power sharing 2 generators 4

[zekeman]

I would like to hook up 2 generators (1000W and 3000W)
to share power in proportion to the size.

I would appreciate any suggestions, caveats, etc. in this regard.

 

[wayne440]

Sell both of them and buy a 4000W set.

 

[electricpete]

I will mention some basic stuff. I'm sure others can say more.

Real load sharing among generators will be dependent upon the speed vs power characteristics of the prime movers and their governors. Part of this is the speed droop (how much does speed drop as power increases). Another part of that is adjusting the no-load speed.

If both prime movers have the same no-load speed, and the same droop at their respective full load, they should share load in proportion to their rating.... assuming there is nothing else tied in. If there is a grid involved, then adjustment of both no-load speeds relative to grid will be a factor in determining load sharing with the grid.

=====================================
(2B)+(2B)' ?

 

[electricpete]

Caveats - don't do this unless all aspects (protection, fault current etc) have been thoroughly reviewed by someone with more knowledge than me.

=====================================
(2B)+(2B)' ?

 

[byrdj]

all you have to do is plug one into the other, that is, if they are both these

http://www.hondapowerequipment.com/products/generators/content.aspx?asset=gg_parallelcapability

 

[electricpete]

Byrdj - your comments are probably more on target than mine.

The first time I read the post, I was thinking 1000KW and 3000KW.

Now after I read yours, I checked again and see it's 1000W and 3000W.

=====================================
(2B)+(2B)' ?

 

[waross]

I have worked on a few parallel generator installations. For that size, follow wayne440's suggestion. Buy the proper sized set.
Pete forgot to mention synchronizing gear which may cost more than trading the sets for the proper sized set.
If you have Honda inverter sets you can try byrdj's suggestion. But, reading the instructions, after the smoke clears from connecting dis-similar sized sets together, then go and buy the proper sized set.
Another option, which may be cheaper, is to split the load between the sets and run them independently.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[dicer]

After connected they will keep themselves synced.
You would need 2 volt meters and make sure you have zero volts between the lines before they are connected.

 

[MiketheEngineer]

Or maybe split your loads??

 

[burnt2x]

Good point MiketheEngineer! A star for you.

 

[zekeman]

Diced,
The gens h ave avr control of voltage which precludes getting the amplitudes the same but it is possible to get the frequency and phases the same.

So, I would think the possibility of syncing the gens is slim. What is your take on this?

 

[catserveng]

Paralleling two generators this size has a lot of drawbacks.

The act of syncrhonizing, matching phase and voltage as close as possible and being able to close the breaker of the oncoming unit quickly enough to assure you actually close while in phase can be done. Depends a lot on how stable your units are, do you have any ability to adjust voltage or speed and regulate it well enough to get a close that won't damage either machine?

OK, so you come overcome all the issues that come with synchronizing two units and get both breakers closed, now what? Is the speed and voltage droop the same for both units? Can it be adjusted? The ability to share both real and reactive power now becomes the issue. For larger units a number of options are available, for units this size, with fixed governors and voltage regulators, not much you can do. You might get lucky in certain load conditions, but it is highly unlikely you'll be able to properly share the load between the units without possible damage to either generator and possibly the loads.

I like the idea of either selling them and buy a right sized unit or split the loads. Unless you have something like the new Honda units that have the built in capability to parallel units, I think paralleing units this size is not feasible.

My 2 cents, Mike L.

 

[waross]

On the larger sets, the governor is a proportional controller. (Some add integral but that does not concern us here.) The no load speed and the speed droop between no load and full load may be accurately adjusted.
When sets are to be paralleled the speeds are matched. If there is a load on one set, then after paralleling the speed setting of the incoming set must be adjusted up to pick up a share of the load. In operation both sets will have the same no load speed setting and the same droop setting. The voltages must be adjusted to properly share the VAR demand.
It is common on the small sets to use a vane in the air path of the cooling fan for a governor. It works pretty well for a single engine but setting the droop accurately may be a challenge. Accurate droop setting is a must for proper load sharing. There will be droop but it will depend on the dynamic drop in cooling air velocity as the engine speed drops and on the linearity of the return spring. Linkage geometry will also have an effect. Dirt buildup on the control vane and in the cooling circuit may have an effect on droop.
I doubt that your voltages are AVR controlled. Most AVRs have voltage adjustments. Many small sets use capacitor excitation. The voltage is raised by the addition of capacitors. In addition, loads containing capacitors such as metal halide ballasts may interact with the excitation system and tend to raise the voltage. If this is not the same for both sets, VAR sharing will be affected.
Although the Honda sets may be paralleled, the sets must be the same size. Here we have two dissimilar sets. Possibly a different brand of inverter type set may be paralleled.
What you don't know won't hurt you, it may kill you.
In this case you may not know about VAR sharing but improper VAR sharing may set up circulating currents which will quickly kill your generators.
Come to think of it, improper load sharing due to mismatched droop characteristics may also kill those little sets.
You may get it to work but be advised; Those of us who have experience setting speeds, droops, and voltage and cross compensation circuits for VAR sharing, shudder at the thought of paralleling machines where droop and voltage are not easily adjusted.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[zekeman]

"I doubt that your voltages are AVR controlled..."

Believe it or not, indeed, they are and that is the problem as I see it, so the 2 gens will be fighting for voltage control...

So, in view of all the great responses I will not try this.

 

[Smokerr]

Good plan!

Hondas as mentioned if you have that need to vary the load so much that at times you want all of it and others when you can reduce to either one of the two seperte units larger or smaller one off.

Otherwise a single larger unit.

 

[MiketheEngineer]

As a structural engineer who knows little about EE - how do power plants (big ones) run multiple generators??

 

[Bramb]

Hi Mike,
big power plants are typically connected to the transmission grid: in this case there isn't a "real" load to be shared between the different generators of one power plant.
Each generator has its own active and reactive power setpoint when it is in parallel with the grid.
The generator control system acts on the AVR and on the prime mover (turbine) speed controller in order to keep actual reactive and active power values close to the setpoints.

If the power plant is partecipating to the grid frequency/voltage regulating program the setpoints are "chosen" by a national (or regional) central control system which is communicate the setpoints to the power plant by means of digital remote communication. The algorithms which run inside the national (regional) central control center realize the load sharing between the different power plants calculating the P/Q setpoints in order to get nominal frequency (which depends by the balance of the active power in the grid) and voltage (which depends by the balance of the reactive power in the grid) values over the network.

If the power plant is not partecipating to the grid f/V regulating program the two setpoints are simply a fixed active power value and fixed cosphi.

In case of industrial power plant connected to an external grid, the concept is pretty the same: the setpoints are fixed in order to have the generator(s) running at the maximum active power (typically is cheaper to feed the industrial loads with the power plant than to buy energy by the grid) and nominal voltage.

In case of industrial power plant running islanded (not connected to any grid) there should be a power plant control system with load sharing facilities (or a dedicated device) which calculate the setpoints of the different generators in order to share the total load avoiding power swings between the machines.

Ciao
Erminio

The difference between overload and short circuit lies in the nature of the fault, not in the value of the current.

 

[waross]

How do big generators work together? No problem, all you need is money. A small engine may use a flap in the cooling air stream and a spring. The large sets have sophisticated governors that cost a lot more. When you are running plants that cost hundreds of thousands of dollars (for a very small one) and well up into the millions for larger plants, the cost of the control gear is negligible.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[TheBlacksmith]

And a careless operator can still get it wrong and do amazing amounts of damage to windings, rotors, couplings, prime movers, breakers, etc.