Transformer Oil Power Factor

[blacklabs]

Can someone explain the meaning of "Transformer Oil Power Factor". An analysis of our transformer oil indicated a sudden change in Power Factor that gradually began to recover. When we refer to PF, we refer to the relationship of real and reactive power. How does it equate with oil analysis?
Thanks for any help.

 

[Skogsgurra]

Don't see that very often. For power people, I think that 'Power Factor' is a bad name since one tends to think cos(delta). It is, instead sin(delta). Tan(delta) is used more often.

Mathematically, they are quite close at the small loss angles you find in oil. If oil contains water, that usually (always) means that tan(d) increases. Getting rid of water (filter or heating) takes tan(d) back again.

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[waross]

When we refer to PF, we refer to the relationship of real and reactive power.

Yes
The formula for real power is I²R and for reactive power is I²X so power factor also expresses the relationship between Resistance and Reactance. A change in power factor may indicate a change in either R or X or both.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[hdp]

http://www.astm.org/Standards/D924.htm

Link to ASTM D-924 test protocol for Oil Dissipation Factor

 

[electricpete]

When we refer to PF, we refer to the relationship of real and reactive power. How does it equate with oil analysis?

The oil power factor test quick easy test that is somewhat sensitive to things like moisture (power factor goes up when moisture goes up). For a Karl Fischer moisture test, you have to send the sample off to a lab somewhere and wait. For the oil power factor test, it can be done in the field with Doble test set and a special cup. Results are available immediately.



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

 

[blacklabs]

Thanks to all.

 

[electricpete]

Don't see that very often. For power people, I think that 'Power Factor' is a bad name since one tends to think cos(delta). It is, instead sin(delta) [emphasis added]. Tan(delta) is used more often.

I think the bolded portion is incorrect. There is no special definition of power factor used for insulation testing, it is the same power factor as used in circuit analysis. Both are cos of angle between voltage and current, which I assume you’re calling delta.

Examples:

Typical motor under load: Current lags voltage by 15 degrees, delta = 15 degrees, pf = cos(delta) = cos(15) = 0.966 (lagging)

Typical insulation specimen: Current leads voltage by 89 degrees, delta = -89, pf = cos(-89) = 0.017 = 1.7%.(leading)

Tan delta is often used abroad to measure the same insulation characteristics that are measured by power factor in US. For typical values small power factor angles below 5%, the two numbers (power factor and tan delta) are virtually the same.

I’m sure this is nothing you don’t already know. Sorry if I have misunderstood something what you were saying.


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

 

[electricpete]

Tan delta is often used abroad to measure the same insulation characteristics that are measured by power factor in US. For typical values small power factor angles below 5%, the two numbers (power factor and tan delta) are virtually the same.

Sorry, I was a little off on this piece. I think tan delta works off a different angle than the power factor angle. It works off the loss angle which is the devitation from pure capacitive behavior... i.e. deviation from 90 leading power factor angle.





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

 

[electricpete]

Let me re-write the whole thing using different notation
Notation:
[*]pf = cos(theta)
[*]theta = power factor angle = angle by which current lags voltage.
[*]delta = loss angle
[*]delta = theta + 90 degrees

Examples
Typical motor under load: Current lags voltage by 15 degrees, theta = 15 degrees, pf = cos(theta) = cos(15) = 0.966 (lagging)

Typical insulation specimen: Current leads voltage by 89 degrees, theta = -89, pf = cos(-89) = 0.017 = 1.7%.(leading)

Tan delta is often used abroad to measure the same insulation characteristics that are measured by power factor in US. For typical values small power factor angles below 5%, the two numbers (power factor and tan delta) are virtually the same.

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

 

[Skogsgurra]

Pete, I used delta for a reason.

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[electricpete]

Thanks, I understand what you were saying now, and I apologize for saying it was “incorrect”. It was certainly correct.

As an excuse for my misunderstanding, I’ll mention that it was somewhat hard for me to follow the logic that suggested “power factor” is a misleading term to apply to insulation...... since power factor is applied in a perfectly consistently manner in both insulation measurements and circuits as I showed above. The less straightforward terminology imo comes from delta and tan delta... new terms to describe something we already know (power factor). It all just depends on what you’re used to, I guess.


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

 

[oldfieldguy]

The effect is pretty easy to envision, really.

You have a sample cup filled with oil. Good oil is almost a perfect insulator, so in the sample cup you have a capacitance (essentially fixed: cup geometry and the dielectric constant of oil) so you have much more capacitive current than resistive, giving a LOW (very low) power factor, because you have a high resistance (real power loss) paralleling a very small capacitance.

It the oil is deteriorated, the resistive component changes due to polarized components in the oil, therefore the resistive component is higher. In any reasonable oil sample, there will be little change in the capacitive element, but with "bad" oil, the resistance goes down, the watts loss goes up, but the capacitance stays more or less the same. Therefore, the power factor INCREASES.

These changes are very small. Doble's literature says that oil with a power factor of 0.3% or less is acceptable for service, 0.5% is considered "doubtful" and should be tested by other means, and 1.0% should be investigated, reconditioned or replaced. Of course, the Doble oil power factor test setup is a field test and not as rigorously standardized as a laboratory test.

As to the change in your transformer, remember that the affinity for oil in the transformer is shared between cellulosic components, the paper and wood used in insulation and support, and the oil itself. Hot oil will absorb oil from "wet" cellulose. As the oil cools, the moisture will migrate back to the cellulose, as long as there is enough time for this to take place. If the change is rapid, the water in the oil may precipitate out of the oil, resulting in visible moisture in the tank.

It's all not very static, and that makes it interesting to interpret.

old field guy