Transport Energy Demands

As in the case of household and industry fuel consumption and carbon emission modeling, transport modeling of light and heavy vehicles has a large degree of uncertainty from real-world operating conditions. Travel demand models quantify traffic volumes through the interaction of transportation supply and demand. To distinguish between supply and demand, generally, travel demand is represented by traffic zones, which produce and attract trips, while the physical road networks represent the travel supply. Supply zones are defined by the population, socioeconomic status, number of households, average household size and employment ratio, among other factors. Transportation models, once they are set up, defined and producing relevant traffic information, must also be calibrated and properly analyzed to identify patterns and to improve the accuracy of the model. Transportation models are constantly changing with the introduction of new variables such as technology improvements and urban developments, and long-term adjustment of these models inevitably improves their accuracy; otherwise they may become redundant if not adapted to the changing environment.

Typically, for passenger vehicles, freight vehicles and aviation, factors such as:

  • • fleet configuration,
  • • operating conditions,
  • • vehicle type,
  • • road conditions or air traffic control, and
  • • driver profiles

are considered during modeling of transport emissions. Transport models provide quantitative information about travel demand and the performance of current and planned developments of transportation systems, providing valuable data and statistics to evaluate alternatives and employ informed decisions on energy-efficient eco-planning and eco-city transportation (Castiglione et al. 2015).

In the passenger and freight transport sector, Gouldson et al. (2015) also contribute significantly to awareness of reducing travel demands through urban planning and identify three main categories of energy savings in the transport sector. These are the shifting of transport modes for passenger transport, improving upon energy efficiency in the public transport sector and electrification of the public transport fleet. The associated considerations and variables of these three categories are presented in Fig. 5 (adapted from Gouldson et al. 2015).

In Fig. 5, the considerations of mode shifting, vehicle efficiency and electrification of fleets are listed. Common in the three categories is determining accurate

Passenger and freight transport sector efficiency improvement calculation datasets as proposed and used by Gouldson et al. (2015)

Fig. 5 Passenger and freight transport sector efficiency improvement calculation datasets as proposed and used by Gouldson et al. (2015)

statistics for urban populations by region, which also contributes to intercity travel demands and goods or service delivery among regions. Essentially, these considerations are used to determine the overall viability of upgrading or replacing existing transportation modes or technologies. The parameters can be used to plan and model feasibility studies in the passenger and freight transport sector.

Less obvious but equally relevant factors also play a role in efficiency modeling. Waygood and Avineri (2016) argue that the gender of the driver plays a significant role in the carbon emissions of vehicles. In this work several hypotheses are investigated regarding the differences between genders and their general concern about climate change. These concerns are based on education level, whether having children affects the attitude of people to climate change as well as how residents of developed and developing countries react to climate change. The results are detailed by Waygood and Avineri (2016); essentially, the results show that there is in fact correlation between socioeconomic characteristics and the concerns of individuals about climate change. These viewpoints can be linked to driver profiles, requiring complex and multivariate modeling techniques to model transport carbon emissions efficiently; again, these models are normally based on specific regions with certain de facto standards applied. In traffic-dense locations, low driving speed, driver frustration and requirements for air-conditioning typically also raise average fuel consumption significantly and lead to adaptations of transport emission modeling of carbon pollution.

Simplified representations of complex mechanisms to quantify the cause and effect of air pollution in the transportation sector require elaborate consideration of the problem. Air pollution from multiple sources, as is the case of pollution models from road vehicles and airplanes, typically entail combining

  • • microscopic (individual behavior needed to provide intelligence to the vehicle),
  • • mesoscopic (collective decision-making in groups or clusters of vehicles; physical, socio-cultural and communities), and
  • • macroscopic (the infrastructure of system-wide goals such as reducing congestion and efficiently using available land and roads, policies and governments)

sources along with data-backed estimations from trustworthy sources in combination with theoretical atmospheric dispersion models (Sanderson et al. 2012). The sources of emissions for on-road motor vehicles can be classified into two categories. The two categories are exhaust emissions, generated as by-products of the fuel combustion process, and evaporative emissions, which result from hydrocarbons emitted from the fuel. According to Delcan (2007), evaporative emissions are temperature-dependent, increase proportional to temperature and occur by four mechanisms, namely

  • • diurnal emissions, which evaporate from the fuel tank as the temperature rises,
  • • running loss emissions from heating of the engine,
  • • hot soak emissions from the heat of the engine when the vehicle is switched off, and
  • • refueling emissions from vapors escaping from the fuel tank during refueling.

Delcan (2007) provides a simplistic vehicle emission estimate through the product of vehicle activity data and the aggregate rate of emissions per unit of the activity; however, the simplistic estimation becomes complex depending on the required accuracy of the emissions evaluation. Activity data are categorized into macroscopic (kilometers traveled and average speed) and microscopic (real-time vehicle operation) activity. Emission factors are categorized by fleet, fuel, trip, environmental, regulatory and driver characteristics.

Road traffic air pollution can also be assessed by means of average emission estimations, as described by Marino et al. (2016). Such models necessitate simple representations to evaluate gaseous pollutants emitted by vehicles in urban environments. Vehicle emission models used to estimate emissions from sources contributing to air pollution, GHG and air toxins, especially in congested urban environments, available from the United States EPA online resource databases, include (these vehicle emission model descriptions are obtained primarily from the official online websites of each model)

  • • the motor vehicle emission simulator (MOVES) to estimate emissions from mobile sources at national, country and project level,
  • • the MOBILE vehicle emissions factor model to predict gram per mile emissions for hydrocarbons (HC), carbon monoxide (CO), oxides of nitrogen (NOx), CO2, PM and air toxins from cars, trucks and motorcycles under various conditions,
  • • assessment and reliability of transport emission models and inventory systems (ARTEMIS) for road, rail, air and ship transport at national and regional levels,
  • • the VERSIT+ emissions model to predict emission factors and energy use factors representative of vehicle fleets, differentiated for vehicle types and traffic situations, taking into account driving conditions,
  • • the computer program to calculate emissions from road transport (COPERT),
  • • the highway vehicle particulate emission modeling software to estimate PM emissions specifically from vehicles driving on highways, therefore at high speeds,
  • • the optimization model for reducing emissions of GHG from automobiles (OMEGA) to estimate technology costs for automobile manufacturers to achieve variable fleet-wide levels of vehicle GHG emissions,
  • • the mobile emissions assessment system for urban and regional evaluation (MEASURE) to improve predictions of CO, HC and NOx,
  • • the traffic emissions and energy (TEE) model, and
  • • the GHG emissions model (GEM), which estimates the GHG emissions and fuel-efficiency performance of heavy-duty vehicles.

These models, as provided by the United States EPA, are used in many cases for studies on global transport emissions, typically adapted for specific situations and presented as case studies by multiple sources. The models can also be divided into categories concerning average speed models on macroscopic level (COPERT, MOBILE, EMFAC), traffic situation models such as ARTEMIS, traffic variable models such as the TEE model, cycle-variable models concerning individual driving patterns such as the VERSIT+ and MEASURE modes and finally, modal models, a microscale approach to engine operation variables. Transport vehicle pollutant emissions are typically based on exhaust gas measurements of vehicles on test benches with varying driving cycles (Marino et al. 2016). Further travel demand traffic models with an emission component, which are given by Delcan (2007), include

  • • the systematic traffic evaluation and planning model (STEP),
  • • the transportation analysis simulation systems (TRANSIMS) model,
  • • SYNCHRO 7,
  • • CORSIM,
  • • PARAMICS, and
  • • VISSIM.

Furthermore, activity-based transportation models (Castiglione et al. 2015) are derived from the daily activities and routines of people in urban, suburban or agricultural environments. Activity-based transportation models are emerging as increasingly more important and applicable for modern eco-city planning and energy-efficient and low-carbon developments. These models predict which types of activities are most likely to be conducted at specific times, with regard to the demographics of the drivers for each activity (for example distinguishing between work and leisure), typical travel times for these activities and potential alternative travelling routes based on the demands of the activity—for example, shopping and leisure activities are more likely to have several alternative route options, whereas in comparison, work activities typically follow specific and unchanged routes for drivers. Activity-based transport models are somewhat similar to more traditional four-step models in the way that for each model,

  • • activity types are created,
  • • for these activities the destinations are defined,
  • • the types/modes of travel are identified, and
  • • routes are estimated.

The four-step transportation model is derived from the transportation system analysis (Manheim 1979), expanded by Florian et al. (1988) and revised by McNally (2007). Essentially, the four steps are:

  • • trip generation,
  • • trip distribution,
  • • mode choice, and
  • • trip assignment.

A detailed description of these four steps is provided by Teodorovic (2015) and provides insightful considerations concerning the four-step model. Activity-based transport models are primarily used and specifically appropriate for permitting a continuous distribution of value-of-travel-time-savings for populations and crucial for freeway pricing assessments. They give stronger accounting for all costs and utilities involved, not only for travelling as its own entity but also for common driver activities and transit routines. The activity-based transport model is essentially based on three theoretical methodologies of activity models, namely

  • • constraint-based models to assess the feasibility of an individual within particular space-time constraints,
  • • a stream of utility-maximizing models based on the idea that individual drivers aim to capitalize on their effectiveness when planning their daily schedule, and
  • • models that imitate judgement heuristics to avoid uncertainties of individuals, effectively maximizing the effectiveness of their daily schedule.

A report presented by Castiglione et al. (2015) provides a detailed discussion of activity-based models and serves as an excellent primer to activity-based models and the primary characteristics that make these models commonplace and essential in eco-city planning.

Regulatory simulation test procedures of real-world environments are required to authenticate and warrant compliance across various standards, specifically focusing on emissions of PM, NOx, HC and CO. Emissions are usually expressed in units of grams of pollutant per unit distance traveled, for example g/km or g/mile. In addition, if measurements are made based on engine dynamometer test cycles, emission units are expressed in grams of pollutant per unit of mechanical energy delivered by the engine with unit g/kWh or g/bhp-hr. High temporal and spatial resolution emissions from cycle-variable models using ubiquitous networks of sensors and nodes are required to simulate air quality in urban spaces on interpretable scales. To allow vehicle manufacturers to comply with emission standards, also known as tailpipe emission standards, government ministries such as the United States EPA certify equipment (such as internal combustion engines) if requirements are met. These requirements and standards are put in place to ensure regulations of country-specific occupational health and safety standards of ambient air quality subjected to tailpipe emissions. Regulation factors are typically sub-divided into categories for

  • • the vehicle type and/or size,
  • • its age in years and accumulated traveling distance,
  • • type of fuel used,
  • • average ambient weather conditions of its primary region, and
  • • the maintenance records and mechanical condition of the vehicle.

Zhang et al. (2016) employ a fuel-consumption and carbon-emission modeling methodology of distance-specific fuel consumption based on two classifications: localized type-approval fuel consumption estimated by model-year or for fleet average under a baseline cycle, which takes into account vehicle specifications, and correcting the average fuel consumption to real-world operating conditions. The modeling approach of the two classifications given by Zhang et al. (2016) are mathematically represented by

FCTA(ry) FCTA(originalr,y) X Csize(ry) (9)

where FCTA(r,y) is the localized type-approval fuel consumption for vehicles manufactured in region r and year y, FCTA(originai r y) is the average type-approval fuel consumption of the total number of vehicles in region r and year y. Csize(r,y) is the correction factor of vehicle sizes (Zhang et al. 2016). The mathematical representation for correcting for real-world operating conditions in the work of Zhang et al. (2016) is given as

FCreal(y) = FCTA(y) X SCF(v) X Clm + FCac (Ta) (10)

where FCreai(y) is the corrected real-world fuel consumption in year y, SCF is the speed (v) correction factor with the average speed of the vehicle fleet as the baseline, Clm is the correction factor of the loading mass from passenger occupancy and FCAC is additional fuel consumption due to air conditioning use as a function of ambient temperature Ta. Zhang et al. (2016) therefore provide an insightful investigation of environmentally specific transport modeling based on external factors such as driving speed, temperature control and average passenger occupancy in highly populated urban driving conditions.

Ehsani et al. (2016) construct a vehicle fuel consumption model for various types of vehicles and vehicle-specific features such as engine type (gasoline or diesel), road types and asphalt efficiency, a range of renewable and non-renewable fuel types used by transportation vehicles, driving style and overall fuel efficiency, as well as wind effects. The model of Ehsani et al. 2016 proposes a top-down approach of a mechanical model, which concerns losses of energy from the effects of gravity, acceleration (driving style), rolling resistance between the tires of the vehicle and the asphalt and aerodynamic resistance of the vehicle. The simplified relationship among these parameters is represented by Ehsani et al. (2016) as

where i represents the type of vehicle, j represents the engine type and k represents the fuel type, Ug is the energy consumed by losses due to gravity, Ui are losses from acceleration, Ur are the losses from rolling resistance, Ud are the aerodynamic losses and Uc are the cornering losses. The research conducted by Ehsani et al. (2016) is therefore specifically focused on mechanical vehicle attributes and all forces acting on the vehicle during movement; this is a predominantly physics-based approach combined with driver interaction and style.

The complexity and specificity of transport emissions modeling are somewhat relaxed in the agricultural sector, where a stricter and definite boundary can be drawn for activity data. Agricultural emissions modeling is, however, not a simple task, but allows more control over the input and output variables to a certain extent. Considerations of agricultural emissions modeling are presented in the following section.

< Prev   CONTENTS   Source   Next >