当前位置:首页 >> 其它语言学习 >>

2011 Transformer PD monitoring case study


Transformer PD Monitoring Case Study
Richard Heywood Matthias Boltze Sacha Markalous Carl Johnstone

Doble PowerTest Ltd, UK bl d

Introduction Because of … ? Increasing age of in-service power transformers f f ? Increasing need for utilities to maximise use of transformers while maintaining system reliability. reliability ? Increasing forced outage costs Condition monitoring and life cycle management of power transformers have received considerable attention, with many new monitoring techniques available, especially it i t h i il bl i ll for partial discharge fault detection and localization. Difficult to decide when to introduce monitoring on a power transformer and how to use monitoring to assist in asset management, particularly for suspect units. g ,p y p
2

Background This story started in late 2008… ?400/275/13kV 750 MVA auto-transformer (no tap changer) p ?Important location in London network ?28 years service ?LV cable through-fault on the attached circuit – may have g y contributed to the failure of a Quad Booster at the other end of the 275kV circuit ?Online PD monitoring system on 400 kV GIS system detected an increase in PD activity on the gas zone connected to this transformer. transformer Concern about the condition of the transformer and the reliability of the circuit. circuit
3

The transformer

4

Step 1: Enhanced DGA sampling DGA i invented i the UK t d in th Over the years many power transformers have been saved by effective DGA analysis therefore Enhanced oil sampling was first decided to check if there as was any indication of possibly developing fault inside the transformer. transformer

5

Step 1: Enhanced DGA sampling
Significant increase in hydrogen and acetylene in main tank 60

H2
50

C2H2

Gas concentratio (ppm) on

40

30

20

10

0 12/06/2007 08/11/2007 Sample date 21/01/2009 06/02/2009

Significant increase in acetylene to 26ppm from trace, and the resample not only confirmed the high level of acetylene but also showed a further increase to 30ppm within two weeks. Such a DGA signature is typical of an arcing/sparking fault in the main tank of a power transformer. It was then decided to carry out PD measurements to investigate further. y g
6

Step 2:GIS PD monitoring

PD activity appeared to be on the Yellow phase of the transformer. p
7

Step 2: Neutral RF CT measurements

8

Step 3: Review of design, history and DGA

AE PD location in a sister transformer in December 1990
9

Step 3: Review of design, history and DGA

Photo from 2001 inspection showing carbon p g marks around loose clamping screws
10

Gas concent tration (ppm)
100 150 200 250 50 0

24/01/1985 26/06/1985 27/11/1985 18/03/1986 24/09/1986 29/12/1986 20/02/1989 06/12/1990 25/10/1991 20/07/1995 21/04/1997 02/12/1999 04/03/2001 26/03/2001 17/04/2001 16/11/2001 13/06/2002 19/07/2004 12/08/2004 22/09/2004 15/02/2005
Sample date

H2

Step 3: Review of design, history and DGA

31/01/2006 31/05/2007 21/01/2009

History of diagnostic hydrogen and acetylene in main tank of transformer T6607

C2H2

19/03/2009 17/04/2009 11/06/2009

11

Step 3: Review of design, history and DGA
History of combustible gases in main tank of transformer T6607
100 90 80 70 60 50 40 30 20 10 0 13/04/2000 11/01/2001 09/04/2001 02/10/2001 21/01/2009 06/02/2009 04/03/2009

%CO

%H2

%CH4

%C2H4

%C2H6

%C2H2

Conclusion was that most likely some of the clamping bolts were loose again after they had been retightened in 2001.
12

Step 4: Winding movement assessment

Clearly more in-depth investigation required. Outage scheduled for off line testing and internal off-line inspection. In view of the recent 275 kV through-fault it was decided to check mechanical condition by carrying out SFRA measurements before any remedial work carried out to retighten clamping bolts.

13

Step 4: SFRA test
10 0 -10 -20 -30 -40
-20 0 -10

HV to LV

A m p litu d e in d B

LV to N

-50 -60 -70 -80 0 50000 100000 150000

-30 30 A m p litu d e in d B -40 -50 -60 -70 -80

200000

Frequency in Hz

No evidence of winding g movement, so can re-clamp.

-90 -100 0 50000 100000 150000

2

Frequency in Hz

14

Step 5: Internal inspection and repair p p p

Photo from internal inspection on the critical transformer in 2007, 2007 B (Yellow) phase HV side clamping bolts
15

Step 6: On-line UHF PD measurement
In order to confirm the effectiveness of the repair, it was then decided to use on-line UHF PD measurements when transformer back to service to ensure it was discharge free.

Doble Lemke PD Measuring System LDS-6/UHF

16

Step 6: On-line UHF PD measurements

UHF drain valve sensor (left) and LDS-6/UHF measuring system (right)
17

UHF PD probe measurements

Only low-level PD activity measured !
18

Postscript:

After return to service following repair the acetylene level in main tank fell to trace (l i t k f ll t t (less than 3 ppm), confirming the th ) fi i th success of re-clamping. Nevertheless, it was decided that the transformer should remain at Condition Category 4 (normal) in the Transformer Asset Health Review, but with periodic on-line GIS PD monitoring and regular routine oil DGA sampling to monitor its condition – and RFI surveys. surveys

19

On-line UHF RFI surveys

20

PDS100
? Scan the site in seconds ? D t t PD at an early stage Detect t l t ? Record PD activity

Non-invasive method Hand held, battery operated, easy to use RF CT and TEV sensors now available

- Doble PDS100 -

21

PD-Smart
Input Unit vPD
PD-Signal g
A D

DAQ

DSP
Filtering Detection

PC

Postprocessing

vU
Voltage Signal
A D

?P

Computer Software User/ Panel

22

PD alternatives ?‘Conventional’ IEC 60270 ?Acoustic i ?UHF ?RFI
Requires direct electrical coupling

Effective for location of ‘shallow’ faults

Most sensitive

Most convenient, very effective
23

Conclusions and Recommendations
?Step-by-step approach use of monitoring/diagnostics recommended to assist management of suspect units ?Monitoring should target suspected fault Monitoring ?Knowledge of design weaknesses and likely failure modes Knowledge invaluable in directing investigation ?Objective is proactive use of monitoring/diagnostics to prevent unplanned outage or catastrophic failure.

24


相关文章:
更多相关标签: