Transcript Document

INTEGRATION OF SYNCHRO-PHASOR
MEASUREMENTS IN POWER
SYSTEMS STATE ESTIMATION FOR
ENHANCED POWER SYSTEM
RELIABILITY
Hassan Ghoudjehbaklou, Ph.D.– Open Systems
International, Inc.
Gary Roskos – Open Systems International, Inc.
AGENDA
•
PMUs and the Smart Grid
•
PMUs and the State Estimation (SE)
•
Effect of PMU in Observability
•
Enhancing Solutions for Unobservable Islands
•
Certification Tests for SE PMU
•
Conclusions – How PMUs Can Help?
APPLICATIONS OF PMUS (WIKIPEDIA)
• Power system automation, as in smart grids
• Load shedding and other load control techniques such as
demand response mechanisms to manage a power system.
(i.e. Directing power where it is needed in real-time)
• Increase the reliability of the power grid by detecting faults
early, allowing for isolation of operative system, and the
prevention of power outages.
• Increase power quality by precise analysis and automated
correction of sources of system degradation.
• Wide Area measurement and control, in very wide area super
grids, regional transmission networks, and local distribution
grids.
PMUS AND THE SMART GRID
Phasor Measurement Units (PMUs) Provide
Synchronized, Wide-Area Power Measurements
• PMUs provide the Magnitude and Angle of all power
measurements at all grid locations simultaneously
• Measurements are available as frequently
as 30 times each second
PMUS AND THE SMART GRID
Thoughtful PMU deployment is a key element to
Smart Grid development at the transmission level,
accomplishing these Smart Grid goals:
Increased System Reliability, Efficiency and Security
•
•
High Quality, Real-time Data
Advanced Analysis, Optimization and Controls
•
•
•
•
Aggregate transmission operations and planning
Enhance grid security and stability (reliability)
Manage losses and congestion
Enhanced Communications Infrastructure and
Data Security
CURRENT EVENTS AND CHALLENGES
PMU Technology is Now Widely Available:
•
PMU Devices are Readily Available From Multiple
Vendors
•
Open Connectivity/Interoperability via
IEEE Standards
•
GPS and Communications Equipment is Affordable and
Hardened for Substation Use
•
Utility Communications Infrastructure is Improving Daily
CURRENT EVENTS AND CHALLENGES
Multiple Active Pilot Projects and Research Programs are
in Place
•
Basic Research in Massive-Volume, Real-Time Data
Processing and Dynamic Applications
•
Basic Research in New Approaches to Grid Stability
•
Strategic Deployment
•
Post-event analysis
•
Model verification
•
Data integrity and visualization
CURRENT EVENTS AND CHALLENGES
Implementation Hurdles
•
Device Deployment (
•
Communications Infrastructure Deployment (
•
Application Development - “Chicken and Egg” Problem
)
•
Can’t Justify Applications without Data
•
Can’t Justify Data Collection without Applications
)
PMU PROJECT LAB
SEL 5076
Sychrowave SW
SEL3306 PDC
OSI monarch
EMS System
OSI PMU IMPLEMENTATION
OpenPMU:
• Brings PMU Data Directly to
EMS
• Initially, Utilize EMS
Development/QA System
for Testing and Comparisons
Visualization Tools
OSI PMU IMPLEMENTATION
Early Lessons Learned (Learning)
•
IEEE Standards Revisions
•
Communications: Security/Redundancy/Failover
•
Visualization Tool Improvements
•
Troubleshooting
OSI PMU IMPLEMENTATION
SRP Research with Arizona State University
• Optimal PMU positioning in electric power system –
based on achieving maximum State Estimation
improvement (Prof. Heydt, Vittal)
• Synchrophasor technology in validation of T-line
impedance parameters (Prof. Tylavsky)
• Decision tree assisted online Security Assessment
using PMU measurements (Prof. Vittal)
• Generator dynamic parameters validation (Prof. Heydt)
PMU IMPLEMENTATION
Current and Future SRP PMU Uses
• Instantaneous State of the Electric System View
• Enhanced State Estimation (Measurement)
• Operator Visualization
• Black Start Visibility
• Line Impedance Derivation
• Disturbance Post-analysis
• Island Phase Angle Studies
PMU IMPLEMENTATION
PMU Observations
• Will be the Most Important Measuring Device in Transmission
System Monitoring and Control
• Will Revolutionize Power Systems Monitoring and Control
• Gradual Migration Towards Full PMU Implementation for the
Transmission Grid
• For Full Potential, a PMU System Must Have Communication
Infrastructure Support Including Coverage and Speed to Match
Streaming PMU Measurements
• WECC Synchronized Phasor Network (DMWG & WAMTF)
• NASPInet
OSI APPLICATION DEVELOPMENT
OSI is Working to Bring PMU Data into the
EMS Environment to Meet Several Goals,
Including:
 Ease of Implementation
 Solution Accuracy
Input Data
System Models
 Solution Speed
Increased Observability
 Development of Enhanced Visualization Tools
Situational Awareness
 Development of Enhanced Dynamic Analysis Tools
Take advantage of a reduced solution cycle
OSI APPLICATION DEVELOPMENT
Short-term Enhancements:
• Enhanced Communications Security
• Enhanced Fail-over Capabilities
• Enhanced Visualization Tools
Current OSI PMU-Specific Development:
•
•
•
•
•
•
-
Enhanced Data Access
Optimized Hybrid State Estimation
Advanced Data Archive/Historian Capabilities
Enhanced Dynamic Stability Analysis and Control
More Real-time and Historical Visualization Tools
Next-generation Data Security Tools
PMU DEPLOYMENT STRATEGIES
Limited Deployment
• Measurement and Model Improvement
• Both sides of a variable device (Phase-Shifter, LTC, DC Line, etc.)
• Measurement or Visibility Problem Areas
• Interconnections
Large-Scale Deployments
• Start at Highest Voltages
• Cover 500kV, then 345kV, etc.
• Grow Contiguous PMU Measurement Areas
• Start at one end and work toward the other
PMU DEPLOYMENT STRATEGIES
Long-Term Goals
• High-Quality, Sub-second State and
Model Measurement
• System state measured, not estimated
• System parameters measured, not calculated
• Dynamic events detectable
• Add Applications to Capitalize on New
Paradigm
PMUS AND THE STATE ESTIMATION (SE)
Effects on:
• Observability
• Solution accuracy for observable islands and
boundaries
• Bad data detection
• Solution accuracy for the unobservable
islands
TOPOLOGICAL OBSERVABILITY
Step 1: Determine the measurement islands.
All islands with PMUs will have the same
group/island number
Step 2: Reduce the effect of bad angle
measurements (Use Median of the angles)
Step 3: All Branches within a measurement
islands will have observable flows
TOPOLOGICAL OBSERVABILITY
Step 4: Enlarge the observable islands using
n-1 rule recursively
Step 5: If voltage/angle of both sides of a
branch are measured, add its calculated flows
as pseudo measurement, for added stability
and accuracy
Step 6: Change unobservable islands to
observables, if all injections are measured or
at most one injection is not measured
PMU measurements added to model studied by P. Katsilas, et. al. (2003)
PMU Voltage/Angle Measurements
Injection Measurement
Flow Measurement
Actual Flows
SE FLOWS (NO PMU)
SE Flows (W/ PMU)
SELECTION OF REFERENCE ANGLES
FOR SE (NO PMUS)
Electric Island 1
U2
Electric Island 2
O2
O1
Main Observable
Island
U3
U1
O1
Main Observable
Island
U4
U1
O3
O2
U2
SELECTION OF REFERENCE ANGLES
FOR SE (NO PMUS)
Action
In Flat start, initial angles are set to zero
Convergence
Good convergence of SE for Observable
islands
Poor convergence for unobservable islands
Accuracy of the SE solution
Good for inner observable island
Poor for close to boundaries
Worst for unobservable islands
SELECTION OF REFERENCE
ANGLES FOR SE WITH PMUS
Electric Island 1
U2
Electric Island 2
O2
O1
Main Observable
Island
U3
U1
O1
Main Observable
Island
U4
U1
O3
O2
U2
SELECTION OF REFERENCE ANGLES FOR
SE WITH PMUS
Action
In Flat start, initial angles of the observable
islands are set to the Median angles of all
PMUs of that island. Initial angles of
unobservable islands are set to zero.
Convergence
Good convergence of SE for Observable
islands
Poor convergence for unobservable islands
Accuracy of the SE solution
Good for inner observable island
Poor for close to boundaries
Worst for unobservable islands
HEURISTIC SELECTION OF
REFERENCE ANGLES FOR SE
WITH PMUS
Electric Island 1
U2
Electric Island 2
O2
O1
Main Observable
Island
U3
U1
O1
Main Observable
Island
U1
O3
O2
U4
Selection of PMU Based Reference Angle for SE
U2
HEURISTIC SELECTION OF REFERENCE
ANGLES FOR SE WITH PMUS
Action
In Flat start, initial angles of the observable islands are set
to the Median angles of all PMUs of that island. Initial
angles of unobservable islands are set to angle reference of
the electrical island.
Convergence
Good convergence of SE for Observable islands
Better chance of convergence for unobservable islands
Accuracy of the SE solution
Good for observable island
Good for close to boundaries
Good for unobservable islands (depends on schedules)
PMU SE CERTIFICATION DATABASES

Following slides present results for series of tests
for Phasor Measurement Units (PMU)
implementation in State Estimation (SE). Four
different databases are considered for this study:
 IEEE-14 (Power Flow solution as PMU Measurements)


Large Customer no. 1 (With actual PMU measurements)
Larger Customer no. 2 (No PMU Measurements)
PMU SE CERTIFICATION TEST 1

Test 1 – Verify Observability and solvability of the PMU
SE with only Phase angle and Voltage Magnitude
Measurements at all buses with no other measurements.
Compare the results with only bus injection measurements
or only branch flow measurements.

Action Summary – All tests completed with solution
matching within the tolerances

Conclusion – When all measurements are good, phase
angles and voltage magnitudes provide good observability
and accurate solution (This fact has been reported by other
researchers as well.)
PMU SE CERTIFICATION TEST 2

Test2 – Introduce some bad angle measurements to the
cases with all phase angle and voltage magnitude
measurements. Note the effect on the solution quality and
convergence.

Action Summary – Initially some tests completed and
bad angles detected. Later Median angle enhancement was
employed for the reference angle of the measurement
islands. That made all cases converge, when only few angles
were bad.

Conclusion – SE solution is very susceptible to bad angle
measurements (As reported by other researchers) and some
heuristics should be deployed.
PMU SE CERTIFICATION TEST 3

Test3 – Use databases with PMU measurements for the
existing large customers (if the large customer does not have
PMU, introduce some PMUs in the model and use phase
angles from a Power Flow solution as measurement.) Verify
Convergence of PMU SE.

Action Summary – Initially some tests completed when
phase angles where small. Later with enhancement for large
angles, all cases converged, when all angles where good.
Using the enhancement of Power Flow for PMU, all cases
converged and good results were obtained for the
unobservable as well as observable islands.

Conclusion – Classical SE and PF need to be enhanced to
handle both large and bad angle measurements..
PMU SE CERTIFICATION TEST 4

Test4 – Verify that adding phase angle and voltage
magnitude measurements actually changes observable
islands.

Action Summary – To observe any change in the
observable island the PMU measurements need to be close to
the boundaries in the unobservable islands.

Conclusion – Not all PMUs directly impact the quality of
the solution of the network. Some have more effect than the
others.
CONCLUSIONS



How PMUs can help SE.
Provides redundant measurement that could enhance
observability and improve quality of the solution for the
observable island.
Provides angle reference for measurement islands that
enhances stability and accuracy of the solution for the
unobservable island.
CONCLUSIONS








What Enhancements are needed for PMU
SE?
Enhancing Observability algorithm for PMU
measurements.
Good selection of PMU phase angles for measurements
islands.
Improved heuristics for handling unobservable islands.
What Other improvements are possible for
PMU SE?
Model verification (parameter estimation).
Real-time State Estimation of a critical sub-network.
Enhanced Visualizations.
QUESTIONS?