MOVES2010 Highway Vehicle Temperature, Humidity, Air Conditioning, and Inspection and Maintenance Adjustments &EPA United States Environmental Protection Agency ------- MOVES2010 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 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. &EPA United States Environmental Protection Agency EPA-420-R-10-027 December 2010 ------- Table of Contents Table of Contents i List of Tables ii List of Figures iii Glossary of Acronyms iv 1. Introduction 1 2. Temperature Adjustments 1 2.1 Data Sources for Temperature Effects 1 2.2 Temperature Adjustment Methodology 2 2.3 Effects of Temperature on Gasoline Fueled Vehicles 3 2.3.1 Temperature Effects on Gasoline Start Emissions 3 2.3.1.1 HC and CO Start Emissions for Gasoline-Fueled Vehicles: 3 2.3.1.2 Temperature Effects on Gasoline NOx Start Emissions 5 2.3.1.3 Temperature Effects on Gasoline PM Start Emissions 7 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 12 2.5.1 Cold Weather CO Requirement 12 2.5.2 Cold Weather HC Requirement 12 2.5.3 Cold Weather PM Effects 14 3. Humidity Adjustments 17 4. Air Conditioning Adjustments 17 4.1 Air Conditioning Effects Data 18 4.2 Mapping Data to VSP Bins 20 4.3 Air Conditioning Effects on Emissions 22 4.3.1 A/C Adjustments for HC, CO and NOx Emissions 22 4.3.2 Full A/C Adjustments for Energy Consumption 22 4.3.3 Uncertainty Analysis 23 4.4 Adjustments to Air Conditioning Effects 25 5. Inspection and Maintenance Programs 28 5.1 Inspection & Maintenance in MOBILE6 28 5.2 Inspection & Maintenance in MOVES 29 5.3 Devel opment of MO VES I/M Factors 30 5.4 Development of MOVES I/M Compliance Inputs 32 6. References 34 Appendix A-Mean Start Emission by Temperature 36 Appendix B - Calculation of Specific Humidity 38 Appendix C - Air Conditioning Analysis Vehicle Sample 39 Appendix D - Toros Topaloglu, Comments 41 Appendix E - ENVIRON International Corporation, Comments 48 Appendix F - Julio Vassallo, Comments 52 Appendix G - Coordinating Research Council Proj ect E-68a Comments 54 ------- List of Tables Table 2-1 Average NOx Emission Results by Temperature 7 Table 2-2 Diesel Vehicle Emissions by Temperature 9 Table 2-3 Phase-In of Vehicles Meeting Cold Weather HC Standard 13 Table 2-4 Multiplicative Increases of PM at 20° Fahrenheit 16 Table 2-5 Exponential Equation Constant Terms 16 Table 3-1 Humidity Correction Coefficients Used by MOVES 17 Table 4-1 Distribution of Test Vehicles by Model Year 19 Table 4-2 Distribution of Tests by Schedule Type 19 Table 4-3 VSP Bin Definitions 21 Table 4-4 Full Air Conditioning Adjustments for HC, CO and NOx 22 Table 4-5 Full Air Conditioning Adjustments for Energy 23 Table 4-6 Fraction of Vehicles Equipped with Air Conditioning 26 Table 4-7 Fraction of Air Conditioning Units Still Functioning By Age 27 Table 4-8 Effect of Heat Index on Air Conditioning Activity 27 Table 5-1 MOBILE6.2 Runs Used to Populate the MOVES IMF actor 31 Table 5-2 I/M Coverage Table Data Sources 33 Table A-l Change in Mean Start Emissions at Various Temperatures 36 Table A-2 Change in Mean Start Emissions at Various Temperatures 37 Table C-l Vehicle Sample for the Air Conditioning Analysis 39 11 ------- List of Figures Figure 2-1 Effects of Ambient Temperature on Changes in Cold-Start NOx 6 Figure 2-2 Logarithm of Bag-2 HC Versus Temperature 8 Figure 2-3 Cold-Start HC Emissions (in grams) with Confidence Interval 10 Figure 2-4 Bag-1 minus Bag-3 CO Emissions (in grams) with Confidence Interval 11 Figure 2-5 Bag-1 minus Bag-3 NOx Emissions (grams) with Confidence Intervals 11 Figure 2-6 FTP Bag 1 PM and FTP Bag 1 NMHC for Tier 2 Vehicles 15 in ------- Glossary of Acronyms A/C ACCF ASM CO EPA F FID FTP g/mi GVWR HC HLDT I/M LOT LDV LLDT MDPV MOVES MSAT NMHC NOx OBD PM RSD SFTP THC VIN voc Air Conditioning Air Conditioning Correction Factor Acceleration Simulation Mode Carbon Monoxide Environmental Protection Agency Fahrenheit Flame lonization Detection Federal Test Procedure Grams per Mile Gross Vehicle Weight Rating HydroCarbons Heavy Light Duty Truck Inspection and Maintenance Light Duty Truck Light Duty Vehicle Light Light Duty Truck Medium Duty Passenger Vehicle MOtor Vehicle Emission Simulator Mobile Source Air Toxics Non-Methane HydroCarbons Oxides of Nitrogen On-Board Diagnostic Paniculate Matter Remote Sensing Data Supplemental Federal Test Procedure Total HydroCarbons (FID detection) Vehicle Identification Number Volatile Organic Compounds IV ------- 1. Introduction The emission rates in the MOVES model database represent a single (base) scenario of conditions for 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 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. This report describes adjustments that affect running exhaust, start exhaust and extended idling emissions. The crankcase emission processes are chained to running exhaust, engine start and extended idling emissions, and thus are similarly affected by the temperature adjustments describe in this report. The impact of fuels, temperatures and I/M programs on vapor venting, permeation and liquid leaks is addressed in a separate report on evaporative emissions. 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 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 adjustments. 2.1 Data Sources for Temperature Effects For this analysis, we used almost entirely "Bagged" tests. Those data sets 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 also 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 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. Information on this study is available at: http://www.epa.gov/otaq/emission-factors-research/420r08009.pdf 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 of gasoline vehicles, 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. For this analysis, we started with the model year groups used in MOVES for start emission rates. By combining several finer model year 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 vehicles meeting the Tier 0, Tier 1, and LEV standards all exhibited similar increases in emissions as the ambient temperature drops from 75° F to 20° F. A single additive adjustment (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 (cold ------- engine and emission control equipment) and the third mode begins with a hot-start (relatively warm engine and control equipment).. 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, IM240, 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 was the model year grouping. 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 of temperature. For simplicity, we only considered functions that were a multiple of "temperature minus 75° F" raised the first, second, or third degree. This produced (additive) adjustments that exhibit zero change at 75 degrees Fahrenheit. 2.3 Effects of Temperature on Gasoline Fueled Vehicles 2.3.1 Temperature Effects on Gasoline Start Emissions The effects of ambient temperature on HC, CO, and NOx start emissions were modeled using the following algorithm: 1. Using additive (rather than multiplicative) adjustments. 2. Calculating these adjustments as simple functions of one of the following measures: — 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 at 75° F (i.e., the nominal temperature of EPA's FTP test). Those coefficients are stored in the MOVES database table named StartTempAdjustment. 3. Setting the value of the adjustments equal to zero for temperatures higher than 75C Fahrenheit. 2.3.1.1 HC and CO Start Emissions for Gasoline-Fueled Vehicles: As described above we used the difference between the Bag-1 emissions and 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, we decided to model the changes in cold-start emissions as a polynomial (linear, or a quadratic, or a cubic) function 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, we set the value of those additive adjustments equal to zero for temperatures higher than 75° F. Thus, 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) Forthel981-1982s: 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 Forthel986-1989s: 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 Note: 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) where: tempAdjustTermA =-1.141434345 R-sqr = 0.99017 Note: 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 Start 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 2-1), 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 t d in d L. 01 W X 0 Z t d +rf n 0 01 01 d 0 6.0 3.0 0.0 -3.0 -20 20 40 60 80 100 Temperature (degrees Fahrenheit) 120 A visual inspection of Figure 2-1 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 analysis results were the ones in which we average together all of the NOx results (from Appendix A) to obtain the following table: ------- Table 2-1 Average NOx Emission Results by Temperature Delta NOx Temp (grams) -20.0 1.201 0.0 1.227 1974 0202 20.7 0.089 2274 iQ;i55 Delta NOx Temp (grams) 31.0 -0.007 40.0 0.876 4878 0127 49.8 0.333 5lTO 0325 Delta NOx Temp (grams) 54.2 0.438 76.3 0.000 9573 0225 97.1 0.370 10578 0543 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 adjustments), this additive adjustment is set to zero for temperatures higher than 75° F. 2.3.1.3 Temperature Effects on Gasoline PM Start Emissions The temperature effects on particulate matter (PM) start emissions modeled in MOVES using a multiplicative (not additive) exponential (not polynomial) adjustment. Thus analysis is included as Chapters 7 and 8 of a separate report ("Analysis of Particulate 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) ro _a O I t 3 - 2 1 0 - 1C 1 ^ -4 - -5 - R « .0 2C * * ^ + .0 3{ * * 3 ^ . * ^ » 4 < ^ • » «? A .0 ' *4C * ^ « * » * <• * *+ ** <» »* »*«» * *« * t » » , * ^ • » ^ »* «* »»* . t t » V *' ^g *»^>5C ^ *v^*» < :^'* > ** * »• * *» » * ^ • V « » t * « * • «e,»«* 6C •i>* * ** »* » * ** •* « , »*« * * * * * » * » ^ * » *0 *^7C ^ * (**** f* »* x V *•*»»*« »#.*« • » < * $« t * «» « • »• i* »>«? » ** * « » ^^^^ ^^. *^ ***^> ^*± * < **V**^ * 5 »»*•» • *» * ^ > » » ^ ,* ^ »0 *»» ,9C • ? ••• h * * * • »•* » » * * « * , .« 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 IM240 tests run in Chicago (as part of Chicago's I/M program) to validate this conclusion. To avoid variable preconditioning, we used only second IM240s when back-to-back IM240s were performed, and for the other EVI240s we examined the last 120 seconds of full duration EVI240s. We found no evidence of a temperature effect between 5 and 95 degrees F. The effect of temperature on hot running HC, CO, and NOx emissions is coded in MOVES using polynomial functions as multiplicative adjustments. In this version of MOVES, we propose to set all of those adjustments equal to 1.0, that is, no change in those running emissions with temperature. This was not the case for PM emissions. Our data did show a temperature effect for running emissions of particulate matter. As for start emissions, the temperature effects on particulate matter (PM) running emissions modeled in MOVES using a multiplicative (not additive) ------- exponential (not polynomial) adjustment. This analysis is detailed in as Chapters 7 and 8 of the "Analysis of Particulate Matter Emissions from Light-Duty Gasoline Vehicles in Kansas City"). [4]. 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). None of these diesel trucks were equipped with after-treatment devices. The Bag-1 minus Bag-3 emissions for those tests are shown in Table 2-2. We stratified the test results into four temperature bands which yielded the following emission values (grams per start) and average temperature value: Table 2-2 Diesel Vehicle Emissions by Temperature (grams per start) Temperature 34.6 43.4 61.5 69.2 Count 6 7 10 2 HC 2.55 2.68 1.69 1.2 CO 2.44 2.03 3 1.91 NOx 2.6 0.32 0.67 0.36 When we plotted the mean HC start emissions (above) versus temperature, we obtained the following graph (where the vertical lines represent 90 percent confidence intervals and the "dashed" line represents a linear regression through the data). ------- 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 adjustment 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 estimate a diesel light-duty vehicle temperature effect for CO and NOx. 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: 10 ------- Figure 2-4 Bag-1 minus Bag-3 CO Emissions (in grams) with Confidence Interval i 30 40 50 60 Temperature (degrees F) 70 Figure 2-5 Bag-1 minus Bag-3 NOx Emissions (grams) with Confidence Intervals i I 30 40 50 60 Temperature (degrees F) 70 11 ------- 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 set the temperature adjustment for start emissions to zero: CO temperature Adjustment = tempAdjustTermA * (Temp. - 75) where: tempAdjustTermA = 0.0 NOx temperature Adjustment = tempAdjustTermA * (Temp. - 75) where: tempAdjustTermA = 0.0 Since gasoline adjustments were set to zero, the temperature effects for diesel running exhaust were also set to zero. Because temperature effects data was not available for heavy duty trucks, the light duty results were extrapolated to these vehicles including thethe extended idling emission process for heavy duty long haul diesel trucks. No attempt has been made to adjust the effects of temperature on emissions to account for the introduction of after-treatment devices (such as diesel particulate filters or oxidation catalysts) that will become more common on future diesel fueled vehicles. 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. 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 TheMobile 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. 12 ------- • 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: Table 2-3 Phase-In of Vehicles Meeting Cold Weather HC Standard Year 2010 2011 2012 2013 2014 2015 / 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 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 could increase 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 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 represent the increases in the generic start emissions. Using the ratio of hot-start 13 ------- to cold-start from our earlier analysis, the increase in NMHC cold-start emissions (as the ambient temperature drops from 75° F down to 20° F) are: • 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), 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 paniculate 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). 14 ------- Figure 2-6 FTP Bag 1 PM and FTP Bag 1 NMHC for Tier 2 Vehicles CM _ CD DQ CO . 8 % V V O A X -3 I -2 i -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 a proportional 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. 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 affects both start and running PM emissions, and that effect (for Tier 2 vehicles) is best modeled by (exponential) multiplicative adjustments of the form: A*(72-t) Multiplicative factor = Q , 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.) 15 ------- Therefore, for Tier 2 vehicles not affected by the MS AT-2 requirements, as the temperature decreases from 72° to 20° F, EPA expects PM emissions to increase by factors of: • 11.10727 for cold-starts and • 5.22576 for hot running. Applying the 30 percent reduction for vehicles affected by the MS AT-2 requirements implies a PM increase as the temperature decreases from 72° to 20° F of: • 7.77509 for cold-starts and • 3.65803 for hot running. Combining this information with the MSAT-2 phase-in schedule from Table 2-3 leads to the following (multiplicative) increases as the temperature decreases from 72° to 20° F: Table 2-4 Multiplicative Increase in PM from 72° to 20° Fahrenheit Model Year 2008 2009 2010 2011 2012 2013 2014 2015 LDVs / LLDTs Start 11.10727 11.10727 10.27423 9.44118 8.60814 7.77509 7.77509 7.77509 Running 5.22576 5.22576 4.83383 4.44189 4.04996 3.65803 3.65803 3.65803 HLDTs/MDPVs Start 11.10727 11.10727 11.10727 11.10727 10.27423 9.44118 8.60814 7.77509 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: Table 2-5 Exponential Equation Constant Terms Model Year 2008 2009 2010 2011 2012 2013 2014 2015 LDVs / LLDTs Cold-Start 0.046300 0.046300 0.044801 0.043175 0.041398 0.039441 0.039441 0.039441 Running 0.031800 0.031800 0.030301 0.028675 0.026898 0.024941 0.024941 0.024941 HLDTs/MDPVs Cold-Start 0.046300 0.046300 0.046300 0.046300 0.044801 0.043175 0.041398 0.039441 Running 0.031800 0.031800 0.031800 0.031800 0.030301 0.028675 0.026898 0.024941 16 ------- We assume that the increases in the PM2.5 emissions apply proportionally to the Elemental Carbon (EC) and Organic Carbon (OC) portions of the PM2.5 emissions. Sulfate PM emissions are not affected by temperature. Although the ARB factors that adjust the start emissions based on soak time were not developed for PM emissions from gasoline-fuel vehicles, the finding that Tier 2 PM emissions are proportional to HC emissions supports our decision to apply the HC soak adjustments 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 adjustment. 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 adjustment. Appendix B shows how the hourly relative humidity values are converted to specific humidity used in this equation using temperature and barometric pressure. Table 3-1 Humidity Correction Coefficients Used by MOVES Fuel Type Gasoline Diesel Fuel Humidity Correction Coefficient 0.0038 0.0026 The diesel humidity correction coefficient is taken directly from the Code of Federal Regulations[6]. The gasoline humidity correction coefficient is carried over from the coefficient used in the MOBILE6 model. 4. Air Conditioning Adj ustments The air conditioning adjustments in MOVES are based on the same data as was used in the previous MOBILE6 model, but the adjustments themselves were recalculated to be consistent with the MOVES methodology. The proposed factors are based on a test procedure meant to simulate air conditioning emission response under extreme "real world" ambient conditions. These factors predict emissions which would occur during full loading of the air conditioning system, and will then be 17 ------- scaled down in MOVES according to ambient conditions in a modeling run. The second-by- second emission data 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. 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. 4.1 Air Conditioning Effects Data The data for the MOVES A/C Correction Factor (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 distribution of test vehicles by model year is shown in Table 4-1. Model years 1990 through 1999 were included. The data set consists of 30 cars and 24 light trucks. No test data were available on other vehicle types (i.e., motorcycles, 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 C. 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 VEST, 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 O 1 54 Table 4-2 Distribution of Tests by Test Cycle 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 Mapping Data to VSP Bins 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. 20 ------- After computing the VSP for each vehicle test / second combination, we assigned the individual secondsto the MOVES VSP bins. These VSP bins are defined in Table 4-3. VSP bins 26 and 36 were not defined because bins 27-30 and bins 37-40 overlap them. Table 4-3 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 Crui se/ Accel erati on; Crui se/ Accel erati on; Crui se/ Accel erati on; Crui se/ Accel erati on; Crui se/ Accel erati on; ; VSP< 0; K=Speed<25 0<=VSP< 3; 1<= Speed<25 3<=VSP< 6; K=Speed<25 6<=VSP< 9; K=Speed<25 9<=VSP<12; K=Speed<25 12<=VSP; K=Speed<25 Moderate Speed Coasting; VSP< 0; 25<=Speed<50 Crui se/ Accel erati on; Crui se/ Accel erati on; Crui se/ Accel erati on; Crui se/ Accel erati on; Crui se/ Accel erati on; Crui se/ Accel erati on; Crui se/ Accel erati on; Crui se/ Accel erati on; Crui se/ Accel erati on; Crui se/ Accel erati on; Crui se/ Accel erati on; Crui se/ Accel erati on; Crui se/ Accel erati on; Crui se/ Accel erati on; Crui se/ Accel erati on; Crui se/ Accel erati on; 0<=VSP< 3; 25<=Speed<50 3<=VSP< 6; 25<=Speed<50 6<=VSP< 9; 25<=Speed<50 9<=VSP<12; 25<=Speed<50 12<=VSP; 25<=Speed<50 12<=VSP<18; 25<=Speed<50 18<=VSP<24; 25<=Speed<50 24<=VSP<30; 25<=Speed<50 30<=VSP; 25<=Speed<50 VSP< 6; 50<=Speed 6<=VSP<12; 50<=Speed 12 <= VSP; 50<=Speed 12<=VSP<18; 50<=Speed 18<=VSP<24; 50<=Speed 24<=VSP<30; 50<=Speed 30<=VSP; 50<=Speed 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 4-3) 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. These large bins are quite different in terms of engine operation and emissions performance. The Braking bin consisted of VSP Bin 0 in Table 4-3, the Idle bin was VSP Bin 1 and the Cruise / Acceleration bin contained the remaining 21 bins. 21 ------- 4.3 Air Conditioning Effects on Emissions 4.3.1 A/C Adjustments for HC, CO and NOx Emissions Full A/C adjustments 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 adjustments are shown in Table 4-4. Measures of statistical uncertainty (coefficient of variation of the mean) were also computed using the standard error of the mean. They are shown in Table 4-4as"MeanCVofCF." Table 4-4 Full Air Conditioning Adjustments 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 4.3.2 Full A/C Adjustments for Energy Consumption 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). 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 adjustments for energy are a function of only VSPBin. The resulting A/C adjustments are shown in Table 4-5. 22 ------- Table 4-5 Full Air Conditioning Adjustments 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 the 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 and 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 as indicated by the coefficient of variation of the mean (mean CV) were calculated using the following formula. The same set of equations were used for each of the three pollutants (although the equations are shown only once). The values of X and Y represent second by second emissions from either HC, CO or NOx. The variable "X" represents emissions with the A/C On and "Y" represents emission with the A/C Off. Given: Mean CV X /Y 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 23 ------- Vz2 = (8Z/8X)2 * Vx2 + (8Z/8Y)2 * Vy2 Where Vz is the variance of Z, Vx is the variance of X and Vy is the variance of Y 8Z/8X is the partial derivative of Z with respect to X 8Z/8Y 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 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 4-3). 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 SA2x = (II (nVeh-1) ) * Sumlx SB2x = ( 1 / (nCell - nVeh) ) * Sum2x SA2y = (II (nVeh-1) ) * Sumly SB2y = ( 1 / (nCell - nVeh) ) * Sum2y 24 ------- And Sumlx = E (YbarVehx - YbarCellx )2 Sum2x = E ( varVehx - ( nMeas - 1) )2 Sumly = E ( YbarVehy - YbarCelly )2 Sum2y = E ( varVehy - ( nMeas - 1) )2 Where The sums (E ) 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. For HC, CO and NOx, detailed VSP was not found to be an important variable in regards to A/C adjustment and A/C usage. However, Full A/C adjustments greater than one were found for all pollutants for both Idle and Cruise / Acceleration modes. For NOx Idle mode, a fairly large multiplicative adjustment 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 adjustment of (2.1123) for CO cruise / Accel was also obtained. These adjustments will double CO emissions under extreme conditions of A/C usage. A/C adjustments of less than or equal to one were found for the Braking / Deceleration mode for all three pollutants. These were set to one for use in the MOVES model. 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 ) 25 ------- 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]. Table 4-6 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. ** 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. 26 ------- Table 4-7 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 + heat!ndex*(ACActivityTermB + ACActivityTermC*heatIndex) Table 4-8 Effect of Heat Index on Air Conditioning Activity -3.63154 0.072465 -0.000276 ACActivityTermA ACActivityTermB ACActivityTermC Heat Index 67.44 70 75 80 85 90 95 100 105 110 AC On Fraction 0.000 0.089 0.251 0.399 0.534 0.655 0.762 0.855 0.934 1.000 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 ) 27 ------- The air conditioning adjustment is a multiplicative adjustment applied to the emission rate after it has been adjusted for fuel effects. Air conditioners are employed for defogging at all temperatures, particularly, at lower temperatures. This secondary use of the A/C along with associated emission effects is not addressed in MOVES2010. 5. Inspection and Maintenance Programs Inspection and Maintenance (I/M) programs are genetically 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. Readers interested in a more thorough treatment of the topic are 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 affected 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 rates being lower. The percentage difference between these two rates is often referred to as the overall I/M reduction or I/M benefit. 28 ------- 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 I/M 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 thel/M benefit of the Phoenix program design assuming perfect compliance. Equation 1 shows this relationship in a mathematical form. Standard I/M difference = EnonM - EM Eq 1 where Enon_M and EM 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 for MOVES came from this source. The selection does not imply any judgment on 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 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, ^p = REjM + (1 ~ R)EnonlM EC1 2 where Ep is the adjusted emission rate for a "target" I/M program, EM is the reference rate, EnonM is the non-I/M reference rate, and R is an aggregate adjustment 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 EnonM, fall between EnonM and EM, or be less than EM- Thus this framework can 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 forR gives leads to Equation 3. This equation shows the I/M adjustment as the ratio of the emission difference between a proposed I/M program design and the Standard I/M Difference E -EnonIM R = — —— Eq 3 29 ------- 5.3 Development of MOVES I/M Factors Early in the MOVES development process it was decided that developing the I/M adjustment factors based on a 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 old model years at young ages (i.e., a 1985 model year at age five). As a result, EPA decided to develop I/M adjustment factors based on the information incorporated 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 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 IMF actors) • Model Year Group • Age Group • IMF actor The IMF actor value 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 IMF actor is the ratio of the mean emission results from these two runs. Equation 4 illustrates the simple formula. 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 (EVI240, EVI147), and OBD-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 30 ------- 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. Table 5-1 MOBILE6.2 Runs Used to Populate the MOVES I/M Adjustment Factor 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 I/M Base IM240 Base (Biennial IM240/147) OBDBase (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 2525/5015 - 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 I/M Design Target I/M Design Target I/M Design Target I/M Design Target I/M Design Target I/M Design Target I/M Design Target I/M Design Target I/M Design Target I/M Design Target I/M Design Target I/M Design Target I/M Design Target I/M Design Target I/M Design Target I/M Design Target I/M Design Target I/M Design Target I/M Design Target I/M Design Target I/M Design Target I/M 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 IMF actor is a function of running and start processes). In addition to the IMF actor, MOVES adjusts rates for particular programs by applying an additional multiplicative "Compliance Factor" (IMCompliance). The IMF actor (R) represents the theoretical effectiveness of a specific I/M program design, relative to the reference design, as described above. Values of the EVIComplianceFactor (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 EVIComplianceF 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 (NEI) [10], and are based on 31 ------- data submitted by individual states in their State Implementation Plan (SIP) processes. The vast majority of the default EVICompliance factors are greater than 90 percent. 5.4 Development of MOVES I/M Compliance Inputs The default I/M Compliance inputs are contained in the EVICoverage table in the MOVES database. The structure of the table is: Pollutant / Process State / County Year Source Use Type Fuel Type (only gasoline fuels) Beginning Model Year of Coverage Ending Model Year of Coverage InspectFreq EVCProgramlD I/M Test Type I/M Test Standards Ignore I/M toggle (user control variable) Compliance Factor The EVICoverage 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. 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 NMEVI model to compute the National Emission Inventory of 2005. The data files listed in Table 5-2 were extracted and processed into the various fields in EVICoverage table. 32 ------- Table 5-2 I/M Coverage Table Data Sources NMIM Data Source MOBILE6 Compliance Rate I/M Outpoints 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 5 shows the relationship. C = M6 Compliance Rate x M6 Effectiveness Rate x (1 - M6 Waiver Rate) Eq 5 MOVES does not have separate inputs for the effect of waivers on I/M benefits. Section 3.10.6.2 of the document, "Technical Guidance on the Use of MOVES2010 for Emission Inventory Preparation in State Implementation Plans and Transportation Conformity" describes how to calculate the MOVES compliance rate to include the effect of waivers. In MOVES, it is assumed that any repairs attempted on vehicles receiving waivers are not effective and do not result in any reduced emissions. Other fields in the EVICoverage table complete the description of the I/M program in effect in each county. The MOBILE6.2 I/M Cutpoints data were used only to determine level of stringency of a state's EVI240 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. 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. 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 EPA420 R 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 HI," 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 Particulate 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. Code of Federal Regulations 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.EVI.001) March 2002. (Available at: http://www.epa.gov/otaq/models/mobile6/r02014.pdf) 34 ------- 10. "Documentation for the 2005 Mobile National Emissions Inventory (NET) 2005, Version 2", December 2008, and National Mobile Inventory Model (NMEVI) 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 Table A-l 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 2Z63 47.55 4978 52.52 6014 77.31 95736 98.06 105706 HC (grams) 36.090 33.018 307560 18.569 157252 18.099 11120 0 -27122 -1.755 -4935 CO (grams) 226.941 254.386 2767341 129.472 126:931 115.776 531517 0 -58:656 -67.555 -86T689 NOx (grams) -0.274 -0.925 -T:445 -0.380 -0:634 0.101 1796 0 1640 1.975 3769 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 22733 49.20 5031 51.43 59715 75.73 95722 97.75 105766 HC (grams) 21.120 23.363 257496 7.782 8:202 9.209 67432" 0 -47659 -5.450 -97958 CO (grams) 231.180 242.806 2537865 109.851 126:239 132.360 1357663 0 -1447116 -174.532 -343:847 NOx (grams) -0.374 -0.252 -0135" -0.066 0065 0.194 -17416 0 1915 1.814 4568 36 ------- APPENDIX A Continued Table A-2 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 2Z48 28.80 48799 50.33 51730 76.20 95:81 97.19 10579 HC (grams) 23.299 17.755 14599 20.594 57213 5.946 67490 0 -1044 -1.209 -1124 CO (grams) 218.857 218.151 2167439 186.549 94414 93.032 95:495 0 -29:275 -35.995 -25T407 NOx (grams) 0.665 -0.017 -07414 -0.126 0513 0.250 0183 0 0903 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 96743 106.29 HC (grams) 27.252 25.087 14011 8.316 0 -0.127 -0139 -0.729 CO (grams) 178.536 147.714 104604 78.525 o -4.257 -5:354 -1.017 NOx (grams) -2.558 -1.360 -0749 0.312 o -0.137 -b769i -0.084 Model Yr Group 1990-2005 1990-2005 _______ 1990-2005 _______ Temp -20 0 20 40 75 HC (grams) 38.164 16.540 87154 4.872 0 CO (grams) 143.260 92.926 567641 33.913 o NOx (grams) 1.201 1.227 1082 0.876 o 37 ------- Appendix B - Calculation of Specific Humidity Equations to convert 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 TQ=647.27-TK " ratio or seciichumidit ~ 4 3 4 7.0 Py I (/j, Fy pv=\ \ 100 J db , (3.2437+0.00588ro+0.0000000117T03) ^=29.92*218.167*10 , =6527.557*10 (3.2437+0.00588ro+0.0000000117ro3) 38 ------- Appendix C -Air Conditioning Analysis Vehicle Sample Table C-l Vehicle Sample for the Air Conditioning Analysis Model Year 1990 1990 1991 1991 1992 1992 1992 1992 1992 1993 1993 1993 1993 1994 1994 1994 1994 1995 1995 1995 1995 1995 1996 1996 1996 1996 1996 1997 1998 1998 1990 1990 1991 1991 1992 1993 1994 1994 1996 1996 1990 Make DODGE NISSAN CHEVROLET FORD CHEVROLET CHEVROLET MAZDA SATURN TOYOTA CHEVROLET EAGLE HONDA TOYOTA CHRYSLER FORD HYUNDAI SATURN BUICK BUICK FORD SATURN SATURN CHEVROLET HONDA HONDA PONTIAC TOYOTA FORD MERCURY TOYOTA JEEP PLYMOUTH CHEVROLET PLYMOUTH CHEVROLET CHEVROLET CHEVROLET PONTIAC FORD FORD CHEVROLET Model DYNA MAXIO CAVAO ESCO GT CAVA LUMI PROT SL CORO CORS SUMMO ACCOO CAMRO LHS ESCO ELAN SL CENT REGA LIMI ESCO SL SL LUMIO ACCO CIVI GRAN PRIX CAMR TAUR GRAN MARQ CAMR LE CHER VOYA ASTRO VOYA LUMI S10 ASTR TRAN EXPL RANG SURB Vehicle Class CAR CAR CAR CAR CAR CAR CAR CAR CAR CAR CAR CAR CAR CAR CAR CAR CAR CAR CAR CAR CAR CAR CAR CAR CAR CAR CAR CAR CAR CAR LDT1 LDT1 LDT1 LDT1 LDT1 LDT1 LDT1 LDT1 LDT1 LDT1 LDT2 Weight 3625 3375 2750 2625 3000 3375 2750 2625 2500 3000 2500 3250 3250 3750 2875 3000 2750 3995 3658 2849 2610 2581 3625 3500 2750 3625 3625 3650 4250 3628 3750 3375 4250 3750 3875 2875 4750 4250 4500 3750 5250 39 ------- Model Year 1991 1994 1996 1996 1996 1996 1996 1996 1997 1997 1997 1998 1999 Make FORD FORD FORD DODGE DODGE DODGE DODGE FORD DODGE DODGE PONTIAC DODGE FORD Model E1500 F150 F150 DAKO PICK D250 RAM GRAN CARA CARA F150PICK GRAN CARA DAKOT TRANSSPOR CARA GRAN WIND Vehicle Class LDT2 LDT2 LDT2 TRUCK TRUCK TRUCK TRUCK TRUCK TRUCK TRUCK TRUCK TRUCK TRUCK Weight 4000 4500 4500 4339 4715 4199 4102 4473 4318 4382 4175 4303 4500 40 ------- Appendix D - Toros Topaloglu, Comments Peer Review of US EPA's 'Draft MOVES2009 Highway Vehicle Temperature, Humidity, Air Conditioning, and Inspection & Maintenance Adjustments" September 29, 2009 As part ofthe MOVES2010 Peer Review process, EPA solicited comments from Toros Topaloglu, Ph.D., P.Eng. on the August 2009 draft of report Draft MOVES2009 Highway Vehicle Temperature, Humidity, Air Conditioning, and Inspection & Maintenance Adjustments. Dr. Topaloglu is an Environmental Systems Specialist at the Ministry of Transportation Ontario, Canada. Dr. Topaloglu's comments are copied below, with EPA response in italics. 1. Introduction The development of MOVES and its predecessor, MOBILE, represent enormous achievements: estimating past, present and future emissions of an infinitely diverse and variable vehicle/driver population under highly variable and ever changing conditions. The US EPA deserves our sincere gratitude for this unparalleled effort, which continues to deliver ever more powerful and user-friendly emission simulators. It has not been easy to think of a few meaningful comments on the above captioned report that describes various adjustments employed in MOVES. I have limited myself to "constructive criticism", assuming that this is what you expect from me and that this will be viewed in a positive vein coming from someone who has a direct and genuine interest in making MOVES as useful as possible. Where I am silent, I fully concur with the adopted approach and its presentation. This happens to be the case for over 99% of the report. In this review, I relied primarily on my personal experience and knowledge but consulted also the relevant MOBILE6 documentation and a few specific publications listed under Section 5 (references). 2. General Comments 2.1. I agree with the empirical/statistical approach adopted in the derivation of the adjustments - given the imprecise nature of cars and the near infinite variability in their population. Some scientist and engineers may feel more comfortable with relations that have a theoretical basis; however, even with the "best" data "the multitude of 41 ------- mechanisms involved in each adjustment make a mechanistic approach very difficult to implement. 2.2. Adjustments for greenhouse gas emission factors may not have been uniformly addressed. The vehicle emissions certification process does not automatically yield adjustments for CC>2, CH4 and N2O emissions. Given the urgency to address Climate Change, MOVES will be called upon frequently to derive more accurate GHG emission factors. 2.3. Adjustments for individual air toxic emission factors may not have been fully addressed. It is conceivable that adjustments for NMHC may not apply equally to each and every air toxic, since they are not formed by identical physical and chemical mechanisms. 2.4. It is not clear that the US EPA adjustments deal fully with up-and-coming technologies such as hybrid, plug-in hybrid and battery-powered electric vehicles. Emissions of these vehicles, where they exist, are less sensitive to variations in ambient conditions, air- conditioning (A/C), and inspection and maintenance (I/M). In fact, they are generally exempt from I/M. The number of electric hybrids in the US fleet exceeds one million already and is expanding rapidly. Hence, it will become progressively more important to account for their characteristics. 2.5. In future updates of MOVES, it may be worthwhile to try and correlate adjustments with major vehicle technologies and fuel types - beyond what is in place. This may improve the ability of the model to simulate future emissions. Vehicle manufacturers often have this information and might share it with the US EPA. 2.6. The accuracy of the adjustments depends, in part, on the representativeness of the test vehicle sample. It is obvious that the US EPA has spent enormous effort to achieve a high degree of representativeness. However, limitations with the test data and, to a lesser extent, unexpected but deliberate efforts to alter the vehicle population such as the recent "cash for clunkers" program of the US government may have somewhat thwarted this effort. Given these factors, it is rather difficult for a regular MOVES user to judge the adequacy of the proposed adjustments. 3. Specific Comments 3.1 Ambient Temperature 3.1.1. I concur with the observation that the principal influence of ambient temperature (Tamb) on emissions is during the warm-up phase of cold-starts. Its influence on warmed-up vehicle running emissions is relatively small - albeit not nil, particularly, under extremely cold conditions when steady-state temperatures of vehicle components (lubricants, tires, etc.) may stay below their "normal" values. 3.1.2. It is not certain that the difference in emissions between Bags 1 and 3 of the FTP cycle can fully account for cold-start emissions under extremely cold conditions 42 ------- (below 32°F) when it takes extended periods of time to reach stead-state. Hence, adjustments based on these data will probably result in underestimates. EPA has seen increased emphasis by manufacturers on decreasing the amount of time it takes to light off the catalytic convenor in order to address tighter emission standards, and since the 1990 's vehicles have had to meet emission standards even at low temperatures. Once the catalytic convenor is fully operational, any small effects from the ambient temperature on emission formation in the engine are easily lost in the catalyst. EPA believes that any temperature effects not captured in the first bag (505 seconds) of the FTP are negligible and existing data bears that out. 3.1.3. The decision to neglect adjustment for ambient temperatures above 75°F is a reasonable but not a perfect one. Reference (1) provides some evidence for less fuel consumption and CC>2 emissions at higher ambient temperatures - perhaps due to less throttling (higher volumetric efficiency). Other emissions are probably also affected, but the test data do not seem to allow for these smaller effects - as noted on page 7 of the report, in the discussion of the Tamb adjustment for NOX emissions. 3.1.4. Part of the difficulty with adjusting for Tamb in the general fleet may be due to the many vehicle parking options: outdoors, unheated indoors, heated indoors or with a plugged-in block heater. If a vehicle is parked outdoors, the wind chill factor might also influence cold-start emissions. The test data do not seem to account for all of these factors. The temperature adjustments in MOVES are intended to represent the effects on vehicle emissions when the ambient temperature to which the vehicle is subjected is known. There may be factors that cause difficulty in determining the appropriate temperature to apply to the fleet, such as the variation of ambient temperature over the area you wish to model. However, these are issues for guidance on how best to use the model for specific scenarios. 3.2. Humidity 3.2.1. One would expect a weak dependence of carbon dioxide emissions on ambient humidity, as reported forNOx. The EPA certification humidity adjustments should, however, account for this effect. 3.2.2. I expect that the EPA certification humidity adjustments are sufficient for inventory work. 3.3. Air Conditioning 3.3.1. The US EPA report indicates that all emission tests with the A/C on were carried out at 95°F only. This implies that the A/C adjustments are not based on emission data obtained at the same temperature, with A/C on and A/C off. If so, according to Reference 1, a significant "error" may have been incurred by not accounting for the co-existing effect of Tamb on some emissions such as those of CO2. 43 ------- On the contrary, because the MOVES model does not apply temperature adjustments to running emissions (except for particulate matter), the comparison ofA/C on emissions at high temperature andA/C off emissions at low temperature allows the A/C correction to incorporate any necessary temperature effect. However, as discussed in Section 2.3.2. our data suggests such effects are not significant. 3.3.2. A/Cs are employed for defogging at all temperatures - particularly, at lower temperatures. This secondary use of the A/C along with associated emission effects do not seem to have been accounted for (according to Ref 1, defogging costs a 1.5 - 7% in CC>2 emissions at 55°F - depending on driving cycle). MOVES does not account for the A/C effects at low temperatures from the use of A/C for defogging. The text has been updated to describe this omission. 3.3.3. Many modern vehicles are equipped with climate control systems, which are usually set by drivers to maintain automatically a preset optimum compartment temperature. The A/C systems of these vehicles switch on when this temperature set-point is exceeded - irrespective of humidity (as is the case with house thermostats). The conditioned air is cold and dry (often reheated with engine coolant). Hence, in these modern vehicles the compressor usage may be largely independent of humidity. The compressor load and hence energy use and some emissions are however very dependent on ambient air humidity. Modeling the behavior of modern A/C systems can be very tricky. As a first cut, MOVES simply addresses the need for A/C based on how comfortable humans will be based on the combination of temperature and humidity. This should adequately capture the need for A/C and the extra loads that result for inventory estimates. A better A/C load model may be developed as our understanding of these systems improves. 3.4. Inspection and Maintenance 3.4.1. The repeated application of MOBILE 6 to predict the relative emission consequences of various I/M program design features appears to involve certain assumptions; viz., all vehicles at a given age have the same odometer reading, are subject to the same deterioration rates, and, if repaired, experience the same emission improvements. It may be worthwhile to test the benefits of replacing these point assumptions with appropriate distributions or variables. Even with the use of distributions, the average impact of 1/M programs on fleet emissions would be the same. We believe the added complexity of using distributions would only add to the complexity of our already complex modeling and provide very limited insight into the benefits of I/M repairs. 3.4.2. Future failure rates will likely be smaller than current ones - largely due to incremental improvements in vehicle technology but also due to the observed shift to inherently low emission vehicles. 44 ------- All modeling of future model years is fraught with uncertainty. EPA has taken the position that improvements in emission performance will only occur if there is an incentive to improve, such as new emission standards. As such, it is reasonable that the existing failure rate, which is already very small, will continue into the future unless there is some regulatory reason why manufacturers would take the time and money to develop solutions that would significantly reduce their failure rates. Even without reductions in failure rates, the benefits of I/M programs will decrease as the emission impact of failure grows smaller on vehicles with new (lower) emission standards. 3.4.3. MOBILE 6 apparently assumed that waived vehicle emission rates are invariably 20% lower than those of failed vehicle emissions (see Ref, 2). Is this assumption carried through in MOVES? If so, it may be worthwhile to re-examine it. In MOVES, vehicles which receive waivers are assumed not to have been repaired at all. The text has been updated to include this information. Waived vehicles are typically a small fraction of the fleet and are difficult to study. Given the limited impact that these vehicles will have on total fleet emissions, determining a more precise impact from waiver vehicles will not be a high priority. 3.4.4. The National Research Council (Ref. 2) raised a number of additional I/M related concerns with MOBILE 6: (a) assumption that vehicles with and without I/M deteriorate at the same rate; (b) no explicit allowance for those vehicles that are repaired before or after inspection but rapidly revert to high-emitter status; and (c) no I/M credit for high emitters that are scrapped or shipped outside of the region. It would be helpful to explain how these concerns were addressed in MOVES. The description of the I/M program effects in the report has been revised to more explicitly address the concerns of the National Research Council. 3.4.5. Another concern in the I/M community, namely the effectiveness of OBD systems and OBD based I/M programs, deserves also a fuller discussion. 4. Editorial Comments EPA has made many changes to the text of the report to address the following editorial comments. 4.1. The term "adjustment", as used in the title of the report, expresses the goal of the effort clearly and concisely. The terms "correction factor" and "adjustment factor", as used in the body of the report, are less clear. First, a correction is not an adjustment. Second, the word "factor" implies a multiplication whereas most of the proposed adjustments are additive. I recommend that the report stick to the term "adjustment" throughout the report. 4.2. This is not a free-standing report. Its contents cannot be fully understood without referring to a series of other reports (at least, the documentation of MOBILE 6). It would 45 ------- have been preferable to have a free-standing report - not for the sake of peer reviewers but for younger MOVES users who haven't witnessed the evolution of MOBILE. 4.3. A Table presenting the principal assumptions made and the resulting effectiveness estimates for major I/M program types would add to the value of the report. 4.4. I am assuming that the final report will include lists of acronyms (with explanations), tables, figures as well as equation numbers, etc. - all the usual pieces that make a report a bit more accessible. 4.5. Minor notes: • Last sentence on page 8 refers to Section 4.1.3, which does not exist. • Section 4.1 apologizes for lack of data with A/C on MC. Do you really mean motorcycles? • Section 4.1, third paragraph, refers to Appendix A for a list of vehicles and their description. It should instead refer to Table 4-1. Also, I don't see any description of the vehicles. • The title of Section 4.3.2 reads "Energy Emissions". It should probably read "Energy Consumption". • On page 30, "OBD" is spelled "OBC". 5. Response to Specific Questions Posed in the US EPA Letter to Me 5.1 The Clarity of the Presentation • The report is well written and very clear to individuals with a technical background in the subject area. It may however require some editing to make it more easily accessible to a wider audience - if this were necessary. 5.2 The Integration of Information from Multiple Areas • The report is based on an enormous volume of previous work and the resulting information. Given the difficulty of condensing this vast volume of information into a relatively compact report, the author(s) have done very well. The information is well integrated. However, as noted in my general comments (Section 1 above), the report is not a stand-alone document. It cannot be fully understood without reading its references. 5.3 The Appropriateness and Completeness of the Literature Discussed • The literature referenced in the report is highly appropriate and sufficiently complete. 5.4 Appropriateness of the Resulting Adjustments 46 ------- • In spite of the inherent complexity of the subject and the limitations of the available data, the author(s) have succeeded in: o Identifying those effects that call for adjustments o Eliminating those effects (variables) that are too insignificant to adjust for o Deriving robust adjustments that reflect the totality of the empirical evidence available and also conform to theory • In my opinion, the adjustments are highly appropriate. The few comments provided in this review are intended to contribute to any future effort to update MOVES and make it as useful as possible to all potential users. 6. References (1) Weilenmann, M.F.; Vasic A-M; Stettler P.; and Novak, P. Influence of Mobile Air- Conditioning on Vehicle Emissions and Fuel Consumption: A Model Approach for Modern Gasoline Cars Used in Europe. Environ. Sci. Technol. 2005, 39, 9601-9610. (2) National Research Council. Evaluating Vehicle Emissions Inspection and Maintenance Programs. The National Academies Press, Washington, DC. 2001. (3) Eisinger D.S. and Wathern, P. Policy Evolution and Clean Air: The Case of US Motor Vehicle Inspection and Maintenance. Transportation Research Part D. 2008, 13, 359-368. Toros Topaloglu, Ph.D., P.Eng. Thank you. 47 ------- Appendix E - ENVIRON International Corporation, Comments ENVIRON Review of EPA Draft Report: "MOVES Temperature, Humidity, Air Conditioner, and Inspection and Maintenance Adjustments" As part of the MOVES2010 Peer Review process, EPA solicited comments from Christian E. Lindhjem of ENVIRON International Corporation on the August 2009 draft of report Draft MOVES2009 Highway Vehicle Temperature, Humidity, Air Conditioning, and Inspection & Maintenance Adjustments. Chris Lindhjem has a PhD. in Chemical Engineering from Rensselaer Polytechnic Institute and has more than 15 years of experience in automotive issues with particular focus on emissions from highway and non-road vehicles, engines, and engine fuels. Dr. Lindhjem's comments are copied below, with EPA response in italics. Christian Lindhjem ENVIRON International Corporation 773 San Marin Drive, Suite 2115 Novato, California 94998 415.899.0700 30 September 2009 Introduction This report appears to gather all of the adjustments to the MOVES basic emission rates into one document despite the seeming unrelated topics discussed. Temperature and humidity are ambient conditions that affect the engine and after-treatment control effectiveness. Air condition loads are influenced by ambient conditions, but in fact are only one of many potential loads. New engine standards and inspection and maintenance programs are primarily emission reduction program credit assessments. Yet despite the varied types of adjustments, it is reasonable to include all adjustments if indeed all adjustments to the model are included in this document. However, as new data or new approaches are required, updating the document could be more difficult or confusing to the reader because the document addresses so many issues. And if this document does not include all such adjustments, it might be confusing to understand all such adjustments split over many documents. 48 ------- Gasoline Vehicle Temperature Adjustments To the extent that temperature adjustments presented here differ from the official test procedures, EPA should make sure that the original data is free of temperature adjustments that the original researchers might have used on the reported data. Often researchers will follow the official test procedures to the letter and adjust the reported values to account for unique test conditions. Overall the method for temperature adjustments seem sound with a few comments noted here that might help the explanation for the reader. EPA correctly separated the start and running temperature adjustments to account for the likely different effects when the engine and after- treatment are at operating temperature. For gasoline vehicle start emissions, hydrocarbon and carbon monoxide effects are presented sufficiently to understand the results. The NOx results were considerably more complex, and there may be reasons for the observed start NOx emission with regard to technology by model year grouping. However, given that the effect is more predominant with older model years, the NOx effect may be less important for most uses of MOVES. For all cases, it would be helpful to put the adjustment estimates in perspective of the base emission rates to demonstrate the relative importance of the temperature adjustment effect, such as inclusion of a description of the percentage effect. For gasoline vehicle running emissions, I agree with the assessment that ambient temperature has little effect on emissions. It might be helpful to note several same-vehicle tests at different ambient temperatures in Figure 2-2, such as by symbol and/or lines, to demonstrate that a temperature effect is not observed with the same equipment. Mixing vehicle tests and temperature conditions tests may mask an effect that could be observed in same vehicle tests. Diesel Vehicle Temperature Adjustments The diesel vehicle start emissions are presented, and I have no dispute with the results for start emissions determined. But this discussion could use some context in terms of vehicle types and applicability. For instance, based on the use of the FTP test cycles to determine the start emissions, I suspect that these 12 vehicles were pickup trucks, light heavy-duty vehicles, so EPA should discuss the relevance of using these vehicles to represent all diesel vehicles. The text of this document describing the temperature adjustments to diesel start emissions has been updated to better address the types of vehicles in the samples used. EPA makes no claim about particulate matter or running emissions temperature adjustments for diesel vehicles, so the report approach is inconsistent to that for gasoline vehicles. In addition, there was no discussion of whether these vehicles used after-treatment devices (either diesel particulate filters (DPF) or oxidation catalysts (OC) or future systems expected for 2010 model years and beyond) and how that might affect the results as was done to incorporate the cold weather CO and HC requirements for gasoline vehicle temperature adjustments. 49 ------- Less is known about the effects of temperatures on diesel particulate matter emissions. The text of this document describing the temperature adjustments to diesel particulate matter emissions has been updated to better address the technologies in the samples used. Cold Weather CO and HC Requirements The methodology to estimate the benefits credited to the light-duty cold weather regulations appears to be a reasonable approach as presented. However, it is questionable if the cold weather regulation adjustments should be applied to high emitters given that the control systems might not be functioning. To the extent that MOVES identifies high emitters, independent temperature adjustments should be applied to high emitters. Since MOVES does not identify high emitting vehicles during calculations, independent temperature adjustments for high emitters cannot be applied. Humidity Without performing additional testing, it is reasonable to use the Federal Register humidity corrections. Because these adjustments would be multiplicative, they would be applicable to the lower emission rates of later model years. Air Conditioning The air conditioning adjustment approach appears to be counterintuitive to approach of MOVES defining power bins to reflect the engine loads. There may be some reasons for this approach given that idle and coasting\braking bins would not otherwise incorporate the auxiliary air conditioner loads. Another reason could be that air conditioner loads would oscillate between VSP bins when the compressor is engaged and disengaged unrelated to the driving demands. The approach presented is easy to follow in concept, but there should be more description of the overall air conditioning effect for sample vehicle types. To help the reader understand how important the air conditioning adjustment is, EPA should plot of the effect with respect to the humidity index, noting the heat index below which there is no air conditioning adjustment. The "ActivityTerm" coefficients for the ' ACOnFraction' estimates should be presented in the document. The text of this document describing the air conditioning adjustments has been updated to better display the effects of the activity adjustments versus the humidity index. Inspection and Maintenance (I/M) Using the I/M benefits from the MOBILE6 analysis is a reasonable approach without an extensive reanalysis of the benefits under the MOVES modeling framework. As with the assessment of new vehicle emission standards, the emission credits estimated for various programs may not be entirely based on a quantitative assessment of each program. Therefore, because the credits assigned have been well vetted under the MOBILE6 plan, it becomes a more accepted approach to use for MOVES as well. 50 ------- Because the MOBILE6.2 benefits only include HC, CO, and NOx and the Figure 2-6 was given as evidence of a relationship between PM and HC emission, PM benefits for I/M programs should also be considered. It would stand to reason, even without direct evidence, that emission reductions of the primary pollutions evaluation would also lead to PM emission reductions when malfunctioning vehicles are repaired. EPA does not yet have sufficient data to estimate PM emission reductions for I/M programs without further evidence that repairs that reduce HC, CO and NOx emissions will significantly affect PM emissions. Errata Numerous changes to the text have been made to address these minor edits. Page 12 above Figure 2-4; "adjustments" has an extra "s" Page 25 below Table ?4-5? (label missing), just above section 4.3.3, "If and when ..." Page 27 above Section 4.4: "A/C correction factors of less than unity or unity were found for. Some Tables have headings and some table headings are missing and references for those tables in the text are not clear. Section 5, Eq. 1, 2, 3 and variable description of nonEVI emissions should read the same, such as "EnonEVI" in all equations and descriptions. Equation 6 or should it be Equation 5? label on last page is incomplete 51 ------- Appendix F - Julio Vassallo, Comments Review of US EPA's "Draft MOVES2009 Highway Vehicle Temperature, Humidity, Air Conditioning, and Inspection & Maintenance Adjustments" September 25, 2009 Additional comments, not part ofthe formal MOVES2010 Peer Review process, were submitted by Julio E. Vassallo on the August 2009 draft of report Draft MOVES2009 Highway Vehicle Temperature, Humidity, Air Conditioning, and Inspection & Maintenance Adjustments. Julio Vassallo is a Chemical Engineer and the Technical Manager of Area new vehicles Approval and Certification in the the Laboratory of Vehicle Gaseous Emission Control (LCEGV) of the Ministry of Environment and Sustainable Development (SAyDS) in Buenos Aires, Argentina. Julio Vassallo's comments are copied below, with EPA response in italics. Page 6: The behavior presented for vehicles without (or deactivated) catalyst (that might be included in the group pre 1981) is different with respect to the NOx emission, those with catalyst. The vehicle can be considered as two reactors in series, combustion reactor in homogeneous phase (cycle Otto engine) and oxidation-reduction catalytic reactor (catalytic converter). The generation of NOx in the engine is principally a function of temperature in the cylinder and the partial pressures of nitrogen and oxygen. Therefore when the engine is cold the issue (without catalyst) is the lowest and increases as the engine warms up (example in doc "Start Emission" vehicles without catalytic converters, emissions EVI240 consecutive test series). In contrast to an engine with catalyst but also emissions start to increase with increasing temperature once you reached the temperature catalyst "light off" decreases again (example in doc "Cold Emission" vehicles with catalyst) Moreover, the emission of NOx is also heavily dependent on power (VSP) Then, depending on which portion of the emission is correlated is provable that the temperature hasn't the same effect (function) for vehicles with catalyst and without catalyst. I think that the start NOx emission (FTP NOx emission Bag3 minus Bagl) of the vehicles without catalyst are highest than those with catalyst and has different start temperature dependence. With CO an HC start emission is different, because both reactors (engine and catalyst) the emission decreases with temperature (example in doc "Start Emission" vehicles with and without catalytic converters, emissions EVI240 consecutive test series). The behavior of catalyst and non-catalyst vehicles is handled in MOVES by having separate temperature adjustments by model year group. 52 ------- Page 7:1 think that is provably if you correlate taking account VSP and the vehicles with catalyst in other group that those without catalyst the R-square coefficient will be better. In MOVES, since the temperature adjustments are grouped by model year, some model year groupings will include catalyst and some non-catalyst vehicles in the correlations. Unless MOVES is redesigned to accommodate separate technologies in addition to model years, a separate correlation for each technology is not possible. Page 8: I agree that to reach working temperature (running) both the emission generated in the engine and removed by the catalyst that should not be so sensitive to the test temperature such as the start While the supply air temperature should influence the reactions of improving combustion efficiency at higher temperatures of income, is provable that the high working temperatures of engine determine less sensibility for that purpose, on the other hand those vehicles with catalyst in the regime temperature will have to be less sensibly since over 90% of the pollutants are converted and that masks any engine inefficiency specially to low exhaust flow (low rpm / VSP). Page 11: The analysis of diesel engine emissions are different from that of Otto cycle, in the case of CO are not as significant and therefore may be more affected by the measurement uncertainty when it comes to a small population, such as that of the reporter. For NOx, in this case normally pre 2007 alone technologies are oxidation catalysts (remove only CO and HC) therefore in this case has only effect the engine and NOx emissions should increase, ie emissions Bag 1 Bag minus 3 should be negative. For example the mean value obtained to 34.6 ° F will have to be negative -2.6? I haven't studies with a diesel emission test series to different ambient temperature in the start, but you have studies for example that about humidity and temperature effects how I adjunct (in page 7 HUMIDITY AND TEMPERATURE CORRECTION FACTORS FOR NOx EMISSIONS FROM DIESEL ENGINES SwRI Project No. 03.30.10.06599). Page 13: as you get this value? This increasing of cold start emission (value 0,5611592) will be in grams per mile?. My doubt is because I think that if you have a total increase of 2.086 g NMHC = 0,43 (M Bagl+MBag2) + 0,57 (MBag2 + MBag3); then the value in grams of the cold start (M NMHC Bag 1) should be higher than 0,5611592. The effects of the MSATrule on the cold temperature adjustment for HC emissions of engine starts will need to be revisited once vehicles compliant with these standards are available for testing. The current adjustment is based solely on the emission standard values. 53 ------- Appendix G- Coordinating Research Council Project E-68a Comments December 3, 2009 Additional comments regarding the adjustments described in the report, "DraftMOVES2009 Highway Vehicle Temperature, Humidity, Air Conditioning, and Inspection & Maintenance Adjustments, that were not part of the formal MOVES2010 Peer Review process, were submitted as part of the Coordinating Research Council (CRC) Project E-68a. Comments from the report that are relevant to the topics covered in the EPA report are copied below, with EPA response in italics. Readers are encouraged to obtain the entire CRC Project E-68a report in order to fully understand the comments in their full context. Correction Factors (Fuels and Temperature) Regarding temperature correction factors, EPA examined recent data and found that cold start HC, CO, and NOx emissions should be adjusted for temperature, but there is no ambient temperature effect on running exhaust emissions. EPA developed additive cold start increments for HC, CO, and NOx that increase with lower temperatures. One concern with the temperature increments is that there is no analysis of how these may change as vehicles age, and the available data seemed to omit the CRC E-74b testing program, which was completed in May of 2009. EPA could utilize the Kansas City temperature data to determine whether the temperature relationships change with vehicle age. Also, the CRC E-74b testing program data could be used to further check the MOVES cold start correction factors. EPA believes that studies, such as the Kansas City study and CRC E-74, which include vehicles of different ages, but do not follow the vehicle fleet over time, are inadequate to conclude that the effects of temperature vary by vehicle age. EPA in cooperation with others, is planning a study specifically designed to follow the vehicle fleet over time and should produce the type of information needed to determine the effects of vehicle age on temperature effects. A second concern is that the method used to develop HC temperature increments for the MSAT rule (which requires lower HC standards at cold temperatures) assumes a compliance margin with respect to the HC standards at 75° F, but no compliance margin with respect to the HC standards at 20° F. As a result, the HC increments for vehicles meeting the MSAT requirements are over-estimated. The method should be revised to include a compliance margin at 20° F to be consistent with the margin currently being utilized at 75° F. EPA believes that any compliance margin at 20 degrees Fahrenheit will likely differ significantly from the margin observed at 75 degrees. Further testing will be needed on vehicles compliant with the MSAT standards to determine the appropriate margin. 54 ------- A third concern is that vehicles subject to the lower MS AT HC standards will very likely have much lower CO emissions as well. Once vehicles are certified to the MSAT cold HC standards, an analysis should be conducted of certification or other data to determine how much the CO increments change for these vehicles as well. CO emission rates already assume the impact of explicit standards for CO emissions at low temperatures. EPA believes that strategies to reduce HC emissions at low temperatures will likely have minimal further impacts on CO emissions at low temperatures. Particulate Matter Emissions for Gasoline Vehicles Temperature correction factors were estimated from the matched vehicle pairs. Unlike HC, CO, and NOx emissions, where the temperature correction factors were only for cold start emissions, EPA found an increase in running PM emissions with decreasing ambient temperatures, albeit lower than for the cold start. The first concern is that the combined MSAT and Kansas City data on matched pairs does not appear to support a cold temperature adjustment for running emissions. Results from other studies such as NFRAQS should be included in the analysis, with special regard to high PM emitters. It is true that data from the matched pairs in the combined MSAT and Kansas City was inconclusive in determine the temperature effect on running emissions. However, using other analysis techniques, EPA was satisfied that a significant temperature effect could be determined. A third concern is that in the draft model, vehicles meeting lower HC standards in response to the MSAT rule currently are not assumed to have lower PM emissions. Since HC and PM emissions seem to correlate well, we believe there will be lower PM emissions with a lower HC standard at cold temperature. In the section on correction factors, we recommend evaluating certification data or other data to examine the effect of cold HC standards on HC and CO emissions. This should be extended to PM as well if possible. EPA has not been able to establish a clear correlation between HC emissions andPM measurements that would support assuming that PM emissions at low temperatures would be significantly affected by changes in the HC standard. EPA will be updating the emission estimates in future versions of MOVES as new data on vehicles certified to the new standards are tested. Summary of Recommendations EPA should utilize the Kansas City data to determine whether temperature correction factors change with vehicle age. Also, the CRC E-74b testing program data could be used to further check the MOVES cold start correction factors. As stated above, we believe the Kansas City data is inadequate for this purpose, but we hope to collect appropriate data to do this analysis in the future.. 55 ------- The Tier 2 cold temperature response should be lower than for Tier 1 vehicles. In addition, the MSAT rules should reduce CO emissions as well as HC emissions. The method used to develop HC temperature correction factors for the MSAT rule should be revised to include margin at 20° F to be consistent with the margin currently being utilized at 75° F. We don't believe these changes are justified based on currently available data. Now that MSAT vehicles are entering the fleet, we hope to gather in-use data on vehicles meeting these standards. The combined MSAT and Kansas City data on matched pairs does not support a cold temperature adjustment for running emissions. Results from other studies such as NFRAQS should be included in the analysis, with special regard to high PM emitters. 56 ------- |