A more accurate method using MOVES (Motor Vehicle Emission Simulator) to estimate emission burden for regional-level analysis

Abstract

The U.S. Environmental Protection Agency’s (EPA) Motor Vehicle Emission Simulator (MOVES) is required by the EPA to replace Mobile 6 as an official on-road emission model. Incorporated with annual vehicle mile traveled (VMT) by Highways Performance Monitoring System (HPMS) vehicle class, MOVES allocates VMT from HPMS to MOVES source (vehicle) types and calculates emission burden by MOVES source type. However, the calculated running emission burden by MOVES source type may be deviated from the actual emission burden because of MOVES source population, specifically the population fraction by MOVES source type in HPMS vehicle class. The deviation is also the result of the use of the universal set of parameters, i.e., relative mileage accumulation rate (relativeMAR), packaged in MOVES default database. This paper presents a novel approach by adjusting the relativeMAR to eliminate the impact of MOVES source population on running exhaust emission and to keep start and evaporative emissions unchanged for both MOVES2010b and MOVES2014. Results from MOVES runs using this approach indicated significant improvements on VMT distribution and emission burden estimation for each MOVES source type. The deviation of VMT by MOVES source type is minimized by using this approach from 12% to less than 0.05% for MOVES2010b and from 50% to less than 0.2% for MOVES2014 except for MOVES source type 53. Source type 53 still remains about 30% variation. The improvement of VMT distribution results in the elimination of emission burden deviation for each MOVES source type. For MOVES2010b, the deviation of emission burdens decreases from −12% for particulate matter less than 2.5 μm (PM_2.5) and −9% for carbon monoxide (CO) to less than 0.002%. For MOVES2014, it drops from 80% for CO and 97% for PM_2.5 to 0.006%.

Implications: This approach is developed to more accurately estimate the total emission burdens using EPA’s MOVES, both MOVES2010b and MOVES2014, by redistributing vehicle mile traveled (VMT) by Highways Performance Monitoring System (HPMS) class to MOVES source type on the basis of comprehensive traffic study, local link-by-link VMT broken down into MOVES source type.

Introduction

The U.S. Environmental Protection Agency’s (EPA) Motor Vehicle Emission Simulator (MOVES) is a state-of-art motor vehicle emission simulator that is capable of estimating mobile emissions at the national, county, and project levels for criteria pollutants, greenhouse gases, and air toxics. It is the official mobile emission model for the State Implementation Plan (SIP) development, transportation conformity determinations, and other purposes. The current approved version is MOVES2010b, which is released on June 2012 to replace MOVES2010a. MOVES2010a builds on the functionality of previous versions: MOVES2004, MOVESDemo, DraftMOVES2009, and MOVES2010. However, the latest version is MOVES2014 released on July 2014. Three new emission control programs included in MOVES2014 and associated with regulations after the release of MOVES2010b are Tier 3 emission standards, heavy-duty engine and vehicle greenhouse gas (GHG) regulations, and the second phase of light-duty vehicle GHG regulations. MOVES2014 also includes new and updated emission data and improvements of graphical user interface (GUI).

The analysis using MOVES requires both non-traffic-related inputs and traffic-related inputs, including annual vehicle mile traveled (VMT) by Highways Performance Monitoring System (HPMS) vehicle class and vehicle population by MOVES source (vehicle) type. To develop traffic-related MOVES inputs for regional emission burden analysis, local link-by-link traffic data, including speed and VMT, are required.

The accuracy of emission burden estimation using MOVES depends on all MOVES inputs. Kall et al. (2014) conducted MOVES model sensitivity analysis on MOVES source population by changing source population and the population percentage by MOVES source type in HPMS vehicle class. It indicated up to 6.9% change for volatile organic compound (VOC) emission, maximum of 4.4% change for carbon monoxide (CO) emission, significantly change of 35.6% for nitrogen oxide (NO_x) emission, 19.7% change for carbon dioxide (CO₂) emission, and the biggest emission change of 66.3% for particulate matter less than 10 micrometer (PM₁₀). However, VMT is also changed in addition to source population. Therefore, the final emission changes may not be only from source population change. Koupal et al. (2013) also performed MOVES sensitivity study based on the MOVES inputs, prepared by 30 states, covering 1,400 counties. It indicated that the VMT change, overall, contributed the highest change in emission, with particularly large increases for NO_x and PM. The increase of daily emission burden in that study could reach 78.6% from VMT fraction change and 15.7% from population change.

The emission burden in EPA MOVES includes both running exhaust, which is the function of VMT, and nonrunning emission (e.g., start, evaporative, extended idle), which is the function of source population (EPA, 2012). Theoretically, the change of source population has impact on nonrunning emission, but not on running exhaust emission. When MOVES was developed, most of the fleet and activity data are calculated based on more readily available traffic data set (EPA, 2010). HPMS traffic data set from Federal Highway Administration (FHWA) was used in MOVES for VMT, VMT growth factor, and VMT distribution among roadway types. However, MOVES has its own vehicle (source) type, which is a subset of HPMS vehicle type, for emission calculation. In MOVES, the base year VMT by HPMS vehicle class is converted to VMT by MOVES source type. The conversion is done using the population fraction by MOVES source type in HPMS vehicle class and relative mileage accumulation rate (relativeMAR). It results two problems. Problem 1 is that the base year VMT distribution is impacted by MOVES source population. The second problem is from relativeMAR. The relativeMAR is calculated as the ratio of the annual MAR to the highest MAR within the HPMS vehicle class based on national-level data set and data set from MOBILE6 (EPA, 2010). It is a list of ratios for all MOVES source types and ages. The relativeMARs are considered as constants for all regions and counties and stored in MOVES default database. Therefore, the derived (converted) VMT for each MOVES source type may not be consistent with the local VMT for every specific region and year. It then results in the impact on the running exhaust. The impact might be very limited if the analysis is for the entire county and the local VMT fraction is close to the derived VMT fraction from default relativeMAR. But it might be much significant for regional analysis, especially transportation conformity determination, since the study area include only partial of the roadway network.

For regional-level analysis using MOVES, the comprehensive traffic study is usually capable of providing more detailed local traffic data, i.e., local link-by-link VMT by MOVES source type, based on the local traffic field survey. In the local traffic field survey, the actual total numbers of vehicles by MOVES source type were counted on each sample roadway. Multiple sample roadways were selected for each MOVES road type. Those local vehicle classifications were then used to calculate the local link-by-link VMT by MOVES source type. Therefore, the base year VMT by MOVES source type is available. It is then necessary to adjust the derived VMT distribution among MOVES source types from the default relativeMAR by using local VMT distribution. The running exhaust emissions, therefore, can be more accurate by adjusting the derived VMT distribution to reflect the local VMT distribution across MOVES source types. This paper will address (1) why and how source population and relativeMAR have impact on the running emission burden, and (2) how to mitigate the impact from source population and relativeMAR using more detailed local traffic data, i.e., local VMT by MOVES source type.

Methodology

Total Activity Generator (TAG), which is the same in MOVES2010b and MOVES2014, is the module in MOVES to distribute VMT from HPMS source (vehicle) type to MOVES source type based on the calculation as shown in eq 1 (EPA, 2009).

(1)

where i is the MOVES source type ID, j is HPMS source type ID, Tf is the travel fraction, Pf is population fraction, and r is the relativeMAR. Based on the travel fraction of each MOVES source type, the VMT fraction by MOVES source type in each HPMS source type is calculated by eq 2.

(2)

where i is MOVES source type ID within same HPMS source type j and . However, more detailed traffic study can provide local VMT by MOVES source type. Equation 3 is used to calculate the local VMT fraction by MOVES source type in each HPMS class.

(3)

where is local VMT directly from traffic study for MOVES source type, i. One variable for each MOVES source type, shown in eq 4, is introduced to describe the ratio of local VMT fraction to the VMT fraction derived in MOVES.

(4)

where i is MOVES source type ID. Equation 5 is derived from eqs 2, 3, and 4.

(5)

Therefore, linear equations (12 for MOVES2014 and 11 for MOVES2010b), shown in eq 6, can be derived from eqs 3, 4, and 5. It is, then, used to solve all variables .

(6)

where i is MOVES source type ID, j is MOVES source type ID other than i and within the same HPMS class with i. Then, adjusted relativeMAR is calculated by multiplying variables with default relativeMAR .

The advantage of applying to relativeMAR is that population fraction by model year is unchanged and that actual source population is also unchanged. Therefore, the running emission rate for each MOVES source type and the total nonrunning emission will not be affected by the change of relativeMAR. The only change is the VMT distribution across MOVES source types, which is adjusted to be consistent with the local VMT distribution.

To analyze the effect of the adjusted relativeMAR, MOVES runs are conducted for a defined custom domain in Illinois and six different counties, King County in Washington, Kitsap County in Washington, Pierce County in Washington, Snohomish County in Washington, St. Thomas in Virgin Islands, and Lake County in Indiana. The calendar years in MOVES runs are 2011, 2012, 2018, 2030, 2035, and 2040. All MOVES runs were grouped into six scenarios. The six scenarios, as shown in Table 1, are designed to cover the uses of original (default) relativeMAR, adjusted relativeMAR, MOVES2010b, and MOVES2014. Scenario 1-1 includes multiple MOVES2010b runs with all available MOVES source types modeled and original relativeMAR used in each run. Scenario 1-2 includes MOVES2010b runs, each of which is for single MOVES source type using original relativeMAR. Scenario 1-3 includes MOVES2010b runs with all available MOVES source types modeled and adjusted relativeMAR used in each run. Scenarios 2-1, 2-2, and 2-3 are MOVES2014 version scenarios 1-1, 1-2, and 1-3 instead of using MOVES2010b, but using MOVES 2014. The difference between scenarios is either number of MOVES source types modeled or the use of adjusted relativeMAR.

Table 1. MOVES run scenarios

		MOVES2010b	MOVES2014
Original relativeMAR	All available MOVES sources	Scenario 1-1	Scenario 2-1
	Single source	Scenario 1-2	Scenario 2-2
Adjusted relativeMAR	All available MOVES sources	Scenario 1-3	Scenario 2-3

Each scenario includes several MOVES runs to estimate total emission burdens of CO and PM_2.5 from all processes for county level using MOVES inventory calculation type. Each MOVES run was setup by following EPA technical guidance for SIP and transportation conformity. Scenarios 1-1 and 2-1 are designed to discover deviations of VMT and emission burden by MOVES source type using original relativeMAR. Because only single MOVES source type is modeled in scenarios 1-2 and 2-2, there would be no VMT conversion from HPMS to MOVES source type in MOVES. Therefore, scenarios 1-2 and 2-2 can be designed to obtain the correct/baseline emission burden by MOVES source type using original relativeMAR. Scenarios 1-3 and 2-3 are designed to demonstrate the effect of adjusted relativeMAR. Scenarios 1-1 and 2-1 include multiple MOVES runs for all alternatives (Build and No Build), counties, and years. Only two MOVES runs with the biggest VMT deviations from each of scenarios 1-1 and 2-1 are selected for other scenarios. All MOVES inputs, especially the source population and base year VMT, are not changed with the change of scenarios. All calendar years in MOVES runs are base years, since all MOVES inputs are developed using local data.

Results and Discussion

VMT results

The results, shown in Table 2, from MOVES2010b runs (using original relativeMAR) in scenario 1-1 indicated that VMT deviations vary from county to county and from year to year. In Table 2, the annual VMT from MOVES activity output table were aggregated to MOVES source type level. The percent changes from MOVES-derived VMT to local VMT (VMT by MOVES source type) were calculated. Table 2 lists maximum and minimum percent VMT changes for each of three regions among all years and all alternatives. As shown in this table, MOVES runs in Washington State have the smallest VMT variations, from −0.2% to 0.1%. The biggest VMT variations are from Illinois and Indiana, ranging from −12.7% to 13.5%. However, this does not mean Washington State will always have the lowest the VMT variations, or that Illinois and Indiana always have the highest VMT variations. Since MOVES uses one set of relativeMAR to distribute the VMT from HPMS source type to MOVES source type, the VMT variations depend on how much the MOVES-derived VMT is close to local VMT. Therefore, the VMT deviations vary from case to case.

Table 2. VMT variations for each MOVES source type from the MOVES2010b runs in scenario 1-1 compared with local traffic

		Washington State		Virgin Islands		Illinois and Indiana
MOVES Source Type ID	MOVES Source Type Name	Min. Difference	Max. Difference	Min. Difference	Max. Difference	Min. Difference	Max. Difference
11	Motorcycles	−0.2%	0.1%	0.0%	0.0%	0.0%	0.0%
21	Passenger Cars	−0.2%	0.1%	0.0%	0.0%	0.0%	0.0%
31	Passenger Trucks	−0.2%	0.1%	0.0%	0.0%	−0.8%	0.2%
32	Light Commercial Trucks	−0.2%	0.1%	0.0%	0.1%	−0.4%	2.3%
41	Intercity Buses	−0.2%	0.1%
42	Transit Buses	−0.2%	0.1%	−1.8%	−0.9%
43	School Buses	−0.2%	0.1%	0.9%	2.0%
51	Refuse Trucks	−0.2%	0.1%
52	Single Unit Short-haul Trucks	−0.2%	0.1%	−3.3%	0.3%	−11.5%	5.3%
53	Single Unit Long-haul Trucks	−0.2%	0.1%	0.0%	0.2%	−1.4%	7.2%
54	Motor Homes	−0.2%	0.1%			0.0%	0.0%
61	Combination Unit Short-haul Trucks	−0.2%	0.1%	−0.9%	0.8%	−12.6%	13.5%
62	Combination Unit Long-haul Trucks	−0.2%	0.1%	−5.1%	5.7%	−2.3%	3.5%

MOVES2014 uses the same method of applying default relativeMAR and MOVES source population fraction to distribute the VMT from HPMS to MOVES. However, MOVES2014 uses different set of relativeMAR. This results different VMT variations compared with MOVES2010b. As shown in Table 3,VMT variations from MOVES2014 are much higher than from MOVES2010b. The highest VMT variation, 97.5% change, is from Virgin Islands for MOVES source type 62.

Table 3. VMT variations for each MOVES source type from the MOVES2014 runs in scenario 2-1 compared with local traffic

		Washington State		Virgin Islands		Illinois and Indiana
MOVES Source Type ID	MOVES Source Type Name	Min. Difference	Max. Difference	Min. Difference	Max. Difference	Min. Difference	Max. Difference
11	Motorcycles	−0.2%	0.1%	0.0%	0.2%	0.0%	0.2%
21	Passenger Cars	−5.6%	−5.2%	−6.7%	−6.0%	−18.9%	−14.7%
31	Passenger Trucks	4.0%	4.7%	36.3%	42.0%	27.5%	34.5%
32	Light Commercial Trucks	4.0%	4.7%	36.3%	42.1%	31.4%	38.7%
41	Intercity Buses	−0.2%	0.1%
42	Transit Buses	−0.2%	0.1%	−1.5%	−0.9%
43	School Buses	−0.2%	0.1%	0.9%	2.0%
51	Refuse Trucks	−12.3%	−11.9%
52	Single Unit Short-haul Trucks	18.1%	18.8%	−36.1%	−32.8%	−34.8%	−26.8%
53	Single Unit Long-haul Trucks	21.2%	44.4%	−28.6%	−20.8%	−18.2%	−13.7%
54	Motor Homes	−39.9%	−39.6%
61	Combination Unit Short-haul Trucks	−35.8%	−35.6%	−14.7%	−11.7%	−54.2%	−47.2%
62	Combination Unit Long-haul Trucks	39.8%	40.2%	79.1%	97.5%	11.6%	15.6%

However, Tables 2 and 3 indicated that the VMT variations for MOVES source type 11 (Motorcycles) in both MOVES2014 and MOVES2010b runs are all less than 0.2%. The reason is that MOVES source type 11 is the only MOVES source type in HPMS source type 10 (Motorcycles) for both MOVES2014 and MOVES2010b. Therefore, there is no VMT conversion from HPMS to MOVES sources trigged in MOVES. And the less than 0.2% change of VMT variations for MOVES source type 11 in both MOVES2014 and MOVES2010b can, therefore, be considered to be not a significant change.

Another interesting finding from Tables 2 and 3 is VMT variation for MOVES source type 21 (Passenger Cars). The VMT variations in MOVES2010b are all very small, close to 0.0%, because of the same reason as MOVES source type 11. In MOVES2010b, source type 21 is the only MOVES source type in HPMS source type 20 (Passenger Car). However, the variations for MOVES source type 21 from MOVES2014 runs are significantly higher than from MOVES2010b. This is because of one change made in MOVES2014. FHWA changed their definitions of 20 and 30 (Other Two-Axle/Four-Tire, Single Unit) classes from passenger cars and other two-axle four-tire trucks to short wheelbase and long wheelbase (EPA, 2014). Following FHWA, EPA MOVES2014 combines HPMS classes 20 and 30 into one new class, 25. Therefore, the base year VMT for HPMS class 25 in MOVES2014 will be distributed to MOVES source types 21, 31 (Passenger Trucks), and 32 (Light Commercial Trucks). This results in the VMT variations for MOVES source type 21 in MOVES2014 runs, not in MOVES2010b runs. It also results in the VMT variations for MOVES source types 31 and 32 being bigger in MOVES2014 runs than in MOVES2010b runs. For all other MOVES source types, Tables 2 and 3 show the VMT variations in both MOVES2010b and MOVES2014 runs because of the VMT conversion from HPMS to MOVES source type.

Since only one MOVES source type was modeled in each MOVES run from scenarios 1-2 and 2-2, no VMT conversion from HPMS to MOVES source type is trigged in MOVES. MOVES-derived VMT distribution should be same as the local VMT distribution even though default relativeMAR was used. This was confirmed in Table 4 by less than 0.2% VMT variations from derived VMT to local VMT in both MOVES2010b and MOVES2014 runs except for MOVES source type 53 (Single Unit Long-haul Trucks) in MOVES2014 runs. Therefore, the emission burdens from scenarios 1-2 and 2-2 can be set as the baseline/actual emissions for emission comparisons. The VMT variations for MOVES source type 53 in scenario 2-2 might be from other calculations in MOVES2014 after the VMT conversion, e.g., source bin (the combination of source type, fuel type, engine size, age, etc.) calculation and emission rate calculation.

Table 4. VMT variations by MOVES source type from MOVES runs in scenarios 1-2 and 2-2

		MOVES2014		MOVES2010b
MOVES Source Type ID	MOVES Source Type Name	Virgin Islands	Indiana	Virgin Islands	Indiana
11	Motorcycles	0.0%	0.2%	0.0%	0.0%
21	Passenger Cars	0.0%	0.2%	0.0%	0.0%
31	Passenger Trucks	0.0%	0.2%	0.0%	0.0%
32	Light Commercial Trucks	0.0%	0.2%	0.0%	0.0%
41	Intercity Buses
42	Transit Buses	0.0%		0.0%
43	School Buses	0.0%		0.0%
51	Refuse Trucks
52	Single Unit Short-haul Trucks	0.0%	0.2%	0.0%	0.0%
53	Single Unit Long-haul Trucks	−29.8%	−29.9%	0.0%	0.0%
54	Motor Homes
61	Combination Unit Short-haul Trucks	0.0%	0.2%	0.0%	0.0%
62	Combination Unit Long-haul Trucks	0.0%	0.2%	0.0%	0.0%

Emission burden results

The same four selected MOVES runs from scenarios 1-1 and 2-1 were rerun using adjusted relativeMAR. Those four MOVES reruns constitute scenario 1-3 and scenario 2-3.

Results from scenarios 1-1 and 1-3 were compared with results from baseline runs, i.e., scenario 1-2. Scenarios 2-1 and 2-3 were compared with scenario 2-2. CO and PM_2.5 emission burdens were selected for the comparisons.

Total emission burdens, including running exhaust emissions and nonrunning emissions, are used for the comparisons. Running exhaust emission is the function of VMT. And nonrunning emission, such as start emission, is the function of source population. Since source population is not changed for the same MOVES runs from different scenarios, nonrunning emission by MOVES source type will not be changed from scenario to scenario. Any change of total emission burden can be interpreted to be the only reflection of the change of running exhaust emission caused by VMT variations.

Table 5 shows the comparisons of VMT, CO, and PM_2.5 from MOVE2010b runs between using original and adjusted relativeMAR. It indicated that the total emission burdens of CO and PM_2.5 from MOVES2010b runs using original relativeMAR have been deviated from the baseline runs for all source types except for MOVES source types 11 and 21. The exception of source type 11 and 22 is because of the negligible variations of VMT for those two sources. Using adjusted relativeMAR, the variations of VMT, CO, and PM_2.5 from MOVES2010b runs are all reduced to 0.0%. This concludes that (1) the adjusted relativeMAR will not change the nonrunning emissions; (2) the adjusted relativeMAR will not change all other calculations in MOVES after the VMT conversion, e.g., emission rate calculation for the running exhaust; and (3) the adjusted relativeMAR is capable of redistributing the VMT from HPMS vehicle classes to MOVES source types, then adjusting the running exhaust emission burden.

Table 5. Comparisons of VMT, CO, and PM_2.5 from MOVES2010b runs between using original and adjusted relativeMAR

		VMT Variation		CO Variation		PM2.5 Variation
MOVES Source Type ID	MOVES Source Type Name	Original	Adjusted	Original	Adjusted	Original	Adjusted
11	Motorcycles	0.0%	0.0%	0.0%	0.0%	0.0%	0.0%
21	Passenger Cars	0.0%	0.0%	0.0%	0.0%	0.0%	0.0%
31	Passenger Trucks	−0.8%	0.0%	−0.3%	0.0%	−0.6%	0.0%
32	Light Commercial Trucks	2.2%	0.0%	0.9%	0.0%	1.9%	0.0%
41	Intercity Buses
42	Transit Buses
43	School Buses
51	Refuse Trucks
52	Single Unit Short-haul Trucks	−11.5%	0.0%	−5.1%	0.0%	−11.0%	0.0%
53	Single Unit Long-haul Trucks	7.2%	0.0%	2.9%	0.0%	6.9%	0.0%
54	Motor Homes
61	Combination Unit Short-haul Trucks	−12.6%	0.0%	−9.8%	0.0%	−12.6%	0.0%
62	Combination Unit Long-haul Trucks	3.5%	0.0%	3.3%	0.0%	3.6%	0.0%

Same comparisons are also conducted for MOVES2014 runs. The results, shown in Table 6, indicated that the deviations of VMT and total emission burden in MOVES2014 runs using original relativeMAR are bigger than in MOVES2010b runs using original relativeMAR except for MOVES source type 11. With the use of adjusted relativeMAR, those variations in MOVES2014 using original relativeMAR are decreased to 0.2% for VMT except for MOVES source type 53 and to 0.0% for total emission burden. For MOVES source type 11, variations of VMT, CO, and PM_2.5 in MOVES201014 runs both using original and adjusted relativeMAR are negligible or zero because of no VMT conversion from HPMS to MOVES source type. However, MOVES source type 21 shows significant variations for both VMT and total emission burden in MOVES2014 runs using original relativeMAR compared with all 0.0% variations in MOVES2010b runs using original relativeMAR. This is the result of the creation of new HPMS class 25 in MOVES2014 for MOVES source types 21, 31, and 32.

Table 6. Comparisons of VMT, CO, and PM_2.5 from MOVES2014 runs between using original and adjusted relativeMAR

		VMT Variation		CO Variation		PM2.5 Variation
MOVES Source Type ID	MOVES Source Type Name	Original	Adjusted	Original	Adjusted	Original	Adjusted
11	Motorcycles	0.2%	0.2%	0.0%	0.0%	0.0%	0.0%
21	Passenger Cars	−14.7%	0.2%	−8.7%	0.0%	−13.1%	0.0%
31	Passenger Trucks	34.5%	0.2%	18.4%	0.0%	28.7%	0.0%
32	Light Commercial Trucks	38.6%	0.2%	19.5%	0.0%	32.2%	0.0%
41	Intercity Buses
42	Transit Buses
43	School Buses
51	Refuse Trucks
52	Single Unit Short-haul Trucks	−34.8%	0.2%	−9.3%	0.0%	−33.5%	0.0%
53	Single Unit Long-haul Trucks	−14.4%	−29.9%	9.4%	0.0%	21.8%	0.0%
54	Motor Homes
61	Combination Unit Short-haul Trucks	−54.2%	0.2%	−42.2%	0.0%	−54.3%	0.0%
62	Combination Unit Long-haul Trucks	15.6%	0.2%	14.9%	0.0%	15.4%	0.0%

The one special case is the MOVES source type 53. As shown in Table 6, it remains −29.9% of VMT variation even though adjusted relativeMAR is used. However, it is same as the VMT variation (shown in Table 4) from baseline runs. Furthermore, variations of total emission burdens for this source type are decreased to 0.0% by using adjusted relativeMAR. This testified that (1) the adjusted relativeMAR did adjust the derived VMT to local VMT, and (2) the adjusted relativeMAR has no impact on the emission calculation. Table 6 also shows the opposite changes between VMT and pollutant emissions for source type 53. The possible reason for this opposite change might be from following calculations in MOVES2014 after VTM conversion, e.g., source bin and emission rate calculations.

Conclusion

The estimation of total emission burdens, including both running exhaust emission and nonrunning emission, using EPA MOVES is essential for SIP development and transportation conformity determination. The running exhaust emissions, affected predominantly by the VMT, are the significant part of the total emission burdens. Therefore, the VMT distribution has significant impact on the estimation of total emission burdens.

The VMT distribution from HPMS vehicle classes to MOVES source types in MOVES is determined by both fraction of MOVES source population and one set of parameters, relativeMAR. This may result in the deviation of MOVES-derived VMT from the local VMT for a specific scenario, i.e., specific county, state, and year. This deviation may be more significant for transportation conformity determination because only portion of region/county roadway network is included in the study area. Therefore, total emission burdens estimated by MOVES are impacted by VMT deviations.

This new approach of adjusting the default relativeMAR is capable of eliminating the deviation of VMT to 0.0% for MOVES2010b and to less than 0.2% for MOVES2014 except for source type 53 in MOVES2014 runs. It is successfully designed to only redistribute the VMT without affecting running emission rate calculation and nonrunning emission estimation. The variations of total emission burdens for both MOVES2010b and MOVES2014 are all eliminated to 0.0% for all MOVES source types. No modification is required for source population and HPMS VMT.

The new HPMS vehicle type 25, introduced into MOVES2014, created the deviations of VMT and total emission burden for MOVES source type 21 and increased both VMT and emission deviations for MOVES source types 31 and 32. The VMT deviation for MOVES source type 53 in MOVES2014 are not eliminated by this approach. This may be the result of other calculations in MOVES2014, e.g., source bin and emission rate calculations.

Additional information

Notes on contributors

Xiaobo Liu

Xiaobo Liu is a licensed professional engineer in environmental engineering, working as an environmental engineer in the New York City office of Parsons Brinckerhoff Inc.