100kw asynchronous generator power factor 2

[wpuryear]

I have a water turbine driven submerged 100kw induction generator that was rewound a couple of years ago. Power factor is running at .57 which we are correcting with capacitors, however the high current between generator and capacitors is causing heat problems.

The field connections are 12 pole, wye connected, adjacent pole, 4 circuit. Would some other connection scheme provide improved power factor or is this irrelevant?

Thanks,
Walt Puryear

 

[waross]

Sounds like undersized wiring.

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

 

[electricpete]

Adding power factor correction in general is intended to reduce the current provided by the source. So I'm wondering why power factor correction capacitors are implicated in high current from the generator. Can you describe the objective / strategy for this power factor correction a little more?

I'm not that familiar with induction generators. This one has a field circuit, I guess that means it's doubly fed?


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

 

[wpuryear]

The turbine and asychronous generator are totaly submerged and the capacitor bank is not. Current between generator and capacitors is 225 amps, however power factor correction with capacitors reduces this to 128 amps going to the grid. A better generator power factor would reduce the reactive power and in turn allow the generator and it's leads to run at lower temperatures.

My question is wheter the wiring scheme within the generator have a pronounced effect on power factor.

Thanks - Walt

 

[electricpete]

ok, I understand the role of the caps now. They supply the reactive requirements needed by the induction generator stator (so the grid doesn't have to).

Now your question:

The field connections are 12 pole, wye connected, adjacent pole, 4 circuit. Would some other connection scheme provide improved power factor or is this irrelevant?

So it's an asynchonous machine with a wound rotor and no external excitation to the rotor, right? [WRONG, SEE NEXT POST] Assuming that is the case:
1 - I don't think you can reconfigure the rotor without a rewind.
2 - Even if you rewind, I don't think rotor winding configuration has much to do with excitation current at all. Consider the magnetizing circuit reactive requirement which is a large part of the total reactive requirement. The magnetizing portion is constant from no-load (where no rotor current) to full load and depends on stator configuration and iron configuration… independent of rotor winding configuration. The only reactive draw you could hope to influence would be the load-dependent vars consumed by the rotor leakage reactance (which is a relatively small portion of total vars consumed in the machine).


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

 

[wpuryear]

It's not a wound but a squirrel cage rotor. It just seems to me that .57 is an awfully low power factor (name plate is .80) and I'm trying to learn why.

The unit was not operational when we purchased the plant due to a short caused by water infiltration, but as I recall, there were considerably more leads connected for each phase (only two per phase now) and thus my question about field connections affecting power factor.

Thanks - Walt

 

[ScottyUK]

Is the machine running at full power? The poor PF is consistent with an induction machine running at less than rated power throughput.

 

[electricpete]

It's not a wound but a squirrel cage rotor.

The field connections are 12 pole, wye connected, adjacent pole, 4 circuit

You threw me off with "field" for an induction motor. That terminology doesn't apply to rotor or stator of induction motor, but it made me think of rotor winding by analogy to synchronous machine.

Since it is squirrel cage machine (vs wound rotor), now we know you're talking about the stator winding. Stator changes could have more influence on reactive requirements than the rotor changes I was envisioining when I thought it was wound rotor.

I assume you are not talking about a rewind but just a change in connection. I don't work with induction generators, so I may not be the one to answer your questions, but some relevant details will include:

What is the nameplate voltage?
What voltage does the machine operate at? How is the voltage controlled?
What value of power factor correction capacitors are attached?


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

 

[waross]

Not enough information. Current on the generator leads, current on the grid leads and current on the capacitor leads would be helpful.
Where was the power factor measured? Power factor on the generator leads and power factor on the grid leads would be helpful.
KVARs of the capacitor bank may be useful.
A note about generator power factor ratings:
A name-plated power factor of 80% means that the generator is rated for 100 kW/0.80 PF = 125 KVA. This means that the load limit on the generator is 125 KVA.
KVA is voltage dependent so divide KVA x 1000 by the rated voltage to get the rated current.
This is the bottom line. This current should never be exceeded at any voltage.
Why a PF of 80%? Many loads operate at a reduced power factor. It is common to apply a prime mover of 100 kW capacity to a generator of 125 KVA. This is to allow low power factor loads (down to 80%) to be powered without overheating the generator. If the load power factor drops below 80% you should limit the current to the rated current.
The minimum capacity of a capacitor bank should be enough capacity to supply the reactive curren demands of the generator and to improve the power factor of the connected load to 80%.
Generators are typically wound with 12 leads to facilitate operation at two voltage classes and to allow re-connection for single phase use.
If a machine is being rewound for a known application it is common to make permanent connections internally and to only bring out 6 or 4 leads.
If the machine has been correctly rewound and connected then the internal connections should not make a difference.


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

 

[wpuryear]

Nameplate voltage is 480
Operating voltage is 480 +/- and is determined by grid voltage
Current on generator leads are 225 amps
Current on grid leads are 128 amps
Current on capacitor leads are 167 amps
Power factor on generator leaks is .57
Power factor on grid leads is .98
Rated output is 100kw
Actual output is 100 to 105 kw

So rated current is 125*1000/480 = 260. Appreciate all the input as I'm still climbing the learning curve.

Walt

 

[electricpete]

So rated current is 125*1000/480 = 260

Assuming you are talking rated current on generator output you need to add a sqrt(3) in there.

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

 

[electricpete]

P = sqrt3 * I * V * pf
I = P / (sqrt3*V*pf) = 100*1000/(sqrt3*480*0.8)=150A

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

 

[electricpete]

Your numbers add up better if your grid power out is at 0.98 leading than if you are at 0.98 lagging. Was that the intent? Can you confirm whether it is 0.98 leading or lagging.

Excess capacitors may tend to drive up voltage at generator terminals for a given system voltage. Do you know actual generator terminal voltage?

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

 

[electricpete]

you might want to check my math on the leading or lagging

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

 

[wpuryear]

I thought generator terminal voltage was determined by the grid voltage. At any rate, I'll read the values in the A.M.

Walt

 

[electricpete]

Your numbers add up better if your grid power out is at 0.98 leading than if you are at 0.98 lagging. Was that the intent? Can you confirm whether it is 0.98 leading or lagging.

Sorry, I was wrong on that piece. When I rechecked (attached) it looks like your current to system is 0.98 lagging as expected. So disregard the quoted comment.

I thought generator terminal voltage was determined by the grid voltage

It depends if there is an impedance between generator and grid (example transformer, possibly transmission line). Particularly var flow through reactive impedance elements creates voltage drops. I was wrong on the thought process that led me to that point (leading or lagging), but still the voltage at generator terminals is more relevant than a remote measurement, particularly if there's a transformer or long line between.

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

 

[electricpete]

Dang, reversed myself back to original conclusion. I had Ic with the wrong polarity in my last attachment.

Attached is 2nd version which suggests you are over-corrected (your caps are sending vars to the system).



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

 

[electricpete]

Attached document3 is a clarification of previous attachment (document2) where I have now corrected the comments labeled "hypothesis" to match the label of the variables (conclusion is unchanged that I_system is 0.98 leading

Attached is final... "document3.pdf"

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

 

[electricpete]

The preceding post had the final attachment (document3.pdf)

It shows your numbers are roughly consistent with I_system "leading" (means the system is providing a portion of the reactive requirements of the generator).

It also provides comparison of nameplate and actual conditions (which are discussed a few posts later in this thread).

[this post has been edited from its original version.]

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

 

[itsmoked]

electricpete; You do realize you can now edit your posts in here? A record is even kept.

Keith Cress
kcress -

http://www.flaminsystems.com