Tranformer assessment and Life extension 3

[AKassem]

We have a 25 MVA - 66/11KV ONAF transformer which has been working for more than 20 years
Now we have to operate the transformer at around 14 MVA ( 56% )only to avoid over temp. trips
My question here is :

1) How to make an assessment for the current status of the transformer and estimate the normal derating factor
at the same ageing conditions ?
2) What are the tests needed for this assessment ?
3) Is there any maintenance that could be done to improve the performance of the transformer ?

 

[ppedUK]

I think your problem might be more of a cooling issue rather than age! Check to see if the cooling fans work. Are the radiators blocked? What is the condition of the oil? Is the temp trip contact set too low? Is the air ventilation system impeded? 20 years is not old for an oil filled transformer that has been looked after!

 

[collies99]

I think the life of the transformer is based on the strength of the paper insulation and and as such an assessment of the Kraft paper in integral to its total life and usefulness. 20 years is teenage class to middle age.

you can do a furalahyde ? test content in the oil. It tell you the breakdown products if Kraft paper. Ultimately you need to obtain a paper sample of the winding, about 1/2 x 1 inch dimension for a tensile strength test. The end of the usefulness of the paper/oil insulation system is when the tensile strength is at around 50%.

that's if my memory still works.

 

[jkristinn]

I agree with ppedUK that this seems to be a cooling problem. I would suspect an oil flow problem rather than air flow problem, since only 56% of the rating can be reached (unless the ambient temperature is above rated). The incorrect working of the temperature alarms is also a possibility.
My suggestion is to make thermal images of the transformer during high loading (e.g. 56% of rated).

If the unit has not been operated at higher temperatures than normal and the dehydrating breathers have been working properly (assuming a conservator type transformer), the aging should not be a problem. A complete analysis of an oil sample (including analysis of moisture, dissolved gases and furan content) could confirm this.

 

[Millbo]

I also agree with ppedUK and jkristinn

How sudden did this condition come on? Are loads and temperatures logged on a regular basis? These values will allow you to plot out any trends.

One suggestion I have it to look at your cooling system thru an IR camera while the unit is loaded and check to see if the cooling system is doing its job.

DGA can give some key indications to what's going on inside the unit. I'd also test the oils properties such as viscosity, which can effect the oils ability to transfer heat.

I've even see where due to oil sampling over time the oil level fell below the top collector pipes of the radiators and could no longer circulate thru them.
This was on a N2 blanketed unit of course.

Let us know your findings.

 

[cmw1971]

If the transformer is tripping on hi temp at 56% rated load, there is something definitely going on. Everyone above makes some good suggestions.

Suggestions:

1) Verify liquid level. Is the level too low? Transformer have leaks?
2) Check liquid temp gauge/winding temp gauge to make sure they are not false trips.
3) Verify incoming voltage has not changed and is now coming in too high. Over exciting the core can cause overheating.
4) If there are pumps, are they operating properly?
5) Oil testing. Moisture, DGA, Furans. If it has not been done, then get a full oil screen done.

 

[squeeky]

Look at the IEC 60076. Try part 7. Loading guide for oil immersed transformers.

Overheating at low loads could be losses from
1. harmonics. Are you running electronic drives or arc furnaces?
2. Circulating earth currents. Do you have multiple transformers with star connections and all with star points earthed. There should only be one reference point. Use the largest transformer. Primary side. If you have an NEC or an NECR on the common bus and this transformer is star connected, remove the star point to earth.
3. Blocked cooling circuit as above.
4. Shorted winding. Get DGA done on the oil. Dissolved gas analysis. The fault will cause heating. The heat will cause the oil to "crack". Different temperatures cause different gasses to be released. The type of fault can be indicated by the gasses.
5. Insulation failure may cause point 4 above. This can be seen in an oil test for Furan. The amount of Furan will give an indication of life left. Warning. If you have had the oil filtered, you will have removed some of the Furan so the result may not be accurate.
6. Contaminated oil. Open a drain valve (if oil level OK) and see what comes out. Oil, you are OK. Milky oil, you have water, problems. Water, you have big problems.
7. Use an infra red thermometer and look at the temperature gradient from top to bottom. Also along the cooling fins. If they are blocked you will see it.
8. Check the load is balanced.
9. If you have power factor correction, you must also check for THD, total harmonic distortion. Modern pfc controllers can have this built in so you may not require a service provider or buy an instrument.

 

[crshears]

OP made no mention of underload or off load tapchangers; does this transformer have either? If so, contact burning or pitting could be present due to bad wiping action or diverter switch operation...poor headboard connections could also be an issue; depending on the nature of the problem, any of these could cause localized heating without dissociating the oil, provided the temperature is low enough and no arcing takes place.

Is the transformer equipped with a gas accumulation alarm? Has the bank been gassing more than usual to date? If so, have gas samples been collected and analyzed? If so, what were the results?

Has any sweep frequency response analysis been performed on this transformer in the past? If so, what sort of comparative changes have occurred to the readings?

ONAF = Oil Natural, Air Forced, so there aren't any pumps, meaning the possibility of an oil valve to a pump having been left closed inadvertently does not exist. But has anyone here seen an ONAF transformer with valves to individual radiators? If these exist and one or more were left closed following rad repair, cleaning or replacement...

Thermovision [infra-red camera imaging] performed with the bank near rated temperature would indeed be an excellent technique for locating hot spots such as those created by impaired oil flow.

 

[stevenal]

But has anyone here seen an ONAF transformer with valves to individual radiators?

Yes. Seen them inadvertently left closed also.

 

[ScottyUK]

Ditto to stevenal's comment.

 

[squeeky]

Valves to isolate radiator are now standard on transformers1600 kV and above in the company I work for. When the regeneration of the oil takes place, you can get the temperature up very fast to drive off the water if you can valve off. Once up to temperature, open the valves in sequence to circulate the oil previously trapped in the radiator.

In Africa it has been discovered that if you mix diesel 50:50 with transformer oil you can run your car. They sometimes spike the radiator to drain the oil. Valves have an advantage.

 

[squeeky]

I have found some pictures of a transformer using an infra red camera. The princile here would be to look at the temperature differential from the top to the bottom. In the first two there is a differential and in the second two hardly anything indicating a failed cooling system. We were actually using the DOT360T Low Frequency Heating unit to drive up the temperature so we could drive the moisture out of the papeer insulation.

This is the first time I have uploaded so if they don't come through plase let me know.