Increasing VAR Output of a Machine or Decreasing VAR Requirement of the System

A

Thread Starter

amritanshup

I am interested to know what are the different ways to increase VAR capacity of the generator?

In short, if the capability curve (D-curve) of the generator is falling short of the VAR requirements required by the regulator, how can the generator expand its VAR absorbing or producing capability?

Thanks,
Amrit
 
Amrit,

Is the grid voltage higher or lower than normal when the regulator is asking for more VArs?

Does the step-up transformer have adjustable taps? If so, is the tap changer an on-line, or on-load, tap changer?

If the transformer has adjustable taps, have you tried changing taps when grid voltage is higher or lower than normal? What were the results?

In general, synchronous generators aren't used for leading VAr service--per the D-curve. Does the regulator ask the unit to go leading very often?
 
Amrit,

There aren't too many things which can be done to increase the VAr capacity of a particular generator, that's a function of the design and construction of the generator. The D-curve represents the limits of the generator's ability to produce or consume VArs as a function of the ability to cool the generator from the heat that develops when producing or consuming VArs. I imagine it's possible through an upgrade of the generator windings, rotor windings and cooling system that something could possible be done to improve the VAr capacity, but it's probably pretty expensive to do one or all of these things. Something to be considered when trying to "absorb" VArs is avoiding pole slipping--which can be very destructive to the generator, the coupling between the generator and its prime mover, and the prime mover.

If system voltage is much higher or much lower than normal then that can have an effect on the ability of a generator to produce or consume VArs. And, that can be the result of a lot of factors--mostly under the control of the grid regulator/operator.

One way to allow a generator to have more "range" is to use a transformer with a tap changer to vary the transformer ratio. There are a couple of types of tap changers, but they're not generally something that can be added to a transformer--they're either already present, or they're not. And, on-line, or on-load, tap changers are not cheap.

Power factor correction capacitors can also be used to help with grid power factor, but they're not cheap either.

Without more information from you, there's not much more can be said.
 
A
Phil and CSA,
Thanks a ton for your help. Sorry for the late reply but I was having trouble accessing my account.

Phil,
I went through the previous post and got great insight.

CSA and Phil,
I further agree with the both of you regarding tap changers being the best possible way to increase VAR capability and was my initial understanding.

Going into more detail,

The generator is connected to the grid. The owner increased the generator real power capability only to realize that the it cannot produce sufficient reactive power (importantly leading reactive power).

What interested me in this case was the application of D-curve to justify inadequacy of the VAR capability and please correct me if I am wrong. From what I understand, the manufacturer provided D curve (in this case GE) is for a particular voltage usually the rated voltage (am I right ?). Further, from what I understand, the generator D curve should expand or contract depending on the system voltage (and therefore the tap-changer). Has anyone actually looked into using dynamic D curves for varying system voltages? Or is it too foolish for me to think that this is not already what's used by the owner and the regulator?

However in this case, as the leading VAR requirement is not met, terminal voltage would not help much because the VAR capability is restricted by the end region heating due to flux.

Your thoughts please. Sorry again for the late reply.

Thanks,
Amrit
 
A
Phil,
My colleague was working on a project wherein the issue that the generator owner was facing was that the generator in its own capacity could not produce enough reactive power for a maximum real power.

The reason I believe that the generator owner even reached in such a situation is that generator real power capability was enhanced. However, the generator when providing that real power is not able to meet the reactive power requirement.

So I was approached by my colleague for comment and was provided with a generator D curve. The generator D curve showed that the required leading reactive power point for maximum real power was outside the D-curve.

My suggestions and discussions with my colleague involved (1) How to increase generator reactive power capability ? ;(2) How to control voltage using on-load tap changer to improve reactive power capability or ; (3) Using capacitors banks to supply additional reactive power

What I did not understand in depth is the actual D curve. Why did only one D-curve exist (My guess is that D curve would expand or contract based on terminal voltage)? Secondly, what are the different ways to modify the D-curves to which I was helped by CSA and you.

Do you have some insight you will like to share?

Thanks,
Amrit
 
amritanshup,
It's not all that common for power producers to be asked to "consume" leading VArs, and when they are it's been my experience that a synchronous condenser is a better choice for such applications.

There's something we're not being told about this application and the grid voltage conditions. You way the owner upgraded the real power output capacity--of the prime mover or the generator, or both?

Was the owner being asked to consume "leading" VArs prior to the power output upgrade?

Dynamic D-curves? You'd have to consult the generator manufacturer about that. I've never see a D-curve drawn with varying terminal voltages, only varying cold gas temperatures. I would suspect the typical D-curve for varying cold gas temperatures would be good for the expected operating range of the generator (usually, plus/minus 5% of rated generator terminal voltage--since that's about all the excitation system is usually good for, and, as you know, it's further limited on the under-excited end for multiple reasons including end-turn/end-iron heating limits and pole slipping avoidance).

At any rate, expensive as it may seem, based on the information <b><i>provided</b></i> a transformer with adjustable taps is likely the best choice. On-line, or on-load, tap changers are pretty expensive, too.

As Mr. Corso says, if you want more suggestions or recommendations you're simply going to have to provide a lot more information.

 
A
Thank you Phil and CSA,

From what I know the owner upgraded the output power capacity and not the prime mover. I can try to dig in for more information and see if I can find any.

You are also right that D curves are also limited by many factors that are not a function of voltage. However, I only thought about dynamic D curves because of reading in the literature about how D curves are developed. A lot of it is dependent on armature and field current and given generators are constant KVA machines, higher voltage will mean lower currents and so was my thought.

Although, I am sure there are many other factors involved in developing D curve (such as end ring heating) because of which only one D curve is used.

Always great to learn from the two of you!

Thanks a lot,
Amrit
 
amritanshup,

I think there are some issues with terms and usage here.

Real power is watts, or kw, or MW--the power producing tangible work (watts can converted to HP). It is the side adjacent to the hypotenuse of the power triangle, usually called "P."

Reactive power is power that doesn't do any real "work"--it's sometimes called "imaginary" power, and it can be leading (feeding leading--predominantly capacitive--load), or it can be lagging (feeding lagging--predominantly inductive load). Its units are VArs, kVArs, MVArs. It is the side opposite to the hypotenuse of the power triangle, usually called "Q."

Apparent power is the "combination" of Real and Reactive power; its units are VA, kVA, MVA. It's usually how most generators are rated, and one takes the power factor into account when calculating the Real power output rating capability of the generator.

For example, a generator rated at 30 MVA at a power factor of 0.80 would be capable of producing 24 MW <i><b>if it were producing power at a 0.80 power factor</i></b>. Apparent power is the hypotenuse of the power triangle--the vector sum of Real- and Reactive power--usually called "S."

In my experience, it's quite rare for the real power capability generator to be rated for the same or less than the prime mover driving the generator.

Most prime movers, under certain conditions, can produce slightly more power (horsepower; torque--watts) than nameplate rated, so in order to prevent the generator from limiting the total output of the generator-set (the generator and the prime mover) the generator is typically rated for more power (Real and Apparent) than the prime mover is capable of producing (Real). The excitation system provides the "power" for producing lagging VArs, by "over-exciting" the generator (the generator rotor, the field).

A synchronous AC generator is a device for converting torque into generator stator amperes--which is how Real power is produced. Increase the torque from the prime mover and the generator (which can't rotate any faster) converts the torque into generator stator amperes, which is what makes the Wattmeter needle move in the positive direction (increasing).

So, unless in this case the generator was "under-rated" before the uprate, it's hard to understand how increasing the Real power output rating of the generator would actually increase the Real power output of the generator-set--since the prime mover provides the real power to the generator which the generator converts to amperes to make the Wattmeter needle move in the positive direction.

It is possible to increase a generator's ability to convert torque to amperes to increase the generator's ability to produce Real power--but generally it would be necessary to increase the torque-producing capability of the prime mover to be able to have more torque for the generator to convert to amperes (over and above the previous rating/power output capability). Just increasing the real power output capability of the generator doesn't mean one can increase the Real power output of the generator--unless there is additional torque coming into the generator (or unless the generator was previously rated for less than the prime mover).

Using the generator in the example above, it would not be uncommon for a prime mover capable of producing 20MW to be driving the 30 MVA generator. If the power factor were maintained at 1.0 (unity--producing purely Real power)--and this is done by adjusting the AVR (exciter) output--the generator would only be capable of producing 20 MW. In other words, at a power factor of 1.0, the Apparent power output of the generator would be 20 MVA, with 0 MVArs.

If the excitation were increased while producing the same 20 MW but the power factor was changed to 0.90 lagging, the Apparent power output of the generator-set would change to 22.22 MVA, with only 20 MW of Real Power and 9.69 MVArs (lagging)--approximately.

When one is looking at the D-curve (it's called a D-curve because there are three sides to the curve which are curved, one which represent the limit of the ability to produce Real power, one which represents the limit of the ability to produce lagging Reactive power, and which represents the limit of the ability to produce leading reactive power.

While it's difficult to draw, it might look like this:<pre> Watts
(+)
|
**************** <--Maximum Real Power
* | *
* | *
* | *
Maximum * | *
Leading * | * <--Maximum Lagging Reactive Power
Reactive * | *
Power---> * | *
* | *
* | *
* | *
(-)---------------------------------------(+) VArs
Leading Lagging</pre>(In reality the upper line is usually has a slight curvature to it, also, but it's difficult to "draw" in ASCII.)

The top "curve" is the limit of the ability to remove the heat when producing Real power--the heat generated by the flow of current in the stator windings of the generator.

The right-most "curve" is the limit of the ability to remove the heat when "producing" lagging Reactive power--the heat generated by the flow of (DC) current in the generator rotor windings.

The left-most "curve" is the limit of the ability to remove the heat when "producing" leading Reactive power--the heat which concentrates in the end-turns of the generator stator windings and the heat in the end-iron of the generator stator. I believe there is also avoidance of pole slipping factored into this "curve"--but that's just a personal opinion and has no scientific basis. (Reducing excitation is how leading VArs are "produced", and reducing excitation too much can result in loss of magnetism and pole slipping--which can be very destructive to a generator, load coupling and prime mover.) It's not generally possible for most synchronous generators I've worked on to "produce" very many leading VArs--at least not per their D-curve (reactive capability curve, or capability curve).

Many capability curves have three or more sets of curves, all with the same general shape, which represent different cold gas temperatures (the temperature of the gas entering the generator--which may just be ambient air--but which affects the ability of the generator to remove heat from the various parts of the generator. And, the hotter cold gas temperature lines ("curves") are closer to the Watts line--meaning less Real- and Reactive power can be produced because the cold gas entering the generator is high so less heat can be removed to less Real- and Reactive power can be produced.

Lagging VArs and lagging power factor are generally considered "positive"--from a generator perspective. I believe this is somewhat opposite at high-voltage switchyards, but that's only what I've been told by switch-yard operators, and I'm not sure they were trained very well.

So, which Reactive power are you talking about--leading or lagging (from the generator's perspective, as in the drawing above)? Or, are you referring to Reactive power at a high-voltage switchyard somewhere outside of the generating plant (which can refer to VArs differently)?

'Cause I'm getting quite confused here.
 
A
CSA,

First thank you on such a detailed explanation.

My information (I had to dig some information):

1. The generator was uprated. This is a combustion turbine generator. I frankly do not know if in this particular case, it is the generator that was uprated or the prime mover or both. From your explanation was guess is that prime-mover must have been uprated too or both generator and prime mover. You must be a better judge of this having greater knowledge of CTG systems.

2. Required reactive power capability post this uprate was not sufficient. I do not know if the reactive power was measured at the terminals or the grid. Requirement was set by PJM in interconnect agreement (My best guess).

3. To best of my knowledge the plant was not able to supply both required lagging or leading reactive power (kVARs or MVARs) for the uprated real power (Operating at different power factors).

Getting back to the D-curve, my understanding is that new uprated generator may not be meeting the reactive power requirements for right and left in the figure.(Operation mode drawn using symbol 0)

I am pretty certain that required leading VARs is a cause of concern for the owner because when my colleague first spoke to me he had mentioned capacitor banks, so I am guessing they need to provide leading power factor).

Watts
(+)
|
**************** <--Maximum Real Power
* | *
O * | * O
* | *
Maximum * | *
Leading * | * <--Maximum Lagging Reactive Power
Reactive * | *
Power---> * | *
* | *
* | *
* | *
(-)---------------------------------------(+) VArs
Leading Lagging

 
A
CSA,

Sorry my figure did not come out as I expected. I was trying to show the region of operation on the D curve that was required by PJM.

Thanks,
Amrit
 
amritanshup,

Here's a link to a website about power factor, and there's a webpage for PF Capacitor, which you should find useful:

http://www.the-power-factor-site.com/Powerfactorcorrectioncapacitor.html

When it comes to generators--that is, from the generator perspective (at the generator output terminals)--lagging VArs feed a lagging load. Lagging loads (inductive load factor) result in reduced system (grid) voltage as well as shifting the AC voltage- and current sine waves out of phase with each other.

A synchronous generator can help to support system voltage and keep the voltage- and current sine waves more in phase with each other when it's over-excited. In other words, by increasing the excitation applied to the generator rotor field above that required to make the generator terminal voltage exactly equal to the grid voltage (a power factor of 1.0, and a VAr-meter reading of 0.0), the generator can "produce" lagging VArs to help feed a lagging load. This also has the effect of "boosting" the grid voltage.

Another way of countering a high inductive load factor is to use power factor correction capacitors. And, fluorescent lighting is also a very effective means of helping to maintain grid voltage and power factor, since many loads are inductive in nature.

So, when someone mentions using power factor correction capacitors it's usually because the load is inductive in nature, and adding capacitance (which is leading in nature) is one way to counter the effects of a large inductive load factor.

BUT, the other method of countering the effects of a high inductive load factor is to over-excite the generator, which results in lagging VArs to feed a lagging load (which is probably why lagging VArs are considered "positive"). Going leading when the grid load is highly inductive isn't going to help the grid.

Here's a link to a wikipedia article about synchronous condensers:

http://en.wikipedia.org/wiki/Synchronous_condenser

Note that a synchronous condenser is considered to be an electric motor--and when a synchronous electric motor is over-excited it is considered to have a leading power factor, producing leading VARs, to supply a lagging load. This is different from a generator--which, when over-excited, is considered to have a lagging power factor, producing lagging VArs to supply (feed) a lagging load.

Many GT-driven synchronous generators have special clutches that allow the GT to be shut down while the "generator" remains connected to the grid, spinning as a motor (actually drawing current--real power) from the grid BUT because it can be, and is, over-excited it can act to counter the effects of a lagging load factor--but it is considered to have a leading power factor, producing leading VArs. But, since it's NOT a generator at this point, but rather a motor, the convention is reversed.

[I don't make this stuff up--it's the way society has agreed to standardize on the terminology.]

I suggest we've beat this subject to near death, and without concrete information from your colleague it's not possible to continue any further.

Further, I maintain there is still some confusion (on someone's part--maybe even mine) about terms and usage. I've been wrong many times in the past, and I'll probably be wrong many times in the future. But, on this--I think (without using a lot of maths and vectors) I think I'm pretty correct. (I may not be describing things exactly correctly when I refer to "inductive- or capacitive load factor", but I think the rest is pretty correct.
 
Top