Transition to Electric Vehicles and the Next Generation Energy System

Transcription

1 Transition to Electric Vehicles and the Next Generation Energy System Aviral Shukla Joseph Pekny Venkat Venkatasubramanian School of Chemical Engineering Purdue University

2 Conventional Petroleum Based Transportation Fuel Dwindling Reserves Green House Gas Emissions Energy Security Trade Deficits Low Internal Combustion Engine (ICE) efficiency Alternative Transportation fuels Biofuel Hydrogen Electricity 2

3 Alternative Transportation Fuels Biofuels Pros Cons Inexhaustible Produced Domestically Doesn t harm the environment 3 Food/fuel? Impact on Water Resources Loss in Biodiversity Life Cycle Analysis Sustainability? Shortage of Infrastructure Availability

4 Alternative Transportation Fuels Hydrogen Pros Abundant Can be used in internal combustion engine or fuel cell Produces water on combustion Cons Currently produced from fossil fuels Large onboard storage tanks High Cost of fuel cells Hydrogen internal combustion engine not reliable Lower overall Efficiency Lack of Refueling Infrastructure 4

5 Grid to motor Efficiency 5

6 Alternative Transportation Fuels Electric Vehicles Pros Electric motors are highly efficient High Torque at low RPM speeds Regenerative breaking Solar/Wind Electricity Well Developed Electric Grid Cons Limited Range Expensive Batteries Relative Lack of refueling infrastructure 6

7 Refueling Station Location Lower energy density of alternative fuels Size and Weight Limitations Smaller Range Chicken or Egg? Maximize return on Capital Need for directing infrastructure investments towards locations for maximum impact Mathematical programming model 7

8 Flow Interception Facility Location Model (FIFLM) Most transportation planning for activity locations at the starting or end points of a trip Useful when activity can only be done at a particular destination Discretionary Services: convenience stores, gas stations etc. Convenient Location increases attractiveness Interception: Placing a facility on a pre-planned route. 8

9 FIFLM: Formulation & Model Details x j y p 1 if a facility is located on node j = 0 otherwise 1 if at least one facility is located on path p = 0 otherwise, j N, p P Maximize n such that x m (1) j p x j y p j= 1 j, for all p P (2) 9

10 Agent Based Modeling with TRANSIMS Vehicles modeled as individual agents Flow equilibration, without shortest path assumptions No zonal aggregations Data Requirements: GIS Map, Origin-Destination Matrices, Travel Mode Information, Diurnal travel Distributions 10

11 Model Overview 11

12 Agent Based Modeling with TRANSIMS Detailed second by second data on each vehicle stored in a relational database: speed, acceleration, location, direction of motion etc. TRANSIMS output>post processing>fiflm input SQL queries for data manipulation 12

13 : Alexandria, VA 2576 Nodes 1632 links 114,943 equations 117,519 variables 13

14 Results Number of Refueling Stations 14 Vehicles Intercepted(%) $12,000, $10,000,000 Vehicles Intercepted (%) $8,000,000 $6,000,000 $4,000,000 Installation Cost $2,000, Maximum Number of Refueling Stations (m) $0 Vehicles Intercepted (%) Installation Cost

15 Comparison with Existing Infrastructure 20 gasoline filling stations 55% traffic Interception 15

16 Comparison with Existing Infrastructure 5 refueling stations 57% Interception

17 : Observations Possible to refuel more vehicles with fewer refueling stations Existing 20 gasoline stations intercept 55% traffic It is possible to intercept >55% traffic with just 5 refueling stations using FIFLM Optimization leads to as much as 75% savings over the existing gasoline stations 17

18 Retrofitting Existing Gasoline Stations Retrofitting existing gasoline stations with alternative fuel dispensing units is more cost effective FIFLM modified to calculate the percentage of vehicles intercepted by successively retrofitting gasoline stations Most of the vehicles can be intercepted with a small retrofitting investment Number of Retrofitted Stations Vehicles Intercepted(%)

19 Summary Large Scale Authentic Transportation Network Congestion Effects in Routing No zonal aggregation Microscopic Agent Based Modeling Relational Database Queries Mathematical Programming/ FIFLM 19

20 Electricity Market Stakeholders Electricity Generators Independent System Operators Consumers

21 Multi-Paradigm Modeling Why is multi-paradigm modeling needed for energy systems analysis? Multiple levels of aggregation and granularity Multiple perspectives Formalisms that best fit the structure of a problem Multiple scenarios / uncertainty management Model components can vary as model objectives change or system understanding increases

22 Multi-Paradigm Modeling of Electricity and Transportation Systems Electricity Supply Agent-based Modeling Electricity Market Industrial and Commercial Demand Historical Correlations Household Electricity Demand Discrete Event PHEV Usage Traffic Simulation Agent-based Modeling

23 Effect of PHEV adoption on Natural Gas demand in electricity generation PHEV Electricity Consumption Commercial Electricity Consumption Total Electricity Production Natural Gas without PHEV Demand Residential Electricity Consumption Industrial Electricity Consumption Natural Gas with PHEV Demand MW Hour

24 Future Work We have partnered with Indianapolis Power Light to implement the model for Marion county. s Best refueling station location Impact of electric vehicles on electricity demand 24

25 Thank You

26 Extra slides