EV research at Chalmers using GPS data from vehicles

Transcription

1 EV research at Chalmers using GPS data from vehicles Frances Sprei Assistant Professor in Sustainable Mobility Energy and Environment, Chalmers Studies in collaboration with: Sten Karlsson, Niklas Jakobsson, Lars-Henrik Björnsson (FRT, Chalmers), Patrick Plötz, Till Gnann (Fraunhofer ISI)

2 Data sets Outline Distribution of daily vehicle kilometers traveled (VKT) Is the second car in a household better suited as a BEV? Maximizing EV usage in 2 car households Next step

3 Data sets Swedish Car Movement Data (SCMD): GPS measurements of randomly sampled conventional cars from households in Western Sweden. German Mobility Panel (MoP): Household survey of individuals movements. Winnipeg data: GPS measurements of 75 household in Winnipeg Canada

4 Summary statistics of data

5

6 Distribution of daily VKT is right-skewed but how?

7 Different distributions affect the likelihood of long-distance driving

8 Estimating days per year with long-distance travel Comparison between estimate & extrapolated data

9 Goodness of fit statistics

10 Next steps PDF of the days requiring adaptation for L = 100, 150 and 200 km using kernel density estimation Looking at measurements length and its influence on distribution

11 Days Requiring Adaptation (DRA) w.r.t. range

12 Nbr of cars Days requiring adaptation range 140 km Annual VKT (10000 km) Source: Sprei, Jakobsson, Karlsson

13 Two car (+) better suited for adoption of BEVs? A household has a first car that is the preferred choice for long distance trips, and a second car that is mainly used for shorter everyday trips. Do second cars have a more suitable driving pattern for being replaced by BEVs? Are they economical as BEVs considering TCO? A household could optimize the use of a BEV so that it takes all the short trips, and the conventional car the long trips.

14 Nbr of cars Two-car households - range140 km 1 st column: 1 car 2 nd column: 1 st car 3 rd column: 2 nd car Annual VKT (1000 km) Source: Sprei, Jakobsson, Karlsson

15 CDF of DRA for different household categories (range 120 km) Swedish data German data Days requiring adaptation

16 Conclusions related to driving needs Share of cars viable as BEVs with a range of 120 km Second car: 30% First car: 10 % Single car: 15 % The share of viable second cars is 20% higher than first cars and 15% higher than single cars for a large part of the DRA spectra. Why? fewer long distance driving days - also lower annual VKT

17 Method economic analysis Annual capital costs: TCO a = a capex + a opex Annual operating costs: a i opex = VKT i c e/c k e/c + k OM a i capex = p LP i (1+p) T 1 SP i (1+p) T k tax + k renti D i

18 Parameters TCO calculated for an electric/diesel/gasoline car given the driving pattern, electric car compared to the cheapest of the conventional cars. Costs annuitized with a discount rate over 8 years. Parameters for A cost for DRA of 60 Euro per day. Swedish direct subsidy of 4400 Euro upon purchase. BEV running costs higher in Germany. Attribute Unit Sweden Germany Electricity price /kwh 0, Gasoline price /l 2, Diesel price /l 2, Battery price /kwh Investment horizon years Interest rate - 5% 5% VAT - 25% 19% Attribute Unit Sweden Germany BEV price w/o battery EUR ,500 Diesel vehicle price EUR ,400 Gasoline vehicle price EUR ,800 O&M BEV EUR/km 0, O&M Diesel EUR/km 0, O&M Gasoline EUR/km 0, Vehicle tax BEV EUR/yr 0 0 Vehicle tax Diesel EUR/yr Vehicle tax Gasoline EUR/yr Rental car cost EUR/day BEV subsidy EUR Attribute Battery capacity Depth of discharge Electric consumption Electric range Conventional consumption (gasoline) Conventional consumption (diesel) Unit kwh - kwh/km km l/km l/km Parameter %

19 Economical BEVs w.r.t. annual VKT Swedish data 1 st column: 1 car 2 nd column: 1 st car 3 rd column: 2 nd car

20 Economical BEVs w.r.t. annual VKT German data

21 Share economic BEVs w.r.t. DRA Swedish data German data

22 Next steps Improvements of parameters for TCO Looking at different ranges

23 Two cars 65 households ( 100 measured) 2 vehicles GPS logged 2-3 months measurements

24 Possible electric driving Maximize EV: % of household driving Example, Household A Electric, maximum possible Shares, electric maximum possible Total distance [km] Car 1 Car 2 electric non-electric non-served Share of total distance [%] Car 1 Car 1+2 electric non-electric non-served Original Unlimited 200km 150km 100km 60km Original Unlimited 200km 150km 100km 60km Range, electric Range, electric

25 Potential electric driving in 2 car households Possible BEV share of 2-car household driving (%) The potential increases considerably by flexible use of the car! max. BEV driving second car only not fulfilled 0 Second car Unlimited 200km 150km BEV range 100km 60km (Preliminary results for 18households)

26 Next step Replacing one of the vehicles with a BEV (VW e-golf) GPS logging both 4-5 months 20 households ( from the 65 previous) Interviews beginning and end Starting April 2015

27 THANKS! QUESTIONS?