Mixing Lead with LFP

[sailingharry]

I'm knee deep in my LFP install, so my mind is filled with info. I'm an engineer, so that's it's own problem. And it's 16F outside today, so I don't feel like going to the boat. So you all are the victims of this situation!


My install isn't really affected by my thoughts. I have a pure LFP bank, with an external contactor-based BMS, so load dump is a non-issue (all charge sources to the LFP with BMS interrupt controls). I still have my Group 31 sealed start battery, and charge is through an ancient Ample Power Isolator Eliminator, which appears to be a 5A B-B charger with temperature compensation. It has charged my start battery since the boat was new (26 years) and seems up to the task.


BUT, discussions here and elsewhere have me pondering the mixing of chemistries, either directly in parallel with the LFP (solves a lot of issues with Drop-In batteries) or with auxiliary/backup batteries on critical loads, or start batteries.


So, the first thing I did was pull some SOC/Volt tables from the Web. Yeah, all of them are bad, there are many variations, and Volts is an awful way to look at SOC anyway. But it's what we have. The table, with an associated graph, is attached.


Some key points that show up:
* LFP max voltage is below AGM max voltage
* LFP full (100%) is very close to AGM bulk/absorption
* LFP at 80% is very close to AGM float
* LFP at 10% is higher than AGM at 10%
* LFP above 14% is higher than AGM at 80%


Conclusions?


- In a direct parallel, an AGM will live almost all of its life between 80% SOC and normal bulk/absorption charge voltages
- As long as the LFP is below 80% SOC, the AGM is being held at about 80% SOC or above, a very happy place for AGM (as long as LFP is above 40%, the AGM won't discharge at all!).
- Recognizing that any AGM mixed with LFP rarely if ever drops below "full," it should live a long life
- An AGM in this use could/should be a Group 31 or smaller (even a motorcycle battery for many applications), and so is exceptionally cheap. Replace every 10 years if it needs it or not.


Charging an AGM from LFP (if not a direct connection) could be done with a B2B, or a ACR (set at say 13.6 or 14.0). However, I note that the LFP is slightly higher than the desired AGM profile.


If you mix in a diode, such as this 80A (with much higher short term currents) from Mouser for $20 https://www.mouser.com/ProductDetail...adaYqpow%3D%3D, then the voltage on the AGM would always be .5-.7V less. This means that if you want to hold the AGM above 12.5V (the 80% number), you have to hold the LFP above 13.2, or about 60% SOC. Since a large percentage of the time the LFP will be well above that number, a start battery could be easily kept above 80% SOC and should hit 100% SOC daily (and since they get very little Ah draw, stay above 95% all the time).


Note, a Victron ARGOFET isolator virtually eliminates voltage drop and could be used (only 1/2, but you could use the other half for another battery) for $100.



As I said, I'm not doing any of this. But, I think that Lead can be easily paralleled with a Drop-In with significant advantages, a start battery can be easily and cheaply charged from LFP via a simple diode (fuses on both ends!), and makes a great solution to keeping critical power up on a load dump (imagine a LFP feeding a diode feeding a 10Ah sealed battery, $25 all over the place, which feeds your house panel and keeps the lights/chartplotter/etc on for 15 minutes or more).


One big thing I noted is that conventional wisdom has it that LFP has a much flatter SOC curve than AGM. For the data I pulled off the Web, this doesn't appear to be true. From 80% to 20%, AGM drops from 12.5 to 11.7, or .8V, and LFP drop from 13.4 to 12.8, or .6V. Yes, 25% less, but not exactly game changing. However, from 90% to 70%, perhaps the "normal range," the difference in drop is even less (.32 vs .30). However, throughout the operating range, LFP runs about 1V higher than AGM, a much more significant improvement.


This higher operating voltage calls into question the ABYC voltage drop standard, which is based on % of input. In an AGM battery at 12V, the max drop of 3% is .36V yielding an acceptable voltage at your VHF of 11.64. However, in an LFP battery at 13V, you can get that same 11.64V with a 1.36V drop or a 10% drop! Maybe ABYC should shift from a "Percent standard" to a "Volts at the device" standard (I know I think about that all the time if it's an issue -- thrusters are designed for way more than 3% drop, and drop on a steaming light is kind of meaningless when the alternator is putting out 14.5V).



OK, any counter thoughts?

[wholybee]

Lots of people have done the math, and you have now repeated it. The math works out, except at high C Rates, which aren't so easy to model, and probably won't happen on a boat.

However, all the standards organizations prohibit it. All the manufactures prohibit it(except possibly one drop in mfgr in the UK, but I bet the cell mfgr doesn't allow it) No professional will install it. And no insurance company would cover it. If you ask, be explicit "I wish to install Lithium batteries in a way not allowed by ABYC or ISO, and against manufactures guidelines, will you cover it if my batteries catch fire?"

With that said, yeah, it works. But I always come back to the same conclusion. Why? There just isn't any huge problem that it solves. Any (IMHO small) issues are easily solved other ways. Many thousands of boats install drop-ins without using hybrid.

There is no benefit. But, you introduce Peurkert into a battery bank that otherwise wouldn't have it. SOC becomes impossible to know, whereas it's very easy to get an accurate SOC with a pure LFP bank. You take a very safe LFP bank and add acid and hydrogen gas to it. You now need to worry about ventilation and acid proof enclosures. There are restrictions about what can safely be mounted near lead acid due to corrosion from hydrogen gas, and everything near needs to be ignition protected.

And LFP is no more expensive than AGM, so you are not even getting a benefit of more capacity for cheaper.

So, my conclusion is always the same, don't do it because it's just a bad idea that solves nothing and makes an otherwise great LFP bank worse. It's only even discussed because one popular YouTube cruiser made a series of videos on it to sell his useless device.

[sailingharry]

Originally Posted by wholybee

With that said, yeah, it works. But I always come back to the same conclusion. Why? There just isn't any huge problem that it solves. Any (IMHO small) issues are easily solved other ways. Many thousands of boats install drop-ins without using hybrid.

There are 3 places to use Lead. One is in direct parallel, which is what you are talking about. I'll agree, it's the least useful place.


Another is your start battery. I'm not sure any suitable LFP start battery can be found for the price of a sealed start battery ($70-$150, depending on what you need). But somehow, they need to be charged -- and using a diode isolator (you may even already have one you aren't using now) is an option. My boat has bascially a B-B, so my start battery is set.



Finally, a battery for critical power. I'm not sure what would qualify as "critical power" on a sailboat, but some people think things like bilge pumps, chart plotter, VHF, AIS, RADAR, running lights, etc are "critical." If "something" happens (thermistor wire breaks and "sensed temperature" goes out of bounds) and the BMS trips the main output, an inline battery fed by a FET isolator would keep all those systems still on line. A cheap sealed battery, 10-20Ah, is well under $100 (the FET and wiring, mounting, etc would cost more!), could be fed by the LFP, and provide a reasonable length of temporary power.


I don't use any of these. My LFP is esentially immune to alternator problems, my start battery has a B-B, and I'm not worried about "critical systems."


Note on "critical systems." My chart plotter has been known to have issues so if I'm in an area where I'm concerned, I have my phone queued up. Radar and AIS are nice, but not something I can't survive 5 minutes without while I resolve the issue. And running lights don't rise to my idea of "critical" (last winter, crossing the Gulf of Mexico, the jib whacked a wire and our port running light went off line -- we didn't even think about fixing it until the next morning).

[donradcliffe]

Is your LFP bank immune to problems at 15 degrees F? What do you do to layup for winter, or leave the boat on a mooring? Or leave the boat plugged into dock power?

[AndyOz]

I believe the main reason paralleling LFP with lead chemistries is frowned upon is because if the lead suffers a shorted cell (a relatively common failure mode), the substantial LFP energy is dumped into the short and greatly increases the chance of a severe fire.

There's ways to manage this, but a direct parallel connection should be avoided.

There's considerable advantages of paralleling, such as avoiding spikes from BMS disconnects, but the chance of burning your boat to the waterline almost certainly outweighs this.

[donradcliffe]

Originally Posted by AndyOz

I believe the main reason paralleling LFP with lead chemistries is frowned upon is because if the lead suffers a shorted cell (a relatively common failure mode), the substantial LFP energy is dumped into the short and greatly increases the chance of a severe fire.

Having lived through one shorted cell with gel batteries, I don't think there is much chance of a fire. The lead battery has a relatively high resistance which limits the current between the batteries. After the short gave me a low voltage alarm at 0300 I got the engine started, and was putting 100 amps from the alternator into the battery isolator. All that current went to the shorted battery until the isolator blew. The battery got warm, but was no where near on fire.

[AndyOz]

Originally Posted by donradcliffe

Having lived through one shorted cell with gel batteries, I don't think there is much chance of a fire. The lead battery has a relatively high resistance which limits the current between the batteries. After the short gave me a low voltage alarm at 0300 I got the engine started, and was putting 100 amps from the alternator into the battery isolator. All that current went to the shorted battery until the isolator blew. The battery got warm, but was no where near on fire.

Depends on the nature of the short. Could be much lower resistance than that, and an LFP battery could supply multiples of that current. If the electrolyte boils dry, the temperature could soar.

[sailingharry]

Originally Posted by AndyOz

Depends on the nature of the short. Could be much lower resistance than that, and an LFP battery could supply multiples of that current. If the electrolyte boils dry, the temperature could soar.

Hm....Interesting point. It's not really a "dead short," it's a "shorted cell," so your AGM becomes a 5-cell battery. With an LFP at, say, 13.5 and ignoring all voltage drops (internal resistance, bus bars, fuses, wires, connections), that results in 2.7V per cell, which is the same as charging a healthy battery at 16.2V. Certainly not good for the AGM, but I'm not sure that a few hours at 16.2 will cause a fire. Days, for sure! But I'm not even sure that typical start-battery-sized AGM could sink 100A with a shorted cell.

[CaptainRivet]

There is 0, absoluetly 0 where lead makes sense nowadays.
Lifepo4 is cheaper, Needs no complex floating and waste of energy, lasts triple, is tons more reliable, has higher Charge efficency, can be kept Stored at any voltage below 3.35V cell voltage….
All Arguments are the lead mind Training over decades and find all Kinds of Strange combinations to justify using lead while everyone knows the Most unreliable component in an electrical system are the lead Battery. So it’s 100% garunteed that lead Battery will let you down, sooner then later.
And gets argumented that a Max 300 cycles lifespan lead is the security backup for a Lithium house with 8000cycles, it’s simply not. It’s physically not possible.
Just throw that unreliable, Heavy and Stoneage Technology lead into the next bin, that‘s where it belongs.

[s/v Jedi]

I have tested this, running a 10kWh LFP bank parallel to a Lifeline L16 AGM bank (years old but same Ah as lithium so also around 10kWh). There were no explosions but after one season the Lifeline bank failed.

After isolating it and observing the cycles of LFP I found that the AGM only handled 20% of the cycles. I did not notice any difference other than numbers on a display.

I have since ran a “cheap” LiTime 10kWh LFP bank parallel to the primary LFP bank. It handled 40% of the cycles with the primary bank doing 60%. The primary bank has Winston cells and simply has less internal resistance so takes the lead.

That said, after one year both LiTime batteries had a BMS failure. LiTime replaced them under warranty but they could not tell me what caused this. My diagnosis was that everything worked except charging. Discharging worked as well as the LVC protection.

If things pan out I will be out again this year with the same setup with the new LiTime batteries.

My primary reference diagrams are still the only setup that I have seen 100% reliable service.

[CaptainRivet]

Originally Posted by s/v Jedi

I have tested this, running a 10kWh LFP bank parallel to a Lifeline L16 AGM bank (years old but same Ah as lithium so also around 10kWh). There were no explosions but after one season the Lifeline bank failed.

After isolating it and observing the cycles of LFP I found that the AGM only handled 20% of the cycles. I did not notice any difference other than numbers on a display.

I have since ran a “cheap” LiTime 10kWh LFP bank parallel to the primary LFP bank. It handled 40% of the cycles with the primary bank doing 60%. The primary bank has Winston cells and simply has less internal resistance so takes the lead.

That said, after one year both LiTime batteries had a BMS failure. LiTime replaced them under warranty but they could not tell me what caused this. My diagnosis was that everything worked except charging. Discharging worked as well as the LVC protection.

If things pan out I will be out again this year with the same setup with the new LiTime batteries.

My primary reference diagrams are still the only setup that I have seen 100% reliable service.

Wow only 1 year life from your hyped Li-time. And the very incompetent, basically useless Technical Li-time service cannot even Tell you what was the exact damage and what caused it, no surprise at all.
Let me tell you whats wrong:
The cheap BMS with back to back installed uncalibrated mosfets couldn‘t handle the surges in your system and one mosfet died after another without you been able to notice it as the Battery is sealed. Then out of nowhere the BMS doesn‘t allow Charge as the last Charge mosfet died but because you have no glue Whats going on inside the Chinese surprise Box even that you don‘t know because 0 Communication.
That‘s what Will Prowse doesn‘t Test but you as their Customer. And that‘s what Nobody can write after one year on Amazon….you get what you pay for.

Buy an Eel-Kit with a JK BMS and get grade A+ cells and for less money then a comparable Li-time you have a Battery that works equal to your Winston primary one.

[s/v Jedi]

Originally Posted by CaptainRivet

Wow only 1 year life from your hyped Li-time. And the very incompetent, basically useless Technical Li-time service cannot even Tell you what was the exact damage and what caused it, no surprise at all.
Let me tell you whats wrong:
The cheap BMS with back to back installed uncalibrated mosfets couldn‘t handle the surges in your system and one mosfet died after another without you been able to notice it as the Battery is sealed. Then out of nowhere the BMS doesn‘t allow Charge as the last Charge mosfet died but because you have no glue Whats going on inside the Chinese surprise Box even that you don‘t know because 0 Communication.
That‘s what Will Prowse doesn‘t Test but you as their Customer. And that‘s what Nobody can write after one year on Amazon….you get what you pay for.

Buy an Eel-Kit with a JK BMS and get grade A+ cells and for less money then a comparable Li-time you have a Battery that works equal to your Winston primary one.

I don’t think so because it was the charge bus that failed, not the discharge bus. There are no surges on my charge bus… the sun comes up slow and steady and goes down the same way and the charge follows that.

Fact is that there was a BMS failure. Warranty is handled by their sales and customer support staff, not by techies and once UPS picked up the boxes for return shipment they were gone and I never got word on what was done with them. If it were my company, the service department would cut open the case, put the contents in a new case with a new BMS, test, seal and put in inventory for warranty replacements. I wouldn’t even examine the BMS because a company is for making money, not doing research for component manufacturers.

Also, Will’s forum has a large membership reporting on these batteries and LiTime has a very good reputation. You are an activist against Will Prowse and LiTime even though what you come up with is inferior. All you care about is cheapest diy options and anything else gets attacked like if your life depends on seeing that your ideas are right and the rest of the world is wrong.

[CaptainRivet]

Originally Posted by s/v Jedi

I don’t think so because it was the charge bus that failed, not the discharge bus. There are no surges on my charge bus… the sun comes up slow and steady and goes down the same way and the charge follows that.

Fact is that there was a BMS failure. Warranty is handled by their sales and customer support staff, not by techies and once UPS picked up the boxes for return shipment they were gone and I never got word on what was done with them. If it were my company, the service department would cut open the case, put the contents in a new case with a new BMS, test, seal and put in inventory for warranty replacements. I wouldn’t even examine the BMS because a company is for making money, not doing research for component manufacturers.

Also, Will’s forum has a large membership reporting on these batteries and LiTime has a very good reputation. You are an activist against Will Prowse and LiTime even though what you come up with is inferior. All you care about is cheapest diy options and anything else gets attacked like if your life depends on seeing that your ideas are right and the rest of the world is wrong.

If a mosfet BMS dies from load surges it's typically the back to back installed charge mosfets that gets too much reverse current and dies while the load mosfet survives as that has much more reserve. Because they are uncalibrated the weakest charge mosfet has to withstand in the first milliseconds the whole reverse current and dies.
And when the load mosfets dies it's mostly an a charge surge that killed them.
Load surges are common, eg startup of a motor like AC and that's why it's mostly the charge side that dies first, especially if the mosfet are of big tolerance spec=cheap.

You are completely wrong, I like Will Prowse very much as he is authentic and simply shows every flaw plus lso straight forward.
But you have to be aware what Will is and can test here and where the limits are. What will tests is simply a tear down video what's inside that box, the general build quality and if it fullfills the spec the seller=eg li-time state when new. That's it.
He doesn't test the longevity and cell quality, he just tests if the overcurrent protection works but doesn't test if the BMS is dead after 10 or 20 or 5 overcurrent protection disconnect....
Andy from offgrid garage is different, he test also the stuff in daily operations like the JK BMS that run flawlessly since 4 years and he continuesly documents eg the status of the cells.

Li-time is just another reseller of a Chinese battery factory, they are not the real manufacturer. It's just manufactured in their name with their sticker on. Its not a bad battery for the money but its also not as good as you lately hype them. Its a good budget battery with good low quality tier 3 cells, thats sufficent for most offgrid and RV uses as well as costal boats but has nothing to do in bluewater boats.
So what they simply do is put the warranty returns in a container and ship them back to the manufacturer and let him deal with that. The manufacturer reimburses the whole return container. That's common practice.

I am sure what you described happens then in China that they simply sell that container to a reseller which dismantles them, replace BMS or cells from other battery, put them into case with another sticker and sell them under a different brand name.

[sailingharry]

Originally Posted by CaptainRivet

There is 0, absoluetly 0 where lead makes sense nowadays.
Lifepo4 is cheaper, Needs no complex floating and waste of energy, lasts triple, is tons more reliable, has higher Charge efficency, can be kept Stored at any voltage below 3.35V cell voltage….

I can buy a lead start battery for under $150, drop it in the place where the old one was, and be good for a decade. How much is a reliable LFP (or LTO) for a start battery? I am sure my house bank can start my engine, and I will have that ability in the existing wiring, but I sure do like having two independent batteries.


Some people stress a lot about the ability of the LFP to trip. Not likely, but possible. For $150 or so, I can buy all the parts to add a 30Ah sealed battery in line to the house panel board. This should run the panelboard and all systems on it for quite a while (probably 2+ hours on my boat). This cheaply and easily solves the "dark ship" risk. While I see loss of all DC using a DIY with an external BMS and direct shutdown of all charge sources BEFORE dump as just not plausible, most packaged units and most FET based BMS don't have this ability and loss of all DC is a real risk. I'm not sure how to mitigate this risk for less, without using a small lead battery.


The direct parallel between lead and LFP does have real benefit. Primarily for low-end (imagine an older 35' boat with minimal systems -- perhaps no solar, a factory internally-regulated alternator, etc). Alternator dump is a real concern for them, and requires some solution. Some BMS (mine) don't control the charge bus for temp -- under temp, trip the main. I've set the under-temp at an unreachable temperature (-17C, about 0F) and will manage under-temp manually, but if you don't have that control in the BMS, under-temp is a realistic dump scenario. But regardless of why the BMS dumps, a parallel AGM is arguably one of the least expensive solutions. Charging the Start and then a DC-DC is perhaps cheaper, easier, and for DC loads under maybe 30A total (that 35' boat), can avoid any load dump as well (the DC-DC can pick up all loads).

[wholybee]

Originally Posted by donradcliffe

Having lived through one shorted cell with gel batteries, I don't think there is much chance of a fire. The lead battery has a relatively high resistance which limits the current between the batteries. After the short gave me a low voltage alarm at 0300 I got the engine started, and was putting 100 amps from the alternator into the battery isolator. All that current went to the shorted battery until the isolator blew. The battery got warm, but was no where near on fire.

I had a shorted cell with FLA before I switched to LFP. I found it because of the smell of gas, and at that concentration was probably explosive.