Maximum amount of SOK SK48V100 batteries

[Joske]

Hallo Everyone, (especialy @HighTechLab )

I don't know if this is a common question but I am sure they are people or businesses out there that maybe have the same questions. So i thought to post the question here as reference material for everyone to see and the future questions.

Ideea:
I have a customer how would like to have a battery storage system. After researching the scope of project his capacity of this battery would be +/- 250MW. A not so common and large setup of battery storage, but for this project the propper solution. Now we are researching the possibility to make a system combining Victron and SOK SK48V100 Rack batterij and it would take a +/- 50 batteries to buffer this amount. But the SK48V100 can only communicate with 16 batteries parallel at the same time.

Question:
A- Is there a way that SOK can agragate more than the stock amount of 16 SOK SK48V100 batterys to victron? if so, how?
B- If not, what would be the solution to combine the BMS communicate with multiple strings of 16x SOK SK48V100 batteries to victron, for instance like a Pylontech Powercube X1?
C- If A-B has no solution, would the SK48V100-NCVT (type without BMS, not our preferente) in multiple strings with multiple Victron SmartShunt would be a solution?
D- If not, what is your thoughts on an other solution on project?

 

[HighTechLab]

Yeah, it can be done with voltage control. I have a guy in Michigan running 100 batteries and he's using voltage control (no communication). He has 12 Sol-Ark 15k's but the premise is the same.

I had a communication hub in the works, but when I started doing large-scale testing, I found it was in no way better, not even marginally vs voltage control. Frankly, the only advantage for communication on the smaller systems is that you don't need a shunt (because the battery has one built in). For larger systems, you really want a nice and precise dedicated shunt; for example the SmartShunt 2000A.

One gotcha on these batteries is that the BMS on the pro model can't accurately measure current less than .7A; this means that with 50 batteries, if you are running less than 2-3kw load or charge, the shunts don't actually track it properly

On most Victron systems that do have communication, I usually end up setting the BMS connected to report SOC/wattage, but don't make it the controlling BMS in DVCC and let the voltage setpoints still do the charge and discharge regulation.

 

[Joske]

@HighTechLab / Dexter: Thanks for the fast reply.

Voltage control with a SmartShunt 2000A is a good solution. Our thoughts with the BMS communication was: keep track on the batteries health such as individually packed- / cell imbalances or errors and let them report via the VRM. If there is no BMS reporting to VRM we can only see possible fault if we are physically on premise. With that said, we can add a SLA so that we can do frequently schedule check up onpremise to keep track on any future problems that can occur while it is aging.

I do have some questions:
A- Witch type of battery does your guy in Michigan use, the SOK NCVT or some other?
B- "One gotcha on these batteries is that the BMS" Do you mean whit "these batteries" large-scale batterie setup? And if so, by using voltage controlled dis/charging and a SmartShunt 2000A, there is no BMS connected because then you can use the SOK NCVT battery that don't have a build in BMS. Or am i not understanding your solution correctly?
B- "BMS on the pro model" What do you mean with pro model? Of the the SOK batteries?
D- What do you mean with "On most Victron systems /.../ report SOC/wattage" DO you mean smaller installations with max 16 SOK batteries?

 

[Adam De Lay]

Yeah, it can be done with voltage control. I have a guy in Michigan running 100 batteries and he's using voltage control (no communication). He has 12 Sol-Ark 15k's but the premise is the same.

Wow, 100 batteries and 12 15ks?!!! Makes a guy a little jealous....

 

[HighTechLab]

A - the one with the screen. Seems a lot of customers are now calling this this "pro" model and the NCBT is the "base" model.
B - No matter how many of the pro models.
B2 - refer to A
D - yes, with max 15 batteries.

Hope we are on the same page now - I've been out sick part of this week so sorry if this is not making sense.

Wow, 100 batteries and 12 15ks?!!! Makes a guy a little jealous....

Pretty sure he was using it to mine crypto...not sure though.

 

[Joske]

This was helpful, thanx for the awnsers and get well soon.

 

[itech20]

Yeah, it can be done with voltage control. I have a guy in Michigan running 100 batteries and he's using voltage control (no communication). He has 12 Sol-Ark 15k's but the premise is the same.

I had a communication hub in the works, but when I started doing large-scale testing, I found it was in no way better, not even marginally vs voltage control. Frankly, the only advantage for communication on the smaller systems is that you don't need a shunt (because the battery has one built in). For larger systems, you really want a nice and precise dedicated shunt; for example the SmartShunt 2000A.

One gotcha on these batteries is that the BMS on the pro model can't accurately measure current less than .7A; this means that with 50 batteries, if you are running less than 2-3kw load or charge, the shunts don't actually track it properly

On most Victron systems that do have communication, I usually end up setting the BMS connected to report SOC/wattage, but don't make it the controlling BMS in DVCC and let the voltage setpoints still do the charge and discharge regulation.

This is a slightly off-topic follow up, but I'm intrigued by the 0.7 A accuracy comment.

I just installed one of your 5 battery rack mount kits. If I extrapolate, it sounds like loads below 200 watts will result in inaccurate internal shunt readings on my batteries.

I take it the main issue here is that the BMS SOC calculation will be inaccurate. True?

The inverter in this case is an EG4 6000XP. I'm unsure if it's currently able to support the hybrid "communication for SOC and output reporting but voltage for charge/discharge control" configuration that you describe in message #2 in this thread.