Problem Load Sharing with 2 Small Gensets

You haven't said why it is a problem that needs solving.

Synchronise the Gensets on No load, check the circulating current (Current flowing between the alternators), if it is more than 10 % of rated Current of one alternator (Thumb Rule), you need to check the terminal voltages of both alternators and make sure that they are SAME. Alos check the speed og both primemovers, it should be SAME.

If on no load the circulation current is within 10 % of Rated Current and the circulationg current is increasing with increasing the load, then you have the problem with Load Sharing.. Check the Voltage V/s Load Characteristics (It should be drooping) of both alternators.. it should be same, Similarly check the Speed V/s Load Characteristics of the Prime Movers...it should be very identical.

1. Both alternators are of same make and model ? if no then please compare the phase angle difference by connecting voltmeter between R Phases of both alternators. The voltage should be zero.

All the best, Please share the findings.. I have not attempted sync of such small gensets..it would be great learning..

For 120VA it hardly seems worth any effort.

Is this a time scale adjustment error or are you trying to make two generators with 200% difference in output frequency to... sync? Or are you measuring neutral current? (Which is not much usefull most of the time.) Anyway in unloaded not perfectly matched "synced" gens at any given moment minor phase differences, (usually monotonic, but that depends on the controller) exist, hence the "zero load" current distortion. S.M.

Nice experiment, way back in time BDC (Before Digital Computers) we used to do the same thing for for loadflow and stability analysis, it was called an "AC Calculating Board", only difference was the use of high frequency generators to keep the circuit elements small.

What your experiencing looks like harmonics. Are your generators exactly matched, do you know the saturation characteristics of your CTs? If the generators aren't perfect twins then there will be incremental differences in the waveforms of each one, when added together when synchronized those differences cause incremental circulating currents between machines, loaded or unloaded.

I believe that you're driving your CTs backwards in an effort to raise the output current to measurable levels. Try to measure their characteristics with a variable AC source and different loadings while using your scope to view the waveforms, you may find that the hysteresis of the core is causing the distortion and creating the harmonics.

You have to ensure that your instrumentation is "spot on" before you start blaming your circuit setup.

Interesting stuff, keep up the good work.

First, more about my setup:

I have full load share and synchronizing control. If you look at the last picture you'll see that the gens are sharing kW and kVAR. KVAR is not as stable as KW, I'll have like 4-5 kVAR go between the gens at any point in time, but it's always low.

At the top of each gen box in the same picture you can see the bias signals to the governor and boltage regulator for load sharing.

The load is purely resistive: 3 x 40w light bulbs, the system is wired in a wye configuration. The voltage regulator is a DECS 100 the "governor" is a VFD since I drive them with electric motors.

To address some of your replies, Yes, the generators, and all other components in the setup are the same between the two generators.

Now, these small alternators... I'm not sure how precise they're manufactured, so not sure if they're "matched". I'll try to find that out.

No, I don't have a gen with twice the speed of the other one :) They're both the same. Picture one shows 2 gens, already in parallel. The graph is color coded and there is a legend at the top. The higher frequency waveform, is the current flowing between the gens at no load.

Yes, I'm driving the CTs backwards to make it look like I'm dealing with a lot of power ( hence the hundreds of kW you see on the screen in pic#4. The CT's were custom made... While they won't deal with currents of magnitudes of many times over their rated current (fault currents), they're fine at rated values. When i have only one generator online, current waveform looks good all the way up to rated power (and more). The distortion only occurs when I parallel the 2 gens.

I'll try to power everything from a fixed power source and see how that goes (if I can get my hands on one). IF it ends up being mismatched alternators, is there anything gI can do to make it better? besides buying new ones :)

Thanks everyone for the input!

Short of rewinding the generators yourself there's very little you can do. If your load is incandescent bulbs you may find that the filament will have an effect on the VAR loading. Depending upon the manufacturer, rating, etc., some have very curly filaments that may present as inductance to your setup, others have fairly straight ones. You might find straight nichrome wire that you can wind as a bifilar coil can be used to cancel any inductance. Is this for school or your own edification?

1/ You have not given us the frequency of your system. 60 Hz??

Theoretically, for equal voltage and true sine wave, synchronised generators will not exchange current. If voltages are different, current will flow between them, depending on voltage difference and the regulation of the machines due to resistance and reactance.

If one generator is driven harder, current will circulate to keep the machines synchronised, this is the essence of a normal AC power grid with inductive machines & transformers. If you wire a neon indicator lamp to the common bus, shine it onto a mark on generator shaft (may need dim room light), you can see the generators shaft angle change with load (moves in direction of rotation with increasing engine power).

In the early days of AC generators much trouble was experienced with machines of different rating or make in parallel until standards for a true near sine wave output were set.

2/ Have you tried loading each machine alone to measure regulation [voltage drop or droop] with load? With what result? As a simple indication, what are the DC resistances phase to phase?

3/ You write there are voltage regulators DECS100 [AVR] - have you tried adjusting voltage settings with machines in parallel?

A good AVR will have adjustment for the voltage regulation (droop) as well as voltage setpoint - but this normally responds to the VARs not the watts & the droop will be small on a resistive (watts only) load. It has been pointed out that wrong polarity of a CT connected to AVR or VAR sharing control would cause circulation not reduce it.

4/ Are you able to measure the DC field currents/voltages of the generators? Applying varying DC field excitation in place of DCS100 output, the two generators should have similar field current against AC output voltage curves.

67model

Before I receive the full ire of the informed crowd, might it be that the IC motor speed of the gensets is in fact not identical. If not, you may be able to microadjust the operating RPM's and observe improving or degrading quality of the waveform. It also could be, as another observed, a minor manufacturing variation between the two genertor windings, which may be overcome by adjusting engine RPM's? Or is that hogwash.

KVAR is regulators. Make sure the sensing wires and droop CT's are phased correctly on the DECS. Also make sure that the 52J/K terminals are connected. You can check phasing with an oscilloscope. Easier way is to look at the metering screen in the DECS and see if the amps are going crazy while under resistive load in single unit operation. Also, you can try to change from droop to CCC or from CCC to droop.