Synchronizing generators of different ratings 2

[AusLee]

Hello,

I have bought a couple of books online based on the abstract and the table of contents, all claim to be about syncrhoneous generators but in the end they do not discuss in a spearate chapter the subject of synchronizing generatorsd of different ratings in detail.

If you have actually read a book or code about this topic, could you please recommend the litterature?

Thank you.

 

[dpc]

You might something of interest here:

http://www.basler.com/html/dwntech.htm#PowerGen

There is nothing inherently unique about synchronizing generators of different kVA ratings. It's done all the time.

 

[waross]

Hi AusLee;
If you set the base speed and the percentage droop the same on all sets, they will share the load in the proportion of their ratings.
If the droop is not set the same, you can parallel a set and adjust the base speed until the sets share the load as you wish. However, if the load changes, the sets may not continue to share well. You can have a condition where a decrease in load throws the load to the smaller set.
There are other schemes than droop to parallel generators, but for a small installation droop is often the most reliable choice. The size of your power plant and the type of load all make a difference in your choice of control schemes.
If you have a modern load-sharing control panel you should be able to program the size of the various sets into the load controller.
I hope this helps.
respectfully

 

[AusLee]

Thank you for your posts.

The link is good, but is one step ahead of me as i need a book that talks about "droop", which all my knowledge about it be from the local technician.

The books i bought are from CRC press and one is about Synchroneous Generators and one about Variable speed generators. Not a single match to a search for "droop". The Electrical Engineering Dictionary, Second Edition, also does not define "droop".

Are there books about this or it's that kind of stuff that i will have to get a manufacturer's training for?

 

[rbulsara]

See this:

http://www.toromontpowersystems.com/ConsultantsCorner/Engine Speed Governors.pdf

You may be better off web searching for governors such as Woodward and may get more information. Call their engineering support for references.

 

[dpc]

The droop that waross is talking about is related to the governor on the prime mover and not the generator per se.

On an ac generator, the kW output is a function of the prime mover. The kVAR output is a function of the exciter/voltage regulator.

 

[AusLee]

I know the primer mover must be rated for the kW of the load + overload + the radiator's fan and some safety factor. I have seen a 2 MW prime mover on a 1600 kVA alternator.

I think i put my hopes too high and got disappointed with the books.

Thanks again for your replies.

 

[waross]

Droop;
I just had a droop related problem. It is a governor setting, but it is the one that lets generators share the load.
What is droop? The amount that the speed changes or droops between no load and full load.
If you are familiar with PID controllers, droop is the same action as proportional band.
Typically, sets with a mechanical governor, and sets with electronic governors set to droop mode will be set to run at 61.8 hz. at no load. (51.5 Hz. in IRC land).
This is 3% droop based on a rated frequency of 60 Hz. As the set loads up, it slows down. At 50% load the set is running at 60.9 Hz. At full load the set is running at 60 Hz.
This is independent of the set capacity. When the sets are assembled and tested, they are load tested and the governor is adjusted to run at rated frequency at full load and 103% of rated frequency at no load.
The speed of the set can be calculated at any load.
When synchronising a set with others, the speed control must be used to synchronise the set. Once online, the speed control is then advanced until the set is taking the proper percentage of the total load. More sets may be added in parallel, and if each set has the speed set properly for the correct proportion of the load, it does not have to be adjusted again until it is time to shed the load prior to going offline.
I think that droop control is the most dependable and operator friendly control scheme for small islanded plants with untrained operators.

If you have to synchronize with the grid, then everything changes. You set your power output with the throttle setting. The grid locks your set in and sets the frequency and the voltage. Beware, now that the grid is controlling the voltage your field excitation controls VAR production.
If you set your voltage on the AVR to match the grid voltage
and the grid voltage drops, you will be over-excited and your VAR production will increase. This can easily cause an over-current condition on the generator. That is why power factor control is used in this situation. It allows the set to maintain the same KW and KVAR output with a fluctuating grid voltage.

Oh, my problem if you're interested. The set ran fine. when the power failed, the set started automatically and picked up the load. (A large service station and convenience store.) When the utility power returned the Automatic Transfer Switch put the load back on the mains and the set shut off. The next time the power failed the set wouldn't start. Instead of a no-load speed of 61.8 Hz. I had 64 Hz.
When the load goes off the set and back on the mains, the set has a little speed spurt before the governor can react.
With the extra 2.2 Hz. the speed spurt was tripping the overspeed safety as the set shut down. These standby sets have a narrow speed window between the overspeed shutdown and the Under Frequency Roll Off which starts dropping the voltage if the speed drops more than 2 Hz. below rated frequency.
Those big Cats are used on a wide range of sets. The governors are set to limit the maximum power out to match the alternastor on the back. I service a couple of Vee 16 Cats. The alternators are rated at 1250 KW. I think that the same engine can drive up to a 2000 KW alternator. I am sure that they are used on 1500 KW alternators.
If you spend much time around diesel sets, you get to know a little about diesel engines and a lot about governors.
Tell us more about your setup, AusLee.
respectfully

 

[itsmoked]

Nice little summary Bill. You answered a bunch of nagging questions I had and some I didn't know I had. Thanks.

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[waross]

You've done the same for me more than once Keith, thanks back at you.
Yours
Bill

 

[AusLee]

Hi waross,

Thank you for the nice information. I can tell you two things pertinent to the installations the contractor has installed:

1. In a 2 x 1600 kW installation now in service, i frequently see in the History of the Alarm list "G1 overspeed". As per what you said, it is due to the fact that the generator is running from full load to no load when the ATS returns the power to the public network. However, the set would start again automatically, no need to go manually clear the alarm. I only need to do a manual intervention when someone stops it with the Emergency Stop button, they like the red knob. So in this case, not a serious problem.

2. In another situation, a purchase order has been placed for one 500 kW and one 1000 kW sets needed to supply an installation of 1100 kW. I get in a lot of trouble if i believe the local suppliers, they say they can do things but they think that sending their people for training is a waste of money. Therefore i needed to ask if these 2 sets can be put in synchro (else i will divide the installation in two MDBs). From previous experience, i "heard" that when sychronizing sets, the droop mode must be turned off, set droop equal to zero, which in my IEC world means that the control of each set must fix the speed to 50 Hz. But as per your post, each set must regulate its speed to create a motor torque appropriate with its share of the load. Hence the two generators do not have the same frequency but they are in synchronism, which i do not understand.

In future contracts, i will put "the installation must be supervised by a specialist certified by the manufacturer and whose task on the project is limited to the commissioning and testing."

 

[waross]

Just a little misunderstanding;
In droop mode (at 3%), a set, regardless of rating, will run at 51.5Hz at zero load. It will run at 50.075 Hz. at 50% load and it will run at 50 Hz at full load.
The frequency varies slightly with the load, but both sets run at the same speed when in synchronisation.
Turning droop mode off. In older conventional governors, droop mode is the basic control setting. It is a parallel to a process controller with only proportional band control.
Autogenous mode is an added feature on more expensive governors. In autogenous mode the governor will respond to load changes with a slight change in speed determined by the droop. The autogenous feature will then detect the error between the desired frequency (50Hz) and the actual frequency.
The autogenous feature will then correct the frequency back to 50 Hz.
Autogenous is a parallel function to Reset or Integral on a PID process controller.
I had an interesting experience at a remote mill running on diesel-electric power. There were two paralleled sets. One was about 500 KW and the other was about 750 KW.
One had a hydraulic governor and the other had an electronic governor.
Both sets were needed to start the 400 Hp. hog motor.
If an attempt was made to start the Hog just after the sets were started, the electronic governor would respond much faster and hog the load. The breaker would trip on the set with the electronic governor.
If the gen-sets were started 15 or 20 minutes early, the oil in the hydraulic governor would have time to warm up and the response time would improve enough to work with the electronic governor.
When both sets were at working temperature the governors worked well together.
The nature of the hog operation was such that it would be idleing with no load and a large amount of waste wood would drop in (gravity feed) and the motor would go immediately to a short term overload condition. When warm, the governors were able to handle this condition without problems.
The governors were running in droop mode.
respectfully

 

[itsmoked]

I see where AusLee is coming from.

How does the droop method work when you have badly matched machines?

Like a 500kW teamed with a 100kW running a (let's say) 400kW load?

Does the lower impedance of the bigger machine help it shoulder 4/5 or a bigger share of the load? If it does share the bigger load why would it not try to run at a much lower speed?

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[davidbeach]

The two machines will run at the same speed, otherwise one will be slipping poles relative to the other and that's not likely. Given that they run at the same speed, then if they have the same droop characteristic they will share the load based on the operating frequency. That 400kW load would be divided with 333 kW on the 500 kW machine and 67kW on the 100 kW machine. Using waross' 3% droop, a nominal 50 Hz system would be running at 51 Hz and a nominal 60 Hz system would be running at 62.2 Hz.

 

[itsmoked]

Most interesting. Thansks

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[waross]

Hi David;
You said it better than I could. I think that the 62.2 Hz. should be 61.2 Hz., but I make too many of those little mistakes to quible. You probably caught it from me.
Thanks
Respectfully

 

[davidbeach]

You're right, it should be 61.2. Somehow the spell checker didn't catch that. ;-)

 

[waross]

A couple of comments on day top day operation of a small plant that may make the operation a little clearer.
The old plant that I was most familiar with had an operating sequence something like this.
1> The load is less than 300 KW, one 350 KW set is online.
As the load changes the operator may adjust the speed setting if he notices that the frequency is off 60 Hz.
2> The load is up to about 300 KW. Another 350 KW set is placed online.
A> The second 350 KW set is started and the governor speed control is adjusted until synchronism is achieved. The breaker is closed. Both sets are now locked together at the same frequency but the second set is not carrying any load.
B> The operator now increases the speed setting of the second set. The set starts to pick up the load. As it does, the first set is proportionately unloaded and both sets speed up slightly. The operator will probably reduce the speed setting of the first set, until the load is shared equally by both sets. Now both speed controls may be moved in the same direction if needed to get closer to 60 Hz.
From time to time, if the operator notices the frequency dropping a little with a load increase, he may increase the speed settings of both sets slightly to get back to 60 Hz.
With the load evenly distributed on both sets, the speed setting will be the same. Now the sets will share any load variations.
C> As the load gets up over 600 KW a 900 KW set is started, synchronized and put online. The speed setting of each 350 Kw set is reduced until it is at a no load condition and it is taken off-line and stopped. As this is being done, the speed control of the 900 KW set will be tweeked to maintain the frequency at 60 Hz.
D> The load is over 800 KW and climbing. A 350 KW set is started and synchronized. Total capacity on-line is now 350 KW + 900 KW = 1250 KW. The governors are now adjusted until the load is shared in the ratio of 350KW/1250KW to 900KW/1250KW. Both governors are tweeked to bring the frequency close to 60 Hz. The sets will now share load fluctuations in proportion to their ratings.
There is no indiction for exact speed setting. Typically there will be a three position, return to center switch on the switchboard. Turn to the left, reduce speed. Turn to the right, increase speed. Spring return to center, off.
The speed is adjusted by balancing the proportion of the load between the sets, and watching the frequency. The operators don't know what the no-load setting may be, but they don't have to know to do their jobs. They do know that the load is shared in the correct proportion, and that the frequency is close to 60 Hz. That's all they need to know.
The operators read and record a number of parameters every 15 minutes. Volts, Amps, Oil Pressure, Coolant Temperature etc.
An option is to use a crystal controlled clock and an old synchronpus motor driven clock and compare the times to fine tune the frequency. The small utility that I am describing didn't bother with this refinement.
By the way, when I first started encountering generators, I didn't see one often enough to justify the purchase of a frequency meter.
I have set the frequency many times with an old electric clock and the second hand on my wrist watch. Alas, the synchronous motor drive clocks seem to be all gone and I now make sure that any multi meter that I buy has a frequency function. I haven't used a clock for about 15 years now and, come to think of it, I haven't seen a suitable clock for about 15 years.
respectfully

 

[itsmoked]

Nice! Now in that imaginary world of societal collapse I will be able to spring into action and man the local power plant...

Now if I had more than a clue about transmission.

Thanks for that further glimpse.

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[matrixabc]

All the new synchronizing generator controllers will parallel any size sets together, via can bus connections.

I used to have to use droop on old machines, but you have voltage droop and speed droop, you also have to make sure they have the same winding pith on the alternator other wise you will get a reaction which will make them difficult to load share correctly , if you fit something like an Intelegen unit made by comAp, you will not need to worry about droop ever again