The group met on Monday, March 13, 2006 at 2:00 p.m. with 36 members and guests in attendance.

Minutes:

Apologies were received from R. Schlesinger, vice chair of C18D, who was not able to attend the meeting.

Correction to minutes of 2005 Fall meeting.  The failure statistics from Mr. Moh from Malaysia should have read as follows:

                67% failed within 12 minutes

                21.5% failed in13-30 minutes

                5.9% failed within 31-45 minutes

                4.9% failed within 46-60 minutes

The times in the 2005 Fall minutes were stated as months and should have been minutes.  The test voltage was 2.2 Uo and test duration of 1 hour.

There were three scheduled presentations followed by a general discussion.

John Hans made a presentation entitled "VLF Testing at Exelon"; the presentation is included as Appendix C-11.  Data were presented of VLF tests on PILC and extruded cables that had been repaired after an in-service failure.  Testing began in 2003 using a sinusoidal voltage waveform.  Two test criteria were used on 15 kV rated cables, 15 minutes at 22 kVrms prior to November 2004, and 30 minutes at 16 kVrms since November 2004 (the latter conditions are specified in IEEE 400.2).  The data allowed an evaluation and comparison of the two test conditions.  A larger number of extruded than PILC cables failed during the actual VLF test.  The percentages of extruded cable failures were similar for each test condition whereas there were a larger percentage of PILC cable failures during the 15-minute at 22 kVrms tests.  Of interest is the number of failures occurring in service after the VLF testing and when they occurred.  Some of the cables tested at 22 kVrms have been service for 850 days while those tested at 16 kVrms have only been in service for a maximum of 500 days.  The cables tested at 22 kVrms showed an almost random failure rate over the 850 days while the number of failures decrease with time for the cables tested at 16 kVrms, which could be interpreted as damage being caused during the test but the voltage and/or time was not sufficient to complete the breakdown process.  During the discussion, comments were raised about the times of the failures during the test and also a comparison of the failures rates of the tested cables and cables not subjected to VLF tests.  John Hans responded that it should be possible to get this information.

Kent Brown made a presentation entitled "VLF Testing at TVA Nuclear- Spring 2006 Update".  The presentation, included as Appendix C-12, is an update of data presented at the 2005 Spring meeting.  The previous results on cables installed in the 1970s were briefly reviewed.  The tests, using a sinusoidal waveform, were tan delta (at voltages up to 2Uo) and withstand (using the levels in IEEE 400.2).  The tan delta tests reduced the scope of planned replacements by 33% and there were 2 failures during the withstand tests.  Since the testing program was initiated there have been no in-service failures.  The testing in 2005 has been carried out on XLPE and black and pink EPR cables using different VLF equipment.  The new tan delta measuring equipment is simple to install and uses wireless technology to transmit data to a laptop computer.  Six cables failed the tan delta test and one cable failed the withstand test.  There have been no failures of the tested cables that have been put back into service.

Craig Goodwin made a presentation entitled "A Method of Optimizing the Testing Time required to perform a VLF Cable Test".  The presentation is included as Appendix C-13.  The test durations specified in IEEE range from 15 minutes to 60 minutes with a recommended duration of 30 minutes.  Although testing at shorter durations may not identify all the defects that could cause early failure, it is commonly used as more cables can be tested.  A procedure was described to optimize the test duration without compromising cable performance and reliability.  This is achieved by monitoring the tan delta during the withstand test.  If the magnitude of the tan delta is low and does not show significant variation with time, the cable is in good condition and the test duration can be reduced to 15 minutes.  If the reading of tan delta is high and/or shows significant variation with time, the test duration can be extended to the full 60 minutes to determine if breakdown is imminent.  Such a test method should reduce the risk of early failures when a cable is returned to service.

The chair has been given a list of contacts at several utilities to obtain VLF test data.

The group met on Tuesday, October 31 from 8:00 AM to 10:15 AM with 38 persons attending.

Minutes:

There were two presentations. 

Dr. N. Hampton (NEETRAC) presented "Reflections on "First Experiences" with IEEE 400", attached as Appendix C-15.  The author raised several issues regarding IEEE 400-2001 (IEEE Guide for Field Testing and Evaluation of the Insulation of Shielded Power Cable Systems) and IEEE 400.2  (IEEE Guide for Field Testing of Shielded Power Cable Systems Using Very Low Frequency (VLF)). The issues were discussed and will be addressed in the revisions of the two documents.  The revisions will ensure that there will be no contradictions in the two documents with regards to VLF testing.  The revision of the VLF document will be more specific about test duration of the acceptance and maintenance tests.  The author pointed out that testing for too short a duration will not find all the weak cables but testing for too long can cause additional deterioration. Other points raised were: test durations at different frequencies, and tan delta criteria for XLPE cables and cables with other insulation systems.

K. Brown (TVA) gave an update of the "VLF testing at TVA Nuclear", attached as Appendix C-16.  The tan delta and withstand protocols used were according to the guides mentioned above. The 30-minute acceptance test was performed on newly installed cable and aged cables were subjected to a 15-minute maintenance test.  The author emphasized the need for proper preparation of the test circuits when making tan delta tests otherwise misleading data may result.  Many retests had to made early in the program but the number has been reduced to 2% in the latest series of tests.  There were no failures during 51 additional withstand tests made since the Spring.  There have been no in-service failures of circuits tested since the program started in 2003.

B. Lanz (IMCORP) mentioned that IEEE 400-2001 is due for renewal asked if anyone has comments on the Guide he or she should send them to W. Thue, Chair of C15W.  It is planned to have a meeting during the Spring ICC.  It was recommended that much of the information on VLF testing, particularly test values and diagnostic criteria, be transferred to IEEE 400.2.

Other points raised at the meeting were:

1.  Acceptance and maintenance test voltages for 45 kV and 69 kV rated cable systems.  For 25 kV and 35 kV systems the peak acceptance voltage is about 3 U_orms.  Using this criterion the acceptance voltages for 45 kV and 69 kV systems would be 78 kV_peak and 120 kV_peak respectively.  Installation and maintenance voltage levels would be about 89% and 75% of these values.  One utility is using 120 kV_peak as a test voltage on its 69 kV system.  More data is needed.

2.  More test data from utilities needs to be presented to determine the optimum test voltages and durations.  J. Hans from Exelon (ComEd) and R. Patterson (NETA) will present data at the Spring meeting.

3.  More data on EPR cables is needed (C. Katz to investigate).