Electric vehicle integration in the grid: experiences from

DTU Elektro10 October 2019 Title

Asso. Prof. Mattia Marinelli, Ph.D.

[email protected]

Center for Electric Power and Energy

DTU Risø Campus

Electric vehicle integration in the grid: experiences from

the Danish V2G projects ACES and Parker

NEDO, Tokyo, 8 October 2019


Outline

• ACES and Parker projects and EV related projects at DTU

• Controlling an EV for the benefit of the grid

• Understanding the impact on the local distribution grid

• Frequency control using unidirectional and bidirectional (V2G) charging

• Consequences in term of battery degradation

• Lessons learned

2


• EDISON 2008 – 2011

• NIKOLA 2013 – 2016

• COTEVOS 2013 – 2016

• ELECTRA 2013 – 2018

• EnergyLab Nordhavn 2015 – 2019

• Parker 2016 – 2018

• ACES 2017 – 2020

• CAR 2018 – 2020

Parker

ACES

CAR

EV (related) projects

at the Center for Electric Power and Energy - DTU

3


4

Distribution Feeder in Rønne

• LV grid: 400 V

• 10/0.4 kV 400 kVA

distribution transformer

• 4 subfeeders: 110 known

load consumptions

• 8 10 kW DC chargers

• Common district heating

Budget: 10 MDKK (=1.4 M€)

Public grant (EUDP): 55 %

Equivalent person-months: 130

over 3y (04/17-03/20)

Public chargers and EVs used in

the demo:

20 Nissan Leaf and env-200

www.aces-bornholm.eu

The ACES project

Across Continents Electric vehicles Services

4


Partners: Nissan, Mitsubishi Corporation,

Mitsubishi Motors Corporation, PSA ID,

NUVVE, Frederiksberg Forsyning A/S,

Insero A/S, Enel and DTU.

Duration: August 2016 to January 2019.

Budget: Two million euros, funding by

EUDP

Thomas Parker, 1843 –

1915

Demonstrate that

contemporary electrical

vehicles can participate in

advanced smart grid

services.

More info:

www.parker-project.com

5


On-board charger

24 - 30 - 40 kWh battery

BMS

+/- 10 kW

3-phase

Grid power

Battery power

3.7 kW

1 phase

• Charging options considered:

• 10 kW DC via external charger bidirectional control

• 3.7 - 11 - 22 kW AC via internal charger unidirectional control

3.7 - 11 - 22 kW

1; 3-phase

6

Investigated AC and DC (V2G) charging/control options

for electric vehicles


• historical driving characteristics of private conventional vehicles from Denmark

• home plug-in behavior of EVs from Japan

7

L. Calearo, A. Thingvad, K. Suzuki, M. Marinelli, “Grid Loading

due to EV Charging Profiles Based on Pseudo-Real Driving

Pattern and User Behaviour,” Transportation Electrification

Transaction, vol. 5, Sep 2019

• Considering realistic

driving/charging behaviour, what’s

the expected loading impact on two

representative distribution grids

assuming a 100% EV scenario?

• If reinforcement is necessary, what

would be a fair value for a load

deferring?

Getting to the local grid – the driving behavior matters

7



Mean distribution: 42% phase a,

33% phase b, 25% phase c.

Assumed load: 40% in phase a,

30% in phase b, 30% in phase c.

• LV grid: 400 V

• 10/0.4 kV 400 kVA

• 4 subfeeders: 127

known load

consumptions

• Common electric

heating (40% of the

houses)

8


• Higher rated power of the

chargers less EVs

charging at same time, but

higher peak consumption.


• In the most conservative scenario, we assumed that the potential plug in

events are split in 4 sets of 25% EVs each: 16:00, 17:00, 18:00 and

19:00. EVs are equally distributed on phases (for the 3.7 kW).

• Throughout the whole study NO work-charging nor public charging is

considered (each customer will solely charge at home).

• Single-phase chargers (3.7 kW); Three-phase chargers (11.1 kW)

9


Aggregator

Grid Services

Energy to/from

Electric GridOCPP based protocol

Bi-directional

Energy Flow

VSL Vehicle System Link Software – Software Defined Charging Station Control signals Bi-directional energy flow

Bi-directional energy flow

CHAdeMo 2.0

Standardized protocol

Distributed Energy

Storage

Frederiksberg Forsyning and Bornholm configurations

10


• The droop is the measure of how

much the machine is sensitive to

frequency changes and is the value

that quantify its contribution to

primary frequency/power regulation.

• ൗ𝛥𝑓𝑓𝑛𝑜𝑚 = −𝑘𝑑𝑟𝑜𝑜𝑝 Τ𝛥𝑃

𝑃𝑛𝑜𝑚

The most straightforward control logic for providing frequency control is to use droop (pure proportional)

controllers – as commonly done in conventional power plants.

Depending on the capability of the device, the characteristic can be symmetric or not and use either

bidirectional (V2G) or unidirectional power flow.

11

Primary Frequency Control – bidirectional (V2G) or

unidirectional flow


A. Thingvad, C. Ziras, M. Marinelli, “Economic Value of Electric Vehicle

Reserve Provision in the Nordic Countries under Driving Requirements and

Charger Losses,” Journal of Energy Storage, Vol 21, 2019.

12

Realizing bidirectional (V2G) control with commercial

vehicles in Denmark – 14 h/day provision


https://brokergraphs.syslab.dk/d/AcNHuXFZk/electric-vehicle-data?orgId=1

Equipment used:

EVSE: charger controller 6-32 A 3-ph

EV: Tesla model S

Controller running:

Frequency control (remote meas)

Response time around 5-6 seconds (mostly

on the EV side)

13

Realizing unidirectional frequency control with private

vehicles in Bornholm (DK) – ACES smart charger


Assessing the impact in term of degradation – how

much driving and V2G services affect battery life

The characteristics of a 40 kWh battery with NMC cells are taken in account.

The EV is subject to a driving pattern of 45 km/day (average driven distance

in DK = 9kWh)

The EV provides 14 h/day of frequency regulation with a ±10 kW reserve

L. Calearo, A. Thingvad, M. Marinelli, “Modelling of Electric Vehicles for Degradation Studies,”

Universities Power Engineering Conference (UPEC), 2019 Proceedings of the 54th International,

pp. 1-6, Bucharest, 3 Sep. – 6 Sep. 2019

14


Quantification of calendar and cycling ageing

A. Thingvad, M. Marinelli, “Influence of V2G Frequency Services and Driving on

Electric Vehicles Battery Degradation in the Nordic Countries,” EVS 31 and

EVTeC 2018, Kobe, 30 Sep. – 03 Oct 2018

15


Calendar and cycling degradation depending on the

initial daily SOC (60 vs 80%)

Recharging at 60% Recharging at 80%

16


Total capacity degradation for 60% SoC case Total capacity degradation for 80% SoC case

➔ Combined operation has the highest degradation.

➔ 1.54% more than driving only and 3.09% more than parked only

➔ Combined operation has the highest degradation.

➔ 1.09% more than driving only and 2.76% more than parked only

Calendar and cycling degradation depending on the

initial daily SOC (60 vs 80%)

A. Lamba, “Technical and economic characterization of electric vehicle battery degradation due to grid service provision,” M.Sc. thesis in Sustainable

Energy, DTU, 7 Aug 2019 (supervisors: M. Marinelli, L. Calearo, A. Thingvad).

17


• Current market structures and framework conditions facilitate demand response/DER

participation to a lesser extent: the most beneficial aspects are hardly leveraged.

• Technology and social aspects are equally important: the focus on social practices will help the

move away from a “technology push” approach to smart Grids.

• Uncertainty in the amount of power&energy provided by heterogeneous sets of units is crucial

for large-scale applications, particularly frequency based services.

• Mimicking the response of conventional power plants when providing balancing services with

heterogonous aggregation of DERs is necessary to replace conventional units.

• Wear of equipment (EVs particularly) seems to be limited, despite heavy usage.

18

Lessons learned


STAY TUNED! –

CHECK RESULTS ON THE PROJECTS WEBSITES

• ACES: www.aces-bornholm.eu

• CAR: www.sbcar.eu

• Parker: www.parker-project.com

19

http://www.aces-bornholm.eu/

http://www.sbcar.eu/

http://www.parker-project.com/


• A. Zecchino, A. M. Prostejovsky, C. Ziras, M. Marinelli, “Large-scale Provision of Frequency Control via V2G: the Bornholm Power System Case,”

Electric power system research, vol. 170, pp. 25-34, May 2019.

• A. Thingvad, C. Ziras, M. Marinelli, “Economic Value of Electric Vehicle Reserve Provision in the Nordic Countries under Driving Requirements and

Charger Losses,” Journal of Energy Storage, Vol 21, 2019.

• L. Calearo, A. Thingvad, K. Suzuki, M. Marinelli, “Grid Loading due to EV Charging Profiles Based on Pseudo-Real Driving Pattern and User

Behaviour,” Transportation Electrification Transaction, 2019.

• L. Calearo, A. Thingvad, M. Marinelli, “Modelling of Electric Vehicles for Degradation Studies,” Universities Power Engineering Conference

(UPEC), 2019 Proceedings of the 54th International, pp. 1-6, Bucharest, 3 Sep. – 6 Sep. 2019

• M. Rezkalla, A. Zecchino, S. Martinenas, A. M. Prostejovsky, M. Marinelli, ”Comparison between synthetic inertia and fast frequency containment

control based on single phase EVs in a microgrid,” In Applied Energy, vol. 210, pp.764-775, 2017.

• S. Martinenas, K. Knezović, and M. Marinelli, “Management of Power Quality Issues in Low Voltage Networks using Electric Vehicles:

Experimental Validation,” Power delivery, IEEE Transactions on, vol. 32, no.9, pp. 971-979, Apr. 2017.

• A. González-Garrido, A. Thingvad, H. Gaztañaga, M. Marinelli, “Full-Scale Electric Vehicles Penetration in the Danish Island of Bornholm –

Optimal Scheduling and Battery Degradation under Driving Constraints,” Journal of Energy Storage, vol. 23, pp. 381-391, June 2019.

• A. Thingvad, M. Marinelli, “Influence of V2G Frequency Services and Driving on Electric Vehicles Battery Degradation in the Nordic Countries,”

EVS 31 and EVTeC 2018, Kobe, 30 Sep. – 03 Oct 2018.

• A. Zecchino, A. Thingvad, P. B. Andersen, M. Marinelli~, “Suitability of Commercial V2G CHAdeMO Chargers for Grid Services,” EVS 31 and

EVTeC 2018, Kobe, 30 Sep. – 03 Oct 2018

20

References and further readings

Documents

Electric vehicle integration in the grid: experiences from