REC Master BMS/2Q Max System Size


  • Good day everyone,

    I hope all is well.

     

    To begin with, OGM staff please note this is a rough duplicate of a contact form I submitted. I realized after submitting the form that posting here in the forum would make the discussion available to the world, available to future readers, and would allow the community to offer any thoughts or ideas as well. Please feel free to respond wherever is most appropriate.

     

    I am investigating REC BMS solutions for a sizeable house in a remote region. The energy storage system is being designed such that the reserve battery capacity allows for a substantial number of so called Days of Autonomy. To accomplish this, the number of REC units involved would be substantial... hypothetically in excess of 30 2Q Slave units once the system is complete. Given the size of the investment, I would like to discuss some particular capabilities of REC equipment prior to commiting to purchase.

    In case my fundamental understanding is flawed I would like to start with the relevant limits as I understand them:

    1. One (x1) REC Master BMS is capable of communicating with a maximum of Fifteen (x15) REC 2Q BMS units at a given time.
    2. Communication between the Master BMS and the Slave 2Q BMS units is performed by way of an RS-485 connection between the Master/2Q units.
    3. Each 2Q BMS in the, let's say, x15 long RS-485 chain must have a discrete network address to identify one unit from the others. This address is selected via a physical rotary switch on each 2Q.
    4. The final (i.e. last in the sequence; ex. #15) 2Q BMS needs to have a termination plug of a given value at a given location as specified in the manual to signal to the Master BMS that it is the final unit in the RS-485 chain.

     

    With that foundation laid, a possibly complicating factor has arisen.

    Paralleling cells at the sub-pack level prior to series connection is not sought at this time for reasons that are worth a thread of their own.

    That is to say, each Battery Pack will be composed of 16 cells in series, for a 16S(48v) configuration. Per the above, that implies a maximum possible system size of Fifteen (x15) such 16S packs.

    There is concern this may not be sufficient reserve capacity. Solutions are being sought.

    One such possible solution is the use of a Secondary bank of x15 16S batteries each with its own 2Q BMS. Consider a scenario where x15 2Q + Battery Packs are in active use and connected to a single x1 Master BMS.

    • A certain Battery Pack out of those fifteen, let's call it Primary Battery #10, reaches some State of Charge level deemed critically low.
    • Its power connections to the system busbar are disconnected.
    • Its communication connections(recalling that this is Primary Battery/BMS #10; so that means the inbound cable from BMS #9 further upstream and the outbound cable to BMS #11 downstream) are both disconnected.
    • Some seconds later another battery pack, fully charged and which we can call Standby Battery #10, is connected power and comms both in place of where Primary Battery #10 just was.
    • Its address is set to the exact same value as Primary Battery #10.

     

    Ideally, from the Master BMS' perspective, all that occured is 2Q BMS #10 was briefly disconnected and then re-connected. The sole difference being that it is now reporting a higher SOC.

    If we assume this switching is done electronically over the course of some number of milliseconds, and for each of the x15 Banks in rapid succession what will the response of the Master BMS be to such transient interruptions?

    There are some non-BMS electrical concerns that also arise from this type of switching- but for now I would like to focus specifically on how the REC Master BMS will handle a temporary interruption of the communication bus in this manner.

    Namely, I am concerned:

    1. If Battery Number #10 is physically disconnected for any length of time, in a similar fashion as Christmas tree string lights when one bulb fails, all 2Q BMS units in series further downstream of #10 will no longer appear to the Master BMS for the duration of the disconnect, given that the chain is physically broken from the point of disconnection onwards.
    2. There would no longer be a terminating plug the Master BMS could detect indicating the end of the RS-485 chain.

     

    If the 2Q BMS, when they cannot properly establish a connection with the Master BMS, revert to their local/individual settings then I expect power interruption to the system bus to be minimal.

    It is also possible to switch the Comms cable going to the Inverter from the Master. So a Primary Master BMS at the head of a battery array would have its Comms cable switched with another Master BMS that is at the head of a second battery array.

    The final point I should mention is that I intend to use Victron's DVCC utility. I believe I could, in a manner of speaking, run the Master/2Q array independent of the Quattro Inverters and use the programmable outputs to signal end of charging and the like but doing so introduces another set of problems which DVCC very conveniently solve.

    It is specifically for the use of DVCC why I am endeavoring to maintain a physical communication channel between a Master BMS and the Inverter at all.

    In summation then, the question is how might an excess of 15 battery packs be used with the REC ecosystem without fully cutting off power to the system bus and with the clarifier that there need not be more than 15 connected at any given time. There just needs be a way to rotate out depleted packs and introduce fresh ones while having features like DVCC function as expected afterwards.

     

    Thank you for your time.

    All the best.



  • Here is some feedback from REC engineering that may address your challenge while giving you a level of system redundancy at the same time.  Hope it helps

    I would go with 2 parallel systems.   So if there are say 20 Slaves I would put 2 x Master BMS with 10 Slaves each and 2 GX devices and distribute Quatros evenly.
    So only the GX device would be an additional cost

Please login to reply this topic!