1. Setting-less Protection: Laboratory Experiments and Field Trials
A.P. Meliopoulos, G. J. Cokkinides
Georgia Institute of Technology
Paul Myrda
EPRI
B. Fardanesh, G. Stefopoulos
NYPA
C. Vournas
NTUA

2. The Substation of the Future PSERC Project 2007-08

3. Overall Approach
From
This To
This

4. Traditional Zone (Component) Protection
Monitor Specific Quantity or Quantities (current, differential current, distance, voltage over frequency, etc.) and Act When the Quantity Enters a Specified Locus (settings).
The Traditional Protection Approach Exhibits Limitations for the Simple Reason that the Specific Quantity that is Monitored Does not Always Represent the Condition of The Component Under Protection.
NERC: #1 Root Cause of System Disturbances is Protection Relaying

5. Complexity, Gaps & Challenges
A Modern Substation May Have Tens of Numerical Relays, each relay has an average of 12 protection functions. Coordination of all these relay functions is quite complex. Experts (humans) and Expert Systems (computers) are needed…
Tools to validate coordination of complex protection schemes are at best inadequate.
Protection Gaps: HIF, Down Conductors, Faults Near Neutrals, Inverter Interfaced Generation, Faults in Series Compensated Lines, etc. etc.
Protection Gaps Result in Fatalities…

6. Basic Questions
Does it make sense to continue designing relays that rely on (typically) three currents and three voltages?
How do we deal with hidden failures?
Is it necessary for each relay to be equipped with data acquisition systems? Should DAQ be separate from relays (logic devices)?
Present Systems and Technologies are Digital – They Provide Tremendous Flexibility – Are the capabilities of the technology used or we simply mimic E/M relays?
Is the technology available to move from zone protection to subsystem (such as substation) protection?
Can we (a) decrease Complexity? (b) identify Hidden Failures?

7. The Setting-Less
Protection Method

8. DSE Motivation: In Search of Secure Protection
Setting-less Protection can be viewed as Generalized Differential protection
Analytics: Dynamic State Estimation (systematic way to determine observance of physical laws)

9. Dynamic State Estimation
Given the dynamic model of a system:
And a set of measurements at times t, t+h, t+2h, …
Compute the best Estimate of the System State at times t, t+h, t+2h, …
We have developed three algorithmic approaches which provide similar results:
Treat actual measurements, virtual, derived and pseudo as data with certain uncertainty
Treat virtual measurements as constraints
Use the standard Extended Kalman Filter algorithm

10. Key Elements of Approach
“Digitization” Separate Data Acquisition from logic devices (relays, recorders, etc.) – Merging Unit Approach
“Objectify” the model and measurements of each component: Starting Point: component physical model.
“Interoperability” at all levels
Each logic device (IEDs) performs:
Protective functions for the component
Validate the “model object”
Perform parameter identification, if necessary
Transmits model objects to any other stakeholder
Extent this structure to the control center

11. “Objectification”: The SCAQCF Model
(State & Control Algebraic Quadratic Companion Form)
Compact
Component
Physical Model
Set of
Algebraic & Differential Linear & Nonlinear Equations
&
Inequalities
Quadratized
Model
State & Control
Algebraic
Quadratic
Compact
Form
(SCAQCF)
Addition of State Variables
Quadratic Integration

12. “Objectification”: Measurement Model
Actual Measurements
Derived Measurements
Pseudo-Measurements
Virtual Measurements
All h equations are quadratic at most

13. Implementation Issues
Typical Sampling Rates  ts = 0.1 ms to 0.5 ms
Challenges
Perform the Analytics at time less than (2ts)
Robust Operation of DSE requires accurate zone model
GPS Synchronized Measurements simplify Dynamic State Estimation (for linear zones it becomes a direct method)
Example: Data Acquisition is performed 4 ks/s  Dynamic State Estimation is performed 2,000 times per sec.
IMPORTANT ADVANTAGE: Relay detects fault within a few sampling periods

14. Protection of Saturable Core Transformers
Event:
14.4/2.2kV, 1000 kVA single- phase saturable-core transformer
A 800kW load connected to the system and generate inrush currents at 0.72 seconds
Followed by a 5% turn-ground fault near neutral terminal of the transformer at 1.52 seconds
Inrush Current
List of Measurements:
Single-phase voltages at two sides
Single-phase currents at two sides
Temperature measurements at selected points
Internal Turn-Ground Fault

15. Protection of Saturable Core Transformers
Inrush Current
Internal Fault
Setting-Less Protection - No Trip
Setting-Less Protection - Trip
Differential Protection – No Trip
Differential Protection - Trip

16. Laboratory Experimentation
Power System Control and Automation Lab (PSCAL)
System Under Test consists of:
Simulator of system
Amplifiers to create signals at usual instrumentation level
Relays or Merging units to perform data acquisition
Station or process bus, and
A personal computer attached to the station bus or process bus (executes the state estimator and the intrusion detection software)
We have allocated $60k of internal funds to upgrade the lab for further facilitating the work under this project.

18. Differential Protection for Turn-ground Fault 5% from Neutral
Graph show:
Diff Current (SV)
Op Current (f,rms)
Restraint Current
Diff Percentage
Trip Decision
Legacy Differential Protection MISSES the Fault

19. Setting-less Protection for Turn-ground Fault 5% from Neutral
Graph show:
Voltage Measrmnts
Current Measrmnts
DSE Residuals
Chi Square Test
Confidence Level
Trip Decision
DSE Execution Time
Setting-less Protection DETECTS IN SUB-msec AND TRIPS IN SUB-cycle.

20. Laboratory
Results & Animation

21. Laboratory Instrumentation
Our Setting-less Relay is an 8-core $2k PC

22. Implementation Overview: Data Flow
The Component is represented with a set of Differential Equations (DE). Measurements from legacy instrumentation or MUs.
The Dynamic State Estimator fits the Streaming Data to the Dynamic Model (DE) of the Component. Protection Logic based on DSE results.
Object Oriented Implementation

23. Implementation Overview: User Interface
MU Data Concentrator
Collects data from MU via Process Bus, Time Align Data and Feed into a Circular Buffer
MU Simulator
Generates Virtual MUs
COMTRADE Data
Enables ability to play back data into the relay from a COMTRADE file
EBP Relay
Performs the dynamic State Estimation, three user selected methods
Device and Measurement Model
Enables the setting of the EBP relay. User input of the protection zone model and measurement list
Reports
A variety of reports and visualizations
Implementation Overview: User Interface
Buffer Utilization Indicator
DSE Execution Time Indicator
100%=Sampling Period
XXXXXXX

24. Example of Reports & Visualizations
Laboratory Results
Example of Reports & Visualizations

25. Another Example of Reports & Visualizations
Laboratory Results
Another Example of Reports & Visualizations

26. EBP: Gatekeeper of Protection Zone Models
Protection is Ubiquitous
Makes Economic Sense to Use Relays for Distributed Model Data Base
Capability of Perpetual Model Validation
A Ubiquitous System for Perpetual Model Validation

27. Field
Demonstration

28. Field Demonstration on NYPA System Monitored Transformer
Marcy Substation
Marcy – Massena Line
Monitored Transformer

29. Conceptual Implementation on a 765/345/13.8 kV Autobank
Selected Hardware

30. Setting-less Relay
Hardware

31. Conceptual Implementation on a 765 kV Line
Selected Hardware

32. Conclusions
Setting-less Protection has been proven in a lab environment in six application areas
Prototype installation at NYPA under a NYSERDA grant
Multiple secondary benefits
Model Validation
Detection of Hidden Failures
Distributed State Estimation
Applications
Project Team
EPRI - Paul Myrda, E. Farantatos
Georgia Tech - A.P. Meliopoulos, George J. Cokkinides, Renke Huang, L. Sun, R. Fan, Z. Tan, Evangelos Polymeneas

33. Τέλος Acknowledgements
The support of DoE, EPRI, NYSERDA, PSERC and numerous industry partners Southern Company, NYPA, USVI-WAPA, PG&E, TVA and Burbank Water & Power is greatly appreciated
Τέλος