In the last few years, scientific consensus is that emission of greenhouse gases (GHGs) into the atmosphere is contributing to changes in the earth's climate. While uncertainty remains over the pace and dimensions of the change, a consensus on the need for action has grown among the public and elected officials. In part, this shift has been accelerated by concern over energy security and rising fuel prices. The new political landscape has led many cities, states, and regions to institute policies aimed at reducing GHG emissions. These policies and emerging initiatives have significant implications for the transportation planning process. The transportation sector accounts for approximately 27% of GHG production in the U.S. (as of 2003) and while the U.S. accounts for only roughly 5% of the world's population, it is estimated that it produces over 20% of the world's GHG emissions. Note that this does not include "lifecycle" emissions that result from the processes undertaken to extract, manufacture, and transport fuel. Carbon dioxide represents approximately 96% of the transportation sector's radiative forcing effects. Unlike conventional air pollutants, carbon dioxide emissions are directly tied to the amount of fuel consumed and its carbon intensity. Therefore, emissions reductions can be achieved by increasing the use of low-carbon fuels, improving fuel economy, or reducing total vehicle miles of travel - often called the three legged stool. (A fourth leg is congestion reduction, at certain optimal speeds). These same factors are related to our use of imported oil, so actions taken to reduce GHG emissions may actually produce benefits in both policy areas. The climatic risks of additional emissions associated with capacity projects must be balanced against the mobility, safety, and economic needs of a community or region. Consequently, this dissertation attempts to quantify the impacts of high-emitting vehicles on the environment and to propose solutions to enhance the currently-used high-emitting vehicle detection procedures. In addition, fuel consumption and emission models for high-speed vehicles are developed in order to provide more reliable estimates of vehicle emissions and study the impact of vehicle speeds on vehicle emissions. The dissertation extends the state-of-the-art analysis of high emitting vehicles (HEVs) by quantifying the network-wide environmental impact of HEVs. The literature reports that 7% to 12% of HEVs account for somewhere between 41% to 63% of the total CO emissions, and 10% are responsible for 47% to 65% of HC emissions, and 10% are responsible for 32% of NOx emissions. These studies, however, are based on spot measurements and do not necessarily reflect network-wide impacts. Consequently, the research presented in this dissertation extends the state-of-knowledge by quantifying HEV contributions on a network level. The study uses microscopic vehicle emission models (CMEM and VT-Micro model) along with pre-defined drive cycles (under the assumption that the composite HEV and VT-LDV3 represent HEVs and NEVs, respectively) in addition to the simulation of two transportation networks (freeway and arterial) to quantify the contributions of HEVs. The study demonstrates that HEVs are responsible for 67% to 87% of HC emissions, 51% to 78% of CO emissions, and 32% to 62% of the NOX emissions for HEV percentages ranging from 5% to 20%. Additionally, the traffic simulation results demonstrate that 10% of the HEVs are responsible for 50% to 66% of the I-81 HC and 59% to 78% of the Columbia Pike HC emissions, 35% to 67% of the I-81 CO and 38% to 69% of the Columbia Pike CO emissions, and 35% to 44% of the I-81 NOX and 35% to 60% of the Columbia Pike NOX emissions depending on the percentage of the normal-emitting LDTs to the total NEVs. HEV emission contributions to total HC and CO emissions appear to be consistent with what is reported in the literature. However, the contribution of NOX emissions is greater than what is reported in the literature. The study demonstrates that the contribution of HEVs to the total vehicle emissions is dependent on the type of roadway facility (arterials vs. highways), the background normal vehicle composition, and the composition of HEVs. Consequently, these results are network and roadway specific. Finally, considering that emission control technologies in new vehicles are advancing, the contribution of HEVs will increase given that the background emission contribution will decrease. Given that HEVs are responsible for a large portion of on-road vehicle emissions, the dissertation proposes solutions to the HEV screening procedures. First, a new approach is proposed for estimating vehicle mass emissions from concentration remote sensing emission measurements using the carbon balance equation in conjunction with either the VT-Micro or PERE fuel consumption rates for the enhancement of current state-of-the-art HEV screening procedures using RSD technology. The study demonstrates that the proposed approach produces reliable mass emission estimates for different vehicle types including sedans, station wagons, full size vans, mini vans, pickup trucks, and SUVs. Second, a procedure is proposed for constructing on-road RS emission standards sensitive to vehicle speed and acceleration levels. The proposed procedure is broadly divided into three sub-processes. In the first process, HE cut points in grams per second are developed as a function of a vehicle's speed and acceleration levels using the VT-Micro and CMEM emission models. Subsequently, the HE cut points in grams per second are converted to concentration emissions cut points in parts per million using the carbon balance equation. Finally, the scale factors are computed using either ASM ETW- and model-year-based standards or engine-displacement-based standards. Given the RS emissions standards, the study demonstrated that the use of on-road RS cut points sensitive to speed and acceleration levels is required in order to enhance the effectiveness of RS. Finally, the dissertation conducted a study to develop fuel consumption and emissions models for high-speed vehicles to overcome the shortcomings of state-of-practice models. The research effort gathered field data and developed models for the estimation of fuel consumption, CO₂, CO, NO, NO2, NOx, HC, and PM emissions at high speeds. A total of nine vehicles including three semi-trucks, three pick-up trucks, and three passenger cars were tested on a nine-mile test track in Pecos, Texas. The fuel consumption and emission rates were measured using two portable emission measurement systems. Models were developed using these data producing minimum errors for fuel consumption, CO₂, NO2, HC, and PM emissions. Alternatively, the NO and NOx emission models produced the highest errors with a least degree of correlation. Given the models, the study demonstrated that the newly constructed models overcome the shortcomings of the state-of-practice models and can be utilized to evaluate the environmental impacts of high speed driving. / Ph. D.

Since the transportation sector is highly responsible for U.S. fuel consumption and emissions, assessing the environmental impacts of transportation activities is essential for air-quality improvement programs. Also, high emitting vehicles need to be considered in the modeling of mobile-source emissions, because they contribute to a large portion of the total emissions, although they comprise a small portion of the vehicle fleet. In the context of this research, the thesis quantifies the environmental impacts of roadway grades and proposes a procedure that can enhance the screening of high emitting vehicles. First, the study quantifies the environmental impacts of roadway grades. Although roadway grades are known to affect vehicle fuel consumption and emission rates, there do not appear to be any systematic evaluations of these impacts in the literature. Consequently, this study addresses this void by offering a systematic analysis of the impact of roadway grades on vehicle fuel consumption and emission rates using the INTEGRATION microscopic traffic simulation software. The energy and emission impacts are quantified for various cruising speeds, under stop and go conditions, and for various traffic signal control scenarios. The study demonstrates that the impact of roadway grade is significant with increases in fuel consumption and emission rates in excess of 9% for a 1% increase in roadway grade. Consequently, a reduction in roadway grades in the range of 1% can offer savings that are equivalent to various forms of advanced traffic management systems. Second, the study proposes a new procedure for estimating vehicle mass emissions from remote sensing device measurements that can be used to enhance HEV screening procedures. Remote Sensing Devices (RSDs) are used as supplementary tools for screening high emitting vehicles (HEVs) in the U.S. in order to achieve the National Ambient Air Quality Standards (NAAQS). However, tailpipe emissions in grams cannot be directly measured using RSDs because they use a concentration-based technique. Therefore, converting a concentration measurement to mass emissions is needed. The research combines the carbon balance equation with fuel consumption estimates to make the conversion. In estimating vehicle fuel consumption rates, the VT-Micro model and a Vehicle Specific Power (VSP)-based model (the PERE model) are considered and compared. The results of the comparison demonstrate that the VSP-based model under-estimates fuel consumption at 79% and produces significant errors (R2 = 45%), while the VT-Micro model produces a minimum systematic error of 1% and a high degree of correlation (R2 = 87%) in estimating a sample vehicle's (1993 Honda Accord with a 2.4L engine) fuel consumption. The sample vehicle was correctly identified 100%, 97%, and 89% as a normal vehicle in terms of HC, CO, NOX emissions, respectively, using its in-laboratory measured emissions. Its estimated emissions yielded 100%, 97%, and 88% of correct detection rates in terms of HC, CO, NOX emissions, respectively. The study clearly demonstrates that the proposed procedure works well in converting concentration measurements to mass emissions and can be applicable in the screening of HEVs and normal emitting vehicles for several vehicle types such as sedans, station wagons, full-size vans, mini vans, pickup trucks, and SUVs. / Master of Science

No description available.

Ocorrem anualmente aproximadamente 600.000 mortes de crianças com até cinco anos, no mundo. Pneumonia é a principal causa e mais de 50 por cento destas mortes são atribuídas à poluição do ar. Ela ainda é responsável pelo aumento do risco de infecções respiratórias, asma, condições neonatais adversas e anomalias congênitas. A poluição do ar também afeta o desenvolvimento cognitivo de crianças e induz o desenvolvimento de doenças crônicas na idade adulta. Entre 70 e 80 por cento da poluição do ar em nações em desenvolvimento são de origem veicular. Objetivando definir critérios baseados em medições com sensoriamento remoto para identificação de veículos automotores leves do ciclo Otto com elevadas emissões de monóxido de carbono, hidrocarbonetos ou óxido nítrico, foram utilizados os dados secundários gerados pela Remote Sensing do Brasil Ltda dos quais foram selecionados 179.142 veículos em uso da frota circulante da cidade de São Paulo com medições completas dos índices de emissão dos poluentes monóxido de carbono (CO), hidrocarbonetos (HC) e óxido nítrico (NO) e ainda velocidade e aceleração do veículo quando da medição e inclinação da pista no local escolhido para as medições. Foram ajustados modelos estatísticos da classe Generalised Additive Models for Location, Scale and Shape (GAMLSS) visando testar a influência do Tipo de Combustível, da Potência Específica do Veículo (VSP) e das Fases do Programa de Controle da Poluição do Ar por Veículos Automotores (Proconve) sobre as emissões de CO, HC e NO, medidos usando o Remote Sensing Device (RSD). As emissões foram então conceitualmente subdivididas em dois grupos: veículos com emissões normais e com emissões anormais, isso para os diversos poluentes em veículos das Fases L3, L4 e L5 que são as fases de interesse para o gerenciamento da qualidade do ar. Variáveis latentes foram definidas para indicarem as distribuições dos veículos em relação a esses grupos e Fases. O algoritmo Expectation-Maximization (EM) foi empregado para identificação dos parâmetros das distribuições. Para determinação dos valores associados aos veículos com elevadas emissões de determinado poluente e fase do Proconve, foi empregado o percentil 98 por cento da distribuição ajustada para os veículos dos grupos com emissões normais. Assim sendo, o Erro de Tipo I foi fixado em 2 por cento sendo que esse percentual foi estabelecido considerando o Erro de Tipo II, de apontar o veículo como tendo emissão normal quando na realidade trata-se de um high emitter. Através desta abordagem foram determinados os valores indicativos de veículos com elevadas emissões segundo o poluente e a Fase do Proconve. Os resultados apontaram decréscimo nas emissões de CO e de HC segundo as Fases do Proconve. Para o NO, o comportamento das emissões não acompanhou as reduções impostas pelas Fases do Proconve. Foi constatado que os veículos de 2005 a 2009, movidos exclusivamente a gasool, foram os que apresentaram as maiores emissões de NO. Diversos possíveis fatores causadores deste comportamento diferenciado do NO foram discutidos neste trabalho. Os dados de qualidade do ar detectaram aumento significativo nas concentrações ambientais de Óxidos de Nitrogênio (NOx) em 2007, quando foi monitorado este parâmetro no período de inverno, o que pode indicar a influência dos high emitters, mas necessita de estudos mais aprofundados para confirmação da causa deste comportamento. / Approximately 600,000 deaths occur worldwide annually for children up to five years of age. Pneumonia is the leading cause and more than 50 per cent of these deaths are attributed to air pollution. It is still responsible for increased risk of respiratory infections, asthma, adverse neonatal conditions and congenital anomalies. Air pollution also affects the cognitive development of children and induces future development of chronic diseases in adulthood. In order to define criteria based on remote sensing measurements to identify Otto cycle light duty vehicles (LDV) with high emissions of carbon monoxide, hydrocarbons or nitric oxide it was used secondary data produced by Remote Sensing do Brasil Ltda, from which 179,142 inuse vehicles were selected, that belongs to the city of São Paulos current fleet. All those vehicles had complete measurements of emission of carbon monoxide (CO), hydrocarbons (HC) and nitric oxide (NO), and also speed and acceleration of the vehicle during measurements, and slope of the track at the place chosen for the measurements. Statistical models of the Generalized Additive Models for Location, Scale and Shape (GAMLSS) class were adjusted to test the influence of fuel type, Vehicle Specific Power (VSP) and of the Brazilian Vehicle Emission Control Program [Proconve] phases on CO, HC and NO emissions, measured using Remote Sensing Device (RSD). The emissions were then conceptually subdivided into two groups: vehicles with normal and abnormal emission, for the various pollutants in vehicles of L3, L4 and L5 phases of Proconve, which were of interest for the air quality management. Latent variables were defined to indicate the distribution of vehicles in relation to those groups and phases. The algorithm Expectation Maximization (EM) was employed to identify all parameters of the distributions. We use the 98 per cent percentiles of the statistical distribution set, for vehicles of groups with normal emissions to determine the limit values for vehicles with high emissions of pollutants and Proconve Phase. Therefore, the Type I Error was set at 2 per cent and this percentage was established considering the Type II Error to point the vehicle as having normal emission when in fact it is a high emitter. Through this approach, the indicative values of vehicles with high emissions according to the pollutant and the Proconve Phase were determined. Results of emissions measured with the RSD technique indicated a decrease in CO and HC emissions according to the Proconve Phase. For the NO, the emissions behavior did not follow the reductions imposed by the Proconve Phases. It was found that newer vehicles year model from 2005 to 2009 exclusively gasohol-powered vehicles, were the ones that presented the highest NO emissions. Several possible causative factors of this differential behavior of NO were discussed in this study. A significant increase in the environmental concentrations of Nitrogen Oxides (NOx) was detected in 2007, when this parameter was monitored in the winter period. This may indicate the influence of the high emitter vehicles, but it requires a more in-depth cause-effect study for confirmation of this behavior.

Ocorrem anualmente aproximadamente 600.000 mortes de crianças com até cinco anos, no mundo. Pneumonia é a principal causa e mais de 50 por cento destas mortes são atribuídas à poluição do ar. Ela ainda é responsável pelo aumento do risco de infecções respiratórias, asma, condições neonatais adversas e anomalias congênitas. A poluição do ar também afeta o desenvolvimento cognitivo de crianças e induz o desenvolvimento de doenças crônicas na idade adulta. Entre 70 e 80 por cento da poluição do ar em nações em desenvolvimento são de origem veicular. Objetivando definir critérios baseados em medições com sensoriamento remoto para identificação de veículos automotores leves do ciclo Otto com elevadas emissões de monóxido de carbono, hidrocarbonetos ou óxido nítrico, foram utilizados os dados secundários gerados pela Remote Sensing do Brasil Ltda dos quais foram selecionados 179.142 veículos em uso da frota circulante da cidade de São Paulo com medições completas dos índices de emissão dos poluentes monóxido de carbono (CO), hidrocarbonetos (HC) e óxido nítrico (NO) e ainda velocidade e aceleração do veículo quando da medição e inclinação da pista no local escolhido para as medições. Foram ajustados modelos estatísticos da classe Generalised Additive Models for Location, Scale and Shape (GAMLSS) visando testar a influência do Tipo de Combustível, da Potência Específica do Veículo (VSP) e das Fases do Programa de Controle da Poluição do Ar por Veículos Automotores (Proconve) sobre as emissões de CO, HC e NO, medidos usando o Remote Sensing Device (RSD). As emissões foram então conceitualmente subdivididas em dois grupos: veículos com emissões normais e com emissões anormais, isso para os diversos poluentes em veículos das Fases L3, L4 e L5 que são as fases de interesse para o gerenciamento da qualidade do ar. Variáveis latentes foram definidas para indicarem as distribuições dos veículos em relação a esses grupos e Fases. O algoritmo Expectation-Maximization (EM) foi empregado para identificação dos parâmetros das distribuições. Para determinação dos valores associados aos veículos com elevadas emissões de determinado poluente e fase do Proconve, foi empregado o percentil 98 por cento da distribuição ajustada para os veículos dos grupos com emissões normais. Assim sendo, o Erro de Tipo I foi fixado em 2 por cento sendo que esse percentual foi estabelecido considerando o Erro de Tipo II, de apontar o veículo como tendo emissão normal quando na realidade trata-se de um high emitter. Através desta abordagem foram determinados os valores indicativos de veículos com elevadas emissões segundo o poluente e a Fase do Proconve. Os resultados apontaram decréscimo nas emissões de CO e de HC segundo as Fases do Proconve. Para o NO, o comportamento das emissões não acompanhou as reduções impostas pelas Fases do Proconve. Foi constatado que os veículos de 2005 a 2009, movidos exclusivamente a gasool, foram os que apresentaram as maiores emissões de NO. Diversos possíveis fatores causadores deste comportamento diferenciado do NO foram discutidos neste trabalho. Os dados de qualidade do ar detectaram aumento significativo nas concentrações ambientais de Óxidos de Nitrogênio (NOx) em 2007, quando foi monitorado este parâmetro no período de inverno, o que pode indicar a influência dos high emitters, mas necessita de estudos mais aprofundados para confirmação da causa deste comportamento. / Approximately 600,000 deaths occur worldwide annually for children up to five years of age. Pneumonia is the leading cause and more than 50 per cent of these deaths are attributed to air pollution. It is still responsible for increased risk of respiratory infections, asthma, adverse neonatal conditions and congenital anomalies. Air pollution also affects the cognitive development of children and induces future development of chronic diseases in adulthood. In order to define criteria based on remote sensing measurements to identify Otto cycle light duty vehicles (LDV) with high emissions of carbon monoxide, hydrocarbons or nitric oxide it was used secondary data produced by Remote Sensing do Brasil Ltda, from which 179,142 inuse vehicles were selected, that belongs to the city of São Paulos current fleet. All those vehicles had complete measurements of emission of carbon monoxide (CO), hydrocarbons (HC) and nitric oxide (NO), and also speed and acceleration of the vehicle during measurements, and slope of the track at the place chosen for the measurements. Statistical models of the Generalized Additive Models for Location, Scale and Shape (GAMLSS) class were adjusted to test the influence of fuel type, Vehicle Specific Power (VSP) and of the Brazilian Vehicle Emission Control Program [Proconve] phases on CO, HC and NO emissions, measured using Remote Sensing Device (RSD). The emissions were then conceptually subdivided into two groups: vehicles with normal and abnormal emission, for the various pollutants in vehicles of L3, L4 and L5 phases of Proconve, which were of interest for the air quality management. Latent variables were defined to indicate the distribution of vehicles in relation to those groups and phases. The algorithm Expectation Maximization (EM) was employed to identify all parameters of the distributions. We use the 98 per cent percentiles of the statistical distribution set, for vehicles of groups with normal emissions to determine the limit values for vehicles with high emissions of pollutants and Proconve Phase. Therefore, the Type I Error was set at 2 per cent and this percentage was established considering the Type II Error to point the vehicle as having normal emission when in fact it is a high emitter. Through this approach, the indicative values of vehicles with high emissions according to the pollutant and the Proconve Phase were determined. Results of emissions measured with the RSD technique indicated a decrease in CO and HC emissions according to the Proconve Phase. For the NO, the emissions behavior did not follow the reductions imposed by the Proconve Phases. It was found that newer vehicles year model from 2005 to 2009 exclusively gasohol-powered vehicles, were the ones that presented the highest NO emissions. Several possible causative factors of this differential behavior of NO were discussed in this study. A significant increase in the environmental concentrations of Nitrogen Oxides (NOx) was detected in 2007, when this parameter was monitored in the winter period. This may indicate the influence of the high emitter vehicles, but it requires a more in-depth cause-effect study for confirmation of this behavior.