LiFePO4 Batteries - Okay Tear Me Apart ;-)

[gael]

Originally Posted by S&S

I setting up to go with LiFePO4 (Thundersky)for the house bank- See the "Running two Alternators" thread. To make best use of the available charge acceptance rate you really need BIG alternators.
The cost amortization works out even at the presently higher pricing. We're setting our BMS for 75% discharge and can expect a 20+year pack life.
What I've found is that alternator charging large banks of LiFe cells takes more finesse if you want reliability, and I still have to build an alternator controller that talks with the BMS. For lower field coil currents than what I'm using one of the Balmar or Ample regulators may work.

Evans, they're not magic but they do provide a higher energy density than lead acid. Working within their constraints I'd expect a good system.
To get the value out, you really have to DIY. Genasun or Elithion seem to be the best BMS's around for marine use.

Nice to see your interest, jallum.

I'm just coming from the "Running two Alternators" thread.
Seem we are ending to the same conclusions.
but still few questions before to jump...
I saw few benchmark Thundersky vs Sky Energy or Heter ... why did you choseThundersky?
I have already send several mails to Alex (genasun) . I feel this the best horse for marine, even if the price of a full battery pack is definitively high. In Europe the average price for a raw 160Ah cell is less than 200$ so 1600$ for a 24V bank.
Which alternator brand and regularor did you select?
Have you a diagram of your electrical system?

Did you notices electric or magnetic interferences when running all these electronics? fyi many I have discarded many LED drivers for this reason and I dnt want to screen all electronics for EMI/RFI

There is also Valence Technology, based in Austin (Tx), which seem to have a nice product line.

Definitively this thread on LiFePO batteries is very valuable.

[S&S]

Originally Posted by gael

I'm just coming from the "Running two Alternators" thread.
Seem we are ending to the same conclusions.
but still few questions before to jump...
I saw few benchmark Thundersky vs Sky Energy or Heter ... why did you choseThundersky?
I have already send several mails to Alex (genasun) . I feel this the best horse for marine, even if the price of a full battery pack is definitively high. In Europe the average price for a raw 160Ah cell is less than 200$ so 1600$ for a 24V bank.
Which alternator brand and regularor did you select?
Have you a diagram of your electrical system?

Did you notices electric or magnetic interferences when running all these electronics? fyi many I have discarded many LED drivers for this reason and I dnt want to screen all electronics for EMI/RFI

There is also Valence Technology, based in Austin (Tx), which seem to have a nice product line.

Definitively this thread on LiFePO batteries is very valuable.

Thanks Gael-
I went with thundersky, initally due to that they had the cell sizes I was looking for. There are some differences between prismatic cell chemistry but as long as you set up your BMS and charger for the cells your using, I think the choice can be made based on what's available and at what cost. Alex's cells are great, but Sky Energy doesn't make big enough cells for my application.
Valance is nice, but very pricey.

Due to wanting to charge the banks fast I went with a C.E. Niehoff C671 (400A @ 15.5V). It uses their stock adjustable regulator.
They make 24V alts. up to 500A. Niehoff also has the advantage of being both brushless and air cooled. What I do have to have made is the controller to douse the field coils of the alt based on a BMS signal. That's being engineered now.
Wiring is straightforward: series to get the voltage, then parallel to get the current. As to wiring the BMS, I'd just follow Alex's recommendations.

Almost all of the electronics are in the lazarette with the batteries so I'm not expecting a big EMI problem. I'll cross that bridge when I get to it.

[jallum]

The batteries I ordered (Thundersky 300Ah's) have begun to arrive. Here's a picture of the first batch, adding up to 300Ah@24v (or 600Ah@12v!)... with a National Geographic for scale. Total cost for this batch: $2640.00, plus shipping. This is going to be good fun. Now, to select a BMS.

[electric1]

Greetings everyone!

I come from DIY Electric Car community where we have used LiFePo4 cells for last couple of years. This is wonderful technology and it beats Lead Acid hands down. Subject of BMS has always been shrouded in a mystery, which deterred many people from these batteries. Many companies make super expensive and complicated BMS systems, but it doesn't have to be either complex or expensive.

As I built my own electric car with LiFePo4 cells, I also designed simple and affordable BMS, which I am now offering for DIY EV market. One of my customers asked if my BMS can be used in marine environment to replace house banks of Lead Acid batteries. Sure it can, its very flexible and simple to integrate with 4 LiFePo4 cells for drop-in replacement of 12V Lead Acid battery, or 8 cells to replace 24V battery bank, etc.

I am attaching wiring diagram here for drop in replacement of 12V Lead Acid battery for a house bank. Also here is the link to my BMS site, where you can find user guide and ordering info.

As for cells, Thundersky and SkyEnergy are 2 popular names, but there is also HiPower which has a new cell model competing with others. I am curently awaiting a test batch of new HiPower cells to confirm their quality. I am working directly with factory in China and planning to bring those cells to US to offer complete kits for marine and EV use.

I have decent experience with LiFePo4 cells by now, so I hope to be able to help your community here if you allow me.

Thanks

[witzgall]

Sir,

Welcome to cruisers forum. it is great to have a BMS builder here to talk to us about how these might work on a boat. The typical cruising sailboat is rather unique when compared to an EV auto. EV autos typically have one charging source, the AC charger. If you were to take my boat as an example, we have FOUR different charging sources:

1. Engine Alternator. Mine is 50a, with an internal regulator. Others have larger alternators, and external regualtion. Balmar and Ample Power both make popular external, regulators.
2. Solar. I have 240w of solar panels, wired to a BZ500hv MPPT solar controller. The panels are putting 24v into the controller. Blue Sky makes to most popular MPPT controllers in the US market, there are many more.
3. Wind generator. 200w 12v wind generator, Air-x. It has it own internal regulator.
4. AC Charger. Used when on shore power, 50a, Iota smart charger.

What the boating industry needs to make DIY LIFEPO4 feasible is a way to make all of these charging sources play nicely with the cells. The electronics should be able to withstand a salt water environment. Finally, for a 12v setup, high voltage needs to be controlled, to avoid damaging electronics, etc. I bring this up as the Thundersky cells can get to over 17v when wired as a 12v pack, and that is too high.

I would be glad to talk this over with you it you have any questions. To let you know when I am coming from, I have a solar array and battery backup on my house, as well as a 24v LIFEPO4 (Skyenergy)pack for my Torqueedo electric outboard dingy motor.

Chris

PS - If you want to continue the conversation, you should register yourself as a Vendor on this forum, as we like to keep things transparent as to a poster's motivations. This will be required by the forum's moderators.

Originally Posted by electric1

Greetings everyone!

I come from DIY Electric Car community where we have used LiFePo4 cells for last couple of years. This is wonderful technology and it beats Lead Acid hands down. Subject of BMS has always been shrouded in a mystery, which deterred many people from these batteries. Many companies make super expensive and complicated BMS systems, but it doesn't have to be either complex or expensive.

As I built my own electric car with LiFePo4 cells, I also designed simple and affordable BMS, which I am now offering for DIY EV market. One of my customers asked if my BMS can be used in marine environment to replace house banks of Lead Acid batteries. Sure it can, its very flexible and simple to integrate with 4 LiFePo4 cells for drop-in replacement of 12V Lead Acid battery, or 8 cells to replace 24V battery bank, etc.

I am attaching wiring diagram here for drop in replacement of 12V Lead Acid battery for a house bank. Also here is the link to my BMS site, where you can find user guide and ordering info.

As for cells, Thundersky and SkyEnergy are 2 popular names, but there is also HiPower which has a new cell model competing with others. I am curently awaiting a test batch of new HiPower cells to confirm their quality. I am working directly with factory in China and planning to bring those cells to US to offer complete kits for marine and EV use.

I have decent experience with LiFePo4 cells by now, so I hope to be able to help your community here if you allow me.

Thanks

MiniBMS link

Attachment 13872

[electric1]

Chris,

thanks for your warm welcome! I have no problem registering as a vendor, but can't see it as an option during registration process, maybe I missed it. Admin can change my account any time, no issues here. My intentions are very transparent, I have nothing to hide. I have a good reputation in EV community and its worth more to me than money. I am fascinated with LiFePo4 batteries and want to help people switch over and drop the lead.

There have been some good remarks earlier in this thread that lead acid is fine for occasional cruisers who don't spend several days at sea and upfront investment in LiFePo4 is not worth for them. I totally agree with this statement. LiFePo4 with today's price and limited availability is only beneficial for those who spend a lot of time running generators to keep their house bank charged. My customer who is installing my BMS on his sailboat spends many nights at sea and claims that he has to run diesel generator at least twice a day, which is a waste of fuel and huge inconvenience. With 200 AH LiFePo4 cells he can easily run for over 24 hours without starting the genset.

As for your points of multiple charging sources, the answer is very simple. Anything that currently charges Lead Acid battery, can charge LiFePo4 and then some. LiFePo4 cells are much less sensitive to charging current and voltage, they can absorb insane charging current, up to 1C without any issues, and up to 3C if done very carefully. I don't think ANY of your charging sources can ever overrun LiFePo4 cells. My customer is swapping his 30 Amp alternator with 100Amp model, which will reduce his diesel runtime 3 times and saving him lots of fuel in a long run. This could not have been done with same size Lead cells since 100Amp would boil them and destroy them.

My idea of simple drop-in replacement is that you don't have to worry about any other components, just treat the LiFePo4/BMS kit as you would treat any other battery.

Voltage control is also easy, LiFePo4 is usually charged to 3.6V - 3.8V per cell depending on the model. BMS will not allow it to charge higher, but that is not even possible since all your charging equipment designed for Lead Acid will never let it charge over 14.5V , which is perfect level for 4 LiFePo4 cells in series.

[imcbride]

Originally Posted by electric1

With 200 AH LiFePo4 cells he can easily run for over 24 hours without starting the genset.

My idea of simple drop-in replacement is that you don't have to worry about any other components, just treat the LiFePo4/BMS kit as you would treat any other battery.

I think the issue being identified is with the way most of us charge our batteries -- a high-powered alternator on the main motive engine. On a "bad" day when we "have to get somewhere", it could mean running the engine (and spinning the alternator) for 9 hours or more. This is different than controlled running of a genset run/stop to maintain battery charge.

AFAIK, all standard alternator controllers will keep trying to stuff more current into the battery bank as long as the alternator/engine is running. Assuming all cells are topped up after a few hours, what will the BMS/LFP battery bank do with this excess current? Lead/SLA/GEL batteries don't mind a little overcharge at low current so it hasn't been a problem. Same goes for charging from solar panels. If the batteries are full the current has to go somewhere.

As an aside, the user manual for your MiniBMS seems to indicate that the "ignition"/current solenoid has to be OFF to activate the BMS high voltage alarm. Doesn't this preclude the very common (if not universal) usage profile of both charging and discharging the battery bank at the same time?

Iain

[electric1]

Iain,

thanks for bringing up these scenarios. I realize that engine sometimes runs for propulsion and not just to charge the battery.

Once the battery is charged and its voltage is equal to charger's voltage ( whether its alternator or solar or mains line is not relevant, any charger has max voltage ) the current is no longer flowing into the battery since there is no voltage difference. Some solar systems switch to a dumping load, like a water heater or what not, alternator can simply run unloaded and not put any mechanical load on the engine since no current is flowing. None of these scenarios will have any impact on LiFePo4, at least no different than on Lead Acid battery. BTW, contrary to your statement, SLA and Gel LA batteries HATE overcharge and will die much sooner if frequently overcharged. Only flooded batteries can sustain overcharge by boiling off extra water, which then has to be added or battery will be damaged.

MiniBMS user guide was writtent for EVs, and BMS usage in marine environment is a little different. As you see in my marine diagram I have a push button for BMS reset instead of permanent "Ignition" connection. This is because in EV BMS is powered by separate AUX 12V battery and manages main traction pack, but in marine use BMS is powered by the same battery its managing.

My BMS has single wire NC loop, so it does not differentiate LVC and HVC events. Its basically "All is Good" signal, which is interrpupted when battery voltage is too low or too high. In case of marine use when battery is charged and discharged at the same time, BMS will simply have no action until charge/discharge balance tips too far into low voltage or high voltage condition. In that case contactor will break the circuit and save the battery while buzzer will get your attention below the deck.

HVC is virtually impossible in marine use since your charging sources are not capable of more than 14.5V after regulators, since they are all designed for Lead Acid battery. BMS is mostly used as insurance policy to protect from overdischarge or as we call it "LVC event".

Again, this BMS was dsigned for EVs, but easily adapts to marine use since its so simple and flexible.

Thanks for asking all the good questions, this has been a great conversation so far.

[imcbride]

Originally Posted by electric1

Once the battery is charged and its voltage is equal to charger's voltage ( whether its alternator or solar or mains line is not relevant, any charger has max voltage ) the current is no longer flowing into the battery since there is no voltage difference.

I might be missing out on some electronics 101 here, but my understanding is that LFP battery banks are charged in a profile of Constant Current then Constant Voltage (CC->CV). Current up to a threshold voltage (~14.5V on a 12V nominal pack) then push current at that voltage until absorbed current falls below ~.01C. Other BMS shops have said that the charge source needs to be disconnected at that point. This brought up the problem that just opening a solenoid on the circuit from the alternator under load would blow its diodes. Also, on a large bank, say 400Ah, .01C is 4A, your MiniBMS shunting is rated at only .75A?

BTW, contrary to your statement, SLA and Gel LA batteries HATE overcharge and will die much sooner if frequently overcharged. Only flooded batteries can sustain overcharge by boiling off extra water, which then has to be added or battery will be damaged.

You're right, they do care. But at the small overcharge we're talking about won't kill them immediately. As a new technology, we don't know how sensitive LFPs are so we're treating them with kid gloves.

BMS usage in marine environment is a little different. As you see in my marine diagram I have a push button for BMS reset instead of permanent "Ignition" connection.
...
BMS will simply have no action until charge/discharge balance tips too far into low voltage or high voltage condition. In that case contactor will break the circuit and save the battery while buzzer

Noted, thanks for the clarification. So the user only needs to reset the BMS after a high/low voltage error condition?

Again, this BMS was dsigned for EVs, but easily adapts to marine use since its so simple and flexible.

I agree that this style of BMS as advertised is simple. This is a refreshing outlook. I think I speak for all of us who are trying to get our heads around LFP battery banks when I say that experience like yours in already deploying these systems is appreciated and increases confidence in adapting them for our own needs.

Two additional concerns I have are protection of the BMS PCBs from the rough (corrosive) marine environment, and secondly the method of connection of the BMS wires to the tabs/PCB of the cell chips. ABYC standards have told us not to use spade connectors or soldered connections for anything. Reasoning as I understand it is to avoid disconnects or mechanical wear in the rough, high movement environment of a boat. Any thoughts on these issues?

[S&S]

You have to remove the contactor and take those wires to at least a cap then to the field coil of the alt.

[witzgall]

S&S, can you elaborate? What do you mean by " At least a cap"? a Capacitor? Why, and how large?

Chris

[CharlieJ]

imcbride-

ABYC standards have told us not to use spade connectors or soldered connections for anything.

Allow me to clarify/amplify the ABYC Standards (E-11)

Spade connectors:

11.14.5.4 Terminal connectors shall be the ring or captive spade types. (See FIGURE 17)
EXCEPTION: Friction type connectors may be used on components if 1. the circuit is rated not more than 20 amperes or the manufacturer's rating for a terminal designed to meet the requirements of UL 310, “Electrical Quick-Connect Terminals”, or UL 1059, “Terminal
Blocks, and 2. the voltage drop from terminal to terminal does not exceed 50 millivolts for a 20 amp current flow,
and
3. the connection does not separate if subjected for one minute to a six pound (27 Newton) tensile force along the axial direction of the connector, on the first withdrawal.

Solder Joints:

11.14.5.7 Solder shall not be the sole means of mechanical connection in any circuit. If soldered, the connection shall be so located or supported as to minimize flexing of the conductor where the solder changes the flexible conductor into a solid conductor.
EXCEPTION: Battery lugs with a solder contact length of not less than 1.5 times the diameter of the conductor.
NOTE: When a stranded conductor is soldered, the soldered portion of the conductor becomes a solid strand conductor, and flexing can cause the conductor to break at the end of the solder joint unless adequate additional support is provided.

Hope this helps.
Charlie

[S&S]

Originally Posted by witzgall

S&S, can you elaborate? What do you mean by " At least a cap"? a Capacitor? Why, and how large?

Chris

Curl E = dB/dt. In other words, you'll get a potentially frying voltage spike if you suddenly cut the field current (small t -> Big E). Using a cap will allow the field to decay more slowly (as the cap discharges through the field coils) rather than just collapsing. You'd have to calculate based on the energy stored in your field coils for a minimum size.

That's at a minimum, for a big alt. it makes sense to hire a EE to come up with a way to get the field to decay slowly. On a small alt. it may not be a problem- I can't say for sure.

In my application, I'm not willing to risk a 400A alt. and 1400Ah of batteries on a direct connection.

[jallum]

Originally Posted by S&S

Curl E = dB/dt. In other words, you'll get a potentially frying voltage spike if you suddenly cut the field current (small t -> Big E). Using a cap will allow the field to decay more slowly (as the cap discharges through the field coils) rather than just collapsing. You'd have to calculate based on the energy stored in your field coils for a minimum size.

That's at a minimum, for a big alt. it makes sense to hire a EE to come up with a way to get the field to decay slowly. On a small alt. it may not be a problem- I can't say for sure.

In my application, I'm not willing to risk a 400A alt. and 1400Ah of batteries on a direct connection.

I was thinking about this... maybe this is me being a dumbass, but what about putting the alternator on a magnetic clutch? Instead of cutting the field wire, you could drop out the clutch and let the field decay as the rotor spins down?

[electric1]

Originally Posted by imcbride

I might be missing out on some electronics 101 here, but my understanding is that LFP battery banks are charged in a profile of Constant Current then Constant Voltage (CC->CV). Current up to a threshold voltage (~14.5V on a 12V nominal pack) then push current at that voltage until absorbed current falls below ~.01C. Other BMS shops have said that the charge source needs to be disconnected at that point. This brought up the problem that just opening a solenoid on the circuit from the alternator under load would blow its diodes. Also, on a large bank, say 400Ah, .01C is 4A, your MiniBMS shunting is rated at only .75A?

Yes, LFP is charged with CC/CV profile to get them to 100% SOC, but in most situations its enough to just bulk charge them to the threshold level and stop, especially when charging from gas powered generator, since you would be wasting fuel on finishing charge. The best part of LFP chemistry is they don't care if you never charge them to 100%, in fact, their lifecycle is even higher if you don't. When charging from mains using CC/CV charger, you can charge to whatever level your charge is capable of until BMS kicks in or charger stops on its own. But when charging at sea from generator, all you need is the bulk charge and then stop. You will end up with 90%-95% SOC, but the best charging efficiency since LFP will suck every available amp from your alternator and there is virtually no Puekert effect. At the end of charge current will reduce to almost zero since there won't be any voltage difference. This happens very quickly on LFP cells, they have extremely flat charge/discharge curves with steep slopes on both ends. One minute you are charging at full current, next minute its all done, literally.

The difference here is that Lead Acid battery never stays at 14.5V, so it will always take some finishing current at this voltage, but LFP can go even higher than 14.5V, so when LFP reaches 14.5V level there will be no more current flowing. I don't see how alternator can be damaged if disconnected while there is no current flowing. Yes, you would not want to trigger BMS during bulk charge, but that is not even possible since there is never going to be HVC event at these voltage levels. Contactor is there to prevent overdischarge when battery is low and there is no generator running. I don't see how battery can get low when generator is running unless your house load is higher than generator output.

As a new technology, we don't know how sensitive LFPs are so we're treating them with kid gloves.

I understand that, I was in the same boat (pun intended) when I spent 11K on my EV LFP pack a year ago. Since then I learned how much less sensitive LFP is compared to LA. As long as you stay under 2C currents and within operating voltage window you really can't hurt these cells.

So the user only needs to reset the BMS after a high/low voltage error condition?

Yes, momentary push to reset BMS and start/stop the charger. Again, I don't see HVC happening in your case since you use chargers made for LA batteries and they don't go as high to reach HVC on a LFP batery.

Two additional concerns I have are protection of the BMS PCBs from the rough (corrosive) marine environment, and secondly the method of connection of the BMS wires to the tabs/PCB of the cell chips. ABYC standards have told us not to use spade connectors or soldered connections for anything. Reasoning as I understand it is to avoid disconnects or mechanical wear in the rough, high movement environment of a boat. Any thoughts on these issues?

My BMS modules are spray coated with protective lacquer, but I would advise to spray coat entire top of the battery bank after BMS is installed with CRC Battery Protector or similar clear coating. Head end control board with buttons and contactor can be installed in a small sealed plastic box. QD Push connectors I use are standard for automotive environment and should be sufficient for marine use, they are super tight and have locking holes. I think there is more vibration in a car than a boat. I would spray coat them with CRC as well.