Search

ABSTRACTLight-duty gasoline vehicle (LDGV) tailpipe emission rates can be quantified based on pollutant concentrations measured using portable emission measurement systems (PEMS). Emission rates depend on exhaust flow. For simplified and micro-PEMS, exhaust flow is inferred from engine mass air flow (MAF) and air-to-fuel ratio. For many LDGVs, MAF is broadcast via the on-board diagnostic (OBD) interface. For some vehicles, only indirect indicators of MAF are broadcast. In such cases, MAF can be estimated using the speed-density method (SDM). The SDM requires an estimate of the engine volumetric efficiency (VE), which is the ratio of actual to theoretical MAF. VE is affected by intra-vehicle variability in the engine load and inter-vehicle variability in engine characteristics (e.g., the type of valvetrain). The suitability of SDM-based estimates of MAF in conjunction with simplified and micro-PEMS has not been adequately evaluated. Therefore, the objectives are to: (1) quantify VE accounting for intra- and inter-vehicle variability; and (2) evaluate the accuracy of SDM-based vehicle emission rate estimation approaches. Seventy-seven naturally-aspirated LDGVs were measured using PEMS. For each vehicle, VE was estimated using three approaches: (1) constant VE calibrated to actual fuel use; (2) average estimates of VE for Vehicle Specific Power modes imputed from OBD data; and (3) modeled VE using multilinear regression (MLR). The MLR models were developed based on engine load and engine characteristics. The best model was selected based on various statistical diagnostics. When engines were under load, variability in VE was most sensitive to variations in engine load. During idling, VE differed between engines depending on engine characteristics. The constant and modeled VE estimation approaches enable the accurate estimation of microscale and mesoscale emission rates, with errors typically within ±10% compared to values imputed from OBD data. Thus, accurate emission rates can be obtained from simplified and micro-PEMS.Implications: Simplified and micro portable emission measurement systems (PEMS) enable widespread measurement of vehicle exhaust emission. As a cost saving measure, they estimate exhaust flow indirectly rather than via measurement, typically based on engine mass air flow (MAF). For some vehicles, MAF is not reported by the on-board diagnostic (OBD) system but can be inferred from other reported variables and volumetric efficiency (VE). However, VE is typically proprietary. Methods demonstrated here for estimating VE enable accurate quantification of emission rates, thereby enabling use of these PEMS for policy-relevant applications such as technology assessments, trends analysis, and emissions inventories.

Spatially varying diesel locomotive fuel use and emission rates (FUERs) are needed to accurately quantify local emission hotspots and their health impacts. However, existing locomotive FUER data are typically not spatially resolved or representative of real-world locomotive operation. Therefore, existing data are of limited use in quantifying the spatial variability in real-world FUERs. The objectives of this work are to quantify spatial variability in locomotive FUERs and identify factors differentiating hotspots from non-hotspots. FUERs were measured based on real-world measurements conducted for the Piedmont passenger rail service using a portable emission measurement system. FUERs were quantified based on 0.25 mile track segments on the Piedmont route. Hotspots were defined as segments in the top quintile of segment-average FUERs. On average, hotspots contributed 40–50% to trip fuel use and emissions. Hotspots were typically associated with low-to-medium speed, and high acceleration and grade. In contrast, non-hotspots were associated with high speed, and low acceleration and grade. Hotspots were typically located near populated areas and, thus, may exacerbate air pollutant exposure. The method demonstrated here can be applied to other passenger train services to assess key trends in hotspot locations and factors that explain the occurrence of hotspots.

Spatial variability in real-world on-road tailpipe light-duty gasoline vehicle nitrogen oxides, hydrocarbon, carbon monoxide, and carbon dioxide emission rates, the locations of emissions hotspots, and factors that explain spatial variability are quantified. A sample of 205 vehicles were measured on four predefined round-trip study routes using Portable Emission Measurement Systems. The trips on each route were divided into segments, averaging 1/4 mile in length. Segment-average emission rates were estimated based on measured 1 Hz emission rates. Emission hotspots are defined as segments with ≥90th percentile of segment-average emission rates. The hotspots have average emission rates 2–4 times greater, depending on the pollutant, than other segments. Hotspots are of heterogeneous characteristics including road attributes and vehicle activity metrics. For example, some hotspots were on arterial roads with an upstream signalized intersection and positive road grade, whereas some hotspots were on interstates with positive grade. Vehicle activity metrics, including average vehicle specific power and relative positive acceleration, help identify the hotspots. To reliably identify a fleet-average hotspot, data are needed for at least 36–130 vehicles, depending on the pollutant.

A vehicle specific power (VSP) modal model and the MOtor Vehicle Emission Simulator (MOVES) Operating Mode (OpMode) model have been used to evaluate and quantify the fuel use and emission rates (FUERs) for on-road vehicles. These models bin second-by-second FUERs based on factors such as VSP, speed, and others. The validity of binning approaches depends on their precision and accuracy in predicting variability in cycle-average emission rates (CAERs). The objective is to quantify the precision and accuracy of the two modeling methods. Since 2008, North Carolina State University has used portable emission measurement systems to measure tailpipe emission rates for 214 light duty gasoline vehicles on 1,677 driving cycles, including 839 outbound cycles and 838 inbound cycles on the same routes. These vehicles represent a wide range of characteristics and emission standards. For each vehicle, the models were calibrated based on outbound cycles and were validated based on inbound cycles. The goodness-of-fit of the calibrated models was assessed using linear least squares regression without intercept between model-predicted versus empirical CAERs for individual vehicles. Based on model calibration and validation, the coefficients of determination (R2) typically range from 0.60 to 0.97 depending on the vehicle group and pollutant, indicating moderate to high precision, with precision typically higher for higher-emitting vehicle groups. The slopes of parity plots for each vehicle group and all vehicles typically range from 0.90 to 1.10, indicating good accuracy. The two modeling approaches are similar to each other at the microscopic and macroscopic levels.

International organizations such as the United Nations have begun to recognize the connection between sport events and air pollution. However, to date, there has been a dearth of research attempting to measure air pollution at sporting events. To address the lack of research on air pollution at sporting events, in this article we developed a methodology for measuring the air pollution at a college football game utilizing both stationary and mobile monitoring systems. Stationary monitors allowed for the creation of a microenvironment wherein we could examine the overall particulate matter exposure within the stadium and tailgate lot areas. Mobile monitors allowed for the examination of specific causes of spikes in air pollution. We found grills, generators, and cars caused pollutant spikes, which were, at times, over 20 times worse than the recognized levels for moderate air quality. Additionally, while the overall pollution in the microenvironment was elevated, pollutant levels during postgame (i. e., where there is a rush of individuals leaving at the same time) were consistently and considerably higher. Implications for practice and future research include event policy considerations and the application of the methodology across alternative sporting venues/contexts.

Passenger train energy consumption is dependent on speed trajectories. The variability of passenger train energy consumption owing to the variability in speed trajectories can help identify ways to reduce train energy use via improved operations. Empirical fuel use data from a portable measurement emission measurement system (PEMS) and empirical speed trajectories measured using a global positioning system (GPS) receiver were used to verify and quantify real-world energy consumption variability and the variability in empirical speed trajectories, respectively. To identify potential realistic speed trajectories that can lead to energy saving (i.e., eco-driving), a Markov chain based speed trajectory simulator was used to simulate inter-run variability in speed trajectories. An energy index model (EIM) was used to compare energy consumption among different speed trajectories. The results show inter-run variability in fuel use associated with inter-run variability in the empirical speed trajectories. There is also inter-segment variability in fuel use related to the segment length and grade. The Markov chain based speed trajectory simulator can simulate realistic inter-run variability in speed trajectories based on calibration using empirical speed trajectories. The number of empirical speed trajectories used for simulator calibration affects the range of simulated inter-run variability. The EIM provides an accurate estimation of the empirical fuel use. Eco-driving, such as reducing the peak speed, can reduce energy consumption without compromising travel time. The methodology shown in this study is not system-specific and can be applied to other passenger train systems.

With more stringent U.S. fuel economy (FE) standards, the effect of auxiliary devices such as air-conditioning (AC) have received increased attention. AC is the largest auxiliary engine load for light duty gasoline vehicles (LDGVs). However, there are few data regarding the effect of AC operation on FE for LDGVs based on real-world measurements, especially for recent model year vehicles. The Motor Vehicle Emission Simulator (MOVES) is a regulatory model for estimating on-road vehicle energy-use and emissions. MOVES adjusts vehicle energy-use rates for AC effects. However, MOVES-predicted FE with AC has not been evaluated based on empirical measurements. The research objectives are to quantify the LDGVs FE penalty from AC and assess the accuracy of MOVES2014a-predicted FE with AC. The AC effect on real-world fleet-average FE was quantified based on 78 AC-off vehicles versus 55 AC-on vehicles, measured with onboard instruments on defined study routes. MOVES2014a-based FE penalty from AC was evaluated based on real-world estimates and chassis dynamometer-based FE test results used for FE ratings. The real-world FE penalty ranges between 1.3% and 7.5% among a wide range of driving cycles. Fuel consumption at idle is 13% higher with AC on. MOVES underestimates the real-world FE with AC by 6%, on average. MOVES overestimates the AC effect on cycle-average FE ranging between 13.5% and 18.5% for real-world and MOVES default cycles, and between 11.1% and 14.5% for standard cycles.

Ambient PM2.5concentrations measured at fixed site monitors (FSM) are often biased with respect to exposure concentrations because of spatial variability and infiltration. Based on comparison of ambient concentrations from 14 FSMs and of exposure concentrations measured indoors and outdoors at two schools in Hong Kong for winter and summer seasons, the magnitude and sources of exposure error based on using FSMs as a surrogate for exposure are quantified. An approach for bias correcting surrogate exposure estimates from FSMs is demonstrated. The approach is based on a proximity factor (PF) that accounts for differences in spatial locations, proximity to emissions and deviation from dominant wind direction, and an infiltration factor (IF) that varies by season. The combination of the PF and IF reduce bias in mean school exposure estimates from ±90% to ±20%. Bias in exposure estimates from using FSMs as surrogates tend to be smaller for which the exposure site and FSM are aligned with wind direction, have similar sampling height, and are in close proximity. The methodology demonstrated to assess concordance between FSMs and exposure measurement sites can be applied more broadly to help reduce exposure error, which may help to interpret seasonal variations in health estimates.

Eco-driving involves alterations to driving style to improve energy efficiency. The observed driving style reflects the combined effects of roadway, traffic, driver, and vehicle performance. Although the effect of roadway and traffic characteristics can be inferred from microscale driving activity data, the effect of vehicle performance on driving style is not properly understood. This paper addresses two questions: (1) how different is an individual driver’s driving style when operating vehicles with differences in performance?; and (2) how dissimilar are the driving styles of different drivers when operating vehicles that have similar performance? To answer these questions, we have gathered microscale vehicle activity measurements from 17 controlled real-world driving schedules and two years of naturalistic driving data from five drivers. We also developed a metric for driving style termed “envelope deviation,” which is a distribution of gaps between microscale activity (1 Hz) and fleet average envelope. We found that there is significant inter-driver heterogeneity in driving styles when controlling for vehicle performance. However, no significant inter-vehicle heterogeneity was present in driving styles while controlling for the driver. Findings from this study imply that the choice of vehicle does not significantly alter the natural driving style of a driver.

ABSTRACTGlobally, 1.3 billion on-road vehicles consume 79 quadrillion BTU of energy, mostly gasoline and diesel fuels, emit 5.7 gigatonnes of CO2, and emit other pollutants to which approximately 200,000 annual premature deaths are attributed. Improved vehicle energy efficiency and emission controls have helped offset growth in vehicle activity. New technologies are diffusing into the vehicle fleet in response to fuel efficiency and emission standards. Empirical assessment of vehicle emissions is challenging because of myriad fuels and technologies, intervehicle variability, multiple emission processes, variability in operating conditions, and varying capabilities of measurement methods. Fuel economy and emissions regulations have been effective in reducing total emissions of key pollutants. Real-world fuel use and emissions are consistent with official values in the United States but not in Europe or countries that adopt European standards. Portable emission measurements systems, which uncovered a recent emissions cheating scandal, have a key role in regulatory programs to ensure conformity between “real driving emissions” and emission standards. The global vehicle fleet will experience tremendous growth, especially in Asia. Although existing data and modeling tools are useful, they are often based on convenience samples, small sample sizes, large variability, and unquantified uncertainty. Vehicles emit precursors to several important secondary pollutants, including ozone and secondary organic aerosols, which requires a multipollutant emissions and air quality management strategy. Gasoline and diesel are likely to persist as key energy sources to mid-century. Adoption of electric vehicles is not a panacea with regard to greenhouse gas emissions unless coupled with policies to change the power generation mix. Depending on how they are actually implemented and used, autonomous vehicles could lead to very large reductions or increases in energy consumption. Numerous other trends are addressed with regard to technology, emissions controls, vehicle operations, emission measurements, impacts on exposure, and impacts on public health.Implications: Without specific policies to the contrary, fossil fuels are likely to continue to be the major source of on-road vehicle energy consumption. Fuel economy and emission standards are generally effective in achieving reductions per unit of vehicle activity. However, the number of vehicles and miles traveled will increase. Total energy use and emissions depend on factors such as fuels, technologies, land use, demographics, economics, road design, vehicle operation, societal values, and others that affect demand for transportation, mode choice, energy use, and emissions. Thus, there are many opportunities to influence future trends in vehicle energy use and emissions.

ABSTRACTFlex fuel vehicles (FFVs) typically operate on gasoline or E85, an 85%/15% volume blend of ethanol and gasoline. Differences in FFV fuel use and tailpipe emission rates are quantified for E85 versus gasoline based on real-world measurements of five FFVs with a portable emissions measurement system (PEMS), supplemented chassis dynamometer data, and estimates from the Motor Vehicle Emission Simulator (MOVES) model. Because of inter-vehicle variability, an individual FFV may have higher nitrogen oxide (NOx) or carbon monoxide (CO) emission rates on E85 versus gasoline, even though average rates are lower. Based on PEMS data, the comparison of tailpipe emission rates for E85 versus gasoline is sensitive to vehicle-specific power (VSP). For example, although CO emission rates are lower for all VSP modes, they are proportionally lowest at higher VSP. Driving cycles with high power demand are more advantageous with respect to CO emissions, but less advantageous for NOx. Chassis dynamometer data are available for 121 FFVs at 50,000 useful life miles. Based on the dynamometer data, the average difference in tailpipe emissions for E85 versus gasoline is −23% for NOx, −30% for CO, and no significant difference for hydrocarbons (HC). To account for both the fuel cycle and tailpipe emissions from the vehicle, a life cycle inventory was conducted. Although tailpipe NOxemissions are lower for E85 versus gasoline for FFVs and thus benefit areas where the vehicles operate, the life cycle NOxemissions are higher because the NOxemissions generated during fuel production are higher. The fuel production emissions take place typically in rural areas. Although there are not significant differences in the total HC emissions, there are differences in HC speciation. The net effect of lower tailpipe NOxemissions and differences in HC speciation on ozone formation should be further evaluated.Implications: Reported comparisons of flex fuel vehicle (FFV) tailpipe emission rates for E85 versus gasoline have been inconsistent. To date, this is the most comprehensive evaluation of available and new data. The large range of inter-vehicle variability illustrates why prior studies based on small sample sizes led to apparently contradictory findings. E85 leads to significant reductions in tailpipe nitrogen oxide (NOx) and carbon monoxide (CO) emission rates compared with gasoline, indicating a potential benefit for ozone air quality management in NOx-limited areas. The comparison of FFV tailpipe emissions between E85 and gasoline is sensitive to power demand and driving cycles.

There are few data on differences in real-world emissions by in-use vehicles when they operate on freeway ramps compared with operations on the freeway itself. The objective of this paper is to quantify the variability in link-based emissions rates for on-ramps and off-ramps in comparison to rates on freeways. Real-world measurements were made with the use of a portable emissions measurement system (PEMS) for selected vehicles, ramps, and freeway segments. The methodology included development of a study design for field data collection of vehicle activity and emissions, execution of the study design, quality assurance of the raw data, and analysis of the quality-assured data. Four light-duty gasoline vehicles were driven on two routes, each composed of on-ramp, freeway, and off-ramp links. Data were collected for morning peak, evening peak, and off-peak time periods. A PEMS test was used to measure exhaust emissions of oxides of nitrogen (NOx), hydrocarbon (HC), and carbon monoxide (CO). The emissions rates for on-ramps were shown to be substantially higher than rates on freeways for NOx, HC, and CO. Some of this variability in emissions rates can be explained by link average vehicle specific power, which can vary by time of day and from one location to another. The variability in emissions rates by route and time of day indicates that there can be complex interactions between traffic flow, road geometry, and emissions rates. Recommendations are offered for additional study and regarding how these results can be used by researchers and practitioners.

Avoided emissions attributable to the reduction in personal automobile trips for passenger rail riders are quantified based on real-world measurements. The North Carolina Department of Transportation (NCDOT) sponsors the Piedmont passenger rail service between Raleigh and Charlotte, NC. Per passenger-kilometer locomotive emissions were quantified based on portable emissions measurement system measured exhaust concentrations and duty cycles, or the fraction of trip time spent in each throttle notch setting of the prime mover engine, from 68 one-way trips of six Tier 0+ and Tier 1+ locomotives, and actual ridership data. Motor Vehicle Emissions Simulator (MOVES) software was used to estimate light-duty gasoline vehicle (LDGV) emission factors. Moving a passenger from an LDGV to a Piedmont train would lead to a net reduction in carbon dioxide (CO2) and carbon monoxide (CO) emissions by 44–94 %, respectively, between Raleigh and Charlotte, based on the assumption that the driver is the only LDGV passenger. However, locomotive nitrogen oxides (NOx), hydrocarbons (HC), and particulate matter (PM) emission factors were 4–11 times higher than for the LDGV, respectively. Delays for either the train or highway vehicles did not substantially alter the key findings. If a Tier 4 locomotive was used, NOx, PM, and HC emission rates would be 90–99 % lower than current NCDOT locomotives. The use of real-world data representative of actual train operations provides an accurate basis for comparing rail and personal vehicle energy use and emissions and for identifying key factors affecting variability in the comparison.

On-road vehicles consume a substantial amount of fuel and emit air pollutants, including carbon monoxide, hydrocarbon, and nitrogen oxides. Increasing concern about air quality has meant increasing interest in accurate estimates of microscale vehicle energy use and emissions (EU&E). Vehicle EU&E rates are related to vehicle-specific power (VSP). For a given vehicle, VSP is a function of vehicle speed, acceleration, and grade, each of which can be categorized as an externally observable variable (EOV). However, internally observable variables (IOVs), such as manifold absolute pressure (MAP) and engine revolutions per minute (RPM), are more predictive of EU&E. Field measurements were conducted for 10 on-road vehicles using a portable emissions measurement system, an on-board diagnostic scan tool, and GPS receivers to measure exhaust concentrations and engine activity and to estimate road grade. VSP and the product of MAP and RPM (PM×R) are useful as indicators of engine load. IOV-based models were developed to predict EU&E rates. The IOV-based models typically performed better than EOV-based models. The models demonstrated could be integrated into next-generation traffic simulation programs, as well as into vehicle engine control units, to provide feedback on real-time emissions.

A study demonstrated a systematic method to assess the degree of concordance between estimated real-world and the U.S. Environmental Protection Agency (EPA) rated fuel economy for light-duty gasoline vehicles based on real-world emissions measurements. Cold engine start effects are accounted for in the real-world fuel economy and in comparison with EPA ratings. Portable emissions measurement systems were used to measure emissions of carbon dioxide, carbon monoxide, hydrocarbons, and nitrogen oxides of 122 vehicles on a specified test route consisting of various roads. Fuel economy of measured vehicles was estimated considering two cases: EPA standard driving cycles and real-world driving cycles. Average cold start emissions measurements from 32 vehicles were used in sensitivity analysis of fuel economy with respect to cold start. Real-world fuel economy estimates and comparison with rated fuel economy are not highly sensitive to cold start effects. On average, both the city and highway ratings are conservative compared with real-world fuel economy. However, there are proportions of vehicles that have lower real-world fuel economy than rated. The method demonstrated should be applied to additional vehicles and real-world driving cycles. Insights from these comparisons could motivate development of a more accurate rating scheme and identification of advice to consumers to improve their fuel economy.

Accurate estimation of vehicle activity is critically important for the accurate estimation of emissions. To provide a benchmark for estimation of vehicle speed trajectories such as those from traffic simulation models, this paper demonstrates a method for quantifying light-duty vehicle activity envelopes based on real-world activity data for 100 light-duty vehicles, including conventional passenger cars, passenger trucks, and hybrid electric vehicles. The vehicle activity envelope was quantified in the 95% frequency range of acceleration for each of 15 speed bins with intervals of 5 mph and a speed bin for greater than 75 mph. Potential factors affecting the activity envelope were evaluated; these factors included vehicle type, transmission type, road grade, engine displacement, engine horsepower, curb weight, and ratio of horsepower to curb weight. The activity envelope was wider for speeds ranging from 5 to 20 mph and narrowed as speed increased. The latter was consistent with a constraint on maximum achievable engine power demand. The envelope was weakly sensitive to factors such as type of vehicle, type of transmission, road grade, and engine horsepower. The effect of road grade on cycle average emissions rates was evaluated for selected real-word cycles. The measured activity envelope was compared with those of dynamometer driving cycles, such as the federal test procedure, highway fuel economy test, SC03, and US06 cycles. The effect of intervehicle variability on the activity envelope was minor; this factor implied that the envelope could be quantified based on a smaller vehicle sample than used for this study.

An empirically based macroscopic method is presented. It estimates and compares the pollutant emissions generated from signalized intersections and roundabouts. This method was built on a large sample size of real-world second-by-second vehicle trajectories, traffic volumes, and other traffic characteristics collected at signalized intersections and roundabouts in six U.S. states. The basis for predicting and estimating pollutant emissions was the concept of vehicle-specific power. The method enables inclusion of emissions standards and vehicle classes, such as Tier 1 (T1) and Tier 2 (T2) passenger cars (PCs) and passenger trucks (PTs). More than 1,980 vehicle trajectories were analyzed. Traffic variables including intersection capacity, demand-to-capacity ratio (d/c), cycle length, green-to-cycle length ratio, signal progression (i.e., arrival type), and number of lanes were included in the model for analysis and comparison between signals and roundabouts. Application of the method to a case study showed that on average under low d/c(<0.7), roundabouts generated lower emission rates than signalized intersections. As demand approached capacity under high traffic volumes, signalized intersections with favorable progression (i.e., most demand arrived during green phase) generally produced lower emission rates than roundabouts. Signalized intersections with poor progression (i.e., most demand arrived during red phase) generated more emissions than roundabouts. Results also showed that during oversaturation periods (when d/c> 1), the amount of produced emissions increased steadily in roundabouts but increased a large amount at signals.

The objectives of this study were to quantify real-world activity, fuel use, and emissions for heavy duty diesel roll-off refuse trucks; evaluate the contribution of duty cycles and emissions controls to variability in cycle average fuel use and emission rates; quantify the effect of vehicle weight on fuel use and emission rates; and compare empirical cycle average emission rates with the U.S. Environmental Protection Agency’s MOVES emission factor model predictions. Measurements were made at 1 Hz on six trucks of model years 2005 to 2012, using onboard systems. The trucks traveled 870 miles, had an average speed of 16 mph, and collected 165 tons of trash. The average fuel economy was 4.4 mpg, which is approximately twice previously reported values for residential trash collection trucks. On average, 50% of time is spent idling and about 58% of emissions occur in urban areas. Newer trucks with selective catalytic reduction and diesel particulate filter had NOxand PM cycle average emission rates that were 80% lower and 95% lower, respectively, compared to older trucks without. On average, the combined can and trash weight was about 55% of chassis weight. The marginal effect of vehicle weight on fuel use and emissions is highest at low loads and decreases as load increases. Among 36 cycle average rates (6 trucks × 6 cycles), MOVES-predicted values and estimates based on real-world data have similar relative trends. MOVES-predicted CO2emissions are similar to those of the real world, while NOxand PM emissions are, on average, 43% lower and 300% higher, respectively. The real-world data presented here can be used to estimate benefits of replacing old trucks with new trucks. Further, the data can be used to improve emission inventories and model predictions.Implications: In-use measurements of the real-world activity, fuel use, and emissions of heavy-duty diesel roll-off refuse trucks can be used to improve the accuracy of predictive models, such as MOVES, and emissions inventories. Further, the activity data from this study can be used to generate more representative duty cycles for more accurate chassis dynamometer testing. Comparisons of old and new model year diesel trucks are useful in analyzing the effect of fleet turnover. The analysis of effect of haul weight on fuel use can be used by fleet managers to optimize operations to reduce fuel cost.

Portable emissions measurement systems (PEMS) are useful for quantification of real-world vehicle activity, energy use, and emissions. However, there is no standard methodology for processing PEMS data; this can lead to errors in reported results. PEMS typically include tailpipe exhaust gas and particle analyzers, Global Positioning System (GPS) receivers, engine sensors, and electronic control unit (ECU) data loggers. The sensitivity of estimated emission rates to random errors in measurements is quantified. Methods are evaluated for identification and correction of improper synchronization of PEMS, ECU, and GPS data streams and for road grade estimation. Estimated fuel use and emission rates for light- and heavy-duty vehicles are sensitive to errors in intake manifold absolute pressure and engine revolutions per minute values and in indicators of air-to-fuel ratio including carbon dioxide and oxygen concentrations. Synchronization can be aided by maximizing the Pearson correlation coefficient between two indicator variables and confirming the result by matching concurrent increases in indicator variables. The effect of improper synchronization on estimated modal emission rates is quantified. Modal average emission rates based on vehicle-specific power (VSP) are more sensitive to improperly synchronized engine versus GPS data. Improperly synchronized data streams result in decreased variability between the lowest and highest modal average emission rates. Estimation of road grade from a linear least squares slope of elevation over a specified distance is demonstrated. VSP-based modal fuel use and pollutant emission rates are less sensitive to differences in road grade than to errors in synchronization.

Human exposure to fine particulate matter of less than 2.5 microns in aerodynamic diameter is causally linked to cardiovascular and pulmonary diseases. In-vehicle exposure may account for 10% to 20% of daily average exposure. However, exposure models are typically based on areawide air quality data that poorly predict in-vehicle concentration. A practical method is demonstrated for conducting field measurements to quantify the ratio of in-vehicle to outside vehicle concentration (I/O) for a wide range of conditions that affect intravehicle variability in exposure concentration. A field data collection study design is developed on the basis of sources of intravehicle variability in I/O that include ventilation air source, window status, fan setting, air-conditioning (AC) use, vehicle speed, road type, travel direction, and time of day. Three replicates of measurements were made for 16 combinations of these factors on 110 mi of roads comprising eight one-way routes between typical commuter origin–destination pairs. Two portable particle monitors recorded in-vehicle and near-vehicle ambient concentrations on 1-min averages for four particle size ranges. The comparability of the monitors was quantified. Near-vehicle concentrations varied with road type, time of day, and traffic conditions. The I/O ratio was approximately independent of near-vehicle concentration and varied with window status, source of ventilation air (fresh or recirculated), and for cases with recirculation and closed windows, fan setting, and AC use. The study design can be extended to additional vehicles to account for potential sources of inter-vehicle variability. Data collected here can be used to improve exposure simulation models.

For environmental emissions certification, locomotive prime mover engines undergo rigorous measurements with the use of federal reference methods (FRMs). However, there are nonregulatory reasons for engine emissions measurement, such as to compare relative differences in fuel use and emissions as a result of engine rebuild or between engines in a fleet. Portable emissions measurement systems (PEMSs) are widely used for highway vehicle and nonroad construction equipment but have had limited locomotive applications. The objectives are to (a) demonstrate an approach for quantifying locomotive emission rates with a PEMS during dynamometer and rail yard engine load tests, (b) compare the emissions measured in dynamometer versus rail yard load tests, and (c) assess the relative change in fuel use and emissions from engine rebuild. Measurements were conducted on 3,000-hp prime movers, including an EMD 16-645 for a GP40 and two EMD 12-710s for F59PHs. Fuel use and PEMS-based emission rates for nitric oxide, carbon monoxide, hydrocarbons, and particulate matter were compared between dynamometer and rail yard load tests and with data from previous literature. Fuel use and oxides of nitrogen (NOx) emission rates after engine rebuild were lower for the GP40 prime mover, and the fuel use and NOxemission rates for the F59PH rebuilt engines were lower than those of the rebuilt GP40 engine. PEMS is not a substitute for locomotive FRMs if compliance certification is needed but provides useful data for comparative assessment.

This paper reports on research that explored how multilane roundabouts located on urban corridors have affected traffic performance and pollutant emissions generated from vehicles. The research also compared the emissions of vehicles moving through the roundabouts as they used either the left or right entry lanes. The methodology can be generalized to measure the emissions of any multilane roundabout. The paper identified a representative speed profile for each speed trajectory type, no stop, one stop, and multiple stops, from field data collected at four multilane roundabouts in Aveiro, Portugal. The vehicle-specific power emissions methodology was employed to estimate the second-by-second emissions generated from a vehicle during different acceleration–deceleration cycles. Congestion-specific vehicle speed profiles for two-lane roundabout approaches were used to develop regression models to predict the percentage of vehicles that would experience different speed trajectory types in the roundabout. The analysis tested hypotheses about how differences in the following characteristics have affected the amount of emissions generated from vehicles in each lane: (a) the speed profiles in each lane (left versus right), (b) the conflicting flows for the left and right lanes, (c) the lane flow, and (d) the overall congestion levels. Under low congestion levels, vehicles in the right lane emitted more pollutant because they had on average higher speed and sharper acceleration and deceleration rates. For high congestion levels, given equal flow rates for the left and right lanes, vehicles in the left lane produced more emissions because vehicles in the left lane experienced longer stop-and-go cycles and had different speed profiles than did vehicles in the right lane.

The purpose of this study is to assess the robustness of relative comparisons of emission rates between fuels and technologies to differences in real-world duty cycles on the basis of in-use measurements of five heavy-duty diesel vehicles (HDDVs). Prior comparisons of biodiesel and ultra-low-sulfur diesel emissions, recent changes in emission standards applicable to HDDVs, and typical emission control technologies used in these vehicles are reviewed. The study methodology includes field measurements with a portable emission measurement system and related instruments and sensors for five HDDVs operated in normal service by professional drivers on multiple round-trip routes within North Carolina. Duty cycles and emission rates are quantified on the basis of the manifold absolute pressure, which is an indicator of engine load. Variability in engine load for each observed round trip is quantified on the basis of the cumulative distribution function of normalized manifold absolute pressure. The effect of variability in duty cycles on fuel-based emission rates for nitrogen oxides, carbon monoxide, hydrocarbons, and particulate matter is evaluated. Comparisons are made for emissions of three trucks operated on B20 biodiesel and ultra-low-sulfur diesel. Furthermore, comparisons of five trucks with model years ranging from 1999 to 2010 illustrate the impact of different emission standards and emission control technologies on real-world emission rates. A key finding is that relative comparisons pertaining to fuels and technologies are robust to variability in observed duty cycles.

It is difficult to assess the effects of engine idling on air pollutant emissions from highway construction equipment because of a lack of combined activity and emissions data. A methodology is presented for quantifying the impact of idling on National Ambient Air Quality Standards criteria pollutant emissions, including nitrogen oxides (NOx), carbon monoxide (CO), hydrocarbons (HC), and particulate matter (PM). The methodology and results are based on field data collected from 35 items of nonroad diesel construction equipment. Engine idle time was quantified in terms of equipment operational efficiency (?), which was defined as the ratio of nonidle time to total equipment use time (nonidle time plus idle time). With ? and the ratio of idle to nonidle emission rates (re) for each pollutant, the percentage increase in the total quantities of each pollutant emitted (NE) was calculated for each item of equipment for the observed values of ? and re. Results showed that as ? increased (or idle time decreased), NEdecreased. Mathematical models with ? as a predictor variable were developed to estimate values of NEfor each pollutant. Regression equations in the form of y = ax3+ bx2+ cx + d were developed for NOxand HC, with R2values of 97% and 87%, respectively. Regression equations in the form of y = mx + b were developed for CO and PM, with R2values of .50 and .65, respectively. Recommendations include using the models as performance measurement tools to assess idle restriction regulations and operational strategies of highway construction equipment.

The purpose of this study was to demonstrate a method for measuring passenger railroad locomotive emissions with the use of a portable emission measurement system (PEMS) based on rail yard load tests of three locomotives, including one GP40 and two F59PHIs. These locomotives have mechanically governed diesel prime mover engines (PMEs) with an approximately 3,000-hp output. Each locomotive has a head end power (HEP) engine that produces approximately 600 hp for generating electricity used in the passenger cars. The engine measurements were based on ultralow sulfur diesel fuel. Each engine was instrumented to measure manifold absolute pressure, engine revolutions per minute, intake air temperature, and exhaust concentrations of selected gases and particles. These data were used to quantify exhaust and fuel flow. The exhaust concentrations of nitric oxide, carbon monoxide (CO), carbon dioxide, hydrocarbons, and particulate matter were measured. The PMEs are operated at each of many throttle notch settings. For the HEP engines, three electrical loads were applied on the basis of power usage for one, two, and four passenger cars, respectively. More than 97% of the raw data survived a multistep quality assurance process. The data obtained from the PEMS for the main engines were found to be comparable on a fuel basis to data reported by others, particularly for oxides of nitrogen and CO. The key results from this work are the establishment of a simplified methodology for future tests and the development of baseline data.

Operational efficiency is a simple way of addressing nonproductive time of diesel-powered construction equipment that performs a construction activity and is defined as the ratio of nonidle activity time to total activity time including idle time. The effects of diesel equipment operational efficiency on total emissions for construction activities have not been fully established; thus a fundamental understanding of operational efficiency is needed to evaluate the effects on total pollutant emissions of diesel-powered construction equipment. This paper presents a methodology to assess the effects of equipment operational efficiency on the total pollutant emissions of construction equipment performing construction activities. This methodology estimates the percentage of additional pollutants emitted due to a reduced operational efficiency, based on the ratio of idle to nonidle emissions. Results for a case study of six bulldozers indicate that pollutant emissions, in addition to the minimum amount required to complete a specific quantity of work, increase as operational efficiency decreases. This conclusion allows the savings resulting from increased efficiency to be determined. Furthermore, additional pollutant emissions for a specified operational efficiency increase as the ratio of idle to nonidle emissions increases. The ratio of idle to nonidle emissions varies according to pollutant and is approximately 1:2 for hydrocarbons and carbon monoxide, 1:5 for nitrogen oxides, and 1:7 for particulate matter.

Fine particulate matter (PM2.5) exposure is associated with short-term and chronic respiratory effects. It is necessary to investigate human exposure to PM2.5to support the assessment of the association between exposure and adverse health effects. The methodology used in the current version of Stochastic Exposure and Dose Simulation Model for Particulate Matter for in-vehicle PM2.5concentration is reviewed. An alternative approach for estimating in-vehicle PM2.5concentration, based on the use of a dispersion model to estimate near-road PM2.5concentration and a mass balance model to estimate in-vehicle concentration, has been specified and applied. The objectives of this paper are to (a) demonstrate the application of the alternative approach to various scenarios, (b) analyze the sensitivity of the modeling results to the key inputs, and (c) evaluate this approach. Typical inputs for the alternative approach are reviewed. Sensitivities of these inputs to the modeling results and estimated human exposure have been analyzed. The in-vehicle PM2.5concentration was estimated to vary by more than a factor of two, depending on factors such as the cabin air exchange rate and filter efficiency of the ventilation system. Recommendations are made for field data that would be useful to better characterize variability in factors that affect in-vehicle PM2.5concentration.

Vehicle emission estimates with high spatial and temporal resolution are needed to estimate near-roadway air quality and human exposure to emissions. The U.S. Environmental Protection Agency's (EPA) MOBILE6 emission factor model is based on test cycles with transient speeds and durations on the order of 10 min. MOBILE6 does not report emission factors for average speeds higher than 65 mph. However, for near-roadway studies, emission factors are needed for short highway segments that represent only a few seconds of vehicle travel time at approximately constant speed and speeds greater than 65 mph. Constant-speed and high-speed correction factors for nitrogen oxides, hydrocarbons, carbon dioxide (CO2), and carbon monoxide were developed on the basis of 59,286 s of EPA dynamometer emissions data for diesel vehicles from which speed–acceleration modal models were derived. The constant speed correction factor is the ratio of the emission factor at constant speed to the emissions factor during transient cycles with an equivalent average speed. For example, the constant speed correction factor for CO2increases from 0.71 to 1 as speed increases from 31 to 78 mph. High-speed correction factors are based on the ratio of emission factors at a constant speed greater than 65 mph to emission factors at a constant speed of 65 mph. For example, at 80 mph, the high-speed correction factor is approximately 1.5 for CO2. Sensitivity analysis of emission estimates to these correction factors was conducted for speeds of 40 to 80 mph.

Limited field data are available for analyses of fuel use and emissions of nonroad diesel construction equipment. This paper summarizes the results of field research that used a portable emissions monitoring system to collect fuel use and emissions data from eight backhoes, six bulldozers, three excavators, four generators, six motor graders, three off-road trucks, one skid-steer loader, three track loaders, and five wheel loaders while they performed various duty cycles. These tests produced approximately 119 h of field data for petroleum diesel and approximately 48 h for B20 biodiesel. Engine attribute data including horsepower, displacement, model year, engine tier, and engine load were collected to determine these factors’ influence on fuel use rates and emission rates of nitrogen oxides, hydrocarbons, carbon monoxide, carbon dioxide, and opacity. Mass per time fuel use rates were developed for each item of equipment, as were mass per time and mass per fuel used emission rates for each pollutant. For petroleum diesel, fuel use and emission rates of each pollutant were found to increase with engine displacement, horsepower, and load and to decrease with model year and engine tier. The results were qualitatively similar for B20 biodiesel. Fuel-based emission rates were found to have less variability and less sensitivity to engine size and load than time-based emission rates. Where possible, development of emission inventories based on fuel consumed, rather than time of activity, is preferred.

AbstractLong-haul freight trucks typically idle for 2000 or more hours per year, motivating interest in reducing idle fuel use and emissions using auxiliary power units (APUs) and shore-power (SP). Fuel-use rates are estimated based on electronic control unit (ECU) data for truck engines and measurements for APU engines. Engine emission factors were measured using a portable emission measurement system. Indirect emissions from SP were based on average utility grid emission factors. Base engine fuel use and APU and SP electrical load were analyzed for 20 trucks monitored for more than 1 yr during 2.76 million mi of activity within 42 U.S. states. The average base engine fuel use varied from 0.46 to 0.65 gal/hr. The average APU fuel use varied from 0.24 to 0.41 gal/hr. Fuel-use rates are typically lowest in mild weather, highest in hot or cold weather, and depend on engine speed (revolutions per minute [RPM]). Compared with the base engine, APU fuel use and emissions of carbon dioxide (CO2) and sulfur dioxide (SO2) are lower by 36–47%. Oxides of nitrogen (NOx) emissions are lower by 80–90%. Reductions in particulate matter (PM), carbon monoxide (CO), and hydrocarbon emissions vary from approximately 10 to over 50%. SP leads to more substantial reductions, except for SO2. The actual achievable reductions will be lower because only a fraction of base engine usage will be replaced by APUs, SP, or both. Recommendations are made for reducing base engine fuel use and emissions, accounting for variability in fuel use and emissions reductions, and further work to quantify real-world avoided fuel use and emissions.

Field data for in-use fuel consumption and emission rates were collected for 15 nonroad vehicles by using a portable emission measurement system. Five backhoes, four front-end loaders, and six motor graders were tested once on petroleum diesel and once on B20 biodiesel. The vehicles represented a variety of engine certification tiers. A methodology was developed for study design, field data collection, data screening and quality assurance, data analysis, and benchmarking of the data. On average, 6.9% of data were lost because of quality issues and more than 3 h of valid data were collected in each test. Time-based emission factors increased monotonically with respect to engine manifold absolute pressure. Fuel-based emission factors were sensitive to differences between operations of engines idling and not idling. Typical duty cycles were quantified in terms of frequency distributions of manifold absolute pressure and used to estimate cycle average emission factors. On average, the use of B20 instead of petroleum diesel led to an insignificant 1.8% decrease in the nitric oxide (NO) emission rate and significant decreases of 18%, 26%, and 25% for opacity, hydrocarbons (HC), and carbon monoxide (CO), respectively. Emission rates decreased significantly in newer, higher-tier vehicles compared with older ones. Fuel use, NO, HC, and CO data were of similar magnitude as independent benchmark data. Specific recommendations were made for future work.

Heavy-duty diesel vehicles contribute a substantial fraction of nitrogen oxides (NOx) and particulate matter (PM) to the on-road vehicle emission inventory. The objectives of this study are to estimate roadway link-based emission rates for heavy-duty trucks for use in emission inventory estimation and to quantify the impact of factors affecting truck emissions. A speed–acceleration modal emission approach is developed from a database gathered via a portable emission measurement system for single rear-axle and tandem-axle dump trucks. Second-by-second real-world truck speed profiles on links are analyzed on the basis of observed patterns of time distributions of speed–acceleration modes. Link-based emission rates are estimated as the product of the fraction of time spent in each mode and the corresponding modal average emission rate. The sensitivity of link-based emission rates to key factors including chassis type, vehicle load, and fuel type is discussed. Single rear-axle trucks have lower emission rates than do tandem-axle trucks for carbon dioxide, PM, nitric oxide (NO), and hydrocarbons (HC) but higher carbon monoxide (CO) emission rates. Loaded trucks have higher fuel use and emissions than unloaded trucks. Replacing diesel fuel with biodiesel fuel for heavy-duty trucks may reduce tailpipe NO exhaust emissions and will reduce emissions of PM, CO, and HC. However, both fuels generate similar CO2emissions. Benchmark comparisons for link-based emission rates show that NO emission rates increase with mean speed. However, link-based CO and HC emission rates were not as sensitive to speed variation as NO emissions. The link-based emission rate approach is recommended to couple heavy-duty vehicle emission inventory estimation with transportation demand models.

AbstractIntegrated gasification combined cycle (IGCC) technology is a promising alternative for clean generation of power and coproduction of chemicals from coal and other feedstocks. Advanced concepts for IGCC systems that incorporate state-of-the-art gas turbine systems, however, are not commercially demonstrated. Therefore, there is uncertainty regarding the future commercial-scale performance, emissions, and cost of such technologies. The Frame 7F gas turbine represents current state-of-practice, whereas the Frame 7H is the most recently introduced advanced commercial gas turbine. The objective of this study was to evaluate the risks and potential payoffs of IGCC technology based on different gas turbine combined cycle designs. Models of entrained-flow gasifier-based IGCC systems with Frame 7F (IGCC-7F) and 7H gas turbine combined cycles (IGCC-7H) were developed in ASPEN Plus. An uncertainty analysis was conducted. Gasifier carbon conversion and project cost uncertainty are identified as the most important uncertain inputs with respect to system performance and cost. The uncertainties in the difference of the efficiencies and costs for the two systems are characterized. Despite uncertainty, the IGCC-7H system is robustly preferred to the IGCC-7F system. Advances in gas turbine design will improve the performance, emissions, and cost of IGCC systems. The implications of this study for decision-making regarding technology selection, research planning, and plant operation are discussed.

AbstractAlthough emission inventories are the foundation of air quality management and have supported substantial improvements in North American air quality, they have a number of shortcomings that can potentially lead to ineffective air quality management strategies. Major reductions in the largest emissions sources have made accurate inventories of previously minor sources much more important to the understanding and improvement of local air quality. Changes in manufacturing processes, industry types, vehicle technologies, and metropolitan infrastructure are occurring at an increasingly rapid pace, emphasizing the importance of inventories that reflect current conditions. New technologies for measuring source emissions and ambient pollutant concentrations, both at the point of emissions and from remote platforms, are providing novel approaches to collecting data for inventory developers. Advances in information technologies are allowing data to be shared more quickly, more easily, and processed and compared in novel ways that can speed the development of emission inventories. Approaches to improving quantitative measures of inventory uncertainty allow air quality management decisions to take into account the uncertainties associated with emissions estimates, providing more accurate projections of how well alternative strategies may work. This paper discusses applications of these technologies and techniques to improve the accuracy, timeliness, and completeness of emission inventories across North America and outlines a series of eight recommendations aimed at inventory developers and air quality management decision-makers to improve emission inventories and enable them to support effective air quality management decisions for the foreseeable future.

AbstractVehicle–specific power (VSP) is useful for explaining a substantial portion of variability in real–world vehicle emissions, such as those measured with portable emissions monitoring systems (PEMS). VSP is a function of vehicle speed, acceleration, and road grade. Road grade is shown to significantly affect estimates of both VSP and of real–world emissions via sensitivity analysis and analysis of empirical data. However, road grade is difficult to measure reliably using PEMS. Therefore, alternative methods for estimating road grade were identified and compared. A preferred method for estimating road grade was explored in more detail based on light detection and ranging (LIDAR) data. The method includes buffering LIDAR data onto roadway maps using a geographic information system tool, defining segments of roadway based on criteria pertaining to vertical curvature, quantification of roadway elevations within the buffered segments, and estimation of road grade and banking by fitting a plane to each segment. Factors influencing errors in road grade estimates are discussed. The method was evaluated by application to selected interstate highways and comparison to design drawing data. The development and application of LIDAR–based road grade data are demonstrated via a case study using PEMS data collected in the Research Triangle Park, NC, area. LIDAR data are shown to be reliable and accurate for road grade estimation for vehicle emissions modeling.

Estimating the emissions consequences of surface transportation operations is a complex process. Decision makers need to quantify the air quality impacts of transportation improvements aimed at reducing congestion on the surface street network. This often requires the coupling of transportation and emissions models in ways that are sometimes incompatible. For example, most macroscopic transportation demand and land use models, such as TransCAD, TranPlan, and TRANUS, produce average link speed and link vehicle miles traveled (VMT) by vehicle and road class. These values are subsequently used to estimate link-based emissions by using standard emissions models such as the U.S. Environmental Protection Agency's MOBILE6 model. In contrast, recent research with portable emissions monitoring systems indicates that emissions are not directly proportional to VMT but are episodic in nature, with high-emissions events coinciding with periods of high acceleration and speed. This research represents an attempt to bridge the gap in transportation and emissions models through the use of the real-world distributions of vehicle-specific power bins that are associated with average link speeds for various road classes. A successful effort in this direction would extend the use of transportation models to improve emissions estimation by using the limited output produced by such models. In addition, the variability of emissions and emissions rates over average speeds for a given facility type is explored, and recommendations are made to extend the methodology to additional facility types.

Diesel vehicles contribute substantially to the emissions of nitrogen oxides (NOx), an ozone precursor, and to particulate matter. Real-world, in-use, on-road emissions of selected diesel vehicles fueled with B20 biodiesel and petroleum diesel were measured during normal duty cycles by using a portable emissions measurement system. Four categories of dump trucks were tested: (a) trucks with a single rear axle and Tier 1 engines, (b) trucks with a single rear axle and Tier 2 engines, (c) tandem-rear-axle trucks with Tier 1 engines, and (d) tandem-rear-axle trucks with Tier 2 engines. Each vehicle was tested for 1 day while it was operating on B20 biodiesel fuel and for 1 day while it was operating on petroleum diesel fuel. On average, there were 4.5 duty cycles per day. Each duty cycle can be represented as a combination of nine operating modes (idle, three levels of acceleration, three levels of cruise, deceleration, and dumping). The average fuel flow and emissions rates on a mass-per-time basis varied substantially among the operating modes. The key factors responsible for the observed variability in fuel use and emissions included operating mode, vehicle size, engine type, vehicle weight, and fuel. Average fuel use and CO2emissions rates were approximately the same for the two fuels; but the average emissions rates of NO, CO, hydrocarbons, and particulate matter PM decreased by 10%, 11%, 22%, and 10%, respectively, for B20 biodiesel fuel compared with those for petroleum diesel fuel. The decrease in NO emissions observed was different from the results reported by others, based on engine dynamometer testing. The difference may be because of the effect of real-world duty cycles versus the standardized test procedures used in dynamometer testing, and further work is required to determine whether the ratio of NO to total NOx is affected by the duty cycle.

AbstractProbabilistic emission inventories were developed for 1,3-butadiene, mercury (Hg), arsenic (As), benzene, formaldehyde, and lead for Jacksonville, FL. To quantify inter-unit variability in empirical emission factor data, the Maximum Likelihood Estimation (MLE) method or the Method of Matching Moments was used to fit parametric distributions. For data sets that contain nondetected measurements, a method based upon MLE was used for parameter estimation. To quantify the uncertainty in urban air toxic emission factors, parametric bootstrap simulation and empirical bootstrap simulation were applied to uncensored and censored data, respectively. The probabilistic emission inventories were developed based on the product of the uncertainties in the emission factors and in the activity factors. The uncertainties in the urban air toxics emission inventories range from as small as –25 to +30% for Hg to as large as –83 to +243% for As. The key sources of uncertainty in the emission inventory for each toxic are identified based upon sensitivity analysis. Typically, uncertainty in the inventory of a given pollutant can be attributed primarily to a small number of source categories. Priorities for improving the inventories and for refining the probabilistic analysis are discussed.

AbstractQuantitative methods for characterizing variability and uncertainty were applied to case studies of oxides of nitrogen and total organic carbon emission factors for lean-burn natural gas-fueled internal combustion engines. Parametric probability distributions were fit to represent inter-engine variability in specific emission factors. Bootstrap simulation was used to quantify uncertainty in the fitted cumulative distribution function and in the mean emission factor. Some methodological challenges were encountered in analyzing the data. For example, in one instance, five data points were available, with each data point representing a different market share. Therefore, an approach was developed in which parametric distributions were fitted to population-weighted data. The uncertainty in mean emission factors ranges from as little as ~±10% to as much as -90 to 21+180%. The wide range of uncertainty in some emission factors emphasizes the importance of recognizing and accounting for uncertainty in emissions estimates. The skewness in some uncertainty estimates illustrates the importance of using numerical simulation approaches that do not impose restrictive symmetry assumptions on the confidence interval for the mean. In this paper, the quantitative method, the analysis results, and key findings are presented.

AbstractThe objectives of this paper are to (1) quantify variability in hourly utility oxides of nitrogen (NOx) emission factors, activity factors, and total emissions; (2) investigate the autocorrelation structure and evaluate cyclic effects at short and long scales of the time series of total hourly emissions; (3) compare emissions for the ozone (O3) season versus the entire year to identify seasonal differences, if any; and (4) evaluate interannual variability. Continuous emissions monitoring data were analyzed for 1995 and 1998 for 32 units from nine baseload power plants in the Charlotte, NC, airshed. Unit emissions have a strong 24-hr cycle attributable primarily to the capacity factor. Typical ranges of the coefficient of variation for emissions at a given hour of the day were from 0.2 to 0.45. Little difference was found when comparing weekend emissions with the entire week or when comparing the O3season with the entire year. There were substantial differences in the mean and standard deviation of emissions when comparing 1995 and 1998 data, indicative of the effect of retrofits of control technology during the intervening time. The wide range of variability and its autocorrelation should be accounted for when developing probabilistic utility emission inventories for analysis of near-term future episodes.

AbstractA study design procedure was developed and demonstrated for the deployment of portable onboard tailpipe emissions measurement systems for selected highway vehicles fueled by gasoline and E85 (a blend of 85% ethanol and 15% gasoline). Data collection, screening, processing, and analysis protocols were developed to assure data quality and to provide insights regarding quantification of real-world intravehicle variability in hot-stabilized emissions. Onboard systems provide representative real-world emissions measurements; however, onboard field studies are challenged by the observable but uncontrollable nature of traffic flow and ambient conditions. By characterizing intravehicle variability based on repeated data collection runs with the same driver/vehicle/route combinations, this study establishes the ability to develop stable modal emissions rates for idle, acceleration, cruise, and deceleration even in the face of uncontrollable external factors. For example, a consistent finding is that average emissions during acceleration are typically 5 times greater than during idle for hydrocarbons and carbon dioxide and 10 times greater for nitric oxide and carbon monoxide. A statistical method for comparing on-road emissions of different drivers is presented. Onboard data demonstrate the importance of accounting for the episodic nature of real-world emissions to help develop appropriate traffic and air quality management strategies.

The effect of arterial traffic signal timing and coordination on vehicle emissions is studied. Traffic signal timing improvement is one of the most common practices for congestion management in the United States. Although the benefits of improved signal timing for reduced fuel consumption are well documented, its effectiveness as a transportation control measure for emissions has not been clearly investigated. An empirical approach based on real-world, on-road vehicle emissions measurements was used. A total of 824 one-way runs representing 100 h and 2,020 vehicle miles of travel were conducted involving four drivers and eight gasoline-fueled light-duty vehicles on two signalized arterials in Cary, North Carolina: Walnut Street and Chapel Hill Road. Modal analyses of the data indicate that emissions rates were highest during acceleration and tend to decrease (in descending order) for cruise, deceleration, and idle. A modal approach is used to quantify the effect of arterial traffic signal timing and coordination on emissions. A key result is that signal coordination on Walnut Street yielded measurable improvements in arterial level of service and emissions reduction. For Chapel Hill Road, emissions were substantially lower under uncongested conditions [level of service (LOS) A/B] than under congested conditions (LOS D/E) for travel in the same direction at different times of day. Findings confirm the utility of signal coordination and congestion management as effective tools for controlling emissions.

AbstractThe quality of stationary source emission factors is typically described using data quality ratings, which provide no quantification of the precision of the emission factor for an average source, nor of the variability from one source to another within a category. Variability refers to actual differences caused by differences in feedstock composition, design, maintenance, and operation. Uncertainty refers to lack of knowledge regarding the true emissions. A general methodology for the quantification of variability and uncertainty in emission factors, activity factors, and emission inventories (EIs) is described, featuring the use of bootstrap simulation and related techniques. The methodology is demonstrated via a case study for a selected example of NOxemissions from coal-fired power plants. A prototype software tool was developed to implement the methodology. The range of interunit variability in selected activity and emission factors was shown to be as much as a factor of 4, and the range of uncertainty in mean emissions is shown to depend on the interunit variability and sample size. The uncertainty in the total inventory of −16 to +19% was attributed primarily to one technology group, suggesting priorities for collecting data and improving the inventory. The implications for decision-making are discussed.

AbstractVariability refers to real differences in emissions among multiple emission sources at any given time or over time for any individual emission source. Variability in emissions can be attributed to variation in fuel or feedstock composition, ambient temperature, design, maintenance, or operation. Uncertainty refers to lack of knowledge regarding the true value of emissions. Sources of uncertainty include small sample sizes, bias or imprecision in measurements, nonrepresentativeness, or lack of data. Quantitative methods for characterizing both variability and uncertainty are demonstrated and applied to case studies of emission factors for lawn and garden (L&G) equipment engines. Variability was quantified using empirical and parametric distributions. Bootstrap simulation was used to characterize confidence intervals for the fitted distributions. The 95% confidence intervals for the mean grams per brake horsepower/hour (g/hp-hr) emission factors for two-stroke engine total hydrocarbon (THC) and NOxemissions were from -30 to +41% and from -45 to +75%, respectively. The confidence intervals for four-stroke engines were from -33 to +46% for THCs and from -27 to +35% for NOx. These quantitative measures of uncertainty convey information regarding the quality of the emission factors and serve as a basis for calculation of uncertainty in emission inventories (Els).

Variability is the heterogeneity of values with respect to time, space, or a population. Variability may be quantified using frequency distributions. Uncertainty arises due to lack of knowledge regarding the true value of a quantity. Uncertainty may be quantified using probability distributions. These two concepts are distinct and, therefore, should be treated separately in an analysis. In this paper, methods for quantifying variability and uncertainty in model inputs, simulating variability and uncertainty in a model, and analyzing the results are presented. The method is demonstrated via an illustrative case study involving emissions characterization. The analysis of input data illustrates methods for characterizing variability and uncertainty when uncertainties arise due to sampling error (i.e., small sample sizes) or measurement error, as well as for dealing with differences in averaging times among multiple datasets. Correlation structures between sampling distributions representing uncertainty in frequency distributions were evaluated using bootstrap simulation. The frequency distributions for variability and the probability distributions for uncertainty were entered into a two-dimensional simulation framework in which the distinction between variability and uncertainty is maintained. The model output is a two-dimensional array of values which are analyzed to characterize the variability and uncertainty in the plant emission rate. The results are analyzed to identify the key sources of uncertainty and variability and to evaluate the likelihood that emissions will exceed limits from possible future regulations. The results are compared to one-dimensional probabilistic simulations. The benefits and limitations of two-dimensional simulation are evaluated. The applicability of the approach to other aspects of risk assessment is discussed.

Integrated environmental control (IEC) concepts involving combined SO2/NOx. removal processes in combination with pre-combustion and combustion control methods offer significant potential for more cost-effective emissions control for coal-fired power plants. IEC system designs involving three advanced flue gas treatment processes in conjunction with conventional and advanced coal cleaning technologies to achieve overall reductions of 90 percent for both SO2 and NOx are compared to a conventional pulverized coal plant design using currently available technology. A new computer modeling framework is used to characterize key uncertainties, and to evaluate overall performance and cost probabilistically using Monte Carlo methods. Results indicate that several IEC systems currently under development offer a good chance of reducing environmental control costs by 20 to 30 percent, although improvements in currently commercial technology are likely to offset some of the gains expected from advanced process designs. The use of probabilistic analysis for research management and utility planning are discussed in the context of the case studies presented.

Technological innovation in emissions control of acid rain precursors has made possible increasingly stringent control requirements for coal‐fired power plants. A key challenge for potential process adopters is evaluation of the uncertainties in performance and cost inherent in any new control technology. Uncertainties can be explicitly characterized using probabilistic modeling techniques such as Monte Carlo simulation. A robust approach to evaluating advanced systems is illustrated via a case study based on the fluidized bed copper oxide process. An engineering performance and cost model of this process was developed. Selected input parameters were assigned probability distributions based on data analysis and expert judgments. The model then was exercised in a probabilistic modeling environment. The modeling applications illustrate how uncertainty may be included in process evaluation. In particular, the likely cost savings and risks of a new technology compared to conventional technology can be estimated under varying design and uncertainty assumptions. The results can be used for decisions regarding process design trade‐offs, technology selection, and research planning in the face of uncertainty.