Reverse Power Engine Shutdown

The question and symptoms seems to imply that power sets with different natural/design frequencies are fed into a common distribution network.

Alternatively you may have a slow acting governor on the fast engine.

Can you please supply more detail.

Hendrik

There are two types of engine connected on the same bus bar and both have a frequency of 60 hz:

Type 1:

Total load of the site was: 25 MW

Load on high speed engines: 11 MW

Load on low speed engines: 14 MW

Low speed engine (600 RPM) (4 numbers - 4 MW each)

High speed engine (1800 RPM) (11 numbers - 1 MW each)

These engines are connected on the same bus bar, the high speed engine are connected to each others through load share cables. an over current appear on one of the outgoing feeders which cause the feeder to trip in addition this fault cause all the high speed engine to trip on reverse power and the low speed engine take all the load until the load shed trip 2 more feeders.

The question is why the high speed engine trip on reverse power and the low speed engine did not (why the low speed engines motorize the high speed engines)

Three more questions. I think I may have a clue.

Was the fault a single phase or 3 phase fault (remember that during a fault the voltage is at/or near zero and voltage is required to maintain synchronization of the machines)

What was the clearing time of the fault? (this is the time period that the machines have no means of synchronization and will drift apart)

Was it a close in fault (not much line impedance to limit the current, fault appears as a direct short to the machines)

Discussion:

Because the speed is different on the two different types of machines, the number of poles on the them is also different.

The angular rotation of the generator shaft position per complete cycle (16.66 milliseconds would be.

180 degrees for the 1800 RPM machines.

60 degrees for the 600 RPM machines.

During the time the fault is present, the machines were running random because the the fault is holding the voltage near zero and there was little or no synchronization voltage/current available to hold the machines in synchronism.

Suppose that during this time the machines drifted 45 (electrical) degrees out of phase with each other. (a normal value) .

When the fault cleared, the 1800 rpm machines would have to make 3 times the shaft angle correction than the 600 rpm machines. The 600 rpm machines (with the smaller required shaft angle correction re-synchronized sooner and won the race. The 1800 rpm machines could not cope with the larger shaft angle corrections required for lock.

This may not be the correct scenario but could be the basis for developing a better one.

Caution, This would only be correct is the above conditions are met.

Snakers

More support to comment #11 (i'm still thinking about it)

If the machines were generating their nameplate rating ....

The 1800 rpm machines were generating 11 MW

The 600 rpm machines were generating 14 MW, and were the dominate power source.

The 600 rpm machines would have at least 3 times the rotation mass of the 1800 rom machines and be much slower responding to torque changes. (you have got to have that much more iron and copper for the slower speed, figure it out)

If the fault sensing relay took 2 cycles to respond, allow a half cycle for the DC trip circuit to make up and another 3 cycles for the breaker on the faulted line to clear. thats 91.66 milliseconds.

During this fault clearing time ......

The machines are running with little, if no synchronization voltage.

The shafts of the 1800 rpm machines will have rotated 990 .3 degrees

The shaft of the 600 rpm machines will have rotated 330 degrees

A shaft error of more that 45 degrees (on the 1800 rpm machine) means the machine has slipped a pole. It would be history from there.

Come on some body.. Shoot this idea down. Have I done something wrong?

Snakers

You'll need to open a new query for this to discuss fresh. Most people reply once, and forget this. Some do not check the box to be notified of new posts so they don't read your challenge.

Hola Snakers,

I am in the large engine power generation business (1 MW up to 10 MW each engine; 1800 RPM 4 pole - 600 RPM 12 pole ) so just by chance I am buying some large generators now and asked this (your) hypothetical question of reverse power from purely a power generation point of view (forgetting drivers). Poles and such. I spoke with the chief designer of a very well known manufacturer of 2 - 50 MW generators. There is no phenomena that will cause this from the generator. It is likely driver inertia and governor response as discussed several times. But I like the way you think !! Good try (and it still might be true since 50% of all electrical engineers designing generators graduated in the lower half of their class!!)

When I was a little kid, I had large ears just like in your photo, The other kids used to tease me and called me Cabbage Ears. As I grew up, the size ratio of my head and ears became more realistic and I also became strong enough to beat the XXXX out of them.

As far as the theory that I have proposed, it really is realistic and has happened a number of times. It takes a close in 3 phase fault where the voltage is forced to near zero and the machines lose the required voltage synchronization signal.

Even though the machines are delivering very high currents, it is very highly reactive and contains little power. The machine will speed up (like when you close off the suction on a vacuum cleaner.) The famous North East Blackout back in the 70s, is an example when the Niagara Falls machines twisted 180 degrees out of phase between Rochester and Syracuse. This appeared as a short circuit to the protective relays between Rochester and Syracuse and separated the Power Authority's West system from the East. This included New York City.

There never was any actual short circuit and no equipment damage except for the Big Allis (1000 MW) machine owned by Con. Ed. that wiped the bearings on coast down because there was no power available to run the lubrication oil pumps.

But as you say, half of the graduation class was in the lower 50 percent. When I was in school, I was a co-op student and also worked in power generation at a local power utility. (they later hired me full time). From the text-book to the practical stand points, some of the instructors may also have been in the lower half.

I'm not going to tell you what grade I squeeeked through in the math department.

Interesting stuff actually. You should make a new discussion topic.

Ears: They told me from behind I looked like a taxi cab with the doors open !!!

If one looks at my profile we can find that I am not qualified for academic discussions. I fetch coffee in such meetings, not participate!

Thanks

The big flywheel of the slow engine will result in a smaller effect on the speed. The revs on the fast engine will howver drop sharp.

Greetings,

WR2 or rotating mass, as mentioned the low-speed typically will have a flywheel to even out the power flow, whereas the high-speed relies on its response to a load change.

The Voltage Regulators and Governors droop could be set to compensate for the difference in response, protective relays dumbed down, but generally speaking it did what you want it to.

I had a similar situation with smaller generators and the grid on a hydro start-up where a 3MW synchronous hydro and two 1MW induction hydro in a common powerhouse and the incoming grid. A tree branch fell on the utility incoming grid tripping the line, the two 1 MW each blew a 1 watt resistor in there over-current relays and the plastic face of some meters blew off about 5 feet across the floor, as the utility closed back in the grid with the fault present and the 3MW trying to idle along. The utility had not repaired a meter that would have prevented them closing into the shorted phase line. Voltages were way above normal...

After 10 years of start-ups you collect a few stories.

Dan

The system inertia, speed of response and system configurations make this a complicated problem to analyze. What you experienced was a miniature black out and would need to be computer modeled to provide your answer. Even then, I doubt that you have enough information for a valid stability study.

I assume that your protective relays did there job and there was no damage to your generation equipment.

Trivia..

I would mention that these problems originated when Tesla introduced AC systems and in the old days these studies was done by hundreds of hand calculated slide rule iterations.

Large grids of today utilize on line computers that constantly track the system stability limits. They can alarm and in some cases take direct action.

I have seen large (slow turning) Hydro machines stay on line long after the steam turbines tripped off./

Another comment, Do you know that a generator can speed up on a close in short circuit.

Remember that with a short circuit you have zero volts at many, many amperes. but regardless of the number of amperes, the power (watts) is volts times amperes and Zero times any number is Zero i.e. No Watts, No power out from the generator. The generator speeds up.

In Actually the high current becomes highly reactive. (Watt less Power)

I'll bet this one gets a lot of comments.

Zero Volts = Zolts and it is dangerous.

As a boy I was friends with the operator of the power station (source of useful scrap). The station had some 600 rpm and 2400 rpm engines.

He had a bucket of water with 1 or 2 electrodes coupled to the system.

He explained it but I forgot the detail and the purpose.

I am not in electricity but I want somebody to explain the possible effect of capacitance and phase shift on the problem of najialassal.

Try examine the droop setting of each one of the two generating sets.

Each of them have to match with the actual load conditions.

Many times, the droops variate due to load variation. This occurs when load frequently varies during a certain period.

In this case, there is a possibility to acquire an automatic droop regulator. This is normally expensive, but for these cases, you will be adequately served.

One manufacturer you may contact is Woodward. They have a good engineering team to suport your doubts.

Regards

Sometimes we can apply the KIS rule; Keep It Simple. Sure there is inertia issues with large engines and reciprocating mass, WR2, etc. People think that the flywheel on a 10 MW engine is huge . . . . but you know that for years now we use 'thin flywheel technology'. I can show you a 10 MW flywheel about 1.25 wide. The real inertia is from the HUGE generator rotor and reciprocating masses inside the engine. Also, high speed engines are highly turbocharged at pressure ratios exceeding 85%. This makes them extremely sensitive to load changes, including load dumps.

But back to the KIS rules. The HS engines were probably set up and supplied by vendor X (Cat? MTU?) and the low speed set up by vendor Y (Wartsila, MANN, ?) and they probably have different reverse power relays with different set points.

Now if that's not it, then we are back to high BMEP, highly turbocharged reactions and Woodward makes a black box that senses load upstream (electrically) of the speed difference and sends a signal to the governors a few milliseconds ahead of the actual speed delta loop reaction.

petro power:

the main question is the reason why the engine trip on reverse power, what is the cause that lead them to trip, even if they have differnet setting of reverse power relay.

Reason of reverse power:

One engine has a power bias caused by a speed bias and a governor pouring some additional fuel into a particular unit giving that unit more power availability (by a ratio of its share of the load) when paralleled to another; i.e. one engine is motoring the other and the power is being consumed trying to rotate the other machine (motoring) instead of going out to the load.

This is why each unit paralleled must react under load in the same manner individually so one machine is getting more 'power' capacity by a governor not adjusted for the same load bias and one unit starts to motor the other (reverse power going between units instead of out to the bus).

So, you probably have automatic load sharing governors, so the speed droop bias under load is eliminated by each governor talking to another, even for grossly misized units. i.e. we can load share perfectly a 1 MW with a 10 MW.

So when your feeder breaker tripped, you INSTANTLY unloaded all units that were under load. But they are all still paralleled on the line side of the feeder breaker (available to motor each other if there is a large speed [power availability] bias between units). The speed went up for a few milliseconds and individual governors that were pouring fuel to each unit underload suddenly reacted to a instant shock speed increase (due to instant no load situation) and the engine governors all reacted as fast as they can, but, as small high speed highly turbocharged units have a much different reaction to load changes (more sensitive) they were easier to 'motor' and you tripped out as the reverse power relay detected power going between machines instead of out to the bus. i.e. the bus load is now zero amps as you tripped the feeder breaker, but as the governors hunted with each other for 8-25 milliseconds before they 'caught' themselves and settled down, amps were flowing between generators motoring one another. The easiest ones to motor trip out on reverse power.

It's like a 100 kilo strong fit man and a 40 kilo boy lifting a box where each take their proportional share of load, are tied together by a taught rope around their waist. Then they throw the box together to the side. The big man will maintain his footing after release. The boy will fall forward half a step before he catches his self as both react differently to load changes and their own mass, reaction responses, inertia, stability, etc. The taught rope jerks the boy back. The boy doesn't jerk the big man forward. So the boy is 'motored'.

I think I see what is happening here. The load tripped the line (went above 25MW power consumption) correct? That means that all the motors were running at full capacity when the fault occurred.

The bigger motors turn at fewer RPMs than the smaller ones because that's the general nature of a motor. When the trip occurred, the large motors sensed the sudden load loss and tried to slow themselves down, probably with the use of a governor. The governor reacted to quickly and tried to slow the big motors down to fast because the little motors governor didn't sense the load drop yet. Under this condition, the smaller motors will appear to drive the bigger ones until the little ones realize what happened and either trip themselves or until they can start to re-power the line (fault cleared) and the big ones realize they are needed again. Although, this looks like reverse power, it's really not.

This was probably done to maintain synchronization of all the motors so they can get back to work faster and not need to be re-synchronized.