Lithium battery & component suggestions?

[john61ct]

Good luck with that.

Many insurance companies require a survey, and in NA market ABYC sets the standards.

If you read the actual contract you will likely find an escape clause for the insurer, if your Science Experiment is the cause of an incident.

 

[john61ct]

Plus if Tesla does not support your use case you may find problems there.

 

[diver dave]

...


3) Buy bare cells and do my own BMS using an industrial PLC. This is the path I'm taking. All the components are available off the shelf from a plethora of manufacturer who have been around for an eternity. So repair down the road is not an issue. It will require some programming on my part, but PLC programming is pretty simple and understandable by any industrial automation person. And costs are reasonable.


I think I have sorted out the PLC (programmable logic controller) part. For a variety of reasons I have picked IDEC for the core PLC, and have been experimenting with a variety of applications for the boat, one of which is a BMS for LFP batteries.


The first stumbling block was per-cell voltage monitoring. There are about a zillion voltage sensing modules available for PLCs, but all of them want to sense voltages referenced off the same ground. That doesn't work to sense voltage on each of a bunch of series wired cells, or at a minimum you would lose a lot of measurement accuracy doing it that way. The issue is that each cell can be measured accurately with about a 5V measurement range. That part if easy. But each successive cell spans a voltage that's higher and higher relative to the battery pack's ground, which is also the ground for the PLC. So, using simplified numbers, the first battery is 0v to 3V, the second is 3V to 6V, the third is 6V to 9V, etc. The measurement range is around 3V for each battery, but the reference voltage climbs with each successive battery. I have only been able to find one voltage sensing device that will handle this, and am about to test it out in real life to see if it works as expected.


.......

I would find this to be an interesting project. I am working now to develop various A to D systems for industrial monitoring. What you reference above is the need for a front end device to monitor cell voltages. You could do a number of "ground based" A to D's, and take a difference of each voltage to arrive at the cell voltage. OR, use isolated A to D's, so you get each cell voltage directly. The design I am doing now, fyi, has a couple dozen inputs, with each having isolation. So, you can measure a 0 to 5V signal accurately, even with 1000V of common mode voltage sitting on it. The output of each ADC feeds into a FPGA and then into the uP. It could be a PLC instead. CAN is a logical comm standard, there is even the marine based NMEA2000, based on CAN.

EMC is another big consideration. You don't want to blow up cells, or raise false alarms if VHF, or HF/SSB is used. The NMEA2k standard contains well defined immunity tests with levels.

Have you written down a general spec sheet for the BMS?

 

[Delfin]

I actually would NOT ask my insurance company. It just gives them the opportunity to say No, or to increase my rates. Unless there is some prohibition against LFP in your insurance contract, then it's covered. I'll wait for them to exclude them on their own rather than prompt them to do so.

True. I believe that if something isn't in the exclusions section of the policy, it is covered unless the loss is due to negligence on the part of the policy holder. I doubt installing batteries used by a great many boats in a professional manner would qualify as negligence, especially since LiFePO4 is arguably a safer chemistry than lead acid.

 

[twistedtree]

True. I believe that if something isn't in the exclusions section of the policy, it is covered unless the loss is due to negligence on the part of the policy holder. I doubt installing batteries used by a great many boats in a professional manner would qualify as negligence, especially since LiFePO4 is arguably a safer chemistry than lead acid.

I tend to agree. It doesn't mean there won't be an argument over it, but I think as long as I can show design and implementation diligence in managing the batteries according to specs, they would have a hard time with it. I don't think it's any different from working on your own boat. I've seen plenty of "professional" work that is way more dangerous than work I do.


My bigger fear is that insurance companies will put a blanket exclusion in policies for Li, probably without consideration for different chemistries. That would be way more of a risk and problem than maybe having a fight down the road, if maybe my boat gets damaged, and maybe it's directly because of the batteries.

 

[twistedtree]

I would find this to be an interesting project. I am working now to develop various A to D systems for industrial monitoring. What you reference above is the need for a front end device to monitor cell voltages. You could do a number of "ground based" A to D's, and take a difference of each voltage to arrive at the cell voltage. OR, use isolated A to D's, so you get each cell voltage directly. The design I am doing now, fyi, has a couple dozen inputs, with each having isolation. So, you can measure a 0 to 5V signal accurately, even with 1000V of common mode voltage sitting on it. The output of each ADC feeds into a FPGA and then into the uP. It could be a PLC instead. CAN is a logical comm standard, there is even the marine based NMEA2000, based on CAN.

EMC is another big consideration. You don't want to blow up cells, or raise false alarms if VHF, or HF/SSB is used. The NMEA2k standard contains well defined immunity tests with levels.

Have you written down a general spec sheet for the BMS?

Isolated sensors with a large common mode tolerance is what I've opted for. That gives full range sensitivity and accuracy for measuring each cell since I'm measuring a 3-3.5V cell voltage using a 5V sensor range/scale with accuracy of 0.1%, or +/-5mv. If I measured each cell relative to ground I would have to measure with at least a 70V range (it's a 48V system), so accuracy at 0.1% would reduce to +/-70mv. And, errors compound since I'd be taking the difference of two measured readings.


All the control is over Modbus since that's the standard for all this industry automation stuff. Some of it is over direct wired twisted pair (RS-485), and other is Modbus over TCP using ethernet. Both are robust in noisy environments, assuming properly constructed.


Written specs? Ok, busted. I'll do a schematic of the whole thing and document set points and behavior. I'm generally pretty good about that sort of thing because I have burned myself coming back to something that I built years earlier and wondering WTF I did.


BTW, my batteries are supposedly waiting for me at the local USP depot, and I hope to pick them up tomorrow.

 

[JDCAVE]

Lithium battery & component suggestions?

Good luck with that.

Many insurance companies require a survey, and in NA market ABYC sets the standards.

If you read the actual contract you will likely find an escape clause for the insurer, if your Science Experiment is the cause of an incident.

Of course. But it’s all up to the surveyor. Having just gone through this process, surveyors are not equal. My surveyor got stuck on one issue that was up to AYBC specs. He wanted me to consult with Transport Canada. So my fuel sight gauges exceed ABYC specs now and are acceptable (but not approved) by TC.

I also learned that you don’t want to offer up stuff to surveyors or they grab onto it and make you do more than necessary. Other things he just plain missed.

I’ve seen TT’s impressive installations first hand, and I can assure you that his background on these matters is considerable and his attention to details and troubleshooting is extensive.

Jim

 

[john61ct]

True. I believe that if something isn't in the exclusions section of the policy, it is covered unless the loss is due to negligence on the part of the policy holder. I doubt installing batteries used by a great many boats in a professional manner would qualify as negligence, especially since LiFePO4 is arguably a safer chemistry than lead acid.

By professionals is different from DIY.

DIY may or may not be done "in a professional manner", what does that even mean?

Yes if your insurer doesn't require a survey, and your system is not the cause of any problem, you're probably fine.

In any case, the immediate above discussion was not about proper prismatic LFP cells at all, which if not "common" are at least very safe and well-understood.

Some were talking about jury-rigging secondhand cells, 18650's, scrapped EV packs, home Powerwall systems etc.

Chalk and cheese, completely different scenarios

 

[john61ct]

I’ve seen TT’s impressive installations first hand, and I can assure you that his background on these matters is considerable and his attention to details and troubleshooting is extensive.

Nothing I say is wrt any one individual.

Of course there are exceptions that prove the rules.

The target for my statements is relatively newbie readers coming across the thread while doing initial research.

Those with solid engineering level skills and a high tolerance for risk already know what they can afford to ignore.

 

[john61ct]

I completely agree that you can trust yourself to regulate the charge profile yourself manually. I was just stating that I cannot.

If I get distracted, the Balmar profile programmed matches the recommendations from Lithionics.

Which is much higher than what those striving for longevity would use.

> I am not sure your statement that SoC is an "arbitrary call" and not objective makes a whole lot of sense. A battery will accept additional charging current, or it won't. If it won't, by definition, it is at or near 100% SoC.

No. Keep increasing voltage and more current will be accepted.

Current falls, and at some fraction of C rate you stop, that is Full.

Waiting for **all** current to stop is harming your bank, even at 3.45Vpc.

As you say vendor definition of Full is 14.4V / 3.60Vpc, Absorb until .02C endAmps.

There is **no rational reason** to push through the shoulder that far.

Why would you only want a fraction of the potential cycle lifetime for all the money you're spending?

 

[diver dave]

Isolated sensors with a large common mode tolerance is what I've opted for. That gives full range sensitivity and accuracy for measuring each cell since I'm measuring a 3-3.5V cell voltage using a 5V sensor range/scale with accuracy of 0.1%, or +/-5mv. If I measured each cell relative to ground I would have to measure with at least a 70V range (it's a 48V system), so accuracy at 0.1% would reduce to +/-70mv. And, errors compound since I'd be taking the difference of two measured readings.


All the control is over Modbus since that's the standard for all this industry automation stuff. Some of it is over direct wired twisted pair (RS-485), and other is Modbus over TCP using ethernet. Both are robust in noisy environments, assuming properly constructed.


Written specs? Ok, busted. I'll do a schematic of the whole thing and document set points and behavior. I'm generally pretty good about that sort of thing because I have burned myself coming back to something that I built years earlier and wondering WTF I did.


BTW, my batteries are supposedly waiting for me at the local USP depot, and I hope to pick them up tomorrow.

Modbus over STP should be fine. Interestingly, I've been running experiments with Modbus vs noise vs baud. My last experiment was in a very high 60Hz field environment, and showed quite a bit of Modbus frame errors, diminishing when I reduced the rate down to 4800 (from 119,2). BUT, this was on a 345kV air system; your mileage will definitely vary for the better in a boat or factory! This was a place where you received a constant shock just touching any grounded surface. A constant shock, not ESD.
My cabling was indiv. shielded TP with overall shield, but only attached on one end. More work to follow there.
Now, I am chasing the viability of CAN, which is essentially the same physical layer, with built-in error correction, vs the primitive Modbus protocol. Anyway, I am not optimistic on CAN even at 250kb, much less the latest 1Mb flavor. Unless I go with fiber. Ok, i'm now rambling.
Yes, you really should do a block diagram, then schematic, data flow diagram, and algo strategy too. But, I suspect you already know this. Is there any plan for commercialization? or just a one off?

 

[MurrayM]

...Those with solid engineering level skills and a high tolerance for risk already know what they can afford to ignore.

Which is why I'll probably go with carbon foam. That, and most components, if needed, should be available in remote locations...we can't afford airfreight.

 

[Delfin]

I completely agree that you can trust yourself to regulate the charge profile yourself manually. I was just stating that I cannot.


Which is much higher than what those striving for longevity would use.



Why would you only want a fraction of the potential cycle lifetime for all the money you're spending?

You're assuming you know more about this chemistry than CMS, whose advice I follow closely, or the U.S. manufacturer, whose measures and recommendations match CMS's. You've yet to demonstrate such expertise.

 

[john61ct]

You're assuming you know more about this chemistry than CMS, whose advice I follow closely, or the U.S. manufacturer, whose measures and recommendations match CMS's. You've yet to demonstrate such expertise.

That's funny, I would be very interested to be informed where anything I've written conflicts with CMS / Maine Sail. I usually get accused of following his reco's too slavishly

I've certainly never read him recommend any charging voltage higher than 3.5Vpc

This thread is long but informative
http://www.cruisersforum.com/forums...or-those-using-them-as-house-banks-65069.html

besides Maine Sail, Bruce @ Ocean Planet knows his stuff, but comes a bit more from a vendor POV.

Also MS' summary notes here
https://marinehowto.com/lifepo4-batteries-on-boats/

 

[john61ct]

And what US manufacturer are you talking about?

wrt charging voltages I have yet to see any one worldwide specify as low as what CMS recommends.

My (conspiracy) theory is that manufacturers would prefer their cells get used up in less than 10 years.

 

[stubones99]

I would be curious to see how Iron Edison batteries stack up against Trojans and similar batteries. Except for their size and requiring a forklift to manage them, it seems like a decent battery with long life, ideal for off-the-grid home use. Not so good for boats though.

https://ironedison.com/48-volt-lithium-battery-for-solar

It looks like SimpliPHI has the same technology in a smaller size...

http://simpliphipower.com/

 

[twistedtree]

Modbus over STP should be fine. Interestingly, I've been running experiments with Modbus vs noise vs baud. My last experiment was in a very high 60Hz field environment, and showed quite a bit of Modbus frame errors, diminishing when I reduced the rate down to 4800 (from 119,2). BUT, this was on a 345kV air system; your mileage will definitely vary for the better in a boat or factory! This was a place where you received a constant shock just touching any grounded surface. A constant shock, not ESD.
My cabling was indiv. shielded TP with overall shield, but only attached on one end. More work to follow there.
Now, I am chasing the viability of CAN, which is essentially the same physical layer, with built-in error correction, vs the primitive Modbus protocol. Anyway, I am not optimistic on CAN even at 250kb, much less the latest 1Mb flavor. Unless I go with fiber. Ok, i'm now rambling.
Yes, you really should do a block diagram, then schematic, data flow diagram, and algo strategy too. But, I suspect you already know this. Is there any plan for commercialization? or just a one off?

I'm operating at a significantly lower voltage


No plans to commercialize - I stopped doing that 10 years ago. My how time flies.

 

[Delfin]

That's funny, I would be very interested to be informed where anything I've written conflicts with CMS / Maine Sail. I usually get accused of following his reco's too slavishly

I've certainly never read him recommend any charging voltage higher than 3.5Vpc

This thread is long but informative
LiFePO4 Batteries: Discussion Thread for Those Using Them as House Banks - Cruisers & Sailing Forums

besides Maine Sail, Bruce @ Ocean Planet knows his stuff, but comes a bit more from a vendor POV.

Also MS' summary notes here
https://marinehowto.com/lifepo4-batteries-on-boats/

Perhaps read through CMS' notes to answer your question.

 

[Delfin]

And what US manufacturer are you talking about?

wrt charging voltages I have yet to see any one worldwide specify as low as what CMS recommends.

My (conspiracy) theory is that manufacturers would prefer their cells get used up in less than 10 years.

Lithionics purchases cells from Taiwan, then matches them and builds the box, provides the warranty and the BMS. To say that they aren't a U.S. manufacturer is to say that Apple isn't one. They provide batteries for the U.S. military who aren't great at everything, but are pretty good at verifying quality control from contractors.

 

[john61ct]

I thought you meant cell manufacturers.

Quote right from that page:

If your cells rest at room temperature for 3+ hours with no load and wind up at 3.4VPC they are full. Getting them full does require a voltage higher than 3.4VPC in order to get them to 3.4VPC rested, but not much higher. Charge to 13.8V or 14.0V and allow the current to tail off to about .025C (2.5A on a 100Ah cell) and your bank is now about as full as you’ll need it to be.

I charge this 400Ah Winston pack to 13.8V and 0.025C current and it still delivers 425Ah’s at a .25C load! This has now been tested and repeated twelve times in complete 100% discharge capacity test to confirm the Ah capacity.

 

[twistedtree]

Quick update. Today I uncrated 32 CALB LFP batteries, 180ah each. And so begins the experiment....

 

[BruceK]

I just read that the supply of lithium has increased significantly,going from "famine to feast". Hopefully that should impact battery prices.

 

[twistedtree]

I just read that the supply of lithium has increased significantly,going from "famine to feast". Hopefully that should impact battery prices.

That would be nice, but of course just after I bought.....

 

[BruceK]

That would be nice, but of course just after I bought.....

Sorry,not when you want to hear it.

A purer form, lithium carbonate, is used medically as a "mood stabilizer" for bipolar etc,any uptick in demand there might balance the supply. Lots of steps between a raw material getting cheaper/more plentiful, and the price of a finished product. Though titanium became much more available when other materials replaced it for aerospace, turning up in watches, high end bicycles, etc

 

[Delfin]

Quick update. Today I uncrated 32 CALB LFP batteries, 180ah each. And so begins the experiment....

With your expertise, it won't be an experiment....

Is matching the cells for identical impedance an issue? Or did the supplier do that already? I was told (I think) that this is something purchasers of large numbers of cells do before combining them into battery packs.

 

[john61ct]

The demand for LFP will keep prices high.

The supply of the metal lithium itself is not an important factor in the pricing of batteries.

Certainly nothing to do with the medication.

Quick update. Today I uncrated 32 CALB LFP batteries, 180ah each. And so begins the experiment....

Christmas!

Link to your vendor?

 

[sunchaser]

TF is fortunate to have TT report on his findings. Seldom will you find in the boating press or online sources an individual who has the budget, business and proven technical background to jump into the middle of today’s technological Marine offerings as it applies to instrumentation and energy arenas.

Adulation aside for a moment TT, will your new build be initially set up to minimize generator run time by utilizing a Li package along the lines of this thread? If so the need for a solar array comes into question whether initially or as an after addition.

 

[Wil]

TT, would you give further details on the purchase of your batteries please?

Who did you buy them from?
Warranty?
Price?
How were they shipped?
Is this repeatable (others can buy from the same entity)?

Thanks, this is hopefully good news for others too.

Quick update. Today I uncrated 32 CALB LFP batteries, 180ah each. And so begins the experiment....

 

[twistedtree]

With your expertise, it won't be an experiment....

Is matching the cells for identical impedance an issue? Or did the supplier do that already? I was told (I think) that this is something purchasers of large numbers of cells do before combining them into battery packs.

I wasn't expecting it, but the batteries came with a sheet characterizing each cell. Their impedences are well matched, and all have tested capacity over 200ah. And they are all sequential serial numbers.

 

[twistedtree]

TT, would you give further details on the purchase of your batteries please?

Who did you buy them from?
Warranty?
Price?
How were they shipped?
Is this repeatable (others can buy from the same entity)?

Thanks, this is hopefully good news for others too.

I bought them from Electric Car Parts Company in Utah. Warranty I think is 1 year, but I assume useless since a bad cell would have to be shipped back to CALB, presumably their China office. So I'm assuming no warranty or support. Price was about $250 per battery including shipping, jumper straps, Credit card fees, etc. They were shipped UPS freight and I picked them up at the local UPS terminal. That made the shipping equal to shipping to a commercial address and avoided the lift gate fee. Plus it was logistically easier for me.


They were well packed in a wood crate, but the shippers still managed to puncture the crate and knock off one of the skids with a fork. BTW, I have had a 100% failure rate with truck shipping. The product hasn't always been damaged, but 100% of the time there is structural damage to the shipping container, and about 50% of the time product damage. FedEx freight is the worst of them all and I won't buy anything anymore that ships via them. I opened the crate at the UPS terminal to inspect the cells and they were fine, so I accepted the delivery despite the forklift hole.


Yes, anyone can buy these, though availability comes and goes, and the selling "dealers" are tiny one man shops. It's very, very far from a mature supply chain.