Generator Excitation & Power Factor 1

It doesn't lag, it leads to offset the lagging PF of the loads.
If a paralleled generator is badly under-excited it will input inductive VARs from the other generators to maintain excitation. When the generator is over excited it is exporting inductive VARs and supplying more than its share of VARs into the system relative to the other generators.
A synchronous condenser is a special synchronous machine which is started like a motor but which does not drive a load. Some do not have a shaft extension. As the excitation is increased, the machine exports VARs to offset load VARs.
Some synchronous motor applications use an oversized motor and over excite the motor to supply VARs to the system.
A synchronous motor may theoretically be used as a synchronous condenser, but in practice the high excitation levels may result in field currents above the current rating of the field windings and over-heating of the field windings. It depends.

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

Hmm, I think the original statement was correct? An over-excited generator exporting VArs to the system is operating at a lagging power factor. Everything is back to front for a generator vs a motor.

The rest sounds good though

Look up the classic model of a synchronous machine, which is an ideal source in series with a reactance. It should be in any decent machines text, along with a vector diagram of how the phase angle between the machine internal EMF, the system EMF, and the line current varies as excitation is varied.

I'd draw it for you but in a text-only forum it is quite difficult. Google [blue]synchronous machine vector diagram[/blue]. This interactive one might be of interest:



----------------------------------

If we learn from our mistakes I'm getting a great education!

BigJohn,

You stated in another thread that you are not an engineer and I suspected that when I read the original post.

Your question can be answered, but you have asked a group of (mostly) engineers or if not engineers then people well versed in this field so in order to try to help you (which I think most engineers are willing to do) I would ask your background so that they can couch their answers in less complicated terms that will aid your quest for knowledge.

And, that said, in the interest of full disclosure, I am a ME who works in the power generation world so sometimes I have to have things explained to me in "not so in depth" EE terms. I learn a lot here.

I don't understand your statement about not being able to adjust the stator voltage. To my EE friends, is that an accurate statement? Once synchronized, isn't the raising of the stator voltage the way power is pushed out onto the grid? Or is it only by the machine trying to "speed up" the grid at synchronous frequency?

BigJohn's assertion has confused me here.

rmw

If the unit is synchronized with the grid, in other words the Generator circuit breaker is closed, and the Grid is large compared to the Generator MVA, then raising the excitation will increase the VAR flow to the grid, but the voltage will only change a relatively small amount. The amount of change of voltage measured by PTs connected to the stator depends on the impedance of the generator transformer (if present). That could be around 10% on the stator side and a lesser percentage on the Grid side, depending on what else is connected on the grid side.

This does not change the MW output, but only the reactive component, the VARS. (or MVARS) Only the governor of the prime mover can change the MW.

The Power factor will change becoming "lagging" as seen by the Power factor meter, as the Generator is seeing a "lagging" load requiring VARS. There is some confusion on PF meters, some old ones having a 360 degree scale where the pointer can do a full circle.

hope this is clear...

rasevskii

Watts and VArs in the same direction is lagging. Watts and VArs in opposite directions is leading.

Using the motor convention (suitable for all loads), VArs in is lagging while VArs out is leading. Any load that requires VArs is therefore lagging and any load that supplies VArs is leading. Generally the only leading load is something capacitive, although anything with an active front end (power electronics) can be leading if so desired.

Using the generator convention (suitable only for sources) VArs out is lagging while VArs in is leading. A lagging source supplies a lagging load. If current into the load lags the voltage then current out of the source also has to lag the voltage.

Instead of worrying about supplying VArs (leading loads and lagging generators) or absorbing VArs (lagging loads or leading generators), I find it much easier to just consider VAr direction in relation to Watt direction. In all cases, if something absorbs both Watts and VArs or supplies both Watts and VArs it is lagging, whether a load or a source. If something absorbs one while supplying the other, it is leading, also whether a load or a source.

RMW,

I'm a utility electrician for a bunch of generating stations, I'm just looking for a non-mathematical explanation for this. I think visually, so I'm trying to visualize what's actually happening to the stator when excitation is changed while the unit is paralleled.

When a paralleled stator is over-excited, it may technically change the voltage measured at the generator leads, but my experience is that value is very small because of the ability of the grid to hold the generator at grid voltage, that's what I meant when I spoke of "infinite bus." I'm not too concerned with this part.

What does change significantly with excitation is the PF reading and that's the part I'm trying to understand.

Here's a question that may help me a bit: If the load on the generator were almost purely resistive (ignoring line reactance, etc.) and the generator was over-excited, what would the power factor meter read? My understanding is that generator PF is determined purely by the load. So what would the increased excitation do to the generator under that circumstance?

Thanks so far.

-John

If the generator was islanded, the voltage would go up. If the generator was tied to the grid, the generator would export VARs. If the generator was able to export more vars than the grid demanded, the voltage would try to rise. The other generators would react by dropping their excitation slightly and become under excited. They would then use the VARs provided by the overexcited generator to provide part of their excitation.
It would be unusual in practice for a large grid to be so close to unity.
On the other hand, The small islanded system that I was associated with was generating with a mix of 600 KW and 350 KW diesel generators. Machines were added or dropped off as the load changed. We seldom went over two x 600 KW.
On a system that small, you could get a lot of interaction between the machines if the excitation was not correctly adjusted. Raising the excitation would often result in a combination of higher voltage on the grid and a shift in power factor and uneven VAR sharing.

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

waross...Nice explanation on the practical results of changing generator excitation.

About the question of generator power factor being leading or lagging...I had to think through several times bacause some of the post were at the same time convincing and confusing. I wonder if this isn't the same type if question as deciding whether current flows from a positive terminal to a negative terminal or the other way.

I say that the power factor of whatever you are looking at is determined by measuring the phase relationship between voltage and current at the terminals of the device. If the current is lagging the voltage at the terminals then the power factor is considered to be lagging.

To picture this in a simple way, picture a single generator (source) connected to a single load. The source voltage and current must be the same as the load. If the load is inductive and the current lags the voltage (lagging power factor) then the source current must be the same. Meters installed at the termanals of the source must read the same voltage, the same current, and the same lagging power factor as meters instaled at the load.

For ideas of what happens when paralleling, refer to waross's post.