A (Smart) Real-time PMU-assisted Power Transfer Limitation Monitoring .

A (Smart) Real-time PMU-assistedPower Transfer Limitation Monitoringand Enhancement System Support Renewables on the Grid Exploring existing transmission infrastructure Enhance control room situational awareness andearly warning systemDr. Hsiao-Dong Chiang

PJM System PJM’s Base-case power system (Stateestimation EMS using CIM-compliance formator PSSE format) Look-ahead scenario (proposed power transfer,look-ahead loads, look-ahead generationdispatch scheme, planned outage schedule) PJM’s On-line Available transfer capabilitymonitoring system and (smart) enhancements(i.e. increase ATC in a smart way)

00Safe marginSafe marginthresholdthresholdMonitoring of critical angle differenceMonitoring of critical power transfer PMUicritical bus #i anglePMUj Criticalbus i Criticalbus jcritical bus #j angle

Monitoring & AnalysisGraphical Scheme50%100%Voltage Security Threat KeyVoltage Violation Type KeyRedDanger of VoltageCollapseVoltage CollapseOrangeDanger ofThermal LimitThermal LimitYellowDanger of VoltageViolationVoltage ViolationBlueSafe

Monitoring & Analysis (Base-Case)Main th

ContingenciesPower System Outageand BlackoutCornell University

ContingenciesShort-circuitby rcuitby non-lightening

Contingencies cause limits on power systemsHard LimitsTransient (angle) instabilityVoltage instabilitySmall Signal Stability

Contingencies cause limits on power systemsSoft LimitsVoltage limitThermal limit

Challenges and OpportunitiesATC Monitoring and EnhancementData issuesSystems Real-time network model of 13,000-bus, 18168braches Real-time data Verification of model and data

Model Validation & Correction Identify critical component and parameters. Validate and correct model structure and parametersusing measurement-based (mismatch-based) approach.Power SystemMeasuredDynamicResponsesWAMSDisturbancey (t )u (t )MeasuredInputPower SystemDynamic Model andData yˆ (t ) ModeledOutputValidation& Correction (t )MismatchError

Challenges and OpportunitiesATC Monitoring and EnhancementComputation ChallengesSystems On-line computation capability N-1 Criteria

Challenges and OpportunitiesATC Monitoring and EnhancementControl ChallengesSystems Optimal control design(priority-based, minimum number of controlactions and minimum amount of control actions) On-line optimization technologies

Problem statementsConsiderations (ATC monitoringsystems)1. ATC of the base-case power system2. ATC of base-case contingencies3. Which ones will cause ATC’s limitation ?(insecure contingencies)4. Which ones will push the system near itslimitations ? (critical contingencies)5. Where are the weak buses, weak areas ?

Architecture of Real-Time Stability MonitoringSystemsOn-line Voltage StabilityAnalysis (VSA)On-line Dynamic StabilityAnalysis (DSA)EnergyManagementSystem (EMS)Diagnostic &ComputationalInformationPhasorMeasurement yMonitoring

Monitoring & Analysis (Base-Case)Main th

Preventive & Enhancement Control PSouth

Real-Time ATC Monitoring SystemCurrent StateSafe marginThresholdvalues

On-line TSA (Transient StabilityAssessment) 12,000 plus buses in system model 1,300 generators 3000 contingencies 15-minute cycle for real-time EMS data 5 minutes in cycle allocated for contingencyscreening target is 1.5 seconds to 2 seconds percontingency

Model for each contingencyDifferential equationsNonlinear algebraicequations

Time-Domain Approach Speed: too slow for on-line applications Degree of Stability: no knowledge ofdegree of stability (critical contingencies vshighly stable contingencies) Control : do not provide informationregarding how to derive effective control

Dynamical behavior, a generator’sangle

Time-Domain ApproachPre-Fault System (Pre-fault s.e.p.)x(t) (Pre-fault s.e.p.)end point of fault-ontrajectoryFault-On System.x fF(x,y)t0 t tclx(t)end point of fault-ontrajectoryfault-on trajectoryt t0t tclNumerical integrationx(t)Post-Fault System.x f(x,y)tcl t t Direct Methods (Energy Function)fault-on trajectorytt t0t tclNumerical integrationinitial point of post-faulttrajectoryt1. The post-fault trajectory x(t)is not required2. If v(x(tcl)) vcr, x(t) is stable.Otherwise, x(t) may be unstable.post-fault trajectoryt tclNumerical integrationtDirect stability assessment is based onan energy function and the associatedcritical energy

History of Direct Methods R&D between 1950s and 1980s werebased on heuristics and did not work. Theoretical foundations were developed in1987 by Chiang, Wu and Varaiya(Berkeley) Practical methods, Controlling UEPmethod BCU method, were developed inthe 1990s.

History of Direct Methods MOD (mode of disturbance) method(1970-1980s) PEBS method (by Kakimoto etc.) Acceleration machine method (Pavellaetc.) Extended Equal Area Criteria (EEAC) Single-Machine-Equivalent-Bus (SIME) BCU method TEPCO-BCU method

Key developments Theoretical FoundationDesign of Solution AlgorithmNumerical MethodsImplementations (ComputerPrograms) Industrial User Interactions Practical system installations

Key developments1. Theoretical Foundation (gaininsights and build belief) Theory of stability boundary Energy Function Theory (extensionof Lyapunov function function) Energy Functions for TransientStability Models (non-existence ofanalytical energy function)

Key developments1. Theoretical Foundation (gaininsights and build belief) Theoretical Foundations of DirectMethods CUEP method and Theoreticalfoundation Theoretical Foundation of BCUmethod

Important Implications CUEP method is the key direct method To directly compute CUEP of the originalpower system model, the time-domainapproach seems to be the only approach These results serve to explain whyprevious direct methods did not work(motivation of developing BCU method)

Fundamentals of BCU MethodWhat: a boundary of stability region basedcontrolling unstable equilibrium pointmethod to compute the critical energyWhy: an effective method to compute CUEP.

Static and Dynamic Relationships?(Step 7)(Step 1) 1 uÝ – ------ U u w x y g1 u w x y u 2 wÝ – ------- U u w x y g2 u w x y w TxÝ – ------ U u w x y g3 u w x y xyÝ z 1 uÝ – ------ U u w x y g1 u w x y u 2 wÝ – ------- U u w x y g 2 u w x y w TxÝ – ------ U u w x y g3 u w x y x yÝ – ------ U u w x y g4 u w x y y MzÝ – Dz – ------ U u w x y g 4 u w x y y(Step 2)(Step 6) 1 uÝ – ------ U u w x y u 2 wÝ – ------- U u w x y w TxÝ – ------ U u w x y x 1 uÝ – ------ U u w x y u 2 wÝ – ------- U u w x y w TxÝ – ------ U u w x y x yÝ – ------ U u w x y yyÝ z MzÝ – Dz – ------ U u w x y y(Step 5) 1 uÝ – ------ U u w x y u 2 wÝ – ------- U u w x y w T xÝ – ------ U u w x y x yÝ 1 – z – ------ U u w x y y MzÝ –Dz – 1 – z – ------ U u w x y y4/22/96 HDC 0 – ------ U u w x y g1 u w x y u 0 – ------- U u w x y g2 u w x y w TxÝ – ------U u w x y g3 u w x y x yÝ – ------ U u w x y g 4 u w x y y(Step 3)(Step 4) 1 uÝ – ------ U u w x y u 2 wÝ – ------- U u w x y w TxÝ – ------ U u w x y x - U u w x y yÝ – ---- yMzÝ – Dz

Fundamentals of BCU MethodBasic Ideas: Given a power system stabilitymodel (which admits an energy function), theBCU method computes the controlling u.e.p. ofthe original model via the controlling u.e.p. of adimension-reduction system whose controllingu.e.p. can be easily, reliabily computed.

Fundamentals of the BCUMethodStep 1: define an artificial, dimensionreduction system satisfying the static as well asdynamic properties.(how ?) explores special properties of theunderlying original modelStep 2: find the controlling u.e.p. of thedimension-reduction system(how?) explores the special structure of thestability boundary and the energy function of thedimension-reduction system.

Fundamentals of the BCUMethodStep 3: find the controlling u.e.p. of theoriginal system.(How ?) relates the controlling u.e.p. of theartificial system to the controlling u.e.p. of theoriginal system with theoretical supports.

BCU Method Explores the special structure of theunderlying model so as to define anartificial, reduced-state model whichcaptures all the equilibrium points on thestability boundary of the original model,and then Computes the controlling u.e.p. of theoriginal model via computing thecontrolling u.e.p. of the reduced-state,which can be efficiently computed without

BSI &TEPCO JointDevelopment 1997 –Present (2011)2005 Development and Implementation of Modelsof Generator Controllers and Phase-shifters forBCU and GBCU Programs Improvement in the Performance of Groupbased BCU Programs2001200019991998199720032002 Study of the precision improvement for theGroup-based BCU Method Feasibility Study of Developing New TimeDomain Energy Indices for TEPCO PowerSystem Development of a Group-based BCUMethod – Part I: Research Development of Improved BCU Classifierfor TEPCO Incorporated Analytical System Study of the Applicability of ImprovedBCU Classifiers for Multi-swing StabilityAnalysis Continual Development of BCU Classifiers(Version 2)2004 Development and Implementation ofGroup-based BCU Program and Study onComputing Method of Energy Margin Indexfor BCU and Group-based BCU Methods Study of Detailed Excitation Models inBCU Program for TEPCO Power System Feasibility Study of Developing Screening Methods toDecide Network Reconfiguration and Network Reloadingfor Maintaining/Improving Transient Stability Feasibility Studies of Developing Time-Domain EnergyIndices for Dynamic Security Assessment Development of a group-based BCU Classifier – Part II:Development Enhancements of BCU Program with TEPCOTransient Stability Models Research into BCU Method for PracticalApplication to Comprehensive Stability Model Extensions of BCU Method to TransientStability Models with Non-smooth Load Models U.S. Patent allowed for issuance 11/02/2004: METHODAND SYSTEM FOR ON-LINE DYNAMICALSCREENING OF ELECTRIC POWER SYSTEM. A Second Patent Application is Pending

TEPCO-BCU TEPCO-BCU is developed under thisdirection by integrating BCU method,improved BCU classifiers, and BCU-guidetime domain method. The evaluationresults indicate that TEPCO-BCU workswell on several study power systemsincluding a 15,000-bus test system.

High-level OverviewSolution for PJM on-line TransientStability AssessmentsEMSData BridgeTo ProvideReal Time Data(BSI)Information exchangeTEPCO-BCU(BSI)DSA Manager& TSAT (PLI)Data Bridge contains common fixeddata for both TEPCO-BCU/TSATand local data required only byTEPCO-BCU or TSATResultDepositoryandVisualization(BSI & PLI)

PJM Evaluation ResultsReliability measure: TEPCO-BCUconsistently gave conservative stabilityassessments for each contingencyduring the three-month evaluation time.TEPCO-BCU did not give overestimated stability assessment for anycontingency. (1)

PJM Evaluation Results Foratotalof5.29millioncontingencies, TEPCO-BCU capturesall the unstable contingencies.Table 1.Reliability MeasureTotal No. ofcontingencyPercentage of capturingunstable contingencies5293691100%

Speed: TEPCO-BCU consumes a total of 717575CPU seconds. Hence, on average,TEPCO-BCU consumes about 1.3556second for each contingency.Table 2. Speed AssessmentTotal No. ofcontingencyComputation TimeTime/percontingency5293691717575 seconds1.3556second

Screening measure: Depending on the loading conditionsand network topologies, the screeningrate ranges from 92% to 99.5%Table 3. Screening Percentage AssessmentTotal No. of contingencyPercentage Range529369192% to 99.5 %

A summary The overall performance indicates thatTEPCO-BCU is an excellent screening toolThese unstable contingencies exhibit firstswing instability as well as multi-swinginstability.Table 4. Overall performance of TEPCO-BCU for on-linedynamic contingency putationspeedon-linecomputation100%92% to 99.5%1.3 secondYes

Proposed PJM Implementation20 ProcessorsTEPCO-BCU3000Contingencies3.75 to 5 Min.ContingencyRanking1 to 3000