Hi All

Our Solar system is made up of the following: 1xMecer MKSII 5KW, 12X355W panels, 2x 3.5KW Pylontech's, and an ICC.

All works fine for 3 months now except for the start-up current of the (brand new Bosch A++) fridge.

During the night when only the fridge runs I see on the ICC that at start-up there is a peak of 2KW for a second then it drops to 90W until it switches off again.

This is no problem during the night, but during the day it means we can only use 3KW continuously, if we use 4KW and the fridge start up the system switches over to the grid for 8 minutes although the inverter is able to handle 10KW for 5 seconds.

I am not sure if the inverter switches over or the ICC does it??

Strange enough when I disconnect the grid it handles the surges without a problem!

On the internet some people advice to get a "soft start" device, others say this does not work.

I can imagine some sort of a capacitor that charges and discharges and so absorbs or lessens the inrush current? does anyone know of a device like that.

Please help/advice.

 

I suspect the battery’s voltage dips to below the threshold you have set for back to mains? A larger battery pack may lessen the effect or if one lowers the back to grid voltage, just enough so it won’t switch, being careful not to go too low and damage the batteries?

i also have some of the Bosch A++ fridges and freezers, two of each, cant say I have seen the large spikes you mentioned? 

Models are 

2 x KSV33NW30 fridges - both set the 2 deg C

2 x GSV33VW30 freezers - 26 deg C

I tend to find pumps are the real problem, where is see spikes, only

long enough for ICC to display briefly and then they subside, for water pumps, pool pumps, jetted bath pump etc, but with the 5kw Axpert no issues as long as one has enough solar

I am not sure where one can buy a soft start device? Not even sure if they are device specific? Hopefully someone else On the forum can advise?

 

 

 

I also don't see major spikes being caused by fridges. At night my system runs quite happily and uses little power with 2 fridges and a deep freeze running. And one of the fridges is a pretty old affair, without the variable speed and other smart stuff that our new LG has.

Things that I see causing spikes are micro waves, electric kettles, pool pumps, hair dryers. OK... and our heat pump, but that is predictable and my system is handling it (it's not low power, but it's significantly lower power than a geyser element).

21 hours ago, Bert sa said:

I am not sure if the inverter switches over or the ICC does it??

It will be the inverters. If you make settings 12 and 29 lower, it might happen less frequently, or not at all. That way, you let ICC do the switching. Setting 29 (battery low cutoff voltage) is silently involved, because the inverter switches to mains when the battery voltage reaches setting 12, OR if it reaches 2 volts (for 48 V systems) above setting 29. It comes back to battery voltage only if the battery voltage is above setting 13, AND also 4 volts (for 48 V) above setting 29. This behaviour is unhelpfully not documented.

There is no patched firmware for MKS IIs. The more complex patched firmwares implement KettleKomp™, which is a crude form of internal resistance compensation.

So if my maths is correct, 12 x 355 = 4260W, so say you get +-4000W out of the panels midday. Probably less in winter. You daytime load is 3 - 4 kW. Just about covered by solar on a good day.

Now the fridge kicks in. All 2kW spike. The inverter sees the extra load, and it is instructed (via you settings, assuming SBU or SUB), to collect the balance from either Battery, or Utility (exclusive of Solar). Hence the switch happens.

So I think you could have a double wammy.

1) You do not have enough Solar W to carry the load - add more panels, seeing your MKS11 can go to 5000W (and a little extra, if you can stomach it). Your mppt's V is rather large, so you need a lot of extra panels for the next string. My guess is 4 minimum. Assuming Voc for your 355W panels is approx 46.8V. So I assuming you have 2P6S PV setup. The will give you 281V. Well below the MKS11's Vmax of 500V. So a 2P8S setup will give you 374 V. Still safe. The total PV array will then be 5680W. If you don't want to over provision, then just add two panels, one in each string (4970W - right on the money). Obviously, you need the roof space and extra mounting hardware for that too! There goes the relaxing weekend you planned! Seeing you have ICC, create a long temp view graph of your panels PV performance (Watts), and then make an informed decision.

2) As already mentioned, the Batteries V may sag just outside your settings, so maybe lower the low voltage setting, as explained by @Coulomb. I would have wanted to believe the Pylontech's should be able to handle. Seem not. But I don't think it's the PT's fault.

Either way, it's the switching characteristic of your MKS11 that doesn't allow you to blend between your 3 sources. That why the King's/Infini's is preferred, but to each his/her own.

Edited August 7, 20197 yr by Sidewinder

23 hours ago, Coulomb said:

It will be the inverters. If you make settings 12 and 29 lower, it might happen less frequently, or not at all. That way, you let ICC do the switching. Setting 29 (battery low cutoff voltage) is silently involved, because the inverter switches to mains when the battery voltage reaches setting 12, OR if it reaches 2 volts (for 48 V systems) above setting 29. 

Hi Coulomb

Thanks for your input.

Setting 12 is set to 45v, setting 29 also 45v, (Pylontech min. discharge voltage is 44.5v)

Is it safe to set setting 29 to 43v to make up for the 2volts it sets it above this setting?

You mentioned: "That way, you let ICC do the switching" , do you mean that I should set the voltage of 12 and 29 even below the minimum recommended  by Pylontech to bypass the inverter and let the ICC do the switching? the only option I see on the ICC settings is the low battery setting that I set to 25% DOD and comes back at 35%

What puzzles me is, during the night from 24h00 til 6h00 the fridge is the only load and when it switches on the ICC measures a 2kw surge (only 1 out of 10 times, I suppose it takes a reading every second and most of the times it misses the surge) that comes strait out of the batteries and NEVER causes the inverter to switch to the grid, as an example: during the day with a load of 3500w (panels also 3500w) the fridge switches on and the 2000w (now 5500w total) surge needs comes out of the batteries but causes the inverter to go to the grid, the voltage drop in the batteries should be the same shouldn't it? this morning the the load was about 2600w, the kettle of 1900w was switched on for 2minutes (load now 4500w )when coincidently the fridge came on and causes the system to go to the grid, see dashboard and battery trend pictures.

Makes me think that the 5500w is the problem although the inverter should be able to handle a 150% overload for 5 sec, and has a surge capacity of 10kw for 5 sec according to the spec sheet.

How can we establish if it is the low voltage or the overload protection that is switching it over???

 

Thank you all for your input.

29 minutes ago, Bert sa said:

Is it safe to set setting 29 to 43v to make up for the 2volts it sets it above this setting?

Yes.

30 minutes ago, Bert sa said:

do you mean that I should set the voltage of 12 and 29 even below the minimum recommended  by Pylontech to bypass the inverter and let the ICC do the switching?

Yes.

Quote

the only option I see on the ICC settings is the low battery setting that I set to 25% DOD and comes back at 35%

Well, set those to what you are comfortable with. It's much better than switching based on a battery voltage. Especially when the firmware doesn't make any attempt at internal resistance compensation.

32 minutes ago, Bert sa said:

Makes me think that the 5500w is the problem although the inverter should be able to handle a 150% overload for 5 sec, and has a surge capacity of 10kw for 5 sec according to the spec sheet.

Unfortunately, the firmware seems to be quite sensitive to overloads, not allowing the full overload promised by the marketing.

thanks Coulomb

I am a little hesitant to set setting 29 below the minimum stated by Pylontech, even if it switches 2 volts earlier then it is set, why is it not switching when it is supposed to according the setting but 2 volts higher?

Does anybody else uses these settings for 12 and 29?

12 minutes ago, Bert sa said:

thanks Coulomb

I am a little hesitant to set setting 29 below the minimum stated by Pylontech, even if it switches 2 volts earlier then it is set, why is it not switching when it is supposed to according the setting but 2 volts higher?

Does anybody else uses these settings for 12 and 29?

hmm now i am a bit confused, I know the pylontech specs show discharge range as low as 45, but yet their recommended setting on an axpert is 47.5 for setting 29?

 

my 12 is 48V and my 29 is 47.5V