Vac-SRR Discussion

A Discussion of all of the issues surrounding Vac-SRR errors written by:
John Berdner  - President SMA America  9/18/02

I will try to explain what we are seeing out there, and what I think is going on. Hopefully this will help your situation and others out there who are seeing the same thing.

We are seeing high voltage issues in a few installations out there. Everything seemed to be fine until the weather got hot and the utilities started bumping up the voltage to deal with big afternoon air conditioning loads. The problem is that the voltage is really too high in the morning when the air conditioning is not running. Typically, we see this issue in multiple inverter residential installations. I think that there are four factors which combine to give us this problem.

The first issue is high grid voltage. Technically the utilities are supposed to provide service voltage within "ANSI Range A" (112 to 126 @ 120 Vac, and 224 to 252 @ 240 Vac). They are allowed to go into "ANSI Range B" occasionally. It is my opinion that every morning all summer long does not qualify as occasionally.

We first saw this on a 10kW residential system in Vacaville, CA. The local utility, PG&E, came out, measured, and ultimately fixed, the high line voltage because it was above 252 Vac (in accordance with the ANSI standard). The system owner had to make a fair amount of noise before they would fix the line voltage, but PG&E did fix it and everything is working fine now. PG&E is pretty good and other utilities have different limits and/or policies.

The second issue is grid impedance. A brief aside: In Germany, they monitor grid impedance as part of the anti-islanding scheme. When we came to the US, I started asking people about grid impedance, and guess what, nobody had any idea what I was talking about. I could not even find a meter manufactured in the US to measure grid impedance, and we had to order one from Germany.

As we drive current back into the grid the voltage has to increase in accordance with Ohm's law (V=IR). At 10kW the peak current would be about 42 Amps. After taking all kinds of impedance measurements, in both residential and commercial installs, in several US States, we found that the grid impedance (and correspondingly voltage rise at 10kW) is roughly as follows:

• 100 Amp Service: 400 mOhms =  16.8 Volt Rise
• 200 Amp Service: 200 mOhms =  8.4 Volt Rise
• 400 Amp Service: 100 mOhms =  4.2 Volt Rise

The third issue is wiring losses which, in our case, show up as more voltage rise. Normally, we recommend that designers try to account for no more than 1 to 1.5% losses on the AC side (2.4 to 3.6 Vac rise). If you design for the NEC max of 3% then you will be at a rise of 7.2 Vac. Remember, this has to include all the TOTAL losses between the inverters and the meter location - wire, connectors, breakers, sub-panels, switches, and so on.

Lastly, the UL/IEEE anti-islanding requirements make the inverters very sensitive to high grid voltages. The UL/IEEE limit where we have to be off-line if 264 Vac RMS. This is under any condition regardless of power level, noise, or spikes on the grid. In actuality, we trip off-line at about 262 to 263 Vac so that we can be sure we are gone by 264 Vac.

We are seeing inverters trip off-line on high grid voltages with the threshold being about 248 to 252 Vac, at rest (i.e. with no inverters producing power), or so. We think this is due to the above factors plus some noise or spikes on the grid. We are working the issue with our colleagues in Germany and UL to see if we can make the inverter a little less sensitive to spikes and noise and still meet the UL/IEEE requirements.

When you start at the grid voltage of 250 Vac, add in another 8.4 Vac for grid impedance (200 Amp service), 3.6 Vac for wiring, and a little bit more for noise, you can see how it is that we have a hard time staying below the UL limit of 264 Vac.

As the PV industry starts putting inverters into the field, we are seeing that the grid is not always within specification. There are good reasons why local utilities pump up the voltage in the summer time and so they are reluctant to drop it back down. If you are above 252 Vac, at rest, then you may be able to get the utility to install a bucking transformer to bring the voltage back down for your house while the voltage remains higher on the feeder. Ideally, have them come out and measure when the inverters are running so the voltage they see will be highest.

Another approach is to get a written waiver from the utility to allow us to bump up the set points beyond the UL limit for that specific installation. In my experience, they are usually reluctant to do this as well. Even if you do get the waiver this means that the inverter has to come back to the factory for a short vacation because UL has dictated that we cannot change system parameters in the field.

As installers, the only thing that can be controlled is wiring losses. The cost to increase the wire sizes to get down below the 1% loss range may be money well spent if you are doing larger residential systems.

I hope that this helps to explain the problem even if it does not give you any really good solutions. If you have any questions about this you can either reach me directly at berdner@dont-want-spam.sma-america.com or contact our technical support department at techinfo@dont-want-spam.sma-america.com or call us at (530) 273-4895.