Solar array inverter to run as island

[eeprom]

Hello,
I'm working with an enphase inverter, intended for PV arrays for grid tie applications. The inverter needs to see the proper grid voltage and frequency for some time before it will come on line. For testing purposes (and safety), I would like to create a simulated grid to which I would connect the inverter. The inverter is designed to push out as much current as the attached PV array can supply.

I have considered 2 methods so far: 1 is to use another inverter tied to a battery bank. The inverter would produce a 120V sine wave for the enphase inverter to synch with. The problem with this is that the enphase inverter is likely to backfeed the other inverter, which would destroy it.

Method 2 is to use a generator. This would work but it would be loud and expensive. And there is still the problem of backfeeding, which would destroy the generator.

Does anyone have any experience in doing this? Any ideas?

thanks

 

[jraef]

I wouldn't use another inverter, just because the PWM of that inverter may affect how the Enphase inverter synchs up. I would use a generator. The Enphase inverter would not "blow up the generator", but it may affect the AVR if it's one that cannot handle harmonics, so make sure when you talk to a generator supplier that you tell them you are going to run an inverter on the output. You also will need a load on the system in order to test anything the Enphase is doing, so run a motor with that generator, then turn on the Enphase and see if it pumps power into the system. I would use a pump in a closed loop tank and a valve so that you can easily vary the load.


"If I had eight hours to chop down a tree, I'd spend six sharpening my axe." -- Abraham Lincoln
For the best use of Eng-Tips, please click here -> faq731-376

 

[davidbeach]

Method 2 with a load bank. Load bank needs to be larger than maximum output of the inverter plus the minimum sustained output of the generator.

But why? The the testing labs have gone through all of this stuff already.

 

[Noway2]

If you have access to IEEE specifications, spec #1547 has a lot of information regarding test setups and procedures. There may be something in there that would be of benefit. It would also give you the criterion that you would need to meet for a grid tie switch.

 

[eeprom]

Noway2,
That is an excellent suggestion. I will do that.

thanks

 

[waross]

Size matters. What is appropriate for a 1 kilowatt test may not be appropriate for a 1 megawatt test.
That said; I would look for a generator and a resistive load bank that is a little larger than the invertor that you want to test. Start the generator and connect the load bank. Then connect your invertor. The heavy resistive load should give you stability and absorb a lot of the harmonics. The output of the invertor should supply the load bank and reduce the load on the generator (and the fuel costs).

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

 

[eeprom]

I agree with the generator idea, and the fact that it should be larger than the inverter and load. The problem is that these panels are to be tested for a 30 day period. A small generator (2.5kW) would be sufficient, but it would have to be refueled several times a day. But this might be the simplest solution.

 

[eeprom]

One other possible malfunction with the generator...if the frequency drifts lower than 59.5 or greater than 60.5, the inverter shuts off.

 

[cranky108]

A little old thinking, but an Motor-Generator set would work as both generator and load. That is if the MG set is a dc motor tied to an ac generator.

 

[eeprom]

We had come up with a similar idea. To use a VFD to control and induction motor. The induction motor drives a synchronous motor. The VFD keeps the output of the synchronous motor at 60Hz. If we have any trouble with this, we can put a tachometer on the synchronous shaft and use the output signal as a feedback to the VFD.

 

[waross]

You may use a 5KW set with the no-load frequency set down to 60.5. With a 50% load, the frequency drop will be about 0.9 HZ. or down to 59.6 Hz.
Fuel consumption will be reduced when the panels are supplying power. You may be able to switch off most of the load during the night.
Whatever you are most comfortable will probably be the best solution.
If you go the VFD route, the VFD should be capable of regenerating or you will need a load bank. You may be able to load the DC bus when the voltage goes high, but wait for an opinion from one of the VFD experts on this.

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

 

[LionelHutz]

I'm not sure the inverter would push power back through any type of M-G set to the grid. The inverter will likely just raise the voltage at the output terminals. It would have to run at a higher output frequency to back feed a generator.

 

[waross]

"It would have to run at a higher output frequency to back feed a generator."
That is why we suggest a load bank and a generator. The load bank takes a fairly constant load. As the inverter supplies power, the load on the generator becomes less. The inverter is capable of running a little above or below rated frequency and it should be possible to locate a generator capable of running within the inverter limits.

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