You are on page 1of 31

Distribution Systems Fault

Analysis
Laurentiu Nastac and Anupam Thatte
Concurrent Technologies Corporation (CTC)

May 25, 2006


Objective
Integrate an intelligent, operational, decision-
support (fault locator) software tool to predict
the location of low impedance, momentary and
permanent (more than one minute) faults in
distribution power systems

2
Background:
Identified Simulation and Modeling Tools
Company / Software Product Name Company URL
Distribution Management System www.dmsgroup.co.yu
Application Software Group (DMS)
Optimal Technologies (AEMPFAST) www.otii.com

Electrical Distribution Design (EDD) www.samsixedd.com


Distribution Engineering Workstation (DEW)
Siemens PTI www.pti-us.com
(PSS/E and PSS/ADEPT )
CYME International (CYMDIST) www.cyme.com

Tekla Corporation (Tekla Xpower) www.tekla.com

RTDS Technologies (RTDS) www.rtds.com

Manitoba Research Center (PSCAD) www.pscad.com


3
Fault Locator Software Integration:
Features
Stand alone tool
Easily integrated in network fault analysis software platform
as a post-processing tool
Easily integrated as module (.dll - dynamic link library) in
DEW/AEMPFAST/PSCAD/CYMDIST
Combined heuristic/genetic algorithm (GA)
Rule-based approach
Use GA to minimize errors between measurements and
expected parameters
Quick runs
Adequate quality (experimentally validated)
Operational decision-support tool
Uses real time data for
Load conditions
Recorded currents
Recloser status
Customer (trouble) calls 4
Fault Locator Software Integration:
Technical Approach
Thevenin-Norton (e.g., Thevenin) analysis to predict
the fault currents in circuit components [1, 2]
Measured faulted circuit data from DTE used to
validate network fault analysis method
Approximately 10% typical errors exist in predicting the fault
currents for distribution systems [1, 2]
Preload-conditions & fault impedance accounted for [1, 2]
Accuracy in predicting fault currents does not significantly
depend on the distance from the fault location to the
substation [3]

[1] L. Nastac et al., 8th IASTED International Conference on Power and Energy Systems,
PES2005, October 24-26, 2005, Marina del Rey, CA.
[2] L. Nastac et al., IEEE 37th North American Power Symposium, October 23-25, 2005, Ames, Iowa.
[3] L. Nastac, Personal Communication, DTE, December 29, 2005. 5
Fault Locator Software Integration:
Network Fault Analysis with DEW
Comparison (% Err) of measured (DTE faulted data) and predicted
results with DEW*

Pre-fault Arcing Err


Loading [Ohm] [%]
0 -12.9
100%
1 -2.1
0 -9.8
75%
1 1.4

% Err = 100 * (Meas Ic Calc Ic) / Meas Ic

Calc Ic calculated post-fault phase C current


Meas Ic measured post-fault phase C current

*Post-fault current includes both fault current and load current


6
Fault Locator Software Integration:
Technical Approach
Algorithm Rules
Rule #1: Compare predicted fault currents with measured /
recorded fault currents (from the substation,
modern reclosers, etc.)
Rule #2: Recloser status (open/close)
Rule #3: Recloser V&I RMS values (if available)
Rule #4: Customer (trouble) call input file (if available)
Rule #5: Time synchronized phase angles and waveforms
as well as transient RMS current and voltage
values (if available)

7
Fault Locator Software Integration:
Procedures
1. Read input data:
Distribution System & Component data
Recorded currents at substation
Recloser information & Trouble call records

2. Compute load and fault currents in the


selected branch components
Compare them with recorded currents

3. Use recloser information to determine


circuit branches related to fault region

4. Knowledge-based engine
Database of signature library A
Heuristic rules / GA minimization

YES
5. Customer (trouble) calls?

6. Perform back-tracking
NO from customers to narrow A
7. Plot fault locations down fault locations
Write fault location report 8
Fault Locator Software Integration:
Technical Approach
Evaluating a GA to perform a minimization for each
circuit branch to narrow down the possible fault
locations
Recloser Substation -relay Branch #1
Fault
Customer call Branch #2

Recloser fault
location
Branch #5

Branch #3

Branch #4

9
Fault Locator Software Implementation:
DTEs Orion Circuit
Experimental

Zoom area

Simulated

Legend
Blue Circuit components
Black Fault currents
Red Predicted faults locations
10
Fault Locator Software Implementation:
Predictions of Faults (DTEs Orion)
Trouble call Recloser Fault location

a) No faults b) Fault currents c) + Recloser & trouble calls


Legend
Blue All components (1078)
Black Fault currents (10% Error) = 88
Red Possible faults locations = 43
Green Recloser
Light Blue Customer (trouble) call
11
Fault Locator Software Implementation:
Predicted Results
Predicted numbers of possible fault locations
DTE Distance Number of Rule #1 Rule #2 Rule #3 Rule #4 Rule #5
Faulted from system
Circuit Name fault components Current Recloser Recloser Customer Other
location status current/ trouble intelligences
to voltage call (GA, ANN,
substation RMS input file waveforms,
[ft] values phase angle)
Orion #2 6500 1078 88 67 Phase II 43 Phase II
July 4, 2003 (all components
selected)
Orion #2 6500 1078 23 19 Phase II 9 Phase II
July 4, 2003 (125 selected*)

*Selected components were defined by DTE as the most likely components to fail
during an outage event
12
Fault Locator: Predicted Results
Predicted number of possible permanent fault locations, assuming 10% errors in estimating
fault currents (recorded fault locations were captured in all predictions)
Predicted numbers of possible fault locations
DTE Distance Number of Number of Rule #1 Rule #2 Rule #3 Rule #4 Rule #5c
from system selected
Circuit fault components components Current Recloser Recloser Customer Other
namea location status current/ (trouble) intelligences
to voltage call (GA, ANN,
substation (RMS input file waveforms,
[ft] values) b phase angle)
Clark 6900 2300 188 12 8 Phase II 3 Phase II

Orion 6500 1078 125 21 17 Phase II 6 Phase II


#1

Orion 6500 1078 125 23 19 Phase II 9 Phase II


#2

Mac 19100 2401 169 23 8 Phase II 4 Phase II

Notes: a. DTEs Orion circuit two different faults that occurred in different times at the same location
b. Rule #3 will be validated in Phase II
c. Rule #5 will be implemented in Phase II 13
Project Budget with Milestones

Phase I (FY06, July 05 June 06): Distribution


Systems Fault Analysis
Budget: $135K
Milestones and deliverables
April 30, 2006: Completed integration of the fault locator
software tool for predicting the locations of permanent
faults in distribution power systems
May 31, 2006: Validate fault locator with DTE measured
data
June 20, 2006: Show results of validation to DTE
June 30, 2006: Final report

14
Project Budget with Milestones (Contd.)
Phase II (FY07, July 06 June 07): Advanced
Fault Analysis System (AFAS)
Budget: $158K
Milestones and deliverables
July 31, 2006: Communicate with AEP on possible
involvement and extension of fault data for longer
distribution line applications
March 26, 2007: Complete integration of AFAS software for
predicting faults in distribution power systems
May 31, 2007: Validate AFAS with additional measured
data from DTE
June 15, 2007: Show results of validation to DTE and other
possibly utility stakeholder
June 30, 2007: Final report 15
Interactions & Collaborations
Phase I (FY06)
DTE Energy (Stakeholder)
Optimal Technologies (AEMPFAST Software)
EDD Inc. (DEW Software)
Phase II (FY07)
DTE Energy (Stakeholder) and other possible utility
stakeholder
Optimal Technologies (AEMPFAST Software)
EDD Inc. (DEW Software)
Nayak Corporation (PSCAD software) (Subcon)

16
Technical and Economic Benefits
Distribution systems fault analysis software will
significantly enhance ability of distribution utilities to
provide protection, operational and planning personnel
with
Improved fault diagnosis technologies that enable anticipating,
locating, isolating and restoring faults/failures with minimum
human input and fast response time
Specifically, current fault analysis software can give:
Improved system analysis (protection, planning and
operational)
Reduced outage time (improved restoration time)
Increased service and component reliability

17
Concluding Remarks
CTC identified and assessed several modeling and
simulation tools that can be successfully applied to
analyze, monitor, manage and control large and
complex energy systems at the distribution level
Comparisons of predicted fault currents with DEW
and AEMPFAST software tools with recorded
measurements from DTE were acceptable
Differences within 10% for pre-fault load current ranging
from 50100% and for arcing impedance ranging from
0.51 ohm
Accuracy in predicting fault currents does not significantly
depend on the distance of the fault location from the
substation
18
Concluding Remarks (Contd.)
Fault Locator Software: Very promising operational,
decision-support tool that can be used to predict
most likely fault locations in power systems
Numerical predictions were fully validated against
measured data from DTE
The numbers of possible fault locations were narrowed
down significantly by the fault locator software
Recorded fault locations were captured in all predictions
The following benefits are anticipated by using this tool
System analysis improvements at protection, planning and
operational levels
Reduction in outage time due to shorter restoration times
Increase in service and component reliability

19
Future Work
Predictive capabilities of the fault analysis software
will be significantly enhanced
PSCADTM software as well as other advanced tools
and algorithms (e.g., fuzzy logic, genetic algorithms,
neural tools, etc.) will be utilized in next phases
Phase II (July 06June 07) of the Distribution Systems
Fault Analysis project (i.e., Advanced Fault Analysis
System or AFAS) All algorithm rules will be implemented
into AFAS to accurately predict fault locations
Phase III (July 07June 08) AFAS will additionally use
transient analyses and data to intelligently anticipate
momentary and permanent faults

20
Acknowledgments
Concurrent Technologies Corporation conducted this
work under DOE cooperative agreement DE-FC02-
04CH11241. Such support does not constitute an
endorsement by DOE of the views expressed in this
presentation. Approved for public dissemination;
distribution is unlimited.
DTE Energy, Detroit, MI R. Lascu, D. Costyk, N.
Carlson, R. Sequin and H. Asgeirsson
Optimal Technologies Inc., Benicia, CA R. Schoettle,
S. Kuloor and T. Mellik
EDD Inc., Blacksburg, VA R. Broadwater and M. Dilek
Nayak Corporation, Princeton, NJ O. Nayak and M.
Griffin
Carnegie Mellon University, Pittsburgh, PA M. Ilic and
M. Prica
21
Contact Information
Principle Investigator:
Dr. Laurentiu Nastac
Concurrent Technologies Corporation,
425 Sixth Avenue, Regional Enterprise Tower
Pittsburgh, PA 15219
Email: nastac@ctc.com
Phone: 412-992-5361

Co - Principle Investigator:
Anupam Thatte
Email: thattea@ctc.com
Phone: 412-992-5376
22
Backup Slides

23
Background:
Assessed Capabilities/Features [1,2]
Basic analysis tools: Load flow, fault analysis, motor
analysis, voltage regulation
More advanced analysis tools: Overcurrent protection,
transient stability, harmonic analysis
Additional features for analysis: GIS, import/export of data,
GUI, help utilities
Optimization tools: Network reconfiguration for loss
minimization via switching, restoration for return supply,
optimal active/reactive power flow, capacitor placement, DG
placement
Equipment/hardware models: Regulators, converters,
motors, batteries, fuel cells, transformers

[1] L. Nastac et al., 8th IASTED International Conference on Power and Energy Systems,
PES2005, October 24-26, 2005, Marina del Rey, CA.
[2] L. Nastac et al., IEEE 37th North American Power Symposium, October 23-25, 2005, Ames, Iowa. 24
Technical Approach: Thevenin Method
Thevenin method to compute short-circuit currents for
unbalanced faults in unbalanced three-phase system [4]
Positive, Z+ and zero, Z0, sequence equivalent system
impedances (in ohms, ) are calculated
2 2
V 3V
Z+= LL Z0 = LL 2Z
(MVA 3 phase )* (MVA 1 phase )* +

VLL nominal line-to-line voltage in kV of distribution system


MVA given short-circuit MVA magnitude and angle

Convert positive and zero sequence impedances into phase


impedance matrix, ZT
Thevenin matrices sum of phase impedance matrices of
each device in between circuit voltage source and fault
location
[4] M. Dilek et al., 2004 Power Systems Conference and Exposition, New York, October 2004. 25

[3] M. Dilek et al., 2004 Power Systems Conference and Exposition, New York , October 2004.
Technical Approach: Thevenin Method

Vanf Ia
Vbnf Ib =0
Post-fault
If Zf
System f I
Vcn c
Model

Phase-to-Phase Fault
f f
I b = 0, I a = I c = I f and V an Vcn = If Zf
Zf arcing impedance

26
Fault Locator Software Integration:
Network Fault Analysis with AEMPFAST
Comparison (% Err) of measured (DTE faulted data) and predicted
results with AEMPFAST

Arcing Pre-fault Err


[Ohm] Loading [%]
[%]
0 0 -11.7
0 50 -6.0
0 100 -1.9
0.5 0 -4.8
0.5 50 0.8
0.5 100 6.0
27
Fault Locator Software Implementation
Strategy
Working with DTE Energy - Several faulted circuits for
integrating, testing and validating the software
Current Implementation: QuickWin Application (Fortran 90)
Final implementation: Lotus Domino Developer v7 Application

28
Fault Report
Fault Report for the Circuit = Orion #2 (1078 components)

A) Fault Current Comparison Using Error [%] = 10

Fault Type No. X Y Status Current


3 8 2293331 466782 100 2526 2291
3 9 2293302 466822 100 2503 2291
3 10 2293297 466878 100 2503 2291
3 11 2293324 466605 100 2463 2291
3 12 2293355 466263 100 2429 2291
3 13 2293880 465094 100 2086 2291
3 16 2293642 465426 100 2205 2291
3 17 2293427 465461 100 2249 2291
3 18 2293414 465286 100 2222 2291
3 20 2293395 465618 100 2291 2291
3 23 2292612 466125 100 2107 2291
3 24 2292640 465894 100 2148 2291
3 25 2292691 465493 100 2146 2291
3 26 2293382 465933 100 2359 2291
3 29 2293373 466145 100 2406 2291
3 34 2294334 466424 100 2324 2291
3 35 2294014 466193 100 2427 2291
3 37 2294192 464546 100 2121 2291
3 39 2294312 463919 100 2041 2291
3 101 2294373 464168 100 2088 2291
3 102 2294150 465150 100 2231 2291
3 106 2294080 465594 100 2305 2291
3 108 2294011 466439 100 2496 2291

Number of possible fault locations = 23 29


Fault Report (Contd.)
B) Recloser Type X Y Current
1 1 2293355 466263 0

Fault Type No. X Y Status Current


3 12 2293355 466263 100 2429 2291
3 13 2293880 465094 100 2086 2291
3 16 2293642 465426 100 2205 2291
3 17 2293427 465461 100 2249 2291
3 18 2293414 465286 100 2222 2291
3 20 2293395 465618 100 2291 2291
3 23 2292612 466125 100 2107 2291
3 24 2292640 465894 100 2148 2291
3 25 2292691 465493 100 2146 2291
3 26 2293382 465933 100 2359 2291
3 29 2293373 466145 100 2406 2291
3 34 2294334 466424 100 2324 2291
3 35 2294014 466193 100 2427 2291
3 37 2294192 464546 100 2121 2291
3 39 2294312 463919 100 2041 2291
3 101 2294373 464168 100 2088 2291
3 102 2294150 465150 100 2231 2291
3 106 2294080 465594 100 2305 2291
3 108 2294011 466439 100 2496 2291

Number of possible fault locations = 19

30
Fault Report (Contd.)
C) Call No. Comp. No. X Y
1 25 2292691 465493
2 25 2292691 465493
3 25 2292691 465493
4 25 2292691 465493
5 25 2292691 465493
6 25 2292691 465493
7 25 2292691 465493
8 25 2292691 465493
9 25 2292691 465493
10 25 2292691 465493
11 25 2292691 465493
12 25 2292691 465493
13 25 2292691 465493

Fault Type No. X Y Status Current


3 12 2293355 466263 100 2429 2291
3 13 2293880 465094 100 2086 2291
3 16 2293642 465426 100 2205 2291
3 17 2293427 465461 100 2249 2291
3 18 2293414 465286 100 2222 2291
3 20 2293395 465618 100 2291 2291
3 23 2292612 466125 100 2107 2291
3 24 2292640 465894 100 2148 2291
3 25 2292691 465493 100 2146 2291

Number of possible fault locations = 9


31

You might also like