transformer highest temperature attained temperature indicator mysteriously spikes twice 2

[electricpete]

Background
* 50MVA H / X / Y transformer 25kv to 13.8kv with LTC.
* transformer has 3 local temperature indicators. oil temperature (mechanical / capillary). X and Y temperatures similar to oil temperature indicator except the temperature is boosted by resistive heater circuit fed by ct's for associated X or Y winding to simulate associated winding temperature.
* white needle gives temperature at the time of the reading, red needle gives highest temperature attained since last manual reset of the indicator.

The values are logged daily. For 1/12, 1/13, 1/14, all the white needles were < 40. 2 of the three red needles were < 40. The X winding red needle
* 1/12/23: 40
* 1/13/23: 85
* 1/14/23: 113

The transformer has an on-line gas monitor. No alarms were received. The trends will be reviewed tomorrow.
The transformer has 2 sudden pressure devices (one for alarm 2 for trip). No alarm received.
No unusual voltage transients were seen on transformer looking at the available monitored voltage on the low side (i think it samples every 10 seconds). Transformer LTC cycle counter will be reviewed tomorrow

Brainstorming possible causes
[ol 1]
[li]Actual increase in oil temperature. Ruled out since oil temperature indicator did not go as high[/li]
[li]Temporary increase in actual load – unlikely since 13.8kv bus voltages showed no unusual transients (monitored with 10-second samples). We could further rule out increase in actual load by reviewing LTC step counts tomorrow.[/li]
[li]Temporary increase in simulated load due to temporary malfunction of the heater circuit[/li]
[li]Unknown mechanical factors causing higher temperature attained red needle to increase[/li]
[li]Temporary electrical fault. Unlikely that a fault would clear itself. The fault scenario could be further ruled by reviewing locally logged parameters[/li]
[/ol]

questions
[ol 1]
[li]REGARDING 4 - has anyone seen spikes in highest temperature attained temperature indicator which were spurious (lilke item 4). is the indicator susceptble to that kind of thing?[/li]
[li]ReGARDING 3 - would intermittent shorting of turns in the ct secondary cause a lower effective secondary turns and lower stepdown ratio and higher secondary current? [/li]
[/ol]

=====================================
(2B)+(2B)' ?

 

[prc]

It may be only malfunctioning of needle. Just tap the side of the meter and see whether the needle is jumping. Chance for item 3 is remote.

 

[Palletjack]

Most winding temperature gauges are out of calibration. I wouldn't trust it as there are too many variables at play. Tap position, loading, calibration, etc…
When is the last time the oil temp guage has been tested and calibrated?
There really isn’t a way that i know of to test or calibrate a winding temperature simulated temperature
Could also be an indication the CT is bad.

 

[thermionic1]

The conventional way of calibrating oil temp / winding temp is to place the temperature sensing probe in a dry well calibrator and check indication. For winding temp, regulated AC current is also applied in several increments. From there, the manufacturer provides a chart for applied current & temperature vs. gauge indication. Pretty standard for Qualitrol and ABB (Wika?) gauges. The WTI CT's are likely checked at the time of installation and at regular intervals, along with any protection and metering CT's. We'll check the entire cooling circuit, verifying that the microswitches activate the respective alarms, fan 1 / fan 2 and trip operations. IIRC, the OP works at a facility that probably takes calibration of devices seriously.

 

[bacon4life]

Is the X winding a tertiary? Any changes in harmonics or imbalance that would increase load on the tertiary?

Other unlikely ideas:
-Sunshine can increase temperature well readings by many degrees, but probably not 60+ degrees.
-Loss of oil in an oil filled well could affect temperature readings.

 

[jghrist]

Could the oil level affect the winding temperature indicator? If there is little oil in the heater well, would the CT current increase the temperature too much?

 

[RRaghunath]

I guess jghrist has a point.

 

[prc]

Raghu, it cannot be correct. In today's models, winding gradient is added to oil temp not by heating the oil in the well. It is accommodated in the indicator part by heating a bellows. Even in those old models, when oil level is low, heater will be exposed to air to that extent and will not heat up oil excessively. In our case, winding temperature is not continuously showing high; only the maximum indicator needle is showing high values.

 

[waross]

A delta tertiary or phantom delta will attempt to stabilize primary voltages and phase angles.
How to explain?
Consider a voltage unbalance on the primary supply of 10% with the transformer off-line.
When the transformer is placed on-line, the unbalance is corrected to 5%.
Phase angle errors are more subtle and may cause similar unbalance issues.
If the KVA rating of the delta is much lower than the KVA of the main windings, the circulating current in the delta may cause unexpected excess heating in the tertiary winding.
A mis-operation of the Load Tap Changer may cause similar heating in a delta winding
on some tap positions.

--------------------
Ohm's law
Not just a good idea;
It's the LAW!

 

[3DDave]

"Temporary electrical fault. Unlikely that a fault would clear itself."

This is not always true. I have seen several cases where differential thermal expansion has caused wires to fatigue and fracture. At normal temperature the assembly holds the broken ends together sufficient to operate correctly. As the temperature elevates the various elements of the housing grow faster than the wire does, eventually pulling them into less and less contact. If the temp rise is slow the connection has an intermediate high-resistance condition before finally going to an open/infinite resistance condition.

When the assembly cools the process reverses and the unit operates as expected.

Symptomatically one might see that this high-resistance/open circuit condition happens with increasing frequency as each cycle does more damage to the contacting ends of the metal.

For example, the famously crashed Lion Air jet had, for a period, been reporting intermittent angle of attack sensor issues. These issues were coincident with the use of the sensor deice heater. Bench testing, at normal temp, showed no problem. It took a microscope to see the place a resolver wire had fractured. In that case it fractured because a strain-relief section of wire had been epoxied, putting the full strain of thermal expansion away from the now ineffective strain relief.

More often they happen with components on circuit boards where there is a mismatch between component expansion and board expansion, causing fractures either at the board solder joint or where the lead enters the epoxy body, possibly of an IC or, most sneaky, within the body of a component such as a capacitor were the internal bonds can fail. These are often found using a local application of "freeze spray" to cause contraction to force electrical connections back into place.

Thermal differential expansion based failures can be difficult to locate.

 

[electricpete]

I appreciate all the great inputs on this.

I'll have to look up if we have a tertiary winding if and when I get a chance. Although the oil temperature did not increase and I'm pretty sure the ct's feeding the heater for the winding temperature indicator only measure external currents.

I have to confess this question arose on a Sunday morning while I was on duty for our plant electrical engineering section (that's the only time I have involvement with transformers any more).
My role was primarily to decide if any immediate actions were needed.
I laid out some possible causes and went out on a limb to say it should be fine, at least until.... the other Engineers get to work on Monday!
By Monday morning I had handed off the issue someone else.
I should have mentioned that a few weeks ago before you guys spent the time replying.
It does make me feel good to know I have so many expert people willing to contribute their ideas when I ask. Including our old friend waross!

Thanks again.




=====================================
(2B)+(2B)' ?

 

[waross]

You may suggest exercising the LTC while watching for current spikes. Are the primary voltages recorded?

--------------------
Ohm's law
Not just a good idea;
It's the LAW!