Breakout #18

Breakout Session #18

Effects of Vehicle Automation on Energy- and Carbon-Intensity

Wednesday, July 20

Room- Union Square 22

Organizers

  • Don MacKenzie, University of Washington
  • Paul Leiby, Oak Ridge National Laboratory

Overview

This session will consider potential effects of automation on the energy and carbon intensity of vehicle travel (MJ/km, gC/MJ). Along with travel distance and mode choices (being addressed in other sessions) these are the major determinants of total energy demand and GHG emissions from vehicles. The session will thus be centered on the question, "Given improved projections of future travel demand and patterns in response to CAVs, what will be the system-wide effects on energy demand and climate change?"

Themes to be addressed include:

  • Opportunities to increase operational energy efficiency through platooning, traffic smoothing, congestion relief, or other measures;
  • Effects of automated, shared-use services on vehicle size, and adoption of efficiency technologies and alternative fuels;
  • Synergies between automation, advanced efficiency technologies, and alternative fuels;
  • Opportunities and limits of public policy in promoting energy efficiency and GHG reductions as outcomes of automation; and
  • Understanding the feedbacks among operational efficiency, vehicle design, and changes in demand in the face of automation

Attendees will then break into facilitated small group discussions to identify key unknowns, research questions, and possible approaches to addressing them.

Goals/Outputs

Short proceedings document synthesizing:

  • Presenters’ materials, content of audience Q&A, and small group discussions
  • Current knowledge regarding potential effects of CAV technologies on energy/carbon intensity of road travel
  • Identified key research questions and possible research approaches

AGENDA

    Setting the framework

  1. Matthew Barth & Caroline Rodier, University of California. “Energy and Emissions Implications of Connected and Automated Vehicles: Results from a National Center for Sustainable Transportation study.”
  2. Energy impacts of automated mobility services

  3. Yuche Chen, National Renewable Energy Laboratory. “Estimate of Fuel Consumption and GHG Emission Impact from an Automated Mobility District.”
    This study estimates the range of fuel and emissions impacts of an automated-vehicle (AV)-based transit system that services campus-based developments, termed an automated mobility district (AMD). The study develops a framework to quantify the fuel consumption and greenhouse gas (GHG) emission impacts of a transit system comprised of AVs, taking into consideration average vehicle fleet composition, fuel consumption/GHG emission of vehicles within specific speed bins, and the average occupancy of passenger vehicles and transit vehicles.

    The framework is exercised using a previous mobility analysis of a personal rapid transit (PRT) system, a system that shares many attributes with envisioned AV-based transit systems. Total fuel consumption and GHG emissions with and without an AMD are estimated, providing a range of potential system impacts on sustainability. The results of a previous case study based on a proposed implementation of PRT on the Kansas State University (KSU) campus in Manhattan, Kansas, serve as the basis for estimating personal miles traveled supplanted by an AMD at varying levels of service.

    The results show that an AMD has the potential to reduce total system fuel consumption and GHG emissions 4% to 14% dependent on operating and ridership assumptions. Even with the increase in travel demand, the PRT system has the potential to reduce system fuel consumption based on higher occupancy and better vehicle performance. The study points to the need to better understand ride-sharing scenarios and calls for future research on sustainability benefits of an AMD system at both vehicle and system levels.

  4. Kara Kockelman, University of Texas. “Shared Autonomous Electric Vehicles (SAEV) Operations across the Austin, Texas Network, with a Focus on Charging Infrastructure Decisions.”
    Autonomous vehicles (AVs) offer a variety of potential benefits, including shared, electrified fleet use. Shared Autonomous Electric Vehicles (SAEVs) can transport passengers or groups of passengers, much like taxis and minibuses, providing substantial cost savings to passengers by distributing the cost of AV ownership, avoiding parking fees, and filling seats by sharing rides. EV and AV technologies are advancing, with longer all-electric ranges - thanks to lower-cost batteries, hands-free and faster charging - thanks to innovations in inductive station pads, and smart ride coordination and route choices, thanks to mapping technologies and vehicle automation. SAEVs will have lower range and longer re-fueling times at fewer locations than conventional-engine SAVs, resulting in different operational outcomes. However, electric power costs less than petrol, and can be generated domestically and using renewables, providing significant national-security and environmental benefits.

    This work simulates the effects of various SAEV fleet sizes in serving the 3-county Austin, Texas metro area using the agent-based simulator MATSim. MATSim is able to track vehicles – not just their passengers, while reflecting activity-based travel patterns and dynamically evolving traffic conditions, with travel time feedbacks and traveler schedule, mode, and destination changes. This study illustrates the differences in optimal charging site investments and schedules relative to a conventional SAV fleet.
  5. Effects of driving style and control

  6. Jun Liu, University of Texas. “Anticipating the Emissions Impacts of Autonomous Vehicles Using the MOVES Model.”
    Connected and autonomous vehicles (CAVs) are expected to have significant impacts on the environmental sustainability of transportation systems. This study examines the emission impacts of CAVs, presuming that CAVs are programmed to drive more smoothly than humans. This work uses the US Environmental Protection Agency’s (EPA’s) Motor Vehicle Emission Simulator (MOVES) to estimate CAVs’ emissions based on driving schedules or profiles. CAV engine load profiles are anticipated to be smoother than those of human-controlled vehicles (HVs), because CAVs are designed to be more situationally aware (thanks to cameras and radar communications) and enjoy faster reaction times and more sophisticated throttle and brake control than HVs. Human drivers tend to demonstrate significant and frequent speed fluctuations and have relatively long reaction times.

    This study uses EPA driving cycles and Austin-specific driving schedules to reflect national, trip-based and local, link-based driving behaviors. Those driving cycles are smoothed using spline functions, to estimate how CAVs may handle; and emissions results suggest that the smoothed CAV cycles deliver lower average emission rates (in grams per mile) for all five species of interest. For example, with gasoline vehicles, smoothing of the Federal Test Procedure (FTP) cycle delivers 11.7% less fine particulate matter (PM2.5), 6.4% less carbon monoxide (CO), 13.6% less oxides of nitrogen (NOx), and 3% less sulfur and carbon dioxide (SO2 and CO2). Reductions using the nation’s more aggressive test cycle (US06) were 28.3% for PM2.5, 24.3% for CO, 23.4% for NOx, and 1.8% for SO2 and CO2. Using Austin link-based cycles, average reductions were 21.2% for PM2.5, 15.3% for CO, 17.2% for NOx, and 8.6% for both SO2 and CO2. While added travel distances by CAVs may more than cancel many of these benefits, it is valuable to start discussing a shift to gentle driving, to obtain these reductions via emerging technologies.

  7. Dominik Karbowski, Argonne National Laboratory. “Impact of Connectivity and Automation on Advanced Vehicles Energy Use.”
    Connectivity and automation are increasingly being developed for cars and trucks, aiming to provide better safety and better driving experience. As these technologies mature and reach higher adoption rates, they will also have an impact on the energy consumption: Connected and Automated Vehicles (CAVs) may drive more smoothly, stop less often, and move at faster speeds, thanks to overall improvements to traffic flows. These potential impacts are not well studied, and any existing studies tend to focus solely on conventional engine-powered cars, leaving aside electrified vehicles such as Hybrid Electric Vehicles (HEVs) and Battery Electric Vehicles (BEVs).

    This work intends to address this issue by analyzing the energy impact of various CAV scenarios on different types of electric vehicles using high-fidelity models. The vehicles—all midsize, one HEV, one BEV, and a conventional—are modeled in Autonomie, a high-fidelity, forward-looking vehicle simulation tool. They are simulated on various CAVs scenarios and modeled by variations of the drive cycle.

    First, a reference fuel consumption value is obtained for steady-state speeds, which estimate an optimal state reached by the highest achievable connectivity degree, in which vehicles never stop and drive at constant speed. Second, Real-World Driving Cycles (RWDCs) are selected from a database of recorded Global Positioning System (GPS) traces in the Chicago area. Energetic criteria are used to select RWDCs representing the average driving style. Different changes to the original speed profiles are then applied to represent the connectivity impact: some stops are removed, speed is smoothed, and strong accelerations are saturated. An overall increase in speed is also investigated to represent improved traffic flow. In each case, the distance remains the same as in the original case, representing the same origin and destination. Finally, a detailed energy analysis is performed, highlighting the close relationship between CAV technologies and powertrain electrification.

    This work shows the synergies between connected vehicles and electrified vehicles. Conventional vehicles and electrified vehicles both can benefit highly from connectivity because they have an equivalent proportional potential fuel consumption reduction. However, on the one hand, because conventional vehicle fuel consumption is the highest, connectivity has a larger absolute energy impact on conventional vehicles. On the other hand, because the best fuel consumption levels for HEVs and BEVs are more easily reached with connectivity, connected electrified powertrains also have high energy savings potential.

  8. Kanok Boriboonsomsin & Matthew Barth, University of California Riverside (invited). “An update on eco-control strategies in the lab and in the field.”
    Connected and automated vehicles offer a wide range of opportunities to reduce energy consumption through smarter vehicle control. These include eco-optimizations of signal approach and departure, cooperative adaptive cruise control, and energy management of plug-in hybrid vehicles. In this presentation, we will review recent research on these strategies including laboratory simulations, test track, and field experiments conducted at UC Riverside.
  9. From local results to national impacts

  10. Don MacKenzie, University of Washington. “Scaling of platooning energy benefits with compatible vehicle adoption.”
    In this work, I estimate how the energy savings from platooning scale with level of adoption of platooning-capable vehicles, under the assumption that mutual compatibility is required for successful platooning. The greater the adoption of platoon-capable vehicles, the more frequently they will encounter compatible vehicles on the highway, and the more time they will spend in platoons. Moreover, average platoon size will increase with the prevalence of compatible vehicles, leading to larger aerodynamic improvements and greater energy savings. I use a simple simulation to capture the formation of platoons of compatible vehicles in a freeway environment with passing. Results indicate that at a 10% adoption of platooning-capable vehicles, each of these vehicles realizes only 40% of the maximum energy savings that would be possible if all vehicles were platoon-capable, translating into approximately a 0.4% reduction in overall energy consumption.

  11. Jeff Gonder, National Renewable Energy Laboratory. “Potential Energy Impacts of Connected and Automated Vehicles: Opportunities and Approaches.”
    Initial estimates of the potential impacts of vehicle connectivity and automation have been reported for several scenarios, mostly at a city or regional scale. A number of these studies were reviewed and combined to estimate approximate bounds on the potential energy use for passenger transportation in the U.S. at low and high levels of connected and automated vehicle (CAV) technologies, considering high and low levels of ridesharing. These bounds are wide due to large uncertainties in potential changes in vehicle efficiency, travel demand, and in the future adoption of CAVs.

    In order to understand the potential impacts of CAVs on energy and greenhouse gas (GHG) emissions, methods are being developed to take results from additional case studies of CAVs, in particular, detailed modeling and simulation results, and expand these results to the U.S. national level. Expansion methods will be applicable to vehicle-level simulation results as well as city- or metropolitan-area transportation system modeling results. This collaborative research effort focuses on three areas: (1) mapping vehicle-level efficiency changes attributable to CAVs capabilities to the entire route network across the U.S., (2) transferring estimated changes in travel patterns, especially in vehicle-miles traveled from activity-based travel demand simulations at a metropolitan area scale to the national scale, and (3) developing analytical methods to model adoption of CAVs technology by different users.

    Methods will be briefly described, and preliminary results will be presented. CAVs use cases and scenarios that present promising opportunities for analysis using these methods will be discussed. To the extent that changes in vehicle efficiency, travel behavior and technology adoption can be influenced by policy, these approaches will allow analysis of the potential influence of policies on the energy and GHG emission outcomes of future CAV deployment.

Breakout 18 / 22