concept of a wide area defense system for the power grid chen-ching liu university college dublin...
TRANSCRIPT
Concept of a Wide Area Defense Concept of a Wide Area Defense System for the Power GridSystem for the Power Grid
Chen-Ching LiuUniversity College Dublin
National University of Ireland, Dublin
Seminar at NTUA and IEEE PES Greece, June 20091
Catastrophic Power Outages
2
Western Electricity Coordinating Council (WECC) system - Aug 10th, 1996 Blackout
PDCI Remedial Action Schemes (RAS) began to actuate. Shunt and series capacitors were inserted.
15:47:40-15:48:57 p.m. Generators at the McNary power house supplying 494 MVAR trip. The system begins to experience “mild oscillations”.
15:48:51 p.m. Oscillations on the POI reached 1000MW and 60-kV peak-to-peak.
Lines throughout the system begin to experience overloads as well as low voltage conditions. Additional lines trip due to sagging.
The WECC broke into 4 asynchronous islands with heavy loss of load.
7.5 million people lost power.
Mild .224 Hz oscillations were seen throughout the system and began to appear on of the PDCI.
15:48 p.m. Keeler-Allston 500-kV line contacts a tree due to inadequate right-of-way maintenance. Additionally the Pearl-Keeler line is forced out of service due to the Keeler 500/230-Kv transformer being OOS.
Initiating events System becomes unstable Blackout
Shunt capacitor banks were switched in to raise the voltage but the oscillations were not being damped.
AZ
CA CO
ID
MT
NENV
NM
ND
OR
SD
UT
WA
WY
With the loss of these 2 lines, 5 lines are now out of service, removing hundreds of MVAR.
3
Eastern Interconnection –August 14th, 2003 Blackout
2:14 p.m. FE’s control room lost alarm functions followed by a number of the EMS consoles.
1:31 p.m. Eastlake 5 generation unit trips and shuts down.
2:02 p.m. Stuart-Atlanta 345-kV line tips off due to contact with a tree.
1:07 p.m. FE turns off their state estimator for troubleshooting.
Initiating events System becomes unstable Blackout
AL
AR
CT
DE
FL
GA
IL IN
IA
KS
KY
LA
ME
MD
MA
MI
MN
MS
MO
NE
NH
NJ
NY
NC
ND
OH
OK
PA
RI
SC
SD
TN
VT
VAWV
WI
265 power plants tripped off line and 50 million people are without power.
Low voltage/ high load conditions and system disturbances propagate through the system tripping transmission lines and generators.
4:08:59 p.m. Galion-Ohio and Central-Muskinghun 345-kV lines trip on Zone 3 causing major power swings through New York and Ontario and into Michigan.
4:05:57 p.m. The loss of 138-kV lines overloads the Sammis-Star line.
2:54 p.m. The primary and secondary alarms servers failed.
3:05:41-3:57:35 p.m. 3 345-kV lines trip due to contact with trees. This overloads the underlying 138-kV system and depressed voltages.
3:39:17-4:08:59 p.m. 16 138-kV lines trip due to overloading.
4:13 p.m. most of the North East and parts of Canada blacked out. There are only a few islands which remain operating.
4
Hydro-Québec Blackout-April 18th, 1988
8:08 p.m. In response to the loss of the Church Fall generation station, a signal to initiate load shedding is sent to the central control center
8:08 p.m. 3200 MW of generation from the Church Falls generation complex is isolated due to isolation of transmission lines
8:08 p.m. Wet snow and freezing rain cause flash over on all three phases at the Arnaud substation
Initiating events System become unstable Blackout
18,500 MW of load was lost because 3,200 MW of automatic load shedding failed to occur
With the loss of the James bay transmission lines the Le Grand network was separated from the rest of the system. Shortly afterward three transformers at Le Grand 4 failed
1.7 seconds after the failure of the load shedding system the first of the James bay transmission lines tripped, followed by the other two
The load shedding signal was not received at the central control center due to a faulty contact in the communications system
15 seconds later the Manicouagan-Montréal transmission system collapsed. This lead to the loss of all DC interconnections as well as 8 generators at the Beauharnois generating station which had been isolated to serve the Ney York Power Authority (NPA)
5
Blackout Propagation(without defense systems)
Complete System Collapse
Triggering Event
6
Occurrences and Extent of Blackouts in North America
Number of customers affected100 102 104 106 108
Num
ber
of B
lack
outs
10
100
1000
7
Strategic Power Infrastructure Defense (SPID)
Design self-healing strategies and adaptive reconfiguration schemes
To achieve autonomous, adaptive, and preventive remedial control actions
To provide adaptive/intelligent protection
To minimize the impact of power system vulnerability
8
SPID System
FailureAnalysis
Self-HealingStrategies
VulnerabilityAssessment
Informationand
SensingReal-Time
Securit
y Robustne
ss
Dependability
Power Infrastructure
•Satellite, Internet•Communication system monitoring and control
Hidden failure monitoring
Adaptive: load shedding, generation rejection, islanding, protection
Fast and on-line power & comm.
system assessment
9
Multi-Agent System for SPID
REACTIVE LAYER
COORDINATION LAYER
DELIBERATIVE LAYER
Knowledge/Decision exchange
Protection Agents
GenerationAgents
Fault Isolation Agents
FrequencyStabilityAgents
ModelUpdate Agents
CommandInterpretation
Agents
Planning Agent
Restoration Agents
HiddenFailure
Monitoring Agents Reconfiguration Agents
VulnerabilityAssessment
Agents
Power System
Controls
Inhibition Signal
Controls
Plans/Decisions
EventIdentification
Agents
Triggering Events
Event/AlarmFiltering Agents
Events/Alarms
Inputs
Update Model CheckConsistency
Comm.Agent
10
Multi-Agent System for SPID
• Subsumption Architecture (Brooks) for Coordination
• Agents in the higher layer can block the control actions of agents in lower layers
Load SheddingAgent
Global View/Goal(s)
R
Under Frequency RelayLocal View/Goal(s) Tripping Signal
Inhibition Signal
11
Cascaded Events
Some Basic Patterns• Line tripping due to overloading• Generator tripping due to over-excitation• Line tripping due to loss of synchronism• Generator tripping due to abnormal voltage and frequency system condition• Under-frequency/voltage load shedding
Identifying the basic patterns of cascaded events and explore how these patterns can be
combined into sequences
12
Cascaded Zone 3 Operations
13
Zone 3 Relay Operations Contributed to Causes of Blackouts.
Heavy Loaded Line
Low Voltage High Current
Lower Impedance Seen by Relay
Loss of Transmission Lines
Other Heavy Loaded Lines
Zone 3 Relay
Operation(s)
Catastrophic Outage
Prediction of Zone 3 Relay Tripping Based on On-Line Steady State Security Assessment
14
Case Relay Status Contingency Description
1 N/A Secure 3 phase fault at bus 1
2 Zone3 Insecure 3 phase fault at bus 2
.. …
N Secure 3 phase fault at bus N
Case Relay Status Contingency Description
1 N/A Secure 3 phase fault at bus 1
2 Zone3 Insecure 3 phase fault at bus 2
.. …
N Secure 3 phase fault at bus N
Case Relay Status Contingency Description
1 N/A Secure 3 phase fault at bus 1
2 Zone3 Insecure 3 phase fault at bus 2
.. …
N Secure 3 phase fault at bus N
Case Relay Status Contingency Description
1 N/A Secure 3 phase fault at bus 1
2 Zone 3 Insecure 3 phase fault at bus 2
.. …
N N/A Secure 3 phase fault at bus N
…
Contingency Evaluation Performed On Line
Every Several Minutes
Contingency Evaluation
Post-Contingency Power Flow
Post-Contingency Apparent Impedance
Corrected Post-ContingencyApparent Impedance
FISFuzzy Inference System (FIS) Developed Using Off-Line Time-Domain Simulations
Impedances Obtained by Power Flow and Time Domain Simulation
15
Post-Contingency Impedance Obtained by Power Flow Does Not Coincide with Impedance Obtained by Time-Domain Dynamic Simulations
Impedance Obtained by Steady State Power Flow
Zone 3 Circle
Correction Term
Impedance Locus
Zon
e 3
Rea
ch
R
X
0
Load Angle
Automatic Development of Fuzzy Rule Base
16
Wang & Mendel’s algorithm is a “learning” algorithm:1) One can combine measured information and human linguistic information into a common framework2) Simple and straightforward one-pass build up procedure3) There is flexibility in choosing the membership function
Pre-determine number of membership functions N
Give input and outputdata sets
In this example, N is 7
(Inp1, Inp2, Out) = (10, 1, -2)(Inp1, Inp2, Out) = (8, 3, -1)(Inp1, Inp2, Out) = (5, 6, -4)(Inp1, Inp2, Out) = (2, 8, -5)…
FIScreated
automatically
17
Case A
86
157
181
182
113
32
31
79
74
66 65
184
75 73
7669
68
70
7782
9591
92
93
94
96
97
98
83
168
169
170
171
114
172
173
111
120
121
122
123
124
125
115
100
101
105106
117
116
132
133
134
104
118
108
107
110
89
90
88
87180 81 99 84
156
161162
85
36 186187
183
185
188
189
ß 21
6
11
18
190
5
17
8 9
10
72914
192193194
139
13
21
20
2627
148147
195
191
163
158 159167 155 165
164
166
16045 44
64
63
37
178176
179177
142
46 4839
230-287kV
109
112
11-22kV
Substation
12
138
28
135
126
127
128
129
130
131
175
174
67
72197
196
198
200
78
199
80
71
Relay Location
Fault Location
Impedance on R-X (Case A)
18
-0.01
0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0 0.02 0.04 0.06 0.08 0.1
R [PU]
X [
PU
]
Impedance LocusZone 3 CircleZ Obtained by Power Flow CalculationCorrected Z Obtained by FISPost-Contingency Z Obtained by Time-Domain Simulation
Pre-fault
Line Tripping by Zone 3 relay
Case A
Z obtained by power flow solution is outside Zone 3 circle.
Load Shedding Studies have shown that the August 10th 1996 blackout could have been
prevented if just 0.4% of the total system load had been dropped for 30 minutes.
According to the Final NERC Report on August 14, 2003, Blackout, at least 1,500 to 2,500 MW of load in Cleveland-Akron area had to be shed, prior to the loss of the 345-kV Sammis-Star line, to prevent the blackout.
19
Automatic Load Shedding
Under Voltage
Under Frequency
Rate of Frequency Decrease
Remedial Action Scheme
20
Adaptive Self-healing:Load Shedding Agent
• A control action might fail
• Unsupervised adaptive learning method should be deployed
• Reinforcement Learning– Autonomous learning method based on interactions with the
agent’s environment
– If an action is followed by a satisfactory state, the tendency to produce the action is strengthened
21
Load Shedding Options
0 50 100 150 200 25058.6
58.8
59
59.2
59.4
59.6
59.8
60
frequency with 20 % load shedding
10% load shedding freq
uenc
y
Time (multiples of 0.02 sec)
22
Adaptive Self-healing:Load Shedding Agent
• 179 bus system resembling WSCC system• ETMSP simulation• Remote load shedding scheme based on
frequency decline + frequency decline rate• Temporal Difference (TD) method is used for
adaptation: Need to find the learning factor for convergence
23
Adaptive Load Shedding Agent
State 1 State 2 State 3
Freq := 59.5
Dec.rate > threshold value
Freq := 58.8
Dec.rate > threshold value
Freq := 58.6
Dec.rate > threshold value
179 bus system24
Adaptive Load Shedding Agent
0 20 40 60 80 100 120 140 160 1800
0.5
1
1.5
2
2.5
a=0.55
a=0.75
Number of trials
Nor
mal
ized
fre
quen
cy
Expected normalized system frequency that makes the system stable
“The load shedding agent is able to find the proper control action in an adaptive manner based on responses from the power system”
“The load shedding agent is able to find the proper control action in an adaptive manner based on responses from the power system”
25
Flexible Grid Configuration to Enhance Robustness
Flexible Grid Configuration can play a significant role in defending against catastrophic events.
Power infrastructure must be more intelligent and flexible.
To allow coordinated operation and control measures to absorb the shock and minimize the potential damages caused by radical events .
26
Cascading Events
• A Cascading Event Refers to a Series of Tripping Initiated by One or Several Component Failures in a Power System– Here the initial component(s) failure is
designated as “shock” to the power infrastructure
27
Simulated Cascading Events (179 Bus System)
• Compute Power Flows after Tripping
– Six lines are found on limit violation – Trip these lines
• Identify New Network Configuration and Solve Power Flows Again
– Fifteen lines are found with limit violations – Trip these lines
• Continue This Simulation Procedure
– Finally system collapses: most transmission lines are tripped and most loads are lost
28
2-Area Partitioning Algorithm (from VLSI)
• Spectral 2-way Ratio-Cut Partitioning – Theorem Given an edge-weighted graph G = (V,
E), the second smallest eigenvalue λ2 of the graph’s Laplacian matrix Q yields a lower bound on the cost c of the optimal ratio cut partition, with c = e(U,W)/(|U|·|W|) ≥ (λ2/n)
– Cut-Size: e(U,W) ≥ (λ2/n) (|U|·|W|)
29
Area-Partitioning Algorithm
5
2 1
6 4
3 1
0.1 0.1 0.1
2
1 1
6-Bus System
30
Area-Partitioning Algorithm
Partition {U, W} Cut Set Size e(U,W)
e(U,W) / (|U|·|W|)
λ2 / n
{(4), (5 6 3 2 1)} 1.1 1.1 / (1 5) = 0.22
{(4 5), (6 3 2 1)} 1.2 1.2 / (2 4) = 0.15
{(4 5 6), (3 2 1)} 0.3 0.3 / (3 3) = 0.0333
0.1966 / 6 = 0.0328
{(4 5 6 3), (2 1)} 2.1 2.1 / (4 2) = 0.2625
{(4 5 6 3 2), (1)} 1.1 1.1 / (5 1) = 0.22
Partition Results of 6-Bus System
31
Flexible Grid Configuration to Absorb the Shock
• Solve Power Flows of Area One – All 35684.71 MW loads are supplied, no line
flow constraints violations
• Solve Power Flows of Area Two – Seven lines on limit violation:
• (Bus158-Bus164), (163-8), (64-163), double lines (16-19), and double lines (150-154)
32
Flexible Grid Configuration to Absorb the Shock
• Use “Power Redispatching & Load Shedding” in Area Two– Totally, 188 + 64.4 + 60 = 312.4 MW load are shed
Bus #
Original Load (MW)
Load Shed (MW)
Load Supplied (MW)
8 239 188 51
16 793.4 64.4 729
154 1066 60 1006
33
Split System into Two Areas
33 32
31 3 0
35
80
78
74
79 66
75
77
76 72
82
81
86 83
84
85
114
115
118 119
103
107
108
110 102
104
109
230 kV 345 kV 500 kV
34
65
71
69
70
87 88
99 36
73
89 90
124 125
169 171
170 172 173
111
120
121 122
123
91 - 94 95 - 98
13 2 133
135 134 104
174, 176, 178
113 100
101
105 106
117 116
68
67
112
180
156 157 161 162 168
167 165
158
159 155 44
45 160
166
163
5 11
6
8
9
18 17
4
3
7
14
12 13
138 139
147
15
19
16
142
37 64 63
153 145 151 152
136 49 48
146 154
149
143
43
175, 177, 179
48 38
57 58
54 51
52 53 42
55 41 62 56
40
39
150
137
61
148 22
23 25
24
20 21
29 28
2
10
164
50
42
47
46
59 60(3)
27 26
141 140
144
48
48
48
34
Flexible Grid Configuration to Absorb the Shock
• Shed Load vs. System Total Load– K=1
– K=2
– K=3
%51.0%100MW 60785.41
MW 312.4
%982.0%100MW 60785.41
MW 597
%755.1%100MW 60785.41
MW 1067
35
Flexible Grid Configuration to Avoid Cascaded Failures
• Step 1 : Compute power flows after initial tripping event(s).
• Step 2 : Convert power network to an edge-weighted graph G, weight of each edge is absolute value of real power flow.
• Step 3: Multilevel graph partitioning with minimum edge-cut.
Network is separated into k areas to minimize generation / load imbalance.
Initial Graph G0
G1
G2
Gk
G2
G1
G0
Projected Partition
Refined Partition
Coarsening Phase
Initial Partitioning Phase
Uncoarsening and
Refinement Phase
•Graph COARSENED to a smaller number of vertices, • Bisection PARTITIONING of much smaller graph,• UNCOARSENING back toward original graph.• Coarsest graph small, Coarsening can be parallelized, Partitioning efficiency high.
36
Emergency Control with Multilevel Graph Partitioning
CommunicationNetwork
Power Transmission Network
Central Control Unit
CB1
CB2
SCU1
SCU2SCU n
Relay1
Relay2
SCU – Substation Control UnitCB – Circuit Breaker
• CCUs acquires system data, generates system separation strategy.• SCUs receive system separation commands from CCU and send breaker opening commands to specific auxiliary relays.
An adaptive relaying architecture for controlled islanding
• Partitioning a 22,000 bus and 32,749 branch system into 2, 3, 4 islands with 0.07s, 0.081s and 0.09s on 2 GHz Pentium CPU and 1GB RAM.
• Fast computational speed makes it possible to determine partitioning strategy and identify new network configuration in on-line environment.
37
Flexible Grid Configuration to Avoid Cascaded FailuresControlled islanding on a 200-bus system:199 buses, 31 generators, 248 branches.
Sequence of cascading events:•At t=0 s, three transmission lines out-of-service.•At t=60 s, line 72-197 tripped due to line fault.•At t=120 s, line 78-196 tripped due to line fault. Generator G70 at bus 70 overloaded.•At t=240 s, generator G70 tripped by over-excitation protection. At t= 260 second, system collapses.
157
113
35
34
33
32
31
79
30
66 65
184
75
70
7782
9591
92
93
94
96
97
98
168
169
170
171
114
172
173
111
120
121
122
123
124
125
115
119
100
101
105106
117
116
132
133
134
104
118
108
102
107
110
89
90
88
87180 81 99 84
156
161162
85
36 186187
183
188
189
ß 21
6
11
18
5
17
8 9
103
2 4
72914
24
16 15
25
19
23
22
193194
139
13
21
20
2627
148147
195
163
158 159167 155 165
164
166
16045 44
64
63
37
38
178176
179177
145
154
146
144
151
142
152143
137
150
149
40
46 4839
59
60
55
4258
43
50
57 54 51 53
52
41
56
140
500kV345-360kV230-287kV
109
112
11-22kV
103
12
138
28
49
47
135
126
127
128
129
130
131
141
175
174
62
67
72
198
200
199
80
71
74
196
19778
69 76
182
181
73
185
6886
83 190
191
192
136
153
61
North Island
South Island
Cut Set Load-Generation (MW)
Bus 174-175, 176-177, 178-179, 155-156, 155-
167, 188-189(1,2)
North: Gen=37862 , Load=37104South: Gen=24517, Load=23794
38
Flexible Grid Configuration to Avoid Cascaded Failures
0
0. 2
0. 4
0. 6
0. 8
1
1. 2
0 30 60 90 120 150 180 210 240 270
Time (sec)
Loa
d B
us V
olta
ge (
pu)
Load bus voltages without/with islanding strategy
0.6
0.7
0.8
0.9
1
1.1
1.2
0 40 80 120 160 200 240 280 320 360 400
Loa
d B
us V
olta
ge (
pu)
Time (sec)
Load bus at North IslandLoad bus at South Island
• System islanding initiated at 241s.• Islanding strategy results in balanced generation / load in both islands.• All loads in the system are served. Load shedding scheme not applied. • Islanding strategy successfully prevents the collapse of the system.
39
Conclusion
• Cascading failures remain a grand challenge
• New communication, information and computer technologies enable wide area protection and control
• Connectivity also brings cyber vulnerability
• “Smart” grid?
• “Self-healing” grid?
40
Further Information
• J. Li, C. C. Liu, “Power System Reconfiguration Based on Multilevel Graph Partitioning,” IEEE PES Power Tech, Bucharest, Romania, 2009.
• K. Yamashita, J. Li, C. C. Liu, P. Zhang, and M. Hafmann, “Learning to Recognize Vulnerable Patterns Due to Undesirable Zone-3 Relay Operations,” IEEJ Trans. Electrical and Electronic Engineering, May 2009, pp. 322-333.
• J. Li, K. Yamashita, C. C. Liu, P. Zhang, M. Hoffmann, “Identification of Cascaded Generator Over-Excitation Tripping Events,” PSCC, Glasgow, U.K., 2008.
• H. Li, G. Rosenwald, J. Jung, and C. C. Liu, “Strategic Power Infrastructure Defense,” Proceedings of the IEEE, May 2005, pp. 918-933.
• J. Jung, C. C. Liu, S. Tanimoto, and V. Vittal, “Adaptation in Load Shedding under Vulnerable Operating Conditions,” IEEE Trans. Power Systems, Nov. 2002, pp. 1199-1205.
• H. You, V. Vittal, J. Jung, C. C. Liu, M. Amin, and R. Adapa, “An Intelligent Adaptive Load Shedding Scheme,” Proc. 2002 PSCC, Seville, Spain, June 2002.
• C. C. Liu, J. Jung, G. Heydt, V. Vittal, and A. Phadke, “Strategic Power Infrastructure Defense (SPID) System: A Conceptual Design,” IEEE Control Systems Magazine, Aug. 2000, pp. 40-52.
41