Undergraduate: Runsha Long Mentor: Hantao Cui Professor: Fran Li

1

Minimize Total Power Loss in Distribution Network Reconfiguration Considering PEV

Charging Strategy

Undergraduate: Runsha LongMentor: Hantao CuiProfessor: Fran Li

2

Outline· PEV Introduction· Data Collection· Minimizing Power Loss

3

PEV

· Plug-in Electric Vehicles· Electricity instead of gasoline· Modeled after Nissan Leaf· Specifications:

· Range: ~84 miles· Battery Capacity: 24 kWh· Efficiency = 3.5 miles/kWh

4

PEV Charging

· Three levels of charging- Level 1: 120V/15A- Level 2: 240V/30A- Level 3: 480V/60A (not used)

5

Data Collection· Daily Travel Data· PEV Load Data· Base Load Characteristics· % PEV Penetration Characteristics

6

Daily Travel Data

· National Household Transportation Survey - nhts.ornl.gov

· 10,000 samples- Start Time- End Time- Trip Duration- Distance Traveled

7

End Times

12:00 AM 3:02 AM 5:21 AM 6:55 AM 8:28 AM 10:01 AM11:34 AM 1:07 PM 2:40 PM 4:13 PM 5:46 PM 7:19 PM 8:57 PM 10:36 PM0

100

200

300

400

500

600

700

Num

ber o

f Veh

icle

s

8

PEV Load Data

· Estimated the percentage of Levels 1 & 2 charging- 35% Level 1- 65% Level 2

· Charging assumed to start immediately upon arrival

9

Sample PEV Load Curve

12:00 AM 1:54 AM 3:48 AM 5:42 AM 7:36 AM 9:30 AM 11:24 AM 1:18 PM 3:12 PM 5:06 PM 7:00 PM 8:54 PM 10:48 PM0

1000

2000

3000

4000

5000

6000

1K2K3K4K

Kilo

Wat

ts (k

W)

10

Base Load Characteristics

· NYISO load characteristic· Scaled load curve to a population size to

180,000 – about the size of Knoxville

11

Base Load Curve

12:00 AM1:55 AM 3:50 AM 5:45 AM 7:40 AM 9:35 AM11:30 AM1:25 PM 3:20 PM 5:15 PM 7:10 PM 9:05 PM11:00 PM100

120

140

160

180

200

220

240

Base Load

Meg

aWat

ts(M

W)

12

Base Load + PEV

12:00 AM1:55 AM 3:50 AM 5:45 AM 7:40 AM 9:35 AM11:30 AM1:25 PM 3:20 PM 5:15 PM 7:10 PM 9:05 PM11:00 PM100

120

140

160

180

200

220

240

Base Load10% PEV20% PEV

Meg

aWat

ts(M

W)

13

Base Load + PEV

12:00 AM1:55 AM 3:50 AM 5:45 AM 7:40 AM 9:35 AM11:30 AM1:25 PM 3:20 PM 5:15 PM 7:10 PM 9:05 PM11:00 PM0

50

100

150

200

250

300

350

400

Load Curve

Base Load10% PEV20% PEV50% PEV100% PEV

Load

(MW

)

14

Minimizing Power Loss· Reconfiguration Model· Delay Strategy

15

Network Reconfiguration

· Distribution network are typically constructed with sectionalizing switches

· Interconnecting lines can be switched on/off· Lines are configured radially outward from

the perspective of a substation

IEEE 33-Bus Distribution System

16

Power Loss

· Power loss in power lines exist based on · As the amount of load on a node increases,

so does the current → Greater power loss

17

Delayed Charging Strategy· Delay charging PEV until after a certain

period of time· Benefits:

- Reduced peak load- Valley-filling

18

Improved Network Reconfiguration· One-time optimization model

reformulated to minimize over a period of time- Enables the calculation of total power loss

during that period· Additional PEV delay constraint added

to allow PEV charge scheduling

19

Model

1 0

2, , , , , ,

( )

, , ,

Minimize subject to :

1)Network radialityconstraints2)Realpower injectionconstraints

[ ( cos sin )]

P , 1,..., ,

3)Reactivepower in

T n

Iit i

Ii t ij ij i t i t j t ij ij t ij ij tj N i

DGi t Di t PEVi t

P

P a g V V V g b

P P i n

-

- -

2, , , , , ,

( )

, ,

min , max

jectionconstraints

[ ( / 2) ( cos sin )]

, 1,...,

4)Voltagemagnitudelimits = , 1,...,

Ii t ij ij shij i t i t j t ij ij t ij ij tj N i

DGi t Di t

i i t i

Q a b b V V V b g

Q Q i n

V V V i n

-

-

20

Model (contd.)

2 2 2, , , , , , ,

2max

2 2sh

2 2

2sh

5)Maximum permissible line current carrying capacity

[ 2 ( cos sin )]

, 1,...,

where

( / 2)

( / 2)

l t l ij i t ij j t i t j t ij ij t ij ij t

l

ij ij ij ij

ij ij ij

ij ij ij ij ij

ij

I a A V B V V V C D

I l m

A g b b

B g b

C g b b b

D

- -

sh

max

, , , ,0

/ 2

6) Delayed Charging Strategy Constraintij ij

k

Ii t DGi t Di t k PEVi t kk

g b

P P P þ -

- -

21

Testing Phase· Optimization model is input into GAMS· Testing carried out on the 33-bus

distribution system with 10%, 20% and 40% PEV penetration

· PEV split into 8 groups· Maximum delay allowed: 5 hours

22

PEV Set Up

23

Results

Total Power LossNo PEV/Base

(kWh) Pen. LevelNo Delay

(kWh) Delay (kWh) Reduction (%)1095.8 10% 1513.8 1480.9 2.17

1095.8 20% 2083.3 2024.6 2.81

1095.8 40% 3680 3451.2 6.22

Delay (in hours)Group 1 2 3 4 5 6 7 810% 5 1 5 4 3 5 0 0

20% 0 4 5 1 1 5 0 0

40% 0 4 5 3 3 0 5 1

24

10% PEV Reconfiguration

25

Analysis· Network configuration remains the

same for delayed charging and unmanaged charging

· Power loss reduction difference between ‘no delay’ and ‘with delay’ increases as penetration level increases

26

10% Penetration Load Curve

12:00

AM

2:00 A

M

4:00 A

M

6:00 A

M

8:00 A

M

10:00

AM

12:00

PM

2:00 P

M

4:00 P

M

6:00 P

M

8:00 P

M

10:00

PM

1.25

1.5

1.75

2

2.25

Base No Delay With Delay

Load

Lev

el (M

W)

27

Conclusion· With the addition of a delayed charging

strategy:- Peak load is reduced by shifting PEV

charging to later hours- Total power loss is reduced in the system- There is no configuration difference

between unmanaged and delayed charging