March 21-23, 2013

DOUBLETREE HOTEL
ANNAPOLIS , MARYLAND

Proceedings of the
54th Annual
Transportation Research Forum

www.trforum.org
MODELING FUEL CONSUMPTION AND EMISSIONS AT SIGNALIZED INTERSECTION
APPROACHES: A SYNTHESIS OF DATA SOURCES AND ANALYSIS TOOLS
Sherief Elbassuoni, University of Idaho
Ahmed Abdel-sRahim, University of Idaho

ABSTRACT

The continuous growth in demand of vehicular traffic has led to a large increase in fuel consumption and emissions,
negatively impacting the environment, public health, and the economy. The main goal of this paper is to synthesize and
document the state-of-the-practice in modeling fuel consumption and emissions at signalized intersection approaches. The
synthesis work presented in this paper focuses on two main areas: fuel consumption and emission modeling tools for
signalized intersection approaches and the sources of emissions inventory and data used in the models. The first part of this
paper includes a background covering different traffic-related pollutants, emission factors and different methods used to
obtain them and examples of the currently available emission inventories. In the second part, a review of currently available
fuel consumption and emission models suitable for modeling traffic operations at signalized intersection approaches is
presented covering three different analysis levels: microscopic, mesoscopic, and macroscopic.

1.0 INTRODUCTION

The main goal of this paper is to synthesize and document the state-of-the-practice in modeling fuel consumption and
emissions at signalized intersection approaches. It is divided into two main sections: the first focuses on sources of vehicle
emissions and fuel consumption data and the second addresses vehicle emissions and fuel consumption modeling capabilities
of different signalized intersection simulation modeling tool covering three analysis levels: microscopic, mesoscopic, and
macroscopic. For each of the models presented in the paper, details on how both the kinematics and kinetics components of
vehicle motion are being represented and modeled will be presented and discussed.

2.0 SOURCES OF VEHICLE EMISSIONS AND FUEL CONSUMPTION DATA

2.1 Emission Inventories

An Emission inventory is a “database that lists, by source, the amount of air pollutants discharged into the atmosphere
during a given time period from a certain activity.” It is an important component in the air quality management process and
is used to determine sources and quantities of different air pollutants and to build emission trends over time (“EPA - Air
Quality Management (AQM) - Emissions Inventory” 2012). Some of the methods of collecting vehicle emissions and fuel
consumption data to build emission inventories include: 1) continuous monitoring of actual emissions measurements, 2)
extrapolating the results from short-term source emissions tests, and 3) merging available activity levels data with associated
emission factors.
An emission factor is defined as a value that attempts to relate the quantity of a pollutant released to the atmosphere
with an activity associated with the release of that pollutant. Emission factors are typically expressed as the weight of a
specific pollutant divided by the unit weight, volume, distance, or duration of the activity emitting the pollutant. (Office of Air
Quality Planning and Standards 2012c).The general equation for emissions estimations is:

Equation 1

Where:
 E = emissions;
 A = activity rate;
 EF = emission factor, and
 ER =overall emission reduction efficiency, %
There are several emission factors’ databases. An example is the Emission Factors (EMFAC) database that includes
emission rates from different types of motor vehicles (passenger cars and heavy-duty trucks) operating on highways,
freeways and local roads in California. EMFAC-2011 is the most recent version of this model. Data from 25 million registered
vehicles in the 2009 vehicles registration data was used to update the populations in each vehicle class for 59 geographical
Page 156
areas. Travel activity data are provided by regional transportation planning agencies while Vehicle Miles of Travel (VMT) and
speed data are obtained from the southern California association of governments, Bay area Metropolitan transportation
commission, San Diego association of governments, and san Joaquin valley councils of government (California Environmental
Protection Agency - Air Resources Board 2011).
Another extensive source for emissions factors is the EPA's Air Pollutant (AP)‐42 Documentation. This documentation
was developed by the Emission Factor and Inventory Group (EFIG). It addresses ozone related pollutants (like total organic
compounds, oxides of nitrogen, and carbon monoxide), hazardous pollutants, and global warming gases. (Office of Air Quality
Planning and Standards 2012c), and (“Compilation of Air Pollutant Emission Factors | Modeling and Inventories | US EPA”
2013)
Examples of emission inventories and their available resources are:
 National Emission Inventory (NEI): NEI is a comprehensive and detailed estimate of air emissions and hazardous air
pollutants from all air emissions sources. It is developed and maintained by the U.S. Environmental Protection
Agency (EPA). NEI is updated every three years based on emission estimates and emission model inputs provided by
state, local and tribal air quality agencies form sources in their jurisdictions. The database is supplemented by data
developed by the US EPA. (“EPA - Air Quality Management (AQM) - Emissions Inventory” 2012), and (Office of Air
Quality Planning and Standards 2012a).
 EPA Clearinghouse for Inventories & Emission Factors: This database contains information on emissions inventories,
emissions factors, software and tools used for emissions inventories, and emissions modeling. All emission inventory
data, tools, and resources could be downloaded from this following source: (EPA 2012)
 Biogenic Emissions Inventory System (BEIS): The Biogenic emission sources are emissions that come from natural
sources, and need to be accounted for in photochemical grid models, as most types are widespread and ubiquitous
contributors to background air chemistry. Often only the emissions from vegetation and soils are included, but other
relevant sources include volcanic emissions, lightning, and sea salt. Biogenic emissions are typically computed using
a model which utilizes spatial information on vegetation and land use and environmental conditions of temperature
and solar radiation. The model inputs are typically horizontally allocated (gridded) data, and the outputs are gridded
biogenic emissions which can then be speciated and utilized as input to photochemical grid models. (Office of Air
Quality Planning and Standards 2012d).
 Emissions Modeling System for Hazardous Pollutants (EMS-HAP): The Emissions Modeling System for Hazardous
Pollutants (EMS-HAP) is an emissions processor that performs the steps needed to process an emission inventory for
input into the ASPENmodel or the ISCST3 model. EMS-HAP is written in the SAS programming language and is
designed to run on any UNIX workstation. The user will need a SAS license and some knowledge of SAS to use this
program. (“Related Programs | TTN - Support Center for Regulatory Atmospheric Modeling | US EPA” 2012)
 NONROAD Vehicle & Engine Emission Modeling: The NONROAD Model is intended for Windows 98 and later. Its
primary use is for estimation of air pollution inventories by professional mobile source modelers, such as state air
quality officials and consultants. NONROAD2008 updates NONROAD2005 to include new nonroad emission
standards promulgated in 2008 related to small gasoline engines and pleasure craft. (EPA - nonroad 2012)
 Sparse Matrix Operator Kernel Emissions (SMOKE): It is a Linux software supported by the “Center for
Environmental Modeling for Policy Development (CEMPD)”at the University of North Carolina at Chapel Hill. SMOKE
is an active open-source development project supported and distributed through the Community Modeling and
Analysis System Center. All required information could be obtained from this following source (SMOKE Version 3.1
2012).
 International Vehicle Emissions (IVE) Model: The International Vehicle Emissions (IVE) Model is a computer model
designed to estimate emissions from motor vehicles. The model is intended to help cities and regions develop
emissions estimates to: focus control strategies and transportation planning on those that are most effective;
predict how different strategies will effect local emissions; and measure progress in reducing emissions over time.
The model makes estimates of local air pollutants (criteria pollutants), greenhouse gas emissions, and toxic
pollutants. (The International Vehicle Emissions (IVE) Model 2012).
 Comprehensive Modal Emissions Model (CMEM): CMEM is a microscopic emission rate database. The model is
based on a total of 315 vehicle tests in the first phase and another 31 additional tests in subsequent phases (M. J.
Barth et al. 2000). CMEM uses a physical, power-demand modal modeling approach based on the vehicle physical
characteristics to generate emission tables. The vehicle emissions testing procedure was based on a second-by-
second performance with pre- and post-catalyst measurements of CO2, CO, HC, and NOx over three separate driving
cycles. These driving cycles are the full 3-bag FTP, EPA’s SFTP Bag 4 cycles (US06), and a newly designed modal test
cycle (MEC01) that focuses on specific modal events. The complete modal emissions model is composed of six

Page 157
 modules which are: 1) engine power demand; 2) engine speed; 3) fuel/air ratio; 4) fuel-rate; 5) engine-out emissions;
and 6) catalyst pass fraction (M. J. Barth et al. 2000).
 Emission Inventory Guidebook: This emission inventory is prepared by the United Nations European Environment
Agency (EEA), and the Task Force on Emissions Inventories and Projections (TFEIP). This guidebook provides a
comprehensive guide to state‐of‐the‐art atmospheric emissions inventory methodology. The guidebook also
supports the efforts of reporting under the United Nations Economic Commission for Europe (UNECE) Convention on
Long‐ Range Trans-boundary Air Pollution and the European Union (EU) directive on national emission ceilings.
(United Nations Economic Commission for Europe 2012), and (TFEIP 2012).

3.0 VEHICLE EMISSIONS AND FUEL CONSUMPTION MODELING CAPABILITIES OF SIMULATION MODELING TOOLS

In this part of the report, different simulation modeling tools are presented. Traffic simulation models could be divided,
based on the level of analysis into these three categories:(microscopic, mesoscopic, and macroscopic) (Yue 2008). In the
following sections, a brief summary of the characteristics of different models, as they relate to modeling fuel consumption
and vehicle emissions at signalized intersection approached will be presented and discussed. The reviews presented include
the model’s emission modeling capabilities, how vehicle kinematics and kinetics are being represented in the model, and how
it could be used for researches related to emissions and fuel consumption.

3.1 Microscopic Simulation Models

Microscopic simulation tools model the movement of individual vehicles on the basis of car-following and lane-changing
theories. They describe both the system entities and their interactions at a high level of detail. These models are also
designed for operations analyses of systems of road facilities (Lieberman and Rathi 1997), (Dowling, Holland, and Huang
2002), and (Alexiadis, Jeannotte,, and Chandra 2004). Using instantaneous speeds and acceleration of individual vehicles,
microscopic models can calculate fuel consumptions. These models are usually used to evaluate individual transportation
projects. There are several microscopic simulation models currently available. Examples of these models include: Advanced
Interactive Microscopic Simulator for Urban and Non-Urban Networks (AIMSUN2), Corridor Simulation (CORSIM),
INTEGRATION, MICSTRAN (Microscopic Simulator Model for Traffic Networks (MicroSim), Microscopic Traffic Simulator
(MITSIM), Probabilistic Adaptive Simulation Model (PADSIM), PARAMICS, SimTraffic, VISSIM, and Wide Area Traffic
Simulation (WATSim).

3.1.1 VISSIM
VISSIM is a stochastic, microscopic, time step, and behavior based traffic simulator. VISSIM’s traffic model is based
mainly on the psychophysical driver behavior model of R. Wiedemann, in which the driver’s reactions are in response to the
relative speed and distance of the preceding vehicle. VISSIM can simulate multi-modal traffic flows including passenger cars,
buses, light rail, trucks, pedestrians, and others. VISSIM was originally developed at the University of Karlsruhe, Germany in
the early 1970s, while the commercial distribution, development, and maintenance was started in 1993 by PTV Transworld
AG and is still done by them today. (L. Bloomberg and Dale 2000), and (Gomes, May, and Horowitz 2004).
VISSIM is formed mainly from two main parts; the first one is the simulator generator part, in which the user can import
aerial photos and schematic drawings and can begin to graphically build the simulation network. The second main part is the
signal state generator (SSG), in which the logic of the signal control exists, and in which the user can define lots of different
signal operations like fixed time, and ramp metering. (L. Bloomberg and Dale 2000), and (Gomes, May, and Horowitz 2004).
VISSIM was used in several of the research work related to the emissions.This paper is just an example for that. (Umedu,
Togashi, and Higashino 2012) proposed a real-time signal control method to minimize CO2 emissions. This method integrates
arrival times from information that consists of position and speed from vehicle to vehicle and vehicle to infrastructure
communications. The paper speculates that the more signals that are controlled together the more efficient control can be
and thus the authors have decided to use this speculation in their system. The process of analyzing emission was use of
mathematic calculation of exhaust output and VISSIM simulation to provide travel times of a corridor.
VISSIM is supporting an add-on called EnViVer. Using the vehicle record data, EnViVer can calculate different types of
emissions like CO2, NOx, and PM10. The used emission model in EnViVer is the microscopic exhaust gas/emission model
VERSIT+ by TNO which is based on emission measurements for about 2,800 vehicles under different driving conditions. The
main idea of EnViver is based on importing the PTV VISSIM record files and then calculating the emissions at spatial detail,
and finally getting the output tabulated or graphed (PTV Group 2012).

Page 158
3.1.2 CORSIM
CORSIM is a microscopic simulation program developed by the Federal Highway Administration (FHWA), and part of the
Traffic Software Integrated System (TSIS) which offers a windows-based interface for running the model. CORSIM mainly
includes two traffic simulation predecessor models, NETSIM and FRESIM. NETSIM is concerned with arterials with at grade
intersections (urban street traffic); while FRESIM is modeling uninterrupted facilities including interstate freeways and grade
separated expressways (freeway traffic). CORSIM executes the simulation and gets back the network efficiency to the user in
predetermined Measures Of Effectiveness (MOEs) .(Minnesota Department of Transportation 2008), (Loren Bloomberg and
Dale 2000) , (Park, Yun, and Choi 2004) , (Hall, Darter, and Rexroad 1993).
Unless CORSIM is not able to optimize phase plans, with combining NETSIM and FRESIM models, CORSIM is capable of
simulating large set of systems including freeways, urban streets, corridor or networks, different intersection control (e.g.
pre-timed signals and actuated ones), almost any surface geometry including turn pockets and number of lanes, weaving
sections, work-zones, and bus operations. (Minnesota Department of Transportation 2008), (Loren Bloomberg and Dale
2000), (Park, Yun, and Choi 2004), (Hall, Darter, and Rexroad 1993), and (Peter Holm et al. 2007).
CORSIM is mainly based on link-node structure network model, where links characterize roadway segments, and nodes
represent intersections, entry and exit points. (Minnesota Department of Transportation 2008), and (Park, Yun, and Choi
2004). CORSIM is a stochastic simulation model, which means that it utilizes random processes to model drivers’ behavior,
vehicles characteristics, and interactions into each run. This will lead to the conclusion that many runs may be required to get
a true picture for the network, and drawing conclusions based on only one run might be not representative. (Hall, Darter, and
Rexroad 1993), (Minnesota Department of Transportation 2008), and (Peter Holm et al. 2007). CORSIM creates tabulated
data for fuel consumption and environmental emissions. It can also tabulate and graph the effect of traffic different control
strategies on acceleration and fuel consumption. Because CORSIM’s data was not updated for many years, FHWA
recommends using the data for comparison analysis only, and not as absolute indications (FHWA 2010).
The newer versions of CORSIM have adopted the Vehicle Transient Emissions Simulation Software (VeTESS) for the
process of calculating fuel consumption and emissions. VeTESS was developed as a vehicle level tool for the simulation of fuel
consumption and emissions for real traffic transient vehicle operation within the EU 5th framework project DECADE (2001-
2003). VeTESS is capable of calculating emissions and fuel consumption made by a single vehicle during a defined ‘drive-
cycle’. For a given driving cycle, VeTESS uses simple mathematical calculations to determine the engine’s operating conditions
from the force on the vehicle. These calculations involve gear ratios and their efficiencies. Using the equation of motion,
Equation 2, VeTESS calculates the total force on the vehicle.

Ftotal = Faccel + Fgrad + Froll + Faero Equation 2

Where:
 Ftotal = The total force acting on the vehicle
 Faccel = The force required in order to cause an acceleration of the mass of the vehicle
 Fgrad = The component of the weight force of the vehicle acting parallel to the slope
 Froll = The rolling resistance
 Faero = The aerodynamic resistance
The force required to overcome the motion resistances is provided by the engine as a torque. After that, driven wheels
convert this torque from rotational to linear motion. VeTESS then evaluates the engine speed and engine torque from the
forces acting on the vehicle and references after that the corresponding values for the emission components using emission
maps (Beckx et al. 2007).
CORSIM was used in many research work related to emissions. This is one example for this type of research. (Kosman et
al. 2003) evaluated both CORSIM and VISSIM for project-level emission modeling. This studied project-levels included the
impacts of traffic flow improvements resulting from changes in signal timing or other roadway improvements. CORSIM was
compared to VISSIM for two different scenarios. In the first scenario, outputs from the two models were compared to spot
speed and average speed data collected from the field. In the second scenario, the models’ predicted emission reductions
were compared to three Congestion Mitigation and Air Quality (CMAQ) projects. Regarding the results of the first scenario, it
was found that both CORSIM and VISSIM under-predicted the mid-block spot speed, with a mixed results for average speed.
Overall predicted emissions (VOC and NOx) in the first scenario by results from CORSIM, VISSIM, and the field studies were
less than 7%. In the second scenario, it was found that CORSIM predictions differences in pre- and post-project speeds are
greater than VISSIM predictions for all three projects. This resulted in causing greater reductions in emissions.

3.1.3 CMEM

Page 159
 The Comprehensive Modal Emissions Model (CMEM) is published by the Center for Environmental Research and
Technology (CE-CERT) in Riverside California. Development began in the late 1990’s with support from the Cooperative
Highway Research council, and continued until the final phase was completed in 2005. CE-CERT has also received some
support from the Environmental Protection Agency (EPA). CMEM is not a traffic simulator, which means that all vehicle
activity data must come from another source. Because of this, microscopic traffic simulation software is often used to
generate traffic network and activity data, and input to CMEM to calculate emissions. Several interface applications used to
link CMEM with microscopic simulation software have been developed and used with success (Barth et al. 2005). CMEM is an
open source, and is available free for download.
CMEM is based on a deterministic physical power demand model, with rates dependent on causal variables such as fuel
delivery system, inspection maintenance effects, and vehicle age. Second-by-second vehicle tailpipe emissions are modeled
as the product of three components: fuel rate (FR), engine-out emission indices (gemission/gfuel), and time dependent catalyst
pass fraction (Barth et al. 2000). Because the rates contained in CMEM are based on physical parameters, in theory it could
be adapted to represent new vehicle technologies. However, the most recent major changes the CMEM emission rate
database took place in 2005, which indicates that the emissions data contained in this software is somewhat outdated.
Because it was desired that CMEM data properly incorporate the physical variables that contribute to emission rates,
existing emission inventories were deemed too generalized by vehicle category and not detailed enough for this purpose.
Instead, emission rates were developed “in-house” from tests conducted on several hundred recruited vehicles at the
University of California Riverside Center for Environmental Research and Technology. CMEM is basically based on a simple
parameterized physical approach. It has six modules that predict engine power, engine speed, air/fuel ratio, fuel use, engine-
out emissions, and catalyst pass fraction. All emission data used in CMEM was collected in the USA, mostly at the Riverside
facility. For the most part, emission data was collected on dynamometer tests, using a number of different drive cycles with
the intent to capture the full range of operation in the resulting emission rates. Much of the data used in CMEM was
collected before 1998, with additional low emitting vehicle data collected in 2002-2003 (M. Barth et al. 2006). The input
operating variables in CMEM model is having some variables. These variables are (acceleration, air/fuel equivalence ratio,
and fuel rate), second-by-second speed, road grade angle, and accessory use (such as air conditioning). the main output for
this model is the instantaneous emission. This instantaneous emission could be calculated as the product of three
components: fuel rate, mass of engine-out emissions per grams of fuel consumed, and catalyst pass fraction which is the ratio
of tailpipe emission to engine-out emission (An et al. 1997). The instantaneous emissions equation is shown in Equation 3.

Equation 3

Where:
FR = fuel-use rate in grams/s;
gemissions/gfuel = grams of engine-out emissions per grams of fuel consumed; and
CPF = the catalyst pass fraction, defined as the ratio of tailpipe to engine-out emission.
The total tractive power requirements (in kW) placed on the vehicle (at the wheels) is shown in Equation 4:

Ptract A • v  B • v2 C • V 3 M • a  M • g • v • sin 

Equation 4

Where:
M = the vehicle mass with appropriate inertial correction for rotating and reciprocating parts (kg),
v = speed (m/sec),
2
a = acceleration (m/s ),
2
g = the gravitational constant (9.81 m/s ), and
= the road grade angle.
P = the engine power output,
ηtf = the combined efficiency of the transmission and final drive,
Pacc = the engine power demand associated with the operation of vehicle accessories,

The fuel rate in any driving cycle for any vehicle model could be calculated using Equation 5.
Page 160
 Equation 5

where,
k = the engine friction factor,
N = engine speed (revolutions per second),
V = engine displacement (liter),
η ≈ 0.4 = a measure of indicated efficiency

and finally the engine-out emission module could be calculated using Equation 6.

Equation 6
if>0 and 0 otherwise, for ф<1.05
if>0 and 0 otherwise, for ф≥1.05

Where:
C0, aCO, aHC, and ϒHC are calibrated constant coefficients that are slightly different from vehicle to vehicle.
a1NOX and a2NOX are engine-out NOX emission indexes in grams of emissions per gram of fuel use under
stoichiometric and enrichment conditions, respectively, and
FRNOX is fuel rate thresholds

One of the examples of using CMEM in the emissions research work is this following paper. (Xia et al. 2012) discussed the
use of the Eco-Signal Operations approach technology. The ECO-approach technology means a traffic signal that broadcasts
its signal phase and timing, and Geometric Intersection Description to the vehicle. An onboard system takes this information
along with vehicle position and speed, and provides speed recommendations to the driver. This speed would be an emissions
reducing optimal speed. Field tests were conducted using a test vehicle equipped with the optimal speed algorithm. Based on
the vehicle speed trajectories during the test, vehicle fuel consumption is calculated using CMEM. During the two tested
scenarios, vehicle emissions were reduced compared to the control, an uninformed driver.

3.1.4 Integration
INTEGRATION is a trip-based microscopic traffic and emissions simulation model. It is developed in its current version
under the direction of Dr. Hesham Rahka at Virginia Tech. The software includes a traffic assignment tool, in which
origin/destination (OD) matrices are entered along with departure time series histograms for each O-D pair. The
INTEGRATION framework can model a range of on road vehicle types, and supports a total of 25 default options including
passenger cars, light-duty trucks, and heavy-duty trucks. Similar to most industry standard traffic simulation software, vehicle
operation is governed by proprietary car following, lane changing, and gap acceptance models (H. Rakha 2010).
INTEGRATION was designed with a strong basis in vehicle operation dynamics, and is intended to accurately represent
acceleration and velocity at high temporal resolution. In addition, the model is designed to offer flexibility in estimating
acceleration rates of both large and small vehicles on varying road types and conditions. This accuracy and flexibility is a
necessity for microscopic power-based emissions modeling (H. Rakha 2010) .
INTGRATION allows the user to specify time series histograms for departures for each origin-destination pare in the
simulation. The software generates individual vehicle departure time schedule from this information before the simulation is
run. Departures can be fully random, or any combination of random and uniform. Calibration of the O-D demand is achieved
using a maximum likelihood approach (H. Rakha and Ahn 2004). Pipes and Greenshields models were combined into a single
regime model, based on desired speed and proximity to followed vehicle. This is referred to as Van Aerde's model after
Michael Van Arde, the original developer of INTEGRATION. The model is calibrated using four parameters based on field data:
free-speed, speed-at-capacity, capacity, and jam density. Position, headway, and speed are computed in 0.1 second time
steps (H. Rakha and Ahn 2004).
INTEGRATION uses a separate acceleration and deceleration logic. Deceleration is computed based on the existing speed,
the speed of the vehicle or object that is necessitating the deceleration, and the time available to decelerate. According to
the developers, this can lead to asymptotic deceleration of a vehicle following another vehicle that is traveling at a constant
speed. In any case, a vehicle will not continue to decelerate once it reaches the speed of the vehicle ahead of it (H. Rakha and

Page 161
Ahn 2004). Vehicle acceleration in INTEGRATION is simulated with vehicle dynamics model, as opposed to the more common
kinematic acceleration models. One of the problems identified with state of practice models is in their ability to represent the
speed/acceleration relationship at low vehicle speeds. Most models allow acceleration to vary as a function of speed, with
higher acceleration in general associated with lower travel speeds. However, there are a number of different ways of
modeling this and is no definitive relationship between these two parameters. According to the developers of INTEGRATION,
industry standard traffic models tend to overestimate acceleration at low speeds, which makes them poorly suited for
emissions modeling. (Rakha and Ahn 2004; Rakha et al. 2001, 2004) INTEGRATION addresses this issue by constraining vehicle
acceleration to the physical limitations of the vehicle as described below.
In the INTEGRATION model, power is computed as the minimum of 1) maximum tractive effort based on tire/road
surface friction and vehicle weight and 2) Engine power, corrected for transmission efficiency and accessory power use. In
addition to power, rolling resistance, aerodynamic drag, grade, and vehicle mass are all considered in acceleration
computations. In addition, the model accounts for the fact that drivers do not typically use the maximum power available in
their vehicles. The configuration differs somewhat for light duty and heavy-duty vehicles (LDV’s and HDV’s). The HDV model
accounts for the loss in power at low speeds due to gear shifting. (H. Rakha et al. 2001) The LDV model is very similar, but
assumes that the low speed power loss effect is negligible for passenger vehicles. (H. Rakha, Snare, and Dion 2004)Both the
LDV and HDV models were calibrated using field data collected at Virginia Tech’s Smart Road test facility in Blacksburg,
Virginia.
For the HDV model, four heavy-duty diesel trucks of model years ranging from 1990 to 1998 were driven over the test
course in the spring of 2001. Ten different load cases were tested on each vehicle. By changing the loading, the researchers
hoped to incorporate the range of power to weight rations that would typically be observed in practice. This data was used to
calibrate the power adjustment factor, which is a linear function relating engine power to vehicle speed. This factor is
intended to capture the average power reduction due to gear shifting as a function of speed, as actual power fluctuates
between the maximum value and zero as gear shifts take place (H. Rakha et al. 2001). Vehicle acceleration data was collected
in the summer of 2001 using 13 test vehicles ranging from subcompact to light duty trucks. This data was used to calibrate
and compare several state of practice vehicle acceleration models, including the dynamics model used in INTEGRATION.
Comparisons were made for speed and acceleration vs. time and distance, as well as acceleration vs. speed. The results of the
comparison strongly supported the vehicle dynamics model over the other state of practice models. This was a relatively
simple validation, and did not involve any real world vehicle interaction or complex maneuvers (H. Rakha, Snare, and Dion
2004).
The emissions model incorporated into INTEGRATION is the Virginia Tech Microscopic energy and emission model (VT-
Micro). While a full discussion of the model structure is outside the scope of this report, the model will be described as it
relates to INTEGRATION. Based on the instantaneous speed and acceleration levels of individual vehicles, VT-Micro, predicts
the instantaneous fuel consumption and emission rates of HC, CO, NOX and CO2 (Ahn 1998), (Hesham Rakha, Van Aerde, and
Ahn 2000), and (Ahn 2002).
One important thing to note is that, when the performance envelope of the VT-Micro data is exceeded in the input speed
and acceleration data, emissions will be computed at the boundary of the VT-Micro data. That is to say, the performance
envelope of the VT-Micro data determines the maximum possible emission rates (H. Rakha, Ahn, and Trani 2004). Although
INTEGRATION computes speed and acceleration in 0.1 second time steps, these values are averaged over 1 second time steps
for emission calculations. According to supporting documentation, this does significantly reduce the accuracy of the
emissions estimates, and is done to reduce the computational load.
Emission rates were collected using dynamometer tests and standardized drive cycles. The first study to develop non-
high emitting vehicle emission rates was conducted in 1996 at the Oak Ridge National Laboratory (ORNL). Test vehicles
included 5 light duty automobiles and three light duty trucks, and were selected to be representative of the sales proportion
in terms of engine displacement, based on sales data from 1995 and 1996. The vehicles used in emission rate development
were subjected to driving tests in order to develop practical performance limits that correlate to actual vehicle capabilities (H.
Rakha and Ahn 2004). The general VT-Micro equations used to calculate the instantaneous fuel consumption and emission
rates of individual vehicles could be seen in Equation 7.

Equation 7

Page 162
Where:
MOEe = Instantaneous fuel consumption or emission rate (L/s or mg/s),
a = Instantaneous acceleration of vehicle (km/h/s),
v = Instantaneous speed of vehicle (km/h),
= Vehicle-specific acceleration regression coefficients for MOEe, and
= Vehicle-specific deceleration regression coefficients for MOEe.

Since the initial release, the VT-Micro model has been updated to include emissions data from 87 additional cars and
trucks. This update is described in documentation related specifically to VT-Micro, but it is stated that the updated model has
been included in the current release of INTEGRATION. (H. Rakha and Ahn 2004)This data were gathered by the EPA using
dynamometer tests at the Automotive Testing Laboratories, Inc., in Ohio and EPA's National Vehicle and Fuels Emission
Laboratory (NVREL), in Ann Arbor, Michigan in the spring of 1997. Vehicles of model years 1986 to 1996 were drafted at
random from inspection and maintenance lanes in Ohio. Vehicles were screened to separate high emitting from non-high
emitting vehicles, and separate rates developed for each group. In the screening process, 60 vehicles were identified as non-
high emitters and 37 as high emitters. Each vehicle was tested on 14 to 16 different drive cycles to insure that the full range
of acceleration/speed combinations was captured. In addition to the typical vehicle categories of light duty and heavy duty,
statistical methods were used to define 5 LDV and 2 light duty truck (LDT) categories.
Emission rates were developed from test data using the methods described in VT-micro model supporting
documentation. (H. Rakha, Ahn, and Trani 2004) Regression equations were developed to represent emission rates as
functions of speed, power, and acceleration. Separate regression equations were developed for positive and negative
acceleration, because engine power is exerted in positive acceleration while none is exerted in negative acceleration. An
optimal temporal shift of approximately 6 – 8 seconds was computed to deal with the time lag between acceleration and tail
pipe emissions.
Aggregate and instantaneous emission rates were validated using a number of drive cycles, but no validation was
performed (at least in the rate development studies) using microsimulation vs. real world data. The documentation of the
rate development studies is focused primarily on the emission vs. acceleration and speed relation, as opposed to the
simulated vs. real world relationship.

3.1.5 PARAMICS
Paramics is a microscopic traffic simulation software package published by Quadstone in Edinburgh Scotland.
Interestingly, there are actually two similar microscopic traffic simulation software titles published under the name Paramics,
the Quadstone product and “S-Paramics” which is published by SIAS in Edinburgh. According to the SIAS company website,
the similarity in both name and underlying model is the result of a partnership between Quadstone and SIAS that was
dissolved in 1998 (SAIS 2012). Here the focus is on the Quadstone product for several reasons. One, Quadstone Paramics has
more functionality in terms of emissions modeling. S-Paramics does have a built in emissions model, but it is out of date and
has limited functionality. In addition, Quadstone Paramics contains “Paramics API”, which is a tool for creating and utilizing
added functionality, such as integrating Paramics with external emissions models (Quadstone 2012).
Paramics allows a good deal of user flexibility in terms of vehicle behavior, signal timing, and data collection (Speirs and
Braidwood 2004). Paramics is advertised as “fully scalable”, which means that a broad range of network scales can be
modeled, from a single intersection to an entire city(Quadstone 2012). Simulated vehicles are represented as “Driver-Vehicle
units” (DVU’s), each of which is assigned a number of physical and decision characteristics including vehicle geometry and
performance parameters, familiarity with the traffic network, aggressiveness and origin-destination information. DVU route
selection can be made in three different ways, including deterministic, stochastic cost weighting, and dynamic feedback
which is a real-time decision structure based on driver familiarity and traffic conditions. The “Advanced” version of the
software also contains an OD estimating tool (Quadstone UGM 2009). Vehicle movement across a network is determined by
Quadstone’s proprietary car-following and lane-change models (Quadstone Paramics V5.0 Technical Notes 2004). For signal
timing, Paramics has the ability to model fully actuated and demand responsive signal plans, with the option to change signal
parameters as the model is running. This allows the user to make changes to signal timing and immediately observe the
effect on traffic flow. Paramics apparently does not include an automatic traffic signal optimization application, but does
include a number of up to date tools designed to aid the user in visualizing and optimizing signal timing and other traffic
control mechanisms (Quadstone 2012).
One notable component of Quadstone Paramicss is Monitor, a proprietary emission modeling tool. However, little
documentation was found that describes either the methodology or underlying data used in this application. Based on the
limited documentation available, it is clear that Monitor utilizes tables that contain vehicle exhaust emissions and fuel
consumption rates as a function of vehicle type, speed, and acceleration (M. Barth, Younglove, and Scora 2005). The software
Page 163
simply looks up values for each second in the analysis to compute pollutant quantities and fuel consumption. It is likely that,
in most cases, emission rates will be provided by the user because of the inadequacy of emission data contained in Monitor.
In one study, emission rates were computed for each second in the analysis using a CMEM plug-in developed in the Paramics
API, which demonstrates the flexibility that the API tool provides (M. Barth, Younglove, and Scora 2005).
Another example for using PARAMICS in the emissions related research work is this following paper. (Boriboonsomsin
and Barth 2008) studied the impact of freeway high-occupancy vehicle lane configuration, continuous access HOV and limited
access HOV, of vehicle emissions. In this paper authors used an emissions modeling methodology that integrates PARAMICS,
a microscopic traffic simulation model, with CMEM, a modal emissions model to estimate vehicle emissions from these two
types of HOV lane configurations. As a conclusion for this paper, it was found that the emissions in freeways with continuous
access HOV lane are consistently lower than emissions from freeways with limited access HOV lane. This result was justified
by mentioning that the dedicated ingress/egress sections on the freeway with limited access HOV lane are having highly
concentrated weaving maneuvers that take place on, which cause acceleration/deceleration events to occur with higher
frequency and magnitude.

3.2 Mesoscopic Models

Mesoscopic models combine the properties of microscopic and macroscopic simulation models. They generally represent
most entities at a high level of detail but describe their activities and interactions at a much lower level of detail than would a
microscopic model provides. Generally, they provide more coverage with less modeling detail than microscopic simulation.
Similar to microscopic models, the mesoscopic models’ unit of traffic flow is the individual vehicle. However, the movement
of these vehicles in mesoscopic models follows the approach of the macroscopic models and is governed by the average
speed on the travel link (Lieberman and Rathi 1997), (Dowling, Holland, and Huang 2002), and (Alexiadis, Jeannotte,, and
Chandra 2004).

3.2.1 The MEASURE Model

Researchers at The Georgia Institute of Technology developed the Mobile Emission Assessment System for Urban and
Regional Evaluation (MEASURE). MEASURE is basically a GIS-based emissions model. It can calculate estimates of HC, CO, and
NOx. Two major modules are included in MEASURE. The first one is the start emissions module, while the second one is the
on-road emission module (Bachman et al. 2000). Refined tree-based regression analysis of vehicle emission test data were
used to calculate the emission rates in MEASURE. In MEASURE, emission rates are a function of pollutant, vehicle model year,
vehicle fuel delivery technology, high or normal emitter vehicle, and modal variables. Source of data is from EPA and
California Air Resource Board.
Vehicle emissions for the On-Road emission module are estimated based on different operating modes. These modes
are: idle, cruise, acceleration, and deceleration. Average travel speed, roadway characteristics, traffic flow, and volume to
capacity ratio are used to build the vehicle operating modes. Vehicle registration data was used to get vehicle cold and hot-
start characteristics distribution, and then estimate the start emission module. The start emission estimates will be based on
start characteristics distribution and start emission rates (Bachman et al. 2000).

3.2.2. DYNASMART-P:
DYNASMART (DYnamic Network Assignment-Simulation Model for Advanced Road Telematics) is currently supported by
the Federal Highway Administration through McTrans. DYNASMART-P provides lots of features like: the capability of
modeling big networks, importing of network and demand data from other planning models, enhanced loading and display
speed for large-scale network datasets, and emissions models for light-duty vehicles (Mehta et al. 2003).
The used emission models in DYNASMART-P are adapted from the look-up tables for fuel consumption and emissions
developed at Oak Ridge National Laboratory (ORNL) in the mid-1990s (ORNL models). These fuel consumption and emissions
models are based on functions of vehicle speed and acceleration. DYNASMART-P is using ONROAD models developed at
Texas Southern University for heavy-duty vehicles. EPA also accepted MOBILE 5 to be also interfaced with DYNASMART-P for
additional analysis. More details for the developed methodology of these used models, the look-up tables, and their
evaluation with other emission models like INTEGRATION could be found in these following resources: (H. Rakha et al. 2000),
(West et al. 1997), and (U.S. Department of Transportation, Washington D.C., FHWA 1999).

3.3 Macroscopic Models

Macroscopic simulation models are designed for operations analyses of systems of road facilities. The simulation in a
macroscopic model takes place on a section-by-section basis rather than by tracking individual vehicles. Macroscopic models
are based on the deterministic relationships of the flow, speed, and density of the traffic stream. In these models, entities
Page 164
and their activities and interactions are described at a low level of detail, while minor details,like the lane change maneuvers,
would not be presented at all (Lieberman and Rathi 1997), (Dowling, Holland, and Huang 2002), and (Alexiadis, Jeannotte,,
and Chandra 2004).

3.3.1 MOVES (Motor Vehicle Emission Simulator)

The most recent release, MOVES2010b, is the EPA’s newest multi-scale on-road emissions model. It has replaced MOBILE
6 as the preferred software for State Implementation Plans and Transportation Conformity Analyses. Although it is
considered by the EPA to capable of modeling at macro, meso, and micro scale, it is not a traffic simulator and vehicle activity
data must be input by the user for smaller scale modeling. In macro scale computes vehicle activity is based on nationwide
“vehicle miles traveled” data, which is not precise enough for mesoscopic and microscopic scale analysis (EPA 2012). That
said, MOVES contains an impressive amount of emissions and vehicle data, and can be used with a microscopic traffic
simulator to model system-level changes. Emissions data for light duty passenger vehicles is based largely on inspection
maintenance test data from the Phoenix Arizona area. This data dates from 1995 to 2005, and was collected from
approximately 62,500 vehicles (EPA 2011a). Heavy-duty and additional light duty vehicle and emissions data is based on a
number of different studies, all using American vehicles and driving conditions. In general, the EPA loosely based data
requirements for each vehicle subgroup on the population of that vehicle type currently on the road. As a result, some
emission rates (such as those for heavy duty gasoline vehicles) are based on a comparatively small quantity and range of
actual data (EPA 2011b). Some emission rates were determined using on road vehicle operation, others were simply
calculated from dynamometer tests using EPA standard drive schedules.
Over 60 different pollutants can be modeled separately in MOVES, which likely the most on any emission model
currently available. Basic emission rates in MOVES such as Fuel consumption, NOx, and CO are a function of vehicle specific
power (VSP), and are in units of mass/time (EPA 2011a). Many minor pollutants and all air toxics are computed as fractions of
major pollutant quantities. The software corrects for a number of scenario variables, including meteorological conditions,
startup emissions, and local inspection and maintenance program effects. Simulation is conducted in 1-second time steps and
generates output in the form of MySQL database tables.
At the micro scale, there are two preferred ways to input vehicle activity information. One, activity is input separately for
each link in the modeling area is input in the form of a second-to-second vehicle velocity profile. This can come from actual
vehicle travel logs, EPA standard drive cycles, or from microscopic simulation output. This is used along with default vehicle
characteristics to generate a time distribution of vehicle operating modes differentiated according to the MOVES
classification system by VSP. Alternatively, the user can provide the operating mode distribution directly as an input table.
Total activity is computed in units of time based on average speed and link geometry, and this is used with the operating
mode distribution to compute the time spent in each emission rate-specific operating mode (EPA 2011a).
MOVES was used in many researches in the last ten years. Here is one example of these researches. (Hallmark and
Mudgal 2012) reviewed a study conducted to compare the vehicle activity of a roundabout to other intersections. It is
suggested that roundabouts are more efficient and this study is to test this hypothesis. The study was completed on-road
instrumentation to record second by second vehicle activity, i.e. instantaneous speed and acceleration. This vehicle activity
data were inputted into MOVES and mathematical equations for the vehicle specific power (VSP) bins are used to estimate
the emissions for the three different traffic controls. The VSP bins separate emissions rates and data is placed into bins based
on the emissions rate recorded. With these bins you can see where the majority of the data points lie. The data shows that
roundabouts have higher vehicle emissions through the intersection however they are not in the “system” as long. This
means more experimentation needs to be completed for vehicle emissions over a time period.
Another example for using MOVES is this following paper. (Qiao et al. 2012) looked at the concept that if vehicles can
communicate with stop signs drivers so can start to slow earlier without physically seeing the stop sign. Preemptively locating
stop signs can result in more fuel efficient stopping. In order to communicate to the vehicles the authors analyzed the use of
Radio Frequency Identification (RFID) devices. RFIDs have already been implemented in the transportation discipline. MOVES
was used alongside incorporated VSP bin distributions and their related emissions rates. Average emissions were found from
the MOVES binning and PEMS emissions data. Emissions rates were estimated for vehicles with and without the RFID. The
vehicle with the RFID had lower emission rates no matter which direction the driver chose to go once leaving the intersection.

3.3.2 Watson Model

A fuel consumption model was developed by Watson et al. using average speed data. the changes in the positive kinetic
energy during acceleration was used as a predictor variable in this model, But the effects of speed changes during the
deceleration phase are not included. (Watson, Milkins, and Marshall 1979). It is worth to say also that the effect of
aerodynamic drag on fuel consumption becomes significant at higher average speeds (at average speeds over 55 km/h), and

Page 165
this is also is not included in the study (EVANS and HERMAN 1978).The fuel consumption-space mean speed relationship
could be seen in Equation 8:

Equation 8

Where,
F = fuel consumed (L/km)
Vs = space mean speed (km/hr)
2
PKE represents the sum of the positive kinetic energy changes during acceleration in m/s ,
Vf = final speed (km/hr) f
Vi = initial speed (km/hr) i
Xs = total section length (km)

SUMMARY AND CONCLUSION

The main goal of this paper is to synthesize and document the different aspects of emissions inventories, models, and
optimization of these models. The synthesis work presented in this paper focused on three main areas: fuel consumption and
emission modeling tools for signalized intersection operation, sources of emissions inventory and data used in the models,
and finally how to optimize these emission models. In the first part of this paper, the authors started in giving some details
about the definition, importance, and methods of calculation of emission inventories, emission factors, and some of the
currently available emission inventories. After that and in the second part, the authors reviewed currently available fuel
consumption and emission models suitable for modeling traffic operations at signalized intersection approaches. The review
in this second part included three different analysis levels: microscopic models, mesoscopic models, and macroscopic models.
In the third part of this paper, the authors provided documentation for the optimization of fuel consumption and vehicle
emission models. The organization of this section consisted of techniques that utilize mathematical models, simulation
packages, and in-the-field systems. All of these topics are supposed to form a good base for those who are interested to take
a comprehensive introduction for the research work done in the fuel consumption and emissions fields.

REFERENCES

Ahn, Kyoungho. 1998. “Microscopic Fuel Consumption and Emission Modeling”. Virginia Polytechnic Institute and State
University. http://scholar.lib.vt.edu/theses/available/etd-122898-094232/.

———. 2002. “Modeling Light Duty Vehicle Emissions Based on Instantaneous Speed and Acceleration Levels”. University
Libraries, Virginia Polytechnic Institute and State University.

Alexiadis, Vassili, Krista Jeannotte,, and Andre Chandra. 2004. Traffic Analysis Toolbox Volume I: Traffic Analysis Tools
PrimerFHWA-HRT-04-038. Washington, DC: FHWA.

An, Feng, Matthew Barth, Joseph Norbeck, and Marc Ross. 1997. “Development of Comprehensive Modal Emissions Model:
Operating Under Hot-stabilized Conditions.” Transportation Research Record: Journal of the Transportation Research
Board 1587 (-1): 52–62.

Bachman, William, Wayne Sarasua, and Randall Guensler. “Geographic Information System Framework for Modeling Mobile
Source Emissions.” Transportation Research Record Record No. 1551: 123–133.

Bachman, William, Wayne Sarasua, Shauna Hallmark, and Randall Guensler. 2000. “Modeling Regional Mobile Source
Emissions in a Geographic Information System Framework.” Transportation Research Part C: Emerging Technologies
8 (1): 205–229.

Page 166
Barth, M., F. AN, T. Younglove, G. Scora, C. Levine, M. Ross, and T. Wenzel. 2000. “Comprehensive Modal Emissions Model.”
https://www.cert.ucr.edu/cmem/docs/NCHRP_Final_Report.pdf.

Barth, M. J., F. An, T. Younglove, G. Scora, C. Levine, M. Ross, and T. Wenzel. 2000. Development of a Comprehensive Modal
Emissions Model. http://trid.trb.org/view.aspx?id=859153.

Barth, Matthew, John Collins, George Scora, Nicole Davis, and Norbeck Joseph. 2006. “TRB-2006-Modeling-extremely-low-
emitting-vehicles.pdf”. Transportation Research Board. http://www.cert.ucr.edu/cmem/docs/TRB-2006-Modeling-
extremely-low-emitting-vehicles.pdf.

Barth, Matthew, Theodore Younglove, and George Scora. 2005. “Development of a Heavy-Duty Diesel Modal Emissions and
Fuel Consumption Model” (January 1). http://escholarship.org/uc/item/67f0v3zf#page-1.

Beckx, Carolien, Luc INT Panis, Rudi Torfs, Davy Janssens, and Steven Broekx. 2007. “The Application of the Simulation
Software VeTESS to Evaluate the Environmental Impact of Traffic Measures.” In Iguassu Falls, Brazil.
http://uhdspace.uhasselt.be/dspace/handle/1942/10533.
Bloomberg, L., and J. Dale. 2000. “A Comparison of the VISSIM and CORSIM Traffic Simulation Models.” In Institute of
Transportation Engineers Annual Meeting, 3–16.
http://www.chinautc.com/information/manage/UNCC_Editor/uploadfile/20081022160007907.pdf.

Bloomberg, Loren, and Jim Dale. 2000. “Comparison of VISSIM and CORSIM Traffic Simulation Models on a Congested
Network.” Transportation Research Record: Journal of the Transportation Research Board 1727 (March): 1–23.

Boriboonsomsin, Kanok, and Matthew Barth. 2008. “Impacts of Freeway High-occupancy Vehicle Lane Configuration on
Vehicle Emissions.” Transportation Research Part D: Transport and Environment 13 (2) (March): 112–125.
doi:10.1016/j.trd.2008.01.001.

California Environmental Protection Agency - Air Resources Board. 2011. “EMFAC 2011- Technical Documentation.”
http://www.arb.ca.gov/msei/emfac2011-documentation-final.pdf.

“Compilation of Air Pollutant Emission Factors | Modeling and Inventories | US EPA.” 2013. Accessed February 18.
http://www.epa.gov/otaq/ap42.htm.

Dowling, R., P. E. J. Holland, and P. E. A. Huang. 2002. “California Department of Transportation Guidelines for Applying
Traffic Microsimulation Modeling Software.” Way 3: 3–2.

EPA. 1992. Procedures for Emission Inventory Preparation.pdfEPA420-R-92-009.

http://www.epa.gov/otaq/models/nonrdmdl/r92009.pdf.

———. 2011a. “Development of Emission Rates for Light-Duty Vehicles in the Motor Vehicle Emissions Simulator
(MOVES2010) Final Report”. EPA. http://www.epa.gov/otaq/models/moves/documents/420r11011.pdf.

———. 2011b. “Development of Emission Rates for Heavy-Duty Vehicles in the Motor Vehicle Emissions Simulator
MOVES2010 Final Report”. EPA. http://www.epa.gov/otaq/models/moves/documents/420b12049.pdf.

———. 2012. “MOVES2010b User Guide”. the United States Environmental Protection Agency (EPA).
http://www.epa.gov/otaq/models/moves/documents/420b12001b.pdf.

———. 2012. “Technology Transfer Network Clearinghouse for Inventories & Emissions Factors.” Accessed November 14.
http://www.epa.gov/ttnchie1/eiinformation.html.

“EPA - Air Quality Management (AQM) - Emissions Inventory.” 2012. Accessed December 27.
http://www.epa.gov/air/aqmportal/management/emissions_inventory/.

EPA - nonroad. 2012. “Nonroad.” Accessed December 28. http://www.epa.gov/oms/nonrdmdl.htm.

Page 167
EVANS, LEONARD, and ROBERT HERMAN. 1978. “Automobile Fuel Economy on Fixed Urban Driving Schedules.”
Transportation Science 12 (2) (May 1): 137–152. doi:10.2307/25767904.

FHWA. 2010. “Appendix A: Introduction to CORSIM Theory.” Appendix A: Introduction to CORSIM Theory. June.
http://ops.fhwa.dot.gov/trafficanalysistools/tat_vol4/app_a.htm.

Gomes, G., A. May, and R. Horowitz. 2004. “A Microsimulation Model of a Congested Freeway Using VISSIM.” In
Transportation Research Board Conference.
http://www.me.berkeley.edu/~horowitz/Publications_files/All_papers_numbered/136c_Gomes_VISSIM_calibration
_TRB2003.pdf.
Hall, K. T., M. I. Darter, and W. M. Rexroad. 1993. Performance of Bare and Resurfaced JRCP and CRCP on the Illinois Interstate
Highway System, 1991 Update. Department of Civil Engineering, Engineering Experiment Station, University of
Illinois at Urbana-Champaign. http://128.174.2.147/publications/report%20files/TES-077.pdf.

Hallmark, S.L., and A. Mudgal. 2012. “Comparison of VSP Profiles for Three Types of Intersection Control and Implications for
Emissions.” In 2012 15th International IEEE Conference on Intelligent Transportation Systems (ITSC), 415 –420.
doi:10.1109/ITSC.2012.6338862.

Kosman, K. P., S. L. Hallmark, S. Poska, and T. Planner. 2003. “Evaluation of Simulation Models for Project-Level Emissions
Modeling.” In Transportation Research Board, Annual Meeting. http://www.ltrc.lsu.edu/TRB_82/TRB2003-
001925.pdf.

Lieberman, E., and A. K. Rathi. 1997. “Traffic Simulation.” Traffic Flow Theory.

Mehta, T., A. Kottapalli, H. S. Mahmassani, and C. Bhat. 2003. “Assessing the Environmental Benefits of Intelligent
Transportation Systems Measures: Methodologies And Applications.” Work 4197: 3.

Minnesota Department of Transportation. 2008. Advanced CORSIM Training Manual- Final Corsim Manual 9-19-09.SEH No. A-
MNDOT0318.00.
http://www.dot.state.mn.us/trafficeng/modeling/resources/CORSIMmanual/final%20corsim%20manual%209-19-
09.pdf.

New Zealand Ministry for the Environment. 2012. “Good Practice Guide for Preparing Emissions Inventories.” Accessed
November 14. http://www.mfe.govt.nz/publications/air/emissions-good-practice-guide-01.html.

Office of Air Quality Planning and Standards, US EPA. 2012a. “Emissions Inventories | Clearinghouse for Emission Inventories
and Emissions Factors | Technology Transfer Network | US EPA.” Accessed December 27.
http://www.epa.gov/ttnchie1/net/2011inventory.html.

———. 2012b. “2002 National Emissions Inventory Data & Documentation.” Accessed December 27.
http://www.epa.gov/ttn/chief/net/2008inventory.html.

———. 2012c. “Emissions Factors & AP 42.” Accessed December 27. http://www.epa.gov/ttnchie1/ap42/.

———. 2012d. “Biogenic Emission Sources.” Accessed December 28. http://www.epa.gov/ttnchie1/emch/biogenic/.

Park, B., I. Yun, and K. Choi. 2004. “Evaluation of Microscopic Simulation Tools for Coordinated Signal System Deployment.”
KSCE Journal of Civil Engineering 8 (2). Transportation Engineering (March): 239–248.

Peter Holm, Daniel Tomich, Jaimie Sloboden, and Cheryl Lowrance. 2007. Traffic Analysis Toolbox Volume I: Traffic Analysis
Tools PrimerFHWA-HRT-04-038. Oakland, CA: Cambridge Systematics, Inc.
http://ops.fhwa.dot.gov/trafficanalysistools/tat_vol1/vol1_primer.pdf.

Page 168
PTV Group. 2012. “VISSIM and Enviver.docx.” Common Add-ons - EnViVer. http://www.ptv-vision.com/en-
uk/products/vision-traffic-suite/ptv-vissim/modules/.

Qiao, Fengxiang, Jinghui Wang, Xiaobing Wang, Jing Jia, and Lei Yu. 2012. “A RFID Based e-STOP Sign and Its Impacts to
Vehicle Emissions.” In 2012 15th International IEEE Conference on Intelligent Transportation Systems (ITSC), 206 –
211. doi:10.1109/ITSC.2012.6338899.

Quadstone. 2012. “Quadstone Paramics | Traffic and Pedestrian Simulation, Analysis and Design Software.”
http://www.paramics-online.com/.
Quadstone Paramics V5.0 Technical Notes. 2004. Technical Notes. Scotland: Quadstone Limited. http://www.paramics-
online.com/downloads/technicaldocs/Quadstone%20Paramics%20Core%20Models.pdf.

Quadstone UGM. 2009. “Paramics North American Update 2009”. Quadstone Limited. http://www.paramics-
online.com/downloads/presentations/paramics_ugm_2009.pdf.

Rakha, H. 2010. “INTEGRATION Software.” INTEGRATION Dynamic Traffic Assignment and Simulation Software.
http://www.filebox.vt.edu/users/hrakha/Software.htm.

Rakha, H., M. V. Aerde, K. Ahn, and A.A. Trani. 2000. “Requirements for Evaluating Traffic Signal Control Impacts on Energy
and Emissions Based on Instantaneous Speed and Acceleration Measurements.” In Washington D.C.

Rakha, H., and K. Ahn. 2004. “Integration Modeling Framework for Estimating Mobile Source Emissions.” Journal of
Transportation Engineering 130 (2): 183–193.

Rakha, H., K. Ahn, and A. Trani. 2004. “Development of VT-Micro Model for Estimating Hot Stabilized Light Duty Vehicle and
Truck Emissions.” Transportation Research Part D: Transport and Environment 9 (1): 49–74.

Rakha, H., I. Lucic, S. H. Demarchi, J. R. Setti, and M. Van Aerde. 2001. “Vehicle Dynamics Model for Predicting Maximum
Truck Acceleration Levels.” Journal of Transportation Engineering 127 (5): 418–425.

Rakha, H., M. Snare, and F. Dion. 2004. “Vehicle Dynamics Model for Estimating Maximum Light-duty Vehicle Acceleration
Levels.” Transportation Research Record: Journal of the Transportation Research Board 1883 (-1): 40–49.

Rakha, Hesham, Michel Van Aerde, and K. Ahn. 2000. “Requirements for Evaluating Traffic Signal Control Impacts on Energy
and Emissions Based on Instantaneous Speed and Acceleration Measurements.” In Washington, D.C.

“Related Programs | TTN - Support Center for Regulatory Atmospheric Modeling | US EPA.” 2012. Accessed December 28.
http://www.epa.gov/scram001/dispersion_related.htm#ems-hap.

SAIS. 2012. “SIAS Transport Planners.” Paramics Microsimulation.

http://www.sias.co.uk/ng/sparamicshome/sparamicshome.htm.

SMOKE Version 3.1. 2012. http://www.smoke-model.org/index.cfm.

Speirs, Ewan, and Richard Braidwood, ed. 2004. “Quadstone Paramics V5.0 Technical Notes.” http://www.paramics-
online.com/downloads/technicaldocs/Quadstone%20Paramics%20Core%20Models.pdf.

TFEIP. 2012. “Task Force on Emission Inventories and Projections (TFEIP).” Accessed November 14. http://tfeip-
secretariat.org/.

The International Vehicle Emissions (IVE) Model. 2012. http://www.issrc.org/ive/.

U.S. Department of Transportation, Washington D.C., FHWA. 1999. Development and Validation of Light-Duty Vehicle Modal
Emissions and Fuel Consumption Values for Traffic ModelsFHWA-RD-99-068.

Page 169
Umedu, T., Y. Togashi, and T. Higashino. 2012. “A Self-learning Traffic Signal Control Method for CO2 Reduction Using
Prediction of Vehicle Arrivals.” In 2012 15th International IEEE Conference on Intelligent Transportation Systems
(ITSC), 421 –426. doi:10.1109/ITSC.2012.6338843.

United Nations Economic Commission for Europe. 2012. “TF Emission Inventories and Projections.” Accessed November 14.
http://www.unece.org/env/lrtap/taskforce/tfeip/welcome.html.

Watson, H. C., E. E. Milkins, and G. A. Marshall. 1979. “A Simplified Method for Quantifying Fuel Consumption of Vehicles in
Urban Traffic.” In . http://trid.trb.org/view.aspx?id=1206155.

West, Brian H., Ralph N. McGill, Jeffrey W. Hodgson, C. Scott Sluder, and David E. Smith. 1997. Development of Data-Based
Light-Duty Modal Emissions and Fuel Consumption Models972910. Warrendale, PA: SAE International.
http://papers.sae.org/972910/.

Xia, Haitao, K. Boriboonsomsin, F. Schweizer, A. Winckler, Kun Zhou, Wei-Bin Zhang, and M. Barth. 2012. “Field Operational
Testing of ECO-approach Technology at a Fixed-time Signalized Intersection.” In 2012 15th International IEEE
Conference on Intelligent Transportation Systems (ITSC), 188 –193. doi:10.1109/ITSC.2012.6338888.

Yue, Huanyu. 2008. “Mesoscopic Fuel Consumption and Emission Modeling”. Virginia Polytechnic Institute and State
University. http://scholar.lib.vt.edu/theses/available/etd-04082008-101942/.

Zallinger, M., T. Anh, and S. Hausberger. 2005. “Improving an Instantaneous Emission Model for Passenger Cars.” In , 167–
176. Graz, Austria.

Page 170