I have an off-grid system with a MPII 5kw with 2 x US5000C pylontech batteries and 3kwp.
I will need to increase my inverter and battery power as I am planing to install a inverter heat pump. I will need at least 10kw of inverter power to cover the heat pump start peak and around 40kw more of battery power to cover 48 hours of non solar production.
40kw of pylontech battery power is very expensive and I will need to select a cheaper battery option. Unfortunatelly a Victron system is not able to manage 2 battery banks yet and I am looking for alternative solutions.
One option would be the âchained MPII designâ that I have seen in Andyâs Off-Grid Garage videos in youtube, where two MPII are chained (ac-out of the first mp2 is connected to ac-in of the second MPII) and you use the Power Assist and relay features to control how both MPII work together.(https://youtu.be/SsTkgoN6ENM?si=QIE6MJlqDA4Vxvx1)
I would connect the big and cheap second battery bank to the first MPII that will emulate the grid and it will assist the second MPII when load demand reach some level.
Of course, I will have 2 diferent systems as both MPII will not see each other and I will need 2 Cerbo to monitorize both systems but they will work as a unique solution and I will be able to manage 2 battery banks.
I have not seen any reference in the community to this chained design that it seems enough robust and simple.
Am I missing any potential big issue by chaining 2 MPII with power assist and relay features enabled?
If it really is an âinverter heat pumpâ, then no current peak is to be expected ⌠unlike with a normal scroll compressor. The frequency is regulated up very slowly and since there is no heavy starting here but only a scroll compressor, there should be no problems when starting.
You can have more than one battery bank connected to a gx, the requirement is that the charge voltage is the same for both banks. And only one controlling bms.
The second inverter can be added to the gx through the mk3 to usb but for monitoring only no control and no information on any overview pages.
This is correct, but people still do it -search cascade set up.
Thanks LX but one batery bank is a Pylontech US5000C bank and the other bank will be most probably a bank of 3 x 304A primatoc bateries with JK BMS.
Do you think that I can run both banks in parallel with only one BMS connected to Cerbo (the Pylontech bank)?
Perhaps I am wrong but my view is that as the Pylontech BMS will dictate to the mppt and the inverter the charge/discharge parameters, the other bank will have to work with wrong or not the best parameters. For instance, pylontech can be full charged and the second bank be only at 50% and mppt will considernrhat the batreries are already charged. Am I wrong?
On the other hand, Pylontech is a 15 cell battery and the other use to be 16 cells batteries and the voltatges will be diferent. I can build the second bank with only 15 cells but I do not know if this will be enough.
With the jk you can program it how you want to, so make it match the pylons. 52.4v when balancing starts, cell number, etc.
Use Bus bar, no daisy chain. You will find unequal amp draw between the batteries but you will find they will even out.
It will be an interesting set up for sure.
For SOC you can decide what to do (on my mix bank also using jk- not pylons- i have a shunt now. No controlling bms) I do find though that the one battery shuts down before others as it seems to discharge faster. But it wakes up again when charging starts up again.
If I can have both battery banks working together then I have to decide how i increase the inverter power.
I could keep the initial design where both inverters are chained but now I can use the parallel design as both inverters share the same battery banks.
Any thoughts?
Would it be better to build a 15 cells battery to match with Phylontech design or it will be enough to setup JK BMS wirh Pylontech charge and discharge parameters?
On the other hand, why do you use a shunt? Your JK BMS does not have CAN suport to connect with Cerbo?
So the SOC of the two batteries is different due to their different capacities.
I also had the concern that the DCL from either battery bank would shut down the system unexpectedly.
So ended up disabling both bms controls and using the shunt and didnât want to use the inverters internal soc to run the show.
Keep to 15s and use the same charge voltages would be an options.
But since you can program balancing where you like with the jk bms run 16s at lower charge voltages.
Which route you choose would be up to you there.
What about the inverters: chained desing or parallel design? Both cases would share the same battery bank with the 2 banks (pylontech and prismatic+JK BMS)
Daisy chaining or cascade is what andy demonstrated. But it is not a supported set up officially.
My advice is donât ESS one off the other if you do decide to do so.
You could always have a system just dedicated to the heat pump and if there are issues you donât loose the whole lot. You are already buying more solar so you almost at a whole system. And 16s battery is more efficient than the pylons anyway.
Apologies LX but I do not understand this sentence:âMy advice is donât ESS one off the other if you do decide to do so.â
My idea was to connect the heat-pump to AC-out2 and disconnect this output when SOC reach some low level to protect the critical loads
Why do tou say that 16S primatic batteries are more efficient? I thought that Pylontech batteries with its propietary BMS are better fine tuned to get the maximum performance of the battery. In 16s, people use a generic bms not especialized for a concrete battery rype.
Yes, this was my initial propossal when I thought that I could not mix 2 battery banks.
As you told me that it is to possible to mix 2 banks but only connect 1 bms and the other bank charge with the parameters of the pyylontech bms, perhaps it would be better the attached design with 2 inverters in parallel and monitor the second battery bank (15s instead 16s to match wirh pylontech voltages) with a shunt.
Would it work?
Could I have an agregated SOC (pylontech plus jkbms bank)?
The soc wonât agggregate unless you use the shunt for the entire system soc.
Both batteries have to use the same charged and float voltages. I see you have that in your design (nice use of excel by the way)
Both the jk (with a driver load from git hub) and the pylon can be viewed for information on the gx and transmit to the dashbord on the vrm.
If you can avoid it use a 10kVa not 2x 5kVa in parrallel (those types of systems can be a nightmare.) especially if you are adding a new unit to an older one.
See this post.
You are right! I forget the issue with MPII not manufactured the same year. I was thinking to add a new 5kw box to minimize cost as 10kw MPII costs the double than a 5kw box.
To avoid manufacturing date issue, I can move to a cascaded design with a new 5kw box as initially planned or, as you proposse, replace the existing one with a 10kw inverter and resell the actual 5kw.
Regarding the agregated SOC topic, if I move the shunt to the negative cable of the 10kw inverter (assuming one inverter only) where both battery banks are already behind, would I have an agregated soc in gx and vem and I will be able to use the agregated SOC to control AC-Out2 ?
My main concern is how to control AC-out2 line to protect critical loads in AC-out1 when you have 2 battery banks. The ideal would be to dedicate Pylontech bank to critical load and JK bank to heat pump in AC-out2 or control AC-out2 with the agregated SOC value of botg battery banks.
Regarding the github driver for JK battery bank, how do you connect JK battery bank to Cerbo if you prefer to have separated bank view in gx and vrm? You already have the Pylontech BMS connected to Cerbo via CAN bus.
