12V Battery - Flooded Lead Acid vs. AGM vs. Lithium

Yeah, not exactly about the Leaf traction battery, but I thought this could still fit into the "battery & charging" category since it's about a battery that is vital to the Leaf operation.

Hoping this topic can be used as a way to share our experiences with changing out the OEM 12V battery in your Leaf with anything other than what the dealership wants to give you.

To start, I replaced my OEM 12V battery in my 2013 Leaf back in 2015. Since then, I've moved it from my old 2013 before trade-in to my newer 2020 Leaf where it continues to function today.

Where I live, the climate is warmer and so far, the temperature never gets below 0F (-17C) and even if it is in the single digits, that might only last a day or two. I do get a lot of events where the weekly temperature is below 32F (0C) for long periods of time, but nowhere near what someone living in Canada, Alaska, northern US get during the winter as my temperatures to them would be a warm spring day. :lol:

Having stated that, I do feel like the Lithium 12V battery gets a bad reputation as being failure prone when temperatures get below freezing. Not so much about the members of this forum, but basically everywhere you google about these batteries, you read about how if you try to charge them below 32F (0C) they can catch fire or damage the battery. I am aware of the technical limitations of the Lithium 12V (LiFePO4) when it comes to temperature extremes. When the temperature of the battery is below freezing, the amount of current it can absorb for charging is greatly reduced. Once the battery drops to -40F/C it will freeze to the point of taking internal damage.

I would never recommend someone who lives in any climate that will dip below -40F/C for any amount of time to use a LFP battery just for that reason, but some areas that experience cold temperatures that never get below 0F (-17C) or even some extremes of -20F (-28C) could still function in the Leaf well enough to get it started and with the somewhat, bizarre charging algorithm of the Leaf, would not suffer any charge problems due to the much higher initial voltage of the LFP.

In another topic, I made the claim that my own small, 20AH LFP should be able to turn on a Leaf hundreds of times before being depleted. The reason being, if the battery was too cold to charge, it should still retain enough capacity to start your Leaf over and over until the climate warms back up and it is able to absorb enough energy to reach full charge levels again. It does sounds like a skeptical claim and I'll be to first to admit I'm wrong if it's really not possible, so I'm just using a thought experiment to make sense of what I said.

Let's say my LFP is in good shape and can actually do the full 20AH of capacity. In theory, the battery nomial voltage is 12.8 V and with 20AH, that means it should be able to output at least 12.8v x 20a = 256 watts of power.
In my other Leaf experiments here: https://mynissanleaf.com/viewtopic.php?f=9&t=32820 I was measuring how much power it takes just to turn on the Leaf. My measurements record that every time I want to turn on my Leaf, it takes about 175 watts of power for 2 seconds to turn it on. In essence, it takes around (175 * (2/3600)) = 0.0972 watt/hours to start the Leaf. The battery is rated for 256 watt/hours, so every time you start the Leaf, you are depleting about 0.0972 / 256 = 0.00038 or 0.038% of the battery capacity with every push. So, in a perfect world, that would give you about 100 / 0.038 = 2,631 times to turn the Leaf on.

If I could never charge my LFP for a week for example because the temperature is always too low, there should be enough capacity to always make sure I can turn the Leaf on. Now this isn't accounting for the vampire drain of the computer in standby mode, which I measured to be around 0.158 watts of power. I haven't done a full day of measurements yet, but I'm sure every-time I access the Leaf with the app, that uses additional power, additional power is used when I sit down in the Leaf before turning it on, etc. All of those vampire drains will eat into that ideal starting total so it will certainly reduce it by some bit, but it seems that without the math, there should be enough left to cover starting the Leaf until the weather is warmer. Once I get a chance to measure a full day of "idle" power usage (I actually have a device that can do that on the 12V system), I'm going with my gut feeling that there should be enough capacity left over to keep one from being stranded in the Leaf even if the outside temperature is very cold for long periods of time.

The big problem I see with the above is that once the Leaf is started, the DC-DC converter will be tasked with recharging the starting battery. The charging system may be very mediocre at this, but you're still looking at the prospect of having to modify the car to not do that unless the 12 volt battery temperature is above, say, 37F. It's doable, but would require a substantial amount of work on the charging system. Maybe a switched diode array linked to a temp sensor and something like an Arduino...? Or maybe an Old School bimetallic thermostat and a relay, with the blocking diode(s)... Electronics are not in my wheelhouse, I'm afraid.

The other problem I see is that, as the LFP battery uses capacity, the closer it gets to undercharged, the less capacity is available for Accessory mode. So that would have to be disabled, still leaving the lights to drain the battery when they are on but the car is Off or in Acc mode.

Then you have the emergency flashers...

LeftieBiker said:

The big problem I see with the above is that once the Leaf is started, the DC-DC converter will be tasked with recharging the starting battery. The charging system may be very mediocre at this, but you're still looking at the prospect of having to modify the car to not do that unless the 12 volt battery temperature is above, say, 37F.

Click to expand...

These LFP 12v marketed for car applications have a BMS. Would the BMS prevent charging if battery temp is too low ?

But even if true, that means that the LFP 12v battery can go days/weeks/months in the winter without getting charged. Then all the loads you mentioned come into play

The one aspect of a "Starting" battery workload for an EV is that it more closely resembles a deep-cycle marine battery application than a traditional ICE starting battery. My own plan is for a deep-cycle (or combo) type replacement when the time comes (AGM or FLA). At the price difference for a comparable Lithium, that's not going to happen to just save a few pounds. My tomcat weighs more than the weight savings.

LeftieBiker said:

The big problem I see with the above is that once the Leaf is started, the DC-DC converter will be tasked with recharging the starting battery. The charging system may be very mediocre at this, but you're still looking at the prospect of having to modify the car to not do that unless the 12 volt battery temperature is above, say, 37F. It's doable, but would require a substantial amount of work on the charging system. Maybe a switched diode array linked to a temp sensor and something like an Arduino...? Or maybe an Old School bimetallic thermostat and a relay, with the blocking diode(s)... Electronics are not in my wheelhouse, I'm afraid.

Click to expand...

That's where I'm still fuzzy about how the Leaf charges the 12V battery too. When I was doing the 12V starting experiment to measure power going out to start the Leaf, my meter could actually measure current in the positive or negative direction. So, starting the Leaf took about +13.7 amps, then when the Leaf was started, I watched the current flow in reverse to the battery. It was only -1 amp, so basically the Leaf was feeding in about 13 watts of power to charge the battery. I shrugged my shoulders and watched if it would go up, but it didn't. :lol:
I should do a test where I power on the windshield wipers to make the voltage go up artificially to the +14V range and see how much power goes into the battery then.

The other problem I see is that, as the LFP's battery uses capacity, the closer it gets to undercharged, the less capacity is available for Accessory mode. So that would have to be disabled, still leaving the lights to drain the battery when they are on but the car is Off or in Acc mode.

Then you have the emergency flashers...

Click to expand...

Yeah, that is a good point, unless someone was paying attention to the power usage *knowing* that the 12V battery was not being "charged", I could see a really bad situation happening with that. I'm also trying to figure out if the reduced charging current for the LFP is an actual physics based limit or just a recommendation to prevent damage. Even if the battery was getting a tiny amount of power when freezing, it would mean it can still in theory be charged back up. :|

SageBrush said:

These LFP 12v marketed for car applications have a BMS. Would the BMS prevent charging if battery temp is too low ?

But even if true, that means that the LFP 12v battery can go days/weeks/months in the winter without getting charged. Then all the loads you mentioned come into play

Click to expand...

The one I'm using is too old to have that feature, the BMS only protects against low voltage, too high voltage, too much current, etc. If the new wave of 12V LFP come with built-in heaters standard, all of this discussion might be moot I guess.

I am curious though if LeafSpy is showing current going in and out of the 12V battery. I've noticed that on the 4th readings screen, it will often show the amp usage as both positive and negative numbers. It is showing what the Leaf is sending to the battery or taking from the battery, I'm not sure? :?

If the new wave of 12V LFP come with built-in heaters standard, all of this discussion might be moot I guess.

Click to expand...

Picture this: a Leaf with LFP battery is left parked for a week, in a driveway that gets down to 0F or lower every night. The internal heater consumes all of the charge. The BMS would also have to have a Smart Sleep mode in which it disconnects the cells from each other until conditions allow for charging. Even then, the BMS would have to use one of those cells for power, and would likely kill it. This has actually happened with some small EV packs. So the car would have to have a battery maintainer on it whenever there was risk of frigid weather lasting more than a day or so.

LeftieBiker said:

If the new wave of 12V LFP come with built-in heaters standard, all of this discussion might be moot I guess.

Click to expand...

Picture this: a Leaf with LFP battery is left parked for a week, in a driveway that gets down to 0F or lower every night. The internal heater consumes all of the charge. The BMS would also have to have a Smart Sleep mode in which it disconnects the cells from each other until conditions allow for charging. Even then, the BMS would have to use one of those cells for power, and would likely kill it. This has actually happened with some small EV packs. So the car would have to have a battery maintainer on it whenever there was risk of frigid weather lasting more than a day or so.

Click to expand...

Oh my mistake, I didn't clarify, the ones I've seen for sale have a heater that runs when you charge it while cold, not when it's sitting in the cold. So if the Leaf runs it's occasional charge cycle for the 12V, it should in theory be able to keep the 12V topped off? :?

Yes, hopefully. Fingers crossed!

Ok, some interesting data. Last night, I had a unique opportunity (for my area anyway) in which the temperatures fell to 19F. I knew this was coming, so I just left the hood up on my Leaf to let as much cold area get into the front where the 12V battery as possible. I also put a glass of water on top of my 12V battery to get an idea that it was really below freezing at that point. I came back out to the Leaf around midnight and the battery was covered in frost, the glass of water was a solid at that point, it was easily well below freezing. So, with everything turned off, I got a LeafSpy capture of what the Leaf was reading as battery voltage and current draw.

As you will notice from the first screen-shot, me sitting in the Leaf with the computer booted up in ACC mode is drawing about 3.54 amps or about (12.88 x 3.54) = 45.6 watts of power. So, make sure everything was off, no headlights, air, fan, unplug all power from the front console, and finally turn the Leaf on and just let it sit idle so I could watch how much power it would put back into the 12 volt battery.
At first, it started at about 10 amps, then quickly ticked down to roughly 0.6 amps and stayed there. I wanted to see if it would change or go back up but it was too cold to be sitting in the Leaf with no heat, so I just walked back into the house as my phone can reach that far with the Bluetooth with Leafspy. Watched the readings for about 15 minutes and it had not changed, it was still doing a "trickle" charge on the 12V battery. Doing the math, it was only feeding about (14.56V x 0.59A) = 8.6 watts of power into the battery. If this was a Flooded Lead Acid battery, I'm not sure if that would be enough to keep it topped off with a proper charge or not given how cold it was, but for a LFP battery getting below freezing means you can only charge at a very, very low rate. According to research papers, the allowable charge rate at –30°C (–22°F) is 0.02C at maximum. The warmer it gets, the amount ticks up a little until you finally reach the freezing point again. That means at -30C/-22F, my battery would max at (256 * 0.02) = 5 watts of safe charging current. Since LeafSpy was reading the motor at 26F, I'll use that for the battery too since it has a lot less thermal mass. I couldn't find a convenient online calculator for charge rate to temperature for the LiFePo4 12V battery. So I just used a research paper that had a -5C (23F) value instead where they tested to see what was the safe charge rate. It listed 0.05C for that temperature. It was a little lower than what I was seeing, but for this thought experiment, should be sufficient.
That means my 12V LFP battery should be able to safely charge at around (256 * 0.05) = 12.8 watts of power at that temperature.

It's interesting that the Leaf appears to be doing a low power trickle charge on purpose, well within the safety limits of my LFP. I'm curious now if that is just a normal behavior for all Leaf and charging the 12V battery or if it is actually lowering the current on purpose because it detects a below freezing outside temperature?
Either way, that may explain why my own and Stanton's LFP batteries have lasted as long as they have. Even when it's below freezing, the Leaf is not damaging our 12V batteries by trying to blast large amounts of current into it when it's below freezing. Either by accident or on purpose. :lol:

Ok, more interesting data.

I waited until 4pm today, plenty of sunshine and time to warm up above freezing for the Leaf. I took another reading with the Leaf on and this time it will gladly feed 5.5 amps or (14.56 x 5.5) = 80 watts of power into my 12V LFP. The only thing different other than the time of day is the temperature of everything. I made sure to turn on all the temperature screens in LeafSpy this time.
It seems to me there is some connection between the temperature the Leaf is reading outside and how much power it will push into the 12V battery. I haven't driven anywhere since last night, the Leaf has just been sitting out in the elements the whole time.

Another odd thing, I only noticed this because I left the hood up all night/day :lol: , when I plugged in the EVSE just now, the radiator fans came on for a few minutes. I didn't have the thought of mind to grab my phone and get a LeafSpy reading fast enough because I was bent over the engine compartment feeling the air blowing and wondering WTF? Why oh why would the radiator fans come on when it's obvious that everything in the Leaf would be very cold from the previous night. :shock:

I think the voltage is temp compensated...from my observations of the DC voltage...the DC to DC convert puts out the voltage called for by temperature (and windshield wipers on/off) and the amps are what the battery can absorb at that voltage etc.

This is another one of those things that a CANBUS bridge might be useful for. Depending on what parameters are available for the DC to DC converter, it should be possible to either tweak the settings or to spoof the outside temperature to manipulate the charging voltage (or possibly to spoof the on state of the wipers)

Dala is coming up with so many cool features that I'm thinking about buying a CANBUS bridge, despite having the original 24 kWh pack, just to be able to tweak things:

• https://github.com/dalathegreat,
• https://github.com/dalathegreat/LeafEnhancer-UserManual

Beyond what's possible using LeafSpy and/or factory settings.

Here's an interesting comment I found on the Tesla Owners Club, from "Ingineer" -- Judging from the avatar, the same Ingineer who was once active on this forum.

"I *highly* advise against putting any type of Lithium-Ion 12v replacement battery in any Tesla. One of the functions the lead-acid battery serves in the system is to function as a load dump and stabilizes the 12v rail. Anytime an inductive load is switched off, the excess energy can appear on the 12v bus and this spike can be dangerous to other devices on the bus if not absorbed. The "dumb" lead-acid battery has no BMS to disconnect it from the bus, so anytime the bus fluctuates or has a large load dump, the lead-acid battery happily obliges to keep the voltage level stable. All these Lithium-Ion replacements have an integrated BMS that is needed to protect the cells inside from overcharge/undercharge and when they become fully charged, the BMS effectively disconnects the internal cells from the 12v bus. This is the same as disconnecting the terminals and now there is no load stabilization, and sudden load dumps or demands have no sink/source. ...
"

Nubo said:

Here's an interesting comment I found on the Tesla Owners Club, from "Ingineer" -- Judging from the avatar, the same Ingineer who was once active on this forum.

"I *highly* advise against putting any type of Lithium-Ion 12v replacement battery in any Tesla. One of the functions the lead-acid battery serves in the system is to function as a load dump and stabilizes the 12v rail. Anytime an inductive load is switched off, the excess energy can appear on the 12v bus and this spike can be dangerous to other devices on the bus if not absorbed. The "dumb" lead-acid battery has no BMS to disconnect it from the bus, so anytime the bus fluctuates or has a large load dump, the lead-acid battery happily obliges to keep the voltage level stable. All these Lithium-Ion replacements have an integrated BMS that is needed to protect the cells inside from overcharge/undercharge and when they become fully charged, the BMS effectively disconnects the internal cells from the 12v bus. This is the same as disconnecting the terminals and now there is no load stabilization, and sudden load dumps or demands have no sink/source. ...
"

Click to expand...

Good info!

Nubo said:

Here's an interesting comment I found on the Tesla Owners Club, from "Ingineer" -- Judging from the avatar, the same Ingineer who was once active on this forum.

Click to expand...

Same guy -- Phll Sadow. He has become perhaps the pre-eminent independent EV hacker on the planet

His comment was from 2019, and it look like it related to a 3rd party LFP vendor named Ohhmu.
I would not presume he would say the same about Tesla designing an LFP for its own cars but it is possible.

In any case, I am in complete agreement with him that generic LFP solutions are early days and without integration assistance from the manufacturers it is at best a crap shoot. An EV in Texas might do OK with a 3rd party generic LFP, but it is a mistake to extrapolate that anecdote to other EVs, other climates, other LFP batteries etc.

Nubo said:

Here's an interesting comment I found on the Tesla Owners Club, from "Ingineer" -- Judging from the avatar, the same Ingineer who was once active on this forum.

"I *highly* advise against putting any type of Lithium-Ion 12v replacement battery in any Tesla. One of the functions the lead-acid battery serves in the system is to function as a load dump and stabilizes the 12v rail. Anytime an inductive load is switched off, the excess energy can appear on the 12v bus and this spike can be dangerous to other devices on the bus if not absorbed. The "dumb" lead-acid battery has no BMS to disconnect it from the bus, so anytime the bus fluctuates or has a large load dump, the lead-acid battery happily obliges to keep the voltage level stable. All these Lithium-Ion replacements have an integrated BMS that is needed to protect the cells inside from overcharge/undercharge and when they become fully charged, the BMS effectively disconnects the internal cells from the 12v bus. This is the same as disconnecting the terminals and now there is no load stabilization, and sudden load dumps or demands have no sink/source. ...
"

Click to expand...

I don't know how Tesla operates their 12V system in that regard, but I would highly disagree with his statement on these reasons.

1) Inductive loads are for AC not DC
2) Lead Acid batteries can not absorb power as fast as AGM or Lithium batteries, they are quite slow actually.
3) The 12V system is being moderated by the DC to DC converter, so the voltage is already being stabilized. In no situation that I see would turning off a device on the 12V system cause the voltage to rise above 16V that would trigger the BMS in modern LFP batteries. The DC to DC converter is just smarter than that because it's not just adjusting voltage, it's adjusting the current as well. You don't need to increase the voltage to increase the current
4) Current is pulled by the device on the 12V system, not forced into it. If the 12V system is pushing out 1000 watts and some device is switched off so that the power draw falls to 200 watts, that 800 watt difference is not going to bounce around the 12V system and settle anywhere, that's not how electricity works. The voltage might rise, but using Ohms Law, the power used by the devices will remain the same, regardless.
5) If what he said was actually true, then all of the failing Lead Acid Batteries in the Leaf would be frying electronics around the world non-stop everyday. Something that would probably make the news or get our attention here?

I'm mean no offense, but coming from a long background in electronics (among other things), everything he said sounds completely made up and would make his Electronics 101 Professor weep. :?

Per this post from sister site MyIMiEV.com, the BMS in at least some LFP battery modules seems to be able to prevent charging without totally disconnecting the battery from the car, i.e. allowing discharging. I'm slightly surprised at this; it must take come clever control of the MOSFETs between the bare cells and the car.

But assuming that this is correct, most of the smarts being attributed to the Leaf DC-DC are probably actually due to the battery module's BMS. The leaf decides on a charge voltage, and regulates the current into the battery to achieve that voltage. But the BMS regulates the impedance of the MOSFETs to achieve its desired charge current, which at low temperatures is presumably very low. So while the battery module terminal voltage may rise to 14.6V or so, which would cause a lot of current to flow into even a very cold lead acid battery, the LFP BMS shrugs it off and ensures that less than an amp of charge current actually flows.

Presumably, as soon as the direction of current flow is out of the battery, the MOSFETs turn on hard, unless the minimum cell voltage is critically low.

I like the idea of the battery warmers that only operate when charging a cold LFP battery module. The alternator or DC-DC sees roughly the expected current flow, but instead of dangerously charging the LFP cells, it warms them instead. As the cell temperature rises, the power can slowly or perhaps abruptly change from heating to charging. That way, no power is wasted warming the cells when there is no charge source present, yet the battery is conditioned to be able to receive charge when it is available.

I did a quick google search and found many 12V LFP that come with heaters built in for cold weather. I was surprised that so many are on the market now. Though what I found was mainly for the larger sizes (100AH, 200AH, etc.), it will probably only be a matter of time before they make their way down to the smaller capacity sizes. The prices do seem to reflect the heater upgrade though, ouch! :shock:

Oh wait, I did find one in the same capacity I have, ouch on the price though (https://relionbattery.com/products/lithium/rb20-lt), it cost a hundred dollars more than what I paid for mine 7 years ago.

My take on Ingineer's take is that he mentioned switching a 12V load. That sudden switch can certainly induce an inductive spike and is one reason why DC-rated switches are more complicated and expensive that AC-only switches.

That said, I'd guess almost any 12V load in a modern car uses something more sophisticated than a simple switch. Even my 2002 VW Passat has a snubber circuit on the relay that controls an air pump for this very reason.

I'm not a SLA battery expert but I'd expect the 'stabilization' provided by it to be more a function of impedance than any energy absorption. Inductive spikes are extremely short so while they can have enough energy to blast a hole in a silicon chip or pit the contacts of a mechanical relay, the actual amount of energy in them is fairly small.

All that said, the proof is in the pudding and it looks like so far Li 12V batteries are doing great in the Nissan Leaf. I put an AGM battery in mine when the time came but that was mostly for cost and ease reasons, not anything technical.