HyDIVE (Hydrogen Dynamic Infrastructure and Vehicle Evolution) model analysis Cory Welch Hydrogen Analysis Workshop, August 9-10 Washington, D.C.
Disclaimer and Government License This work has been authored by Midwest Research Institute (MRI) under Contract No. DE- AC36-99GO10337 with the U.S. Department of Energy (the DOE ). The United States Government (the Government ) retains and the publisher, by accepting the work for publication, acknowledges that the Government retains a non-exclusive, paid-up, irrevocable, worldwide license to publish or reproduce the published form of this work, or allow others to do so, for Government purposes. Neither MRI, the DOE, the Government, nor any other agency thereof, nor any of their employees, makes any warranty, express or implied, or assumes any liability or responsibility for the accuracy, completeness, or usefulness of any information, apparatus, product, or process disclosed, or represents that its use would not infringe any privately owned rights. Reference herein to any specific commercial product, process, or service by trade name, trademark, manufacturer, or otherwise does not constitute or imply its endorsement, recommendation, or favoring by the Government or any agency thereof. The views and opinions of the authors and/or presenters expressed herein do not necessarily state or reflect those of MRI, the DOE, the Government, or any agency thereof.
Overview Brief overview of spatial, dynamic model Example intermediate static calculations Key unknowns, next step, and insights
Main Feedback Structure: Initial Model (1) Vehicle Incentives (2) Fuel Production/ Purchase Incentives (1) See Struben, 2006a for model exposition and analysis (2) to be incorporated (research at both MIT and NREL is envisioned to b e part of a larger model including additional relevant dynamics.) Station Incentives
Spatial, dynamic interdependence From Struben, Welch, Sterman (2006). See Struben (2006a) for detailed discussion of initial model.
Vehicle Utility Utility of vehicle for driving, which affects vehicle sales, will be affected by: added drive time to limited refueling stations time to refuel vehicle: f (refueling rate, queue time) probability, cost of running out of fuel Above are endogenously calculated dynamically and spatially for different trip lengths (next 2 slides) Other vehicle attributes (e.g., price, performance, etc.) to be incorporated with future development.
Trip distribution a driver of spatial diffusion Fraction of Trips Taken 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 Trip Distribution Frequency* 642,292 trips from National Household Transportation Survey, V4.0, July 2005. 0-15 15-30 30-45 45-60 60-75 75-90 Trip Length (Miles) 90-105 105-120 120-135 135-150 * Data still needs to be inspected and reviewed for proper context. <http://nhts.ornl.gov/2001/html_files/download_directory.shtml> GPS data from southern CA also being sought from SCAG. NHTS data to be inspected/reviewed.
Limited station coverage reduces driving convenience, which will affect sales & miles driven Desired Annual Driver Miles 6,000 5,000 4,000 3,000 2,000 1,000 0 0-15 0-15 Desired Annual Driver Miles vs. Trip Length 15-30 15-30 calculated for each location and each time step in model 30-45 30-45 Vehicle utility is reduced due to additional effort/risk. The effect on potential sales needs to be better understood. 45-60 45-60 60-75 60-75 75-90 Trip Length (Miles) 90-105 105-120 120-135 Gasoline 600 H2 Stations 100 H2 Stations Gasoline 100 600 20 H2 Stations 20 H2 Stations 10 H2 Stations 75-90 90-105 105-120 120-135 135-150 135-150
Insights and Questions Consumer sensitivity to reduced vehicle utility (due to limited coverage) is a major driver of dynamics, but not yet well understood. Spatial diffusion beyond urban areas may be difficult. The following slides illustrate how we are beginning to improve understanding the above.
Directional trip distribution now modeled Station profitability/growth are calculated based on endogenously calculated vehicle demand and trip distribution Additional drive time, refueling time, out of fuel risk are calculated based on level of H 2 station coverage Example trip distribution for one driver living here
Trip distribution affects station profitability Trip destinations and refueling events are skewed largely, on average, toward drivers home location. L.A. Fraction of trips N Profitability difficult for stations far from vehicle owners Advantage: large % of refuelings are covered by stations close to drivers Disadvantage: stations far from drivers home location may be required for adequate utility, but may be very unprofitable.
Expected Distance to Nearest Gasoline Station 44 mi.
Example: Expected Distance to Nearest H 2 Station (20 Stations) 44 mi.
Example: Extra Distance to Nearest H 2 Station (20 Stations) Extra time to station readily calculable assuming an average speed of travel (currently use a constant value, but could relax if spatial data are obtained) 44 mi.
Expected Distance to Nearest Station Expected Miles to Station 5 4.5 4 3.5 3 2.5 2 1.5 1 0.5 0 "Expected" Distance to Nearest Station* (for most "station dense" cell in LA -- 16x22 mi. cells) 0 100 200 300 400 500 600 700 H2 Lighthouse Stations * Probabilistic distance to nearest station, which is a function of station density within each cell and across adjacent cells. Gasoline: ~1/3 mile
Fraction of driving time 1 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 Potential calibration data Hydrogen Learning Demonstration Project Radius Histogram: Vehicle EcoCar1 Fraction of Refuelings 1 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 Theoretical Range = 260 miles. Total detected 1 refuelings = 674 Range Histogram: EcoCars Created: Jul-11-06 9:59 AM 0 0 10 20 30 40 50 60 70 80 Distance from nearest refueling station Created: Jul-11-06 10:14 AM 0 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 Fraction of theoretical range b/w refueling 1. Due to data noise, some refueling events are not detected. Number of Trips 70 60 50 40 30 20 10 Trip Length Histogram: All Vehicles % of non-idle trips < 1 mile = 17% Drivers DO tend to drive vehicles less/differently with reduced station coverage. 0 0 10 20 30 40 50 60 70 Trip Length (miles) Analysis of learning demo driving behavior could be useful for calibration (if future industry approval is obtained), particularly for low levels of station coverage.
Unknowns and next steps Unknown: purchasing sensitivity to reduced vehicle utility Desired Annual Driver Miles vs. Trip Length Desired Annual Driver Miles 6,000 5,000 4,000 3,000 2,000 1,000 0 0-15 15-30 30-45 Vehicle utility is reduced due to additional effort/risk. The effect on potential sales needs to be better understood. 45-60 60-75 75-90 Trip Length (Miles) 90-105 105-120 120-135 135-150 f (additional drive time, additional refueling time, out of fuel risk) Gasoline 600 H2 Stations 100 H2 Stations 20 H2 Stations 10 H2 Stations Next Steps: discrete choice analysis techniques envisioned to be used for quantification of this sensitivity (collaborate with auto OEMs) broaden model boundary, include additional relevant dynamics policy/strategy analysis
Insights and Questions Incentives may need to differ for urban, rural, intercity, or interstate stations to avoid over/under subsidization. Difficult to satisfy simultaneously high % coverage for vehicle owners and profitability for station owners. Consumer sensitivity to reduced vehicle utility (due to limited coverage) is a major driver of dynamics, but not yet well understood. Spatial diffusion beyond urban areas may be difficult... could require continued gov t/market seeding in different cities/states (e.g., intercity/interstate/regional networks). Early in model development process. Additional development and analysis of consumer sensitivity to driving convenience measures will shed more light.
Related References Struben, J. (2006a). Identifying Challenges for Sustained Adoption of Alternative Fuel Vehicles and Infrastructure. MIT Sloan School of Management Working Paper (abstract at http://web.mit.edu/jjrs/www/current%20research.htm, paper soon available) Struben, J. (2006b). Transitions towards Alternative Fuel Vehicles: Learning and Technology Spillover Trade-Offs. Under Preparation. Struben, Jeroen and Sterman, John, (2006c) Transition Challenges for Alternative Fuel Vehicle and Transportation Systems. MIT Sloan Research Paper No. 4587-06 http://ssrn.com/abstract=881800 Struben J. Welch C. and Sterman J. (2006) Modeling the Spatial Co-Evolutionary Dynamics of Hydrogen Vehicles and Refueling Stations. Poster presentation at the National Hydrogen Association Conference. Long Beach, CA. http://www.nrel.gov/hydrogen/pdfs/nha_poster_welch.pdf Welch, C. (2006) Lessons Learned from Alternative Transportation Fuels: Modeling Transition Dynamics. National Renewable Energy Laboratory. NREL/TP-540-39446. www.nrel.gov/docs/fy06osti/39446.pdf
Backup Slides
44 mi.