Detection of Incipient Faults Using Waveform Analytics

Panel SessionIEEE PES General Meeting

Washington, DC, 2014

Normal Operation Broken

Electrical Feeder Operational Paradigms

Major Event- Outage- Line Down- Fire

Time

TraditionalThinking

Reality Normal Operation Broken

Incipient Failure Period(hours, days, weeks)

Detecting incipient failures offers a paradigm shift in the operation of distribution systems.

Restoration

Restoration

2

Current Ranges of Various EventsCu

rrent

Ran

ge(L

ogar

ithm

ic)

FailingClamp

Failing Line Switch

ArcingSvc Cable

HiZ Fault

SLG Fault

P-P Fault

Conventional Triggering Threshold

Missed Opportunities

Power Line Incipient Failure DetectionFundamental Principle

• Graph shows current during “normal” feeder operations.…what is it? …what is happening (if anything)?

• Analytics report this specifically as a failing clamp. Failing clamps can degrade service quality, drop hot metal particles, which can start fires, and in extreme cases burn down lines.

• Conventional technologies do not detect incipient failures such as this one.

On-LineWaveformAnalytics

Analysis of Low Magnitude Failure Signatures Using DFA Analytics

IEEE PES General MeetingWashington, DC, 2014

Jeffery A. Wischkaemper, Ph.D. Research Engineering Associate II

Texas A&M UniversityCollege Station, TX 77843-3128979-575-7213, jeffw@tamu.edu

mailto:jeffw@tamu.edu

DFA Foundational Research• A Decade of Field Research

– Instrumented dozens of circuits at 10+ utilities– Created largest database of failure signatures in existence– Analyzed waveform anomalies and correlated with failure events– Discovered unique signatures for specific failures– Developed automated reporting to deliver actionable information

• Self-Imposed Constraints– Conventional sensors– Substation equipment only; distributed electronics not required

• Result: Improved power system reliability, operational efficiency, and safety enabled by advanced monitoring of electrical signals

Monitoring Topology

SubstationTransformer

FailingApparatus

High-fidelity DFA devices, connected to conventional CTs and PTs, one per circuit.

Line recloser*tripped 8% ofphase-A load twice,but reclosed and didnot cause outage

Breaker lockout caused by fault-induced conductor slap

Inputs: Substation CT and PT Waveforms

*Analytics applied to high-fidelity substation waveforms report on hydraulic line reclosers, switched line capacitors, apparatus failures, etc, without requiring communications to line devices.

OutputsWaveform Analytics

Failing hot-lineclamp on phase B*

Failed 1200 kVARline capacitor*(phase B inoperable)

On-Line Signal Processing and

Pattern Recognition

Analytics

(Performed by DFA Device inSubstation)

Waveform-Based Analytics – Behind the Scenes

Documented Failures•Voltage regulator failure• LTC controller maloperation•Repetitive overcurrent faults• Lightning arrestor failures• Switch and clamp failures•Cable failures

– Main substation cable– URD primary cables– URD secondary cables– Overhead secondary cables

• Tree/vegetation contacts– Contacts with primary– Contacts with secondary services

• Pole-top xfmr bushing failure• Pole-top xfmr winding failure•URD padmount xfmr failure• Bus capacitor bushing failure•Capacitor problems

– Controller maloperation– Failed capacitor cans– Blown fuses– Switch restrike– Switch sticking– Switch burn-ups– Switch bounce– Pack failure

• A DFA participant utility uses one-way paging to switch line capacitors. After each page, the system monitors substation VARs to verify 1) that the bank has switched and 2) that it is balanced.

• On 11/29/2013, a DFA device began detecting unusual transients suggesting pre-failure of a capacitor bank.

• Trouble tickets indicated no problem. Researchers and the utility continued to monitor the situation.

• The transient occurred 500 times over the next 2-1/2 months. • After 2-1/2 months, increasing event activity suggested the problem

might be accelerating toward failure, prompting corrective action.

Capacitor Vacuum Switch Pre-Failure

Theory and Analysis• Normal capacitor switching causes two phenomena.

– A short-lived high-frequency transient– A step change in voltage (yes, even at the bus!)

• Each subject event caused a transient, but no step change.– This indicates the events were not during switching.

• Each event caused a high-frequency spike in current and voltage.

– The current and voltage spikes had the same polarity (i.e., when voltage spiked up, current spiked up).

– This indicated a “reverse” event. For “forward” events, voltage and current spikes have opposing polarities.

– From the DFA’s perspective, a “reverse” event is one occurring on a different circuit or on the bus itself.

Ten Seconds of RMS Bus Voltages(no steady-state change)

Bus Voltage andFeeder Current

Capacitor Vacuum Switch Pre-Failure (cont’d)

• Graph shows the number of events on each day (11/29/2013 – 2/12/2014)

• There is no definitive trend.

• “Peaks” weakly suggest a slight increase in activity over time.

Capacitor Vacuum Switch Pre-Failure (cont’d)

Statistical Analysis #1: Number of transient eventsrecorded per day, during 75-day period.

• Graph shows the frequency of events as a function of time of day, cumulatively for the 75-day period.

• Events occur at all times of day but most frequently during the middle of the day.

• 64% occur during 25% of day.(319 of 502 events between 10:00 and 16:00)

• 47% occur during 17% of day.(238 of 502 events between 11:00 and 15:00)

Capacitor Vacuum Switch Pre-Failure (cont’d)

Statistical Analysis #2: Number of transient eventsrecorded as a function of time of day.

• On February 14, the utility decided to open the fuses of all five of the adjacent circuit’s banks to confirm the “pre-failure capacitor bank” diagnosis.

• A crew found an anomaly at the first bank.– The bank’s “closed” current should be 30 amps.– The paging system showed the bank as “open.”– A hot-stick meter showed 0.7 amps through the

“open” switch.– First bank’s fuses were pulled.– Other four banks were left in service.– DFA system was watched for five days. Absence of

additional transients validated diagnosis.• The pre-failure switch failed a hi-pot test. A

vacuum interrupter expert performed a root cause analysis.

Capacitor Vacuum Switch Pre-Failure (cont’d)

Magnetron-Based Vacuum

Tester

Capacitor Vacuum Switch Pre-Failure (cont’d)

Red/Green Indicatoron Normal Phase

Red/Green Indicatoron Pre-Failure Phase

The switch has a sight window with a red/green position indicator. The pre-failure switch’s indicator had clear signs of rubbing against the mechanism housing.

Capacitor Vacuum Switch Pre-Failure (cont’d)

Red/Green Position Indicator Mechanism on Normal Phase

Red/Green Position Indicator Mechanism on Pre-Failure Phase

NormalBent Indicator

Rubbing Housing

The root cause of the pre-failure was the indicator rubbing and binding, preventing the switch contacts from parting fully. Current (0.7A) flowing through the gap is believed to have caused progressive internal damage to the vacuum interrupter.

Summary and Conclusions• Vacuum switch failures have multiple

potential consequences.– Least severe: unbalanced capacitor operation– Most severe: rupture of switch or capacitor

• This pre-failure example persisted 2-1/2 months before intervention.

– Normal operational practices did not and do not detect this kind of pre-failure.

– Pre-failure detection provided time (in this case, 2-1/2 months) to correct the condition and preempt full failure.

• Waveform analysis provided only notice and opportunity to correct proactively.

Capacitor Vacuum Switch Pre-Failure (cont’d)

Magnetron-Based Vacuum

Tester

Failing-Clamp Alarms from DFA Analytics(2,333 Episodes over 21-Day Period)

Time

Hard-to-Diagnose Trouble• Customers on a lateral experienced service trouble (e.g., lights out, flicker) four times in

a 40-hour period.• This “cost” the utility four complaints, four truck rolls, and two transformer replacements

– all on overtime and mostly unnecessary.• DFA analytics detected and reported the cause (“failing clamp”) weeks before the first

customer complaint. Crews were unaware of the DFA report, however, so their response was conventional.

Hard-to-Diagnose Trouble (cont’d)

Electrical variations caused by the clamp failure were minor, but on-line analytics diagnosed them properly. A crew knowing to look for a clamp failure can respond more efficiently and

effectively and fix the right problem the first time.

Hard-to-Diagnose Trouble (cont’d)

Summary• DFA technology applies sophisticated waveform analytics

to high-fidelity CT and PT waveforms, to provide heightened visibility, or awareness, of circuit conditions. This enables improved reliability, operational efficiency, and safety.

• The DFA system automates the analytics process, so as to deliver actionable intelligence, not just data.

• DFA is a data-driven technology that embodies multiple functions.

• Utility companies have used DFA to demonstrate the avoidance of outages and improvements in operational efficiency.

Detection of Incipient Faults Using Waveform AnalyticsElectrical Feeder Operational ParadigmsCurrent Ranges of Various EventsPower Line Incipient Failure Detection�Fundamental PrincipleAnalysis of Low Magnitude Failure Signatures Using DFA AnalyticsDFA Foundational ResearchMonitoring TopologyWaveform-Based Analytics – Behind the ScenesDocumented Failures�Capacitor Vacuum Switch Pre-Failure�Capacitor Vacuum Switch Pre-Failure (cont’d)�Capacitor Vacuum Switch Pre-Failure (cont’d)�Capacitor Vacuum Switch Pre-Failure (cont’d)Capacitor Vacuum Switch Pre-Failure (cont’d)Capacitor Vacuum Switch Pre-Failure (cont’d)Capacitor Vacuum Switch Pre-Failure (cont’d)Capacitor Vacuum Switch Pre-Failure (cont’d)Hard-to-Diagnose TroubleHard-to-Diagnose Trouble (cont’d)Hard-to-Diagnose Trouble (cont’d)Summary

/ColorImageDict > /JPEG2000ColorACSImageDict > /JPEG2000ColorImageDict > /AntiAliasGrayImages false /CropGrayImages true /GrayImageMinResolution 150 /GrayImageMinResolutionPolicy /OK /DownsampleGrayImages true /GrayImageDownsampleType /Bicubic /GrayImageResolution 300 /GrayImageDepth -1 /GrayImageMinDownsampleDepth 2 /GrayImageDownsampleThreshold 1.50000 /EncodeGrayImages true /GrayImageFilter /DCTEncode /AutoFilterGrayImages true /GrayImageAutoFilterStrategy /JPEG /GrayACSImageDict > /GrayImageDict > /JPEG2000GrayACSImageDict > /JPEG2000GrayImageDict > /AntiAliasMonoImages false /CropMonoImages true /MonoImageMinResolution 1200 /MonoImageMinResolutionPolicy /OK /DownsampleMonoImages true /MonoImageDownsampleType /Bicubic /MonoImageResolution 1200 /MonoImageDepth -1 /MonoImageDownsampleThreshold 1.50000 /EncodeMonoImages true /MonoImageFilter /CCITTFaxEncode /MonoImageDict > /AllowPSXObjects false /CheckCompliance [ /None ] /PDFX1aCheck false /PDFX3Check false /PDFXCompliantPDFOnly false /PDFXNoTrimBoxError true /PDFXTrimBoxToMediaBoxOffset [ 0.00000 0.00000 0.00000 0.00000 ] /PDFXSetBleedBoxToMediaBox true /PDFXBleedBoxToTrimBoxOffset [ 0.00000 0.00000 0.00000 0.00000 ] /PDFXOutputIntentProfile (None) /PDFXOutputConditionIdentifier () /PDFXOutputCondition () /PDFXRegistryName () /PDFXTrapped /False

/CreateJDFFile false /Description >>> setdistillerparams> setpagedevice