Withstand Tests More than Meets the Eye Nigel Hampton, Joshua - - PowerPoint PPT Presentation

Fall 2008 ICC Meeting - Subcommittee F: Field Testing and Diagnostics

1

Withstand Tests More than Meets the Eye

Nigel Hampton, Joshua Perkel, JC Hernandez, Miroslav Begovic, John Hans, Ron Riley, Pete Tyschenko, Christian Linte, Frank Doherty, George Murray, and Leeman Hong

NEETRAC Georgia Institute of Technology, ECE Commonwealth Edison PEPCO Consolidated Edison

Fall 2008 ICC Meeting - Subcommittee F: Field Testing and Diagnostics

2

Acknowledgements

• Thanks to CDFI Participants for the data presented here.
• This material is based upon work supported by the

Department of Energy under Award No DE-FC02- 04CH1237 and CDFI.

• Any opinions, findings, conclusions or recommendations

expressed in this material are those of the author(s) and do not necessarily reflect the views of the Department of Energy.

Fall 2008 ICC Meeting - Subcommittee F: Field Testing and Diagnostics

3

Outline

• Introduction/Motivation
• Length Issues
• Withstand Testing Process

– Ramp Up – Hold

• “Ramp Up” Diagnostic Features
• “Hold” Diagnostic Features
• Recommendations
• Conclusions

Fall 2008 ICC Meeting - Subcommittee F: Field Testing and Diagnostics

4

Introduction

• Withstand tests are frequently used by utilities who

employ diagnostic tests. – As of 2006, approx. 33% of CDFI member utilities employing diagnostic tests use withstand techniques.

• Withstand tests have been defined as “Pass/Fail” only.
• Utilities maintain records that are much more detailed.
• Cable Tested in the last five years: > 4495 miles.

Fall 2008 ICC Meeting - Subcommittee F: Field Testing and Diagnostics

5

Motivation

Utility records from withstand tests contain much more information than the result of the test. – Test Voltage (including voltage at failure) – Time on Test – Segment Length – Segment Insulation – Segment Location – Failed Equipment Type

How to use this information in a diagnostic manner?

Fall 2008 ICC Meeting - Subcommittee F: Field Testing and Diagnostics

6

Results of Withstand Tests

Retests Percentage

• f Tests

[%] 1 2 3 100

618 Conductor Miles Tested (one utility feeder cable system) No evidence of cascading failures for IEEE 400.2 testing practices.

Fall 2008 ICC Meeting - Subcommittee F: Field Testing and Diagnostics

7

Length Effects

• Withstand tests can be used to test long lengths of cable.
• Comparison of withstand failure on test rates must include

length adjustments.

2000 ft. 500 ft. 500 ft. 500 ft. 500 ft.

Censored Failure

Fall 2008 ICC Meeting - Subcommittee F: Field Testing and Diagnostics

8

Withstand Test Process

HOLD RAMP UP Time Voltage t = 0 tTest

Fall 2008 ICC Meeting - Subcommittee F: Field Testing and Diagnostics

9

Withstand Test Process

HOLD RAMP UP Time Voltage t = 0 tTest

Fall 2008 ICC Meeting - Subcommittee F: Field Testing and Diagnostics

10

Why “Hold” and “Ramp Up” Phases?

Focus is generally on the “Hold” phase but “Ramp Up” is important too.

Time on Test [Minutes] Survivors [% of Total Tested]

5 4 3 2 1 100 80 60 40 20

Extruded 16 PILC 16 PILC 22 XLPE 22 Insulation [kV] Voltage

Large percentage of failures

• ccur during Ramp Up portion.

Fall 2008 ICC Meeting - Subcommittee F: Field Testing and Diagnostics

11

“Ramp Up” Phase Diagnostic Features

Getting up to test voltage is half the battle

Fall 2008 ICC Meeting - Subcommittee F: Field Testing and Diagnostics

12

“Ramp Up” Data

Voltage Time HOLD RAMP UP Vf tTest

Fall 2008 ICC Meeting - Subcommittee F: Field Testing and Diagnostics

13

Ramp Up Failures

• Failures during the “Ramp Up” phase have accounted for

as much as 70% of the total failures on test.

• The failure mechanism during “Ramp Up” phase is

different from the “Hold” phase mechanism.

Fall 2008 ICC Meeting - Subcommittee F: Field Testing and Diagnostics

14

“Ramp Up” and “Hold” Failure Mechanisms

Time on Test [Minutes] Failures on Test [% of Total Tested]

100.0 50.0 10.0 5.0 1.0 0.5 0.1 20 15 10 5 1

“Ramp Up” Failures “Hold” Failures Two different failure mechanisms

Fall 2008 ICC Meeting - Subcommittee F: Field Testing and Diagnostics

15

Voltage [kV] Failures on Test [% of Total Tested]

50 40 30 20 10 5 1 10 5 3 2 1 IEEE 400.2 Recommended Test Voltage

Weibull Curve – “Ramp Up” Failures (VLF)

Evidence of two failure modes.

Fall 2008 ICC Meeting - Subcommittee F: Field Testing and Diagnostics

16

Weibull Curve – “Ramp Up” Failures (DC)

Voltage [U0] Failures on Test [% of Total Test]

10.0 1.0 0.1 80 70 60 50 40 30 20 10

In this case, 60% of the tests produced a failure before reaching the target test voltage. Two failure modes present during “Ramp Up” portion that are voltage dependent.

Fall 2008 ICC Meeting - Subcommittee F: Field Testing and Diagnostics

17

“Hold” Phase Diagnostic Features

Time is everything

Fall 2008 ICC Meeting - Subcommittee F: Field Testing and Diagnostics

18

“Hold” Data – Failure During Test

Voltage Time HOLD RAMP UP

tf

Time on Test tTest

Fall 2008 ICC Meeting - Subcommittee F: Field Testing and Diagnostics

19

“Hold” Data – Test Passes

Voltage Time HOLD RAMP UP

tf

Test Passed

tTest

Fall 2008 ICC Meeting - Subcommittee F: Field Testing and Diagnostics

20

Analysis of Times on Test

PASS Segments

Censored Times on Test

NOT PASS Segments

Failure Times Use data to construct Weibull curves for different areas These curves represent a diagnostic feature.

Fall 2008 ICC Meeting - Subcommittee F: Field Testing and Diagnostics

21

Time on Test [Minutes] Failures on Test [% of Total Tests]

100.0 10.0 1.0 0.1 20 10 5 3 2 1

1000 Feet 500 Feet NONE Adjustment Length

Length Effect on Failures on Test

Time on Test [Minutes] Failures on Test [% of Total Tests]

100.0 10.0 1.0 0.1 20 10 5 3 2 1

1000 Feet 500 Feet NONE Adjustment Length

Time on Test [Minutes] Failures on Test [% of Total Tests]

100.0 10.0 1.0 0.1 20 10 5 3 2 1 4.1% 2.4% 17.2% 30

1000 Feet 500 Feet NONE Adjustment Length

≈ 900 Conductor Miles

Fall 2008 ICC Meeting - Subcommittee F: Field Testing and Diagnostics

22

System Subset (Length Adjusted)

Time on Test [Minutes] Failures on Test [% of Total Tested]

50.0 10.0 5.0 1.0 0.5 30 20 10 5 3 2 1 4.6% 13.5% 1 30

4.6% of tests led to failure during “Ramp Up.”

Curve includes tests from four areas

• f a single utility.

Fall 2008 ICC Meeting - Subcommittee F: Field Testing and Diagnostics

23

Weibull Curves by Area (Length Adjusted)

Time on Test [Minutes] Failures on Test [% of Total Tested]

100 10 1 40 30 20 10 10.0% 22.9% 17.8% 29.0% 30

1 2 3 4 Area

Fall 2008 ICC Meeting - Subcommittee F: Field Testing and Diagnostics

24

Area Failures on Test [% of Tested]

4 3 2 1 Overall 35 30 25 20 15 10

Separation of Regions

Area 1 is clearly different from the others.

Fall 2008 ICC Meeting - Subcommittee F: Field Testing and Diagnostics

25

Recommendations

• Defined “Ramp Up” procedure should be employed with

each test.

• Detailed records should be maintained.

Fall 2008 ICC Meeting - Subcommittee F: Field Testing and Diagnostics

26

Conclusions

• Withstand tests consist of two phases:

– “Ramp Up” – “Hold”

• A significant percentage of failures occur during “Ramp

Up” phase.

• Useful diagnostic features can be derived from withstand

data. – Voltage at failure (“Ramp Up”) – Time on test at final test voltage (“Hold”)

• Results can be used to prioritize areas for action.