SMA too sophisticated to be any good?

Hi and thanks for information gentlemen. 
The genset is a GENERGY Robust series 3 KW rated 2.6 continuous. It uses a standard alternator with electronics for control. It is not an inverter type.. I have had to contact the factory to check that I could tie neutral to earth to comply with the SMA requirement and they were most prompt and helpful. I have not yet asked them about voltage regulation strategies. The genset is new. It shows between 233 and 231 on the front panel and 51 and 52 Hz. I have the SI set up for 230 V 50 Hz I have opened up both voltage and frequency limits in the SB programming for the generator from their defaults to near the widest limits. I can monitor through the SI logged data what it sees from the genset and it matches what I see on the genset front panel. I have added a load to the genset in parallel to the SI and have seen it switch through the genset supply for just a second then the SI throws  an error and resets. The error is always over voltage. I understand that I may have to use a third party charger but it really irks me. I have paid £8K for this kit and connecting a SI to a generator to charge the batteries if there is not enough sun is a basic bit of functionality IMHO. Also 48V chargers here are really expensive and I will need a shunt so the SI can monitor the charging which it can't do itself. I bought the kit from a reputable firm inthe UK, they are supporting me but   when I asked for a generator spec I was told "any modern genset with voltage and frequency in spec".  What I now see from the SMA thoughts on genset requirements which link I posted above indicates that this advice was no good and if I had known how how picky SMA were with genset I would have gone down a Outbacker Victron route. To be frank I really don't understand the difference between a grid and a buzz bar. I have a suspicion that the SI is trying to control the power output of the genset by varying the frequency of the AC which it does when connected to the SB. I would like to ask the genset company what their voltage regulation method is but if I send them the incomprehensible stuff SMA produced and I posted above they will just laugh at me. In the manual set that SMA produced there is of course no mention of genset compatibility requirements. Sorry if I sound P***** Off but I am. Thanks for helping. Spike

Thanks for your comments. First of all Bill thanks for yours and here is my reply. I have detailed manuals. Part of my problem with SMA is that the information is spread over five separate documents and now with the generator 'expectations' extended to a sixth which is called a 'whitepaper'. I have ploughed through all of them. The only generator performance curve specs are in the 'whitepaper'. I have access to many thresholds and have maxed these out temporarily as an attempt to get past the overvoltage error. Although not explicitly stated I stumbled across mention of a frequency switch to carry out load shedding so at least on the battery inverter to PV inverter side frequency is used to manage power. The generator comes in to the battery inverter on a different connection to the PC inverter but the manual says when the generator is on the output voltage and frequency of the battery inverter is dictated by the genset. There is no option to change power demand scheme.

Other advice, and thanks too.

Battery size. I have programmed the C10 level into the unit but since the PVs aren't connected yet and the battery inverter cuts out before using the generator to do any charging  C10 or C20 is a bit moot at the moment IMHO. As for battery capacity my domestic non water pumping demands are very modest and the capacity allows for a five day disconnect, is that not enough? The whole idea was to use the AC predominantly as it is produced to pump water have I really broken some basic design rule here? If I have then why does SMA produce units of this size and specify their use in this way?  If SMA had produced a list of qualified generators or put a genset spec on their technical spec then I would have been forewarned about this problem and gone down a more traditional DC coupled route. Also guys, before the bricks get thrown back, I did post in Beginners Corner so please forgive me for buying kit from a reputable company specified to work in a particular way, reading the manuals, connecting it as specified and then expecting the damn thing to operate. Thanks for your time guys, Mark 

Just to close this off. An installer visited, found nothing wrong with my installation or the genset but reproduced the overvoltage error from the Sunny Island. We tried an old Mitsubishi 3kw genset that he brought with him so it seems something must be wrong with the waveform on my genset even though on a scope it looked fine. Thanks for your comments