Further to David's comment; VARs c 1

[waross]

Further to David's comment;
VARs cause increased line current which causes increased I[sup]2[/sup]R losses.
If you send the VARs back to the source and cancel them there, you save some losses in the generator and avoid reduced KW capacity of the generator.
BUT
You haven't reduced transformer losses or line losses.
The examples that crshears and I shared were plants that were at one time local to the city served. As more remote generation became available the plants were mothballed. as the load grew, the plants were put back into service as synchronous condensers.
These old generators were at the load end of the transmission lines, not at the source end.
Generating the VARs locally served two purposes;
1. Reducing the line losses and at the same time reducing the line voltage drop.
2. With sufficient capacity, VARs could be sent back up the transmission line with the result of raising the voltage. The raised voltage allowed a greater line capacity before the OLTCs bottomed out.


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

 

[prc]

Thank you Waross! In India too, there are proposals to put back old generators as condensors instead of scrapping them,

 

[waross]

It depends on the location of the old generators. If the old generators are close to the loads then they may save enough transmission line losses to be worth while.
If the old generators are near the new generators, they will not help with line losses but they will improve the power factor of the load presented to the newer generators.
This increased generator capacity depends on the ability of the prime movers to run the generators at power factors above the rated 80% PF.

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

 

[crshears]

This increased generator capacity depends on the ability of the prime movers to run the generators at power factors above the rated 80% PF.

?????

Not quite sure what you're saying, Bill, as "above the rated 80% PF" can be taken two ways . . .

The big concern in my view is to always run within the [combined] generator capability curve; the only time I have seem the "generator capacity depend[ing] on the ability of the prime movers to run the generators at power factors above the rated 80% PF" is when the prime mover mechanical power rating exceeds the generator MVA rating, meaning that when it is desired to operate with the generator either absorbing or producing [lagging] VARs the energy input to the prime mover [wicket gates, fuel rack or gas or steam admission valves] have to be dialled back from 100% via the governor to ensure generator limits are respected.

In instances where the prime mover is relatively undersized for the generator, the end result is that near unity power factor the stator currents are always quite comfortably within the limits, implying that even at excitation values resulting in maximum [lagging] VAR production or absorption the stator current limits are essentially never in danger of being exceeded. Even field current limits are not in my experience contingent upon prime mover capability.

If I've misunderstood something, please correct me.

CR

"As iron sharpens iron, so one person sharpens another." [Proverbs 27:17, NIV]

 

[waross]

I'm remembering our islanded plant, out on the island.
The load ran around 80% pF.
At times the load would be just over the capacity of our 600 kW set (0.8 PF rated)
We considered adding capacitors so that we could we could run the set at unity power factor and avoid starting a second set.
That is, run the 600 KW set at unity PF and at the KVA rating of 750 KVA.
For proof of concept we used seat of the pants engineering.
When the load was above 600 KW and two sets were running, we over-excited the second set until the first set was at unity PF. Then we advanced the governor on the first set.
We got up to 600 KW and no more. The diesel engine was not capable of supporting more than 600 KW output.
That was the end of that idea.
That was the experience behind my earlier post.

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

 

[crshears]

Hmmm; interesting, and illustrative. Thanks.

CR

"As iron sharpens iron, so one person sharpens another." [Proverbs 27:17, NIV]

 

[Fischstabchen]

Warros,

The size of the generators you are bringing up have little impact on the grid. No one would operate a 600 kW generator as a synchronous capacitor unless it was for some special application that needed fast transient control or voltage independent var support.

Var control on large synchronous generators is not limited by the prime move in my experience but by the rotor and stator current limits.

Large generators could easily be used as synchronous condensers due to them always have a strong connection to the grid so that the vars could spread. There is little chance of this happening due to the fact that utilities are no longer vertical. Generation and transmission are separated and the way transmission planning is carried out transmission fixes transmission problems. Most regions don't compensate for vars and even if they did, on a MVAR per dollar expense, capacitor banks are much cheaper. A bigger argument for keeping old units online would be for system inertia.

 

[waross]

I guess that you didn't understand my posts.
Yes, you can use a generator to generate VARs at the source, but that only reduces the losses of the generators.
VAR generation at the source does not reduce line losses nor reduce the losses of transformers downstream from the VAR injection.
The point I tried to make and the point of my last post is that theoretically, if your load is at unity power factor rather than the rated 80% or less, you may generate kW up to the KVA rating of the generator without overheating the generator.
But that only works if the prime mover has the capacity to produce the extra kWs.
In my example, we often had to run a second generator when the load went just above 600 kW, often for several hours.
We considered adding a capacitor bank so we could handle the load with one generator.
As proof of concept, we did our test when the load was over 600 kW.
We found that our prime mover was not capable of more than 600 kW.
As far as larger grids, I don't give a damm how big a generator is, you may improve the power factor to reduce generator losses, but you can't get more kW out than the kW that your prime mover puts in.
Often cities had local generation that they outgrew in time. Power was transmitted from remote generation sites and the local generation was mothballed.
It is those local generator sets that are a prime candidate for VAR generation
They are typically close to the loads where they will produce the most benefit.


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

 

[Fischstabchen]

Var generation at sources does decrease losses everywhere in the system. If you increase var generation, your increasing system voltage. If you increase system voltage, all of you capacitor banks increase their Var production by the square of the voltage increase. As system voltage increase, current flow for the same power flows decreases proportionally and reduces reactive var consumption of transmission line and transformers. Pull up a power flow software package and it is something easy to test and see.

 

[crshears]

As system voltage increase, current flow for the same power flows decreases proportionally . . .

Hmmm . . .

It's that "for the same power flows" that might be a bit of a red herring; although it is true that the formula for watts states that at higher applied voltages less current flow is required to achieve the equivalent amount of power, in real life the current flow of systems loads will tend to increase with applied voltage.

By way of example, consider load reduction by means of voltage reduction, namely, brown-outs; the rule of thumb used to calculate this is that for every 2% voltage reduction applied at the customer level, there will be a 1% reduction in primary demand.

I would much prefer to say that at higher system voltages the transfer limits and capabilities of circuits will improve, which is why independent system operators almost invariably choose to operate their systems near the upper acceptable limit of voltage.


CR

"As iron sharpens iron, so one person sharpens another." [Proverbs 27:17, NIV]

 

[Fischstabchen]

csshears,

What you are saying doesn't applying as much anymore due to more load like lighting shifting to power electronics or constant power. Previously, there were transient problems that solved themselves by the voltage dipping and reducing load thereby limiting the transient effect but much of the system doesn't operate like that anymore. With natural gas being cheap and abundant, even electric home heating is on the decline in a lot of areas.

 

[waross]

If you increase var generation, your increasing system voltage.

You may be able to sell that to an MBA, but an engineer will ask; Why don't I just turn up the AVR a little instead of incurring the cost and inconvenience of bringing old generators on line and maintaining them?
If VARs are added at the source I suspect that the reason is that the prime mover is overpowered and improving the power factor allows more kW to be produced.
I further suspect that the voltage is at the maximum allowable before the VARs are added and that the AVR trims the generator output voltage down to the same value as before the VARs were added.
My experience with paralleled sets is that all the parameters are in balance (it may not be the balance that you desire) and if you change one parameter, the other parameters shuffle around to maintain the balance.
You may not see this directly. With a grid connected set, the change may be accommodated somewhere else on the swing set.
I have worked in a couple of islanded plants and it is readily apparent that when you change the voltage setting or the throttle setting of one set, you see a corresponding reaction on the other set(s).


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

 

[Fischstabchen]

Warross,

That is how it is done. When a voltage schedule is set, the units are more often than not run above 1.0. Often 1.02ish. NERC doesn't allow non-contingency voltages above 1.05. Capacitor banks are a much cheaper form of VAR generation, though, and you don't run into any issues of needing to reduce real power output. Another perk that you get from running the generator voltage higher is that your generator swings less on disturbances and it is more stable, which may push it critical power output higher. If you go back to your equal area stability graphs, a high voltage means there is less angular difference for the same real power output.

Running generators to be just synchronous condenser is not a cheap way of putting VARs into the system. You run into the problem in that a large synchronous generator is one large unit so if it is out of service, you lose all of it. If you have problems with a set of capacitor banks, it can be isolated and you maintain some capacity and have some n-1 capability. I am not aware of regions that compensate for people to provide vars, either. The only time I have even heard someone even suggest a synchronous condenser was in a situation needed a tremendous amount of VARs and it needed to be continuously controlled or having a large number of steps to prevent any step from causing a 2% voltage change. The utility ended up going with a power electronic thyristor system.

Vars introduced to the system have 3 limits. Rotor and stator current capability and plant voltage needs. The step up transformer is tapped so that the needed range is gotten from the voltage regulating capability of the plant. I don't know why you keep bringing up prime mover when I have never seen the prime mover be a limiter on anything but real power output. The generator has the capability described by the generator capability curves and is sized to not limit the plant. It has more real power capability than your prime mover.

Voltage schedules are developed so that the plants work together and you don't have some plants lagging and others needlessly leading. Unless, you are an older unit that got grandfather in with var control, all plants have a voltage schedule and they have to notify the transmission operator if for any reason they can't maintain their schedule. I have heard of plants that intentionally ignored their voltage schedule to avoid any real power limitation that can be seen on their capability curve. The plant just scheduled 1.00 pf output to try to keep their numbers good.

 

[crshears]

Fischstabchen and waross, I'm getting confused; are the two of you agreeing, disagreeing, or just talking past each other?

One thing I'm not hearing crystal clear is the crucial point of whether you mean the "hard" actual limit on kW output of a given unit as opposed to its nameplate rated output, as these can differ; it seems Bill's reference is to learning empirically that the single islanded generator had a hard limit of 600 kW, rather than [ as was vainly hoped ] being rated at 600 kW but being capable of producing just a wee bit more real power than that.

Addition via edit: Bill, on that island and having a load of just > 600 kW for several hours, what did you end up doing? Would it have worked to use the UFRO principle and have all the clocks lose some time for a while, making it up later by running at a frequency above nominal to effect time error correction? Or would the fur have flown if you had tried something like that?

 

[waross]

Hi CR. Our operators checked and recorded parameters every 15 minutes. kW output, Volts, Amps, coolant temperature, and oil pressure.
When the KW output reached a given threshold, another set was started and put online. When the KW output dropped to a given value, A set would be taken offline. I was not on-site. If there was a problem they had to wait for me to arrive on the next scheduled flight from the mainland.

Would it have worked to use the UFRO principle and have all the clocks lose some time for a while, making it up later by running at a frequency above nominal to effect time error correction?

I haven't seen a synchronous motor driven clock for a long time. What I am seeing is crystal controlled digital clocks.
This was in the third world and our operators struggled with meter multipliers. There is no way we were going to make their job more complicated by counting cycles.
VAR location:
There is more than one reason to add VARs to a system. Reducing system losses is one reason. If the VARs are introduced close to the load, the saving in losses is the greatest, and the advantage in system capacity is greatest.
Consider power factor correction at a motor and consider the motor current versus the line current.
There may be a reason to introduce VARs at the source but the saving in system losses is marginal at best.

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

 

[waross]

I'm getting confused; are the two of you agreeing, disagreeing, or just talking past each other?

I am wondering that as well.

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

 

[bacon4life]

Fischstabchen- There are multiple examples over the last few years of US electric utilities removing the prime mover from a generator and converting it to a synchronous generator. GE even has a webpage advertising the ability to do the conversion.

Although shunt capacitors support steady state voltage, they are quite bad during transients. The reactive output from shunt capacitors is proportional to the square of the applied voltage, so the output from a shunt capacitor drops precipitously during a voltage dip. Synchronous condensers have basically the opposite characteristics; during a voltage dip, the reactive output from a synchronous condenser will jump to its maximum short term rating.

Although perhaps a tangent, my utility recently retrofited one of our hydro units to add the ability to operate as a synchronous condenser. This allow the unit to be be run as spinning reserve in a weak part of our system. It turns out to be much more efficient to run one unit at high load and one unit in SC mode than run 2 units at low load. During any large transient event, the SC unit automatically switches modes and picks up load within a few seconds.

 

[Fischstabchen]

Bacon4life,

It is in my experience that synchronous condensers on the grid are as common as quadrature boosting transformers. There are specific applications but I invite you to trackdown someone with a regional planning model and do a sort through system capacitor banks. They are few and far between and they are much more common in an industrial setting where an extra motor is sitting around as a spare or their was as a process change. The only times I have heard them actually be considered was in a case that had a transient stability issue that needed fixed var support and a large industrial customer that wanted the utility to provide all their VARs because they didn't have room and that came out to like 400-500 MVAR, that needed steps or control to prevent voltage flickering. I suspect that as poorly as vars spread through the system, there aren't that many sites in general with a strong enough connections to accomodate large VAR injections without causing overvoltages. VARs flow poorly through the system without causing problems compared to MWs.

I said this earlier but transmission fixes transmission problems. Generation doesn't fix transmission problems. Generation works with transmission but if you talk with them, their world ends at their fence. Synchronous condenser might be more prevalent if vertical utilities were still a thing. The way things are, transmission would have no incentive to seek out generation to convert a station to a synchronous condenser because it would void the transmission investment needed to resolve the problem. Most transmission companies are given a rate of return on investment. Not doing justifiable projects would reduce their investment size and returns. It is not unusual for projects to be justified or designed solely to increase the investment. Anyone talking about saving money doing this instead of that is living in a different world than where utilities exist. Industrial customers live in your reality of trying to get the best bang per buck spent.

 

[bacon4life]

Fischstabchen,
Within the Western USA, there are still many utilities that are substantially vertically integrated. Perhaps that is why our are experiences differ.

I just checked my regional model. The model has about the same number of Satic Var Compensators (SVC) and quadrature boosting (i.e. phase shifting) transformers. There are 5 times more SVCs that dedicated synchronous condensers. This could be interpreted that SCs constitute approximately 15% of the dynamic voltage control devices. Most of the dedicated SCs were installed by San Diego Gas and Electric within the last decade.

Currently our regional model does not contain enough information to easily detect which generators have the ability to disconnect the prime mover and run as an SC. For doing post-event analysis, it can be a very manual process to track down which operating state these units were in.

 

[davidbeach]

To greatly condense bacon, the West ain’t Texas.

I’ll see your silver lining and raise you two black clouds. - Protection Operations