Why is bulk/absorption voltage used?

[hwy17]

I'm trying to get down to some charging basics here and interested in any resources or comments that explain the basis of 2 and 3 stage charging.

Why would we want to use a higher voltage than our end goal saturation voltage?

Why would we not just constant current / constant voltage all the way to the float voltage in one big single stage?

If I had to guess, it's something to do with speed, and pure float voltage charging would be slow, like top balancing is. So maybe bulk / boost / absorb is used at a higher voltage to sort of tip the charge into the battery faster than we could otherwise?

And is there really, actually, an intelligent way for a charger to sense when it should go to float? My Schneider 100 600 MPPT charge controller has a timer setting to end bulk. I do not like to see a timer.

What will happen if I set my MPPT to charge at 54v for all stages? Would I find I'm missing out on production because I can't get enough amps into the battery?

 

[Hedges]

I don't know the chemistry of it, but the technology for lead-acid has been around and in development for over 100 years.
What the good manufacturers say to do for their batteries is best. Each type and various brands has its own unique chemistry.

Page 19

https://sunxtender.com/pdfs/Sun_Xtender_Battery_Technical_Manual.pdf

There are various charge controllers and inverter-chargers which implement parameters like these.

Timer to end bulk? I expect bulk to end based on voltage. Then a timer for absorption. Some manufacturers spec time according to DoD (SunXtender does, using average DoD which for us would be a guess).

"Why would we not just constant current / constant voltage all the way to the float voltage in one big single stage?"

I do see that for standby (UPS) use, other parameters for frequent cycling. So probably speed of charging.

 

[hwy17]

Timer to end bulk? I expect bulk to end based on voltage. Then a timer for absorption.

I think I have some terms mixed up on stage 1 and 2 so yes, timer to end absorption.

 

[Steve_S]

Absorb is a Higher Voltage @ Constant Current which tops off the cells in a battery until the resistance drops down to the EdAmps/TailCurrent value at which time float takes over with Constant Voltage / Variable Current allowing the cells the gentle finishing they need. Absorb is typically only 0.2 to 0.5 Volts higher than Float is set to. Absorb is usually set with a Timer, mine is set to 60 Minutes but more often than not EndAmps is reached beforehand and that kicks absorb off. This is the proper operation for these values.

LFP, simply put requires only CC/CV charging. Bulk/Absorb (not everyone uses the same terms) is Constant Current Constant Voltage while Float is Constant Voltage Variable Current.

Float will also service the Inverter demand, so if the inverter all of a sudden demands 100A and solar can supply it, it will use that first before the batteries. Float will ramp up to provide it. IF the 100A demand can't be met by float, the balance will be drawn for the batteries and once the demand is gone, float will increase it's amperage to top off the batteries as best as it can with teh solar power that is available.

There are many charging modes depending on the chemistry in question. My inverter (Samlex) has 9 modes, up to 5 stage charging, it can be confusing & overwhelming in some cases.

The KICKER: Terminologies between different companies vary (no global set standard as such), this makes it confusing when looking at say Victron vs EPEver or Midnite solar for example. Bulk & Absorb get mixed up a LOT... everyone pretty much agrees on FLOAT. Not ALL SCC's or even Inverter/Chargers even have an EndAmps/TailCurrent value that can be set. (mostly the "Value" Brands).

 

[hwy17]

Thank you for the answers. I will keep re reading and chewing on them to see if I'm missing any part of the logic.

Here's a picture that illustrates my question probably better than my words. I am wondering why we do A instead of just B.

To me B is more satisfying, because the charger is never trying to raise voltage higher than I want the bank to rest at. A seems to invoke a lot of hazards around ensuring the absorption stage does not continue too long.

So not to beat the horse, but is the basic answer just that B would be too slow? My favorite reality would be that A is an outdated holdover from lead acid and yes, we can just do B. But it does not seem like B is the standard recommendation from LFP battery manufacturers and pre programmed LFP chargers do A not B. (consider these two voltage graphs that both end at 54v)

 

[krby]

Bulk Is for faster charging. On any reasonable charger, "Bulk" really shouldn't be "constant voltage NO MATTER HOW MUCH current". It's really "current no higher than X with a voltage no higher than Y". This is usually called constant current (CC)
X should be no more than the max charging amps spec'd for your battery bank. Y is typically that max bulk voltage spec'd for your battery bank.

You can absolutely charge the battery at a lower voltage, it will just take longer overall. Old school car battery trickle chargers do this.

 

[hwy17]

Also, regarding my picture above, how can I make my MPPT do B? Should I set it to 2 or 3 stage? I'm thinking it's 3 stage with all voltages set to 54v.

 

[Hedges]

B is just A with same voltage for absorption as float. As you say.

I'm thinking it's 3 stage with all voltages set to 54v.

Lithium, if you hold at 3.65V forever, it will gradually charge to full, maybe over-full beyond 100%.
Charged to a lower voltage, current will taper down and might take forever to reach full charge. Which isn't bad, since it doesn't have to reach 100% to avoid deterioration (as lead-acid does.)

Most people here building batteries did parallel top-balance to 3.65V, stopping at around 0.01C charge rate. Then used the pack with maybe 3.5V charge voltage.
Voltage needs be held in a certain range to allow BMS to balance.
I'm not sure, but that voltage, or the SoC the battery would eventually reach, could be subject to faster deterioration especially when hot.

Further complicating things, loads or clouds could cause voltage to temporarily drop, causing SCC to shift to a different phase of charging.

Some "absorption" followed by "float" might give your BMS a chance to balance, then hold battery in a good state.

 

[hwy17]

I was considering a really tight upper voltage range. Like balancing at 3.35 or 3.4 with absorption and float at 3.4 or 3.45.

I might just discover why people don't do what I'm thinking of. I'm still lacking any hands on experience with running a daily cycled battery, but after carefully building my battery I won't like the idea of throwing higher voltages at it, until I find out that I have to.

 

[Steve_S]

Keep in mind that all battery chemistries have TWO Voltage ranges, even FLA. The allowable voltages where no harm/damage is done if within that voltage range and Working voltage range where the deliverable power comes from. Pushing batteries to the top of the "allowable" range while not causing harm directly, it is stressing the chemistry as that is NOT the normal working range... in a sense you are prematurely aging them.

Overthinking & trying to do it more than what is the norm will lead to Blue Air Generation & Hair Pulling let alone the $$$ factor when it goes poofdah. There are a few victims here that learned that the costly way.

Quick handy graphic for you. Yes you've seen it before.

 

[hwy17]

But aren't I being safer and more gentle to my battery by staying strictly within 3.45v with no charging stages above that at all?

I realize I am greatly increasing the risk of hair pulling.

Or are you saying I could do this at even lower, 3.3-3.4?

It is any charging stage in the 3.5v+ range that scares me and I'm wondering if I can stay out of there entirely, in all stages.

 

[SeaGal]

I don't see any gain in trying to apply techniques designed for LA to LFP cells.

At the end of the day we want to:-
- charge up as fast as possible to chosen voltage
- not age cells by too fast charge, esp when nearly full and/or colder
- give time for balancer to do something useful
- charge at high enough voltage so balancer can operate in upper knee, where voltage becomes a measure of fullness

My control software does this by telling the inverter the max voltage and max current to charge at. Then reduces charge current in steps as SOC goes above 90% towards 100%, finishing with trickle charge till highest cell gets to my pre-set value of 3.45V.

For me charging to 3.45V with balancing starting above 3.4V achieves those requirements well.

 

[hwy17]

I don't see any gain in trying to apply techniques designed for LA to LFP cells.

At the end of the day we want to:-
- charge up as fast as possible to chosen voltage
- not age cells by too fast charge, esp when nearly full and/or colder
- give time for balancer to do something useful
- charge at high enough voltage so balancer can operate in upper knee, where voltage becomes a measure of fullness

My control software does this by telling the inverter the max voltage and max current to charge at. Then reduces charge current in steps as SOC goes above 90% towards 100%, finishing with trickle charge till highest cell gets to my pre-set value of 3.45V.

For me charging to 3.45V with balancing starting above 3.4V achieves those requirements well.

Nice. So do any of your charge stages target voltages above 3.45? (whether they actually reach them or not)

 

[SeaGal]

IIRC my software tells the inverter it can charge at max voltage of 56V (that value is fixed).

But, in practice, that voltage would never be reached because of the current limitations.

Obviously to get current to flow from inverter to battery the inverter will have to be at V=IR volts higher than the pack, where I is the charge current and R is the resistance of the cables between the two.

 

[fat_old_sun]

What will happen if I set my MPPT to charge at 54v for all stages? Would I find I'm missing out on production because I can't get enough amps into the battery?

In principle, this would work. However, it'll take a very long time (100 hrs or more) to complete the process since you're approaching the target voltage asymptotically. Despite widely used, voltage based charging is not the correct approach. However, it is easy to implement. Since we're tying to reverse a chemical reaction with electrons being one of the participants, the correct approach is one based the charge returned. If you took out X Coulomb during discharge, provide X+y during the charging process to fully reverse the reaction; the amount y is a correction for columbic inefficiency.

 

[Steve_S]

I charge to 3.450VPC, float at 3.438? and they settle nicely at 3.390-3.400 pending on whats going on. All 6 Packs too no less.
My JK's start balance @ 3.200 and keep the edges off as they happen so by the time I hit float the differential is typically no more than 0.20 and that does not take much to level up.

A note on Active Balancing.
Active does NOT waste power, it only moves power from High Cells to Low Cells "WHEN" the cell differential exceeds the set limit and WHEN it is above the trigger voltage. Otherwise it sits and does nothing. The best starting point is at the LFP Nominal Voltage of 3.200v.

As an FYI, I am the same Steve_S that edited & wrote the English JKBMS manuals.

 

[hwy17]

I won't have an active balancer to start, only 350ma of passive balancing (Orion).

 

[featherlite]

I ... wrote the English JKBMS manuals.

Kudos!
Are you fluent in Chinese?

 

[hwy17]

Ok so I'm trying to remain open minded and humble about my still relatively lack of experience, but I've done two full charges on my new DIY pack and I'm still skeptical about multi stage charging for Lifepo4.

When I charge with a single stage CC/CV at 55v I get full CC mode right up until 90%+ charge and only in the final hour or so does the charger go into CV and the current tapers.

It seems like adding a higher target voltage stage in the middle would not actually make a difference to the charge curve, even if the middle stage was 58.6v or whatever is commonly recommended, the charger would still be acting in CC mode and the battery would be happily taking the full amperage at somewhere below the final float voltage.

So I am still entertaining this idea that you can just single stage charge Lifepo4 with a single target voltage, the final float voltage. If there is any room where a higher voltage charge would get energy into the battery faster, without actually overshooting the float in the end, it seems like a very narrow window, maybe 30 minutes and very little to gain in that window.

I will remain open minded. But I include open mindedness to the idea that this entire multistage charging concept is just a confused carryover from lead acid. If the charge voltage is not actually exceeding the float voltage at any point in the charge curve, then these mid stage voltage targets are immaterial, they're theoretical allowances for the charger that are not affecting the reality of the charge curve.

 

[hwy17]

Here's another illustration of what I'm trying to say. Don't take this for a detailed analysis, or like I'm trying to draw the literal shape of a charge curve. This is just to illustrate the idea of two different charge profiles, multi stage and single stage, but both are achieving the same actual charge curve in reality. I am just wondering if this is the reality. For the multistage charge profile to actually do anything the actual charge curve in red would have to at some point rise into the first stage above the final float, otherwise bulk/absorb voltage is a pointless parameter.