Power Grid Synchronization and Reactive Power

| Once a generator is connected/synchronized to a grid and everything is | running normally. Can there be a reverse reactive power flow from the | grid to the generator? What would happen in cases of reactive power | flow? How is this prevented/protected? Is this a part of | synchronization or does it come under transmission?

This is something that reportedly happened (reverse reactive flow) during the 2003 blackout in the midcentral to northeast US.

Without more explanation, these posts do not make sense to me.

Reactive current output from the generator can be either positive or negative, depending upon the power factor of the load. Whichever it is, it consists of power flow from the rotational energy of the shaft into electrical output for about half a cycle and return of electrical power to the shaft half a cycle later. If this exchange is large enough, ohmic heating in various windings can become large enough to cause damage.

In any even, such reactive (non)power is normal, albeit costly, in power systems. I can picture situations when such exchange can increase out of control. But even real power can exceed safe operational limits.

Ext question?

Bill

-- Fermez le Bush--about two years to go.

Yes there can be a reverse reactive flow. This is not a problem as long as it is within the capability of the generator*. It won't affect the prime mover. Reactive output of the generator is controlled through the excitation system -trying to raise voltage will increase reactive generation (lagging pf when seen as a generator) and lowering voltage will decrease reactive flow or make it negative (leading pf). Often an unloaded synchronous machine is run as a motor and its reactive adjusted as desired to maintain a given voltage at its terminals or at nearby loads.

*The capability or allowable operating range is different in the lagging and leading regions as the limits are dictated by different factors -VA and field current limits in the lagging region (+var output) and VA and a "stability limit" in the leading region (low field current, -var output) and is lower in the leading region.

On Thu, 22 Mar 2007 12:59:04 -0700 Salmon Egg wrote: | On 3/22/07 9:40 AM, in article snipped-for-privacy@news1.newsguy.com, | " snipped-for-privacy@ipal.net" wrote: | |> On 22 Mar 2007 07:23:04 -0700 snipped-for-privacy@yahoo.com wrote: |> |> | Once a generator is connected/synchronized to a grid and everything is |> | running normally. Can there be a reverse reactive power flow from the |> | grid to the generator? What would happen in cases of reactive power |> | flow? How is this prevented/protected? Is this a part of |> | synchronization or does it come under transmission? |> |> This is something that reportedly happened (reverse reactive flow) during |> the 2003 blackout in the midcentral to northeast US. | Without more explanation, these posts do not make sense to me.

What I read was that the grid was actually looped in a circle involving portions in Canada, and that real power was flowing along a different path than reactive power, which resulted in some portions of the grid having real power going in one direction and reactive power going the other way over transmission line(s). How that could get established, I do not know. I do not recall how long that situation existed, but what I do recall is that the catastrophic failure took place within 2-3 minutes of this. What I wonder is what caused this. Could it have been a change in frequency that had happened just before? Could it have been some voltage or load spikes seen earlier from such things as loss of a transmission line due to a sag fault into a tree. I guess you'll have to google up all the reports (I didn't keep the URLs) yourself and use your professional understanding to make better heads and tails out of it than I possibly could (electricity is mostly a hobby for me, not my profession).

| Reactive current output from the generator can be either positive or | negative, depending upon the power factor of the load. Whichever it is, it | consists of power flow from the rotational energy of the shaft into | electrical output for about half a cycle and return of electrical power to | the shaft half a cycle later. If this exchange is large enough, ohmic | heating in various windings can become large enough to cause damage. | | In any even, such reactive (non)power is normal, albeit costly, in power | systems. I can picture situations when such exchange can increase out of | control. But even real power can exceed safe operational limits.

What about a situation where power demand is very high and more power is being drawn from longer distances (say far into the midwest or central parts of the US). How well does reactive load pass over that distance? Could it be that closer generators had to take on more of the reactive load relative to their power production?

reactive) can flow in or out of that generator. You can not separate the reactive power from the active!! both powers use real current amperes!!

The reactive power is always a component of the total power. So, yes you can have reactive power flowing from the grid to the generator, and that means that the current reverse direction and starts flowing from the grid to the generator, converting it into a motor!!!

That would trigger all king of protective devices, etc . . . . and that has nothing to do with the synchronization process when you connect the generator to the grid. This can happen when the voltage of the generator is too low vs. the grid.

As mentioned by other reply, the reactive power out of the generator is controlled by its excitation, which vary the output current without effecting the real mechanical power driving the generator.

--------- I don't have the info on the '03 problem so I won't hazard an analysis. I'm sure that, just as with the '65 outage, that IEEE (PAS) has something on this. Reactive flow and power flow are not necessarily in the same direction. For example a transmission line may have a real power flow in one directions and at the same time, depending on terminal conditions: a) + reactive into the line at each end b) + reactive out of the line at each end c) + reactive into the line at one end and + reactive out at the other end This is determined mainly by voltages and line inductance and capacitance.

As far as drawing reactive from a distant source- can be done but there are problems in terms of voltage regulation (also in some extreme cases , voltage collapse- not good.) and a rule of thumb would be "generate reactive as close to the load as possible") Hence the use of capacitors or other capacitive sources to counter inductive loads. Closer sources will tend to take on more of the reactive load. The problem is that, if some sources go too far lagging (+ reactive as a generator) and exceed system reactive needs, then other sources may go leading and suck vars. This will require a reduction of internal voltage generated (i.e reduce field) and a reduction in the maximum power transfer- reducing stability limits. Push it to far, and a bump happens and it is time to get out the candles. It is best to try to keep all generators near the same power factor or, at least, all of the same sign of reactive output.

---------------------- Sorry, Al, I disagree strongly. Your last paragraph below

"the reactive power out of the generator

is correct but doesn't jibe with your statement above.

"Reactive power" is a consequence of having inductive or capacitive loads. The average power due to this is 0. The only effect on "real" power is the effect on (I^2)*R losses which increase as one goes either lagging or leading -. "reactive power" because its average is 0 will not affect the prime mover except for the losses which will always be an additional load on the prime mover and cannot make it motor. Only a reversal of the real power will try to do this. You do have real amperes and you do have real voltages. However, you also have real phase differences. --

Don Kelly snipped-for-privacy@shawcross.ca remove the X to answer

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

REACTIVE POWER IS ZERO!!! What you call reactive power is an exchange of power at a rate of 120Hz on a standard 60Hz. IT AVERAGES ZERO. The current associated with this does heat conductors by R*I^2 losses. That loss, however, IS REAL POWER.

Bill

-- Fermez le Bush--about two years to go.

Thank you so much, everyone, for all the information.

Thanks, LS

No,no,no!!!!!

It becomes a motor only if the total power summed over all phases is into the machine.

In any given phase, the instantaneous power will have a positive value sometimes during a cycle and a negative value at other times depending on the phase relation of the voltage and current. At 1 pf there will be no negative portion of the power waveform. At 0 pf the power waveform will be centered on the 0 power axis and will be positive during 2 quarters of the cycle and equally negative during the other 2 quarters of the cycle. (The product of the 2 sine waves is a double frequency sine wave.)

If you calculate the instantaneous phase power as the product of the instantaneous voltage (L-N) and the instantaneous (Line) current, then add the 3 values to get the total power, you will find that for a *balanced* 3 phase system (voltage and current sinusoidal and equal from phase to phase but displace by *exactly* 120 degrees from phase to phase) the total power will be absolutely constant regardless of the power factor. Introduce *any* unbalance and the total power will have a little ripple at 2x operating frequency. (And the shaft torque will have a little ripple at 2x electrical frequency.)

If all phases are at 0 pf the total power is 0. The machine simple shuttles what may be large amounts of instantaneous power from phase to phase. The turbine supplies the losses and excitation. If it doesn't, then the pf won't be 0.

The machine does not motor until the total (sum of all phases) electrical power out of the machine is negative at the same instant! It doesn't happen in a balanced system. To actually be a motor the input electrical power supplies the losses (including excitation) and any mechanical load.

I know that, and I should have said "reactive volt amperes" but the term "reactive power" is commonly used (real, reactive, and apparent power terms and the "power" triangle were established before you and I were around) and, as you indicate, are not strictly correct.

What I did say is (note the quotation marks):

" "Reactive power" is a consequence of having inductive or capacitive loads. The average power due to this is 0. The only effect on "real" power is the effect on (I^2)*R losses which increase as one goes either lagging or leading -. "reactive power" because its average is 0 will not affect the prime mover except for the losses..... "

This seems to be in line with what you said.

I agree--we agree.

In a sense, consider power in a parallel resonant circuit. There is lots of circulating power, but very little is being dissipated.

Bill

-- Fermez le Bush--about two years to go.

I do agree with you Fred!!! and the original question did not mention any imbalances nor instantaneous conditions. I also know that, way before the grid could drive a generator in a power plant, multiple protective devices will be triggered to stop that from taking place . . . .

Actually, my experience in power-plants disagrees, --slightly--. They *do* motorize, *then* protective relaying trips them.

Most protective relaying schemes for generators include a reverse-power trip (ANSI device code number 32-). These activate to trip the generator output breakers when power flow reverses in the unit, after a short time delay.

A typical large steam turbine-generator unit (rated at say, 800 MW) will reverse power trip when power flow reaches about -40 MW after about 15 to 30 seconds.

On many trip signals, the turbine-generator undergoes what's known as a 'sequential trip'. The turbine is tripped for some reason (low oil, thrust wear, loss of vacumn, etc...) and the generator is not tripped directly from the same signals. Instead, it is tripped later, ('sequentially') by the generator's reverse power relaying as I explained above. This prevents the residual steam in the separators / reheaters from overspeeding the unit.

Westinghouse came out with a notice about one series of reverse power relays they have that warns if there is a very low power factor on the unit, it may not operate at all. This could happen if there was a small lagging power factor (say, about 0.9) when running at full load, then you lose the steam supply. The MW will drop to zero and the unit will begin motorizing, but the MVAR load may not drop enough to avoid the very low, non-tripping pf condition.

In either case, operators are trained to look for it and prepare to trip the unit manually.

Diesel-gen units also have a 32- device and although I can't speak from direct experience, I expect gas turbines do as well (probably hydro as well).

Motorizing a large generator isn't particularly risky for the generator if it's already synchronized. It can take such operation indefinitely. A couple of old steam units in a oil-burner near me have disconnected the turbines from the generators and run the generators as 'synchronous condensers' often. This is essentially 'motorizing' the generator 24/7.

The risk is to the 'prime mover' (turbine or engine). Steam turbines don't mind it for a short time, but with little/no steam flow, the blading can actually overheat (windage/friction). And if an 8 or 10 foot long blade grows just a small amount percentage to heating, clearances can be reduced and a rub occurs. And that's a bad thing.

Diesel-engines don't like being driven by the shaft, but I'm not exactly sure what would fail first inside the engine.

daestrom

----------- Ah, the problem is not the reversal of the energy flow as far as the synchronous machine is concerned (or even an induction machine) but it does exist in attempting to turn electricity back to coal or oil (wind and BS are not problems -any government can do that)- which is something that science or engineering has not been able to do.-but hey, physical limitations shouldn't get in the way of "what if" concepts. Shame on you for using common sense!

reverse-power trip

..

generator if

I'm not a protection designer, but for the mid-size hydro units that I've seen protection schemes, none of them have reverse power. There's even a stunt called "synchronous condensor" mode in which you close the intake gates and blow the water out of the unit with compressed air, to run it as a synchronous capacitor...since sometimes not all the water is out, a little reverse power flows.

Bill

Probably more like -4MW. I really don't think you can get to 10 MW let alone 40 unless you break vacuum on the turbine. (Maybe not even then.)

40 MW is a lot of heat to put into the last few stages of a turbine. Even taking out generator no load losses of say 8MW, 32MW is a lot. (I'm thinking of 2 pole machines here. Not really calibrated to 4 pole.) (Are you thinking of cross compound 2 shaft nukees?)

Dead on truth. ANY MECHANICAL TRIP SHOULD BE SEQUENTIAL!!!!! We used a logic string from valve limit switches to 'complete' the trip sequence.

I never figured out what to do if a turbine valve sticks or, more likely, one of those fussy limit switches fails to close.

I hate reverse power relays. Up until the 80's my company used turbine exhaust hood temperature to trip when (if) motoring. Sometime in the 80's the relay manufactures sold a bill-of-goods to our planning and relay department and 32's started appearing. After that, the operators were so afraid of setting off the 32 that they'd trip at up to 10% load when taking a machine off line. Someday there will be a stuck valve and turbine parts will joint the debris cloud in orbit.

Absolutely true. (I wrote the 'motoring' section for IEEE Std 95 in the '80s) A friend of mine on the same committee wrote the following section on 'inadvertent energization' which is the most serious form of 'misadventure' (his really great British word for it) that can befall a large generator.

I've seen 2 rotors from generators that were inadvertently energized (and pictures of a few others). One, a 72MVA 4 pole was taken from 0 to about 300RPM in 'a few seconds'. It had shaft mounted retaining rings. Any place that wedge had 'crawled' out and touched the ring, it was burned by a heavy arc. Some wedges were also burned. 3 weeks down time to chemically etch the burned metal.

The second was a 800MVA 2 pole that went from 1200 to about 3400 when the generator breaker flashed over during shutdown. It was motored for about 5 minutes while the plant got the dispatcher to trip the far end of the line. The rotor body was discolored and there were little stalactites of metal at the ends of the cross pole slots. Some wedges were just starting to come out of the slots. The breaker was open as far as the controls were concerned and the 'fault' current wasn't that much different from normal load current so all the normal protective schemes were blind to the problem. Nobody wanted to crank open a 345KV switch under load. I don't blame them. I don't know how long the outage was.

I did a couple studies to try to convert some old (1930's) generators to synchronous condensers. The sticking points were:

1. How to start the thing (these were 62.5MVA 4 pole turbine generators)
2. The thrust bearing is in the turbine.
3. How to do it all really cheap.

If I had a dollar (as my dad used to say) for every time I said those words, I could have retired a couple years earlier. The risk to an eng> I'm not a protection designer, but for the mid-size hydro units that

A pumped storage plant that I helped place in service about 40 years ago actually has a scheme to depress the water out of the turbine with compressed air (also mainly used during pump starting) and run the rather large machines as synchronous condensers. I don't think it's ever been used that way. You do have to supply the machine's losses (F+W, excitation, core, I^2

*R) which may be 0.5% of rating. You also have to supply water to the turbine seals.

I was at the plant once during a pump start when the dispatcher asked us to hold off loading until the hour (about 15 minutes away). I asked the operator to stop the sequence with the machine synchronized to the system and the pump/turbine spinning in air (about 10 MW load or so, nearly all generator losses). At 7 minutes til we started venting air. At about 3 minutes til we had

50MW load (pumped primed), 3 minutes to open the guard valve and as the turbine wicket gates opened we went from 50 to 250MW in about 5 seconds, right on the hour. When the plant was first started, everybody was afraid that we'd boil the water with that 50MW waiting for the guard valve to open, but in 1000's of pump starts it's never happened. I never figured out why.

Good point. The idea is that it isn't the *generator* that cares, it's whatever is driving it. Some hydro 'turbines' are even meant for this (think 'pumped storage').

daestrom