Hello Wrenches,
We've installed 2 sunny island systems, and were planning to install three or four more, until we identified what I consider to be a serious functional deficiency with the product. In commissioning the first two inverters (different customers, but around the same time) I found that the Sunnyboy inverters that were feeding the island would error out for AC frequency tolerance exceeded when the inverter was in backup mode (no grid signal present) and the batteries were mostly charged with charge current coming in from the array. The SI frequency modulates to reduce battery charge current and control battery voltage to absorption or float voltage - this causes the SB to error, wait 5 minutes (because frequency returns to normal once charging stops) then restart and error out again. This is particularly problematic with a smallish battery bank and largish array as voltage can go high fast. When I first reported this to SMA tech support it was described as a communication problem. Apparently if the two inverters are communicating, the SB is informed by the SI that it needs to operate in off-grid mode and it should disregard the normal grid frequency tolerance. So after a few return trips with new RS-485 cards and checking and double checking the com wires to no avail, I was told that this was a firmware bug and that a fix could be expected in some months, but with no solid time frame commitment. An alternative was presented that the SB can be reprogrammed to run in off-grid mode permanently, but since our systems have an inverter bypass transfer switch in then we could not do that, because if that switch is switched then the SB would be connected directly to the grid without the isolation of the SI, and would take responsibility for anti-islanding. That's clearly a no-no. The transfer switch at the time was a requirement of the utility (all loads must be able to be served directly by the utility if necessary). This requirement was just removed by the PRC, but there are serious design issues that are unacceptable in my eyes with removing the bypass. 1) If the sunny island breaks, there is no option to power the critical loads directly from the grid. We sell battery backup systems so that people have reliable power to their critical loads. This would be clearly a better designed system if the inverter could be bypassed. The other backup systems, like outback and Xantrex (DC coupled) have easily integrateable and supported transfer switches for this purpose - as they should. 2) If the inverter was out for repair (say it's out of warranty, and has to be sent for repair, not just quickly swapped out) the sunny boy would be down too as there is no alternate path to the grid. With a bypass/transfer, the customer could still be producing power and net metering while the backup system is down. 3) Because the SI is 120Vac, and the SBs are all 240 V, this means there must be a transformer, or two inverters (and most people don't need 10Kw of backup power). Some customers would choose to forgo the automatic backup in favor of eliminating the efficiency loss of the transformer in day to day grid-tie operation by leaving the bypass switch activated except during a power outage, and manually activate the backup system only when it's needed. This is a less critical issue, but having the option would be better. Page 122 of the SI manual warns that if the batteries are full, the frequency shifts and the inverter shuts down, but this is not quite true. If the battery is small relative to the array size, the voltage will shoot up fast, even when it is at a lower state of charge. It appears to be an all or nothing charge - no tapering. I watched as one system continued to go down In SOC until it got to the low eighties. I did not stick around to see how low it would get. I reconnected the grid and promised the customer to try to get to the bottom of it. 95% of the time during this test the SB was completely disconnected. Because it waits 5 minutes to reconnect, it doesn't even work as well as the old series charge controllers (like the Trace C30A). When we were designing the systems, we sent one-line drawings to SMA to make sure it would work the way we wanted, and they were given the thumbs up. This includes a more complex arrangement on a 3-phase service that has not been installed yet and is on hold. We do not want to install any more of them until this is resolved. We have customers with battery-less systems that would like to retrofit. We've told many of our customers that if they want to add battery backup later they can. I'm anxious to see the fix. Has anyone else seen this problem? Does anyone have a solution? Do any of you have systems that might have this problem, but haven't been put to the test? Am I missing something in the programming and tech support just hasn't caught it? Any help would be greatly appreciated Thanks, Mark Drummond Field Manager Positive Energy, Inc. NABCEP Certified PV installer
_______________________________________________ List sponsored by Home Power magazine List Address: [email protected] Options & settings: http://lists.re-wrenches.org/options.cgi/re-wrenches-re-wrenches.org List-Archive: http://lists.re-wrenches.org/pipermail/re-wrenches-re-wrenches.org List rules & etiquette: www.re-wrenches.org/etiquette.htm Check out participant bios: www.members.re-wrenches.org
