Lithium battery system thoughts

[joe_db]

Do you need to clamp these together to keep them from bulging and getting ruined?

[lat 64]

Hi all.

Just this year I put together an e-bike with lithium-ion battery and also got a camper with AGM house battery,
What an education!

Everything said here was discovered by myself as I got up to speed on these technologies.
upshot for me is:

• Hi-tech lightweight batteries for fun on the bike,
• Heavy tolerant systems for the camper and boat.

A telling consideration is found in this ironic question:
Why would one care about weight savings in a boat with 5,000 lbs of lead in the keel?

I agree with Neil; gotta keep these systems from running our lives, but we do need to keep the beer cold.

Cheers,

Russ

[edwardc]

Most commercially available lithium batteries can be connected in parallel, I don't see any reason you can't make two batteries with two BMSs. Speaking of, are you SURE you have a good source for cells? I looked into that and it seems a crapshoot at best, apparently almost all cells are rejects that Renogy and the rest of the battery companies didn't want.

SURE? No. Fairly confident, yes.

I’ve spent the last year reading all the reviews (and I mean real technical reviews, not “fanboy” stuff) particularly those who have bought large quantities of cells from multiple vendors. With enough time and data points, some trends started to appear. I did this because about a year ago, I tried to order some cheap cells for testing, direct from China. After waiting for them for 8 weeks, I had the vendor put a trace on them, and the shipping company reported them as “lost”. (But I did get a quick full refund).

At the present time, the Chinese company “Doucan”, while not the cheapest, seems to have a reputation for delivering quality product. Their sales rep Jenny Wu in particular, has a rep for never lying and delivering what was promised. And best of all, they can deliver LiFePO4 cells out of their warehouse in Houston!

So, on Jan 2 I took the plunge and ordered eight 280ah cells from them.
FIVE DAYS LATER, they were sitting on my living room floor!! They arrived well packaged in thick foam liners, two to a box, sealed up watertight.

Initial inspection all looked good, every cell showing no signs of terminal scratching or bulging, both of which are signs of previous use. Every cell had the OEM QR code laser etched on it. I decoded a mfg date of Jun 2021. A more detailed evaluation and measurement is in progress. Wish me luck!

[joe_db]

Hi all.

Just this year I put together an e-bike with lithium-ion battery and also got a camper with AGM house battery,
What an education!

Everything said here was discovered by myself as I got up to speed on these technologies.
upshot for me is:

• Hi-tech lightweight batteries for fun on the bike,
• Heavy tolerant systems for the camper and boat.

A telling consideration is found in this ironic question:
Why would one care about weight savings in a boat with 5,000 lbs of lead in the keel?

I agree with Neil; gotta keep these systems from running our lives, but we do need to keep the beer cold.

Cheers,

Russ

Weight savings is not why I would get lithium batteries, the vastly longer lifespan would be.

[edwardc]

Weight savings is not why I would get lithium batteries, the vastly longer lifespan would be.

That, and greater total energy storage in the same volume.

[joe_db]

Ed - are you using contactors or the BMS itself for switching?

[edwardc]

I had been leaning towards the REC Active BMS, which is set up for driving contactors, but nothing is cast in stone yet.

Other contenders were the Orion, and Stuart Pittaway's DIY BMS project.

The thing that gives me pause with the REC Active is that it wants to be both the BMS and the central charge controller. Seems like too many eggs in one basket. I like the redundancy of a federated architecture, where each charge source is responsible for its own regulation.

[joe_db]

Any progress reports?
My batteries are shot, so I need to do SOMETHING real soon. I can physically fit a 300ah battery, so that would be say about 275ah or so of usable capacity at something like 3,000 cycles. The same package gets me 200ah of lead-acid capacity, at only 100ah usable for long life and probably 200-400 cycles if I use about 150 out of the 200.

[msmith10]

The only progress I've made is deciding to stay with deep cycle lead acid for now. Lithiums clearly aren't ready for the uneducated masses, of which I am a member. I'm not ready to fuss with all the details at this point.

[edwardc]

Thanks for the nudge.

To refresh, I have eight EVE 280AH LiFePO4 prismatic cells.

I ended up ordering the Rec Active BMS. The manufacturer was very quick to reply to my questions, and I now understand how to configure the BMS to operate as part of a distributed system.

In addition to the BMS, I ordered a pair of 500A contactors, one for the charge bus and one for the load bus, and the WiFi adapter so I could configure and monitor the BMS without needing the Windows-only software package (I'm a Mac guy).


Initially, I strapped all the LiFePO4 cells in parallel, and hooked them to a regulated supply. I brought them up to full charge in steps. For instance, I initially set the supply to 3.0V and limit it to 4.5 amps (Its only a 5A supply), and let this sit until the current dropped off to ~100 mA. (this took a LONG time (days) due to the MASSIVE AH capacity of eight paralleled cells, 2240AH!). Then I would bump it up a few tenths of a volt and repeat. Each cycle took less time. By the time I was pushing all the cells up into their fully-charged state, the last cycle took just over an hour. This process "top balances" all the cells.

I next rewired all the cells in a 2P4S (2 parallel 4 series) configuration to create a 12V 560AH battery pack. I wired up the bms sense leads to each of the 4 series cell-groups, connected the charge and load contactors to the appropriate drive leads from the BMS. The charge bus was connected to my supply (now setup for the correct charge voltage) and the load bus was connected to a 1000W inverter. The inverter was connected to a small heater and a collection of incandescent light bulbs to bring the total up to around 900W. This would produce a DC load current of around 75A.

I since setting all this up had brought the battery pack down below full charge, I first tested the high limit charge cutoff. Turned on the supply and let it charge. As soon as the pack reached the charge max voltage that I had set, the Charge contactor shut down as designed. Keep in mind that this is NOT the normal charge voltage, but the "Never Exceed" voltage for the pack. In normal operation, this should never happen. Its just a protection to keep a failure elsewhere from overcharging and destroying your batteries. Meanwhile, the load contactor remained connected, which would allow you to drain the pack down to a safer level.

Next test was a crude capacity test. With the supply turned off, I activated the inverter. The idea was to see where the pack capacity was (according to the BMS monitor) when the load contactor cut off. This test was a reasonable success. It ran about 7 1/2 hrs. The problem was that the inverter shut down due to low voltage before the pack was down to what the BMS considered the 0% State-of-Charge for the pack. This inverter cutout happened within a few Amp-Hours of of the spec value of 560AH, so I considered the capacity test to be a success. The BMS monitor showed all the cells individual voltages stayed within the proper amount, so it looks like the cells I received are reasonably matched.

After a few minutes of sitting, the pack voltage "rebounds" some. This allowed me to temporarily reconfigure the low voltage cutout to a higher voltage than the point where the inverter cuts out. Turned the inverter back on, and this worked perfectly, shutting down the Load contactor, but leaving the Charge contactor on.

So far, I'm satisfied with my purchases. Still need to configure a relay to cutoff the Balmar alternator controller in case of a BMS shutdown of the charge contactor and test that for all cases.

After that, the next step is to start creating a new battery box area in the boat. The old battery location is effectively in the engine compartment, and gets too hot for the LiFePO4's max operating temp of 140 deg F.

[sastanley]

Thanks for the nudge.

To refresh, I have eight EVE 280AH LiFePO4 prismatic cells.

Ed, what is the normal voltage of each cell? I've seen some 100Ah 'lithium' batteries popping up on web advertisements that look like normal 12v batteries, like Dakota and Renogy, and I think are designed as 'drop in', from a form factor anyway, replacements for 12v lead acid. But I doubt that is the case. Trying to begin getting educated on the subject.

[joe_db]

Ed, what is the normal voltage of each cell? I've seen some 100Ah 'lithium' batteries popping up on web advertisements that look like normal 12v batteries, like Dakota and Renogy, and I think are designed as 'drop in', from a form factor anyway, replacements for 12v lead acid. But I doubt that is the case. Trying to begin getting educated on the subject.

*There is no such thing* as "drop in" lithium batteries no matter what the ad says. The lithium batteries do work in 12 volt systems, so in a sense you could just install them and see what happens. The most likely result of that would be ruined batteries, a burned up alternator, or both. They have some unique issues you need to deal with for a successful installation. I'll post my idea for a lower budget install soon, but the first thing to remember is the BMS (battery management system) does NOT actually manage the batteries the way you might think it does. It takes care of cell balancing, but other than that the charge and discharge disconnects are emergency last-ditch ways to save the battery. It does not do anything like manage charging voltages for a correct charge.

[joe_db]

This is my idea of a lower cost system. The Victron DC-DC charger is configurable to charge lithium batteries correctly.

[edwardc]

Ed, what is the normal voltage of each cell? I've seen some 100Ah 'lithium' batteries popping up on web advertisements that look like normal 12v batteries, like Dakota and Renogy, and I think are designed as 'drop in', from a form factor anyway, replacements for 12v lead acid. But I doubt that is the case. Trying to begin getting educated on the subject.

Shawn,

The LiFePO4 chemistry has a nominal cell voltage of 3.2v per cell. Thus, a pack of 4 series cells has a nominal voltage of 12.8V

But this is slightly misleading. The cells have an almost FLAT charge/discharge curve (see attachment), pushing up to a max of 3.65v/cell right at the end of safe charging, and dropping down to 2.5v/cell right at 0% State-of-Charge. This corresponds to pack voltage limits of 14.6V and 10V respectively, which is close enough to typical lead-acid limits.

Keep in mind that these are the "never exceed" limits, and normal pack operation will not see them unless there is a failure somewhere. For example, I will set my chargers limits around 13.2V (3.3V per cell). Because of the "S" shaped charge/discharge curve, I'm only sacrificing a few percent of theoretical storage capacity at the top, but research has shown that this extends the cell's cycle life. More on LiFePO4 charging in a later post.

So, from a purely voltage perspective, the packs look like "drop-in replacements" for lead acid. But that's before you consider the possible actions of the Battery Management System (BMS). At any time, for one of many reasons (over voltage, undervoltage, over temp, under temp, over current), the BMS can abruptly disconnect the battery from the rest of the system in order to protect it. This can cause havoc with the rest of the electrical system unless there are features specifically designed in to handle it. For example, if your alternator is charging at full capacity and the BMS disconnects, its just like flipping the main battery switch off while the engine's running. The output of the alt will try to spike up to infinity, blowing the alt's diodes and potentially every bit of (expensive) electronics that's turned on at the time.

This is the biggest problem with these "drop-in" packs. They have a BMS that's sealed up inside the case, with no way to communicate with the outside world. The more expensive packs, such as Victron, have outputs, such as relays or CANBus, that can be used to signal external chargers of a disconnect so that they can be gracefully shut down.

Another problem with these inexpensive packs is that they use a BMS that uses inexpensive MOSFET switches to disconnect the pack. This often limits them to relatively small charge and discharge currents (50A), which negates one of the advantages of LiFEPO4 batteries.

So, in order to successfully use LiFePO4 in a boat, you either have to design a system using raw cells and an external BMS that can disconnect the chargers when necessary, buy an expensive single-vendor fully integrated system from a company (such as Victron or Mastervolt) that understands and handles these issues, or design a system (such as Joe's low-cost example) that never allows the alternator to become un-loaded while in operation.

Since I have the necessary electronics knowledge, and am an incurable DIY'er, I chose the first approach. YMMV

[joe_db]

One more reason I am looking at the DC-DC charger is a big lithium bank can look like a dead short and be very hard on alternators.