Smart Grid EMS with Source Grid-Demand Synergy€¦ · Smart Grid EMS with Source-Grid-Demand Synergy Boming Zhang, Ph D, Prof., IEEE Fellow Dept. of Electrical Engg. Tsinghua University

Smart Grid EMS with Source-Grid-Demand Synergy

Boming Zhang, Ph D, Prof., IEEE Fellow Dept. of Electrical Engg.

Tsinghua University

12th EPCC Workshop, Bedford Springs, PA USA, 2-5 Jun 2013

2

• Smart Grid EMS is needed for modern EPCC which manages energy and controls power flow in a source-grid-demand coordinated way so as to accommodate more renewables in the network.

• Distributed autonomy and centralized coordination will be the key feature of the Smart Grid EMS.

• Distributed autonomy(Agent): many distributed μ-EMSs are built to control local objectives such as substation, wind farm, EV charger, μ-network, etc.

• Centralized coordination: a higher level EMS to coordinate μ-EMSs to achieve an overall benefit of security, economy and electrical quality.

• All these tasks will be implemented by the family of Smart Grid EMS.

Summary

Outline

Background and key scientific problems 1

Contents and methods

2

Preliminary results 3

4

3

Outline

Background and key scientific problems 1

Contents and methods

2

Preliminary results 3

4

4

5

2 1 3

6

7

•

•

8

•

•

9

EMS

10

(3) EMS suitability

11

12

• This proposal has been approved by Special Fund of the National Priority Basic Research of China (973 Project)

• Fund: YMB 38 million Ruans • Duration: Jan 2013 -- Aug 2017

Outline

Background and key scientific problems 1

Contents and methods

2

Preliminary results 3

4

13

2.1 What we have to do

14

Title

Title

Title

Title

Title

Title

15

EMS

W-EMS

SVC/STATCOMSVC/STATCOM

P-EMS

SVC/STATCOMSVC/STATCOM

minU

J (U ),U ={u0,u

1,u

2,...,,u

M}

s.t. xk+1

= f (xk,u

k,r

k),k = 0,1,..., N -1

g(xk,u

k) £ 0,k = 0,1,..., N -1

MPC

16

g1

T ¶J1

*

¶x+¶L

1

¶u1

= 0

gN

T ¶JN

*

¶x+¶L

N

¶uN

= 0

ì

í

ïïï

î

ïïï

1

10

min ( , , )

= ( ( ), ( ), , ( ))

ii N

u

i N

J u u

L x t u t u t dt

L

L

f (x)+ gi(x)u

ii=1

N

å = 0

Game model

17

1

1 1

1 11 1

1 1

1 11 1

TT T T

T T

Y Y Y YI Y Y I

1

1 11 1 1T T

V Y I Y V

1

1

1

Ki

TT TT TT

i

Ki

T T T

i

T TT T

Y Y Y

I I I

V Y I

2

1

1

i

TT Ti ii iT

i

T Ti ii i

Y Y Y YI Y Y I

1

i ii i iT T

V Y I Y V

k

1

1

K

TT TK KK KT

K

T TK KK K

Y Y Y YI Y Y I

1

K KK K KT T

V Y I Y V

1 1,TT T Y I

TV

,i i

TT T Y I

TV

K

TIK

TTY

TV

18

0

( )[ ( ), ( ), , ] [ ( ), , ] ( )

(0)

( ) [ ( ), , ] [ ( ), , ] ( )

[ ( ), ( ), , ]

x x

y y

xy

d tt t t t t t

dt

t t t t t t

t t t

X F X U θ G X θ ε

X X

Y F X θ G X θ ε

0 F X Y θ

% % %

%

% % %

%

0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 533

33.5

34

34.5

35

35.5

36

t(s)

/(°

)

19

PV Wind

Thermal Hydro

EV

1 2min [ ( ), ( )]

s.t. ( )

( )

J J

00

x xg xc x

20

Distributed autonomy

Centralized coordination

Distributed autonomy

21

D-EMS S-EMS T-EMS V-EMS

B-EMS

H-EMS

W-EMS

μ-EMS

P-EMS

EMS Suitability

22

2.2

•

EMS with S-G-D synergy

Outline

Background and key scientific problems 1

Contents and methods

2

Preliminary results 3

4

23

3.1 Autonomic Voltage Control in Hebei

24

May 12th, 2012

5 kV higher

Wind EMS— C

Hebei Province

Hongda, with control 67 1.5MW WGs 1 MVR SVC

Batou, without control

Guyuan area: installed wind generation capacity >3GW daily generation 1GW

Autonomic control points: 21 wind farms

Coordinated control point: Area control center

Coordinated Voltage Control in Hebei

Before control (11-13-2011)←→ After control (10-18-2012)

27

• online rolling MW schedule

3.2 MW dispatch and control to accommodate large scale wind power integration

Real-time control Day ahead Rolling schedule Real-time

dispatch

24h 4h 15m 10s

• Model Predictive Control(MPC) is used in each time level

28

Close Loop MW Control in Qian-An Wind Farms in Jilin Province

1

2

Qian-An WF • 27 WFs, installed wind

generation capacity 2936MW, 36% of max load and 62% of min load

(2011 ) 2012

Wind power Spillage

3.3 Wind

29

NC

PC

Abandoned wind energy is reduced by 43%

IEE

E 14 nodes case

UC CC

UC CC

3.4

30

S-SE Time Consuming(ms)

FC 23.8

YQ 10.9

SZ 5.2

DT 17.3

Pic. Trans. <5.9s

Model Trans. <4.6s

RT Data Import 0.2s

CC-SE Time Consuming 0.7s（1260 nodes）

0.00%

0.20%

0.40%

0.60%

0.80%

1.00%

估计前 估计后

0.00%

0.05%

0.10%

0.15%

0.20%

估计前 估计后

Monte Carlo simulation results( 1000,RTDS)

Before SE After SE

Before SE After SE

Conclusion

31

Distributed Autonomy - Centralized Coordination

Synergy

New G of EMS

Thanks！