who is in control? ElectroDacus VS Victron

[alvise mj]

Hello everyone new to the forum. I am setting up a system with a Victron multiplex 24/3000, Victron Battery protect to Orion 24x12, 3 x 340W panels 2 DSSR20 with 3.2v LiFEPO4 8S=24V;

My question is do I make ElectroDacus the primary management of the battery (overcharging , low battery, temperature etc) or I set that in the Multiplex and keep SMBS backup if something go wrong or something that Victron can not detect? 

Anyone want to share the parameters setting of the 2 units ?

If I use the multiplex to do the work  i will probably need to add a shunt and Victron BMV-712 Battery Monitor with Battery Temperature Sensor, is that worth it ?

[Dacian Todea]

Of course the Electrodacus SBMS0 is the BMS (Battery Management System) and it needs to be in control of the battery.  None of the Victron equipment will be able to see anything other than total battery voltage that is completely useless when protecting a Lithium battery where you need to know each individual cell voltage and protection is based on that.

I guess you have the 3000VA multiplus (not multiplex) and for that here is the spec on how it will integrate with an external BMS including SBMS0 that is compatible with the two signal setup 
https://www.victronenergy.com/upload/documents/Manual-Connecting-other-lithium-battery-systems-to-Multis-and-Quattros-EN.pdf

The BMV-712 will be useless as the SBMS0 will do all that plus quite a bit more.

[ask first]

SBMS0 is the future. 100 years ahead of its closest competitor. 

[TheoW]

Funny how quiet this thread is.

Of course as with many things there are lots of choices.   For a number of reasons I've gone the second route.  Basically all of the charge sources manage themselves and the SMBS0 is my backup safety device should something go wrong with an individual cell.  I've had my battery bank running for about two and a half years and other than my testing, it's never hit an over or under condition with an individual cell.  I charge (and discharge) my bank somewhat conservatively, stopping charging when the pack hits 27.6 (3.45/cell).  In doing this my actual capacity tests over the years have shown that I still get 400 AH out of my pack which is fine for me so I'm not worried about pushing the shoulders of the curve to perhaps get slightly more.

I had the victron equipment (and another BMS) up and running before getting the SMBS0 and it was operating just fine so I didn't see any need to change all to let the SMBS0 control everything.  Also I do prefer to have redundancy and safety backup systems.... and really want to control safety high and low cut off separately.

The SMBS0 is a great little BMS and solar charge controller, but I am also a big fan of the Victron equipment and I find that the BMS 712 combined with my other victron equipment, specifically the open source Venus OS provides much more capability than the monitoring of the SBMS0.  Specifically the connectivity of the Victron stuff is great, bluetooth connectivity, CAN connection, free logging to their web portal, the ability to change charging profiles anywhere I can connect to the internet, email notifications of high/low or loss of shore power and more.

Bottom line, your setup your choice there are lots of ways to have a good setup depending on your exact needs and wants.

Cheers,

ted

[John W]

Thanks for this info as just starting down this path.

Was planning the "usual" Victron setup of Multiplus, Solar charger, Cerbo, Touch screen, and lynx bus/shunt.  For batteries was going to go with 8 BattleBorn 12V/100Ah in a 24V configuration.  In that setup batteries would be around $8000 USD.

Now looking at the DIY battery side which would be with 8 of the 400Ah Winston cells with the ElectroDacus controller.  That's $4000 for the batteries, plus for the solar side could eliminate a $600 Victron MPPT charge controller and go with the DSSR's.  

But in both configurations the Victron side would be the user facing controls with the ElectroDacus BMS just being a way to get more fine tuning of what's going on under the hood of the battery banks/solar.

Packaging of the Winston is lot more compact as well which helps in limited space.

Decisions, decisions......

[Dacian Todea]

Ted,

No the charge sources can not manage themself and need to be controlled by the SBMS0 unless you want to overcharge and damage an expensive battery.

What you say is anecdotal evidence and wanted to make this clear if anyone reads your comment.

No matter how equal the cells are there will always be a bit of imbalance even if very small and one of the cells will always be the first to be fully charged.

Up to around 3.4V it takes forever but when cell is almost full and can no longer accept charge the voltage will increase very fast so you can have a situation where all cells are almost fully charged say 3.4V x 7s = 23.8V and if you set the charge limit to 27.6V then the remaining cell can be 27.6 - 23.8 = 3.8V and while LiFePO4 is fairly robust and can handle that level of overcharge it will overtime degree especially that cell.

Also absorption and float are detrimental so charging should stop as soon as any cell gets to 3.55V (default for LiFePO4 on SBMS0) or 3.65V if you want to push to where manufacturers recommend (but there is no gain in doing that). 

 

John,

All those Battleborn and similar 12V battery's are not only expensive but also a bad idea as they allow for overcharge of individual cells not to disconnect the battery when full as that will also cut out the load.

The popular cell now are the 280Ah CATL for just around $100 each so a 8s2p for 24V 560Ah will cost just around $2000 that is half the cost of Winston

There are many sellers for those but here is someone that ordered and made a video about the cost https://www.youtube.com/watch?v=21YY4HHdQYI

[TheoW]

Hmm, I don't see any difference in the core of how each setup would work.    As I said the SBMS0 is the backup device should something go wrong with an individual cell.

For example the charger is charging the pack to 27.6 and the SBMS0 is watching individual cells and switches a High voltage disconnect should any cell get to high (3.55 in my setup).  So a setup like this will not overcharge any cells either.

[Dacian Todea]

TheoW,

The difference is that charger can only see total battery voltage (useless info) while a BMS (any BMS including SBMS0) can see each individual cell voltage and thus is the only device that can properly protect the battery by stopping the charge or discharge when any of the cells gets to the set limits in this case default are 3.55V all chargers should stop and 2.8V all loads should stop. The chargers and in same way loads can only see total battery voltage and that is not useful as one of the cell can be outside the limits and based on total battery voltage is not possible to know.

Also keeping the battery forced to say 27.6V is detrimental (it will accelerate aging) as Lithium of any type dislikes float charging. With SBMS0 charging is done at full current the charger is capable of providing until a cell gets to 3.55V then battery is considered fully charged and charging will completely stop and not be enabled until battery SOC drops enough to require reenabling charging and this is a much more favorable charging for Lithium batteries.

And yes if the EXT IO4 set as type 1 (HVD) is used to disable charger no cell will be overcharged but by setting the charger voltage to low you are keeping the battery forced at full charge for many hours still detrimental may be reducing life of the battery to half (maybe not that drastic as LiFePO4 is fairly tolerant) but if you had any other type of Lithium like NMC or LiCoO2 then float charging on those will have had a way more drastic effect.

If SBMS0 where to fail it will fail as open circuit meaning chargers and loads will all be disabled (assuming correct installation) and if the remote ON/OFF wires are broken (most likely fault) then again the charger lr load controlled by that will be OFF as open circuit means OFF and it will again be safe.

[TheoW]

sorry still not getting a difference, I'm in agreement that a BMS is the only way to monitor cells and that is needed.  Sounds like you're saying stop charge if a cell hits 3.55 and I'm saying stop charge if a cell hits 3.55.

Not sure why you mention float but agreed float is not good for LifePO4.  I don't float my batteries.

I guess I hadn't given to much direct thought about the fail safe mode of the SBMS0, but that sounds like another potential reason have separate paths for my install.  I wouldn't really want one failure to stop charging in my systems, I really want to have a way to keep power to my bilge pumps, navigation lights, nav systems, windlass, etc operational. So a single point of failure seems be less that desirable for me.

Thanks,

ted

[Dacian Todea]

Ted,

Not all cells will get to 3.55V at the same time one of them will always be first to get there while all the other may be just around 3.4V

Based on your 27.6V setting one of the cell can get to around 3.8V and that is ignoring any possible cell failure.

Your system seems like a supervised system (meaning a person is close to the system and can intervene) so in the unlikely case where a charger fails to work either because the charger failed or the remote ON/OFF signal got broken an audio/visual alarm can be used to say alert the user that battery is below 30% SOC or whatever you consider significant.

In case of high reliability two separate systems including separate battery/charger's will be ideal the isolated DC-DC converters from each battery with the output of those paralleled supplying essential items.

There are about 7 years since I'm offgrid with different versions of SBMS and I was never with an empty battery tho I can not say that I have critical systems unless you count the fridge critical and I do not have any other backup source just PV solar. Yes I was a few times close mostly my mistake and I had twice damaged equipment from lightning (indirect but very close) still nothing significant most expensive was first time a raspberry Pi that was logging data form SBMS4080 trough 2000V capable optical isolation.

[Marinepower]

Dacian,

for what it's worth.....

TheoW's approach is common in the marine world for retro fit DIY LFP.  Charging is done based on pack level voltage on the assumption that the cells are relatively well balanced at the top of the charge curve.  Staying out of the top of the knees helps in this regard. If pack level charging to an average 3.45v and a cell goes out of balance from the pack's average by more than 100mv at the end of the charge (thereby charging that cell over 3.35v)  , the BMS triggers an audible alarm and the charge bus relay cuts off (or it remotely turns off all the charge sources).  So no one cell can go above 3.55v (or whatever the BMS's set HVC) without charging being stoped.  

I'm not trying to suggest this is necessary the best approach.  Please don't flame me. Ideally all charge sources would have been designed from the ground up for LFP charging and would monitor each cell during charging.  This part of the beauty of what you have created (and what the market was lacking) with the SBMS0 and the DSSR20.  But we are stuck trying to retrofit legacy charge sources designed for SLA/AGM (with tampered charge curves) to work with LFP charging (with remote cut offs while bulk charging).

[Dacian Todea]

Marinepower,

The max allowed delta can not be as small as just 100mV as that will be easily triggered during cell balancing and even if cell balancing is disabled during loads or higher charge rates is easy to exceed 100 or even 200mV delta and there is no delay for that so even a few ms exceeding that max allowed delta will trigger a fault.

It is not a good idea for all charge sources and all loads to have internal BMS to measure all cells. Just a single BMS with signals to stop charge or discharge like the SBMS0 should be good enough and quite a lot of devices chargers and inverters have this capability or can be easily modified to handle that.

There is no problem for Lead Acid designed chargers and inverters to work great with LiFePO4 as long as ON/OFF can be implemented they do not need any special requirement.

Lead Acid charges are set typically for constant current (bulck charge) until they get to 14.4V (28.8V) and since that will never happen with 4s or 8s LiFePO4 even if all cells get at the same time to 355V (next to impossible) then the charger will work in constant current mode only and when first cell gets to 3.55V so full battery the EXT IO4 will be open circuit and that should turn OFF the charger. 

Charger will then resume (turn ON) once the battery has discharged sufficiently to accept a charge again.

[PeterBC]

So I'm planning my first real deployment with a Victron Quattro and Smart Solar MPPT, and 24 Lishen 272AH cells in a 3P x 8S 24V config. I do worry about  passive cell balancing of a 816AH 'cell'  at 140ma being inadequate, but time will see. 

I've got my first test SBMS0 now, and am playing with some test 6AH LIFEPO4 36250 batteries to get familiar. Already destroyed my cheap bench DC 5A Supply  leads by being a bit careless and causing a brief short .. good lesson to respect my bigger 280AH Lishen cells when they eventually come. 

I'm happy with default SBMS0 high volt disconnect of 3.50V .. so what pack V should the Quattro be set to stop charging ?  A bit below the per cell max (=28.0V) ? Would 27.6 be reasonable ?  I fully understand the SBMS0 needs to stop if a cell goes above 3.5 even if pack voltage is within bounds (even at say 27?). Is it expected that it would be rare for the SBMS0 to cause a charge stop as long as the cells are reasonably balanced and the pack target is set appropriately ?  If regularly happening (i.e. a cell is over 3.5v but the pack below say 27.6V) should one take action ? Is adding a small cell in parallel to a weak cell to boost its AH capacity a solution to prevent it frequently limiting the pack ? 

For long inactive periods (eg winter) when the boat is mostly on shore power, how do I tell the Victron (or SBMS0) to limit charging to say 70% SOC ? The SOC limiting seems to say that it will go to 100% every day initially and only on any subsequent charging limit to say 70% ? 

[Bernd]

I´m currently running a SBMS0 with 24V 150Ah systems which will be updated to 32cells with each 280Ah....

the only possible answer to me is: SBMS0 needs to control charge and discharge.

According to dacian the balancing amperage should be fine even up to a 24V +/- 3000Ah battery.

The 3 parallel cells will balance themselves.

I run DSSR20 so I´m not sure about the voltage you need to set for charging, but i would expect 28V. The main topic is that the SBMS0 must be able to turn charging OFF at any time.

A single cell can be out of limits at any time and you need to ensure that the cells are always within their limits.

In case you have a defective cell, you should replace it and not add another cell to the pack to extend the capacity or equalize the defective cell.

Dacian will for sure let us know how to set the values in sbms to stay at 70% for long time absence.

[Dacian Todea]

PeterBC,

Hope to answer most of your questions.

The 816Ah battery is not a problem to be keep in balance by the SBMS0.

The default disconnect is 3.55V and the SBMS0 is the one that will stop the charging always not the Victron that should be set at around 28.8V so above the 28.4V (absolute max that SBMS0 will stop the charging).

Pack is only as good as the lowest capacity cell in that pack but since you get new cells they should all be fairly equal in capacity. Even if they are the best cells in the world they will not be perfectly equal so one cell (group of 3 cells in your case) will always be the first to get to fully charge but that is not a problem as SBMS0 will just stop the charging when any of the cells gets to 3.55V default limit since that pack will be fully charged.

When you will be connected to shore power the battery will charge to full 100% once a day and for the remaining of the day will no longer be charged. That will not be a problem for LiFePO4 so no need to worry about that. 

[PeterBC]

Dacian,

Thanks for your prompt reply. So I do notice the default advanced settings in my unit, Software V 5.0d differ a bit from the online Manual V0.99.  3.55V vrs 3.50V over voltage, and 2.80V vrs 2.85V under voltage, and delays of 6s and 3s respectively vrs 1s each in V0.99 manual page 31.  All good. Nice to see we are squeezing a bit more from the batteries. 

I'm happy to let the SBMS0 control the charging, just a different mind set to the Daly etc BMS's which have to interrupt possibly high loads or charging currents directly not via signaling the external controllers/loads.

I was under the impression that I'd be better to 'store' over winter my batteries with a lower state of charge than 100% or near 100% ? More like 70% ?  Perhaps a 'storage' mode detected when no significant loads for a relevant period (1week ?) to then only charge to a lower voltage like 3.4V or so, or to a specific SOC %  ?  In my case I like to keep the system on to allow remote monitoring of several aspects of my boat. There will be a base load of about 75W , but with a safety low voltage disconnect to shutdown the inverter and any other loads should I loose power etc for a sufficient period. 

On the remote aspect, I would like a little more control via the browser interface. I'd like to be able to change any settings I can from the device itself. Probably not needed I grant , but I'm trying to keep maximum flexibility.  

[Dacian Todea]

Peter,

Those photos in the user manual are old since I think version 2.0 of the software so that is a few years ago when there was no SBMS0. I just did not had the time to update the photos and they are good enough to explain the functionality.

That sort of BMS you mention is good for devices that do not need to charge and discharge at the same time like maybe a e-bike since you will not want the loads disconnected from battery if battery is fully charged.

75W is a significant base load and for example my house at night uses around 30W.  Thus if you need those 75W for monitoring equipment then that can not be considered storing. Battery will be fully charged once a day then not recharged for the remainder of the day unless it drops below a SOC limit that you can set. It is almost 2kWh/day energy consumption so fairly significant.

The 100% charge is needed to re calibrate the SOC and if you do not do that for long periods even if you have the current shunts properly calibrated it will drift as much as 1 or 2% per day and that can add up over a few weeks quite significantly.

Charging to 3.4V is completely unknown in therms of SOC. It may be that in some conditions 3.4V means just 20% SOC and in some other condition's it may mean 99% SOC and this is no exaggeration.

The internal battery impedance can vary hugely with temperature and together with the charge rate it makes absolutely impossible to guess SOC based on voltage unless you get to 3.55V that is about the minimum voltage that will guarantee 100% SOC even 3.5V will not guarantee full charge in some cases and it may be off by more than 30 or 40% SOC  

[PeterBC]

Daican,

All good succinct responses as usual.. thanks..

75W is a significant base load and for example my house at night uses around 30W.  Thus if you need those 75W for monitoring equipment then that can not be considered storing. Battery will be fully charged once a day then not recharged for the remainder of the day unless it drops below a SOC limit that you can set. It is almost 2kWh/day energy consumption so fairly significant.

 

I understand that 75W is significant. Almost 50W is the base load of the Quattro inverter being on. Unfortunately I havent had time to try and convert any monitoring equipment to a 24V DC source so rely on std 220V AC chargers for Internet, Wifi Access Points, POE switch , Camera, and Weather Stn.  I only leave on if I have a reliable shore power connection for winter. 

One option is to just use a small AC inverter for winter (with much lower min draw) and turnoff the Quattro. The quattro can of course just supply shore AC to some circuits and not draw that load from the battery, but I'm unsure now if it draws its minimum load from shore AC too, and not the battery ? (on reflection I think this is the case so when on shore power there is no load on the battery).  

I'm still inclined its considered harmful, or less optimal,  to keep the battery topped to 100% every day when on shore power ?

The 100% charge is needed to re calibrate the SOC and if you do not do that for long periods even if you have the current shunts properly calibrated it will drift as much as 1 or 2% per day and that can add up over a few weeks quite significantly.

Charging to 3.4V is completely unknown in therms of SOC. It may be that in some conditions 3.4V means just 20% SOC and in some other condition's it may mean 99% SOC and this is no exaggeration.

Agree it was silly to suggest a 3.4V target as surrogate for 70% or less SOC.  If there is a reasonable draw from the battery then it could be let run down to about 70% and then topped back to 100% etc  and that would be OK. But if in low power drain mode perhaps best not to charge back to 100%. I realise that may let SOC calculations drift.. a challenge to meet both objectives it seems.  

The internal battery impedance can vary hugely with temperature and together with the charge rate it makes absolutely impossible to guess SOC based on voltage unless you get to 3.55V that is about the minimum voltage that will guarantee 100% SOC even 3.5V will not guarantee full charge in some cases and it may be off by more than 30 or 40% SOC  

 

If possible a 'storage mode' could be that it allows the battery to slowly drop to 70%  or even actively helps it drop.. but I suppose 140ma max draw at 3.4V is only 0.5W , and 30% SOC drop of my 816AH battery would take 12,000 Hrs to do that !  50W would only take 120hrs which would be reasonable, but then how to stop that dropping is a challenge if no charge source available. I'm thinking ahead to when I may have to winter up north and perhaps dont have, or lose shore power.  I suspect I'll have to disconnect the battery in such a case, and make do with the engine 24V lead acids alone. Need to keep them charged then which currently is from the house 240V supply via  8A Victron Blue Smart chargers. Catch 22. Of course I could arrange DC -DC charging of those systems. 

Excuse my thinking aloud here .. 

 

[Dacian Todea]

There is storage condition in your case at almost 2kWh/ day. Even if you remove that to 1kWh/day that is still a normal working system.

The battery will sure drop with that 75W load even with a smaller load (not sure if sown to 70% until next day as it depends on your battery capacity but sure substantial drop).

Here is my battery capacity test and my battery is fully charged every day to 100% even multiple times https://groups.google.com/g/electrodacus/c/nKWp_6tjMNM/m/siO4LH--BwAJ

[PeterBC]

I dont believe the 2kwh will come from my battery, but rather the Shore AC supply. Keeping a LiFEPO4 at 100% for 6+ months, with almost no drops at all, is very different thing to cycling 70-100% each day I understand.  I'd like to have remote monitoring so I can view the status, cell volts etc, but still believe its more beneficial to get them down to 70% or so vrs 100% for that period. When I'm back on board it will cycle deeply most days I expect. 

[Dacian Todea]

Even if that is the case your battery will still be OK in fact better than cycling form 100% to 70% every day.  If you do not need the backup functionality you can just disconnect the battery and use just the grid.

Here is a cycle life for LiFePO4 so almost no cycle (5% DOD is minimum) means hundred of thousands of such cycles. And yes calendar aging that will be the only significant degradation will be just slightly higher than normal when cells are charged to 100% once a day but not as significant as you may think. LiFePO4 is not the same as other chemistry type  LiCoO2 or NMC where keeping the cells at 4.2V fully charge will have very significant effect on degradation.

LiFePO4 will be charged to 3.55V once a day but withing minutes the voltage will drop below 3.4V even without any load and will likely stay around there just under 3.4V for the rest of the day with load being maybe just the SBMS0 if you are sure inverter/charger will not use anything from the battery.

[PeterBC]

I still read several items that say its bad practice to leave LiFEPO4 'stored' at high SOC.   Perhaps this is from the NMC chemistries and others, which DEFINITELY should not be kept at high SOC for maximum life span and capacity retention.  This is more a boat thing, where many boats can spend 6-9 months of the year in storage. (mostly winter and cold weather). 

http://jes.ecsdl.org/content/163/9/A1872.full.pdf



This indicates a capacity loss of approx 5% for 9 months storage at 100%SOC .  Best results in fact at close to 0% SoC. ! 
I think we are all scared by experiences of leaving Lead Acids at low SoC and killing them quickly ! 
Capacity loss increases significantly at higher temperatures. The same article shows NCA and NMC cells much more markedly affected either by capacity loss of increased internal resistance due to storage at high SoC. 

I've seen other articles say this loss is reversible by holding them again at a 100% SoC for some 'time'.


I realise this is contradictory to the life cycle charts that show very high cycle lifes for small Depths of Discharge.  Its hard to argue that if 1% DoD from 100% to 99% had the maximum cycle life of over 200,000 cycles that storing it at 100% period, could damage ? Although I notice your chart stops at 5% DoD, and doesnt quote a 0% DoD ... Even 5% DoD at 1/4 C is substantial cycling vrs 0%  ? 

[Dacian Todea]

Sorry I do not have the time to read the entire paper to see how the test was conducted but their results are incorrect unless I'm missing something. My total capacity degradation for same A123 brand of LiFePo4 that they tested is 4.8% after 36 months and this is a period 4x longer than their test period.  While the battery was cycled the battery was above 70% for at least 80% of this time and between 40 and 70% for less than 20% of this period.

For calendar aging it is irrelevant if cell is cycled of just kept unused.  I saw quite a few other studies where they did not had anywhere close to this results so their test method was likely flawed or they may have an explication for this graph in the paper that I did not had time to read.

What they show there is a 3% degradation at optimum +25C storage temperature and 50% SOC over a period of just 9 months and this is totally inconsistent with any other test or manufacturer spec. Sony for example for their LiFePO4 provided a 0.5% to 1% as calendar aging over one year of storage at +25C and I think for A123 systems it was 1 to 2% per year (do not think they mentioned the storage SOC) probably the variation is related to that.

Not even the worse GBS LiFePO4 that had very bad performance and spec will not degrade anywhere close to that rate.

I had a 19.5Ah A123 cell stored for over 7 years at close to full SOC and capacity loss was around 5% mark