1

European Association for the Development of Renewable Energies, Environment and Power Quality (EA4EPQ)

International Conference on Renewable Energies and Power Quality (ICREPQ’10)

Granada (Spain), 23rd to 25th March, 2010

Preparation of full paper for the International Conference on Renewable Energies and Power Quality

A. Author1, B. Author2 and C. Author1

1 Department of Electrical Engineering E.T.S.I.I., Vigo University

Campus of Lagoas – Marcosende, 36310 Vigo (Spain) Phone/Fax number:+0034 986 812685, e-mail: A.Author@uvigo.es, C.Author@uvigo.es

2 European Association for the Development of Renewable Energies, Environment and Power Quality (EA4EPQ)

Mailing address Phone, fax, e-mail

Abstract. The abstract in placed in the left-hand column and starts at about 11 cm from de top of the page. It should give concise information on the aims, the methods, the results and the conclusions of the paper. Its length should not exceed 200 words.

Key words Please write the main key words of your text (3 to 5) separated by commas. 1. Introduction Authors of accepted abstracts will be requested to submit a full paper for inclusion in the International Conference on Renewable Energies and Power Quality CDRom and also an extended abstract of one sheet. The full papers are due January 20, 2010. The full papers have to be sent as a pdf-file or a word file (.doc) to one of the following e-mail addresses: icrepq@icrepq.com, icrepq@uvigo.es or: donsion@uvigo.es Please carefully read and follow these instructions to ensure a uniform quality and appearance of all contributions. Clearly explain the nature of the problem, previous work, purpose and contribution of the paper. 2. Page Layout Papers should not exceed six double column A4 (21x29.7cm) pages including figures, tables and diagrams. On the last page try to balance both columns.

A. Language English (check carefully for correct grammar and spelling) B. Margins Set the top and bottom margins to 2 cm, and the left and right margins to 1.7 cm. The spacing between the two columns should be 1.25 cm

1) Justification: Full 2) Headers and Footers: None 3) Page Numbering: None

C. Type Sizes Manuscripts must be typed with a proportional serif typeface such as Time New Roman. The type sizes to be used for the different parts in the paper are shown in Table I.

Table I. - Type Sizes

SIZE APPEARANCE

Title 14 pt Bold centred Author 11 pt Centred Abstract 9 pt Main text 10 pt Primary heading 12 pt Bold, left Secondary headings 10 pt Italic, flush left Tertiary headings 10 pt Italic, indented Tables 9 pt Centred Table and figure captions 9 pt Centred References 9 pt

D. Line Spacing Singled spaced, with a double line space between paragraphs. Allow an extra half space above a line

Deterministic and Probabilistic Assessment of theImpact of the Electrical Vehicles on the Power Grid

E. Valsera-Naranjo1, A. Sumper1,2, P. Lloret-Gallego1, R. Villafafila-Robles1, A. Sudria-Andreu1,21Centre d’Innovacio Tecnologica en

Convertidors Estatics i Accionaments(CITCEA-UPC), Departament d’Enginyeria Electrica,Universitat Politecnica de Catalunya.

EU d’Enginyeria Tecnica Industrial de Barcelona,Comte d’Urgell, 187; 08036 Barcelona, Spain. http://www.citcea.upc.edu

2IREC Catalonia Institute for Energy ResearchJosep Pla, B2, Pl. Baixa. 08019 Barcelona, Spain

http://www.irec.cateduard.valsera@citcea.upc.edu, sumper@citcea.upc.edu, lloret@citcea.upc.edu,

roberto.villafafila@citcea.upc.edu, sudria@citcea.upc.edu

Abstract—This paper analyzes the impact of the charge of EVs(Electrical Vehicles) on a power grid. In order to simulate thebehavior of the EV charging on the grid a model of its batteryhas been found (EV’s charging curve). As the way of the EVis charged from the grid affects critically to the voltage levelsand to the saturation of the lines two modalities of charge arepresented: not-controlled charge and controlled charge. Finally,the simulations have been performed from two points of view:deterministic analysis and probabilistic analysis.

I. INTRODUCTION

The recent awareness about fossil fuels and the environmenthas arisen more sustainable alternatives regarding means oftransport. Consequently, hybrid vehicles and pure electricalvehicle have become the main alternatives for green trans-portation. This new trend has caused market activation and itis expected that hybrid and electrical vehicles will constitutethe majority in private transport.

Moreover, one kind of hybrid vehicles, the plug-in electricvehicles (PHEV), will not only charge their batteries, butPHEV will also be able to inject power to the networkwhen required, as the electrical vehicles (EV) do. This factsuggests that EV penetration will affect current power systemperformance. Then, it is necessary to study some scenariosof penetration of such vehicles into the electrical network inorder to maintain security and quality of power supply withinstandard limits.

On one hand, in order to analyze the impact of EVsintegration into the power grid, aspects related to storagetechnologies (namely batteries) and the charging process ofsuch storage devices have to be studied. In order to analyze

the impact of EV on the power grid, load flow calculations ina standard urban network are performed following. To do that,different penetrations of EV into the network are considered.The results of simulations are shown and commented lately.

On the other hand, an European standard which defines thetype of connector to be used by EVs is needed. Until a newstandard gets develop, it is proposed to use a SCHUKO (CEE7/4) connector type for currents up to 16A. In addition, to slowcharging, the output values of the charging station should beup to 16 A per plug, 230 V ± 10%, and 50 Hz ± 1% [1].

II. ELECTRICAL VEHICLE

In order to analyze the impact of the EVs into the grid, amodel of its charging curve is needed. From all the existingmodels of EV, the charging curve chosen is from the Mit-subishi I-MIEV [2]. This vehicle has an autonomy of 160 kmand its batteries are made of Li-ion (50Ah 16kWh 330V) [2];Fig.1 shows the charging curve of the EV when its capacity at20%. The battery is charging at full power during 2 hours andthen decreases its charging power exponentially since the hour4; at this time it is considered that the battery is completelycharged.

The way of how EVs are charged from the grid has acritical influence on its impact on the voltage levels and on thesaturation of the lines. Considering this fact, two modalitiesof charge are presented: not-controlled charge and controlledcharge. In the not-controlled charge EVs start the charge asthey park. In the controlled charge, EVs only can charge duringa determined period of time of the day (low-load period).In both modalities of charge, the number of EVs charging

https://doi.org/10.24084/repqj08.704 1505 RE&PQJ, Vol.1, No.8, April 2010

80

100

[%]

60

80

ng pow

er 

20

40

Chargin

0

0 0,5 1 1,5 2 2,5 3 3,5 4

Time [h]

Fig. 1. Charging curve of the EV

is defined establishing different scenarios of EV penetrationover the vehicles acceding to the city (Fig.2). The curve ofthe movement of the vehicles has been obtained from anestimation of the curve of Barcelona [3] and has been shiftedtwo hours earlier in order have a better representation of theDenmark’s vehicle movement (in this estimation, populationhas also been considered).

700

800

400

500

600

EV

100

200

300

0

0:00

0:30

1:00

1:30

2:00

2:30

3:00

3:30

4:00

4:30

5:00

5:30

6:00

6:30

7:00

7:30

8:00

8:30

9:00

9:30

10:00

10:30

11:00

11:30

12:00

12:30

13:00

13:30

14:00

14:30

15:00

15:30

16:00

16:30

17:00

17:30

18:00

18:30

19:00

19:30

20:00

20:30

21:00

21:30

22:00

22:30

23:00

23:30

Hour

Fig. 2. Movement of vehicles

During the charging process of the EV, will be chargingsimultaneously EV with different states of charge. Therefore,a model is needed to take this fact into account in the steady-state simulations. Eq.1 models the superposition of the demandcaused by the EV in different charging states, where R ispenetration of EV, Pev is the maximum power which a carcan charge, i is the counter associated to the actual hour, j isthe counter associated to the previous hour, ∆vak,k−1 is theincrease of EV between the hour i and the hour j and Cevkis the charging state of the EV (Fig.1).

PHi = R · Pev ·i∑

k=1

(∆vak,k−1 · Cevk) (1)

Applying this model to the vehicle movement curve, the de-mand curves for the both charging modalities can be obtained(Fig.3)

III. POWER GRID

The scenario of the simulations is a part of the Danishsample grid (Fig.4) [4]. This grid, of 400 MW of short-circuitpower, has 3 wind turbine generation units of 630 kW atbusbars B013, B015 and B017. In addition, there are threecombined cycle units of 3 MW each one at the busbar B005,but for purposes of the study has remained disconnected.Loads are at the busbars B005, B010, B011 and B012. Due tosimplify the interpretation of the simulations and the results,EVs have been separated from the other consumptions.

35000

40000

45000

2,50000

3,00000

ent

[MW] EV's charge power (R= 0.1)

EV's charge power (R= 0.2)Vehicle movement

20000

25000

30000

1,50000

2,00000

e movem

e

ge pow

er 

5000

10000

15000

0,50000

1,00000

Vehicle

EV's cha

rg

00,00000

0:00 1:00 2:00 3:00 4:00 5:00 6:00 7:00 8:00 9:00 10:0011:0012:0013:0014:0015:0016:0017:0018:0019:0020:0021:0022:0023:00

E

Hour

(a) Not-controlled charge

35000

40000

45000

12

14

16

ent

 [MW]

20000

25000

30000

6

8

10

e movem

ge pow

er

5000

10000

15000

2

4

6

Vehicle

EV's cha

rg

EV's charge power (R= 0.1)EV's charge power (R= 0.2)Vehicle movement

00

0:00

0:30

1:00

1:30

2:00

2:30

3:00

3:30

4:00

4:30

5:00

5:30

6:00

6:30

7:00

7:30

8:00

8:30

9:00

9:30

10:00

10:30

11:00

11:30

12:00

12:30

13:00

13:30

14:00

14:30

15:00

15:30

16:00

16:30

17:00

17:30

18:00

18:30

19:00

19:30

20:00

20:30

21:00

21:30

22:00

22:30

23:00

23:30

E

Hour

(b) Controlled charge

Fig. 3. EV’s power demand for the not-controlled and controlled chargemodality

L00

External Grid B001

B005

B009B008B007

B010

L01EV00

EV01

CHP

B006

B011

L02 EV02 L03 EV03

WT

B012 B013

WT

B014 B015

WT

B016 B017

B002 B003 B004

Fig. 4. Danish power grid

In order to study the most critical situation the Winter(week)case has been chosen from all the possible situations ofelectrical demand because it has the highest demand in everyhour (Fig.5). Therefore, the wind power generation curve isfrom this period. Fig.6 depicts a representative daily generationfor the three wind power generation units.

14Winter (week)

Summer (week)

Winter (weekend)

10

12( )

Summer (weekend)

8

10

W]

6

Power [M

W

4

2

0

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23

Hour [h]

Fig. 5. Electrical demand

In the deterministic and in the probabilistic analysis thevoltage results presented are from two busbar. The first busbaris B005 and it has been chosen because L00 is always the loadwith the highest power. The second busbar chosen is B014 inorder to have the behavior of a closer busbar to a wind power

2https://doi.org/10.24084/repqj08.704 1506 RE&PQJ, Vol.1, No.8, April 2010

0,75W]

0,69

0,71

0,73

tion

 [MW

0,63

0,65

0,67

r Gen

erat

0,57

0,59

0,61

ind Po

we

0,55

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23

Wi

Hour

Fig. 6. Wind power generation

generation unit.

IV. DETERMINISTIC ANALYSIS

In this section, the result from executing a deterministicanalysis are shown. In this kind of analysis all loads arepreviously determined. The results from the two typologies ofcharge for the EV are presented for the different penetrationsof EV.

A. Not-controlled charge

Fig.7 and Fig.8 show respectively the voltage levels forbusbar B005 and B014 for the case of not-controlled charge.

1

1,005

0,995

1

0,99

u]

0,985

Voltage

 [pu

0,975

0,98

0,97 EV 0%

EV 10%

0,965

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

Hour [h]

EV 20%

Fig. 7. Voltage levels for the busbar B005 for the different penetrations ofEV in the case of not-controlled charge for the deterministic analysis

B. Controlled charge

Fig.9 and Fig.10 show respectively the voltage levels forbusbar B005 and B014 for the case of controlled charge.

V. PROBABILISTIC ANALYSIS

From the compiled data during the week days of the winterseason a mean and an standard deviation have been foundfor each hour of the day. In order to perform a probabilisticanalysis 50 load flows for each hour have been executedgenerating random powers with the mean and the standarddeviation founded previously.

1

1,005

0,995

1

0,99

u]

0,985

Voltage

 [pu

0,975

0,98

0,97 EV 0%

EV 10%

0,965

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

Hour [h]

EV 20%

Fig. 8. Voltage levels for the busbar B005 for the different penetrations ofEV in the case of not-controlled charge for the deterministic analysis

1,05

1

0,95

u]

0,9

Voltage

 [pu

0,85

0,8

EV 0%

EV 10%

0,75

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

Hour [h]

EV 20%

Fig. 9. Voltage levels for the busbar B005 for the different penetrations ofEV in the case of controlled charge for the deterministic analysis

1,05

1

0,95

u]

0,9

Voltage

 [pu

0,85

0,8

EV 0%

EV 10%

0,75

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

Hour [h]

EV 20%

Fig. 10. Voltage levels for the busbar B005 for the different penetrations ofEV in the case of controlled charge for the deterministic analysis

3https://doi.org/10.24084/repqj08.704 1507 RE&PQJ, Vol.1, No.8, April 2010

A. Not-controlled charge

Fig.11 and Fig.12 show mean voltage levels for the differentpenetrations of EV in the case of not-controlled charge. Inboth figures are the calculated mean voltage levels from theprobabilistic analysis and the mean from the deterministicanalysis. In order to have a most detailed view of the criticalbusbar, in this case B014, Fig.13 show the the range of voltagelevel that can be reached at this busbar.

1,005

0 995

1

0,99

0,995

u]

0,985

Voltage

 [pu

0,98

EV 0% (mean)

0,975

EV 10% (mean)

EV 20% (mean)

EV 0% (measured mean)

EV 10% (measured mean)

0,97

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

Hora [h]

EV 20% (measured mean)

Fig. 11. Voltage levels for the busbar B005 for the different penetrations ofEV in the case of not-controlled charge

1,005

0 995

1

0,99

0,995

u]

0,985

Voltage

 [pu

0,98

EV 0% (mean)

0,975

EV 10% (mean)

EV 20% (mean)

EV 0% (measured mean)

EV 10% (measured mean)

0,97

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

Hora [h]

EV 20% (measured mean)

Fig. 12. Voltage levels for the busbar B014 for the different penetrations ofEV in the case of not-controlled charge

B. Controlled charge

Fig.14 and Fig.15 show mean voltage levels for the differentpenetrations of EV in the case of controlled charge. Ashappens in the not-controlled charge, the B014 is the busbarwith the lowest voltage levels, Fig.16 show the the range ofvoltage level that can be reached at this busbar.

VI. CONCLUSION

The objective of the paper is to analyze the impact of charg-ing EVs from a Danish grid. Therefore, different penetrationsof EV and charge modalities are proposed. In order to cover

1

1,005

0,995

0,985

0,99

u]

0,98

0,985

Voltage

 [pu

0 97

0,975

0,965

0,97

MIN

AVG

0,96

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23

Hour [h]

MAX

Fig. 13. Voltage levels for the busbar B005 for the different penetrations ofEV in the case of not-controlled charge

1,05

1

0,95

u]

0,9

Voltage

 [pu

0,85

EV 0% (mean)

0,8

EV 10% (mean)

EV 20% (mean)

EV 0% (measured mean)

EV 10% (measured mean)

0,75

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

Hora [h]

EV 20% (measured mean)

Fig. 14. Voltage levels for the busbar B005 for the different penetrations ofEV in the case of controlled charge

1,05

1

0,95

u]

0,9

Voltage

 [pu

0,85

EV 0% (mean)

0,8

EV 10% (mean)

EV 20% (mean)

EV 0% (measured mean)

EV 10% (measured mean)

0,75

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

Hora [h]

EV 20% (measured mean)

Fig. 15. Voltage levels for the busbar B005 for the different penetrations ofEV in the case of not-controlled charge

4https://doi.org/10.24084/repqj08.704 1508 RE&PQJ, Vol.1, No.8, April 2010

1

1,05

0,95

1

0,9

u]

0,85

Voltage

 [pu

0,75

0,8

0,7 MIN

AVG

0,65

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23

Hour [h]

MAX

Fig. 16. Voltage levels for the busbar B005 for the different penetrations ofEV in the case of not-controlled charge

the maximum situations two analysis have been performed:

deterministic analysis and probabilistic analysis.On one hand, results from the simulations show that not-

controlled charge amplifies the demand of the line at the hoursof the higher electrical demands. On the other hand, simula-tions show that with the controlled charge a high penetrationof EVs can be charged without the need of investments in thepower grid infrastructure.

REFERENCES

[1] Valsera-Naranjo Sumper Lloret-Gallego Villafafila-Robles, Sudria-Andreu. Electrical vehicles: state of art and issues for their connectionto the network. Electrical Power Quality and Utilisation, 2009. 1

[2] http://www.mitsubishi-motors.com/special/ev/index.html. 1[3] Dades Basiques 2008 v.02. Ajuntament de Barcelona. Direccio de Serveis

de Mobilitat., 2008. 2[4] Roberto Villafafila-Robles. Probabilistic modeling in normal operation

and the control of distribution systems with renewable source based DGunits. PhD thesis, Universitat Politecnica de Catalunya, 2009. 2

5https://doi.org/10.24084/repqj08.704 1509 RE&PQJ, Vol.1, No.8, April 2010