2015 PES GM CascadingPhenomenon Final To Print

UnderstandingCascadingPhenomenon:
MethodologiesandIndustryPracticefor
AnalysisofCascadingFailures
2015IEEEPESGeneralMeetingDenver,CO
Sponsoredby
CascadingFailureWorkingGroupunderIEEEPESComputerandAnalyticalMethods
Subcommittee
IEEEPESReliability,RiskandProbabilityApplicationsSubcommittee
TableofContents
1. OverviewofCascadingOutagesPhenomenon
2. MethodologiesforAnalysisofCascadingOutages
3. FrameworksandCurrentToolsforDetection,Prevention,
andMitigationofCascadingOutages
4. EmergingTechnologiesforDetection,Prevention,and
MitigationofCascadingOutages
5. IndustryExperienceintheAnalysisofCascadingOutages
6. RestorationfromCascadingFailures
7. AnalysisofPastBlackoutsCausedbyCascadingOutages:
LessonsLearned
1
Session1
1. OverviewofCascadingOutagesPhenomenon:
1.1 WhatisaCascadingOutage?
Instructor: IanDobson
1.2 CascadeOutageMechanisms
Instructor: VladimirTerzija
Session1.1
WhatisaCascadingOutage?
Ian Dobson
Cascading Failure Working Group Tutorial

PES General Meeting
Denver CO
July 2015
2
DEFINITIONS
PES Working Group: Cascading failure is a sequence of

dependent outages of individual components
that successively weakens the power system
NERC: The uncontrolled successive loss of system elements

triggered by an incident at any location.
Cascading results in widespread electric service interruption
that cannot be restrained from sequentially spreading
beyond an area predetermined by studies.
USEFUL DISTINCTION
Cascading = initial outages + propagation of outages
Whycascadingisahardproblem
- rare, unanticipated, dependent events
(common, well understood failures already mitigated)
- huge number of possibilities and combinations
(impossible to analyze all cases)
- dozens of mechanisms (described in later talks)
- three sources of uncertainty in cascading:
initial state, initial outages, and progress of cascade
- cascades can spread far
- power law in distribution of blackout size implies
Large blackout risk > small blackout risk
(need to study the complicated large blackouts)
3
NERCblackoutsizedatashowsheavytail
102
Large blackouts more
likely than expected; 101
distribution
Probabilty
caused by cascading
Large blackouts rare, 100
but have high impact
and significant risk 10-1
Statistical problems
assessing rare events; 10-2
normal statistics do 10-4 10-3 10-2 10-1 100
not work Load shed/ Power served
Blackout size
Compromises
Analyze:
singleblackoutafteritoccurs
onlysomefailures(highriskorinitial)
onlyafewcascadingmechanisms
simplifiedorprobabilisticmodeling
userealtimeinfotoreduceuncertainty
overallcascadingbutnodetails
4
OverallapproachestoModelingandSimulation
Selectalimitedsubsetofmechanismstomodel
1) Enumeratelikelysequencesinsuitablystressed
conditiontoguidemitigation
ofteninadeterministicframeworkwithaselected
contingencylist
OR
2) Evaluateriskbysamplingfromuncertainties
Simulations need to be validated with data

- reproduce qualitative form of statistics
ApproachestoDataAnalysisandMetrics
Lookatindividualblackoutsequences
Estimatemeanblackoutsize
heavytailproblem:meanisnotusefulbecause
unrepresentativeandhighlyvariableunlesslarge
blackoutsarecensored
Estimatedistributionofblackoutsize
allowsmitigationofbothsmallandlargeblackouts
empiricalapproachtakestoolong(decades)
butcanestimateinitialfailuresandpropagation
andthenestimatedistributionofblackoutsize
5
Worstcasecascading: forexample,
SequenceofoutagesinWesternblackout,July21996
from NERC 1996 blackout report
QuantifyingtypicalcascadingfromTADSdata
TADS=TransmissionAvailabilityDataSystem;
StandardoutagedatathatisreportedtoNERC
Automaticlineoutagescanbegroupedintoinitial
outagesandgenerationsofpropagatingoutages
Canestimateaveragepropagation,whichisthe
averagenumberofchildrenoutagesperparent
outage.
Forexample,eachlineoutageproducesonaverage
0.25lineoutagesinnextgeneration.
CanlocateoutagesonagridconsistentwithTADS
Dobson, IEEE Trans power systems Nov 2012
6
Utility network
that is consistent
with TADS data
For this cascade,

red lines outage in
order shown by
the generation
numbers
7
Probability Predicting cascading
0.05 failure extent from
0.04
average propagation
in utility data,
probability
0.03 (Dobson, IEEE Trans.
Power Systems, 2012)
0.02
0.01
0.00
0 20 40 60 80 100
line outages
Number of line outages
Probability distribution of total number of line outages assuming 5 initial line
outages
Keychallengesincascading:
whattomodelandsimulateandinwhatdetail?
requiredsamplingand/orchoiceofscenarios
validationandinterpretationofsimulationand
modelresults
quantifyriskwithmetricsforcascadingthatcanbe
evaluatedfromrealandsimulateddata
actionsthatmitigatebothlargeandsmallblackouts
8
Session1.2
CascadeOutageMechanisms
Vladimir Terzija,
Negar Shams and Peter Wall
vladimir.terzija@manchester.ac.uk
Outline
1. Power System Blackouts and Cascading Outages Overview
2. Stability and Security Concepts
3. Most Significant Contributing Factors to Blackouts
4. System State and Situational Awareness
5. Stages and Mechanisms of Cascading Blackouts
6. Example: Northern Ireland Blackout
7. Cascading Outage Simulation Examples with and Without
Controlled Islanding
8. Conclusion
9
BlackoutsandCascadingOutages
PowerSystemBlackout:thelossofsupplytotheloadinthe
entirepowersystem,orpartsofthesystem
CascadingOutage:asequenceofeventsinwhichaninitial
disturbance,orasetofdisturbances,causesasequenceof
oneormoredependentcomponentoutages
Each blackout has its own specific nature

large number of scenarios and black out paths
BlackoutandCascadingOutages
Electricpowersystemshavebecomeextremelycomplexand
difficulttooperateandprotect
Itisimpractical anduneconomicaltodesignpowersystemsto
bestable forevery possibledisturbance
Theeconomicandsocialimplicationsofblackoutscanbe
disastrousanddifficulttoquantifyinadvance
Powersystemscannotbeprotectedfromcatastrophicnatural
disasters,buttheconsequencesmightbelimited.
10
RecentMajorBlackouts
Location Date Duration Populationaffected
Brazil 11/03/1999 5hours 97,000,000

India 02/01/2001 12hours 226,000,000 2
USA/Canada 14/08/2003 96hours 50,000,000
Italy 28/09/2003 18hours 56,000,000
Spain 29/11/2004 240hours 2,000,000
Indonesia 18/08/2005 7hours 100,000,000
UCTE 04/11/2006 2hours 15,000,000
Brazil&Paraguay 10/11/2009 7hours 87,000,000
Brazil 04/02/2011 16hours 53,000,000
India 31/07/2011 48hours 620,000,000
1
Bangladesh 01/11/2014 10hours 150,000,000 3
Pakistan 26/01/2015 9hours 140,000,000 4
BlackoutsIncidentsSurvey
Blackout BlackoutIncidents
Voltage Frequency Cascade System Lossof

Location Date
Collapse Collapse Overload Unsymmetrical Synchronism
USA 9/11/1965
France 19/12/1978
USA 02/07/1996
USA 07/08/1996
Croatia 12/01/2003
USA&
Canada
14/08/2003
Denmark &
Sweden
23/09/2003
Italy 28/09/2003

Greece 12/07/2004

Australia 14/03/2005

Total:10 6 1 7 1 2
11
BlackoutandCascadingOutages
Ablackoutcandescribedbasedontheeventsthatcausedthe
finallossofsupplyandtheseareusuallystabilityrelated
However,theactualpathtoablackoutinvolvesawiderange
ofcomplex,interdependentmechanisms,e.g.:
Changesinpowerflowafteroutages(riskofcascadeoverloads)
Protectionfailures
Proximitytosecuritylimits(probabilityofprotectionfailureincreases
asthesecuritymarginfalls)
Controlactionstoresolveonethreattostabilityinstressedconditions
mayhaveadetrimentalimpactonasecondformofstability
Stabilitystudiesdonotaccountforallinteractionssosecure
operationcannotguaranteeablackoutwillnotoccur
StabilityConcept
Power system stability: is the ability of an electric power system, for a given
initial operating condition, to regain a state of operating equilibrium after
being subjected to a physical disturbance, with most system variables
bounded so that practically the entire system remains intact .
Power Systems
Stability
Classification
12
SecurityConcept
Power system security:
Power System Security refers to the degree of risk in a power
systems ability to survive imminent disturbances without
interruption to customer service.
N1 Security Criterion:
The worst single contingency should not result in the violation
of the emergency ratings or stability limits of other system
components.
Security margin
Security Margin stands for the amount by which system loads
or power transfers can change before a security violation, such
as an overloaded transmission line, is encountered.
MonitoringSecurityMargins
Dynamic Security Assessment
should be done in smallest
intervals possible following each
successful state estimation
Security Margins should be then

monitored for all stability types
System Operator should be

alerted if a security margin is
found to be insufficient
IEEE 11 Bus bar Network Dynamic Voltage Security

Assessment results using DSATools software
13
WideAreaMonitoring
VISOR
visorproject.org.uk
Blackouts:ContributingFactors
Operating under stress (near peak demand, small security
margin)
Aged assets (e.g. 2005 Moscow blackout)
Inadequate reactive power reserve increased risk of voltage

collapse (e.g. 2003 USA and Canada)
Some important equipment out of service (e.g. 2006 UCTE)
14
Blackouts:ContributingFactors
Environmental factors such as wind , thunderstorm , fire, etc.
(e.g. 1996 USA, increased amount of dust due to fire led to
flashovers)
Protection system failure (due to wrong settings or hidden

failure modes)
Human factors (e.g. 2006 UCTE)
Inadequate situational awareness
SituationalAwareness
Inadequate situational awareness is one of the most
important contributing factors to cascading outages (e.g.
2003 Canada USA,)
The difference between scheduled power exchange and
actual power flows (e.g. 2006 UTCE)
Systemstateawarenesscanhelpwhendesigning
earlywarningsystems
Lack of real time data prevents operators from taking quick
remedial actions
15
SystemStates
Allconstraintsaresatisfied
N1Security
Normal
Constraints are generation
adequacy (load and
Allconstraintsaresatisfied
N1security
generation balance) and
Alert Increasedprobabilityofdisturbance component overload
Constraintsareviolated
N1security
Emergency
Increased probability of
disturbance can be due to
Constraintsareviolated adverse weather conditions
N1security
Inextremis RapidCascadingoutages
SystemStates ControlActions
Generation re-dispatch
Normal
Tie line re-scheduling
Capacitor switching
Restorative Alert
Fast HVDC power control
Blocking of tap changers
Controlled islanding
Inextremis Emergency
Under frequency and
Under voltage load
Disturbance shedding
Control Action
16
CascadingBlackoutMechanism
Cascading blackouts can be divided into several phases:
Precondition
Initiating event Steadystateprogression
Cascade events Triggeringevent
Final state Highspeedcascade
Restoration
However, not all blackouts have all the phases listed above
(Croatia, 2003)
The precondition state can affect the severity and speed of
outage propagation following a disturbance
BlackoutPhases
Operating Back to
in N-x N-x
Security Security
Italy Blackout September,2003
17
InitiatingEvents
Initiating events vary between different blackouts:
Short circuit
Overload
Protection hidden failure (dependability based or security based)
Loss of generation, or intertie line
Power systems are generally designed to withstand single or
double failure without any operating limit violation (according
to N1, or N2 criterion)
However, depending on the severity of the failure, power
system might enter alert or emergency state
CascadingStages
Initiatingeventoftencausespoweroscillationsandvoltage
fluctuations;leadingtohighcurrentsandlowvoltages
Thismightbedetectedasafault(e.g.byZone3,distancerelays)
Furtheroverloadonneighbouringlinescausingmorelinestotrip
Formationofunintentionalislands
Frequencyorvoltagecollapse
18
SteadyStateProgressionPhase
In this phase the progression of cascade events is slow
System is still able to maintain balance between generation
and demand
Effective implementation of control actions can stop the
spread of extreme conditions or a blackout before the
situation becomes uncontrollable
Crucial opportunity for system operators to respond as time is
available and the system state is not too extreme
E.g. Intentional Controlled Islanding
IntentionalIslandingMechanism
Initiating Uncontrolled
Event System Separation
Remedial Remedial
Actions Actions
Taken Fail
Healthy Steady State High Speed Healthy

Blackout Restoration
System Progression Cascade System
Triggering Verify Feasibility of

Event Islanding Solution
Post-Islanding Stable
Pre-Islanding Healthy
Corrective Islanded
Actions System
Actions Operation
Determine Plan
Implement Resynchronise
Necessity Islanding Islanding Solution Islands
of Islanding
19
NorthernIrelandExample
1st Moyle Interconnector run back

2nd Moyle Interconnector run back
Frequency (Hz)
Under Frequency
Load Shedding
Time (s)
NorthernIrelandExample
Beforetheevent,thesystemwasoperatingat49.98HzandthedemandontheESB
network(Ireland)was3,302MWofwhich377MWwasimportedfromNIE(N.Ireland).
At10:22aspecialprotectionscheme(SPS)incorrectlydetectedtheseparationofthe
NIEandESBsystemsandinstigatedtherunbackoftheMoyleDCinterconnectorto
Scotland.Note: TheSPSisdesignedtopreventanoverfrequencyinN.Ireland
followingthelossoftheNIEESBACinterconnector. ThepowerflowonMoylechanged
fromimporting115MWtoexporting168MW,i.e.equivalenttothelossof238MWof
generation.Thiscausedthefrequencytodropto49.52Hzandreducedthepowerflow
ontheNIEESBACInterconnectorfrom377MWto215MW.
Note: Thesystemgenerationavailabilitywas4880MWbeforetheeventwhichis
sufficientforsystemrecovery.
Thesystemfrequencythenrecoveredto49.83Hz.
20
CascadingOutagesExample
1
Wind Speed >= Cut Out Speed
Disconnection of Wind Farms
InitiatingEvent
Time(s)
0 1 2 3 4 5 6 7 8 9 10
1
InitiatingEvent
Time(s)
0 1 2 3 4 5 6 7 8 9 10
21
2
Line trips due to

overload
3 Line trips due to
overload
22
4
Line trips due to

overload
5
Generator trips
due to out of
step operation
23
6
Line trips due to

overload
7
Line trips due to

overload
24
7
8
Generator trips
due to out of
step operation
25
9
Generator trips
due to out of
step operation
10
Generator trips
due to out of
step operation
26
CascadingOutageExample
3
G1 L3 L6 L9 L13 L16 L17
2 G3
V olta ge M agnitude (p.u.)
G4 1.1
R o t o r A n g le (ra d )
1 G6
G8
0 G9 1
-1
0.9
-2
-3 0.8
0 1 2 3 4 5 6 7 8 9 10 0 1 2 3 4 5 6 7 8 9 10
Time(s) Time(t)
60.8
60.6 G1 G3 G4 G6 G8 G9
60.4
Frequency (Hz)
60.2
60
59.8
59.6
59.4
0 1 2 3 4 5 6 7 8 9 10
Time(s)
27
ControlledIslandingExample
1
InitiatingEvent
Time(s)
0 1 2 3 4 5 6 7 8 9 10
2
Intentional
tripping of the
lines
28
2
3
Generator trips
due to out of
step operation
29
3
4
Line trips due to

overload
30
0
L1 L3 L4 L9 L11 L14
V o lt ag e M a g n itu d e (p .u .)
-0.5
1.1
R o to r A n g le (ra d )
-1 G1
G3 1
-1.5 G4
G6
G8 0.9
-2
G9
-2.5 0.8
0 1 2 3 4 5 6 7 8 9 10 0 1 2 3 4 5 6 7 8 9 10
Time(s) Time(s)
60.2
60.1
Frequency (Hz)
60
G1
59.9 G3
G4
59.8 G6
G8
59.7
G9
59.6
0 1 2 3 4 5 6 7 8 9 10
Time(s)
Comparison
Without With
Controlled Controlled
Islanding Islanding
Capacity Capacity
Element AssetsLost Element AssetsLost
Lost(%) Lost(%)
Tripped Tripped
1,8,9,10 33.9 1 13.2
Generators Generators
Unsupplied Unsupplied
1,3,4,6 42.0 6 18.1
Loads Loads
1,2,3,4,11, 1,2,4,9,
NonEnergised 17,22,23, NonEnergised 11
28.5 9.1
Lines 24,25,26, Lines
27,28
31
Conclusions
Variousinstabilitytypescontributeinterdependently
toblackoutpropagation
Humanerrorandlackofsituationalawarenessare
amongstthemostsignificantblackoutfactors
Poorsituationalawarenessmaycauseincorrectdecisions
Humanerrormaybehiddene.g.relaysettingsandis
difficulttoquantify,particularlywhenunderpressure
Increaseduncertaintyfromnewtechnologies
Greaterinterconnectionofsystemsforcommercial
reasonsgreaterscaleandriskofblackouts
Conclusions
Studyinteractionsofdifferentformsofstability
duringstressedsystemconditions
Improvedvisualisationanddecisionsupporttools
canreducehumanerrorandhelplimitthescopeof
thecascade
Realtimedynamicsecurityassessmenttocombat
increaseduncertaintywithoutincreasingmargins
Controlledislandingforlimitingspreadofblackouts
andacceleratingrestoration
32
CascadeOutageMechanisms
Vladimir Terzija,
Negar Shams and Peter Wall
vladimir.terzija@manchester.ac.uk
Session2
2. Methodologies forAnalysisofCascadingOutages:
2.1 OnlineMethodsforDirectControlandOperator
Support
Instructor: Mladen Kezunovic
2.2 AssessingandReducingCascadingFailureRisk
Instructor: PaulHines
33
Session2.1
Onlinemethodsfordirect
controlandoperatorsupport
Mladen Kezunovic
TexasA&MUniversity
Understanding Cascading
Phenomenon: Methodologies and
Industry Practice for Analysis of
Cascading Failures
PES General Meeting, Denver, USA
2015 Mladen Kezunovic. All rights reserved.
Outline
Introduction
Consequencesofcascades
Realtimevsplanningdecisionsupporttools
Implementationofonlinetools
Examples
Conclusions
34
Introduction
Introduction
Stagesofcascadingsequences:
Uncontrolled
Initiating Steadystate Transient
islandingand
contingency progression progression
blackout
35
Introduction
Consequencesofcascades
36
RealtimevsPlanningdecision
supporttools
Planningandoperationsplanning:
Vulnerabilityanalysis
RobustAdaptiveTopologyControl
Realtime:
Relaymisoperationdetection
Correctionoftheswitchingstate
Proposedapproach
System Monitoring Local Monitoring
and Control and Control
System Status
Security Analysis Real-time Fault Analysis
Routine-based Local Monitoring Neural Network
Event-based Synchronized Sampling

Security Control Disturbance Report
Relay Operation Monitoring
Steady-State
Transient Stability Event Tree Analysis
Measurements Control Measurements Control
Power System and Protection System

37
ProposedApproach
Weighted
System
Vulnerability
Tool
Analysis
EnhancedFault
Local DetectionMethod
Tool duringPower
Swing
System FastSearch
Methodfor
Tool ParallelCorridor
ImplementationExample
Entergy System
Routine-based Event-based
System Monitoring and Control Tools
Power System Modeling
Entergy system was

targeted for Data Requirement: contingency
implementation PSS/E Raw Data analysis
Data Requirement:
Power Flow Results vulnerability
Data Requirement: System Dynamic Data analysis
PSS/E Raw Data; Simulating the results
System Dynamic Data; for the selected control
Model Selection tools
5. Assess 4. Security Control

Stability Performance
NCF
Steady-state Stability GDF
Analysis
LDF Local Monitoring and Control
Transient Stability
Analysis SMLS
38
Localtools
NeuralNetworkBasedFault
DiagnosisandClassification
39
SystemAnalysisFramework
System Monitoring
NCF
Evaluation
GDF
Disturbance?
N
Y LDF
Identification
Control
Methods SMLS
Vulnerable?
N
Y
SecurityControlScheme
Power System Model
Update Network
Parameters
N Violation Found?
Y
NCF Control
Y
Problem Solved? System Island
N N
GDF Control
Y
Problem Solved?
Y SMLS Control
Problem Solved?
N N
LDF Control Problem Solved?
Y
40
StabilityAnalysis
Power System Model
Update Network
Parameters for
Distrubance
Steady-state Stability
Analysis
N
Stable?
Y
Transient Stability
Analysis
Y
Stable?
Find Optimal Control
Publications
Journalpapers:
1. P.Dutta,A.Esmaeilian,M.Kezunovic,"TransmissionLineFaultAnalysisUsing
SynchronizedSampling,"IEEETrans.PowerDelivery,Vol.29,No.2,pp.942950,
April2014.
2. C.Pang,M.Kezunovic,"FastDistanceRelaySchemeforDetectingSymmetrical
FaultDuringPowerSwing,"IEEETransactionsonPowerDeliveryvol.25,no.4,
pp.22052212,October2010
3. H.Song,M.Kezunovic,ANewAnalysisMethodforEarlyDetectionand
PreventionofCascadingEvents,ElectricPowerSystemsResearch,Vol.77,Issue
8,Pages11321142,June2007.
Conferencepapers:
1. A.Esmaeilian,M.Kezunovic,"EvaluationofFaultAnalysisToolunderPower
SwingandOutofStepConditions,The46thNorthAmericanPowerSymposium
(NAPS),Pullman,WA,USA,September2014.
2. C.Pang,M.Kezunovic,StaticSecurityAnalysisBasedonWeightedVulnerability
IndexinPowerSystem,IEEEPESGeneralMeeting2011.
3. C.Pang,M.Kezunovic,ImprovedSymmetricalFaultDetectingSchemeDuring
PowerSwing,CIGREB5Colloquium2011,Lausanne,Switzerland.
41
Publications
M.Kezunovic,C.Zheng,C.Pang,"MergingPMU,Operational,andNonoperational
DataforinterpretingAlarms,LocatingFaultsandPreventingCascades,"HICCS43,
Hawaii,January2010.
M.Kezunovic,J.Ren,C.Pang,"ImprovingRelayPerformancebyOfflineandOnline
Evaluation,"CIGREActualTrendsinDevelopmentofPowerSystemProtectionand
Automation,Moscow,Russia,September2009.
C.Pang,M.Kezunovic,"DetectionToolsforDisturbancesandProtectiveRelay
OperationsLeadingtoCascadingEvents,"IEEEPESGeneralMeeting,Calgary,Canada,
July2009.
C.Pang,M.Kezunovic,WaveletBasedMethodforFaultDetectionandClassificationof
TransmissionLinesduringPowerSwing,MedPower 2008,Thessaloniki,Greece,Nov.
2008
C.Pang,M.Kezunovic,InformationManagementSystemforDetectingCascading
Events,PowerCon 2008&2008IEEEPowerIndiaConference,NewDelhi,India,Oct.
2008
M.Kezunovic,C.Pang,J.Ren,Y.Guan,"NewSolutionsforImprovedTransmissionLine
ProtectiverelayPerformanceAnalysis,"MELECON08,Ajaccio,France,April2008.
M.Kezunovic,C.Pang,"ImprovedTransmissionLineProtectionDuringCascading
Events,"CIGREB5Colloquium,Madrid,Spain,October2007
QuestionsandComments
Thankyou!
MladenKezunovic
kezunov@ece.tamu.edu
42
Session2.2
AssessingandReducing
CascadingFailureRisk
Prof.PaulHines
UniversityofVermont
IEEEPower&EnergySocietyGeneral
Meeting,July2015
Overview
Whatisrisk?
AssessingRisk
MitigatingRisk
43
Whatisrisk?
Riskhastwocomponents
Probability (likelihood,chance):
Howlikelyisaeventtooccur?
Consequence (impact,cost):
Whatarethe(negative)outcomesofanevent?
Anoftenusefulmeasureofoverallrisk:
Risk=Probability*Consequence
Acommonalternative:
Risk=Probability*Vulnerability*Consequence
Considerabadchoice
Youlostabetandhavetochoosebetweentwo
consequences:
Ifyouroll6youlose$10,ifyourolla2youlose
$1000
Ifyouroll7youlose$10,ifyourolla5youlose
$300
44
Oringeneral
Riskisthesumofprobabilitytimesconsequenceoverall
possiblebadthingsthatmighthappen.
Consideralistofpossible10contingencies,eachofwhich
producesablackoutofsizeS(c):
Butnotallblackoutsare
madeequal
100 year blackout

size: 186 GW
~1/2 of Eastern
Interconnect
45
Weneedtothinkaboutbigblackouts
differentlyfromsmallones
Largecascadingfailuresare
extremelyunlikely,buthave
enormousconsequences.
Socialconsequencesarelikely
greaterthanthesumoflostload.
Weneedtothinkdifferently,
carefullyabouttheriskoflarge
blackouts.
ConventionalReliabilityRisk
Analysis(e.g.,GEMARS)
Takeamodelofthesystem
Defineprobabilitiesforthecomponentoutages
Randomlyselectoutages(andcombinations)usingthese
probabilities
Taketheaverageblackoutsizeasameasureofrisk(Lossof
LoadExpectation)
1in10years
Worksfinefornormaldisturbances.Rarelyproducesbig
blackouts.Fartoocomputationallyexpensiveforcascading
failureanalysis.
46
Sourcesof uncertainty
Failureprobabilitiesofcomponents
Operatorsandenergymanagementsystems
MajoroperationsproblemsinAug.2003(US),Nov.2006(Germany),
Sept.,2011(US)
Uncertainresponsetocomponentstoovercurrent,undervoltage
conditions.
Modelingallofthesepossibilitiesisnotpossible.
Allmodelssimplifytheseuncertaintiestosomeextent
Whichsimplificationsareappropriate?
SplittingMethod
Kim,Bucklew,Dobson,TransonPowerSystems,2013
Insteadofsimulatingfullcascades,simulatecascadesinchunks,
observingthelikelihoodofgrowthfromeachstage
FarmoreefficientthanMonteCarlo.Usefulindescribingriskin
stochasticmodels
47
Alternativemethod,basedon
findingdangerouscontingencies
PooyaRezaei,PaulHinesandMargaretEppstein,
EstimatingCascadingFailureRiskwithRandom
Chemistry,IEEETransactionsonPowerSystems,2015.
MargaretJ.EppsteinandPaulD.H.Hines.ARandom
ChemistryAlgorithmforIdentifyingCollectionsofMultiple
ContingenciesthatInitiateCascadingFailure.IEEE
TransactionsonPowerSystems,vol.27,no.3,2012.
Start with a grid model.
48
Findmanyoftheoutagecombinationsthat
causeblackouts(themalignancies)
TheRandomChemistryalgorithm
40
outages
20
outages
10
outages
5
outages
2-3
outages
49
Usetheresultstoquantify
blackoutrisk
The estimated number of
malignancies of size k
Blackout sizes
Probability
of (multiple)
The number of contingency
malignancies of size k
found by RC
ComparingRCtoMonteCarlo
50
Nowthatwecanestimateblackout
risk,whatinsightcanwegain?
Riskvs.load,givenSCOPF
Adding the SCOPF

changes the results
substantially from prior
work showing a phase
transition in risk vs. load
51
Why?
AthighloadlevelsSCOPFleaveslargermarginsonlong
interareatielines(toallowforpotentialcontingencies)
Total absolute flow on lines with large (>200MW)

base case flow
Canweusethisinsightto
reducerisk?
Takethe3linesthatcontributemosttoblackoutrisk
Redispatchgeneratorstoleavemoremarginbetweenthe
flowontheselinesandthelimit(cutthelimitinhalf)
Fuelcostsincreaseby1.6%
Large(S>5%)blackoutriskdecreasesby61%
Verylarge(S>40%)blackoutriskdecreasesby83%
Perhapswewouldbebetteroffwithouttheselines?
52
Conclusions
Cascadingfailuresrequirenewapproachestoriskanalysis.
Themanyuncertaintiescomplicatetheproblem.
However,deepinsightandtestablemitigationstrategies
canresultfromcarefulanalysis.
AssessingandReducing
CascadingFailureRisk
paul.hines@uvm.edu
53
Session3
3. FrameworksandCurrentToolsforDetection,
Prevention,andMitigationofCascadingOutages:
3.1 FrameworksandIndustryToolsforAnalysisof
CascadingFailures
Instructors: MiloradPapicandMariannaVaiman
3.2 AnalyticalApproachesandToolsforPreventionand
MitigationofCascadingOutages
Instructor: MariannaVaiman
Session3.1
FrameworksandIndustryToolsfor
AnalysisofCascadingFailures
MiloradPapic,IdahoPowerCo.
MariannaVaiman,V&REnergy
54
Overview
Driversbehinddevelopingthetoolsforcascading
analysis
Frameworksforanalysisofcascadingoutagesas
modeledinthetools
Industrypracticesforanalysisofcascadingoutages
Industrygradetoolsdesignatedforcascadinganalysis
Researchgradetoolsdesignatedforcascading
analysis
Futureofcascadingoutagestools
DriversbehindDevelopingtheTools
Compliancewithstandards
Improvingreliabilityofthesystem
Reinforcingthesystemandminimizingtheimpactof
futureblackouts
Studyingpastblackoutsandlearnfromthelessons
We had a power outage today ... and my PC, TV, DVD, &
surround sound music system were all shut down. Then I
discovered that my phone battery was flat and I couldn't
charge it. To top it off it was snowing outside. So I couldn't
play golf and I couldn't fish. I went into the kitchen to
make coffee and then I remembered that this also needs
power. So does the microwave. So popcorn won't happen.
So I talked with my wife for a few hours. She seems like
such a nice person.
55
ContributionofCascadingOutages
toBlackoutRisk
Cascadingfailurescontinuetocontribute
significantlytoblackoutrisk:
Changesintransmissionsystem Increaseddependencyinpower
operationspolicy systemoperationonagreater
Continueddifficultyinobtaining numberofindividual,independently
permitsfornewtransmission owned,entities
lines Limitedflexibilityinmuchofthe
Increasedneedforquantified existinggenerationcapacity
economicjustificationofactions Increaseduncertaintyinpower
bypowersystemplannersand transfersduetouncertaintyinwind
operators generation
AssessmentofCascadingOutages
Sincecascadingisverycomplicatedandcomplete
enumerationofallpossibilitiesisimpossible,there
arenecessarilycompromisesandlimitationsin
assessingcascadingriskandimpact
Normally,atradeoffbetweenaccuracyandspeedis
required
Theresultofthistradeoffislikelytodependonthe
timescaleinwhichcascadesareanalyzed
56
SomeoftheAssumptionsinSimulating
CascadingOutages
Modelonlyasubsetofinitiatingevents.
Modelonlytheinitialstagesofcascading.
Modelonlythemostprobable,ormostconsequential
entirecascadingsequences.
Assumethatcascadesproceeddeterministically.
Q: What would you call a power failure?

A: A current event.
FrameworksforAnalysisofCascading
Outages
Cascadinganalysis
frameworkconsistsof
twocomponents:
Identifyinginitiating
eventsthatmayleadto
cascading(steadystate
contingenciesandfaults);
Identifyingthespreadof
thecascadingoutages.
57
SimulationofCascadingOutagesfrom
SteadyStatePerspective
Fastsequentialcontingencysimulationisusedtoidentify
potentialcascadingmodes
Outagesareconsecutivelyapplieduntil:
(1)Systemfailstosolveduetovoltageinstability;
(2)Thermal/voltageviolationsarealleviatedordropbelowthe
thresholds.
Islandsareidentified
Lossofloadandgenerationismonitoredandreported
Probabilitiesofinitiatingeventsandconsequencesmaybe
added
AnalysisofCascadingOutagesdueto
ThermalOverloads
58
AnalysisofCascadingOutagesdueto
VoltageProblems
CombiningSteadyStateandTransient
StabilityAnalyses
Thetoprankedcascadingeventsasdeterminedbythe
steadystatescreeninganalysisareprocessedthrough
dynamicsimulations
Fastfaultscreeningmethodologymaybeusedto
identifypotentialmostseverefaultsinthesystem.These
faultsarealsoprocessedthroughdynamicsimulations
Ifasteadystatecascadecausedsystemcascading
instability,itisnotbeinganalyzedusingdynamic
simulations
59
StabilityAnalyses(cont.)
Initiatingeventsthatcausedcascadingchainsthatdidnt
leadtosteadystateinstabilityareprocessedthrough
dynamicsimulations
Thefollowingquestionsareaddressedduringanalysis:
Whetherrelaysoperatedornot;
Whattheimpactongeneratorsis.
Undervoltageloadshedding(UVLS),underfrequencyload
shedding(UFLS),generatortripping,andotherprotection
actionsshouldbeimplementedduringsimulationof
cascadingoutages
CascadingfromTransientStability
Perspective
TheeventisNOTclassifiedasacascadingoutageincase
ofthefollowing:
Dampingishigh,
Relayshavenotoperated;
Sustainedoscillationsareofsmallamplitude.
Q: Why did the lights go out?

A: Because they liked each other!
60
CascadingfromTransientStability
Perspective(cont.)
Theeventisclassifiedasacascadingoutageincaseofthe
following:
Sharpdropintransientvoltagesinalargepartofthenetwork,
Sharpdropinfrequencyfollowedbysystemseparation,
Islandsareformedasaresultofcascadingand/orprotection
operation,withsignificantamountofload/generationwithinthe
island,
Disconnectionoflargeamountofgeneration,
Disconnectionoflargeamountofload,
Sustainedoscillationswithlargeamplitude,whichdonotcause
protectionsystemtooperate.
CascadingFailureWorkingGroup
SurveyofIndustryPractices
Asurveyconsistsof15questions;someofthemare:
Howoftenistheanalysisofcascadingoutagesperformedinyour
organization?
Inwhichdomaindoesyourorganizationstudycascadingevents?
Iscascadingoutageanalysisanautomatedprocess?
Doyouanalyzecascadingoutagesusingsteadystateanalysistools?
Doyouanalyzecascadingoutagesusingdynamicsimulations?
Doyouusehistoricaloutagedatainyourorganizationtoanalyzecascading
outagesandtheirconsequences?
Doyoumodelortakeintoconsiderationtheprobabilitiesofinitiatingevents?
Doyouusethemetrics/indicatorstoassesstheeffectsofcascadingoutages?
Doyouapplymitigationmeasurestoalleviateconsequencesofcascading
outages?
61
NERCStandardsthatAddressSomeofthe
CascadingOutageAspects
TPL0014 Transmission System Planning Performance Requirements
CIP0141 Physical Security
CIP0025.1 Cyber Security BES Cyber System Categorization
PRC0232 Transmission Relay Loadability
PRC0241 Generator Frequency and Voltage Protective Relay Settings
EOP0023.1 Capacity and Energy Emergencies
EOP0032 Load Shedding Plans
TOP0012 Transmission Operations
TOP0042 Transmission Operations
FAC0033 Transmission Vegetation Management
FAC0102.1 System Operating Limits Methodology for the Planning Horizon
FAC0112 System Operating Limits Methodology for the Operations Horizon
IRO0081 Reliability Coordinator Operational Analyses and Realtime Assessments
IRO0101a Reliability Coordinator Data Specification and Collection
SoftwareToolsforAnalysisofCascading
Outages
Commerciallyavailabletools
Research gradeavailabletools
Methodology:analytical&MonteCarlo
ACorDCpowerflowmodel
Steadystate&transientstabilityanalyses
Analysistimescales:realtimeoperation;operational
planning;longtermplanning.
Limitationsonthesizeofthesystem
A lot of progress has been made in the area of getting
computers to write code. Our latest system is so advanced that
when it runs into a road block it blames the previous computer.
Copyright 2013 Mark W. Lund
62
CommerciallyAvailableTools
AC/DCPOWER MAX.NUMBER
CASCADING TOOL METHODOLOGY
FLOW OF BUSES
ASSESS Analytical+Monte DCorACsteadystate Practicallimitof

byRTE,France&NationalGrid,UK Carlo plusdynamic around2000
simulation buses
CAT Analytical AC 64,000
byCommonwealthAssociates,Inc.
(CAI),USA
ISAP Analytical+Point AC Largesttested
byRicerca sul Sistema Energetico EstimateMethod gridwith1500
S.p.A.(RSE),Italy buses
POMPCM Analytical ACsteadystateplus NoLimit
byV&REnergy,USA dynamicsimulation
TransCARE(SuccessortoTRELSS) Analytical ACorDC 75,000

byElectricPowerResearchInstitute
(EPRI),USA
ResearchGradeAvailableTools
AC/DCPOWER MAX.NUMBER OF
CASCADING TOOL METHODOLOGY
FLOW BUSES
HIDDENFAILURE, MonteCarlo AC 300
USA
MANCHESTER MonteCarlo AC 1,500

byTheUniversityofManchester,
UK
OPA MonteCarlo, DC 1000
byORNLPSERCAlaska,USA complexsystem
PSA MonteCarlo ACorDC 64,000

byLosAlamosNational
Laboratory,USA
TAM MonteCarlo AC 24
byTexasA&MUniversity,USA
63
ASSESSbyRTE,FranceandNGT,UK
Providesasinglesoftwareenvironmentinwhichtheusercanspecify,quite
precisely,averywiderangeofuncertainties,andallowtheirimpacttobe
exploredquitesystematically
Allowstheapplicationofstatisticalanalysisanddataminingmethodstothe
specificneedsofpowersystemsecurityassessment(systemplanningand
operationplanning)
Allowsthepossibilityofmodelingsequencesofevents,whetherindependent
orconsequentialtothecurrentstateofthesystem,andvaryingarangeof
systemparameterssuchasprotectionsettings,lineratings,faultclearance
timesetc.
Assessesthepossibilityofcascadingoutagesoccurringandtheirimpact
Generatessystematic/randomsituationsmodelingtheuncertainties.Itis
possibletomodeluncertaintiesonanyvariabledefiningthestudiednetwork
(forexample:consumptioninazone,linesandgeneratorsavailability/
unavailability,generationofawindturbinesite,devicesmaintenance,load
shedding,excitertimeconstant,etc.)
CATbyCAI
CAT(CascadeAnalysisTool)isapartoftheTransmission2000suiteof
programsdevelopedbyCommonwealthAssociatesInc.
CATautomaticallyrunsasetofcontingenciestodeterminethepotential
toinitiatefacilityand/orloadlossesbeyondinitialcontingency.
Foreachcontingency,theToolchecksthepostcontingencyoperating
stateagainstuserspecifiedcriteria.Themostcommoncriteriathatmight
beusedinananalysisare:thermaloverloadcriterion,lowvoltage
criterionandvoltagechangecriterion.
Forcontingenciesthatcausethepowerflowtodiverge,orifanystepis
takentorelieveaviolationcausesdivergence,loadisdroppedatthebus
associatedwiththedivergenceandanotherattemptismadetosolvethe
case.Theprocessrepeatsuntilcertainconditionsarereached.
64
ISAP(IntegratedSecurityAssessmentPlatform)
Importfrom
Deterministicstaticanddynamicsecurityassessment(SSA&DSA): PSS/E,
MATPOWER,
Component/contingencyorientedSSAindices,gridloadability,staticvoltage
IEEECDF
stability
Transientstabilityindices,SecurityOperatingLimits(SOL),CriticalClearingTimes
Ranking lists
UnifiedSecurityanalyses:howmuchtimeforcontrolbeforeinstability/cascading
starts?
Assessmentoftechnicalriskinpowersystemoperation:
EffectofrenewableandloadforecastuncertaintiesoninitialPSstate Angle based Instability Impact Index (dt = 10 minutes)
Impact value
Usingprobabilisticmodelsofnaturalormanmadethreatsand
0 0.002 0.004 0.006 0.008 0.01 0.012
N-1_CRCP211_TIMP211
SSB2_CORP211_stuckCB_FAULT_ON_L_PP2204
SSB2_CORP211_BUSFAULT_MSig_CB_L_PP2232
N-1_CORP211_SFMP221
SB_CMRP211_HB1_stuck_BusCoupl
SB_CMRP211_HB2_stuck_BusCoupl
SSB1_CRCP211_stuckCB_FAULT_ON_T_CRCP211A2
vulnerabilitiestoidentifymostdangerouscontingencies andtheir
SSB1_CRCP211_stuckCB_FAULT_ON_L_PP2226
SSB2_CORP211_BUSFAULT_MSig_CB_L_PP2209
SSB2_CORP211_no_signal_to_one_CB
SSB2_CORP211_BUSFAULT_stuckCB_L_PP2232
N-1_CORP211_SFMP231
SSB1+DT_CRCP211_L_248 _stuckCB
SSB1+DT_CRCP211_L_211 _stuckCB
probabilities SSB2_CORP211_BUSFAULT_stuckCB_L_PP2209
SS C C S
SB_CMRP211_BDP_OOS
SB_CMRP211_BDP_OOS
SSB2+DT CORP211 L 210 stuckCB
S C
histograms
ContingencyImpactevaluatedbyspecialSSAandDSAindices
Bubble plots
Cascadingsimulatortoassessblackoutrisk: 700
Bubble plot of risk contributions
71
isorisk = 0.0001
CascadingpathisanalyzedbyACPowerflowtoolwithsteadystatemodelsof 600
Contingency Impact, Loss of Load in MW

controls 500 isorisk = 1e-07
400 78
isorisk = 1e-10
Investigatingcascadingpaths & modelinguncertaintiesdueto: 300
(1)Hiddenfailures,(2)protectionsettings,(3)Operatorsdelayto 200
89
deploycontrols(loadshedding) 63 58
57
100 isorisk = 1e-13
427273 70
62
61
60
59 51
50
55
54
46
45
44
43
49
48
53
52 30 56
29
28
27
36
35
34
33 75
7677
65
67
47 8788
8485
91
93
99
101
41
40
39
38
2582 92
24
31
37 90
98
100
6417
6620
19
18
15
14
11
10
74 8021
16
7 86
83 81
795
63
41
2
0 isorisk = 1e-16
-15 -10 -5 0
10 10 10 10
Contingency Probability
POMPCMbyV&REnergy
PCM(PotentialCascadingModes)isadesignatedtool
foranalysisofcascades
Offersvariousoptionstoidentifyinitiatingevents:
AlistofN1and/orN2contingenciesidentifiedas
aresultoftheanalyticalclusterapproach;
AlistoffaultsbasedontheFastFaultScreening;
Auserspecifiedcontingency;
Combinationoftheabovelists.
Combinessteadystateandtransientstabilityanalysis
Modelsrelayoperation
Automaticallyidentifiedvarioussetsofremedialactions
topreventthecascades.
Veryfast canbeusedinplanning,operations,realtime
Outputincludescriticalcascades,load/generationloss,
islandreporting.
65
TransCAREbyEPRI
TransCARE(TransmissionnetworkContingencyAnalysisandReliability
Evaluation)isusedforreliabilityassessmentofcompositegenerationand
transmissionsystems.
TransCAREcanautomaticallyplacebreakersanddeterminePCGs
(ProtectionandControlGroups).APCGisasetoflines,transformerand
generatorsprotectedbyacommonprotectionsystemapproximatinga
primaryzoneofprotection.
CascadingfailureanalysisinTransCAREaimstocapturethecascadepath
startingfromanusersuppliedinitiating(triggering)event.Thresholdvalues
forthermaloverloads,lowvoltagesatwhichanoverloadedlineor
generatoristripped,orabusloadisdroppedalsoneedbespecified.
Themodelsimulatesthecascadingprocessasasequenceofquasisteady
statesystemconditionscausedbyasequenceoftrippingevents.
WhenalineistobetrippedthenthePCGthelinebelongstoistripped.
MainoutputresultsfromprogramareriskindicesexpressedinMWand
criticalcorridorsidentifiedbysimulatingthecascadingsequencesofevents.
Future:WhyPMUs?
PMUsareusedforWideAreaMeasurementSystems
Bringnewmonitoringcapability
Functionalitiestopredictcascadingoutagesinclude:
Earlydetectionofevents
Variationsofreactive/activeinjections
Complementstheinformationcomingfrombreakerstatussignals
Voltagestabilityanalysisoninterfaces/corridors
UsestheV,Imeasurementsatbothendsofonelinecorridorandthe
maximumpowertransfercomputation
Providesthevoltagestabilitymarginwithrespecttomaximumtransfer
condition
PhaseAngleMonitoring
Monitorshighangledisplacements,todetecthighlyloadedlines
Oscillatoryanalysis
Predictsunstableoscillationswhichmaytriggerlinetrippings
66
PMUApplicationsforCascadingPrediction:
theUSExperience
Phase angle and voltage monitoring
Using PMU measurements to compute
system steady-state stability limits and
predict proximity to voltage collapse
Source: M.Ya. Vaiman, M.M. Vaiman, S. Maslennikov, E. Litvinov, X. Luo, Calculation and
Visualization of a Power System Stability Margin Based on the PMU Measurements, 2010
IEEE SmartGridComm:31 - 36
UseofPMUsforAnalysisofCascades
Predictionofslowcascadingoutages:
Maybeanalyzedfromsteadystatestabilityperspective
Themostsensitivephaseanglesareidentifiedinreal
timeforeachscenario/interface/corridor:
Thesequantitiesaremonitored,reportedandvisualized
Maychangeovertimeasthesystemconditionschange
Theaccuracyofthelimitvaluescomputedofflinemay
beimprovedbyusingrealtimePMUmeasurements
Thesevaluesareadjusteddynamically
Fortransientstabilityrelatedoutages,addingtime
synchronizedmeasurementsofthegeneratorrotor
anglewillenablenewprotectionandautomatedcontrol
67
FutureCascadingOutagesToolsWill:
Usesynchrophasors toproviderealtimetoolsforoperatorsand
reliabilitycoordinators
Automaticallyperformanalysisofcascadingoutagesthatare
transientlylimited
Fullysimulateprotectiverelaysoperation
IncorporateSPSandoperatingprocedures intotheanalysis
Modeltheactionsofhumanoperatorsandcomplexinteractions
Obtainasufficientlyuniformandnumerousjointsampleofinitial
conditions,initiatingevents,andeventsequences
Addressinsystematicwaystheuncertaintiesinthepowersystem
modelforplanningandoperationstudiesbyincorporatingrisk
basedandprobabilistictechniques
Questions?
68
Session3.2
AnalyticalApproachesandTools
forPreventionandMitigationof
CascadingOutages
MariannaVaiman,V&REnergy
Overview
1. MethodologiesforMitigation
2. SpecialProtectionSchemesCurrently
DeployedforMitigationofCascades
3. SuccessfulAndUnsuccessfulMitigation
ofCascadingOutages:LessonsLearned
4. FutureDevelopmentsintheAreaof
MitigatingCascadingOutages
5. Conclusions
69
1.MethodologiesforMitigation
CanWeDoBetter?
70
AMethodologyforMitigationinUS
Identifypossible
initiatingevents,
theirspread,
andseverity.
Identifyexisting
resourcesinthe
systemthatmight
besufficienttopreventacascadingoutage.
Applyeffectiveislandingtechniques.
Ifablackoutcantbeprevented,identifyaneffective
blackstarttechnique.
MitigationMeasuresforCascading
Outages
Determineoptimalmitigationmeasures:
Topreventcascading:
Applymeasuresbeforethecascadingstarts
Applyaftertheinitiatingeventandateachtier,asneeded
Tomitigatetheconsequencesofcascading:
Aftercascadinghasoccurred
71
AnalysestoPreventCascading
Generation
trippingtoprevent
cascadingmaybe
determinedbased
oncombinationof
steadystateand
transientstability
analyses
CorrectiveActionsforSteadyStateAnalysis
Automatedanalysis
Correctiveactionsarebasedonauser
definedpriorityschedule
Availableremedialactionsinclude:
MWDispatch
MVarDispatch
CapacitorandReactorSwitching
TransformerTapChange
PhaseShiftersettings
LineSwitching(InandOut)
LoadCurtailment
OptimalCapacitor,ReactorPlacementand
Size
72
UseofCorrectiveActions
duringCascadingAnalysis
Correctiveactionsare
automatically
determinedand
appliedduringanalysis
ofcascadingoutages
Determineremedialactions
toalleviatevoltagecollapse
aftercascading
Mitigatetheeffectofa
cascadingoutageafterallthe
tiersincascadinghappen
Loadcurtailmentmeasureis
used
Theapproachmaybeused
forrankingcascading
outages
73
AMethodologyforMitigationin
Europe
EuropeanNetworkofTransmissionSystemOperators
forElectricity(ENTSOE)providesrecommendationsfor
automaticactionstopreventtheContinentalEurope
SynchronousAreaorpartsofitfromthelossofstability
andcascadingeffects
ENTSOEhasproposedclassificationmethodologyaimed
torankgriddisturbances.
Afourdegreescalehasbeensuggested:
Rangingfromlocaleventswithloweffectonreliabilityto
widespreadandmajorincidentsinonetransmissionsystem
operator(TSO)
PreventiveMeasuresforCascading
Outages
DEMO
74
2.SpecialProtectionSchemes
CurrentlyDeployedforMitigationof
Cascades
RemedialActionSchemes
NERCGlossaryDefinitionofRAS:
Anautomaticprotectionsystemdesignedtodetectabnormalor
predeterminedsystemconditions,andtakecorrectiveactions
otherthanand/orinadditiontotheisolationoffaulted
componentstomaintainsystemreliability
NERCstandardsPRC012 PRC017addressissues
relatedtoRASundertheProtectionandControlcategory
RAS=SIPS=SPS
RAS RemedialActionSchemes;
SIPS SystemIntegrityProtectionSystem;
SPS SpecialProtectionSystem
75
RemedialActionSchemesinWECC
Thereareover190
RASinWECC
transmission
system:
Designed,
maintained,and
evaluatedin
accordancewith
theWECCRAS
Guideand
Proceduretosubmita
RASforassessment
UseofRASatWECC
RASisusedatWECCto:
Ensureadequatesystemreliability,
Maintainorincreasethetransmissionsystem
capability,
Mitigatelowprobability/highconsequencesystem
eventsresultingfromNERCCategoryCandD
contingencies,
Preventeventsspreadingoutacrosslargeregionsor
systemwidebasis.
76
TheMostCommonRASinWECC
Generationtrip,
Brakeinsertion,
Fastvalve/genramp,
HVDCramp;
Configuration
changes/islanding,
Loadshedorrejection,
Excitationforcing,
Shuntcapacitor/reactor
switching,
Seriescapacitor/reactor
switching
TypesofRASinWECC
WECCidentifiesthreetypesofRASs,dependingon
theirpotentialimpact:
62%areLocalAreaProtectionScheme(LAPS)
LAPSisusedtomeetanowner'sperformancerequirementswithintheir
system.
31%areWideAreaProtectionScheme(WAPS)
WAPSisneededtomeetWECCperformancerequirementsand
operatingstandards.
7%areSafetyNet(SN)
SNschemeprovidesdefenseagainstextensivecascadingorcomplete
systemcollapse.
SNisintendedtominimizetheimpactofextremeeventswhensuch
impactscannotbeentirelyavoided.
77
SpecialProtectionSchemesinERCOT
TheERCOTOperatingGuidesdescribeSPSsas
"protectiverelaysystemsdesignedtodetectabnormal
ERCOTSystemconditionsandtakepreplannedcorrective
action(otherthantheisolationoffaultedelements)to
provideacceptableERCOTSystemperformance
SPSactionsinclude:
Changesindemand,
Changesingeneration,or
Systemconfiguration.
AnSPSdoesnotinclude:
Underfrequencyorundervoltageloadshedding.
SpecialProtectionSchemesinItaly
ItalianDefensePlanconsistsoffourLinesofDefense
Includesremedialactionsaimedat:
Preventingcascadetrippingandconsequentuncontrolled
networkseparations:
Fasttrippingofcriticalgeneratingunits;manualemergencytrippingofMV
&HVloads;blockingofonloadtapchangers;
SystemforAutomatic Sheddingtoavoidcascadingon"criticalsections"
definedassetsof400kVlinessothattheircascadetrippingcould
evolvetonetworkseparation
Limitingtheimpactofnetworkseparationincasemeasures
identifiedabovedonotmeettheirtarget.
Remedialactionsaredefinedbyofflinesteadystate
andtransientstudies
78
3.SuccessfulAndUnsuccessful
MitigationofCascadingOutages:
LessonsLearned
CaseofUnsuccessfulMitigation 1
In2012,thelargestcaseofunsuccessfulmitigation
occurredinIndia
Lossofnearly700millioncustomers
Sequenceofevents:
Aseverelyweakenedsystemcoupledwithlargeunscheduled
interchangesledtohighlyloadedtielines.
Loadencroachmenttrippedthesetielinesafterinadequate
operatorreliefactions.
Theresultingpowerswingssplitupthesystem:
Linescontinuedtotripfromunderfrequency/overvoltageactions;
Thiseventuallycausedtotalcollapseofallthreegrids.
79
CaseofUnsuccessfulMitigation 2
In2011,asignificantdisturbanceinWECC(WesternUS)
Disconnectionof2.7millioncustomers
Sequenceofevents:
ThesystemwasnotoperatinginanN1securestate:
peakdemandhoursandlowerthanpeakgenerationcombinedwitha500kVline
trip
Theabovecausedsizeablevoltagedeviations,equipmentfailureanda
cascade
SONGSseparationschemeoperated,reconfiguringtheSONGS230kV
switchyardandisolatingtheSONGSgeneratorsontotheSCEsystemtothe
north.ThisreconfigurationeffectivelyseparatedallfiveSouthofSONGS230
kVtransmissionlinesfromtheSONGSunitsandtheSCEsystem
Coordinationissueswiththeexistingprotectionsystems
contributedtotheeventprogression
CaseofSuccessfulMitigation 1
In2008,arareeventontheUKnetworkresultedin
frequencybeingoutsidelimitfor9minutes.
Sequenceofevents:
Twolargegeneratorstrippedwithin2minutes,followedbytwo
moreunits;
Thislossandfurthertrippingofembeddedgenerationinthe
distributionsystemcaused frequencytodropto48.795Hz.
Frequencydropwasstoppedbyloadsheddingschemes
andNationalGrid:
Restoredsystemfrequency;
InstructedaffectedDistributionNetworkOperatorstorestorethe
droppedloadwithinarangeof2040minutes
80
CaseofSuccessfulMitigation 2
In2006,amajordisturbanceinEurope
Sequenceofevents:
TheeventwasinitiatedwithaplannedoutagebyE.ONNetz
OnetielineconnectingE.ONNetz andaneighboringTSOuseddifferent
relaysettingsineacharea
TielinetrippedandinitiatedcascadesintheUCTEsystem
Duetoovercurrentdistanceprotectionandoutofsynchronismrelays
TheUCTEsystemsplitintothreeasynchronousareas
BlackoutwasavoidedduetotheactionsofTSOsintheir
individualcontrolareas:
Restorationofnormalfrequencywaswithin20minutes
LessonsLearned
Commonfactorinmanyblackouts lackofcoordination
andinformationbetweenTSOsoperatinginan
interconnectedregion
Recommendationspointtowards:
Increasedcoordinationbetweenoperatorsintermsofprotection
settings,realtimeexchanges,systemstudies,andsystem
conditionsofneighboringTSOs;
NeedforRASandSPStobeproperlycoordinatedfor
protectionwithintheTSOsandwithinterconnectedregions.
Maintenancepracticeandschedulesplayanimportantrolein
blackoutprevention.
81
4.FutureDevelopmentsintheAreaof
UseofPhasor MeasurementUnits(PMUs)
forPreventionofCascades
PMUmeasurementsallowforfasterandmoreaccurate
relayoperationandenablingRAS
Wideareaoscillationdampingcontrol
Advanceddefensefunctions,likecoordinated widearea
loadshedding actions,controlledislanding,etc
Noconsolidatedsolutionssofar
82
ImportanceofCoordination
Coordinationbetweenutilitiesprovidestheopportunityfor
futuremitigationmeasures:
HarmonizationamongUFLSs;developingastandardforthe
blockingofOnLoadTapChangersandforUnderVoltageLoad
Shedding(UVLS)intheCESynchronousArea.
Effectivecoordinationbetweendifferentprotection
schemesisalsoimportant:
Consequencesofanactionortheircombinationsunder
contingenciesinstressedsystemconditionscouldbe
significant,
Difficulttomodel,computeandunderstand.
Conclusions
Advancedresearchandtechnologydevelopmentinthe
areaofmitigationandpreventionofcascadingoutagesis
neededtoreducetheiroccurrences,spreadandimpact
Effectofvariousprotectionschemesshouldbeanalyzed,
andtheirsettings/operationshouldbecoordinated
Synchrophasortechnologywillplayasignificantrolein
preventing/mitigatingcascadingoutages
83
Session4
4. EmergingTechnologiesforDetection,Prevention,
andMitigationofCascadingOutages:
4.1 WideareaMeasurementSystemstoPrevent
PropagationofCascadingFailures
Instructors: Damir Novosel andVahid Madani
4.2 UseofSynchrophasorTechnologyforAnalysisof
CascadingOutages
Instructors: Vahid Madani andDamir Novosel
Session 4.1
Wide-area Measurement Systems to
Prevent Propagation of Cascading
Failures
DamirNovosel QuantaTechnology
VahidMadani PacificGas&ElectricCo.
Tutorialon
UnderstandingCascadingPhenomenon
84
AdvancedWarningSystems toPreventBlackouts
MultipleContingencieswithComplexInteractions
GlobalPowerOutagesHighlights CommonThreads
Italy, 2003: 57M people NE-US Can, 2003 50M o Brazil & Paraguay, S. California, Arizona, India, 2012 620 Million -
people 2009, 87 Million People Mxico, 2011, US, Largest outage ever
o Heavy import to Italy
o 2 hours before 2011 2.8M people o Heavy power transfers
o Three key
disturbance 500kV line transformers short o Heavy import to San due to weak inter-
disconnect Diego regional corridors due
circuited in Brazil
to planned and forced
o Heavy power flow in o Crucial high voltage o 500kV line trips outages
region lines disconnected - o Generation tripped in o High Loading on 400 kV
Instantaneous loss of Mexico link, due to
14 GW of power,
o Heavy N-S flow on Path unscheduled
climbing to 28GW
44 interchange, not
corrected
o One 380 kV line sags into a o One 500 kV line sags into o Power loss causing a o Two transformer o 400kV line trip on
tree and disconnects a tree and disconnects Secondary incident, units & two 161kV zone 3 relay
shutting down power lines trip on overload disconnecting North
o Heavy load through o Heavy load 230+115 kV
generation at Itaipu from South
parallel line that sags into lines o More 161 kV lines
(largest hydroelectric
tree and tripe
o 230+115 kV lines dam in operation) and units (Mexico) o Heavy load on
o 220kV/110 kV trip on disconnect due to shared with Paraguay trip (loss of 420 MW) transmission lines
overload, isolating Italy overload causing cascading
o Path 44 trips on
o 345 kV lines trip on failures
overload
overload
o Voltage declines and o Voltage declines and o Voltage declines, o Voltage declines, power o Voltage declines
power units trip power units trip causing cascading units trip San Onofre causing trip of all major
separations power loss and others power stations
o Power oscillations and o Power oscillations and
voltage decline cause voltage decline cause o Power oscillations & o Power oscillations and
cascading separations cascading separations voltage decline cause voltage decline cause
Blackout occurred in
cascading of 230 kV further cascading
Blackout occurred in 2.5 Blackout occurred in 3 minutes
lines separations
minutes minutes
System restored in
Blackout occurred in 11 Blackout occurred in
System restored in System restored in
~3.5 to 6 hours minutes
3 minutes
~5 hours ~1-2 days
System restored in
80% of System restored in
~1 day 15 hours
85
Recommendations
After2003U.S.&2006EuropeanOutages
August 14, 2003 Outage: U.S.-Canada Power System Outage Task Force Report
"A valuable lesson is the importance of having time-synchronized system data recorders. The Task
Forces investigators labored over thousands of data items to determine the sequence of events, much like
putting together small pieces of a very large puzzle. That process would have been significantly faster and
easier if there had been wider use of synchronized data recording devices"
Recommendation 12a The reliability regions, coordinated through the NERC planning committee,
shall within one year define regional criteria for application of synchronized recording devices in
power plants and substations
November 4, 2006 Disturbance - UCTE Final Report System
UCTE-wide Awareness System
" On November 4, this was true more than ever the information about the split of the system into three
areas was available to some operators with significant delay. This issue might be solved via a dedicated
central server collecting the real-time data and making them available to all UCTE TSOs. In this way, each
TSO will obtain within a few minutes essential information about disturbances beyond their own control
area. "
Recommendation 4 - UCTE has to set up an information platform allowing TSOs to observe in real
time the actual state of the whole UCTE system in order to quickly react during large disturbances.
November2006Europe:SynchronizedData
Before After
Separation Separation
86
India,July30and31,2012
WeakInterregionalCorridorsdueto
multipleoutagesoftransmissionlines
intheWestNorthinterface.
o 400kVBinaGwaliorAgrawastheonly
August 2006: mainACcircuitavailablebetweenWest
North synchronized
With Central Grid Northinterfacepriortothegriddisturbance
HighLoadingon400kVBinaGwalior
Agralinkduetounscheduled
Northeast
North
interchange
East
o Inadequateresponsetoreduceloading
West
Lossof400kVBinaGwaliorlinkon
South Before 1991:
Five Regional Grids zone3distancerelaycausedtheNorth
Five Frequencies
toseparatefromtheWest
Source: Report from the Enquiry Committee on Grid Disturbances
in Northern Region, India, September 2012
IndiaBlackout KeyFindings
Bettervisualizationandplanningof
thecorrectiveactions
DeploymentofWAMPAC
Betterregulationofinterchanges
Bettercoordinatedplanningofoutagesof
stateandregionalnetworks,specifically
underdepletedconditionofthe
interregionalpowertransfercorridors
ActivationofprimaryfrequencyresponseofGovernors
Adequatereactivepowercompensation,specificallydynamic
Underfrequencyanddf/dtbasedloadshedding
Avoidmissoperationofprotectiverelays
Source: Report from the Enquiry Committee on Grid Disturbances in Northern Region, India, September 2012
87
SanDiegoBlackout Sep.2011
Over30'majorelement'operationsin11minutes
Weaknessesintwobroadareas
o Operationsplanning
o Realtimesituationalawareness
Contributingfactors
o Notstudyingimpactofsub100
kVfacilitiesparalleltoEHV
o FailuretorecognizeIROLs
o Notstudying/coordinatingeffects
ofprotectionsystemsandRASs
duringcontingencyscenarios
Outage
started
o Notprovidingeffectiveoperator
here toolsandinstructionsfor
reclosinglineswithlargephase
angledifferences
Source: Bob Cummings,

NERC
PMU Measured Phase Angles

Hass.- Rest of WECC
N.G. Trip System
SONGS
SONGS
Coachella Separation
Valley Trip
Miguel
Ramon
Trip
Using Tesla as the reference angle, zeroed
Source: Bob Cummings, NERC
88
PreventingBlackouts
Widespreadelectricoutages
areasymptomofstrategies
forgridmanagement
Analysisofrecentdisturbances
revealscommonthreads
o Learnfromthepastandprovenmethods
tomitigate
New York City on October 31, 2012 o Blackoutpropagationshould
Photographer Iwan Baan
Image published in New York Magazine bearrested
o Restorationtimecouldbereduced
Use of Synchronized measurements for Improved Situational

Awareness and Control
Not possible to avoid multiple contingency initiated blackouts
HOWEVER
PreventingBlackouts
The Probability,
Size and
Impact of
Wide Area Blackouts
can be
REDUCED !!
89
WideAreaMonitoring,Protection&Control
BridgingtheControlGap Finally!
Timeframes of Grid Management
Protection PhasorbasedWideAreaControl
SIPS / Stability Active
ControlRoomEMS/DMS/WAMS
Microgrids
SPS Control Management
EmergingMarket OngoingWAMSEMS
200-600ms 3-15 seconds

Phasor Primed Automated
16-100ms 15 minutes
0.6-3s Dispatch Operator Dispatch
Equipment
Protection Phasor Triggered Human-in-the-Loop
Thermal Limits
Frequency Stability
Oscillatory Stability
Long-Term Voltage
Stability
Short-Term Voltage Stability & Voltage Rise
N-x Transient Stability
Transient Stability
Local & Differential Fault Protection
90
SystemIntegrityProtectionScheme(SIPS)
Preventpropagationof
disturbancesforsevere
systememergencies
causedbyunplanned
operatingconditions
Lastlineofdefenseto
improvesystemsecurityandpreventdisturbancepropagation
Couldhelpbetterutilizesystemmargins
StabilizeSystemforEquipmentOutages,N2orbeyond
o Initiatepreplannedseparationofthepowersystem
o Preventoverloadingofthelines
o Arrestvoltagedecline
ExperienceswithSIPS
IEEEPSRC
Report 2009
91
SIPSDesign
Schemetypes
o Eventbased
o Parameterbased
o Responsebased
o Combinationoftheabove
Distributedvs.central
schemes
o Combined,scalable,
multilayerapproachrecommended
SCE PACI RAS Scheme
Costofdifferentmeasures
o Switchingofshunt/FACTSdevices,startofgasturbines,etc.
preferabletosystemseparationandloadshedding
RequirescoordinationofSIPSinthegrid
Example:VoltageStabilityMonitoring
Dynamic, model-based simulation tools:
Voltage Stability Assessment based on State Estimation
contingency analysis
Tracking the relative distance from voltage instability
boundary continually in real-time
o Distance to the nose of the PV curve
o State Estimation based
stability boundary
Important to validate
model correctness
Source: ABB
92
Example:VoltageStabilityMonitoring
Power - Voltage Curve
Measurement-based indicators: 245
Voltage (kV)
240
Monitor available reactive power levels 235
230
(capacitor/reactor reserves, tap-changers) 225
220
Singular Value Decomposition (SVD) 170 180 190 200
Power (MW)
210 220
Sensitivity analysis
Maximum power transfer
Distance of the load's apparent impedance |Zapp | = |ZThev |
to the Thevenin impedance: VIP*, REI, Point of collapse
RVII (Real-time Voltage Instability Thevenin Load
Indicator)
E ZThev Zapp
*K. Vu & D. Novosel, Voltage Instability Predictor (VIP)
- Method and System for Performing Adaptive Control
to Improve Voltage Stability, Patent Issued April 2001.
Thank you
Questions?
93
Session 4.2
Use of Synchrophasor Technology for
Analysis of Cascading Outages
VahidMadani PacificGas&ElectricCo.
DamirNovosel QuantaTechnology
Tutorialon
UnderstandingCascadingPhenomenon
DOE/NASPISynchrophasorRoadmap 2012
94
PMUs APPLICATION LEVELS
F, V, and Angle
SYSTEM
F, V, Angle, P&Q
AREA
F, V, Angle,
INTER AREA P-V Curve,
Imp. Loci R-X,
LINKS AGCC
F, V, Angle,
SUBSTATION AGCC, P-V,
Q-V, R-X,
POWER PLANT P-F
PMU Value Mapping

Initiative Description
Overview of PMU capabilities and breakdown into PMU functions
PMU Capabilities PMU Functions Value Proposition
Baseline Normal Phase

Fast Disturbance Replay Adaptive Switching Mitigate prevent Black
Analysis / Assessment Angle trending
Out/large incidents
State Estimation
Small signal analysis tool Fast out of step indication Detect developing
enhancement
Instabilities
Situational Awareness
Voltage stability Distributed State Estimator
Islanding detection Run Grid closer to
Assessment
Physical limits
Monitoring / alarming Visualization Dynamic
Post Event Analysis SS Linear state estimator
Oscillation Assist PSS tuning
Visualization abnormal Closed Loop Grid Control

Protection & Control Visualization line overload Support during islanding
Angle/ frequency/volt
conditions
Monitoring Monitoring Wide Area Disturbance
Angle/frequency/volt Line overload Locator Improve state estimator
solution
Alarming dynamic
Post Event Analysis Enhanced Fault Location Improve Look-ahead
oscillation
Analysis
Small Signal Oscillation Wide Area / Local Area
Detection Oscillation monitoring
190
95
New mantra: Model, Measure, Monitor, Mitigate!
Electrical
Grid
Control Room Operations

Engineering Analysis
Wide-Area situational awareness Wide Area Control
Big Data Analytics
and coordination across seams.
Create and manage robust real- Risk assessment, data mining,

time variable stability limits. baselining for grid vulnerability.
Co-ordinated control hierarchy,
Add Operational Response Analysis tools for condition centralized (via EMS/DMS), or
Guidance monitoring, model validation, decentralized as appropriate.
to Situational Awareness for control tuning, post-event
Critical Conditions analysis, compliance monitoring.
OperationalUseExample NYISO
Voltageangledifferencesacross4regions(NYISO,PJM,MISO,IMO)
Thetrafficlightsrepresentingthekeymetricelements
o LeftisinternalNYISOcontrolarea,
o Rightisexternalcontrolareas AngledifferenceundertheHealthcolumnshouldbe
equaledtozeroandlightsupifthesumofthefouranglesexceedsacertainthreshold
Violationmessage
indicatoralso
appearsonthe
EMSSCADAsystem
Source: NYISO
96
PacificGas&ElectricApplications
Situational Awareness, Visualization and Alarming
(angles and voltages; overloads and oscillations)
Voltage Stability Management
Enhanced Energy Management Systems
o Adding synchrophasor measurements to existing SE
o Tracking dynamic changes and contingency analysis
System Restoration
Post-Disturbance Event Analysis, MW Transfer
including Fault Location Margins
Controls
Weak elements (i.e. prone to

Operator and Engineering Training, voltage instability)
Dispatch Training Simulator Remedial Actions
Provide interfaces with EMS and

with third parties
PMUOperationalApplications1(2)
SituationalAwareness,Visualization,
andAlarmingforOperators
o Abnormalanglesandvoltages
o Lineoverloads&oscillations
monitoring
o Systemrestoration
EnhancedEMS
o Addingsynhronized
measurementstoexistingSE
o Trackdynamicstatechangesof
asystemduringdisturbances
o EMSinterfacesandwiththirdparties
97
PMUOperationalApplications2(2)
RealtimeVoltageInstabilityIndicator
(RVII)andContingencyAnalysisbased
onPMUavailability
LinearandDistributedState
Estimation
PostDisturbanceEventAnalysisfor
PlannersandEngineers
OperatorandEngineeringTraining,
DispatchTrainingSimulator
CognitivetaskandPerformance
Analysis
PhaseAngleMonitoringandAlarming
RelativeAngleswithrespecttocommon
reference
o Thedifferencebetweenthemeasuredvoltage
anglebyandthevoltageanglemeasuredata
referencebus
AngleDifferencesbetweenapairofnodes
o Computedasdifferenceinthevoltageangle
betweentwolocations:typicallybetweenthe
sourceandsinkareasofthesystem,or
acrossaknowncorridororinterface
AngleRateofChangetodetectsudden
disturbancesinthesystem
o Computedaschangeinvoltageangleovera
userdefinedtimeperiod(e.g.1s)
o Representsrelativelyfastchanginganglesin
time(e.g.pre andposteventanglechange
duringadisturbance)
98
VoltageStabilityAssessment
Real-time Alerts
Real-Time
Real-TimeVoltage
VoltageInstability
InstabilityIndicator
Indicator(RVII)
(RVII)
MEASUREMENT-BASED
PMUs
Equiv. Imped. (Zeq)
PMU measurements Reactive Margin (Qmargin)
Estimate equivalent
parameters in real-time
EMS & Online-DSA Applications Predict HOW to respond / Advance

Arming (accurate model).
MODEL-BASDED
CA
Predict Qmargin changes under
VSAT worst case contingency.
Provide recommendations on
corrective actions.
RAS
OfflineAnalysisUseCases
PostEvent Quickerpostmortemanalysis.
Sequenceofevents&rootcause
Analysis analysis.
Dynamic Dynamicmodelverification.
Model Generatormodelcalibration.
Synchrophasor benefits for
Loadcharacterization.
Validation Post-Event Analysis
Phasor dataarealsovaluablefor
investigationofgriddisturbances,improving
Assessdynamicperformanceofthegrid. boththespeedandqualityofanalysis.
Baselining Steadystateangularseparation.
Systemdisturbanceimpactmeasures. Inthecaseofthe2007Floridablackout,
NERCinvestigatorsusedphasor datato
createthesequenceofeventsand
determinethecauseoftheblackoutinonly
twodays;incontrast,lackinghighspeed,
Compliance Primaryfrequency(governing)response. timesynchronizeddisturbancedataittook
manyengineeryearsoflabortocompilea
Monitoring PowerSystemStabilizer(PSS)tuning
correctsequenceofeventsforthe2003
blackoutintheNortheastU.S.andOntario.
NERC RAPIR Report, 2010.
99
PostMortemForensicstoIdentifySequenceofEvents(TimeDomain)
Example:May30th,2013Event
Example:RingdownAnalysis
UsingHTLSalgorithm.
Processed@5samples/second.
Sortedbyproprietaryindex(RI).
Firstvoltageswingnotconsidered. 2
0
0
100
InterareaOscillationsModesObservedinColombia
Inter-area mode at 0.49Hz
(Colombia-Ecuador). Opposing
phase in South
Governor common-mode:
whole system oscillates in
coherent phase
KeyDeploymentSuccessFactors
Presentsynchrophasor deploymentisonly
Findingakiller
tipoftheicebergforongoingreliability
application!?
improvementsandbenefitrealization
AssureLifecycleQualityofMeasurements Requires
TOstotakeOwnershipandRealizeownBenefits
Baselining toProvideNorms:HistoricalData/Simulations
UpdatesofApplicationandDesignRoadmaps
o Systemexpandabilityasmeasurements& applicationsgrow
o Systemintegrationwithotherenterprisesystems
EngineeringandOperatorGuidelinesandTraining
DataandInformationExchangeAcrossInterconnections
101
PG&E Proof of Concept Facility
Compliance and Risk Management
In
y
log
te
Ensure
r
no
op
Performance
ch
er
a
Te
bi
What are gaps,
lit
on
y
and how to close them?
m
m
Is compliance integrated ?
Co
Verify Controls
Do we have the right metrics?
Understand the Risk

Whats the risk?
PG&E POC along with other established Whats the potential impact on safety, finances, Company
reputation, DOE, customers, shareholders?
test facilities have provided the platform for Know Compliance Requirements
gathering the knowledge to provide the Do we know what they are?
How do we demonstrate compliance?
industry with direction and a fast track How do we manage changes to requirements?
process for maturing the standards and

Production Level Technology Systematic Processes 2
Source: DECLARATION OF DR. VAHID MADANI IN SUPPORT OF PETITION FOR

DECLARATORY ORDER OF PACIFIC GAS AND ELECTRIC COMPANY SEEKING RATE
RECOVERY FOR SMART GRID PROJECT
Thank you
Questions?
102
Session5
5. IndustryExperienceintheAnalysisofCascadingOutages:
5.1 NERCStandardsApplicabletoAnalysisofCascadingOutages
Instructor: MiloradPapic
5.2 DominionVirginiaPowerCIP0141MethodologyOverview
Instructors: RyanQuintandEmanuelBernabeu
5.3 CurrentindustrypracticeforDetection,Analysis,Preventionand
MitigationofCascadingOutagesfromtheOperatingHorizon
Perspective
Instructor: Dede Subakti
5.4 CascadingAnalysisforAutomatedIROLAssessmentatISONew
England
Instructor: EugeneLitvinov
Session 5.1
NERC standards applicable to
analysis of cascading outages
Milorad Papic,
Idaho Power
mpapic@idahopower.com
103
OUTLINE
1. AnOverviewofStandards
2. WhoisNERC?
3. NERCstandardsapplicabletoCascading
4. MethodologytoAssessCascading
OutagesatIdahoPower
1.AnOverviewofStandards
ISO (InternationalOrganizationforStandardization)istheworldslargest
developerofvoluntaryInternationalStandards.
NEMA (TheNationalElectricalManufacturersAssociation)develops
codesandstandardsthataregenerallyappliedinNorthAmerica
IEC (InternationalElectrotechnical Commission)iscreatedtostandardize
electricalandelectricallyrelatedequipmentacrosstheworld.
ITU (InternationalTelecommunicationUnion)istheUnitedNation
specializesagencyforinformationandcommunicationtechnologies(ICTs).
ANSI (AmericanNationalStandardsInstitute)coordinatesU.S.standards
withinternationalstandardstoachieveuniformconformance.
IEEE (InstituteofElectricalandElectronicsEngineers)developsown
standards(i.e.IEEEStd8591987R2008,IEEEStd7622006,etc.)
NERC (NorthAmericanElectricReliabilityCorporation)developsand
enforcesmorethan150ReliabilityStandards.
104
2.WhoisNERC?
In1968,intheaftermathofthe1965Northeastblackout,NERCwas
formedasanotforprofitorganizationwhosepurposesareto
promoteandhelpensurethereliabilityofthebulkelectricsystems
inNorthAmerica.
NERCworkswitheightRegionalEntities(RE)toimprovethe
reliabilityofthebulkpowersystem.TheseREincludemembers
frominvestorownedutilities;federalpoweragencies;ruralelectric
cooperatives;state,municipal,andprovincialutilities;independent
powerproducers;andpowermarketers.
TheNorthAmericanElectricReliabilityCorporation(NERC)is
governedbyanelevenmemberindependentBoardofTrustees.
NERC'sjurisdictionincludesusers,owners,andoperatorsofthe
bulkpowersystem,whichservesmorethan334millionpeople.
2.WhoisNERC?(cont)
TheNERCkeyprograms,whichimpactmorethan1,900bulk
powersystemownersandoperators,arebasedonfourpillars
ofcontinuedsuccess:Reliability,Assurance,Learningand
RiskBasedApproach
NERConeofStrategicGoalsinnextthreeyearsistodevelop
clear,reasonable,andtechnicallysoundmandatoryreliability
standardsinatimelyandefficientmanner.Thesestandards
establishthresholdrequirementsforassuringtheBESis
planned,operated,andmaintainedinamannerthat
minimizesrisksofcascadingfailures,avoidsdamagetomajor
equipment,orlimitsinterruptionsofbulkelectricsupply.
105
2.NERCRegions
3.NERCReliabilityStandards
Electricsystemreliabilityandsecurityarebecoming
nonnegotiableissuesintermsoftheirimpactonsocietyand
theeconomiesofmostnations.
Thedevelopmentofnewreliabilitystandardsandrevisionsto
existingstandardsareongoingactivities.
NERCReliabilityStandardsarebasedoncertainreliability
principlesthatdefinethefoundationofreliabilityforNorth
Americanbulkpowersystems.
TheintentofthesetofNERCreliabilitystandardsistodeliver
anAdequateLevelofReliability.
106
(cont..)
TheBulkElectricSystem(BES)willachievean
adequatelevelofreliability(ALR)whenitisplanned
andoperatedsuchthat:
TheBESremainswithinacceptablelimits
TheBESperformsacceptablyaftercredible
contingencies
TheBESlimitsinstabilityandcascadingoutages
TheBESsfacilitiesareprotectedfromseveredamage
TheBESsintegritycanberestoredifitislost
3.PlanningandExtremeEvents
Planningeventsincludenocontingency,singlecontingencyandmultiple
contingencies(categoryPotoP7)
Extremeeventsincludelossofmultiplefacilitiessimultaneously
HighImpact,LowFrequency(HILF)Eventsmaysignificantlyaffectthe
reliabilityoftheBESandcouldcauselongterm,catastrophicdamageto
thebulkpowersystem.
CascadingEventistheuncontrolledsuccessivelossofsystemelements
triggeredbyanincidentatanylocation.Cascadingresultsinwidespread
electricserviceinterruptionthatcannotberestrainedfromsequentially
spreadingbeyondanareapredeterminedbystudies.
AdverseReliability TheimpactofaneventthatresultsinBulkElectric
SysteminstabilityorCascading.
InterconnectionReliabilityOperatingLimit(IROL):ASystemOperating
Limitthat,ifviolated,couldleadtoinstability,uncontrolledseparation,or
CascadingoutagesthatadverselyimpactthereliabilityoftheBulkElectric
System.
107
3.Severityvs.Probabilityofan
Event
Planning Protection/CIP
BaseCase
Development
Operation Emergencies/Restoration
108
3.FiveGroupsofNERCReliability
Standards
ModelingStandardsestablishconsistentmodelingdatarequirementsand
reportingproceduresfordevelopmentofcasesnecessarytosupportanalysisof
thereliabilityoftheinterconnectedtransmissionsystem.
Planningstandardsspecifytechnicalanddesigncriteriaandproceduresinthe
planninganddevelopmentoftransmissionsystems,suchasNERCTPL
(TransmissionPlanning)standards.
Operationsstandardsspecifytheoperationstoprotectthereliabilityandsecurity
ofpowersupplyandoperationundernormalandabnormaloperatingconditions,
suchasNERCTOP(TransmissionOperations)standards,NERCIRO(Interconnection
ReliabilityOperationsandCoordination)standards,andNERCVAR(Voltageand
Reactive)standards.
Protectionstandardsspecifythecoordinationandresponsibilitiesofthe
protection,suchasNERCPRC(ProtectionandControl)standards,andNERCCIP
(CriticalInfrastructureProtection)standards.
Emergencystandardsspecifytheprocedures,implementingplans,and
responsibilitiesrelatingtooperatingemergencies,suchasNERCEOP(Emergency
PreparednessandOperations)standards.
3a.BaseCaseDevelopment
MOD0100 SteadyStateDataforTransmissionSystem
ModelingandSimulation
MOD0120 DynamicsDataforTransmissionSystem
ModelingandSimulation
FAC0083 FacilityRatings
PRC0061 AutomaticUnderfrequency LoadShedding
PRC0211 UnderVoltageLoadSheddingProgramData
TPL0014 SystemPerformanceUnderNormalConditions
MOD0140 DevelopmentofSteadyStateModels
MOD0150 DevelopmentofDynamicModels
109
3b.PlanningStandards
. TPL0014 standard basic requirements:
R3.2 states that Studies shall be performed to assess the impact of the
extreme events...
R3.3 states that contingency analysis should Simulate the removal of all
elements that the Protection System and other automatic controls are
expected to disconnect for each Contingency without operator intervention,
including: Tripping of generators where simulations show generator bus
voltages or high side of the generation step up (GSU) voltages are less than
known or assumed minimum generator steady state or ride through voltage
limitations. Tripping of Transmission elements where relay loadability limits
are exceeded
R3.5 Identify the planning and extreme events in Table 1 which are expected
to produce more severe system impacts, and evaluate their consequences
including Cascading . Conduct a cascading evaluation to develop possible
actions to reduce the Cascading likelihood or mitigate the consequences
and adverse impacts.
3b.PlanningStandards
R4.5. Those extreme events in Table 1 that are expected to produce more severe System
impacts shall be identified and a list created of those events to be evaluated. The
rationale for those Contingencies selected for evaluation shall be available as supporting
information. If the analysis concludes there is Cascading caused by the occurrence of
extreme events, an evaluation of possible actions designed to reduce the likelihood or
mitigate the consequences of the event(s) shall be conducted.
R5. Each Transmission Planner and Planning Coordinator shall have criteria for
acceptable System steady state voltage limits, postContingency voltage deviations, and
the transient voltage response for its System. For transient voltage response, the criteria
shall at a minimum, specify a low voltage level and a maximum length of time that
transient voltages may remain below that level.
R6. Each Transmission Planner and Planning Coordinator shall define and document,
within their Planning Assessment, the criteria or methodology used in the analysis to
identify System instability for conditions such as Cascading, voltage instability, or
uncontrolled islanding.
110
3c.OperationStandards
TOP0012 Transmission Operations standard purpose is to
prevent instability, uncontrolled separation, or cascading
outages that adversely impact the reliability of the
Interconnection by ensuring prompt action to prevent or
mitigate such occurrences.
TOP0042 TransmissionOperationsstandardensuresthat
thetransmissionsystemisoperatedsothatinstability,
uncontrolledseparation,orcascadingoutageswillnotoccur
asaresultofthemostseveresingleContingency.R3.Each
TransmissionOperatorshalloperatetoprotectagainst
instability,uncontrolledseparation,orcascadingoutages
resultingfrommultipleoutages,asspecifiedbyitsReliability
Coordinator.
TOP0081 ResponsetoTransmissionLimitViolations
standardensuresthatTransmissionOperatorstakeactionsto
mitigateSOLandIROLviolations.
R1.TheTransmissionOperatorexperiencingorcontributingto
anIROLorSOLviolationshalltakeimmediatestepstorelieve
thecondition,whichmayincludesheddingfirmload.
R2.EachTransmissionOperatorshalloperatetopreventthe
likelihoodthatadisturbance,action,orinactionwillresultinan
IROLorSOLviolationinitsareaoranotherareaofthe
Interconnection.
R3.TheTransmissionOperatorshalldisconnecttheaffected
facilityiftheoverloadonatransmissionfacilityorabnormal
voltageorreactiveconditionpersistsandequipmentis
endangered.
111
IRO0082 Reliability Coordinator Operational Analyses and Realtime
Assessments prevents instability, uncontrolled separation, or cascading outages
that adversely impact the reliability of the interconnection by ensuring that the
Bulk Electric System is assessed during the operations horizon.
IRO0091 Reliability Coordinator Actions to Operate within IROLs standard
prevents instability, uncontrolled separation, or cascading outages that adversely
impact the reliability of the interconnection by ensuring prompt action to prevent
or mitigate instances of exceeding Interconnection Reliability Operating Limits
(IROLs).
IRO0102 Reliability Coordinator Data Specification and Collection prevents
instability, uncontrolled separation, or cascading outages that adversely impact
the reliability of the interconnection by ensuring the Reliability Coordinator has
the data it needs to monitor and assess the operation of its Reliability Coordinator
R1. The Reliability Coordinator shall have a documented specification for data and
information to build and maintain models to support Realtime monitoring,
Operational Planning Analyses, and Realtime Assessments of its Reliability
Coordinator Area to prevent instability, uncontrolled separation, and cascading
outages.
IRO017
3d.Protection/CIPStandards
PRC0232 Transmission Relay Loadability purpose is that
Protective relay settings shall not limit transmission loadability
CIP002 CIP011 Critical Infrastructure Protection (CIP)
Cyber Security Standards specifically pertain to the
identification and protection of critical cyber assets that
support the reliability of power systems .
CIP0141PurposeistoidentifyandprotectTransmission
stationsandTransmissionsubstations,andtheirassociated
primarycontrolcenters,thatifrenderedinoperableor
damagedasaresultofaphysicalattackcouldresultin
widespreadinstability,uncontrolledseparation,orCascading
withinanInterconnection.
112
3e.EmergencyStandards
EOP0023.1 Capacity and Energy Emergencies
R3.4.2 The Reliability Coordinator shall use its best efforts to ensure that
revising SOLs or IROLs would not result in any cascading failures within the
Interconnection.
EOP0032 Load Shedding Plans
R1. After taking all other remedial steps, a Transmission Operator or
Balancing Authority operating with insufficient generation or transmission
capacity shall shed customer load rather than risk an uncontrolled failure
of components or cascading outages of the Interconnection.
EOP0052 SystemRestorationfromBlackstart Resources
EOP0062 SystemRestorationCoordination
EOP0111 EmergencyOperationsisanewstandardthatconsolidates
requirementsfromthreeexistingEmergencyOperationsstandards:EOP
0012.1b,EOP0023.1andEOP0032
4.GenericScenarioofaCascading
Event
0 SystemStateBeforeCascading
Event
1 ContingencyConditions
2 TriggeringEvents
3 PowerFlowSurges,Voltage
problems,Overloads
4 ProtectionSystemTripsLines,
Transformers,Generators
(InsecureSystemState)
5 SystemSeparation,Instabilityand
VoltageCollapse
6 BlackoutEvent
113
4.IdahoPowerMethodologyto
AssessCascadingOutages
4.WECCTransientVoltage
PerformanceCriteria(adopted1998)
114
Questions
?
Milorad Papic
Tel: (208) 388-2343
Email: mpapic@idahopower.com
Session5.2
DominionVirginiaPower
CIP0141MethodologyOverview
RyanD.Quint,PhD
ElectricTransmissionPlanning
DominionVirginiaPower
KyleThomas EmmanuelBernabeu,PhD
OperationsEngineering AppliedTechnologies
DominionVirginiaPower PJMInterconnection
115
DominionProfile
PrimaryOperatingSegmentsOverview
Dominion Virginia Power Dominion Energy Dominion Generation
Electric Transmission Gas Transmission Utility Generation

6,400 miles of transmission lines Together with Gas Distribution, 19,600 MW of capacity
operates one of the largest natural gas Balanced, diverse fuel mix
Favorable regulatory environment storage system in the U.S.
Favorable regulatory environment
10,900 miles of pipeline in six states
Electric Distribution
57,000 miles of distribution lines
Well positioned in Marcellus and Utica Merchant Generation
Shale regions 4,000 MW of capacity, including
2.5 million franchise retail nuclear, gas and renewable power
customer accounts in VA and NC Gas Distribution
Active hedging program for energy
21,900 miles of distribution pipeline
revenue/margins
and 1.3 million franchise retail natural
gas customer accounts in OH & WV Dominion Retail
Retail Gas & Products/Services
Blue Racer Joint Venture
1.5 million non-regulated customer
Utica Shale midstream services
accounts in 10 states*
*Excludeselectricbusinessbeingsold.
DominionEnergyProfile
LeadingproviderofenergyandenergyservicesintheMidwest,
NortheastandMidAtlanticregionsoftheU.S.
~27,600MWofelectricgeneration
6,000+milesofelectrictransmission
12,000milesofnaturalgastransmission,
gatheringandstoragepipeline
942billioncubicfeetofnaturalgasstorage
operated
CovePointLNGFacility
2.4millionelectriccustomersinVAandNC
1.3millionnaturalgascustomersinOH&WV
2.1 million non-regulated retail customers in 13 states
116
DominionVirginiaPowerFootprint
TheImportanceofCascadingAnalysis
Critical Interconnections:
Nuclear Generation Facilities
Cove Point Natural Gas Facility
Key government agencies
Government contractors
Quantico Marine Corps Base
Norfolk area largest Navy base in world
Data centers 50% of internet traffic
2.5M retail customer accounts in VA and NC
117
CascadingMethodology
CIP0141RequirementR1Implementation
Introduction
PreCIP014Objective: CIP014Objective:
Identifycriticalsubstations. instability,uncontrolled
Determinedfromthreemetrics: separation,orcascading.
Loadflowviolations Cascadingalgorithm.
Generatorresponse
Busvoltageangle
Physical
Relativeranking. Attack
Stable Collapse
NO Cascade Cascade
Bounded Collapse
118
Cascading:LogicOverview
P c1 | N k1
Corrective Actions
P c2 | c1
P c P c1 | N k1 P c2 | c1 ... P ck | ck 1
Implementation
MostsoftwaretoolsdoNOTincludetunablecascadinganalysis
Transientandvoltagestabilityanalysisormetrics
Probabilistictripping&cascading
Modifiablesolutionengine
RemedialActions
PythonwrapperalgorithmdevelopedaroundPSS/eengine
PythonWrapperAlgorithm
Tripping Probabilities HardSolution
Corrective Actions CorrectiveActions
TrippingProbabilities
Contingency Definitions CascadingPaths
Cascading Metric Definitions
PowerFlowEngine
ConventionalSoftware
119
HardSolution
BlownUpsolutioneither:
Lackofoperatingstate
Mathematicalissue
Example:
25outof+200casesfailedtoconverge
Definition:HardSolution
Gentlytransitionbetweenoperatingstatestoavoid
mathematicalinstability
Allcasesconverged
Hard Solution parameters are all tunable.
CorrectiveActions
AfterinitialNk1event:
Identifyviolations
LinearProgramming
Generationredispatch
Loadshedding
Switchshuntelements
IsCAfeasible?Limit:
Nonconsequentialloadloss
Generationredispatch
Frequencydecay
Corrective Actions Feasibility Limits are all tunable.
120
Cascading:LikelihoodofTripping
Line
Thermal
Tripping Probability
TX 1
Probabilities CascadingLogic
Thermal
Relay Index=1
Protection
Transformer Thermal Loading
Tripping Probability
1
Index=1
Relay Protection
The Probability curves are all tunable.
CascadingPaths
Definitions:
Cascadingevent:aparticulardisconnectionofacircuitelement.
Cascadingpath:aparticularsequenceofcascadingevents.
Cascadingdepth:numberofstepsintheCascadingpath.
Chainreaction:numberofpossiblecascadingevents.
Chain reaction = 3
P(C3) Cascading Path
P(C1) P(C4) Cascading

P(C6) Path
NK1
P(C5) Probability
P(C2) P(C7)
Depth 1 Depth 2 Depth 3
The Cascading path parameters are all tunable.
121
PossibleCIP014CascadingOutcomes
Collapse: Immediatesystem
collapseafterNk1 Physical
Attack
Cascadingpossibilities: N-k1
NoCascading
BoundedCascade:aftera
certainnumberofcascading Solution Found Collapse
events,itstopspropagating.
UnboundedCascade:
cascadingeventsresultin No Cascading Cascade
systemcollapse
Bounded Collapse
CascadingPaths
OriginalIdea:Removehighestprobabilityelementaftereachcascading
event
1%
Cascading
45% 30%
Path
60%
NK1 Probability=
50% 13.5%
40% 40%
Depth 1 Depth 2 Depth 3
Question:Whatabouttheotherpossiblepathsthatcouldbetaken?
Whatisthedifferencebetween45%and40%?Isthisthetruemaximum
likelihoodpath?
Answeredthesequestionsbyfindingallpossiblepaths,creatingnew
CascadingTrees
122
CascadingTrees:Introduction
Definitions:
Cascadingevent:aparticulardisconnectionofacircuitelement
Cascadingpath: aparticularsequenceofcascadingevents
CascadingTree: alternativecascadingpaths
Node:distinctsystemstate Cascading Path
TerminalNode:cascadingoutcome
N-K1
N3
Cascading Tree
N7
Cascading Event
P(Ci)
N4 Terminal Nodes
N1 N8
N0 N5
N2
N6
BuildingCascadingTrees
Algorithm:
Iterativegrowthalongthetreenodes
Trackingdistinctsystemstates(nodes)
Efficiency&datastorage(parallelprocessing)
CascadingOutcomes(stoppingrules):
Blownup Similar
LargeMismatch
Bounded
Lowlikelihoodpath<0.001%
Maximumnumberofnodesexceeded
123
PreliminaryExampleTrees
Terminal nodes bounded
Node size = cascading likelihood
Overall P(cascading) between nodes
Blown up and large mismatch cases
*LineThermalOverloadOnly
PreliminaryExamples:CascadingTrees
CascadingTree:
Definedcascadingpropagationdirection
Allpossibleoutcomes
124
CascadingAnalysis Dynamics
Dynamicscaneitherbe:
Builtintocascadingalgorithmdirectly
Usedasverificationforhighprobabilitycascadingcases
Considerations
Nk1tripping faulttype
Relaying&nearandfarendtriptimes
DistanceZ1/Z2,POTT,DCB,etc.
OutofStep&ReversePowerProtectionandTripping
BusVoltageProfile
LineZEncroachment
Outofstepprotectionforislandedgeneration
CascadingAnalysis Dynamics
Swinging Voltage V < 0.85 pu

Due to Generator
Instability
Swinging Out-of-step
Generators Generators
125
OngoingWork
TransformerThermalandProtectionprobabilityfunctions
CascadingTreeMetrics
DynamicSimulations
Maxlikelihoodpaths
Pathsneargeneratingresources
VoltageStabilityIndices
Maxlikelihoodpaths
Pathsnearvoltagesensitiveregions(e.g.loadcenters)
SummaryofMethodology
1. Pythonwrapper functionaroundloadflowsolutions
2. Intelligenthardsolutiontomitigatenumerical
instability issues
3. DeterministicNk1 outage
4. Probabilisticcascadingmethodologyusingindices
5. Cascadingpaths/trees todetermineallpossibilitiesof
cascadingandtheirassociatedprobabilityof
occurrence
6. Assessingvoltagestabilityusingmetrics
7. Enablingariskinformeddecisionprocessofcritical
stationsbasedoninstability,uncontrolledseparation,
andcascadingevents
126
ResiliencyandSecurityMeasures
Risk=Probability*Consequence
Reducingriskbyminimizingthelikelihoodofevents
DominionResiliencyStrategy
Thethreatshave
fundamentally
changed
NewNERC
Standardsrequire
ustomakechanges
Dominionisleading
theindustry
Vitalfornational
security
127
EnhancedSecurityFacilities
PerimeterBarriers
o Anticut
o Anticlimb
o Antiram
BallisticProtection
AccessControl
Lighting
Increasedelectronicsurveillance
SecurityMonitoring&Response
CorporateSecurityControl
Center
24x7Monitoringand
Management
Alarmmanagement,
assessmentandresponse
Cyberpatrols
RapidResponse
Interfacewithcontrolcenterand
regionaloperationscenters
InterfacewithCorporateSecurity
Response
InterfacewithStateandLocal
GovernmentAgencies
128
FuturePlannedPMU ThankYou!
Installationstobe
incorporatedinto
systemdeveloped
underthisgrant.
R.D.Quint
257 ryan.d.quint@dom.com
Session5.3
CurrentIndustryPracticeforAnalysis,
PreventionandMitigationof
CascadingOutagesfromtheOperating
HorizonPerspective
Dede Subakti,CAISO
UnderstandingCascadingPhenomenon:
MethodologiesandIndustryPracticeforAnalysisof
CascadingFailures
129
QuickRecapfromNERCStandardRequirement
TOP0022
R6.EachBalancingAuthorityandTransmissionOperatorshallplanto
meetunscheduledchangesinsystemconfigurationandgeneration
dispatch(ataminimumN1Contingencyplanning)inaccordancewith
NERC,RegionalReliabilityOrganization,subregional,andlocalreliability
requirements
TOP0042
R2.EachTransmissionOperatorshalloperatesothatinstability,
uncontrolledseparation,orcascadingoutageswillnotoccurasaresultof
themostseveresinglecontingency.
N-1/N-2 Secure
EnsuringN1/N2SecurityinOperatingHorizon
Inthepast,therearethreewaystomodelcongestion
Method1:UtilizationofPath/Interface
Thiscanbeusedtoensuresystemreliabilityforboththermaland
stabilitylimitation
Method2:UtilizationofLinearizedDistributionFactor
Thisiscommonlyusedtoensuresystemreliabilityforthermal
limitation
Method3:UtilizationofNetworkApplication/RealTimetool
Thiscanbeusedtoensuresystemreliabilityforboththermaland
stabilitylimitation
130
Example SimplifiedSystemDiagram
Un 1 Line A
Un 2
Line B
Un 3
Line C
NormalRating EmergencyRating
LineA 100 150
Line B 100 150
Method1:UtilizationofPath/Interface
Normal Emergency
LineA 100 150
Line B 100 150

Utilizing Proxy Path/Interface Limit:
Line A + Line B + Line C < Import Limit
Off line Transfer Analysis is performed utilizing powerflow cases
Line A
Line B
Line C
Challenges:
Its a proxy limit calculated based on large number of assumption
The limit is the minimum of thermal, voltage, stability limit
(Pacific Southwest blackout in 2011)
131
Normal Emergency
LineA 100 150
Line B 100 150
Utilizing LODF
Line A + (Distribution Factor * Line B) < Rating of Line A
Line A
Line B
Line C
Challenges:
LODF is calculated off line and it is based on system topology.
LODF changes when the system topology changes
(Northeast blackout of 2003)
ImplementationofNomogramisMathematicallythesame
Line A + (Distribution Factor * Line B) < Rating of Line A
Nomogoram
1600
1400
1200
N-1 Insecure Operating Region
Flow in Line A
1000
800
600
400 SafeOperating
Region
200
0
0 500 1000 1500 2000 2500
Flow in Line B
132
Method 2 can be expanded if there are Remedial Action Scheme (RAS) that arm
Generation and/or Load to trip to relief overload
The equation would be expanded

Line A + (Distribution Factor * Line B)
GSF * Armed Gen Drop
LSF*Armed Load Drop < Rating of Line A
GSF = Generation Shift Factor

LSF = Load Shift Factor
Method2expandedforRAS
Line A + (Distribution Factor * Line B) GSF*ArmedGen
LSF*ArmedGen<RatingofLineA
TheGSFandLSFshiftsthesafeoperatingregion
Barre-Lewis_NG
1600
RAS expand
1400
the safe operating
1200 Region
Flow in Line A
1000
800
600
400 SafeOperating
Region
200
0
0 500 1000 1500 2000 2500
Flow in Line B
133
Normal Emergency
LineA 100 150
Line B 100 150

Line A
Line B
Line C
For Steady State Analysis = Real Time Contingency Analysis (RTCA)

Tool automatically do calculation with the input:
What to monitor ( Line A)
What to simulate (Contingency of Line B)
What is the rating (Rating of Line A)
RTCA does not calculate Transfer Capability. It is not uncommon to set
RTCA to alarm at 90% so that operator knows how much margin they
have
Normal Emergency
LineA 100 150
Line B 100 150

Line A
Line B
Line C
For Voltage Stability = Real Time Voltage Stability Analysis

Tool automatically perform PV or QV analysis utilizing State Estimator
solution
Need to define Transfer Scenarios
Need to define Reliability Criteria to be monitored
Need to define contingency to be simulated
134
Normal Emergency
LineA 100 150
Line B 100 150

Line A
Line B
Line C
For Transient Stability = Real Time Transient Stability Analysis

Tool automatically perform time domain transient analysis utilizing State
Estimator solution
Need to define Transfer Scenarios
Need to define Reliability Criteria to be monitored
Need to define contingency/fault switching to be simulated
Need dynamic data representation
EnsuringtooperatewithinSOL
(fromNERCwhitepaper)
Itisimportanttodistinguishoperatingpracticesandstrategiesfromthe
SOLitself.Asstatedearlier,theSOLisbasedontheactualsetofFacility
Ratings,voltagelimits,orStabilitylimitsthataretobemonitoredforthe
pre andpostContingencystate.Howanentityremainswithinthese
SOLscanvarydependingontheplanningstrategies,operatingpractices,
andmechanismsemployedbythatentity.Forexample,oneTransmission
Operatormayutilizelineoutagedistributionfactorsorothersimilar
calculationsasamechanismtoensureSOLsarenotexceeded,while
anothermayutilizeadvancednetworkapplicationstoachievethesame
reliabilityobjective
Thiswhitepaperdiscusshoweachmethodcouldbeusedtoachievethe
ultimategoalofreliability,thatis:acceptablesystemperformanceforpre
andpost contingencystate.
135
SOLExceedance(fromNERCwhitepaper)
SOLexceedanceoccurswhenacceptablesystemperformanceas
describedinapprovedFAC0112isnotoccurringinRealtimeoperations
asdeterminedbyRealtimeAssessments. Inotherwords,unacceptable
systemperformanceasindicatedbyRealtimeAssessmentsequatesto
SOLexceedance.AnSOLisexceededwhenanyofthefollowingoccuror
areobservedaspartofaRealtimeAssessment:
ActualflowonaFacilityisabovetheFacilityRatingforanunacceptabletime
duration
CalculatedPostContingencyflowonaFacilityisabovethehighestavailable
FacilityRating
ActualbusvoltageisoutsideacceptablepreContingency(normal)busvoltage
limits
PostContingencybusvoltageisoutsideacceptablepostContingency
(emergency)busvoltagelimits
DefinedtransientorvoltageStabilitylimitsareexceeded(techniquesfor
determiningandobservingStabilitylimitscanvary)
Comparingthethreemethods
Normal Emergency
LineA 100 150
Line B 100 150 Line A
Line B
Line C
Method Accuracy Setup/ Monitoringfor

Study Operator
Method1:Path/Interface Less Harder Very Accustomed

Method2:LODF/Nomogram
withmultipleinput
Method3:Advanced
Applications More Easier Fairly New
136
OperationalExperiencesforCascadingIdentification
PlanningcriteriavsOperatingcriteria:
Forexample:Thefollowingareplanningcriteriafortransientdynamic
simulation
InOperatingHorizon:
Howapplicablearecriteriasuchastransientvoltagedropandtransient
frequencydropforalocalarea?
Theremaybeaneedtoidentifyandclassifywidespreadsystemarea
limitationversuslocalarealimitation
Identifyingriskinrealtime(i.e howwidespreadtheimpactis)isstilla
significantchallengewithcurrenttechnology
OperationalExperiencesforCascadingIdentification
Whiletoolsandvisualizationshowifalimithasbeenexceeded(orhowmuch
marginoperatorshave),itisalsoimportanttoidentifyhowbigtherisksareoncea
limitisexceeded.
Doesexceedingalimitmeanthatifacontingencyoccurs,weareriskinglossof
1,000MWofload;or,ifacontingencyoccur,weareriskingreducedequipment
lifeexpectancy
Understandingtheriskwillhelpoperatortomakedecision(i.e.doestherisk
warrantdrasticmeasures/actionsbytheoperators)
CurrentproxyofmeasuringpotentialcascadinginOperatingHorizon:
Postcontingencythermaloverloadabove125%ofhighestemergencyrating.
RealTimeContingencyAnalysisshowingdivergence(ornonsolution)
RealTimeVoltageStabilityAnalysisshowingoperatingpointbeyondthenose
curve
137
Session5.4
CascadingAnalysisfor
AutomatedIROLAssessmentat
ISONewEngland
EUGENELITVINOV
ISONewEngland
Outline
NERCdefinitionofCascading,SOLandIROL
NERCreportingrequirementsonIROLviolations
MotivationofISONEseffortsonautomatedandaccurate
IROLviolationsassessment
CurrentISONEsprocessinIROLanalysisandreporting
PotentialCascadingModetool
AutomatedonlineandofflineIROLassessment
BenefitsofautomatedIROLassessment
Conclusions
138
NERCDefinition
Cascading
Theuncontrolledsuccessivelossofsystemelementstriggeredbyan
incidentatanylocation.Cascadingresultsinwidespreadelectricservice
interruptionthatcannotberestrainedfromsequentiallyspreadingbeyond
anareapredeterminedbystudies.
SystemOperatingLimit(SOL)
Thevalue(suchasMW,MVar,Amperes,FrequencyorVolts)thatsatisfies
themostlimitingoftheprescribedoperatingcriteriaforaspecified
systemconfigurationtoensureoperationwithinacceptablereliability
criteria.SOLsarebaseduponcertainoperatingcriteria.Theseinclude,but
arenotlimitedto:
FacilityRatings(Applicablepre andpostContingencyequipmentorfacilityratings)
TransientStabilityRatings(Applicablepre andpostContingencyStabilityLimits)
VoltageStabilityRatings(Applicablepre andpostContingencyVoltageStability)
SystemVoltageLimits(Applicablepre andpostContingencyVoltageLimits)
NERCDefinition
InterconnectionReliabilityOperatingLimit(IROL)
ASystemOperatingLimitthat,ifviolated,couldleadto
instability,uncontrolledseparation,orCascadingoutagesthat
adverselyimpactthereliabilityoftheBulkElectricSystem
InterconnectionReliabilityOperatingLimitTv
ThemaximumtimethatanIROLcanbeviolatedbeforethe
risktotheinterconnectionorotherReliabilityCoordinator
Area(s)becomesgreaterthanacceptable.EachIROLsTv shall
belessthanorequalto30minutes
Isconsequenceof Yes IROLviolation Yes ReportableIROL
systemeventmore IsTv >30min?
UpdateTv violation
severethancriteria?
139
NERCReportingSOLandIROLViolations
StandardTOP007
R2:FollowingaContingencyorothereventthatresultsinanIROL
violation,theTransmissionOperatorshallreturnitstransmissionsystem
towithinIROLassoonaspossible,butnotlongerthan30minutes
ComplianceMonitoring:
TheReliabilityCoordinator(RC)shallreportanyIROLviolationexceeding
30minutestotheRegionalReliabilityOrganizationandNERCwithin72
hours.TheRCshallreportanySOLviolationthathasbecomeanIROL
violationbecauseofchangedsystemconditions;i.e.exceedingthelimit
willrequireactiontoprevent:
Systeminstability.
Unacceptablesystemdynamicresponseorequipmenttripping.
Voltagelevelsinviolationofapplicableemergencylimits.
Loadingsontransmissionfacilitiesinviolationofapplicableemergencylimits.
Unacceptablelossofloadbasedonregionaland/orNERCcriteria.
MotivationofISONEsAutomated
IROLAssessment
IROLcostsofTransmissionOwnercompliance
Improvetheefficiency,accuracyandconsistencyofIROL
assessmentbothofflineandonline
ReducethenumberofmonitoredIROLinterfacesandthe
numberofIROLviolationscausedbyconservativecriteria
NERCdefinitionofreportableIROLviolationisverygeneric
andallowswiderangeofsubjectiveassumptionsinpractical
implementation
Forexample,howmuchMWoftrippedload/generationtobeconsideredas
IROLviolation?
140
Motivation:IROLCosts Transmission
OwnerCompliance(Contd)
EachIROLinterfacehasmultiplepotentialcostimpactsoneach
transmissionprovider
NERCCIP,FAC,physicalsecuritystandards
EachIROLinterfaceremovedisasignificantsavingtotheNewEngland
transmissionowners
Example:oneIROLinterfacewiththreelines(30totallinemiles),six
substationsandthreeseparatelimitingcontingentelements(sixmore
stations)
CIPcompliancecost@appr.$250Kperstation(12stations)=$3M
Physicalsecuritycost@appr.$2Mperstation(1to12stationexposure)=$2M
$24Mpotentialcost
Vegetationmanagementcost@appr.$300Kmoreevery4years(morefrequently)
Exampletotalcost@appr.$3Mto$27MperIROLinterface,withaddedongoing
costs(i.e.ongoingcompliance)
ISONEsIROLMonitoringandReporting
AnySystemOperatingLimit(SOL)thathasbeenidentifiedbytheISOtohavean
adverseimpactbeyond theNewEnglandTransmissionSystemisconsideredtobe
aninterconnectionreliabilityoperatinglimit(IROL)
ISONEsOP19listsreliabilitycriteriaandexamplesofunacceptablesystem
performancethatmayresultinanIROL
TodayOperationsmonitorthirtythreepredefinedIROLinterfacesinrealtime
throughInterfaceLimitCalculator(ILC)
IfanIROLviolationhasoccurredformorethan30 minutes,theISOisresponsible
toreporttheeventtoNERCandNPCCwithin48hours
Areportingcouldbeunnecessaryifdetailedanalysisindicatesthattheeventdoesnot
constituteIROL
ViolationofIROLinterfacelimitsasacriteriaisahighlevelapproximation;
engineersperformdetailedtechnicalanalysisofIROLeventsonaquarterlybasis
toverifythecascadingconsequence
141
CurrentQuarterlyEngineeringAnalysisofIROL
Violations(thermalandvoltage)
OnlyconsidersN1thermalviolations
Process:
IdentifytheperiodwhenNOSOinterfaceflowexceedsN1thermallimit
LookintoControlRoomOpralog Eventtodeterminethecriticalcontingencyandlimiting
element
RequestEMSOnCalltorestoreRTNETcasefortheidentifiedperiod
UseEMSStudyPowerFlowforcascadinganalysisbymanuallyapplyingthecritical
contingency,solvingpowerflow,checkingtransmissionloadingandtrippingthe
transmissionelementiftheMVAflowisoverSTEratingbycertainamount
Solvepowerflowagainandrepeattheprocessuntil:
Voltagecollapse,or
Thereisnomoreoverloading
ConfirmIROLviolationifvoltagecollapsesoruncontrolledcascadingfailures
resultinginalossofmorethan1200MWofloadorgeneration
CurrentQuarterlyEngineeringAnalysisofIROL
Violations(transientstability)
Process:
IdentifytheperiodinwhichthestabilityinterfaceMWflowexceedsthe
stabilitylimit
Checkthetopologyassociatedwiththespecificinterfacetransferlimitand
identifythelimitingcontingencieslistedintheoperatingguides
AdjustthePSS/Eoperationsplanningbasecase(e.g.,unitstatus,plant
outputs,loads,linestatus,etc.)sothattheMWflowsovertheviolated
interfaceandadjacentinterfacesarematchedtotheEMSRTNETasclosely
aspossible
Simulatethelimitingcontingenciesontheabovesimulationcaseand
evaluatethesystemstabilitywithrespecttothetransientvoltageand
angularstabilitycriteria
142
MajorIssuesofCurrentIROLViolation
Assessment
Onlineprocessuseshighlevelapproximatedconservativecriteria
(violationofinterfacelimits)whichcouldreportfalseIROLviolations
Quarterlyofflineprocess,mostlymanualandtimeconsuming
Theanalysismaypossiblyproduceinaccurateresultsbecausetheexact
realtimeoperatingconditionscanhardlybereplicatedbytheoffline
model
Currentpracticereliesonengineeringjudgmentandthereforemay
produceinconsistentstudyresultsbydifferentengineers
Needaplatformwithadvancedtoolsandwelldefinedauditableprocesses
whichenabletheISOtoevaluateIROLviolationsautomatically,accurately
andconsistently
PotentialCascadingMode(PCM)tool
PCMisacustomizedtoolfortheISONEbyV&REnergy
SteadystateanalysisoffastdevelopingcascadingeventswhenOperator
hasnotimetoreact
InputdataiscompatiblewithRegionOfStabilityExistence(ROSE)tooland
usesnodebreakerEMSmodel
StateEstimation(SE) EMSpowerflowcases
Contingencies Contingencies
On-line Off-Line
Settingsfor
OnlinePCM OfflinePCM
Cascadinganalysis
IIS
Results Archived Results
Webbasedviewer Reports SEcases Reports
143
ProcessofPCMAnalysis
Power flow case
Initiating contingencies
(N-1 and N-2)
Thermaloverload TripLines
Lines0N0 0N0
Tier0
Applycontingency i TripLoads
Lowvoltages
Loads0K0 0K0
Thermaloverload TripLines
Tier1 Lines1N1 1N1
Solvepowerflow
Lowvoltages TripLoads
Loads1K1 1K1
Takenextcontingency ContinueTierstill
i=i+1 cascadestops
Report
Violationof TotalMWof
Voltage
Stability Islanding trippedload
collapse
interface andgeneration
PCMSettings
Adjustablelineandtransformer
trippingthresholds
Adjustablevoltagemagnitudeand%
thresholdsforloadtripping
Adjustablevoltagethresholdsfor
trippingofGeneration
Sensitivityanalysis:automaticallyrun
severalscenarioswithdifferentPCM
settingsforcascadetriggers
Scenario1 Medium
Report1
Linetrippingcriteria:LineFlow >1.05*STE probability
Loadtrippingcriteria:Vmin <0.87p.u. cascade
Scenario2 High
Report2
Linetrippingcriteria:LineFlow >1.10*STE probability
Loadtrippingcriteria:Vmin <0.85p.u. cascade
144
IdentificationofCriticalCascadeinPCM
Mainidea:classificationofCascadeasCriticalistheidentificationofIROL
violationbasedonmeasurableconsequences
ExistenceofaCriticalCascadelongerthan30minutesmeansIROL
violationreportableevent
ConceptofCriticalCascadeisaconsistent,quantifiableandauditable
processofIROLviolationanalysis
ConceptofCriticalCascadeisapracticalinstrumenttosatisfygeneric
NERCrequirementsofIROLcompliance
AttributesofCriticalCascade
CriticalCascadesatisfiesatleastoneofthefollowingcriteria
Systemwidevoltagecollapseoccursuponapplyinginitiatingcontingencyorasthe
resultofcascadingtripping
IslandingofthesystemandtotalMWofloadinseparatedislandsisgreaterthanpre
definedthreshold
ActualinterfaceMWflowduringcascadeexceedsstabilityinterfacelimitbypre
defined%level
TotalMWlossofloadexceedspredefinedthreshold
TotalMWlossofgenerationexceedspredefinedthreshold
CascadepropagatesbeyondBalancingAreafootprint
EachcriterionforaCriticalCascadeisconfigurableinPCMtool
Enable/Disable
Thresholdvalue
145
ScenariosinCascadinganalysis
Cascadingstudyisdeterministicperdefinedtrippingcriteria
Trippingcriteriacanbedefinedonlyapproximately(duetolackof
informationonrelaysettings,loadcomposition,operatoractions)
RiskofcascadingcanbeevaluatedbyrunningseveralcascadingScenarios
forthesameinitiatingcontingencieswithdifferenttrippingcriteria
TrippingcriteriaforScenarios
Scenario Line Transformer Loadvoltage Load
%ofrateC %ofrateC p.u. %tripped
HighProbability 130% 130% 0.8 30%
MediumProbability 115% 120% 0.85 40%
LowProbability 105% 110% 0.9 50%
SomesettinginROSEPCMsoftware
Criteriaforidentificationof
CriticalCascade
CriteriaforinclusionofPCMs
intosummaryreport
146
OnlineROSEPCMGUI
Highlevelresultsofthelatestrun ColorcodedseverityofPCM
results;historicalview
DetailreportforaScenariodisplayedin
SummaryReportwindow
SummaryReportfora
selectedScenarioandTime
UnderstandingofGUI
SelectionofRealTime Filteringoptionsfordisplayingresults
orHistoricalview
Selectionof
historical
intervaland
scaleforview
Colorcodedresultsofsingleonline
cascadinganalysisrun,aboutevery3min
NocriticalPCM;noPCMisreportedinSummaryreport
NocriticalPCM;atleastonePCMisreportedinSummaryreport
CriticalPCM;atleastcriticalPCMisreportedinSummaryreport
147
Understandingofresults
Lowriskcriticalcascade.CriticalPCMisdetectedonly
inLowProbability scenario
Understandingofresults,cont.
Highriskcriticalcascade.CriticalPCMisdetectedinallthree
scenarios
Initiatingcontingencyshouldbemitigatedtoreducetheriskof
blackout
148
Understandingofresults,cont.
MorelinesandloadweretrippedinLowProbability scenarioin
Tier0comparingtootherscenarios
Thoseextratrippingwereservingasremedialactions
preventingtodevelopcascadeintocriticalone
OnlineTSATatISONE
System components
oData Preparation Tool (DPT)
interfaces with EMS
merges external equivalent model
prepares dynamic models
consistent with planning case
oDSA Manager
scenario setup
job scheduler
results display
oTSAT Servers
execute TSAT processes (multi-
core support available)
Functionality
Offlinestudy(simulationmodelcorrespondingtohistoricalevents)
Modelvalidation:comparesimulationresultsagainstPMUmeasurements
Posteventanalysis:investigatethecauseandwhatifanalysisofhistoricalevents,for
exampleIROLviolations
Onlinestudy(simulationmodelupdatedcontinuously)
Transientstabilityanalysis&remedialactioncomputation
Automatedcomputationofinterfacelimitsbasedontransientstability
149
AutomatedIROLViolationAssessment:Stage1
Automate existing quarterly IROL assessment process
Off-line PCM tool for thermal and voltage IROL analysis
Off-line TSAT tool and real-time model for stability IROL analysis
PIquerytoidentify
periodsofIROLviolations
StateEstimationcases
StateEstimationcases
StateEstimation
archivedbyROSE
archivedbyROSE
archivedcases
OfflineTSAT OfflinePCM
ResultsofIROL ResultsofIROLanalysis:cascading
analysisfor studybasedonthermaland
transientstability voltageconditions
AutomatedIROLViolationAssessment:Stage2
Deploy on-line IROL assessment
On-line PCM module to predict the exposure to cascading failures in
terms of thermal overloading, voltage violation and voltage collapse,
and recommend preventive actions to halt the cascading
On-line TSAT in production to assess transient stability and calculate
transient stability interface limits in real-time
EMS:
StateEstimatorcases
Studycases
Powerflowmodel
OnlineTSAT Contingencies OnlinePCM
Transientstabilityanalysis Recommendation
ResultsofIROL
andinterfacelimits ofpreventive
identification
calculation actions
TimerforIROLviolations
andIROLreporting
150
RoadmapforIROLanalysis
Period 2015 2016 20172018

Goal Automatequarterly Onlinesteadystate Onlinesteadystate
IROLanalysis IROLanalysis andtransientIROL
analysis
Tasks DeployPCMoffline DeployPCMonline Deployautomated
DeployTSAToffline Pilottransient transientcascading
cascadinganalysison analysisonline
line
AddedFunctionality Abilitytodo Onlinecascading Comprehensiveon
automated,offline analysisforsteady linecapabilityforIROL
IROLanalysisfor state analysis
historicaland future Onlinetransient Abilitytodosimilar
studyperiods IROLanalysisin analysisforhistorical
test/pilotmode andlookahead
periods
UseofCascadingAnalysisresults
StateEstimationcase Setofcontingenciesbeyondnormal
or contingenciesusedindispatch
Studycase N2
Stuckbreaker
PCMtool
Recommendationson
preventiveand
Resultsofcascading correctiveactions
identificationand
classification
Preventionofcascading:
IROL Situational
addcriticalcontingencies
Reporting awareness
intodispatch
151
BenefitsofUsingPCMforThermalandVoltage
IROLAnalysis
AutomatedprocessofpotentialcascadinganalysisusingEMSState
Estimatorcases
Comprehensivestandardizedreporting
WelldefinedcascadingcriteriamakingIROLviolationanalysisaconsistent
(independentfromthesubjectiveengineeringjudgment)andauditable
process
EnhancedsituationalawarenessbyreliableidentificationofPCMsinitiated
bycontingenciesbeyondthoseusedfortraditionaldispatch
ReducecascadingriskbyincludingcriticalPCMcontingenciesintodispatch
Multiscenarioanalysistoqualitativelyidentifytheprobabilityof
cascadingashigh,mediumorlow
BenefitsofUsingTSATforTransientStability
IROLAnalysis
Significantlyreducecasepreparationtimefromdaystominutesforoff
lineanalysis,andautomaticcasepreparationonline
Accuraterepresentationofsystemoperatingconditionsusingrealtime
model,eliminatingtheinaccuracyanddiscrepanciesintroducedbyusing
offlinemodel
Fastercomputationalspeed(executiontimeofTSATisabout1/3to1/2of
PSS/E)
Userfriendlygraphicinterfaceforengineeringanalysisandresults
presentation
Additionalanalytictools,suchasProny analysismodule,allowengineers
toconductmorecomprehensiveanalysis
Automaticallycalculatestabilityinterfacelimitwithuserspecified
stressingpatternandstabilitycriteria(e.g.,transientvoltagesag,damping,
lossofsynchronism,etc.)
152
BenefitsofonlineIROLAssessment
DramaticreductioninmanualeffortsforassessmentandreportingofIROL
events
IncreaseaccuracyinIROLassessmentandenhancesituationalawareness
ConsistentandauditableIROLassessmentprocess
AutomaticonlineIROLidentificationandclassificationwithouttheneed
tohaveapredefinedlistofIROLinterfacesandpracticallyeliminatingthe
needforofflineIROLanalysis
EliminatetheriskthatexistingpracticeofpredefinedsetofIROL
interfacesmaynotcoverallpotentialIROLsituations
ReducethenumberofreportableIROLeventsbyreportingonlyanalyzed
andconfirmedIROLviolationsandprovidingonlinerecommendationsto
mitigatepotentialcascadingevents
Conclusions
DramaticimprovementtheefficiencyofIROLanalysisprocess
PCMandTSATtoolsareproposedtoautomatetheexistingquarterlyIROL
analysis
FutureonlineIROLassessmentwill
improveaccuracy,efficiencyandrobustnessoftheanalysis
reducethenumberofreportedIROLviolations
improvesystemreliabilityandsituationawareness
PotentialtoreducethenumberofreportedIROLviolations
PotentialtoreducethenumberofmonitoredIROLinterfacesandachieve
significant$$savingsforTOs
Overallgoalistoimplementaplatformwithadvancedtoolsandwell
definedauditableprocesseswhichenabletheISOtoevaluatetheIROL
violationsautomatically,accuratelyandconsistently
153
Session6
6. RestorationfromCascadingFailures:
6.1 TransmissionSystemRestorationinRealTime
Instructor: VijayVittal
6.2 ChallengesofElectricSystemRestorationin
NewYorkCity
Instructor: MichaelForte
Session6.1
TransmissionSystemRestoration
inRealTime
VijayVittal
IraA.FultonChairProfessor
ArizonaStateUniversity
154
Outline
Introduction
Constraintsinpowersystemrestoration
Systemsectionalizingandautomatic
restorationpathselection
Casestudy
Conclusions
NERCCriteria
1. Preparedness
Developarestorationplantoreestablishtheelectricsystem
Ensuretheavailabilityandlocationofblackstartcapability
Verifytherestorationprocedurebyactualtestingorsimulation
2. Restoration
Restoresystemfrequencytonormalandmaintainadjusted
interchangeschedules
Notificationtootherentitiesasthestepsoftherestorationplanare
implemented
3. Response
Theplanmustaccountforthepossibilitythatrestorationcannotbe
completedasexpected
155
Motivation
1. Off-line restoration planning
Only for specific systems or areas
Rely on experience of dispatchers to select restoration path
No on-line computer tool for system restoration
2. Decision support tool is desirable

Need near real-time system information to make decision
Difficult to predict changing network configurations and loading levels
Need analytical tool to check system constraints
Decision support tool
SystemRestorationStrategy
Power System Restoration
Transmission Customer/
Load
System
Strategy
Module Constraints
Checking
Parallel Black-Start Power Balance Voltage Stability

Restoration Constraint Constraint Constraint
Restoration Path Thermal Constraint Stability Constraint Voltage Constraint
Selection
156
SystemReliabilityCriteria
Identifyvulnerabilities,assessrisksandsatisfy
reliabilitycriteria
Thermalconstraintsontransmissionlines
Sustainedovervoltagesduringearlyrestoration
Switchingsurges
Unstablephenomenonofselfexcitation
Maintainingsteadystateandtransientstabilityduring
restoration
UseofPowerTransferDistributionFactors
(PTDFs)
Thisideawasoriginallydevelopedincontingency
analysisandevaluatedfortheremovalofbranches
orthelossofgeneratorsatspecificnodes
Thenoveltyoftheapproach istoplansystem
restorationbycalculatingthePTDFsforcandidate
linestobeclosed
Todeterminethecorrectsequenceforenergizingthe
lines,PTDFsandweightingfactorsareused
Theuseofthisideainsystemrestorationisunique
157
ProposedMethods:
PTDFBasedRestorationPathSelection
1. Thermalconstraintontransmissionlines
Relievetheoverloadedtransmissionlinestopreventthesystemfrom
furtherdegradation
Allthenewlyrestoredlinescanbedividedintotwocategories:
Radiallineswhichwillcreateabranchbetweenanexistingnodeanda
newnode
Loopclosurelineswhichwillcompletepathsbetweentwoexisting
nodes
2. Stabilityconstraintinsideeachsubsystem
Systemhasappropriatesecuritymarginsoitcanwithstandthe
disturbancesduringtherestorationprocess
ThismethodisbasedontheNERCN1 criterionandsystemtransient
securityanalysis
3. Voltageconstraintwhenenergizingtransmissionline
ThermalConstraints
1. TheRPIisthesumoftheproductsofthe Radiallines whichwill
weightingfactorandpowerflowchangein createabranchbetweenan
eachexistingtransmissionline.Thecandidate existingnodeandanewnode.
radiallinewiththelowestRPIisrestoredfirst
NL
RPI 1 p 1
S p p
2. Loopclosurelinesshouldbefirstevaluatedfor
restorationinordertoreleasethestressonthe Loopclosurelines which
linesthatarealmostattheirlimit. willcompletepathsbetween
add_ij [(Zbus )ln (Zbus )mn]
new new *
alreadyexistingnodes
lm,n *
zlm
old
[(Zbus )in (Zbus
old old
) jn]* {[(Zbus )il (Zbus
old
) jl ] [(Zbus
old
)im (Zbus
old
) jm]}*
lm,n * *
Zloop zlm

q
RPI 2 add _ line S p p
p 1
158
PathDeterminationAlgorithm
Scenario2(Mostlyinurbanarea)
Restorationofanewradiallinecan
resultinlimitviolationsonheavily
loadedlines.Loopclosurelinesshould
befirstevaluatedforrestorationin
ordertoreleasethestressonthelines
thatarealmostattheirlimit
Scenario1(Mostlyinruralarea)
Ifpowerflowonalltherestored
linesarenotclosetotheirlimits,the
existingcurrentpercentagewillbe
usedasaweightingfactoronthe
changeinpowerflow(obtained
fromPTDFs)toevaluateeach
possibleradiallinepathrestoration
StabilityConstraints
N1 criterionandareadeterminationalgorithm
Duringtheearlystagesofrestoration,generationandloadinthesystemare
keptataverylowleveltomaintainsystembasicoperation
TheremightbeseveralradiallinecandidatesthathaveRPIvaluesthatare
closetoeachotherbutcouldresultintherestorationprocessbringingbackto
servicetotallydifferentloadareas
Theloadareastoberestoredshouldbedefinedbysystemoperatorsbasedon
systemconfigurationorloadpriorities
Iftherearemorethanoneloadareas,thesequenceinwhichtorestorethe
loadareasisdeterminedbasedontheNERCN1 criterionandsystem
transientsecurityanalysis
SoftwarepackageTSAT(DSATools)isusedtoassesstheseverity
ofapossiblecontingencyineachloadarea
Theseverityofacontingencycanbeassessedusingthetransient
stabilityindex(TSI).(max isthemaximumangleseparationofanytwo
generatorsinthesystem) 360
TSI max 100 100 TSI 100
360
max
159
VoltageConstraints
Basically,sustainedovervoltagescanbecontrolledby
absorbingthereactivepowergeneratedbythelightly
loadedtransmissionlines
havingsufficientunderexcitationcapabilityonthegenerators
pickinguploadswithlowpowerfactor
switchingonshuntreactors
adjustingtransformertaps
operatingFACTScontrollers,suchasSVCsandTCRs
Beforeenergizingthetransmissionlinesselectedusingthe
RPIapproach,sustainedovervoltagesshouldbeevaluated
tomakesurenovoltageviolationsoccur
RealisticSystemDisturbance
Theproposedmethodistestedonstormrelatedoutagesin
theWesternregionoftheEntergysystem
Intheoperationsplanningcaserepresentingsummerpeakconditions,
theWesternregionloadwasapproximately1900MW
Followingthestormeventatotalof11linesweretrippedseparating
theWesternregionfromtherestoftheEntergysystem
Crankingpowerandloadareadefinition
Outsidesource LewisCreekunitswhicharethemajorgenerating
unitsintheaffectedareacouldnotmeettheimmediateloaddemand
andsomeofthemsustainedminordamageduringtheevent
Loadareasaredefined TheWoodlandsareawhichhasahigh
concentrationofresidentialandcommercialloadsisdefined
asAreaI
160
June15,2005WesternRegionDisturbance
Area II
Area I (Woodlands
--Conroe Area)
LoadareasboundaryinWesternRegionoftheEntergysystem
ProposedSystemRestoration
Step1:Sending
crankingpowertothe
generators
Step2:Stability
constraintsanalysis
Step3:RestoreAreaI
Step4:Checkthe
stabilityconstraintsin
AreaIandrestoreAreaII
Finishsystem
restoration
161
Intermediatesteps(1)
Thefinaldecisiononrestoringthe
linescouldbelefttotheoperators
basedontheirexperienceand
safetyconsiderations
Theerrorbetweenactualpower
flowandthePTDFpredictedpower
flowerroriswithin6%
60
50
40 97714 - 97721
MW
30 97478 - 97541
97476 - 97543
20
97543 - 97471
10
97471 - 97461
0
Before adding a Restore line to Restore line to Restore line to Restore line to
line 97466 97544 97458 97475
Line candidates
162
Intermediatesteps(2)
LineSecurity(97456) Jayhawk
(97542)andlineCaneyck
(97471) Pechck (97543) are
closetotheirthermallimit
Thealgorithmchoosesline
Cleveland (97475) Jacinto
(97476)whichconnectstwo
busesthathavealreadybeen
energized
160
140
120
100
Power flow before closing
MW
80
new line
60 Power flow after closing
40 new line
20
0
97456-97542 97543-97471 97475-97476
Transmission lines
163
Sustainedovervoltagecheck
Beforeenergizingthe
transmissionlines,sustained
overvoltages shouldbe
evaluatedtomakesureno
voltageviolationsoccur.
Reducegeneratorterminal
voltages
Connectshuntreactivesource
164
ComparisonoftheLoadCurve
Amoreefficientapproachintermsofthenumbersofoperationstorestore
theunservedloadinthesystemthantheactualrestorationoperation
Thetransmissionthermallimitandtransientstabilityconstraintsarealso
satisfied
Theremaybepracticallimitationsinenergizingsomeofthelinesdueto
damagesthatcouldhaveoccurredduetotheoriginaldisturbance
2500
Actual operation
2000 New plan
1500
MW
1000
500
0
1 5 9 13 17 21 25 29 33 37 41 45 49 53 57 61 65 69 73
operations
Summary
Providesasystematicmethodfordevelopingan
automaticrestorationpathselectionprocedure
Thesuggestedapproachusesthepowertransfer
distributionfactoralgorithmandweightingfactors
todeterminetherestorationsequence
Therestorationperformanceindicesareeffective
duringtherestorationofthetransmissionsystem
astheyprovideguidancetotheoperatorsonhow
transmissionlinesshouldberestored
Theproposedrestorationstrategyisdemonstrated
onarealisticpowersystemdisturbance
165
Session6.2.
ChallengesofElectricSystem
RestorationinNewYorkCity
MichaelForte,ChiefEngineer TransmissionPlanning
ConsolidatedEdisonofNewYork
Agenda
Unique System
UniqueAspects
Characteristicsof Restoration
ofNewYorkCity
ConEdisonSystem Challenges
166
UniqueAspectsof
NewYorkCity
Facts NYC
TransitSystem
CityStats
Veryhighpopulationdensity Subway
5.3millionaverageweekdayriders
2800schools&colleges
1.6billionridersannually
60majorhospitals 840milesoftrack
Around6000highrise 468Subwaystations
buildingsbetween115ft to 1.2millionpeopletravelinto
1776ft Manhattanbycars,trucks&
Financialcapitaloftheworld taxiesduringpeakhours
Art,cultureandfashion 110millionpassengersserviced
capital atNYCairportsannually
167
ComparisonofNYCwithotherMajor
MetropolitanCities
Population Size Density

City
(million) (Squaremiles) (Person/Square mile)
NewYork 8.4 469 18,000

Chicago 2.7 234 11,500
LosAngeles 3.9 503 7,800
Denver 0.6 155 4,000
Unique
Characteristicsof
ConEdison
System
168
ConEdisonSystemDesign
Characteristics
Alltimepeakdemandrecordof13,322MWsetonJuly19,2013at5p.m.
Insomeareas,LoadDensityisgreaterthan2000MW/SquareMile.Largecustomersinclude
NYSE,EmpireStateBuildingandWorldTradeCenter.
HighvoltageTransmissionStationsandfeeders,either345kVor138kV
2,297distributionfeedersand62areasubstations
NetworkSystem Underground
95,757milesofundergroundfeeders
Networks:25 450MW
The95,757milesofundergroundfeedersin
RadialLoads:1 25MW
theConEdisonsystemcouldwraparoundthe
MostlyUnderground
Earth3.8timesorcouldstretchupanddown
MultipleFeeders2nd contingency
Broadway7,302times
protection
Impervioustohighwind
Highreliability
HighReliability
GenericHierarchyofPowerFlowandVoltage
Levels
169
System
Restoration
Challenges
Challenges
RestorationTime Restorethesystemaspromptlyaspossible
BlackstartCapacity Havingadequategeneratingunitstorestorethe
powergridintooperationwithoutrelyingonexternalsources
DielectricFluidPressure Maintaindielectricfluidpressureon
undergroundfeeders
VoltageControl Maintainpropervoltagelevels
NetworkLoad Mustenergizeentirenetworksimultaneously
170
RestorationTime
Challenge
Restorethesystemaspromptlyaspossible
Issue ActionTaken
UtilizationofGenerationMix Gas, Developedrestorationplansthat
SteamandNuclear dependmainlyongasturbines
SimpleCycleGasTurbine 10to60
minutesstartup
CombinedCycleGasTurbine 60to
180minutesstartup
Steam 8to24hoursstartup
Nuclear 2to3daysstartup
GenerationMix
8,000
7,000
Net Summer Capacity (MW)
6,000
5,000
4,000
3,000
2,000
1,000
0
1965 1977 2003 2015
SteamElectricUnits GasTurbineUnits NuclearUnits
171
BlackstartCapacity
Challenge
Havingadequategeneratingunitstorestorethepowergridintooperationwithout
relyingonexternalsources
Issue ActionTaken
ConEdisonmustrelyonthirdparty Rulesandtariffsweredevelopedto
generation enableConEdisontohaveadequate
blackstartresources
ReliabilityRuleonBlackstart
StatewideRule
Objective:torestorethesystemafterablackoutaspromptlyaspossible
Blackstartresourcesmustbetestedannually
GeneratorOperatorsmustbetrainedannuallyonrestorationprocedures
NewYorkCity(ConEdison)LocalRule
IfConEdisondemonstratesthattheprovisionofblackstartservicesbyaunit
materiallyaffectsrestoration,blackstartbecomesmandatoryfornewand
repoweredunits,aswellasforexistingunitsnotcurrentlyprovidingtheservice,
unlesstheunitcandemonstratetotheNYISOthatitwouldbeunreasonably
burdensometodoso
172
DielectricFluidPressure
Challenge
Mustmaintaindielectricfluidpressureonundergroundfeeders
Issue ActionTaken
Thereisariskoffeederfailure,ifthe Ineventofcompletelossofpower,an
feederremainsinservicewiththe emergencydieselgeneratorisutilized
fluidpressureaboveorbelow tooperatethepressurizationplant
predeterminedlevels
DielectricFluidFilledFeeder
173
VoltageControl
Challenge
Energizesystemwithinvoltagecriterialimits(between0.95 1.05pu)
Issue ActionsTaken
Energizedundergroundfeeder Shuntreactorsusedtoprotect
generatescapacitivereactive transmissionlinesfromhighvoltagelevels
power byabsorbingreactivepower,thus
Capacitivereactivepowerraises stabilizingthevoltagesalongtheline
systemvoltage Loadflowsimulationoftransmissionand
Undergroundfeedercharging
couldbeashighas300Mvar
networkpickuptomeetvoltagecriteria
limits
VoltageControlDevices
174
NetworkLoad
Challenge
Mustenergizeentirenetwork(25 450MW)
Issue ActionsTaken
Highinrushcurrentswhilepickingup
ProperlydesignedRestorationStepsto
anentirenetwork
balancegenerationandloadateach
step
LoadflowsimulationofRestoration
Stepstoassurefeasiblesolutions
TypicalNetworkConfiguration
175
Summary
Outstanding System Wide Reliability
UniquenessofNewYorkCity
Financialcapitaloftheworld
UniquenessofConEdison
Extensivetransitsystem
VeryhighPopulationDensity
System
LoadDensityashighas2000
SystemRestoration
MW/SquareMile
Challenges
MostlyUndergroundSystem
RestorationTime
NetworkSystem
BlackstartCapacity
DielectricfluidPressure
VoltageControl
NetworkLoad
Session7
7. AnalysisofPastBlackoutsCausedbyCascading
Outages LessonsLearned:
7.1 CascadingCharacteristics
Instructor: RobertCummings
7.2 PracticalStepstoMonitoringforCascadingOutages
Instructor: BrettWangen
176
Session7.1
CascadingCharacteristics
RobertW.Cummings
NERCDirectorofReliabilityInitiatives
andSystemAnalysis
Wherewereyou?
5:16:11pm 1965Northeast
Blackout
9:29pm 1977NewYorkBlackout
3:48pm 1996WesternBlackout
3:05:41pm 2003Northeast
Blackout
177
SomeTrendsinEventAnalysis
1. Protectionsystemmisoperations(39)
2. UnexpectedGen.TurbineControlAction(33)
3. Transmissionequipmentfailures(18)(mostinitiating)
4. Voltagesensitivityofgen.aux.powersystems(13)
5. HumanError(12)
6. Neartermloadforecastingerror(6)
7. Wiringerrors(incidental)(5)
8. Relayloadability(5)
9. Interareaoscillations(5)
10. SPS&RASMisoperations(5)
TwoTopDisturbanceElements
33 Unexpectedgeneratorturbinecontrolactions
29in8eventssinceAugust2007
2unitsCAUSALinasystemseparationevent
13 Voltagesensitivityofgenerationauxiliarypowersystems
(notincludedabove)
Problem:THESEBEHAVIORSARENOTMODELEDORSTUDIED
Turbinecontrols(nongovernor)arenotmodeled
Typicaldynamicanalysis onlyanalyzingt=0tot=20seconds
178
NORTHEASTBLACKOUT
AUGUST14,2003
2003BlackoutSignature
MW/MVAr kV
4000 300
MW
3000 250
kV
2000 200
1000 150
0 MVAr 100
-1000 50
-2000 0
16:10:38 16:10:40 16:10:42 16:10:44 16:10:46 16:10:48
Time - EDT
179
MajorPathtoClevelandBlocked
ONTARIO
4:08:59 - 4:09:07
PM
GenerationTrips
ONTARIO
180
CascadeMovesintoMichigan
4:10:36 PM
PowerTransfersShift4:10:38.6PM
181
NortheastIslandSeparatesfromEI
North of Lake
Superior
4:10:43 4:10:45
PM
EndoftheCascade
182
2003BlackoutAnalysis 16:10:36 16:10:38
16:10:38.6
16:10:41.9 16:10:45.2 16:10:48 16:10:50 16:10:56
16:10:49
16:10:39.5 16:10:42
5,000
New York into Ontario
4,000
New York into New England
3,000
Ontario into Michigan
Power Flows (MW) 2,000
1,000
0
-1,000
-2,000 PJMinto New York
-3,000
Detroit, Cleveland separated from W. MI.

16:10:30 16:10:40 16:10:50 16:11:00
Time
1630 MW Detroit generation trips
NY and New England separate
Ontario reconnects with West NY

Ontario splits from West NY
Cleveland cut off from PA
Cleveland separates
from Toledo, islands
Split complete between

NY separates from NJ
Start of split between
NY separates from PA
East and West NY

East and West MI
16:10:30
to
16:11:00
64
63 NY-West
Frequency (Hz)
62
61
Lambton ONT-MI
60
59
58 NY-East
57
16:10:30 16:10:40 16:10:50 16:11:00
Time
2003 Blackout Simulations

3 Months to Build
140
Sammis-Star
120 345 kV
% of Normal Ratings (Amps)
100 CantC-Tidd
345 kV
80
Star-S.Cant
345 kV
60
Hanna-Jun
345 kV
40
Hard-Chamb
138 kV
W.Ak-PV Q21
138 kV
E.Lima-N.Lib
138 kV
E.Lima-N.Fin
345 kV
Hard-Chamb
Bkr Failure
W.Ak 138 kV
138 kV
Dale-W.Can
138 kV
Cham-W.Ak
345 kV
Star-S.Cant
138 kV
Clov-Torrey
138 kV
Babb-W.Ak
CantC Xfmr
345 kV
345 kV
Hanna-Jun
20
0
Outages
183
SomeKeyElementsofCascade
Vegetationmanagement
Relayloadability
Miscoordinationofgeneratorcontrolsand
systemprotection
Generatorunderfrequencyprotectionand
under/overspeedcontrolsnotcoordinated
PoleSlipping
UFLSfailures
Constrainedbyextremeundervoltage
Timedelaystoolong
AngularSeparationAnalysis
0
-10
-20
Normal Angle ~ -25
-30
-40
-50
-60
Relative Phase Angle
-70
-80
-90
-100
-110
-120
-130
-140
-150
-160
-170
15:05:00 15:32:00 15:44:00 15:51:00 16:05:00 16:06:01 16:09:05 16:10:38
Time (EDT)
Reference:
Browns Ferry Cleveland West MI
184
369
16:05:50to16:08:52
Western MI
Sammis - Star
Muskingum
Ohio Central -
Detroit Galion 15
Degrees
NJ East Lima
Fostoria Central
Ontario
NY
Cleveland
Detroit
16:08:50to16:10:50
out of Western MI
Synch
NY NJ 40
Degrees
NY-PJM
Separation
Cleveland
Separation Cleveland Ontario
185
LOSANGELESBLACKOUT
SEPTEMBER12,2005
EventOverview
2,877MWofloadlostinLosAngeles,
Burbank,&Glendalesystems
35circuitstripped
Causedbyhumanerror
186
KeyElementsofCascade
Severeundervoltage(72%on203kV)
Severeoverloadsforupto30minutes
Someunitsfailedtostartduetoauxiliary
equipmentproblemscausedbyundervoltage
Interactionofgeneratorprotectionand
RemedialActionSchemes
Protectionequipmentfailures
UPPERMIDWESTSYSTEM
SEPARATION
SEPTEMBER18,2007
187
EventOverview
NorthDakota,Minnesota,Manitoba,&
SaskatchewanseparatedfromEastern
Interconnection
SaskatchewanthenseparatedfromManitoba
andNorthDakota
OverfrequencytrippingofgenerationatBoundary
SaskPowerloss~900MWofload&generation
Underfrequencytripping
SaskPowerreconnectionhamperedbyhigh
voltageonopenendedlines
TheIsland
188
Mechanicalconductorfailure
Protectionsystemcommunicationschannelproblem
protectionmisoperationresulted
Undervoltageloadshedding
Slowspeed115&161kVwalkingcascade low
voltages&highcurrent
Highspeeddynamiccascade outofstep
Overfrequencygenerationtripping relaysettings
Overvoltageloadshedding
Underfrequencyloadshedding
Interareaoscillations(0.5Hzfamily)
FLORIDASYSTEMDISTURBANCE
FEBRUARY26,2008
189
EventOverview
Delayedclearingof138kVSLGfault
progressedtoa1.7second3fault
Lossof1,350MWloadnearfault
Lossof2,500MWofgenerationnearfault
Lossof2,300MWmoreloadbyUFLSprogram
Lossof1,800MWmoregenerationacrossthe
Region
Oscillatoryeffectsacrossinterconnection
FloridaUFLSActivation
FRCC RC Visibility
FRCC
EventSummary RC Footprint
Impactcontainedwithin
FRCC
Majorityofloadrestored
within2hrs
Nomajorequipment
Actuation of
damagereported UFLS
NothermalO/L
2Nuclearunits
(trippedasdesigned)
Generation Trips
Location of
138 kV 3 fault
190
InterconnectionwideImpact
Turkey Point (FPL)
Calloway
/ Rush
Island
TVA
Dorsey (MH)
Protectionsystemturnedoff
Singlefaultprogressedtothreefault 1.7seconds
Nuclearunitstrippedonundervoltageasdesigned
Eightturbinesunexpectedtrips;
auxiliarybusvoltageprotection
rateoffrequencychange burnerleanblowout
phenomenon
ProperUFLSaction preventedsystemseparation
Interareaoscillations
1,000MWswingsinTVA500kVsystem
600MWswingsinNewEngland345kVsystem
12kVswingsonOntario230kVsystem
191
PACIFICSOUTHWESTDISTURBANCE
SEPTEMBER8,2011
EventOverview
11minutecascading
2.7millioncustomersoutinAZ,S.CA,MX
Initiatedwhensingle500kVlinetripped
Powerredistributed,increasingflowsthrough
underlyingsystems,causingvoltagedropsand
equipmentoverloads
Rippleeffectledto
trippingoflinesandgenerators,
UFLSactivation
RASactivations
Trippingintertieseparationscheme
192
Phase1 PreDisturbance
Hot, shoulder season day; some
generation and transmission
outages
High loading on some key

facilities: H-NG at 78% of normal
rating; CV transformers at 83%
44 minutes before loss of H-NG,

IIDs RTCA results showed loss
of CV-1 transformer would load
CV-2 transformer above its relay
trip point
15:27:39: APS technician

skipped a critical step in isolating
the series capacitor bank at
North Gila substation; H-NG trips
Phase2 TripofHNG500kVLine
H-NG 500 kV trips at
15:27:39
15:27:39 15:28:16 APS tells WECC RC line

expected to be restored
quickly
H-NG flow redistributes: 84%

to SCE-SDGE (Path 44);
remainder to IID, and WALC
CV transformers immediately
overloaded above relay
settings
Path 44 at 5,900 amps; 8,000

amp limit on SONGS
separation scheme
193
Phase3 TripofCVXfmrs
15:28:16 CV-2 and CV-1
230/92kV transformers trip on
overload relays
15:28:16 15:32:10
Severe low voltage in WALC
161 kV system
Loading on Path 44
increases to 6,700 amps
Phase4 RamonXfmr Trip

15:32:10 Ramon 230/92kV
transformer trips on overload
relay
15:32:10 15:35:40
15:32:13 Blythe-Niland
161kV line trips
15:32:15 Niland CV 161kV
line trips
IID undervoltage load

shedding; loss of generation
and 92 kV transmission lines
Severe low voltage in WALC

161 kV system
Loading on Path 44
increases to 7,800 amps;
settles at 7,200 amps
194
Phase5 YumaSeparates
Yuma AZ Separates from IID
and WALC when Gila and
Yucca transformers trip
15:35:40 15:37:55
Yuma load pocket isolated on
single tie to SDG&E
Loading on Path 44
increases to 7,400 amps after
Gila transformer trip; to 7,800
amps after Yucca
transformers and generator
trip
Phase6 HighSpeedCascade
El Centro Pilot Knob 161kV
line trips; all IID 92 kV system
radial from SDG&E via S-Line
15:37:55
WALC 161 kV system voltage
returns to normal
Path 44 exceeds 8,000 amp

setting and timer starts
195
S-Line RAS trips IV
generation and S-Line,
isolating IID from SDG&E
15:37:58 15:38:02.4
Path 44 loading increases to
as high as 9,500 amps and
reduces to 8,700 amps after
isolating IID
7PhasesofEvent
196
SONGS separation scheme
operates; forms SDG&E,
CFE, Yuma combined island
15:38:21.2
Combined island frequency
begins to collapse
Phase7 YumaSeparates
Yuma island separates from
SDG&E when IV-NG 500 kV
trips on underfrequency
15:38:23.12
APS UFLS operates, but
insufficient to stabilize load
pocket
Yuma load pocket collapses
197
Phase7 CFESeparates
Underfrequency
SDG&E and CFE UFLS
continues
15:38:23.13 CFE separates from SDG&E
Phase7 CompleteBlackout
IID, Yuma, SDG&E, and CFE
blacked out
By 15:38:38
198
ParalleloverloadingofunderlyingnonBESelements
RelayloadabilityonnonBESelements
Loaddropoutonseverundervoltage
UnintentionaltrippingofRAS
Trippingofintertieseparationscheme
FullUFLSactivation
Unexpectedgeneratorturbinecontrolactions
Standingangleonopen500kVlinetoolargetoreclose
Questions?
199
Session7.2
PracticalStepstoMonitoringfor
CascadingOutages
BrettWangen
DirectorofEngineering
PeakReliability
AboutPeakReliability
PeakReliability
(Peak)isthe
Reliability
Coordinator(RC)for
theWesternUS,
BritishColumbia,
andNorthernBaja
California,Mexico
AESOisadjacentto
PeaksRCArea
200
PeakReliabilityRole
ReliabilityCoordinatorisPeaksprimary
function
BenefitofafullWesternInterconnectionview
andmodel
Seasonalcoordination,operationsplanning,
andrealtimeoperations
Provideservicestoimprovereliabilityofthe
WesternInterconnection
September8th,2011
Aboutthisevent
7,835MWloadlostinPacificSouthwest
Lossof500kVlineandunexpectedRASaction
contributedtosevereoverloadsandtrippingof
equipmentin92kVtransmissionsystem
FlowsincreasedNSintoSouthernCalifornia,
eventuallyreachingalevelforanothersignificant
RAStriggeredequipmenttripping
11minuteevent
201
RegionImpactedbyEvent
DriverstoPacificSWEvent
Howdidthiscascadingeventhappen?
LackofunderstandingofBESimpactsofsub100
kVtransmissionsystems
1,100alarmsin11minutesintheRCcontrolroom
Mostpathflowsinitiallywithinlimits,masking
severity
RemedialActionSchemesandunexpectedactions
Lackofunderstandingofequipmentrelaysettings
202
ChallengestoPredictingCascading
Modelingrequirements
Clearunderstandingofequipmentthatimpactthe
BES,includingsub100kVnetworks
CoordinationandmodelingofRASinassessments
WesternInterconnectionhasover250recognizedRAS
andmorethatlookandsmelllikeaRASbutarent
calledRAS
Awarenessofrelaysandrelaysettings
Needforstrongalarmingandsituational
awarenesstools
PracticalApproachtoIdentify
CascadingRisks
Moreoperationsplanningstudies
Nearrealtimeassessmentsusingvoltageand
transientstabilitytools
Cascadingoutagetestingusingapost
contingencythresholdsuchas125%
Assumeequipmenttripsatthislevel
Performadditionalstudiestodeterminewhere
trippingstops
203
CurrentPeakMethod
Test Data Not Actual
FutureSteps
MajorimprovementsinRASmodelingand
coordinationacrossplanningandoperations
horizons
VendorRASmodelingcapabilitiesarenot
sufficient
UnderstandwhatothernonRASautomatic
actionsexistonthesystem
Modelcoordinationtoensureappropriate
levelsofmodelingexist
204
FutureSteps(contd)
Desirearealtimetooltoeventuallytellus
whenthesystemisatriskofcascading
Withouttherightnetworkrepresentation,RAS,
andrelaymodeling,thetoolwouldneverwork
Peaksfocusisonensuringthedataandmodel
isaccurate
ItcantakehoursordaystoverifyRASand
relaysettings howwouldanautomatedtool
beaccurate?
NERCStandardsDirection
NERCIRO,TOP,andrealtimetoolsstandard
developmentisastepintherightdirection
Forcesentitiestohaveoperatingplansandto
havetheabilitytoperformrealtime
assessments
PRCstandardsworkwillputagreaterfocuson
RAScoordination
205
ConcludingRemarks
Riskremainsforcascadingoutages
Resistancetochange manyfeelthatour
collectiveprocessesaregoodenough
becausewehavebeenplanningandoperating
thesystemthiswayformanyyears
Collaborativeeffortsaroundtheindustry,
includingwithvendors,areneededtoaddress
theriskofcascadingoutages
Questions?
206

2015 PES GM CascadingPhenomenon Final To Print

Uploaded by

Copyright:

Available Formats

2015 PES GM CascadingPhenomenon Final To Print

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

2015 PES GM CascadingPhenomenon Final To Print

Uploaded by

Copyright:

Available Formats

UnderstandingCascadingPhenomenon:

Cascading Failure Working Group Tutorial

PES Working Group: Cascading failure is a sequence of

NERC: The uncontrolled successive loss of system elements

Simulations need to be validated with data

from NERC 1996 blackout report

For this cascade,

Each blackout has its own specific nature

Brazil 11/03/1999 5hours 97,000,000

Voltage Frequency Cascade System Lossof

Security Margins should be then

System Operator should be

IEEE 11 Bus bar Network Dynamic Voltage Security

Aged assets (e.g. 2005 Moscow blackout)

Inadequate reactive power reserve increased risk of voltage

Some important equipment out of service (e.g. 2006 UCTE)

Protection system failure (due to wrong settings or hidden

Human factors (e.g. 2006 UCTE)

Inadequate situational awareness

Blocking of tap changers

Italy Blackout September,2003

Healthy Steady State High Speed Healthy

Triggering Verify Feasibility of

1st Moyle Interconnector run back

Wind Speed >= Cut Out Speed

Disconnection of Wind Farms

Wind Speed >= Cut Out Speed

Disconnection of Wind Farms

Line trips due to

Line trips due to

Line trips due to

Line trips due to

Wind Speed >= Cut Out Speed

Disconnection of Wind Farms

Line trips due to

PES General Meeting, Denver, USA

2015 Mladen Kezunovic. All rights reserved.

2015 Mladen Kezunovic. All rights reserved.

2015 Mladen Kezunovic. All rights reserved.

2015 Mladen Kezunovic. All rights reserved.

2015 Mladen Kezunovic. All rights reserved.

2015 Mladen Kezunovic. All rights reserved.

Routine-based Local Monitoring Neural Network

Event-based Synchronized Sampling

Transient Stability Event Tree Analysis

Measurements Control Measurements Control

Power System and Protection System

2015 Mladen Kezunovic. All rights reserved.

Power System Modeling

Entergy system was

5. Assess 4. Security Control

2015 Mladen Kezunovic. All rights reserved.

2015 Mladen Kezunovic. All rights reserved.

2015 Mladen Kezunovic. All rights reserved.

2015 Mladen Kezunovic. All rights reserved.

Find Optimal Control

2015 Mladen Kezunovic. All rights reserved.

2015 Mladen Kezunovic. All rights reserved.

2015 Mladen Kezunovic. All rights reserved.