BMS integration to Victron + system design

dewki

New member
Joined
Aug 12, 2019
Messages
6
hello!

I ask for assistance in selecting a BMS for a home energy storage system (on grid, solar buffering).
I'm not quite sure what information to provide. I've estimated the nightly demand is approx 7kwh (calculated for 10hr use).

Already installed:
SolarEdge SE9K three phase AC coupled interver. 11.2kWp total panels.

I am looking to couple the SolarEdge with the Victron infrastructure.
To be purchased:
-Victron infrastructure (Venus control unit) https://www.victronenergy.com/panel-systems-remote-monitoring/venus-gx
-A Multiplus inverter/charger. Charge capacity around 35-70A depending on model: https://www.victronenergy.com/inverters-chargers/multiplus-ii
-Winston LiFePO4 prismatic cells. Capacity somewhere between 200Ah to 400Ah in 48V nominal setup. (24V might become relevant depending on multiplus pricing). Final budget will decide the size of cell. Estimated 10-14kw setup.
-BMS that can talk to the Victron control unit.

Questions:
1. I've found Batrium and REC that have some sort of verified "unsupported compability" with Victron. If there are vastly better options (price or performance), please link?

2. Does the required balancing capability of the BMS differ between a 18650 based bank, and a bank based of 400Ah cells. Any rules of thumb or way to navigate a suitable bms?


Cheers
 
>2. Does the required balancing capability of the BMS differ between a 18650 based bank, and a bank based of 400Ah cells. Any rules of thumb or way to navigate a suitable bms?
Batrium (among other things) is well suited manage/balance *large* cells. Doesn't matter if they are LiFePO4 or 18650. Batrium can manage a lot of cells and large cells. 400ah is middle of the road for Batrium. Batrium is so flexible you can add more than 1 longmon (cell device) per cell and increase balance speed. The rumor is that a single longmon has been successfully deployed with up to a 1000ah cell. My own cells are 260ah and balance very easily.

I'm not familiar with Victron equipment so can't comment other than to say that Batrium control options are quit extensive.
 
Batrium works great. In conjunction with Victron i cant comment

Wm5 for a single bank is generally enough. Wm4 + longmons and you can have 10000ah if you want with enough longmons.

All depends on how much you need to balance. Note that 18650 is just a form-factor and nothing else...

I run around 400Ah per longmon. Its in the middle recommended figure. Read this: https://support.batrium.com/article/91-cell-monitor-pack-design-considerations
 
Definitely go 48V & as large batteries as you can afford.
You'll regret it later if you go with 24V for the size of system you're doing.
If you have the time and inclination, you can build with 18650's but going with prismatic cells in an all new system is a good choice for a faster build.
Many folk here use new or second life 18650 cells (& respect to them!) but you have to put in the time to go that way.
 
Thanks a lot for the suggestions guys!

@daromer thanks for the link. that provide all the sense of security I was looking for. Indeed does Batrium look promising, particularly in the longer run.
Interestingly the REC BMS also provides 1.3A / cell worth of passive balancing and thus would be comparably suitable for my size of cells unless I miss something?
https://www.rec-bms.com/BMSSMA.html

@Redpacket 48V - check. I will adhere to that advice.
Unfortunately I am under some time crunch to get this battery solution up and running (must be comissioned and inspected before end of 2019).
I like the idea of working with smaller amounts of cells, particularly after watching a few videos by @daromer where bank capacity has been mentioned. It is indeed appealing to grow using smaller cells rather than prisms (ie, the increments that make sense are smaller).

I was hoping for someone to give me some reassurance with the Victron-infrastructure integration.

Maybe I should reconsider??
I am located in Sweden. Maybe @daromer has some fitting suggestions?

Maybe I have overestimated the necessity of the BMS to communicate with the charger? Please share your thoughts.
Currently I am thinking that without proper communication between inverter&charger and BMS the overall system design is suffering considerably.

Looking forward to further discussion!
 
Victron is a well proven provider, Ive used their products in commercial remote installs, not with a Batrium BMS though.

My home system which is under construction is using a Victron Multiplus, Victron Venus and Batrium Watchmon 4.

Batrium has information on configuration settings to communicate with the Venus.

I dont have proven results with the combination yet though sorry.


image_fuhqtt.jpg
 
Victron (Venus device running ESS, Multiplus, MPPT, energy metering etc etc ) & Batrium or REC work well .

But - dont expect to be able to use the full bypass current of either without additional cooling (a well built battery won't require much balancing), dont expect to be able to configure such a system with negligible prior knowledge, any existing grid tired installation using a inverter that supports frequency shift control can be simply integrated, dont assume 14s normally 48v is the correct battery string configuration (build the battery based on your inverter/chargers), dont under estimate the cost of fusing/protection.

REC support is superb and very quick - Batrium support seems to still be as you'd expect from a very small company - rightly so, both companies assume some competence.


image_tianfo.jpg

image_britdm.jpg

image_wazigm.jpg
 
"SolarEdge SE9K three phase AC coupled interver"
 
From batrium you can balance 10A from a single cell if you want. Just add 10 longmons as example.

For continues 1A per longmons is rather normal. (Depending on ambient temperature and how you mount them)
 
daromer said:
From batrium you can balance 10A from a single cell if you want. Just add 10 longmons as example.

How do you ensure that all mons turn on at exactly the same voltage ?
 
Huh? Not sure what ur question is about? Do you need to have Them 100% synced? They Will balance when needed. Andif cells are above threshold they Will all start to bleed off and you get the full current
 
daromer said:
Huh? Not sure what ur question is about? Do you need to have Them 100% synced? They Will balance when needed. Andif cells are above threshold they Will all start to bleed off and you get the full current

..... in reality, only one mon starts bypassing (the one with the lowest turn on voltage) - if you dont limit current when in bypass, and actively cool, multiple mons is ok, but the few fractions of a volt difference in their calibration and no ability to sync all mons makes this a somewhat impractical and expensive configuration.

If all cells are above the bypass voltage then yes they'll all go into bypass immediately, but again the slight difference in calibration will see them all coming off bypass a few fractions of a voltage apart.

Jaron did hint at enabling user calibration of mons ages ago, which would have made multiple mons on the same cell a more usable setup.
 
Not sure it would be a problem. IF you need more than 1 mon to balance the they Will turn on One after another. Think of it going 1a then 2a.
I have used it like that for testing and i know other doing it without issues.

As Said i dont see the issue with it
 
Sean said:
daromer said:
Huh? Not sure what ur question is about? Do you need to have Them 100% synced? They Will balance when needed. Andif cells are above threshold they Will all start to bleed off and you get the full current

..... in reality, only one mon starts bypassing (the one with the lowest turn on voltage) - if you dont limit current when in bypass, and actively cool, multiple mons is ok, but the few fractions of a volt difference in their calibration and no ability to sync all mons makes this a somewhat impractical and expensive configuration.

If all cells are above the bypass voltage then yes they'll all go into bypass immediately, but again the slight difference in calibration will see them all coming off bypass a few fractions of a voltage apart.

Jaron did hint at enabling user calibration of mons ages ago, which would have made multiple mons on the same cell a more usable setup.

It would be nice if there was a sync socket on the mons so if you wanted to stack them up, you would just need to connect a sync cable between them
 
HugF You just connect them in series and sync them in software. Yes you would see more longmons as such and each 2 that is on same basically give you the same values. The part showing the multiple longmons though can for some people yield a disadvantage of course.

Sean why is it relevant to give you name on people who have done it? As I said: Its more then possible to run multiple longmons to increase the continues balancing capacity.
 
daromer said:
HugF You just connect them in series and sync them in software. Yes you would see more longmons as such and each 2 that is on same basically give you the same values. The part showing the multiple longmons though can for some people yield a disadvantage of course.

Sean why is it relevant to give you name on people who have done it? As I said: Its more then possible to run multiple longmons to increase the continues balancing capacity.
A sync cable would be more logical. You wouldn't need to have them display as a separate mon in the chart...
 
Since they are separate units totally i dont see the reason. If so it would be better to group them in the UI. You still want the data from each mon so you know they all work as intended. With that said its all about how they are designed and I prefer how they are currently designed.

With that said I dont see any real life setup where you would really need higher balancing than 1 mon can provide :) Since i rather have separate battery banks.
 
Jon said:
Victron is a well proven provider, Ive used their products in commercial remote installs, not with a Batrium BMS though.

My home system which is under construction is using a Victron Multiplus, Victron Venus and Batrium Watchmon 4.

Batrium has information on configuration settings to communicate with the Venus.

I dont have proven results with the combination yet though sorry.

image_fuhqtt.jpg


Jon, Looks like we are looking into similar setups :)
Thank you for chiming in.


Sean said:
Victron (Venus device running ESS, Multiplus, MPPT, energy metering etc etc ) & Batrium or REC work well .

But - dont expect to be able to use the full bypass current of either without additional cooling (a well built battery won't require much balancing), dont expect to be able to configure such a system with negligible prior knowledge, any existing grid tired installation using a inverter that supports frequency shift control can be simply integrated, dont assume 14s normally 48v is the correct battery string configuration (build the battery based on your inverter/chargers), dont under estimate the cost of fusing/protection.

REC support is superb and very quick - Batrium support seems to still be as you'd expect from a very small company - rightly so, both companies assume some competence.


image_tianfo.jpg

image_britdm.jpg

image_wazigm.jpg

Sean, thisis solid advice.
I am indeed a newbie to this area. ButI do have a lot of experience in12VDCfrom a decade'sbuilding and designingcar audio applications (some of which have carried battery packs >1000Ah, fused far beyond 1000A). I'vebalanced and assembled packs of super capacitors with balancing circuits.Some of the skills learned I expect tocarry over to some extent, likecable sizing calculations and fusing practices.
Anyways, I put this out there to let you guys have an idea where I come from.

Thank you for confirming my concerns regardingcharge voltage. I've identified this issue.The MultiPlus 48'scharging property from the spec sheet (although it is reducable in settings AFAIK), is 57.6V Absorption, 55.2V Float, 52.8V Storage mode.
Quick calculation showed this might be problematic. Here is how:

The Winston260-400AhCells arespecified to a operating voltage 3.8V - 2.8V (80% DOD). Recommended charging at 3.65V per cell.
For the curious:https://www.ev-power.eu/Winston-40Ah-200Ah/

Quickly noting (3.65 charging volt):
16S: 58.4V <<< too high
15S: 54.75V << within range for the Multiplus
Evenat a charge voltage of3.8V, the 15S is still applicable (57.0V).

completelycharged said:
"SolarEdge SE9K three phase AC coupled interver"

I anticipate you're indicating that the Multiplus alone are a single phase modulewhich potentially would cause problems?
Actually, this isa supported configuration when running ESS (Energy Storage System assistant).

https://www.victronenergy.com/live/ess:design-installation-manual#single_phase_ess_in_a_three-phase_system
ESS also would require 3 phaselike:https://www.victronenergy.com/accessories/energy-meter#pd-nav-secondimage

I've been asking suppliers of inverter/charger solution that would be capable of running three phase AND not being locked into a ecosystem of a Plug n' Play battery solution which comes in at approx 5-6kWh for the same price that I'd be able to build an approximately doubled size battery bank. So far, no other model has been available.

That's one reason I've come to stick with the idea of Victron.
Another is the potential future integration with Node-Red that is in the makings..

https://www.victronenergy.com/accessories/energy-meter#pd-nav-secondimage
Cheers!
 
daromer said:
With that said I dont see any real life setup where you would really need higher balancing than 1 mon can provide :)

I agree, but you previously mentioned on this thread that others are using multiple mons to achieve a higher bypass current, I assume with success, who are they ?


@Dewki - have a really close look at the max/min settings for bulk/absorb etc before you finally commit to string length - max bulk on the MPII 48v is actually 64v, as per the larger MPPTs such as the 250/100 - which suits a 16s string with a max cell voltage of 4v perfectly - obviously chemistry dependent - I cant recall the min but its helpfully very low.

Single phase ESS on a multiphase supply works very well, it's very well supported - we have some Schneider equipment arriving shortly that also perform well which you might want to add to your list for consideration.

Node Red has been with Venus for a couple of releases.

https://github.com/victronenergy/node-red-contrib-victron
 
Back
Top