11519RE: [hreg] Hybrid Generator and Solar
- Jan 25, 2013
From the sounds of your situation the PV would be almost unnecessary if the loads were constant and the genset was able to provide 100% of the power with the grid down. With the grid available and the generator off the PV would only power a small percentage of the loads in the first two examples.
You are correct that the generator would never see the inverter output under those circumstances. If it ever were to happen that the AC loads dropped below the PV inverter output with the generator running the generators control circuitry would very likely be damaged from the grid-tied (GT) inverter attempting to backfeed to the generator. (I think this is why there’s the general paranoia you mentioned against such system design.)
Typical residential systems are usually designed with a critical load sub-panel fed from the grid or the generator. A transfer switch will isolate the generator feed from the grid so they are never accidently both feeding the homes main or critical loads simultaneously.
If batteries are in the mix, the inverters could continue to charge them if a SMA Sunny Island or Outback Radian etc inverter/charger (I/C) is used along with a generator. As you alluded to, the phase-shift ability of the I/C allows the GT inverter and PV working only as long as the batteries need charging and not otherwise feed parallel current to the loads.
In an AC coupled grid-tied system with battery backup the I/C will isolate the generator from the GT inverter because it cannot be assured that the loads will always consume >100% of the inverters output.
It’s designed to be homeowner proof.
North Texas Renewable Energy Inc
NABCEP PV 031310-57
People always ask about grid-tied solar systems running with a generator. Most people (I’ve talked with most of the inverter manufacturers) will tell you it can’t happen and give a long list of reasons why. However, the world needs this solution because there are so many places that need solar, but a huge battery bank is not practical. So I tried it myself and I have proof that you can run solar grid-tie with a generator. Here’s what I’ve seen work and how to make it work.
Here’s the general conditions:
1) The generator should be able to cover the loads independently of the solar (big gen, small solar).
2) The load power draw must never be greater than the solar power input plus a minimum output from the generator.
3) The inverter must be able to match the inverter frequency and voltage through a wider range than what would be seen from the Grid.
4) There are some fine voltage/frequency/reactive power controller points in #3 that require a longer explanation, but I’m not going to put that in this e-mail
Example #1: 100kW Genset with 36kW PV inverters. The Genset was an expensive, high quality machine that could hold a very steady frequency and voltage. Even still, the voltage and frequency of the generator would drift outside of the limits imposed by UL1471/IEEE1547. The Inverters were SMA Sunny Boys which actually let you adjust the voltage and frequency limits, so we went in and widened the limits. Also, a load management system made sure there was always at least 30 kW of load on the system. Worked great even through load changes!
Example #2: 350kW Genset with 100kW PV inverter. This Genset was an expensive, high quality machine that could hold a very steady frequency and voltage. The PVPowered inverter was not changed, however, the output was limited to 10kW when the grid was down and the generator was running. The system was not well automated and tested with a steady load of greater than 30 kW, so the frequency and voltage didn’t move around and stayed in the IEEE1547 window. It worked.
Example #2: 5kW Genset with 600W PV inverter. The Genset was the cheapest thing on the market. The inverter was a Chinese non-UL1741 grid-tie inverter with a very wide voltage and frequency window. These ran together just fine in short term testing. I could tell that switching loads on and off caused the generator frequency to change and the inverter was a little slow (about a second, 60 cycles) to resynchronize with the generator. I had motor and lights running as loads and could see no flicker to motor sounds during load switches. However, the power meter would blib an unusual power factor reading indicating that the sine waves between the inverter and generator were not lined up perfectly for a second during those load changes. Didn’t hurt anything short term, but still investigating long term effects.
Anyone out there that has found another solution? Is anyone else out there interested in these solutions?
- << Previous post in topic Next post in topic >>