ENDORSING PARTNERS

Transport modelling and
simulation for next
generation infrastructure
development www.isngi.org

The f...
Transport modeling and simulation
for next generation infrastructure
development:
Connecting vehicle to electricity networ...
Outline
• Background

– Next generation infrastructure and car in Japan

• Battery charging behavior
– At home
– Within tr...
Next generation infrastructure
• Council for Science and Technology Policy, Japan
states the need for next generation infr...
Areas of next generation infrastructure
• Smart energy community

– Energy management system utilizing information technol...
Passenger car ownership in Japan
Light motor passenger car (~ 0.66L)

5

Small passenger car (~2L)

4

Ordinary passenger ...
Passenger car sales ranking in Japan in
2012
Rank

1
2
3
4
5
6
7
8
9
10

Model (Automaker)
Prius (Toyota)
Aqua (Toyota)
Mi...
Electric vehicles and Plug-in hybrid
vehicle in Japan

i-MiEV
2009

Leaf
2010

Prius plug-in hybrid
2012

More energy effi...
Battery charging at home
• Analysis on charge timing choice behavior of
plug-in hybrid vehicles in Toyota City, Japan
– Th...
Smart Melit(Smart Mobility & Energy
Life in Toyota City)project
–
–
–
–

Toyota City, Japan
67 new houses, some with plug-...
Smart house

Visualization by HEMS
(home energy
management system)

PHV charger

PHV

DRP (demand
response point) portal

...
Example of electricity demand curve
PHV charge

Summer

Air cond.

PHV charge

Winter
Air cond.

Heat pump
water heater Sc...
DRP (demand response point)
• Peak pricing by point system
• Low at daytime (solar energy) &
high at evening (more activit...
Distribution of returning home timing
No charge

Charge

Time of day

Many cars return home at 18 to 20 o’clock,
which pot...
Charging time is shifted by
demand response point system
Other

Cheapest
period

Just after
came home

With demand
respons...
Charge timing choice model
• Multinomial logit model

No charge

Just after
came home

Cheapest
timing
Cheapest timing
bef...
Charge timing choice model
Alternative

Variable
Constant
No charge Drive distance (<24 km)
Long distance dummy (>24 km)
P...
Sensitivity of the estimated model
Base case: High energy conscious male driver
returned home in evening after 10 km drive...
Battery charging within trip
• The timing of mid-trip electric vehicle charging
– This is a part of the results obtained b...
Locations
2000
1800
1600
1400
1200
1000
800
600
400
200
0
2013.06
2013.07

2013.04

2012.06
2012.07
2012.08
2012.09
2012.1...
Trade-off between battery size and fast
charger density

How to optimize battery size & fast charger
deployment?
• Drivers...
Data
•
•
•
•

Investigator: Japan Automobile Research Institute
Sample: 252 company cars & 247 private cars
Survey period:...
Distribution of SOC at normal charge
20%

Company cars
個人車両(N=492)
Private cars
法人平均(N=252)

18%
16%
14%
12%
10%
8%
6%
4%
...
Distribution of SOC at fast charge
25%
20%

法人平均(N=252)
Company cars
個人車両(N=492)
Private cars

15%
10%
5%
0%
~10% ~20% ~30...
Stochastic frontier model of SOC at
fast charging within trip
• Driver avoids running out of power
Actual remaining electr...
Distribution of subjective minimum
and actual remaining charge
personal-use vehicles on working day
9

actual

Percent of ...
Distribution of subjective minimum
and actual remaining charge
commercial-use vehicles on working day
Percent of samples (...
Vehicle to grid
• Impact of electric vehicles on electricity
demand curve in Nagoya, Japan

– This is a part of the resear...
Time axis

Micro simulation model of individual’s
activity-travel pattern
【Staying】

<Activity Choice>
Staying

Commuting ...
Interaction between activity-based
model & dynamic traffic assignment
Time axis

Equilibrium state is calculated
【Staying】...
Scenario analysis at Nagoya, Japan
 Nagoya Metropolitan Area
<Population in 2020>
Over 8.0 million
<Zone>
520 zones
(Nago...
Distribution of electricity demand
Electricity demand curve
2.5

Spatial distribution of electricity demand

GWh/h
Total

...
Scenarios for EV charging/discharging
• Case 0: No EVs
• Case 1: Charge at home immediately after
returning home
• Case 2:...
Impact on electricity demand curve
2.5

GWh/h
Case_0
No EVs

Case_1
Charge

at home
Case_2 at home midnight
Charge
Case_3 ...
Spatial distribution of impact
CBD area

Suburban area

500

-200

400

-400

300

-600

200

-800

100
0

-1,000

1,000

...
Conclusions
• Battery charging at home causes significant
electricity demand, but the timing can be
controlled by peak pri...
Upcoming SlideShare
Loading in …5
×

SMART International Symposium for Next Generation Infrastructure: Transport modelling and simulation for next generation infrastructure development

521 views
401 views

Published on

A presentation conducted by Professor Toshiyuki Yamamoto, EcoTopia Science Institute, Nagoya University. Presented on Wednesday the 2nd of October 2013.

Transport system is one of the fundamental infrastructures, and a significant source of energy consumption and environmental load. Electric vehicles are expected to reduce the latter negative effects, and improve the functionality of the transport system. However, the limited driving mileage of the electric vehicles requires different electric charging system than gasoline refueling system, and the lower driving performance might potentially
cause negative effects on traffic flows. An electric charging behavior model is developed and the efficiency in battery
capacity usage is investigated in this study. We also developed a simulation system to investigate electricity demand distribution across time of day and space within urban area considering charging behavior of electric vehicles. In addition, the effects of introducing electric micro-cars into traffic are investigated by microscopic traffic
simulations considering lower maximum speed of the micro-cars. These modeling and simulation tools enable rigorous evaluations of future transport system.

Published in: Technology, Business
0 Comments
2 Likes
Statistics
Notes
  • Be the first to comment

No Downloads
Views
Total views
521
On SlideShare
0
From Embeds
0
Number of Embeds
1
Actions
Shares
0
Downloads
0
Comments
0
Likes
2
Embeds 0
No embeds

No notes for slide

SMART International Symposium for Next Generation Infrastructure: Transport modelling and simulation for next generation infrastructure development

  1. 1. ENDORSING PARTNERS Transport modelling and simulation for next generation infrastructure development www.isngi.org The following are confirmed contributors to the business and policy dialogue in Sydney: • Rick Sawers (National Australia Bank) • Nick Greiner (Chairman (Infrastructure NSW) Monday, 30th September 2013: Business & policy Dialogue Tuesday 1 October to Thursday, 3rd October: Academic and Policy Dialogue Presented by: Professor Toshiyuki Yamamoto, EcoTopia Science Institute, Nagoya University www.isngi.org
  2. 2. Transport modeling and simulation for next generation infrastructure development: Connecting vehicle to electricity network Toshiyuki Yamamoto Nagoya University 2
  3. 3. Outline • Background – Next generation infrastructure and car in Japan • Battery charging behavior – At home – Within trip • Vehicle to grid – Impact on electricity demand curve • Conclusions 3
  4. 4. Next generation infrastructure • Council for Science and Technology Policy, Japan states the need for next generation infrastructure • Features of next generation infrastructure – Smart: information technology to forecast, control and optimize infrastructure system – System: value added as system in addition to strength of products and technology itself – Global: business strategy toward global deployment 4
  5. 5. Areas of next generation infrastructure • Smart energy community – Energy management system utilizing information technology – Renewable energy, decentralized generating plant, etc. • Intelligent transport system – Communication networking among people, vehicles and road utilizing information technology – Navigation system, car sharing, LRT, etc. • Next generation infrastructure in other areas – Water supply, goods distribution, medical care, etc. – Integrated system 5
  6. 6. Passenger car ownership in Japan Light motor passenger car (~ 0.66L) 5 Small passenger car (~2L) 4 Ordinary passenger car (over 2L) 3 2 1 2010 2005 2000 1995 1990 1985 1980 1975 1970 0 1965 10 million vehicles 6 Year Source: MLIT 6
  7. 7. Passenger car sales ranking in Japan in 2012 Rank 1 2 3 4 5 6 7 8 9 10 Model (Automaker) Prius (Toyota) Aqua (Toyota) Mira (Daihatsu) N BOX (Honda) Fit (Honda) Wagon R (Suzuki) Tanto (Daihatsu) Move (Daihatsu) Alto (Suzuki) Freed (Honda) HV: hybrid vehicle Sales 317,675 266,567 218,295 211,156 209,276 195,701 170,609 146,016 112,002 106,316 Engine type HV HV Light motor Light motor Small / HV Light motor Light motor Light motor Light motor Small / HV Source: Nikkei Newspaper 7
  8. 8. Electric vehicles and Plug-in hybrid vehicle in Japan i-MiEV 2009 Leaf 2010 Prius plug-in hybrid 2012 More energy efficient, but more electricity dependent 8
  9. 9. Battery charging at home • Analysis on charge timing choice behavior of plug-in hybrid vehicles in Toyota City, Japan – This is a part of the results obtained by joint research with Toyota Motor Corporation 9
  10. 10. Smart Melit(Smart Mobility & Energy Life in Toyota City)project – – – – Toyota City, Japan 67 new houses, some with plug-in hybrid Prius HEMS (Home Energy Management System) DRP (demand response point) system 10
  11. 11. Smart house Visualization by HEMS (home energy management system) PHV charger PHV DRP (demand response point) portal PHV charging monitor 11
  12. 12. Example of electricity demand curve PHV charge Summer Air cond. PHV charge Winter Air cond. Heat pump water heater Scheduled to fill-up at 4:00 12
  13. 13. DRP (demand response point) • Peak pricing by point system • Low at daytime (solar energy) & high at evening (more activity at home) Feb. Example of DRP May Aug. Nov. 13
  14. 14. Distribution of returning home timing No charge Charge Time of day Many cars return home at 18 to 20 o’clock, which potentially cause peak demand 14
  15. 15. Charging time is shifted by demand response point system Other Cheapest period Just after came home With demand response point W/O demand response point 15
  16. 16. Charge timing choice model • Multinomial logit model No charge Just after came home Cheapest timing Cheapest timing before the next vehicle use – 12 Prius plug-in hybrid vehicles – 2011/10/1 to 2012/10/31 – 4615 cases Other Didn’t change the setting of ontimer previously set, or by mistake 16
  17. 17. Charge timing choice model Alternative Variable Constant No charge Drive distance (<24 km) Long distance dummy (>24 km) Price for energy conscious person Just after Price for energy unconscious person came home Return home at daytime (9-16) Constant Price for energy conscious person Cheapest Price for energy unconscious person time Housewife dummy Return home at evening (17-23) Constant Other Return home at evening (17-23) Same as the last charge dummy Log-likelihood (0) Log-likelihood at convergence Adjusted rho-square Coef. 1.34 ** -0.10 ** -0.38 ** -0.044 ** -0.065 ** 0.70 ** -0.69 ** -0.016 ** 0.001 0.66 ** 1.41 ** -0.96 ** 0.65 ** 2.21 ** -5774 -4415 0.233 17 ** 1%, * 5%
  18. 18. Sensitivity of the estimated model Base case: High energy conscious male driver returned home in evening after 10 km drive Electricity price No DRP (20.9 JPY) Evening price 20.9 -> 28 JPY + Midnight price 20.9 -> 10 JPY No charge 35% 36% 34% Just after came home Cheapest timing Other 47% 8% 38% 18% 19% 7% 42% 18% Charge timing is easier to change than the timing of air conditioner usage, etc. 18
  19. 19. Battery charging within trip • The timing of mid-trip electric vehicle charging – This is a part of the results obtained by the Project Consigning Technology Development for Rational Use of Energy (Promotion of aggregation and sharing of probe information) – The dataset was provided by Japan Automobile Research Institute (JARI) 19
  20. 20. Locations 2000 1800 1600 1400 1200 1000 800 600 400 200 0 2013.06 2013.07 2013.04 2012.06 2012.07 2012.08 2012.09 2012.10 2012.11 2012.12 2013.01 2013.02 2011.12 2012.01 2012.02 2012.03 2012.04 2011.09 2011.10 2011.03 2011.04 2010.12 2010.10 2010.04 2009.11 Fast charger deployment in Japan Year.month Source: CHAdeMO Association 20
  21. 21. Trade-off between battery size and fast charger density How to optimize battery size & fast charger deployment? • Drivers charge battery before empty • Charging behavior should be understood 21
  22. 22. Data • • • • Investigator: Japan Automobile Research Institute Sample: 252 company cars & 247 private cars Survey period: 2 years (2011.2-2013.1) Survey area: 42 out of 47 prefectures in Japan • Built-in data logger with GPS & communication unit: clock time, location, vehicle state (driving, normal charging, fast charging), odometer reading, use of air-conditioner & heater, state of charge 22
  23. 23. Distribution of SOC at normal charge 20% Company cars 個人車両(N=492) Private cars 法人平均(N=252) 18% 16% 14% 12% 10% 8% 6% 4% 2% 0% ~10% ~20% ~30% ~40% ~50% ~60% ~70% ~80% ~90% 90%~ SOC: state of charge Company cars are charged at the end of the working hours 23 regardless of SOC
  24. 24. Distribution of SOC at fast charge 25% 20% 法人平均(N=252) Company cars 個人車両(N=492) Private cars 15% 10% 5% 0% ~10% ~20% ~30% ~40% ~50% ~60% ~70% ~80% ~90% 90%~ SOC: state of charge Battery capacity is not fully utilized 24
  25. 25. Stochastic frontier model of SOC at fast charging within trip • Driver avoids running out of power Actual remaining electricity to start charging = ≥ Subjective minimum electricity + • Inefficiency is added to minimum electricity Actual remaining electricity Subjective minimum electricity Inefficiency • Stochastic cost frontier model is applied 25
  26. 26. Distribution of subjective minimum and actual remaining charge personal-use vehicles on working day 9 actual Percent of samples (%) 8 minimum 7 6 5 4 3 2 1 0.2 0.8 1.4 2.0 2.6 3.2 3.8 4.4 5.0 5.6 6.2 6.8 7.4 8.0 8.6 9.2 9.8 10.4 11.0 11.6 12.2 12.8 13.4 14.0 14.6 15.2 15.8 0 Remaining electricity (kWh) • Subjective minimum remaining charge has peak at 3.6kWh • 1.5kWh of average inefficiency is estimated 26
  27. 27. Distribution of subjective minimum and actual remaining charge commercial-use vehicles on working day Percent of samples (%) 8 actual 7 minimum 6 5 4 3 2 1 0.2 0.8 1.4 2.0 2.6 3.2 3.8 4.4 5.0 5.6 6.2 6.8 7.4 8.0 8.6 9.2 9.8 10.4 11.0 11.6 12.2 12.8 13.4 14.0 14.6 15.2 15.8 0 Remaining electricity (kWh) • Same peak of minimum remaining charge • Larger (1.8kWh) average inefficiency is estimated 27
  28. 28. Vehicle to grid • Impact of electric vehicles on electricity demand curve in Nagoya, Japan – This is a part of the research results funded under the Environmental Research and Technology Development Fund by Ministry of the Environment 28
  29. 29. Time axis Micro simulation model of individual’s activity-travel pattern 【Staying】 <Activity Choice> Staying Commuting Commuting to school to work Private Business Returning home 【Returning home】 Railway 9 p.m. 【Private】 Walking 【Staying】 <Destination Choice> Destination Destination 1 2 ・・・・・ Destination 5 p.m. 【Staying】 s 2 p.m. <Mode Choice> Car Railway Route Route Bus 【Staying】 Bicycle & Walking Another Office <Route Choice> 1 2 ・・・・・ Route k 【Business】 Walking 8 a.m. 【Staying】 【to Work】 Restaurant Railway Work place Home • Nested logit model of activity type, destination, mode and rail route choice at each time period • Sequence of activity-travel pattern is simulated 29
  30. 30. Interaction between activity-based model & dynamic traffic assignment Time axis Equilibrium state is calculated 【Staying】 OD demand 【Returning home】 Railway 9 p.m. 【Private】 Walking 【Staying】 5 p.m. 【Staying】 2 p.m. 【Staying】 【Business】 Walking Another Office 【to Work】 Restaurant Railway 8 a.m. 【Staying】 Work place Home Travel pattern OD travel time Link travel speed • Parking location, parking duration, and SOC of each EV are simulated along time of day 30
  31. 31. Scenario analysis at Nagoya, Japan  Nagoya Metropolitan Area <Population in 2020> Over 8.0 million <Zone> 520 zones (Nagoya City is divided into 259 Average area is 1.3km2) Osaka Tokyo Nagoya <Road Network> Link:22,466 Nagoya City Node:7,600 10% of vehicles are assumed to be replaced by EV, which means 472,000 EVs 31
  32. 32. Distribution of electricity demand Electricity demand curve 2.5 Spatial distribution of electricity demand GWh/h Total 2.0 Home Office Store 1.5 Other 1.0 0.5 0.0 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 Time of day • Total demand is about 31GWh • Higher demand at midday and at CBD 32
  33. 33. Scenarios for EV charging/discharging • Case 0: No EVs • Case 1: Charge at home immediately after returning home • Case 2: Charge at home during midnight • Case 3: Charge at workplace immediately after arriving at work • Case 4: Charge at home during midnight and discharge at workplace during daytime (until the remaining charge at 5 kWh) 33
  34. 34. Impact on electricity demand curve 2.5 GWh/h Case_0 No EVs Case_1 Charge at home Case_2 at home midnight Charge Case_3 at workplace Charge Charge Case_4 & discharge 2.0 1.5 1.0 0.5 0.0 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 Time of day • 0.1GW of daytime demand (9:00 to 16:00) can be cut by vehicle to grid at workplace 34
  35. 35. Spatial distribution of impact CBD area Suburban area 500 -200 400 -400 300 -600 200 -800 100 0 -1,000 1,000 600 800 500 600 400 400 300 200 200 0 100 0 Charge/Discharge (kWh/h) 0 No. of the EV Parked (vehicles) 600 700 200 Charge/Discharge (kWh/h) No. of the EV Parked (vehicles) 700 -200 3 4 5 6 7 8 9 1011121314151617181920212223242526 tiem period 3 4 5 6 7 8 9 1011121314151617181920212223242526 time period Home Home Workplace Other Charge/Discharge Workplace Other Charge/Discharge • Impact is quite different between CBD and suburban area 35
  36. 36. Conclusions • Battery charging at home causes significant electricity demand, but the timing can be controlled by peak pricing • Battery capacity is not fully utilized, and measures to improve efficiency are needed • Potential to cut down peak demand by vehicle to grid at workplace 36

×