High PD value on 13.8KV, 12,5 MW steam turbine generator

Hi Parimalam, welcome.

It has been a few years since I dealt with large motor assessment and repair, and I am surprised that you have not received any posts so far, but I think you would do well to get input from several sources that offer these services professionally.

Partial discharge (PD) when identified at high levels, especially when a continuous monitoring has shown a dramatic increase in levels, is a very good indicator of serious degradation in the insulation or cracks in the bars etc. Sometimes errors in placing the wedges or flaws in the metals can begin to produce these discharge patterns, and usually serious consequences are pending.

Here are a few quick links I found that may help you.

Definition information and a supplier of equipment and services specific to your need

Hope you get what information you need in a timely manner.

Regards, CJ

Parimalam,

A high PD reading does not mean that the generator needs rewinding asap. As with all other methods of testing and diagnostics, PD is not only a science but rather it is also an art.

For example, Equipment A might show a higher PD value compared to Equipment B. However, experience does not always show that Equipment A failed sooner than Equipment B. In fact Equipment A lasted to its normal lifespan as if its PD value is within normal range.

Additionally, was this equipment tested before with PD? If it was tested, what where the results? Compared to the past PD results, is the present PD value too high? Is the rate of increase a cause of alarm?

Now, this is where the experience of the testing engineer comes in. Historical PD data would greatly help in analyzing the present one. A wise testing engineer would not depend on only just one testing method. Were there tests performed on this generator by your specialist? What are these tests? What did the results imply?

Additional test is highly recommended to confirm the PD findings. An FDS (frequency domain spectroscopy) or DFR (dielectric frequency response) if you want the other terminology, is strongly recommended for this situation.

Dear CJMcGill - Power user, myn1996 - Associate, Instruform - Participant,

I thank all of you for taking time and providing guide line, and I am very happy to hear the responses and now I will further explore as per your recommendations and also to look for the test such as FDS (frequency domain spectroscopy) or DFR (dielectric frequency response). can you please advice me the best method for PD measurement, as of now we have measured with online PD measurement only once using toshiba equipment and used internal single winding temperature resistor as a sensing element for this test. Also we found lots of white powder on running turbine generator and this white powder because of PD activity is it not??. will CO2 cleaning on running generator to remove these white powder is necessary, suppose we use CO2 cleaning is there any chance that we may damage insulation further leading to PD??? advice required.

Thanks and best regards

K.Parimalam

By now you have read the excellent information from Siswanto, with great detailed photos. (GA)

I noticed the mention of white powder specifically in connection with grease when loose ties on the leads were found.

My instincts lead me to recognize that this powder will help in the identification of locations affected by the PD, and cleaning may erase the evidence. As far as cleaning, that would depend on the powder properties and whether the powder itself will cause further problems.

I will leave the further discussion about the powder to more experienced individuals with specific knowledge

I have over 40 years of PD testing experience and would support what Myn1996 & CJMcGill recommend. PD limit value standards are set for new machines and have only been set in the last 5 years after some 20 years of trial. For life expectancy you need to ascertain the change from previous results/reports to evaluate your situation. That is why when testing for service life expectancy the inception, extinction voltages and the discharge shape are important readings & observations and to get some consistency the same professional & experienced tester should be used on a regular basis if possible. Unfortunately PD testing is still both a science & a seat of the pants feel. Make preparations by all means, have coils available to limit your outage time, but don't panic you could still have years of service from the machine.

Hi to All,

It is simply great and I never get such valuable information any were but in this forum it is amazing and I very much thankful to all of you. I am going through all the information and suggestion and will take time for me to understand fully. in the mean time I will reply few answers to the questions posted here.

this plant 2 x 12.5MW, 13.8KV steam machines (FSO) running continuously for the past 10 years and now one already failed and also the failure damaged the core laminations and end up with the big job, re-winding and lamination repairs and the job is ongoing at present. My question here - is therey any way we could have prevented core damage due to coil grounding??? in terms of fast switching protection relay etc..????

These machines never checked for PD and now only after failure we have checked the running TA and found PD value is very high (there was no earlier reading for comparisson) . presnelty we are running this TA only with 10% station load, due to fear. My question here, running less load will have less PD???, even though voltage is same?? curious to know some say no change in PD as the voltage is the creteria for PD - I am thinking in other way - by having low load, I reduce the temperature and also vibration of coil some extend reduced can run longer period to cater the station load till we repair another failed TA. need advice.

Can any one can suggest onlind PD meter economical model brand for purchase in order to keep it at site for frequent monitoring

rest in next post.

Dear parimalam

As you know, PD is associated with the high voltage stator conductor bars, PD would be still produced when generator is energize even no load, of course yes, reduced load would be reduced also the acceleration of PD, because you have reduced also the vibration and thermal.

My suggestion, when you scheduled to shut down, do inspect on your winding,

- Inspect the PD by off line method (probe test is recommended, its designed to check and identify which coil bars are have high PD level, so you can repaired, in my opinion the high PD level in your winding is caused by semiconductive deterioration). You can also check the PD activity by using endoscopic (borescope), check the PD via vent holes (air duct) of winding

- Provide also Tangent Delta test, is useful to analysis of your winding

- Stator wedge tightens, loose on this parts will create big problems on your winding, such as:

o Deterioration on semiconductive layer, due to the stator bar coils will movement entire slot

o Loose end winding blocking and ties, its can create endwinding deterioration on insulation, electric field concentration on end of core can occurring when stress grading failure, may also can create delamination on scarf join , scarf join is insulation tape join between overhang to end winding area, this is the weakness insulation join.

o When loose wedges was found, its can also create insulation voids acceleration , when PD is caused by the winding voids, you can't do anything to repair, except rewinding

Note : to repair semiconductive layer, use materials as bellow:

§ Entire slot , use : "corona painting" , i.e. Von Roll Isola product type P 8003 or equivalent.

§ Endwinding, use : "corona painting", i.e. von roll isola product, type P 8001 or equivalent

§ Clean first and dry out before applied these materials

o When you found stator wedges loose, replace and improve the materials, provide with radial springs if the space is possible to accommodate it, for wedge material you can improve by NEMA G11 material or equivalent, and for radial spring improve to class F

- Re tight or replace the endwinding and blocking coil ties by using bending tape class F, also replace if you found loose or missing of blocking coils, the function of Blocking coils are used to determine the clearance between the coil side at the end region, to eliminate mechanical stresses arising coil movement, that will be resulting coil touching each other at the end-winding, Blocking coils are almost made from solid insulation materials (such as textolite, felt, etc) which is soaked in resin.Blocking coil is always held in place with ties. Often solid separator tend to loose after long period of operation, in severe cases it will fall from the winding.

- Inspect all ties on the surge ring support, nd-winding and surge-rings are secured with ties, as with the blocking coils ties, when exposed over a long time operation, ties tend to "dry" , become loose due to lose volume, brittle, broken, etc re tight or replace it by using bending type and soaked with resin

- Inspect tightens of support structures, found sludge or greasing or white powder is indicate loose, retight.

- Inspect the dust under winding close to core, extract it and check it, found metal dust on it, indicate stator core loose, check the core looseness, re tight the bolts support based on manufacturer specification, found bent, broken core teeth, repair it and applied epoxy / resin or RTV on its to prevent or minimize the core vibration.

- Inspect the core by using Digital EL CID method, found hot spot, repaired as soon, its must be done when you perform rewinding, recorded the result as based line, do regularly in 5 years.

Core repair and test, see the following link .

http://cr4.globalspec.com/thread/32530/How-to-Repair-a-Partial-Core-Stator-Generator

PD Instruments.

You can browsing in the internet, its many supplier can provide, or our colleagues here may can help you to provide the technical specification and supplier for the instruments.

PD instrument for off line test , PD Probe may the cheapest , and its useful to quick identified of PD location,

Good Luck.