Evaluating effects of eco-driving at traffic intersections based on traffic micro-simulation

Eco-driving is an initiative driving behavior which aims to reduce fuel consumption and emissions from automobiles. Recently, it has attracted increasing interests and has been adopted by many drivers in Australia. Although many of the studies have revealed considerable benefits in terms of fuel consumption and emissions after utilising eco-driving, most of the literature investigated eco-driving effects on individual driver but not traffic flow. The driving behavior of eco-drivers will potentially affect other drivers and thereby affects the entire traffic flow. To comprehensively assess and understand how effectively eco-driving can perform, therefore, measurement on traffic flow is necessary. In this paper, we proposed and demonstrated an evaluation method based on a microscopic traffic simulator (Aimsun). We focus on one particular eco-driving style which involves moderate and smooth acceleration. We evaluated both traffic performance (travel time) and environmental performance (fuel consumption and CO2 emission) at traffic intersection level in a simple simulation model. The before-and-after comparisons indicated potentially negative impacts when using eco-driving, which highlighted the necessity to carefully evaluate and improve eco-driving before wide promotion and implementation. 1.0 Introduction In Australia, the automobile is one of the major sources of greenhouse emissions, and has contributed 69.2 million tons of greenhouse emissions in 2008 with an increase of 26% over the last twenty years (Australian Government Department of Climate Change and Energy Efficiency, 2010). Such a large amount of emissions has created significant environmental stress on communities. Hence, many strategies have been promoted to alleviate environmental problems caused by automobiles. “Eco-driving” is one such strategy, aiming to reduce exclusive fuel consumption and greenhouse emissions by modifying or optimizing drivers’ behavior. Compared with hardware-based strategies, such as using high-efficiency engines and green energy, eco-driving is especially advantageous as it can be applied to all kinds of vehicles on the road immediately. As a result, eco-driving is receiving increasing popularity among the drivers and the authorities are rolling out a variety of eco-driving programs. Representative eco-driving involves various driving behaviors, such as maintaining a steady speed, avoiding heavy acceleration and deceleration, well anticipating the traffic flow ahead, and minimising idling time. These behaviors will tend to smooth vehicle movements and avoid ATRF 2011 Proceedings 2 unnecessary fuel consumption, thereby reducing greenhouse emissions. In addition, eco-driving involves vehicle maintenances, such as monitoring tyre pressure, adopting regular service and avoiding unnecessary weights. Based on proper maintenances, the vehicles can secure optimal mechanical conditions, thus improving fuel efficiency from the respective of engine operations. The benefits of employing proper vehicle maintenances can be easily recognised because the vehicle condition is internally predefined regardless of the vehicle operations on roads. However, the benefits from adopting particular eco-driving behavior will involve somewhat uncertainty because the vehicle operations on road are subject to the complexity of traffic conditions. For instance, under congestion, steady speed is absolutely not easy to maintain. Therefore, emphasis should be paid on whether a particular eco-driving behavior will be able to benefits the environments under different scenarios and how such the behavior will affect traffic conditions on roads. To this day, advantages of eco-driving dominate the evaluation results among the majorities of the existing studies. Hornung (2004) used a driving simulator to collect the drivers’ behavior and fuel consumption data before and after an eco-driving training course. The fuel consumption was found to be 17% lower after adopting eco-driving among a group of seventy-nine participants. SenterNovem (2005) uncovered the 5-10% fuel saving on average that can be achieved according to the field test results and real experience from a wide spread eco-driving project (TREATISE) conducted in Europe. An Australian eco-driving trail also reported that a 27% reduction in fuel consumption was able to be achieved by fully trained eco-drivers (Symmons et al., 2009). Although the results from existing studies indicated eco-driving as an effective strategy to save fuel consumption and reduce greenhouse emissions, most of the current studies are conducted under specified testing environment and constrains, such as vehicle type, time of day, driving router, design of evaluation, etc. (Symmons et al., 2009). Evaluations subject to local and limited conditions are difficult to provide substantial evidence for a nationwide promotion and implementation. More critical, litter literature has investigated the impacts of eco-driving on traffic flow. The common design of eco-driving studies focuses on before-and-after comparison with respect to only individual vehicle. On the basis of an individual vehicle, the performance of eco-driving is quite deterministic because the core of eco-driving is to efficiently operate the vehicle. However, when considering traffic flow with a mix of driving behavior, in fact the situation will be much more complicated. Eco-driving will affect not only the eco users but also the other drivers in the traffic flow because the vehicles are inter-related. As shown in Figure 1, assuming two typical driving styles on the road, i.e. normal driving and eco-driving, because of the interactions between different driving styles, the actual vehicle operations will be different with regard to not only the driving style itself but also the position in the traffic flow. This implies the importance of evaluating eco-driving on the basis of traffic flow. Hence this paper will introduce an evaluation method to investigate impacts of eco-driving on traffic flow, which is expected to provide useful information to improve existing eco-driving strategies, to identify where and when to implement eco-driving, and to cooperate with other 1 Training programme for local energy agencies and actors in transport and sustainable energy actions Evaluating effects of eco-driving at traffic intersections based on traffic micro-simulation 3 traffic control strategies. In section two, the applicability of using microscopic traffic simulator is explored by an example of simulation model. In this example, a particular eco-driving behavior, moderate and smooth acceleration, is studied. The impacts of eco-driving in the example is concluded by measuring both traffic performance (travel time) and environmental performance (fuel consumption and CO2 emission). In section three, the evaluation results and potential implementations of this study are discussed. Figure 1: Potential driving operations with normal driving and eco-driving in a traffic flow 2.0 Evaluating eco-driving based on traffic micro-simulation Summarised by Smit et al. (2010), evaluating eco-driving involves quantification of driving behavior and its corresponding impacts on vehicle performances (i.e. the fuel consumption and greenhouse emissions). As this paper will consider the impacts on traffic flow, an extra component, traffic flow characteristics, must be added on. The framework to quantify driving behaviors, traffic flow and environmental performance are shown in Figure 2. The driving behavior is the basic unit which physically determines the environmental performance from respective of individual vehicle. The traffic flow is composed by individual driving behavior but it is subject to various conditions, such as the flow speed and traffic signal control. Consequently, the performance of the traffic flow is the product of driving behavior and traffic flow conditions. Figure 2: the framework to quantify effects of eco-driving Generally, there are two approaches to assess the impacts of driving behaviors; they are field tests and traffic simulations. When considering the impacts on traffic flow, traffic simulations have obvious advantages over field tests. Traffic simulations are able to replicate various traffic conditions in computers, while the field measurements will be labor-intense and time-consuming in testing different scenarios. In addition, a traffic simulation can generate Driving Behavior Individual driving behavior (e.g. speed and acceleration profile) Traffic Flow  Traffic flow characteristics (e.g. flow speed)  Traffic control (e.g. traffic signals) Environment  Emission Estimation (e.g. CO2)  Fuel consumption Eco User Eco User Non-Eco User Non-Eco User Operation A Operation B Operation C Operation D