Graham Walmsley avatar image
Graham Walmsley asked

Size of MPPT with Multiplus-ii 48V/8000VA 230V and 6400W solar


I am in the UK and designing a self-consumption ESS system with no feed in allowed. I am new to this and been spending lots of time reading, but there is one issue I can't find the answer to.

Sorry if this is long winded, but want to explain my thoughts.

I am looking at the following configuration:

  • Multiplus-ii 48V/8000VA/110 230V inverter/charger
  • Cerbo GX Controller with ET112 Energy Meter (to support the no feed in I hope)
  • 4 x PylonTech US3000C 48V batteries in parallel (total 14.2Kw)
  • MPPT 250/100 connected to 16 x 400w sunpower maxeon 3 panels in a 2S/8P configuration (6400W vs MPPT rated 5800W)

If I have done my maths correctly, the panels could produce (via the MPPT) a maximum of 151A charging current at 48V (the input amps are well within the limits of the MPPT with voltage at 174V and 27A) but this will be limited by the max 100A output of the MPPT if the panels are reaching maximum. Due to the panels facing south east and being on a 20 degree roof, I don't expect to ever reach peak output but could get to 90-95%.

The 4 x PylonTech batteries in parallel will take up to 148A charge (4 x 37A) so again, charging is limited by the MPPT. I will be adding additional battery units in the future as budget allows.

What are my options?

If I split the panels over 2 x MPPT 250/85 I would get the max output from the panels which would allow charging of the batteries faster but is there then an issue with the 110A charger limit of the Multi-plus? Does the 110A only refer to the Multiplus charging the batteries but the inverter can draw more current (i.e. 6400W/48V being 134A ish) so would be able to use more of the output from the panels to support the load rather than being limited by the 100A of the single MPPT 250/100?

Thanks in advance for any help or advice.

Multiplus-IIMPPT ControllersLithium Batterysolar
2 |3000

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

3 Answers
wkirby avatar image
wkirby answered ·

The two MPPT chargers would then allow the full array power to charge the batteries with a combined total of 134A (6400W/48V).
The 110A limit of the MultiPlus isthe maximum that it can charge from AC input and does not affect the MPPT's output.
The inverter can draw more current at full chat, 166A. So perhaps four Pylontech modules would not be enough if you do get the inverter working at maximum output. Five Pylontech modules would be a good match.

Additionally, you mentioned at the beginning that you want to use ESS but you are not allowed to feed in. ESS is not for you in this case because ESS is a grid parallel topology and it will feed in a little at it's control loop responds to changing grid conditions. Any grid parallel system would require permission from the DNO.
The solution is to not run ESS and not apply a grid code to the system. This way it would not feed in and you could set it up to ignore AC-IN until you need it for charging the batteries when required.

2 |3000

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

Graham Walmsley avatar image Graham Walmsley commented ·
Good catch on the inverter being 166A at 48V, I had worked on the sustained 6400W figure but from some comments it could go as high as 8,500W for brief periods (177A) so adding a 5th module would bring the battery pack up to 185A @ 48V which should cover it.

On the ESS front, I watched the presentation from when ESS was released, and it says you can do no feed in when using a Cerbo GX with an ET112 Energy Meter (which I have). I should have mentioned that earlier but was trying not to complicate things too much :-)

Has this now changed?

I will update my original post with the additional components.


0 Likes 0 ·
wkirby avatar image wkirby ♦♦ Graham Walmsley commented ·
Although it is possible to tell ESS to not feed in excess PV, set feed in to 0W, this is a best endeavour. The grid Voltage fluctuates constantly and the ESS can only act in response and so small amounts of feedback are inevitable. A modern smart meter would sense this right away and potentially land you in it.

Regardless, the DNO will still regard an ESS setup as a grid parallel installation and it would require permission.
Using the Ignore AC configuration will allow complete disconnection from the grid supply until it is needed to charge the battery and as such it would be a battery charging appliance.

0 Likes 0 ·
Graham Walmsley avatar image Graham Walmsley wkirby ♦♦ commented ·

Thanks for your update. Any experience with DNO accepting connection of a self install grid parallel system?

From the ESS manual, the feed in options are shown as :

  1. Energy can be sold back to the grid
  2. Feed-in is allowed, but not rewarded
  3. Feed-in is absolutely not tolerated - even for a few seconds: there are certain prepaid meters in South-Africa that will disconnect from the grid when they detect Feed-in.
  4. Feeding-in results in inflated bills because the electricity meter can only count in one direction - up.

I was hoping to do option 3, but it sounds like that may not be possible, even using the MPPT and grid monitor solution. The inverter isn't big enough for all my loads, so I was hoping to do critical loads via the inverter and have excess solar power help with the non-critical loads if the batteries were full and spare capacity.

The configuration I was going for was from the ESS manual.


So if the only way to achieve feed-in not tolerated to not run grid parallel, why is it in the ESS manual?

Sorry, I must be missing something.

0 Likes 0 ·
Phil Gavin avatar image
Phil Gavin answered ·

A minor tweak in your calculation:

Pylontech's charge at 53.2V, so power through the MPPT will be 5320W @100A.

Empirically, I have found that I can over panel at non-ideal slopes and directions. New panels get dirty quickly too and drop in performance.

I put non-ideal pitched Easterly and Westerly PV strings in parallel through a single MPPT and I can nearly double the theoretical number of panels without the MPPT clipping. ( and that is in South Africa, not the UK).

Another suggestion is to use 4s4p panel configuration as higher voltages have fewer copper losses..

I do believe, the 110A charging limit of the Multi refers to charging done by the Multi, that is to say, charging current drawn from the rectified AC power input.

The MPPT charging 100A capability will be additional to this.

2 |3000

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

wkirby avatar image wkirby ♦♦ commented ·
@Phil Gavin Check out the specifications of the PV modules, they are quite unusual. These are 104 cell PV modules, VOC 75V! 3S wouldn't divide into 16 so well, so I think the 2S8P configuration is good, or 2S4P if splitting over two MPPT controllers.
0 Likes 0 ·
Graham Walmsley avatar image Graham Walmsley wkirby ♦♦ commented ·
Yes, two in a string using the MPPT calculator comes out at 173.8V max, so 3 in a string exceeds the 250V limit on the MPPT and I would prefer not to go to the RS if possible due to space.
0 Likes 0 ·
Phil Gavin avatar image Phil Gavin wkirby ♦♦ commented ·
Quite right, I missed that.
0 Likes 0 ·
Graham Walmsley avatar image Graham Walmsley commented ·

Thanks for your response. Going off the data sheet for the US3000C it says recommended charge is 37A @ 48V, so with 4 in parallel I believe the BMS in the battery stack will tell the GX and thus the MPPT to use 148V max charging current.


If I follow @WKirby suggestion of going to 5 US3000C modules (which I will) then I think the charging current will go up to 185A @ 48V which is higher than my 6400w panels can provide even when using 2 x MPPT 250/85 each with 2S4P.

Am I doing my maths wrong somewhere? I am new to this so any advice would be great.


0 Likes 0 ·
keithbrown avatar image
keithbrown answered ·

Just get the new 450/200 RS and you'll have plenty of room to play around with strings and parallels. Always place panels in parallel up to 20-25a Then string those together assuming you can afford thicker wire. Alternatively, Use SunPower 435 panels and use one 100/20 Victron charger per panel. This would be the DC equivalent of micro inverters. I know of no other brand of panel that has 60+ volts per panel. Other brands would require 2 panels in series to get sufficient voltage.

The answer you are likely looking for is 2 150/70 or 250/70 chargers. Always leave a little room for upgrades and future changes.

2 |3000

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

Graham Walmsley avatar image Graham Walmsley commented ·
Thanks for your response. My two issues with going down the single 100/20 route for each panels are:
  1. Space - 16 100/20 units and all the bus bars to connect it all together will probably need more space than I have on the garage wall for all this
  2. Voltage - My understanding is the battery charging won't start unless the voltage from the panels is high enough, and with all the panels in parallel on a 48V system that would mean I won't get any charging from the panels unless there is bright sunlight (Vmpp is 65V) so massively reduces the utility of the panels.

Am I correct on the voltage thing?

0 Likes 0 ·
kevgermany avatar image kevgermany ♦♦ Graham Walmsley commented ·
Voltage, battery actual +5 to start charging. Vmpp is max power point voltage. Once charging starts the controller will adjust to get max power from the panels depending on the amount of light. Voc is the one to worry about. This is effectively the highest the panels will reach under standard conditions and is the value that can kill the controller. Beware, this increases as temperature drops, there should be a value/K or degree centigrade in the specs, also the Voc temp, usually 25C.

With your panels, nothing to worry about for charge start.
0 Likes 0 ·