How to configure Victron BlueSmart charger to ensure battery cycling between 40% and 90%?

[Atom Tan]

I have an AmpereTime 12V 50Ah battery that I intend to use as a DIY UPS to power some 12V network gear. My plan is to keep it simple, and just have fused 12V leads from the battery feeding the loads and a Victron BlueSmart IP65 charger plugged into the wall to keep the battery charged. I will be several days travel away from this configuration once it is installed, so reliability and safety is important.

However, I'd like to ensure that the battery is cycled to about 40% SoC, and to only charge to 90%. I know that I am risking a situation where the battery is at a relatively low SoC when needed, but I'm accepting that risk.

To achieve this, my plan is to configure the Victron BlueSmart charger with the following settings:

Absorption Voltage: 13.3V
Float Voltage: 13.2V
Equalization: Disabled

I think that should charge to 90% or thereabouts.

What I'm less sure about is how to keep the charger from charging until the battery has dropped to about 40% SoC. if this were a Victron SCC, I'd set a re-bulk voltage offset of .20 volts from the 13.3V "fully" (90%) charged figure.. That would be 13.1V, which should be about 40% SoC.

The wall chargers do not have a re-bulk voltage offset parameter. I can set absorption voltage, float voltage, and storage voltage.

I considered just putting a smart plug between the charger and the wall outlet and just turning it on at a regular interval to charge, but that carries other risk (failure of the smart plug to turn on, overload of smart plug) I'm not willing to accept.

I'm open to a different charger if that solves the problem.

 

[soylentgreen]

I am confused - if this is a UPS, then the only time the battery would be in use would be when there is no 120VAC power. If there is no power, then the battery charger will be off.

Are you not using this as a UPS but in some other mode, such as cycling the battery daily?

 

[Atom Tan]

It's sort of a DIY UPS. There's no active load transfer if utility power fails. Rather, the loads, which are 12V DC, will be served full time from the battery, and the battery will be charged from the BlueSmart charger. I would like to allow the battery to regularly draw down somewhat before it is charged.

 

[soylentgreen]

It's not easy to determine the State of Charge (SOC) of a LFP battery based on voltage, especially when there are loads on it. You could probably do some experimentation - let the battery draw down a bit, then unhook all the loads, let the battery rest a while, then measure the resting voltage and figure out SOC. Repeat as needed, and eventually you may figure out the offset, such that "when I measure X voltage, with my loads, this is equivalent to a resting voltage of Y (with no loads), which equals a SOC of Z". Then you can use that X voltage as the re-bulk setting.

 

[Atom Tan]

That's helpful as I think about how to quantify SoC.

However, the problem is implementation. Victron's BlueSmart charger doesn't offer a re-bulk voltage offset configuration parameter. That parameter only seems to be present on their solar charge controllers.

 

[mikefitz]

13.3 volts is not the charge voltage for 90%, 13.3 is the resting voltage for aproxametly 90%. Using that voltage as a charge voltage will, get the battery in the region of 50 to 60 %.
Not charging a lithium battery over 3.4 volts per cell may cause cell inbalanced and reduce the effective capacity.
The Blue Smart charger will look after your battery using the default lithium setting. It really is smart and once the battery is charged, it drops to a low maintaining voltage, (13.2 from memory). Thus no need for re bulk as used in solar chargers. In the maintenence mode the Blue Smart charger checks for battery status and recharges if necessary.
The concept of not charging to a high SOC to prolong battery life is based on research data graphs produced many years ago, these can be found on the internet. This research was on early generation lithium ion cells and is not relevent to the current generation of LFP cells.
Charging to a high SOC and then holding the battery at a high voltage for long period may reduce service life, but Victron chargers prevent that.
Whilst it may be possible to use low charge volts with a correctly balanced DIY battery, a battery like AmperTime may/will have inbalanced cells. Unless the charge voltage is higher than the balance threshold, 3.4 to 3.45 typically, the battery cells will never balance.

Limit your discharge with a Victron battery protect,

BatteryProtect - Victron Energy

Disconnect the battery from non essential loads before it is completely discharged with Victron Energy's BatteryProtect. Find a dealer near you.

www.victronenergy.com

Set this to 13.0 volts as low SOC limit, perhaps 12.5 volts if high loading. Use for any loads except inverters.

Mike

 

[Atom Tan]

Thanks. That is exactly what I needed to know.

 

[Rocketman]

The other option - I think- is to install a BMV712. It has a relay.

You may be able to set the relay to open at 100% SOC, and then close at 40%. Hook the relay outputs to a relay that controls the charger. (I don’t remember if the IP65 can be controlled by relay inputs or if you will need to have a separate relay to control the 120v power to the charger.

Let the battery hit 100%, then turn off the charger. That way there is a little time for the cell balancing - but the battery spends very little time at 100%. You could even set the BMV712 to hit 100% at 14.2v - which is 3.55v per cell.

Note:I have NOT setup the above plan before - so verify it can be done before spending money.

Good Luck

 

[soylentgreen]

Can you say more about your goals? You mention "reliability" and "safety" - I think having your battery stay fully charged at 100% SOC would be the most reliable and safe way to go, as mentioned by others.

I have a Victron IP22 and it does a great job - once the battery is fully charged it drops to a very low power mode (drawing less than 5W from the wall) so it's very efficient.

I think if you try to add more parts and custom rules and features, you are likely to reduce reliabilty and safety.

 

[Atom Tan]

Sure. Regarding my goals:

Reliability
I mean here that the configuration will work, work consistently, and work without intervention in as many circumstances as possible.
Specifically, I want to ensure that:

1. The 12V loads will be served directly from the battery, with appropriate fuse protection.
2. As long as there is AC power, the battery is getting charged (modulo any SoC cycling I may choose to do).
3. If there is not AC power, the loads will continue to run until the battery is exhausted/goes below a set minimum threshold.
4. If the battery is exhausted/goes below a set minimum threshold, and the loads are dropped, that the battery will remain "serviceable", such that when AC power is restored, the battery will begin recharging and begin serving the loads again, without any intervention.

Safety
I mean here that this setup won't burn the place down or smoke the loads.

I do have a BMV712 on the AmpereTime battery. I can set a "discharge floor" on it, but that only seems to affect time remaining calculations, and has no effect on discharge past that threshold. I do think I'll need a BatteryProtect to handle that.

 

[ampsarus]

I would like to build same to power a 12V router and a couple IP cameras. One camera will have the battery and associated gear in close view.
If it was a FLIR camera (it won't be) that would be better.

 

[RCinFLA]

Cells will eventually get severely out of balance without a periodic absorb level full charge to about 14.2 vdc. Self-discharge variance in cells will drive the imbalance over time. You can get up to about 0.5% of cell capacity per month difference in cell self-discharge rate variance. Self-discharge rate on cells increases the warmer ambient temp the cells are at, along with greater variance in the amount of self-discharge.

The BMS will not balance a cell until it exceeds 3.4 vdc.

To do this you should have individual cell chargers set to 3.30 vdc. This will still not ensure perfect balance but will keep them in tighter SoC range spread.

LFP cathodes do not degrade significantly due to full charging like nickel based Li-ion cathode chemistries. The iron provides lattice vertical support in the cathode that does not exist in other li-ion cathode chemistries, preventing lattice collapse damage when most of the lithium leaves the cathode at full state of charge.

Weakest part of LFP cells is the negative electrode graphite expansion and contraction over SoC. The graphite expands about 11% at full SoC, being stuffed with lithium-ions that came from the cathode lithium depletion. It is the continued expansion and contraction over many charge-discharge cycles that degrades LFP cells.

 

[soylentgreen]

• 1. The 12V loads will be served directly from the battery, with appropriate fuse protection.
• 2. As long as there is AC power, the battery is getting charged (modulo any SoC cycling I may choose to do).
• 3. If there is not AC power, the loads will continue to run until the battery is exhausted/goes below a set minimum threshold.
• 4. If the battery is exhausted/goes below a set minimum threshold, and the loads are dropped, that the battery will remain "serviceable", such that when AC power is restored, the battery will begin recharging and begin serving the loads again, without any intervention.

1. easy, just wire the loads to the battery with a fuse. I personally like MRBF for small systems like this.
2. your victron IP65 will do that just fine.
3. All LFPs with BMS are designed to cut the loads at a safe voltage, leaving the battery in a very-low-but-not-dead state. If you don't trust your battery to do this, you could put a backup system in place, but I think it would be better to just buy a higher quality battery.
4. your victron IP65 will turn on when power returns, and the battery will start charging again.

Since you have a BMV 712 you might as well use it, but I don't think you really need it, and you don't need to do any fancy SOC cycling.

As mentioned by RCinFLA, the only worry is that you make sure your LFP gets charged high enough so the BMS can do balancing. If you look at the IP65 manual on page 10 it shows that the LFP battery setting has Absorption of 14.2, float at 13.5 and storage at 13.5. I think that's perfect.

What I'm not sure is whether this will happen if the battery sits at 100% SOC for long periods - e.g. you set up the system, it charges to 14.2 and balances, and then spends days or weeks at 13.5, never doing a 14.2 absorb. Would it eventually get out of balance? Hmm...

Turns out, Victron gear is really super:

Try this:
- open victron connect app and connect to your IP65
- enable Advanced settings
- click Advanced Battery settings
- battery preset: User Defined
- click Expert Mode

This will reveal some hidden features (which can also cause trouble, so be careful)

You could probably get what you want by enabling Repeated Absorption like this:

As shown here, this setting will automatically give the battery 1 hour at the Absorption voltage (14.2v) every month, but the rest of the time it will sit at the Storage voltage of 13.5v.



Now, I'm not sure you need to do this extra configuration, but it's certainly possible!

 

[Atom Tan]

I think I have zeroed in on my config for maintaining this AmpereTime 50Ah battery using a Victron BlueSmart IP65 based on everyone's comments above.

Feedback welcome:

Absorption: 14.3V
Float: Disabled 13.8V (I see that Victron has disabled this in their preset, but I can't seem to do the same.) (Figured it out)
Storage: 13.2V
Bulk time limit: 24 hours
Rebulk Current: Disabled
Absorption Duration: Fixed
Absorption Time: 1hr
Repeated Absorption: Every 7 days

 

[RCinFLA]

A lithium-ion battery only draws a small amount of self leakage current when a 'float' voltage from a charger is applied. Any float voltage below the LFP cell's potential of 3.43v is an acceptable float level. You also have to consider maintaining cell SoC balance for the maximum cell float voltage on a series stacked cell battery.

A lot of people have heard Li-Ion batteries don't like to be held near full charge state. This is true for nickel based cathode cells like used in cell phones and EV's, but LFP cathode is very rugged and very tolerant of being held at full charge.

There is a small electrolyte degradation rate increase at higher cell float voltage level. This is due to slightly accelerated parasitic side chemical reactions driven by higher cell voltage that breaks down electrolyte. Electrolyte breakdown rate greatly accelerates when cell voltage gets outside of electrolyte high stability range of 0.1v to 4.3v.

Electrolyte degradation also accelerates at higher ambient temps so keeping cells below 35 degs C is even more important than floating cell at high state of charge for LFP cells. Overcharging voltage causes bloating gas from the parasitic chemical reactions that breaks down electrolyte.

The gas bloating is not the worse side effect, just the most visible. The really bad side effect is the hydrocarbon tars that are left behind on the electrodes clogging up the electrode pores hindering lithium-ion migration.

 

[Atom Tan]

Testing and observations so far:

The battery has been serving load for 8+ hours with the charger config above, and with the BatteryProtect set to disconnect at 11.8V and reconnect at 12.8V. I chose this cutoff to maximize runtime while still avoiding a hard BMS low voltage cutoff.

The BMV shows the battery is currently at 13.03V. The BMV calls that a 35% SoC so the calibration is a bit off, as it also indicates that 33Ah have been drawn out of 50 available, which is closer to 33%. The BMV says that current draw is between 3.5A and 4.5A.

The BlueSmart IP65 went to Storage mode after fully charging the battery and has remained there. It is putting (when I just looked) 0.4A to 0.5A into the battery. I'd sort of expected it to re-bulk once the battery dropped below the 13.2V Storage figure, but it has not.


A bit later:

I think the current the BlueSmart is putting into the battery is increasing, as I am now seeing current readings of 0.7A to 1A from the BlueSmart, but I'm not watching it all the time so that may be normal. Yep, as I watch it more, it does seem to be creeping up, seeing peaks of 1.3A, and now 1.7A. The Victron manual says that re-bulk will occur "if the charge current is maintained at the maximum charge current for 4 seconds." Since this is a 15A charger, I guess I need to see 15A into the battery for more than 4 seconds before it'll re-bulk.

If this holds true, I will definitely be seeing a fairly complete SoC cycle with this configuration, and may adjust the settings to re-bulk a bit earlier, as getting below 40% SoC is not ideal for my use case.


Questions:

If I understand you correctly, @RCinFLA, you are recommending that I should re-enable float at a voltage below 13.7V, but relatively high, so probably 13.6V?

 

[Atom Tan]

BlueSmart is now putting 3.4A into the battery, so we're definitely on the curve up to re-bulk.

Since my load is only 3A to 4A, I wonder what will happen once the charger's current in equals the current out. Will it get to 15A in and re-bulk, or will the Storage mode just keep pumping a few amps in over the next 10 hours?

 

[RudyGreene]

Instead of a "smart" plug, why not use a mechanical timer outlet? I have some that have been running continuously for years without failure. If your loads are consistent, you can get it dialed in to hit your SOC levels fairly accurately.

 

[Atom Tan]

I'm down this rabbit hole now, and I'm going to keep going.

I hadn't seriously considered a mechanical timer outlet for three reasons. One, I didn't have a sense that they are as reliable as you suggest, but that may well be my own ignorance. Two, I was concerned about higher amperage loads through them, as I've only had experience with the ones made for lamps. Three, it felt like another added piece of complexity and thus possible point of failure.

I am aware that this third reason may seem faintly ridiculous given what I have set up.

 

[Atom Tan]

BlueSmart is now putting 3.4A into the battery, so we're definitely on the curve up to re-bulk.

Since my load is only 3A to 4A, I wonder what will happen once the charger's current in equals the current out. Will it get to 15A in and re-bulk, or will the Storage mode just keep pumping a few amps in over the next 10 hours?

Having had some coffee, I now suspect it'll never get to re-bulk and it will just continue to pump 3A to 4A into the battery until it gets to a nearly charged state, but don't understand what that state will be, nor whether the cells will balance in this situation. I suspect not.

I now believe this is exactly what the "Rebulk Current" configuration option is intended to address. I suspect that without that option configured, it will never re-enter bulk (constant current) or absorption (constant voltage) mode. The Repeated Absorption configuration option will, of course, run as configured.

It also seems like Storage mode isn't of any real utility to me in this use case where the battery will be actively used.