previous adjustment factors), this additive adjustment factor is set to zero for temperatures
higher than 75° F.
2.3.1.3 Temperature Effects on Gasoline PM Emissions
The effects on both cold-start and running emissions of particulate matter (PM) associated
changes in ambient temperature will be modeled (in MOVES) using a multiplicative (not
additive) exponential (not polynomial) adjustment factor. The analysis for that factor is included
as Chapters 7 and 8 of a separate report ("Analysis of Parti culate Matter Emissions from Light-
Duty Gasoline Vehicles in Kansas City"). [4]
2.3.2 Temperature Effects on Gasoline Running Emissions
The test data analyzed to determine the effects of different ambient temperatures on running
emissions consisted of:
1. Bag-2 of the FTP for vehicles tested at multiple temperatures,
2. US06 for vehicles tested at multiple temperatures, and
3. Remote sensing data (RSD) on a random sample of vehicles
tested at Kansas City over a wide range of temperatures.
4. FTP and IM240 tests on a random sample of vehicles tested
at Kansas City
Those test data suggest that there is very little variation in those running emissions of HC,
CO, or NOx. Regression analyses found that the coefficients (slopes) were not statistically
significant (that is, the slopes were not distinguishable from zero). This is consistent with what
we found in our analysis of the Kansas City data. This lack of correlation between running
emissions and ambient temperature is illustrated (as an example) by the following graph of the
HC data:
-------
Figure 2-2 Logarithm of Bag-2 HC Versus Temperature
CM
D)
O
^-2
-3
-4
-5
.0 2C
*
*
*
.0 34
* *
3
* *
* *
t
* *
.0 ' V
ğ Ğ
ğ * *
* * ^
* * *
* ğ *
> **
ĞĞĞğ
* *
*
* ğ
* ^
ğ/ğğ..
*
^g **Ğ>5C
V*A**'
^ ** *
*^** '
. ** Ğ
ĞĞ * *ğ
** *
A. {*, 6C
\: >*
, ĞĞ **
*
*ğ \
*
iĞ *^7C
(*** ?*
* >
* ğ<
* ğ*
Ğ
0* ğ*Ğğ ^
^A:";
ğ\V*
%/****
I *' *
*
*
,ğ
:(fğ\9c
* ğ
V
\ Ğ
*
*
.t) 10
D.O
Temp
In this plot, each point represents a single FTP Bag-2 test result from the Kansas City
program. A visual inspection of this plot of the natural log of the FTP Bag-2 HC emissions
suggests no strong relationship between the hot-running HC emissions and the ambient
temperature.
The CO and NOx plots are similar in that they also do not indicate a significant trend.
We looked at the second-by-second test data from IM240s run in Chicago (as part of
Chicago's I/M program) to validate this conclusion. To avoid the issue of preconditioning, we
used only second IM240s when back-to-back IM240s were performed, and for the other IM240s
we examined the last 120 seconds of full duration IM240s. We found no evidence of a trend /
effect between 5 and 95 degrees F.
The effect of temperature on hot running HC, CO, and NOx emissions will be modeled in
MOVES using polynomial functions as multiplicative adjustment factors. In this version of
MOVES, we propose to set all of those adjustment factors equal to 1.0, that is, no change in
those running emissions with temperature.
This was not the case for PM emissions which are discussed in Section 4.1.3.
-------
2.4 Effects of Temperature on Diesel Fueled Vehicles
We were able to identify only 12 diesel-fueled vehicles with FTPs at multiple temperatures
(nine passenger cars and 3 light-duty trucks). However, only two of those 12 vehicles were
tested at temperatures within the normal FTP range (68° to 86° F). The Bag-1 minus Bag-3
emissions for those tests are given below. We stratified the test results into four temperature
bands which yielded the following values (grams per start):
Count HC CO NOx
6 2.55 2.44 2.60
7 2.68 2.03 0.32
10 1.69 3.00 0.67
2 1.20 1.91 0.36
When we plotted the mean HC start emissions (above) versus temperature, we obtained the
following graph with 90 percent confidence intervals (and a "dashed" linear regression line).
Figure 2-3 Cold-Start HC Emissions (in grams) with Confidence Interval
30
40 50 60
Temperature (degrees F)
70
The dashed (blue) line in the figure is a linear regression line having as its equation:
HC = (-0.0420985982 * Temperature ) + 4.22477812 R-sqr = 0.9040467
Transforming this equation into an equation that predicts the (additive) change/adjustment in
the cold-start HC emissions from light-duty diesel-fueled vehicles (in the MOVES format), we
obtain:
-------
HC temperature Adjustment = tempAdjustTermA * (Temp. - 75)
where: tempAdjustTermA = -0.0420985982
The coefficient associated with this temperature adjustment term is statistically significant
although its coefficient of variation is relatively large (23.04 percent).
Again, this HC adjusstment factor represent the difference of Bag-1 minus Bag-3 and must
be adjusted to estimate the cold-start HC emissions.
It proved more difficult to repeat this approach for the cold-start CO and NOx emissions
from those same diesel-fueled light-duty cars and trucks because the cold-start CO and NOx
emissions did not exhibit a clear trend relative to the ambient temperature. Plotting the mean CO
and NOx cold-start emissions versus ambient temperature (with 90 percent confidence intervals)
produced the following two graphs:
Figure 2-4 Bag-1 minus Bag-3 CO Emissions (in grams) with Confidence Interval
i
30
40 50 60
Temperature (degrees F)
70
10
-------
Figure 2-5 Bag-1 minus Bag-3 NOx Emissions (grams) with Confidence Intervals
0
30
40 50 60
Temperature (degrees F)
70
Statistical analyses of both the diesel cold-start CO and NOx emissions failed to produce
coefficients that were significantly different from zero. Therefore, for both cold-start CO and
NOx adjustments from light-duty diesel-fueled vehicles, we propose to use:
CO temperature Adjustment = tempAdjustTermA * (Temp. - 75)
where: tempAdjustTermA = 0.0
NOx temperature Adjustment = tempAdjustTermA * (Temp. - 75)
where: tempAdjustTermA = 0.0
That is, neither the CO nor the NOx start emissions for diesel-fueled vehicles will vary with
changes in the ambient temperature. This includes all emissions from the extended idling
emission process for heavy duty long haul diesel trucks.
2.5 Cold Weather Effects
There are two sets of regulations that can affect our estimates of emissions at low
temperature (i.e., at 20 degrees Fahrenheit), namely the cold weather CO requirement and the
cold weather HC requirement.
11
-------
2.5.1 Cold Weather CO Requirement
The cold weather CO requirement for the 1994 and newer model year LDVs and LDTs limits
the composite FTP CO emissions to 10.0 grams per mile at a temperature of 20 degrees
Fahrenheit. However, the FTP test results used for our analysis (for those model years) were
from vehicles that were certified as meeting that cold weather composite CO requirement. Thus,
the temperature adjustments (based on regressions of those FTP results) already incorporated
that cold weather CO requirement into MOVES.
2.5.2 Cold Weather HC Requirement
The recently signed Mobile Source Air Toxic (MSAT-2) rule included a limit on low
temperature (i.e., at 20 degrees Fahrenheit) non-methane hydrocarbon (NMHC) emissions for
light-duty and some medium-duty gasoline-fueled vehicles. Specifically:
For passenger cars (LDVs) and for the light light-duty trucks (LLDTs) (i.e., those with
GVWR up to 6,000 pounds), the composite FTP NMHC emissions should not exceed 0.3
grams per mile.
For heavy light-duty trucks (HLDTs) (those with GVWR from 6,001 up to 8,500 pounds)
and for medium-duty passenger vehicles (MDPVs), the composite FTP NMHC emissions
should not exceed 0.5 grams per mile.
These cold weather standards are to be phased-in beginning with the 2010 model year,
specifically:
Phase-In of Vehicles Meeting Cold Weather HC Standard
Model Year
2010
2011
2012
2013
2014
2015
LDVs / LLDTs
25%
50%
75%
100%
100%
100%
HLDTs / MDPVs
0%
0%
25%
50%
75%
100%
To incorporate this set of HC requirements into MOVES, we must first determine its impact
on the start emissions (both cold-start and hot-start) as well as on the running emissions for each
class of vehicles.
We already observed that changes in the ambient temperature do not have a significant effect
on the running THC emissions. Therefore, we will assume that the full impact of this
requirement will be on the start emissions.
Our earlier analysis of temperature effects on the emissions of Tier-2 vehicles was based on a
single gasoline-fueled passenger car and three light-duty trucks that were each FTP tested at
12
-------
zero, 20, and 75 degrees Fahrenheit. The average nonmethane HC (NMHC) composite FTP
emissions at 75° F were:
0.02 (0.0180) g/mile for the passenger car and
0.04 (0.0353) g/mile for the heavy light-duty trucks (over 6,000 GVWR).
Considering the MSAT-2 standards (0.30 and 0.50, respectively), this would mean the
NMHC composite FTP emissions increasing by no more than 0.28 grams per mile (i.e., 0.30
minus 0.02) for LDVs/LLDTs and by no more than 0.46 grams per mile for HLDTs/MDPVs as
the ambient temperature drops from 75° F down to 20° F.
Applying those increases in NMHC emission rates to the composite FTP (which simulates a
trip of 7.45 miles in length), those rates convert to total NMHC increases of 2.086 grams (for
LDVs/LLDTs) and 3.427 grams (for HLDTs/MDPVs). Since a composite FTP is composed of a
7.45 mile driving cycle plus a generic engine start (57 percent hot-start and 43 percent cold-
start), those increases must represent the increases in the generic start emissions. Using the ratio
of hot-start to cold-start from our earlier analysis, this results in increases in NMHC cold-start
emissions (as the ambient temperature drops from 75° F down to 20° F) of:
0.5611592 grams for the LDVs/LLDTs and
0.9219045 grams for the HLDTs/MDPVs.
Since the analysis for the MSAT-2 rule assumed that increase in NMHC is linear with
temperature (decreasing 55 degrees from 75 down to 20), then those rates convert to decreases in
total NMHC per cold-start of:
-0.0102029 grams per degree F for the LDVs/LLDTs and
-0.0167619 grams per degree F for the HLDTs/MDPVs.
These are the rates (slopes) that we propose to use in MOVES for cold-starts (i.e., starts that
follow a 12 hour engine soak). For the seven shorter soak periods (that MOVES uses as
opModes), we will continue to use the ARB soak adjustments for HC emissions for catalyst
equipped vehicles to estimate those HC emissions (following the seven shorter soak periods).
2.5.3 Cold Weather PM Effects
The MSAT-2 rule (signed February 9, 2007) does not explicitly limit cold weather emissions
of particulate matter (PM). However, the Regulatory Impact Analysis (RIA) document [5] that
accompanied that rule noted there is a strong linear correlation between NMHC and PM2.5
emissions. That correlation is illustrated in the following graph (reproduced from that RIA) of
the logarithm of the Bag-1 PM2.5 versus the logarithm of the Bag-1 NMHC (for various Tier-2
vehicles).
13
-------
Figure 2-6 FTP Bag 1 PM and FTP Bag 1 NMHC for Tier 2 Vehicles
Q_
CD
m
CD
i
CO
I
o
+
v
V O £
V ^> o
v v
X
-3
A
-2
\
-1
0
1
Bag1 NHMC-ln(g/mi)
Plot Icons are Vehicle-Specific
Therefore, the limitation on cold weather HC (or NMHC) emissions is expected to result in
an ancillary reduction in cold weather PM2.5 emissions. In the MSAT-2 RIA (Table 2.1 .-9),
EPA estimated that this requirement would result in a 30 percent reduction of VOC emissions (at
20° F). Also, in the RIA, the ratio of PM to NMHC equaling 0.022 was used to estimate that
PM2.5 reduction. (The 95 percent confidence interval for that ratio was 0.020 to 0.024.)
Applying the same analytical approach that was used in the RIA means that a 30 percent
reduction in VOC emissions would correspond to a 30 percent reduction in PM emissions at 20°
F (for Tier-2 cars and trucks).
EPA's earlier analysis (for MOVES) [4] indicated that ambient temperature does affect the
rate of running PM emissions as well as start PM emissions, and that effect (for Tier-2 vehicles)
is best modeled by (exponential) multiplicative adjustment factors of the form:
A*(72-t)
Multiplicative factor = & , where "t" is the ambient temperature
and where A = 0.0463 for cold-starts and
0.0318 for hot running
(See Table 12 in Reference [4], page 46.)
14
-------
Therefore, for Tier-2 vehicles not affected by the MSAT-2 requirements, EPA expects (as the
temperature decreases from 72° down to 20° F) the PM emissions to increase by factors of:
11.10727 for cold-starts and
5.22576 for hot running.
Thus, applying that 30 percent reduction for vehicles that are affected by the MSAT-2
requirements produces estimates (as the temperature decreases from 72° down to 20° F) of PM
emissions increasing by factors of:
7.77509 for cold-starts and
3.65803 for hot running.
Since the vehicles affected by the MSAT-2 requirements begin to be phased-in starting with the
2010 model year, EPA expects the following (multiplicative) increases (as the temperature
decreases from 72° down to 20° F):
Multiplicative Increases of PM at 20° Fahrenheit
Model Year
2008
2009
2010
2011
2012
2013
2014
2015
LDVs/
Start
11.10727
11.10727
10.27423
9.44118
8. 60814
7.77509
7.77509
7.77509
LLDTs
Running
5.22576
5.22576
4.83383
4.44189
4.04996
3.65803
3.65803
3.65803
HLDTs /
Start
11.10727
11.10727
11.10727
11.10727
10.27423
9.44118
8.60814
7.77509
MDPVs
Running
5.22576
5.22576
5.22576
5.22576
4.83383
4.44189
4.04996
3.65803
Solving for the corresponding constant terms so that the preceding exponential equation will
yield these increases, gives us these "A" values:
15
-------
Constant Terms
Model Year
2008
2009
2010
2011
2012
2013
2014
2015
LDVs/
Cold-Start
0.046300
0.046300
0.044801
0.043175
0.041398
0.039441
0.039441
0.039441
LLDTs
Running
0.031800
0.031800
0.030301
0.028675
0.026898
0.024941
0.024941
0.024941
HLDTs /
Cold-Start
0.046300
0.046300
0.046300
0.046300
0.044801
0.043175
0.041398
0.039441
MDPVs
Running
0.031800
0.031800
0.031800
0.031800
0.030301
0.028675
0.026898
0.024941
We assume that these same magnitude increases in the PM2.5 emissions also apply to the EC
and OC emissions.
Although the ARB factors that adjust the start emissions based on soak time were not
developed for PM emissions from gasoline-fuel vehicles, the fact that the ratio of PM emissions
to the HC emissions are almost constant suggests that we can apply the HC soak adjustment
factors to the start PM emissions.
3. Humidity Adjustments
In EPA's previous emissions model (MOBILE6), only gasoline vehicle exhaust NOx
emissions were adjusted for humidity. MOVES adjusts both gasoline and diesel vehicle exhaust
NOx emissions. The base exhaust emission rates for NOx in all modes and all processes are
multiplied by a humidity correction factor. This factor is calculated using the following formula:
K = 1.0 - ( (Bounded Specific Humidity - 75.0) * Humidity Correction Coefficient)
The bounded specific humidity is in units of grains of water per pound of dry air. The
specific humidity is not allowed to be lower than 21 grains and is not allowed to be larger than
124 grains. If the specific humidity input exceeds these limits, the value of the limit is used to
calculate the humidity correction factor. Appendix B shows how the hourly relative humidity
values are converted to specific humidity used in this equation using temperature and barometric
pressure.
lumidity Correction Coefficients Used by MOVES
Fuel Type
Gasoline
Diesel Fuel
Humidity Correction Coefficient
0.0038
0.0026
16
-------
The diesel humidity correction coefficient is taken directly from the Combined Federal
Register [6]. The gasoline humidity correction coefficient is carried over from the coefficient
used in the MOBILE6 model.
4. Air Conditioning Adjustments
Revised air conditioning exhaust emission correction factors are included in the MOVES
model. The proposed factors are based on testing of 54 vehicles and 625 driving cycle tests in
calendar years 1997 and 1998. All "A/C On" testing was done at a nominal temperature of 95 F,
using a test procedure meant to simulate air conditioning emission response under extreme "real
world" ambient conditions. These factors are meant to predict emissions which would occur
during full loading of the air conditioning system, and will be scaled down in MOVES according
to ambient conditions in a modeling run. The second-by-second emission data from each
individual vehicle-cycle combination were analyzed using the MOVES methodology of binning
the data according to vehicle characteristics (source bins in MOVES) and vehicle specific power
bins (operating modes in MOVES). The results of the analysis showed statistically significant
and consistent results for three bin combinations (deceleration, idle and cruise/acceleration) and
the three primary exhaust pollutants (hydrocarbon, carbon monoxide and nitrous oxides). This
report shows the results of the analysis for the air conditioning adjustments used in MOVES for
HC, CO, NOx and energy consumption.
Past studies conducted in 1997 and 1998 as part of the Supplemental Federal Test
Procedure (SFTP) rulemaking development process indicated that vehicle fuel consumption and
exhaust emissions increase substantially when the air conditioner is in operation. During these
studies vehicles were tested for exhaust emissions under full usage temperature, humidity and
solar loading conditions, and at baseline conditions. These studies provided data that was
subsequently used to develop multiplicative correction factors that represent full or maximum
A/C system usage. In the MOBILE6.2 model these maximum A/C correction factors were
scaled down so as to model more normal levels of A/C demand [7].
The past analysis work was fairly complex and the reports present considerable detail in
regards to the vehicle testing protocols, the work to correlate data between the two tests sites and
expected real-world results, the data analysis and development of correction factors that can be
used to model a range of ambient conditions. For a detailed discussion of the test data and the
subsequent data analysis the reader is referred to the MOBILE6 correction factor report [8]. The
MOVES analysis also differs considerably from the MOBILE6 model analysis. The previous
analysis focused on the development of detailed mathematical algorithms which were inserted
into the MOBILE6.2 model and the adjustments were only applied to exhaust emissions of
oxides of nitrogen (NOx). The MOVES model is a data driven and empirical model which
contains simple data relationships of highly detailed modal data.
MOVES will make adjustments to total energy consumption and exhaust running HC, CO
and NOx emissions separately for each operating mode. The criteria pollutants (HC, CO and
NOx) are only affected for passenger car, passenger truck and commercial light truck source
types. Energy consumption is affected for all source types. The same adjustment values are
used for all source use types affected within a pollutant type.
17
-------
4.1 Air Conditioning Effects Data
As mentioned in the previous section, the data for the MOVES A/C Correction Factors
(ACCF) was collected in calendar year 1997 and 1998 in specially designed test programs. In
the programs the same set of vehicles were tested at standard FTP test conditions (baseline) and
at a nominal temperature of 95 F. Use of the same set of vehicles and test cycles should
eliminate most of the vehicle and test procedure variability and highlight the difference between
a vehicle operating at extreme ambient conditions and at a baseline condition.
The data used to develop the MOVES ACCF consisted of 54 individual cars and light
trucks tested over a variety of test schedules. Overall the database consisted of a total of 625 test
cycles, and 1,440,571 seconds of emission test and speed / acceleration data. Because of the
need to compute vehicle specific power on a modal basis, only test results which consisted of
second by second data were used in the analysis. All second by second data were time aligned
and quality controlled checked.
The model year breakdown of the data is shown in Table 4-1. It shows that all of the
vehicles were 1990 through 1999 model years. It consists of 30 cars and 24 light trucks. No test
data were available on other vehicle types (i.e., MC, heavy trucks, etc). The individual test
cycles which the vehicles were run on are shown with the test counts in Table 4-2. The data
shows a nice balance between different test cycles, and cars and trucks. Unfortunately, the study
does not contain any pre-1990 model years. A complete list of the individual vehicles and a
basic description is shown in Appendix A.
Only vehicles which were coded as having an emission test with the A/C system on were
selected. The A/C On tests and the A/C Off (default for most EPA emission tests in general)
were matched by VIN, test schedule and EPA work assignment. The matching ensured that the
same vehicles and test schedules were contained in both the A/C On sample and the A/C Off
sample.
18
-------
Table 4-1 Distribution of Test Vehicles by Model Year
Model Year
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
TOTAL
Count
5
5
6
5
7
5
13
4
3
1
54
Table 4-2 Distribution of Tests by Schedule Type
Schedule Name
ART-AB
ART-CD
ART-EF
F505
FTP
FWY-AC
FWY-D
FWY-E
FWY-F
FWY-G
FWY-HI
LA4
LA92
LOCAL
NONFRW
NYCC
RAMP
ST01
TOTAL
Count
36
36
36
21
21
57
36
36
36
36
36
23
35
36
36
36
36
36
625
19
-------
4.2 Method for Calculating Air Conditioning Effects
The overall dataset consisted of a sample of vehicle tests with the A/C system on and a
sample of vehicle tests with the A/C system off. Both samples consisted on the same vehicles
and all tests were modal with a data sampling of 1 hertz (second-by-second data collection).
Prior to analysis the data for each vehicle / test cycle combination was time aligned to insure that
the instantaneous vehicle operating mode was in-sync with the emission collection system.
Following time alignment, the vehicle specific power (VSP) was calculated for each vehicle test
/ second combination. This was done using Equation 1.
VSP = 985.5357* Speed* Acoeff/Weight +
440.5729 * SpeedA2 * Bcoeff /Weight +
196.9533 * SpeedA3 * Ccoeff/Weight +
0.19984476 * Speed * Accel + GradeTerm Eq 1
Where
VSP is the vehicle specific power for a given second of operation in units of KW / tonne.
Speed is the instantaneous vehicle speed for a given second in units miles / hour.
Accel is the instantaneous vehicle acceleration for a given second in unit of miles/hr-sec
Weight is the test vehicle weight in pounds.
Acoeff = 0.7457*(0.35/(50*0.447)) * ROAD_HP
Bcoeff = 0.7457*(0.10/(50*50*0.447*0.447))*ROAD_HP
Ccoeff = 0.7457*(0.55/(50*50*50*0.447*0.447*0.447)) * ROAD_HP
Where
ROAD_HP = 4.360117215 + 0.002775927* WEIGHT (for cars)
ROAD_HP = 5.978016174 + 0.003165941 * WEIGHT (for light trucks)
GradeTerm (KW/tonne) = 4.3809811 * Speed * Sin(Radians(GradeDeg))
Where
GradeDeg is the road grade in units of degrees. This term is zero for dynamometer tests.
4.3809811 (mA2 * hr / (sA3 * miles) =
9.80665(m/sA2) * 1609.34(m/mile) / 3600(secs/hr)
KW / tonne = mA2 / sA3
9.80665(m/sA2) is the gravitation constant.
After computation of VSP for each vehicle test / second combination, the individual VSPs'
were grouped into the VSP bins. These VSP bins are defined in Table 3. VSP bins 26 and 36
were not defined because bins 27-30 and bins 37-40 overlap them.
20
-------
VSP Bin Definitions
VSP Label
0
1
11
12
13
14
15
16
21
22
23
24
25
26
27
28
29
30
33
35
36
37
38
39
40
Definition
Braking
Idling
Low Speed Coasting; VSP< 0; K=Speed<25
Cruise/ Acceleration; 0<=VSP< 3; 1<= Speed<25
Cruise/ Acceleration; 3<=VSP< 6; K=Speed<25
Cruise/ Acceleration; 6<=VSP< 9; K=Speed<25
Cruise/Acceleration; 9<=VSP<12; K=Speed<25
Cruise/Acceleration; 12<=VSP; K=Speed<25
Moderate Speed Coasting; VSP< 0; 25<=Speed<50
Cruise/ Acceleration; 0<=VSP< 3; 25<=Speed<50
Cruise/ Acceleration; 3<=VSP< 6; 25<=Speed<50
Cruise/ Acceleration; 6<=VSP< 9; 25<=Speed<50
Cruise/ Acceleration; 9<=VSP<12; 25<=Speed<50
Cruise/ Acceleration; 12<=VSP; 25<=Speed<50
Cruise/ Acceleration; 12<=VSP<18; 25<=Speed<50
Cruise/ Acceleration; 18<=VSP<24; 25<=Speed<50
Cruise/ Acceleration; 24<=VSP<30; 25<=Speed<50
Cruise/ Acceleration; 30<=VSP; 25<=Speed<50
Cruise/ Acceleration; VSP< 6; 50<=Speed
Cruise/Acceleration; 6<=VSP<12; 50<=Speed
Cruise/ Acceleration; 12 <= VSP; 50<=Speed
Cruise/ Acceleration; 12<=VSP<18; 50<=Speed
Cruise/ Acceleration; 18<=VSP<24; 50<=Speed
Cruise/ Acceleration; 24<=VSP<30; 50<=Speed
Cruise/Acceleration; 30<=VSP; 50<=Speed
21
-------
4.3 Air Conditioning Effects on Emissions
4.3.1 A/C Correction Factors for HC, CO and NOx Emissions
An average emission result for each pollutant (HC, CO and NOx) with and without A/C
operation was computed for each VSP Bin. This resulted in 69 (23 VSP bins x 3 pollutants)
pairs of emission averages. However, preliminary analysis of the data grouped into the 23 bins
(defined in Table 3) showed unsatisfactory statistical results. In the general, no trends were
evident across VSP bins or within similar subsets of VSP bins. The trends were highly erratic
and the results were generally not statistically significant. In addition, most of the bins labeled
30 or higher had very few data members. An analysis of cars versus trucks was also performed,
and showed no statistical difference between the two.
To produce more consistent results, the individual VSP bins were collapsed down to three
principal bins. These are the Braking / Deceleration bin, the Idle bin and the Cruise /
Acceleration bin. All three of these large bins are really quite different in terms of engine
operation and emissions performance. The Braking bin consisted of VSP Bin 0 in Table 3, the
Idle bin was VSP Bin 1 and the Cruise / Acceleration bin contained the remaining 21 bins. Full
A/C correction factors were generated for each of the nine VSP Bin and pollutant combinations.
This was done by dividing the mean With A/C emission factor by the mean Without A/C
emission factor for each of the VSP Bin / pollutant combinations. The Full A/C correction
factors are shown in Table 4. Measures of statistical uncertainty (coefficient of variation of the
mean) were also computed using the standard error of the mean. They are also shown in Table 4
in the column labeled Mean CV of CF.
Full Air Conditioning Correction Factors for HC, CO and NOx
Pollutant
HC
HC
HC
CO
CO
CO
NOx
NOx
NOx
Operating Mode
Braking / Decel
Idle
Cruise / Accel
Braking / Decel
Idle
Cruise / Accel
Braking / Decel
Idle
Cruise / Accel
Full A/C CF
1.0000
1.0796
1.2316
1.0000
1.1337
2.1123
1.0000
6.2601
1.3808
MeanCVofCF
0.48582
0.74105
0.33376
0.31198
0.77090
0.18849
0.19366
0.09108
0.10065
22
-------
4.3.2 Full A/C Correction Factors for Energy Emissions
The use of a vehicle's A/C system will often have a sizeable impact on the vehicle's energy
consumption. This was found statistically by analyzing the available second by second data on
CO2 and other gaseous emissions, and converting them to an energy basis using standard EPA
vehicle fuel economy certification equations. The vehicle emission data were binned by VSPBin
(see above for explanation of VSPBin). A mean value was computed for each combination of
VSPBin. Separate analysis was done as a function of sourcebinid (combination of vehicle type,
fuel type and model year), and the results were not statistically different versus sourcebinid
given the relatively small sample sizes. As a result, the A/C correction factors for energy are a
function of only VSPBin. The resulting A/C correction factors are shown in Table 5.
Full Air Conditioning Correction Factors for Energy
VSPBin
0
1
11
12
13
14
15
16
A/C Factor
1.342
1.365
1.314
1.254
1.187
1.166
1.154
1.128
VSPBin
21
22
23
24
25
26
27
28
29
A/C Factor
1.294
1.223
1.187
1.167
1.157
1.127
1.127
1.127
1.127
VSPBin
30
33
35
37
38
39
40
A/C Factor
1.294
1.205
1.156
1.137
1.137
1.137
1.137
Only very small amounts of data were available for VSPBins 26 through 29 and VSPBins
37 through 40. As a result, the data from these bins was averaged together and binned into two
groups. The resulting group averages were used to fill the individual VSPBins. This averaging
process has the effect of leveling off the effect of A/C at higher power levels for an engine. This
is an environmentally conservative assumption since it is likely that engine power devoted to an
A/C compressor probably continues to decline as the overall power demand of the engine is
increased. In fact, in some newer vehicle designs the A/C unit will be shut off by an engine
controller, if the driver demands a very high level of power from the vehicle. If an when new or
additional data become available on this issue, EPA will re-evaluate the assumption of a constant
A/C factor for the high VSPBins.
4.3.3 Uncertainty Analysis
Measures of statistical uncertainty -coefficient of variation of the mean (mean CV) were
calculated using the following formula. The exact set of equations were used for each of the
three pollutants (although the equation are shown only once). The values of X and Y represent
second by second emissions from HC, CO and NOx. The variable "X" represents emissions
with the A/C On and "Y" represents emission with the A/C Off.
Given:
23
-------
Z X /Y
MeanCV = SEZ/Z
Where Z is the ratio of A/C On emissions (X) to A/C Off emissions (Y)
SEZ is the standard error of Z
Mean CV is the coefficient of variation of the mean
Vz2 = (5Z/5X)2 * Vx2 + (5Z/5Y)2 * Vy2
Where Vz is the variance of Z, Vx is the variance of X and Vy is the variance of Y
5Z/5X is the partial derivative of Z with respect to X
5Z/5Y is the partial derivative of Z with respect to Y
(Vz / Z)2 = ((1/Y2)*VX2) / (X2/Y2) + ((X2/Y4) * Vy2) / (X2/Y2)
This equation reduces down to:
(VZ/Z)2 = (Vx/X)2 + (Vy/Y)2
And ultimately to:
SEZ / Z = SQRT [ (SEZ / X)2 + (SEZ / Y)2 ]
The variance term is defined as:
Vz = (1/Y)2 * Sy2x + (-X/Y2) * (-X/Y2) * Sy2y;
Where
X = A/C On emissions
Y = A/C Off emissions
The term Vz represents a contribution from both the X and Y emissions terms (A/C On and A/C
Off). The terms Sy2x and Sy2y also include variance contributions of the "across sample
variance" and the "within a given vehicle test" variance. The "across sample variance" is the
standard variance of the sample and is computed within a given sourcetype (vehicle type such as
car, light truck, heavy truck, etc) and operating mode bin (one of the 23 VSP bin types - See
Table 3). The "within a given vehicle test" variance is the additional variance due to the fact
that each vehicle test contributes hundreds or even thousands of test data elements. Because two
data elements may come from the same vehicle, they are not strictly independent of each other.
Sy2x = SA2x/nVeh + SB2x/nCell
Sy2y = SA2y/nVeh + SB2y/nCell
24
-------
2
SA2x = ( 1 / (nVeh-1) ) * Sumlx
SB2x = ( 1 / (nCell - nVeh) ) * Sum2x
SA2y = (II (nVeh-1) ) * Sumly
SB2y = ( 1 / (nCell - nVeh) ) * Sum2y
And
Sumlx = Z ( YbarVehx - YbarCellx )2
Sum2x = Z ( varVehx - ( nMeas - 1) )
Sumly = Z ( YbarVehy - YbarCelly )2
Sum2y = Z ( varVehy - ( nMeas - 1) )2
Where
The sums ( Z ) are across sourcetype and operating mode.
nMeas Count of data elements within a given sourcetype, operating mode and vehicle
test.
nVeh Count of data elements within a given vehicle test
nCell Count of data elements within a given sourcetype and operating mode
varVeh Variance for each vehicle test. Separate values for both X and Y are calculated.
YbarVeh Mean emission rate for each vehicle test. Separate values for both X and Y are
calculated.
YbarCell Mean emission rate for each sourcetype and operating mode. Separate values for
both X and Y are calculated.
Except for broad groupings, VSP was not found to be an important variable in regards to A/C
correction factor and A/C usage. However, Full A/C correction factors greater than unity were
found for all pollutants for both Idle and Cruise / Acceleration modes. For NOx Idle mode, a
fairly large multiplicative correction factor of 6.2601 was obtained. This large factor reflects the
relatively low levels of NOx emissions during idle operation. A moderately high multiplicative
A/C correction factor of (2.1123) for CO cruise / Accel was also obtained. This correction factor
will double CO emissions under extreme conditions of A/C usage. A/C correction factors of
less than unity or unity where found for the Braking / Deceleration mode for all three pollutants.
These were set to unity for use in the MOVES model.
25
-------
4.4 Adjustments to Air Conditioning Effects
The adjustments for each operating mode are weighted together by the operating mode
distribution calculated from the driving schedules used to represent the driving behavior of
vehicles. Average speed, road type and vehicle type will affect the operating mode distribution.
weightedFull AC Adjustment = SUM( fullACAdjustment*opModeFraction )
Since not all vehicles are equipped with air conditioning and air conditioning is normally not
on all of the time, the full air conditioning effect on emissions is adjusted before it is applied to
the emission rate. The SourceTypeModelYear table of the MOVES database contains the
fraction of vehicles in each model year that are equipped with air conditioning [7].
Fraction of Vehicles Equipped with Air Conditioning
(ACPenetration)
Model Year
1971*
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000**
Passenger Cars
0.592
0.592
0.726
0.616
0.631
0.671
0.720
0.719
0.694
0.624
0.667
0.699
0.737
0.776
0.796
0.800
0.755
0.793
0.762
0.862
0.869
0.882
0.897
0.922
0.934
0.9484
0.9628
0.9772
0.980
0.980
All Trucks and Buses
0.287
0.287
0.287
0.287
0.287
0.311
0.351
0.385
0.366
0.348
0.390
0.449
0.464
0.521
0.532
0.544
0.588
0.640
0.719
0.764
0.771
0.811
0.837
0.848
0.882
0.9056
0.9292
0.950
0.950
0.950
* 1971 model year fractions are applied to all previous model years.
26
-------
Fraction of Vehicles Equipped with Air Conditioning
(ACPenetration)
Model Year
Passenger Cars
All Trucks and Buses
** 2000 model year fractions are applied to all later model years.
Motorcycles are not adjusted for air conditioning.
The fraction of vehicles whose air conditioning is operational varies by age of the vehicle
and is stored in the SourceTypeAge table of the MOVES database.
Fraction of Air Conditioning Units Still Functioning By Age
Age
1
2
3
4
5
6
7
8
9
10
Functioning
1.00
1.00
1.00
0.99
0.99
0.99
0.99
0.98
0.98
0.98
Age
11
12
13
14
15
16
17
18
19
20
Functioning
0.98
0.98
0.96
0.96
0.96
0.96
0.96
0.95
0.95
0.95
Age
21
22
23
24
25
26
27
28
29
30
Functioning
0.95
0.95
0.95
0.95
0.95
0.95
0.95
0.95
0.95
0.95
An equation is used to predict the fraction of those vehicle owners who have air conditioning
available to them that will turn on the air conditioning based on the ambient temperature and
humidity (heat index [7]) of the air outside their vehicles. The heat index values are stored in the
ZoneMonthHour table of the MOVES database.
ACOnFraction = ACActivityTermA
+ heatIndex*(ACActivityTermB + ACActivityTermC*heatIndex)
The fraction of vehicles equipped with air conditioning, the fraction of operational air
conditioning and the fraction of air conditioning use are used to adjust the amount of "full" air
conditioning that occurs in each hour of the day.
AC Adjustment = 1+ ( (weightedFullACAdjustment-1)
* ACPenetration*functioningACFraction*ACOnFraction )
The air conditioning adjustment is a multiplicative adjustment applied to the emission rate
after it has been adjusted for fuel effects.
27
-------
5. Inspection and Maintenance Programs
Inspection and Maintenance (I/M) programs are generically any state-run or locally
mandated inspection of highway motor vehicles intended to identify those vehicles most in need
of repair and requires repairs on those vehicles. Since these programs are locally run, there is
great variability in how these programs are designed and the benefits that they generate in terms
of emission reductions from highway motor vehicles.
5.1 Inspection & Maintenance in MOBILE6
Because MOVES draws heavily on the approaches developed for MOBILE6.2 to represent
the design features of specific I/M programs, it is useful to briefly review these methods. The
reader who is interested in a more thorough treatment of the topic is encouraged to review the
relevant MOBILE documentation [9].
The MOBILE6.2 model used a methodology that categorized vehicles according to emitter
status (High emitters and Normal emitters), and applied a linear growth model to project the
fraction of the fleet that progresses from the Normal emitter to the High emitter status as a
function of age. Average emission rates of High and Normal emitters were weighted using the
High emitter fraction to produce an overall average emission rate as a function of age, model
year group and vehicle type. The emissions generated represented the emissions of the fleet in
the absence of I/M (the No I/M emission rate).
A similar approach was used to generate I/M emission rates. In this case the initial starting
point for the function (where age=0) was the same as the No I/M case. However, the effects of
I/M programs and associated repairs were represented by reductions in the fraction of high
emitters, which consequently affects the average emission level of the fleet. Balancing these
emissions reductions due to I/M repairs were the re-introduction of high emitters in the fleet due
to deterioration of vehicle emission control systems after repairs. The underlying I/M and non-
I/M deterioration rates were assumed to be the same.
With the passage of time, the non-I/M and I/M emission cases diverged from each other with
the I/M function being lower. The percentage difference between these two functions is often
referred to as the overall I/M reduction or I/M benefit.
5.2 Inspection & Maintenance in MOVES
The MOVES emission rates contain estimates of emission levels as a function of age, model
year group and vehicle type for areas where no I/M program exists (the mean base rate, or the
non-I/M reference rates) and for an area representing the "reference I/M program" (the I/M
reference rates). The I/M reference rates were derived using data from the enhanced program in
Phoenix, Arizona, and represent the design features of that program. The difference between the
non-I/M and I/M reference rates are assumed to represent the I/M benefit of Phoenix program
design assuming perfect compliance. Equation 1 shows this relationship in a mathematical form.
28
-------
Standard I/M difference = Eno;m - E;m Eq 1
where Enon_iM and EIM are the non-I/M and I/M reference rates, respectively.
The Phoenix program design was selected as the reference program because virtually all of
the underlying data came from this source, and not due to the strengths or weaknesses of this
specific program. In MOVES, it is this general I/M design which is the model, not the actual
Arizona I/M program as it is operated.
The object of this modeling process is to generate a general model which can be used to
represent all I/M programs in the United States. This goal was achieved by comparing
individual program designs against the reference program for purposes of developing adjustment
to the "standard I/M difference" representing design features differing from those in the
reference program This concept is shown mathematically in Equation 2,
Ep=RE1M+(l-R)EnonlM Eq2
where Ep is the adjusted emission rate for a "target" I/M program, EIM is the reference rate,
EnoniM is the non-I/M reference rate, and R is an aggregate adjustment factor representing the
difference in average emission rates between the target program and the reference program.
Depending on the value of R, Ep may be greater than EnoniM, fall between EnoniM and EIM, or less
than EIM. In general, this framework can, in concept, represent target programs as more effective
or less effective than the reference program. In MOVES, R is referred to as the "IMFactor."
Re-arranging Equation 2 and solving for R gives leads to Equation 3a and 3b. These
equations show the I/M adjustment factor to the ratio of the emission difference between a
proposed I/M program design and the Standard I/M Difference
F - F
R = S. ^L Eq 3
-"IM
5.3 Development of MOVES IMFactors
Early in the MOVES development process it was decided that developing the IMFactors
based on the basis of completely new analysis would prove infeasible. A major obstacle was a
lack of suitable emissions and I/M program data representing the full range of program designs.
Data sets for certain I/M programs (i.e., transient test based programs) were generally quite
complete and robust. However, mass emission results and random vehicles samples were quite
scarce for other test types such as the Acceleration Simulation Mode (ASM), steady-state, idle
tests and OBD-II scans. This situation was particularly true for combinations of old model years
at young ages (i.e., a 1985 model year at age five). As a result, EPA decided to develop
IMFactors based on the representation of relevant design features in MOBILE6.2.
Mechanically, this step was achieved by running the MOBILE6.2 model about 10,000 times over
a complete range of pollutant-process combinations, inspection frequencies, calendar years,
vehicle types, test types, test standards, and model year group / age combinations. The mean
29
-------
emission results for each combination were extracted from the output and utilized. The
IMF actor table includes the following fields:
Pollutant / Process
Test Frequency
Test Type
Test Standard
Regulatory Class
Fuel Type (Only gasoline/ethanol fuels have IMFactors)
Model Year Group
Age Group
IMF actor
The IMFactor was computed for each combination of the parameters listed in the IMF actor
table. A separate MOBILE6.2 run was done for each parameter combination (Target design, Ep
and a second set of runs were done describing the reference program (Reference design, ER).
The IMFactor is the ratio of the mean emission results from these two runs. Equation 4
illustrates the simple formula.
E
R =-£- Eq4
p M
The Reference program has inputs matching the Phoenix I/M program during the time in
which the data used in the MOVES emission rate development were collected (CY 1995-2005).
The Reference design represents a biennial frequency with an exemption period for the four most
recent model years.. It uses three different I/M test types (basic idle test for MY 1960-1980,
transient tailpipe tests for MY 1981-1995 (IM240, EVI147), and OBC-II scans for MY 1996 and
late). Each of these test types became the Reference for the respective model year groups.
The specific combinations of MOBILE6.2 runs performed are shown in Table 5-1 below.
Each of these runs represents a particular test type and test standard design which was expressed
as a ratio to the standard reference tests. The first four runs represent the Non I/M reference and
the three Phoenix I/M references. A set of these runs were done for each calendar year 1990
through 2030, for cars, light trucks and heavy-duty gasoline vehicles and for pollutants HC, CO
and NOx.
30
-------
Table
5-1. MOBILE6.2 Runs Used to Populate the MOVES IMFactor Table
RUN#
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
Description
Non IM Base
IM240 Base (Biennial IM240/147)
OBD Base (Biennial OBD Test)
Basic Base (Loaded - Idle Test)
Biennial - IM240 - Phase-in Outpoints
Annual - IM240 - Phase-in Outpoints
Biennial - IM240 - Final Outpoints
Annual - IM240 - Final Outpoints
Biennial - ASM 2525/50 15 - Phase-in Outpoints
Annual - ASM 2525/50 15 - Phase-in Outpoints
Biennial - ASM 2525/50 15 - Final Outpoints
Annual - ASM 2525/50 15 - Final Outpoints
Biennial - ASM 2525 - Phase-in Outpoints
Annual - ASM 2525 - Phase-in Outpoints
Biennial - ASM 2525 - Final Outpoints
Annual - ASM 2525 - Final Outpoints
Biennial - ASM 50 15 - Phase-in Outpoints
Annual - ASM 50 15 - Phase-in Outpoints
Biennial - ASM 50 15 - Final Outpoints
Annual - ASM 50 15 - Final Outpoints
Annual - OBD -
Annual - LOADED/IDLE
Biennial - IDLE
Annual - IDLE
Biennial - 2500/IDLE
Annual - 2500/IDLE
Type
Non I/M Reference
I/M Reference
I/M Reference
I/M Reference
Target IM Design
Target IM Design
Target IM Design
Target IM Design
Target IM Design
Target IM Design
Target IM Design
Target IM Design
Target IM Design
Target IM Design
Target IM Design
Target IM Design
Target IM Design
Target IM Design
Target IM Design
Target IM Design
Target IM Design
Target IM Design
Target IM Design
Target IM Design
Target IM Design
Target IM Design
The MOBILE6.2 database output option was chosen for all runs. This step produced large
sets of results which were further stratified by facility-cycle / start process and age. This output
format necessitated additional processing of the facility rates into composite running and start
factors (in MOVES the IMFactor is a function of running and start processes).
In addition to the IMFactor, MOVES adjusts rates for particular programs by applying an
additional multiplicative "Compliance Factor" (IMCompliance). The IMFactor (R) represents
the theoretical effectiveness of a specific I/M program design, relative to the reference design, as
described above.
Values of the IMComplianceFactor (C) are specific to individual programs and represent its
overall operational effectiveness and efficiency, aside from the effectiveness inherent in its
design. Variables which impact the IMCompliance factor include waiver rates, compliance rates
and overall operational efficiency. Default IMComplianceF actors are provided in the MOVES
database, but alternate values may be entered by the user for specific analyses. The default
factors were taken from the 2005 EPA National Emission Inventory (NET) [10], and are based on
data submitted by individual states in their State Implementation Plan (SIP) processes. The vast
majority of the default IMCompliance factors are greater than 90 percent.
-------
5.4 Development of MOVES IM Compliance Inputs
The default I/M Compliance inputs are contained in the IMCoverage table in the MOVES
database. The structure of the table is:
Pollutant / Process
State / County
Year
Regulatory Class
Fuel Type (only gasoline fuels)
Beginning Model Year of Coverage
Ending Model Year of Coverage
InspectFreq
IMProgramID
I/M Test Type
I/M Test Standards
Ignore I/M toggle (user control variable)
Compliance Factor
The IMCoverage table structure shows that the EVI Compliance Factor is a function of
numerous variables that include geography, time, vehicle type / fuel / coverage factors, program
test frequency and specific I/M test / I/M test standards types. The Ignore I/M toggle is a user
feature that allows the user to completely disable the effects of I/M for one or more of the
parameter combinations.
For state SIPs, it is expected that the state will enter their own set of Compliance Factors
which reflect current and expected future program operation. The data in the default MOVES
table is likely out of date (i.e., 2005 NEI), and has not been cross referenced or updated with
recent state I/M program designs / changes.
The underlying data used to construct the default Compliance Factors were taken from
MOBILE6.2 input files used in the NMIM model to compute the National Emission Inventory of
2005. The following data files were extracted and processed into the various fields in
IMCoverage table.
32
-------
Table 5-2. I/M Coverage Table Data Sources
NMIM Data Source
MOBILE6 Compliance Rate
I/M Cutpoints
MOBILE6 Effectiveness Rate
Grace Period
Model Year Range
Test Type
Vehicle Type
MOBILE6 Waiver Rate
MOVES I/M Coverage Parameter
Used in the MOVES Compliance Rate
Calculation
Used to determine MOVES I/M Test Standards
Used in the MOVES Compliance Rate
Calculation
Used in MOVES to Determine Beginning
Model Year of Coverage
Used in MOVES to Determine Ending Model
Year of Coverage
Used to determine MOVES I/M Test Type
Used to determine MOVES Regulatory Class
input
Used in the MOVES Compliance Rate
Calculation
As seen in Table 5-2, MOBILE6.2 and MOVES do not have exactly compatible parameter
definitions. Extraction and processing of the MOBILE6.2 inputs for all of the individual states
was required. The MOBILE6 compliance rate, waiver rate and Effectiveness rate were used to
determine the MOVES Compliance Rate. The new MOVES Compliance Rate is a broader
concept that incorporates three separate MOBILE6.2 inputs. Equation 6 shows the relationship.
C = M6ComplianceRatexM6EffectivenessRatex(l-M6WaiverRate) 6
The MOBILE6.2 IM Cutpoints data were used only to determine level of stringency of a
state's IM240 program (if any). The MOBILE6.2 Test Type inputs provided a description of the
specific I/M tests performed by the state and test standards for the ASM and Basic I/M tests.
The MOBILE6.2 inputs of Grace Period and Model Year Range were used to determine the
MOVES Beginning and Ending model year data values for each I/M program. The MOBILE6.2
Vehicle type input was mapped to the MOVES regulatory class.
33
-------
6. References
1. E. Glover and D. Brzezinski, "Exhaust Emission Temperature Correction Factors for
MOBILE6: Adjustments for Engine Start and Running LA4 Emissions for Gasoline
Vehicles," EPA Report Number EPA420R 01 029 (M6.STE.004), April 2001.
(Available at: www. epa.gov/otaq/models/mobile6/m6tech.htm)
2. E. Glover and P. Carey, "Determination of Start Emissions as a Function of Mileage and
Soak Time for 1981-1993 Model Year Light-Duty Vehicles," EPA Report Number
EPA420-R-01-058 (M6.STE.003), November 2001.
(Available at: www. epa.gov/otaq/models/mobile6/m6tech.htm)
3. Stump, F. D., D. L. Dropkin, S. B. Tejada, C. Loomis, and C. Pack, "Characterization of
Emissions from Malfunctioning Vehicles Fueled with Oxygenated Gasoline-Ethanol (E-10)
Fuel - Part III," US EPA's National Exposure Research Laboratory (NERL), U. S.
Environmental Protection Agency, Washington, D.C., EPA Report Number EPA/600/R-
01/053 (NTIS PB2004-106735), July 2002.
(Available at: http://www.epa.gov/nerl/nerlmtbe.htm#mtbe7c)
4. "Analysis of Parti culate Matter Emissions from Light-Duty Gasoline Vehicles in Kansas
City" EPA Report Number EPA420-R-08-010, April 2008.
(Available at: http://epa.gov/otaq/emission-factors-research/420r08010.pdf)
5. "Regulatory Impact Analysis for Final Rule: Control of Hazardous Air Pollutants from
Mobile Sources" EPA Report Number EPA420-R-07-002, February 2007, Chapter 2, pages
2-15 to 2-17.
(Available at: http://www.epa.gov/otaq/regs/toxics/fr-ria-sections.htm)
6. Federal Register 86.1342-90 (Page 309).
(Available at: http://www.access.gpo.gov/nara/cfr/cfr-table-search.html)
7. "Air Conditioning Activity Effects in MOBILE6", EPA Report Number EPA420-R-01-054
(M6.ACE.001), November 2001.
(Available at: http://www.epa.gov/otaq/models/mobile6/m6tech.htm)
8. "Air Conditioning Correction Factors in MOBILE6", EPA Report Number EPA420-R-01-
055 (M6.ACE.002), November 2001.
(Available at: http://www.epa.gov/otaq/models/mobile6/m6tech.htm)
9. "MOBILE6 Inspection / Maintenance Benefit Methodology for 1981 through 1995 Model
Year Light Vehicles", USEPA Office of Transportation and Air Quality, Assessment and
Standards Division. EPA Report Number EPA420-R-02-014 (M6.IM.001) March 2002.
(Available at: http://www.epa.gov/otaq/models/mobile6/r02014.pdf)
34
-------
10. "Documentation for the 2005 Mobile National Emissions Inventory (NEI) 2005, Version
2", December 2008, and National Mobile Inventory Model (NMIM) County Database
(NCD) for 2005 V2.
(Available at: http://www.epa.gov/ttn/chief/net/2005inventory.html)
35
-------
Appendix A: Mean Start Emission by Temperature
Change in Mean Start Emissions at Various Temperatures
By Model Year Group
Relative to 75° F
Model Yr
Group
Pre-81
Pre-81
Pre-81
Pre-81
Pre-81
Pre-81
Pre-81
Pre-81
Pre-81
Pre-81
Pre-81
Temp
19.75
20.67
22.63
47.55
49.78
52.52
60.14
77.31
95.36
98.06
105.06
HC
(grams)
36.090
33.018
30.560
18.569
15.252
18.099
11.120
0
-2.122
-1.755
-4.935
CO
(grams)
226.941
254.386
276.341
129.472
120.931
115.776
53.617
0
-58.656
-67.555
-86.689
NOx
(grams)
-0.274
-0.925
-1.445
-0.380
-0.034
0.101
1.790
0
1.640
1.975
3.769
Model Yr
Group
81-82
81-82
81-82
81-82
81-82
81-82
81-82
81-82
81-82
81-82
81-82
Temp
19.36
20.69
22.33
49.20
50.31
51.43
59.15
75.73
95.22
97.75
105.00
HC
(grams)
21.120
23.363
25.496
7.782
8.202
9.209
6.432
0
-4.659
-5.450
-9.958
CO
(grams)
231.180
242.806
253.865
109.851
120.239
132.360
135.063
0
-144.116
-174.532
-343.847
NOx
(grams)
-0.374
-0.252
-0.135
-0.066
0.065
0.194
-1.416
0
1.915
1.814
4.568
36
-------
APPENDIX A Continued
Change in Mean Start Emissions at Various Temperatures
By Model Year Group
Relative to 75° F
Model Yr
Group
83-85
83-85
83-85
83-85
83-85
83-85
83-85
83-85
83-85
83-85
83-85
Temp
19.32
21.00
22.48
28.80
48.99
50.33
51.30
76.20
95.81
97.19
105.79
HC
(grams)
23.299
17.755
14.599
20.594
5.213
5.946
6.490
0
-1.044
-1.209
-1.124
CO
(grams)
218.857
218.151
216.439
186.549
94.414
93.032
95.495
0
-29.275
-35.995
-25.407
NOx
(grams)
0.665
-0.017
-0.414
-0.126
0.513
0.250
0.183
0
0.903
0.868
-1.010
Model Yr
Group
86-89
86-89
86-89
86-89
86-89
86-89
86-89
86-89
Temp
-20
0
20
40
75
95.03
96.43
106.29
HC
(grams)
27.252
25.087
14.011
8.316
0
-0.127
-0.139
-0.729
CO
(grams)
178.536
147.714
104.604
78.525
0
-4.257
-5.354
-1.017
NOx
(grams)
-2.558
-1.360
-0.749
0.312
0
-0.137
-0.091
-0.084
Model Yr
Group
1990-2005
1990-2005
1990-2005
1990-2005
1990-2005
Temp
-20
0
20
40
75
HC
(grams)
38.164
16.540
8.154
4.872
0
CO
(grams)
143.260
92.926
56.641
33.913
0
NOx
(grams)
1.201
1.227
1.082
0.876
0
37
-------
Appendix B - Calculation of Specific Humidity
Equations to convert from relative humidity in percent to specific humidity (or humidity
ratio) in units of grains of water per pound of dry air (ref CFR section 86.344-79, humidity
calculations).
Inputs:
TF is the temperature in degrees F.
Pb is the barometric pressure.
Hrei is the relative humidity
T0=647.27-TK
mtio or specifichumidity
idit ~ 4347. 0 Py l(Pb Py )
JA / 1 Y
JA
^=29.92*218.167*10
=6527.557*10
(-TO/TK)
t
l+0.00219r
2437+0.00588ro + 0.0000000117T0:
l+0.00219rn
38
------- |