LiFePO4 charging - the overcharging question

[booster]

What you are describing is the Achilles heal of a lot of charging systems. If the monitor/control sees the systems loads, all the accuracy of return amp charging is out the window.

The first question to ask is if the Victron is reading the amp in/out of the battery, with a shunt of inductive, right at the battery, no loads seen. The way you describe it, with the loads showing on the meter it may not be that way.

The goal in return amp charging is to look only at the amps to the batteries and ignore everything else, so if you turn on a load and it shows on the meter, that is an immediate question as to what is going on.

I think I would call Victron and ask them how it works for one thing. Also if it isn't on a battery shunt, it is not going to work correctly for sure. If you see the loads changing when you go to shut off, you are definitely not triggering just off of battery charging amps like you want to.

Your observations are very relevant, and definitely worth checking out!

[ptourin]

They provide a shunt, and it's on the negative lead to the battery. Again, I'm not sure I'm seeing things right, but if it's in series with the only neg lead, it's got to be showing both inbound and outbound current, no? How else is it going to calculate %SOC? For example, if I have a lot of loads turned on, perhaps the monitor will read -15A, if I flip the charger on I'll then see +30A, and if I then turn all the loads off I'll see +45A. So it appears that I'm seeing the difference between inbound and outbound.

I didn't believe that such systems could distinguish between the two - I didn't believe the monitor could say, "now I'm seeing a 15A load, and I'm also seeing 45A of charging current" - I thought it'd only say, "I see a total of +30A". I hope I'm wrong about this, because it'd make more sense. I just don't see how the monitor can see charging and load current and calculate SOC, and at the same time watch the charging current separately and say, "I just saw it drop below 20A and the voltage is above 13.4, so I'm declaring that it's 100% SOC time".

I can call Victron North America support - there are 2 guys there that are very sharp and who have a good deal of lith Phosphate experience, since Victron sells batteries.

[booster]

OK, I think I see what you are explaining now with the load and charge.

You have a shunt, it is in the right place if it is in the negative battery cable right after the battery (no other connections between the shunt and battery), so the monitor should be able to give you accurate SOC (as accurate as the SOC readings can be calculated) and charging amp readings.

Maybe the best thing to do is look at it like a water piping. The battery is a tank full of water, the charger is a well with pump, and the loads are faucets. It would look like a tee setup with the faucets on one end, the well on the other, and the battery in between on the tee leg (your shunt would be in the leg by the tank. If you turn on a faucet with the well off, water runs out of the tank and would run the meter in a minus direction. If the faucets are off and the well is on, water would get pumped into the tank (if it wasn't full) and the meter would read positive. If you have the well on with water going to the tank and then open a faucet the well will then pump water out the faucet, AND IT WILL CONTINUE TO FILL THE TANK AT THE SAME TIME. The tank will not even know the faucet is open if the well can supply enough flow, so the meter will still read positive. The meter would only go negative if the faucets were using more water than the well could supply, then the tank would be supplying the amount the well couldn't do.

Same thing happens with the battery and charger. If you have just loads on the battery with no charger running, you will have negative on you meter. If the battery is being charged by a charging source, you will read positive. If you turn on a load, while the charger is running, the battery will not know the load is there if the charger is big enough to run the load and charge the battery at the same time so the meter reading will not change with the load on or off. If the load is big enough that the charger can't supply it and the charging current at the same time, then the charging current will drop or go negative as the battery will be supplying some of the load current.

I think this is what you are describing, so your meter should be very capable of controlling things. From the numbers you gave, you must have a 45 amp charger. It would have been supplying 30 to the battery and 15 to the loads. It would put 45 to the battery with the loads off. With charger off, loads on you were reading a minus 15 amps for the loads with power coming from battery.

[ptourin]

Yes, you've got it exactly. The charger is actually a 60A charger, but in practice, if the batteries are down around 50% or so and I turn the charger on, I see about 45A max on the monitor. Since it's a CCCV charger, I assume that even if the batteries were down to 20%, I'd still see that 45A when I turned the charger on.

Your description makes perfect sense - I've just never seen one part of it in practice. I usually turn on the charger when the battery is around 40-50%, and I usually have small loads. So I see the monitor move from perhaps -3A to perhaps +42A. That's as expected.

What I haven't done is watched the monitor while the charger is on, and added lots of DC load. And I've never thought about this before - are you saying that as I add load, at first the charger will supply it as well as the charging current, and the monitor reading will stay the same - charger current into the batteries remains the same and there's no outward current? But as the load gets too big for the charger to handle, the battery starts supplying current and the monitor reading then goes less positive, or goes to negative?

If that's true, what actually happens? Does the charger supply power at first because its voltage is higher than the battery voltage? - and as it gets to its limit, does its voltage drop until it's down to the battery voltage, so the battery then starts supplying it? Trying to understand how the system works dynamically....

[booster]

I think you are well on the way to understanding how it all works!

With the charger putting amps to the battery, it will be at full charge voltage and whatever amps (as long as the amps are less than the charger capacity). As you add load, the charger just puts out more amps to cover the increase, and it will hold full charging voltage until it hits it's maximum amps output. At that point the voltage will start to drop rapidly, because the charger can't keep up, and when that voltage gets to the voltage of the battery, the battery will start to give up amps to the loads to help the charger (probably about 12.8ish volts), and your monitor current will go negative because the battery is being discharged, even though the charger is one.

Hope this makes sense with what you are seeing.

[ptourin]

So this moves into the territory of how you configure a monitor. In order to read 100%, the Victron expects first of all that the voltage climbs over the Charged Voltage setting - I've set that at 13.4 as it seems to be the fully charged rest voltage for the batteries - the charger will certainly pull it over that voltage - it'll try to pull the battery to 14.6v.

Second, the Tail current has to drop below the set point and stay there for a time setpoint. Default is 4% for 3 minutes. The Victron guys suggested that I could set this up to 10% - I think their reasoning is to set it conservatively so as to be gentle on the battery, unless I have reason to try to push it to the ULTIMATE CHARGED STATE <g>... But it seems funny - 20A for a 200Ah battery bank seems very high, considering that the charger usually seems to be giving the batteries ~45A max. I'm not sure why they'd want such a high value.

But I haven't yet charged long enough to find out how the charger current behaves as we get near full charge - I only got the RV home last week, so I now have a few weeks to play with it before it goes into winter storage - we'll see how much I can mess with it in the meantime.

The 13.4 setting seems to make sense - once the charger is off, I want the batteries to keep reading 100% until I drop below their fully charged rest voltage. So it remains to decide about the tail current and charged detection time, as it's that value that'll trigger the relay and turn charging off. When to turn charging back on (or trigger the audible alarm) seems much less critical since this hasn't been a very deep cycle use pattern - at least so far.

Also trying to decide whether I have the nerve to charge them, disconnect them and leave them in the RV all winter - the specs say that they should be quite happy and barely discharge over 4 months, as long as they're disconnected so the monitor current doesn't make them creep down...

[booster]

Normally the voltage you use for the "get to" voltage is the charger absorption voltage because until the charger gets to that voltage it is running at full capacity output and has reduced voltage. In your case I think that is 14.6 volts on PD CC/CV charger. Once it gets there, they usually hold for a time regardless of current because non lithium batteries like the hold time to some extent. If you have the tail current set right, the hold time can be very short, and probably should be for lithium. You do have to get to the set voltage however to be able to get a consistent tail current to trigger off of. If the voltage is different, the tail current will change.

Since the lithium like to stopped slightly less than full per most folks, you may need a fairly high tail current, and short hold time at 14.6v. You might want to call the manufacturer of the batteries and ask them what % of capacity they want for tail current at whatever SOC you want to stop, maybe 95%. They should be able to give you that, but be sure to tell them at 12v system voltage.

The cool thing is that if you use the tail current to stop at 95% full, the monitor will think that it is 100% full based on the voltage and tail current, and show that you are done charging. If you reduce the battery bank size in the monitor settings to the same 95% of what you have it will be accurate for the amount of ah you have a available in calculating SOC, so very useful.

[ptourin]

That all makes sense, and I ran a test last night that fit well with what you wrote. I now have the Charged Voltage setpoint at 13.4 and the Tail Current at 4%. As I understand the Victron monitor, it reads 100% SOC when both setpoints are reached and held for a time period (set now to 4 minutes). It will always reach the voltage setpoint first, as the charger voltage ramps up to 14.6. Then the current will drop to the 4% setpoint and the monitor will say "you're there - 100%". I believe the Victron tech guys suggested 10%, agreeing with your comments above that it may be good not to try to pull every last Ah out of the lith batteries.

When I got the RV back to VT a few days ago I flipped the charger on and charged the batteries until the monitor read 100% - then I turned the charger off and also turned the DC disconnect switch off - so the only load on the batteries was the monitor, which is a tiny draw. Here's my test from last night:

Initial check: rest voltage 13.6 (I assume that's a bit high because the surface charge hadn't completely gone away yet) - monitor still read 100% of course

I then turned on the DC switch and saw around -0.50A on the monitor - my parasitic load. Then I turned the charger on. In less than 10 minutes the voltage climbed and the current went way down, as expected:

initial reading +22A @ 13.6V
after only 2 min. +5A @ 14.48V
another 2 min. +2A @ 14.51V
another 2 min. +1A @ 14.53V

I stopped here, as this is exactly what I'd expect when the batteries are near the big elbow in the charging graph up near 100% SOC - voltage jumps up fast, current drops towards zero.

I was running out of time (translate: late evening, falling asleep <g>...), so I didn't do a good load test - I'll do this again in a few days. But I did switch on a bunch of lights, and I didn't see a current change on the monitor. I assume that if I measured charger output with my clamp meter I'd see the increase since the charger would be supplying the load current. Next time I'll load the system up and see if I can max the charger and make the monitor current change as the batteries take over some of the load. That would absolutely confirm what you described.

So - back to how to manage the charger. It looks almost sure that you're right about the monitor, and that I can use it to control the charger. Its config params are very flexible, and I can trigger the relay off the SOC or other params.

I'd think that a few relays would give me a very simple system. I've got the internal monitor relay (tiny control relay, 60V max @ 1A max). I've got a 40A power relay with a 12vdc coil - I'm thinking of using it to switch the AC power to the charger. I've got the generator transfer switch. I can use a small AC coil relay to break the monitor relay circult unless there's AC present - so that if the monitor relay shuts to request recharging but there's no AC present, the power relay current stays open and no current is drawn other than the monitor relay. When there's a request and there is AC available, the monitor relay contacts close the power relay, and its contacts close the AC circuit to the charger.

it would be nice to wire it so if the monitor calls for recharging and there's no shore power, it auto-starts the generator. I don't know how to do that - my Onan LP generator doesn't now auto-start and I don't know how to get relay closure for the generator. But even controlling the charger from the monitor is much more than I've got now.

Again, thanks for the thoughts - it's nice to be able to talk about it, then go out and test it. Having a good time - I hope the snow holds off so I can keep playing for awhile <g>...

[eastbay]

Excellent discussion- and I've learned a lot, thanks!

For my application I am leaning towards having one Lifepo4 100 ah battery that will travel between a Class B, boat, and off grid cabin. For charging it will usually be either driving down the road in the RV or from a Honda 2000i, rarely plugged in to shore power.

My past practice has been sort of the opposite of this thread in terms of battery care and feeding- start the generator and let it run a full tank of gas through a smart charger (or what passed for one then) if I'm going ashore, or less than that if not, but only putting into the battery whatever random amounts required to keep a reasonable amount of juice in the battery.

Needless to say this was quite hard on traditional batteries, I have no experience with AGM's; but it seems to me that a Lifepo4 would be perfect for this kind of abuse- am I correct? Would they be comfortable living in a state between 20% and 90% charge? If the answer to this is yes, than why all the worry? It sounds to me like spending the $$ for Lifepo4 means that you don't have to worry about coddling your battery, or am I wrong and will ruin a very expensive item?

An anticipated big plus for me also is lugging the thing from the boat to the car to the cabin- 30 pounds instead of 65! I should also point out that my boat is a small trimaran, very sensitive to weight, so the weight savings means a lot to me there as well.

[wincrasher]

Eastbay - yes, I think you understand things well enough. You just need a charger with enough voltage to charge the battery - LiFePo4 drop-ins need 14.6 volt constant charge.

But yes, you don't need to top them off. They can be in any state of charge just fine. You can't charge them below freezing, which doesn't sound like a problem you need to worry about.

If you are just running 12v items (not running an inverter), then you probably don't need much in the way of monitoring, other than a voltage meter.

[ptourin]

The answer is - yes and no <g>...
Pros: They don't mind sitting with partial charge, whether 20% or 90%. They can be discharged or charged at a high rate without damaging them - and discharging them way down to the 20% area doesn't damage their chemistry and shorten their life to the degree that it happens with flooded cell batteries. They're light. They maintain a higher voltage until they're almost totally discharged. They will theoretically take many more charge/discharge cycles over their lifetime than flooded cell batteries.

Cons: If you ever discharge them totally, even once, they're gone. There's some question whether they can be damaged by being left on a charger once they're fully charged - wish I knew the answer to that. They can be discharged below freezing (at reduced capacity), but they will be damaged if you charge them below freezing (battery temp, not ambient temp). It's hard to understand just how they should be charged in order to protect them and get best use out of them. Different brands have slightly different recommendations for charging, and very different types of internal battery management circuitry. They're very expensive!

Generally, you charge them with a CCCV (constant voltage constant current) charger. When they're low and you start to charge them, the charger feeds the battery at its rated current and the voltage slowly climbs to the rated voltage - so my 60 amp charger hits the battery with somewhere near 60 amps until the charger voltage rises to 14.6 volts, its maximum charging voltage. At that point the battery is fairly well charged. Then the charger voltage stays at 14.6 and as the battery gets closer and closer to 14.6, the charger current drops until it gets down near or at zero. By the time the current is down to maybe 10% of the rated battery capacity (10A in your example), you're nearly fully charged. Look up Progressive Dynamics PD9100 series chargers - those are their lithium chargers.

So if none of the cons are show stoppers for you, the pros make the batteries very desirable if you can afford them.

[Boxster1971]

Just read the latest posts about lithium battery issues on the sailboat Cruisers forum.
LiFePO4 Batteries: Discussion Thread for Those Using Them as House Banks - Page 313 - Cruisers & Sailing Forums

A major marine supplier, OceanPlanet Energy Inc., has switched to Lithionics batteries after Genasun dropped their lithium product due to QC problems with CALB cells.
OPE-Li3 Marine Lithium Battery Systems - Bruce Schwab Energy Systems

If you are interested in lithium batteries for an application similar to RVs the sailboat cruiser community is a good place to get information.

[ptourin]

I finally had some time to play this evening. I started with around 95% SOC - batteries had been disconnected for a week. I started loading them up, ended up with about 21A load - so around 1/10C. I watched for several hours and saw the system go down about 40+ Ah - final reading 78% SOC.

What surprised me is that when I took my final readings while loaded up, the voltage read 12.96. As soon as I took the big loads off, it rose to 13.14. I'm not surprised to see the voltage drop under load (though this wasn't such a big load!) and rise again when most of the load is gone, but I was surprised that it dropped under 13V.

You guys who have played with them a lot more than I have - is this what you see as you load them up?

[ptourin]

A bit of additional stuff... This morning I had to make 2 short drives of about 10 minutes each. When I started, I was still reading 78% SOC of course, and the voltage had recovered to 13.28v. By the time I got home I was back up to reading 100% SOC (I have the Victron set to read 100% at 13.4v). The alternator lead hits the battery before the DC shutoff switch - it charged with the switch off, so I may want a separate shutoff for the alternator lead. The max charge current I saw was 60A - that's more than the ~40A I saw the last time I made a quick check and I like it a lot better!

The big take-away that I didn't expect is that even though the overall voltage sag with discharge is less than with flooded cells, there's still a lot of voltage sag and recovery as you load up the batteries and unload them again.

[markopolo]

Awesome info. Keep it coming - I'm learning a lot

[booster]

Originally Posted by ptourin

A bit of additional stuff... This morning I had to make 2 short drives of about 10 minutes each. When I started, I was still reading 78% SOC of course, and the voltage had recovered to 13.28v. By the time I got home I was back up to reading 100% SOC (I have the Victron set to read 100% at 13.4v). The alternator lead hits the battery before the DC shutoff switch - it charged with the switch off, so I may want a separate shutoff for the alternator lead. The max charge current I saw was 60A - that's more than the ~40A I saw the last time I made a quick check and I like it a lot better!

The big take-away that I didn't expect is that even though the overall voltage sag with discharge is less than with flooded cells, there's still a lot of voltage sag and recovery as you load up the batteries and unload them again.

I think most our systems charge fo the alternator with main disconnect off, and our new setup will also. That is why you probably have seen a lot of discussion here about manual separator switches so we can turn off the engine charging when the batteries are full. If you don't want continuous charge voltage on your lithiums, you would need to do the same thing.

Good observations!

[ptourin]

Any suggestions for a switch? I've only seen 60A max on that line, but I suppose the switch contacts should be able to handle the full alternator output. All those RV-type rotating switches seem extremely expensive!

[booster]

Originally Posted by ptourin

Any suggestions for a switch? I've only seen 60A max on that line, but I suppose the switch contacts should be able to handle the full alternator output. All those RV-type rotating switches seem extremely expensive!

The rotating switches are in the $20 range, I think. The problem with them is you need to have the big wires going to them, so you can't have them on the dash conveniently. The remote actuated, 300 amp battery switches are really expensive, in the $150 range for a Blue Sea, but they draw no power except when switching, and have nice remote switch with them. It is what we used because we wanted the control at the dash.

[ptourin]

I'm not so sure I care about dash actuation. It seems to me that once you start watching your monitor, you have a pretty good idea about your SOC. So I'd think that the alternator switch would remain off unless you were going to drive somewhere and wanted to charge. For example, I can't envision driving down the road, suddenly deciding that my batteries were down, hitting the switch and - BAM - slamming 60A of charging over to the batteries as I drove <g>...

I'm doing the Progressive Dynamics charger manually now, by throwing the breaker to it - if only because I haven't decided on something better yet. I could leave it that way and be entirely manual. But I do want to wire the battery monitor relay to kick the charger on/off automatically when I'm on shore power.

But maybe that's also an argument for not using a switch on the alternator line - it could be to set up so that when the monitor requests charging current, the system connects the AC feed to the charger and also closes the alternator relay. If I understand correctly, I wouldn't need any kind of fancy logic to choose charger or alternator - after all, if I leave the charger breaker on and head down the road tomorrow, that's how the system works now - if there's shore power, the charger is on, and if the engine is running, the alternator is putting out current - and both can be happening at once.

Am I being simplistic here again?

[booster]

Originally Posted by ptourin

I'm not so sure I care about dash actuation. It seems to me that once you start watching your monitor, you have a pretty good idea about your SOC. So I'd think that the alternator switch would remain off unless you were going to drive somewhere and wanted to charge. For example, I can't envision driving down the road, suddenly deciding that my batteries were down, hitting the switch and - BAM - slamming 60A of charging over to the batteries as I drove <g>...

I'm doing the Progressive Dynamics charger manually now, by throwing the breaker to it - if only because I haven't decided on something better yet. I could leave it that way and be entirely manual. But I do want to wire the battery monitor relay to kick the charger on/off automatically when I'm on shore power.

But maybe that's also an argument for not using a switch on the alternator line - it could be to set up so that when the monitor requests charging current, the system connects the AC feed to the charger and also closes the alternator relay. If I understand correctly, I wouldn't need any kind of fancy logic to choose charger or alternator - after all, if I leave the charger breaker on and head down the road tomorrow, that's how the system works now - if there's shore power, the charger is on, and if the engine is running, the alternator is putting out current - and both can be happening at once.

Am I being simplistic here again?

You are thinking just right, it appears. If you trigger off your monitor, like you will with the shore charger, with a relay for the alternator, you will have that covered too, and not have to do the manual turn on and off. Your monitor is different than the rest of us have, as we don't have access to any output contacts.