LiFePO4 Batteries - Okay Tear Me Apart ;-)

[S&S]

Originally Posted by DotDun

How does the LiFePO4 hold up to heat? Batteries in tropical climates succumb to heat before death from cycles, at least in my experience.

As long as it's under 200*C they should be fine.

[S&S]

Originally Posted by Boracay

I seem to remember something about Lithium batteries sometimes catching fire.

Web reference here.

I do accept that there are lots and lots of sources stating that the new batteries don't burn, but we are talking about new technology, very high currents and almost continual charging.

Charging lead acid batteries still isn't perfect after how many years?

There is an interesting Wikipedia entry here.

Different chemistry. Li ion (cobalt) batteries can suffer thermal runaway and go boom. they do, hovever have significntly higher energy density.

My biggest concern with LiFe batts. was slagging the alternator.

[ardoin]

All the BMS systems are designed for bikes, cars, drills, and toys. There is no BMS for sale that will deal with the complexity of boats (or none that cost less than 18K to 30K). It shouldn't be hard to build a BMS for a boat since boats do not put the stress on these batts that they are designed to handle.
But boats have some unique features that current BMS's don't address. E.g., boats have two batt banks and there are some special safety rules for these banks (never want the engine batt to go dead). Boats also have manual methods to join the batt banks (horrible for LifePO4). And many boats have charge relays that will transfer current from one bank to another as needed.
I'm looking into this carefully. Cell-based batt mgnt looks like it is simple, the difficulty is the integration of the cell-based mgnt with the overall boat charging systems -- how to control the two Balmars, solar panels, wind gen, batt charger, and charge relays in such a way that you won't fry a cell or leave a batt bank dead. while at the same time not requiring the skipper to constantly monitor a complex system. Unfortunately, LifePO4s cannot be treated like normal Pb-batts in the complex charging environment of a boat. But the advantage boats have is that the BMS can be simple in some ways since the a boat does not have a giant electric motor (car) sucking the batt dry in a few seconds (unless you are arc welding on your boat).

Another advantage that has not been stressed is that these batts can provide more Amp Hrs per cycle than others and this will reduce the number of charge cycles.

Still digging into the BMS side of this problem... hmmm maybe there is a market for a reasonably priced BMS for cruising boats, I might need help in the design

[S&S]

Ardoin, you may be overthinking this. All a BMS does really is ensure that the cells remain in balance. There's no "special design" for boats. It really doesn't matter what the charging source is as long as the applied voltage is within the bank parameters. An overvoltage and undervoltage cutout is no different regardless of the application. The only time you need active control of a charging source is with a rotating machine- else a simple voltage activated relay is sufficient. Regulation is simpler too, these like constant voltage so step regulators are not necessary.
The simplest solution is to have two alternators and two banks. I looked into this and cross charging is pretty complex. At the end of the day seperate banks with an interlock to be able to spin the starter off either bank is probably enough.

[fastcat435]

There is a difference between a BMS for bikes and many other small appliances and for yachts.
We work with Mastervolt Lithium Ion ( LifPo ) batteries and the bms integrated in each battery is built to last in the humid salty marine environment , contrary to what you think cross charging is not very complex, we use dc cross chargers in between the batteries and this all is done automatically. we also us dc to dc converters to cross charge from 152 nominal to 25.2 volt nominal cross charging from the propulsion units to the house units. We are extremely happy with the used LifPo units and have been using these special types over 2 years now. ( yes they are expensive in initial layout however keeping in kind that they last in excess of 10 years and the fact that on our specific Catamaran we save over 1150 kilo ( 14 % of the boats loaded weight ) it helps to make the cat faster with less wind and saves us energy.

I have attached a few pictures of our testing setup where 6 x 25.4 volts Mastervolt units are in series creating 152 volts needed for our Green Motion propulsion system as you can see cross chargers are on top of these units and we have tortured these units to see what would happen , discharged one unit complete , charged the lot with in one hour , discharged many times with in one hour ( 160 amp H discharge ), these tests have been performed for a long period to see what the effect would be after many discharges and charges and all we could see is in the beginning that they became even more powerful.

In our case not only the individual cells are balanced but also the batteries are balanced , the positive effect is that we get more useful power out of the complete setup since there are no peak or low voltage cells in the setup.
Gideon

[jallum]

Originally Posted by fastcat435

I have attached a few pictures of our testing setup where 6 x 25.4 volts Mastervolt units are in series creating 152 volts needed for our Green Motion propulsion system

Respect. Someone has serious toy budget!

Yesterday, I took the plunge and ordered 8 of the 300Ah ThunderSky batteries I linked to at the top of the thread. I want 16, so I'll have to wait for the rest... but 8 is enough for 24v, so I can play in the meantime!

My boat is 24v for the house-bank, and 12v for the starter bank. Rather than muck about with all of the cross-charging kit, I thought it would be simpler (in my case) to leave the stock 12V alternator dedicated to charging the starter bank, and have a much larger 24v alternator dedicated to the house bank. The two systems don't really need to be hooked together, except perhaps in an emergency. My modest 12v house-needs come from a single 24-12v DC-DC converter, for clean, clean power.

For those of us not using these batteries for propulsion, I think it's a much more forgiving situation, and we might be over-thinking things a little.

I'm thinking of a simple, per-cell BMS that signals at high and low voltage. At the high set-point, a simple circuit could taper off the field on the alternator to set it down easy, and the inverter/charger is similarly easy to manipulate. Low set-point would open the circuit for protection. I think I'll pot the BMS boards in a compatible epoxy to "marinize" them and keep a couple extra for spares. Pretty simple, pretty cheap.

If a better management system rears it's head, then I'll switch and not look back.

[fastcat435]

Hallo Jallum you better get rid of your 12 volt alternator and use a dc to dc converter to charge the starting battery, it is very inefficient to leave the 12 Volt Alternator on , at best it is 50 % efficient but if your batteries are full it is still costing you in excess of 1 hp and creates extra heat and noise, just try to find a suitable 24 volt alternator to put in place of the 12 volt unit and you will be better of .
What make engine is it ? Do not epoxy the bms in they generate some heat that needs dissipating just spray them with a moist preventing spray so they can cool , epoxy is high in insulation value so it will make the bms hot.

[S&S]

Originally Posted by fastcat435

I have attached a few pictures of our testing setup where 6 x 25.4 volts Mastervolt units are in series creating 152 volts needed for our Green Motion propulsion system as you can see cross chargers are on top of these units and we have tortured these units to see what would happen , discharged one unit complete , charged the lot with in one hour , discharged many times with in one hour ( 160 amp H discharge ), these tests have been performed for a long period to see what the effect would be after many discharges and charges and all we could see is in the beginning that they became even more powerful.

In our case not only the individual cells are balanced but also the batteries are balanced , the positive effect is that we get more useful power out of the complete setup since there are no peak or low voltage cells in the setup.
Gideon

Nice! The rig I was mentioning earlier was ganging alternators to charge multiple banks- too many parts for me if I'm mixing chemistries. Cross charging if they're all LiFe banks is pretty straightforward. Functionally, they all balance the same way-It's a matter of casing to keep the salt out.

Originally Posted by fastcat435

Hallo Jallum you better get rid of your 12 volt alternator and use a dc to dc converter to charge the starting battery, it is very inefficient to leave the 12 Volt Alternator on , at best it is 50 % efficient but if your batteries are full it is still costing you in excess of 1 hp and creates extra heat and noise, just try to find a suitable 24 volt alternator to put in place of the 12 volt unit and you will be better of .
What make engine is it ? Do not epoxy the bms in they generate some heat that needs dissipating just spray them with a moist preventing spray so they can cool , epoxy is high in insulation value so it will make the bms hot.

The manufacturers I talked to recommend a conformal coating (not epoxy) for potting these. I agree a thick coating will retain too much heat. I'd recommend some sort of plastic cover over to top of the cells and the BMS- that should help a lot.

[jallum]

Originally Posted by fastcat435

Hallo Jallum you better get rid of your 12 volt alternator and use a dc to dc converter to charge the starting battery, it is very inefficient to leave the 12 Volt Alternator on , at best it is 50 % efficient but if your batteries are full it is still costing you in excess of 1 hp and creates extra heat and noise, just try to find a suitable 24 volt alternator to put in place of the 12 volt unit and you will be better of .

Superior reasoning rules the day. I'll look into that DC-DC converter. The heat concerns me more than a little lost hp, since the boat's engine compartment is kind of tight... and who likes noise?

I'm not really planning to rely on the 24v alternator (160A), more just to top things off when the engine happens to be running. For primary charging, I'm looking into a high-amp DC genset to replace the 8kW Westerbeke AC generator I have now, or perhaps... well, perhaps that's another thread.

What make engine is it ?

It's a brand new Yanmar 4JH4-HTE (110hp/turbo), pushing a 57' Swan. It replaced an ancient (and abused) Perkins 4.236 (88hp/naturally-aspirated).

Do not epoxy the bms in they generate some heat that needs dissipating just spray them with a moist preventing spray so they can cool , epoxy is high in insulation value so it will make the bms hot.

Quite right. Heat can be an issue. The epoxy I'm referring to is thermally conductive, and is designed for potting electronics -- "conformal coating", though a tad less specific, is perhaps a better choice of words. The modules I'm looking at are of the purely monitoring-type (not shunt-type), so don't generate much in the way of heat.

[fastcat435]

For your DC to DC converter look at Victron energy type Orion 24/12-30 we use these and they work fine
http://www.victronenergy.com/upload/...4%20-%20EN.pdf

It is better to look at a BMS that not only monitor but also equalize since you will have more power available and the batteries wil last longer

Greetings

[S&S]

Originally Posted by fastcat435

It is better to look at a BMS that not only monitor but also equalize since you will have more power available and the batteries wil last longer

Greetings

If you can't equalize the cells you're asking for trouble. Look into shunts.

[jallum]

Originally Posted by S&S

If you can't equalize the cells you're asking for trouble. Look into shunts.

Why is top-equalization with shunts so important? (Can you point me to some literature?)

Wouldn't it be better to "bottom-equalize" the bank by draining the cells individually to the same set-point, and charging them together? Wouldn't it then be better to monitor each cell's voltage, and just cut off charging when the least-performant cell reaches capacity? For sure, this would limit the capacity of the bank to N times that of the weakest cell... but if all of the cells are pretty evenly matched, it wouldn't be much of a loss.

It seems to me, that in order to push in that last bit of current into your highest-performant cell, all of the shunts on the other cells are going to be producing a lot of heat... and if pushed too hard, couldn't they become an ignition point for a fire? If the cells are pretty evenly matched, then you wouldn't stress the shunts terribly, but one could also argue that it's because they're not really doing much for you.

[ardoin]

Over heating the shunts is why I think you will need to control the charging sources from the BMS. If you want to get a max saturation into the batts, you will have to balance the cells until the all get up to power, but just shunting to shed power would be a fire risks (or serious heat source) when you are pumping 100+ Amps into them.
So cell management with some feed back to control the charging sources appears to be the best way to charge them properly. Just relying on the voltage of the entire bank to regulate the source could cause a cell to overcharge or the shunt to overheat. But then again, at the low Amps a typical yacht will draw, what is the chance that the cells will be that far out of balance after a 70% or 80% draw?
I like the concept of dc-dc for cross charging between banks. I currently use a dc-dc system to power many of the electronics with a constant voltage.

[S&S]

Originally Posted by jallum

Why is top-equalization with shunts so important? (Can you point me to some literature?)

Wouldn't it be better to "bottom-equalize" the bank by draining the cells individually to the same set-point, and charging them together? Wouldn't it then be better to monitor each cell's voltage, and just cut off charging when the least-performant cell reaches capacity? For sure, this would limit the capacity of the bank to N times that of the weakest cell... but if all of the cells are pretty evenly matched, it wouldn't be much of a loss.

It seems to me, that in order to push in that last bit of current into your highest-performant cell, all of the shunts on the other cells are going to be producing a lot of heat... and if pushed too hard, couldn't they become an ignition point for a fire? If the cells are pretty evenly matched, then you wouldn't stress the shunts terribly, but one could also argue that it's because they're not really doing much for you.

As ardoin says you need to drop input current until the cells equalize. You can't be sure that all cells will (at least initially) accept at the same rate or come to the same voltage with a given input. Although, according to a conversation with the rep, the balancing will be at smaller currents as the bank gets worked in, at least initially it'll take some time for the top cell to bleed off enough to come into equalization with the other cells. Trickle charging (~1A) is what's been recommended to me.
Pumping high current into an unbalanced bank will overload the shunts or cause the top cell to go over the max voltage, damaging it.
That's why the BMS is rigged to interrupt the charger if it senses a cell imbalance. (usually a relay) Putting a relay in line with an alternator output, however, can have the usual consequences if it opens under load.

[jallum]

Originally Posted by S&S

As ardoin says you need to drop input current until the cells equalize. You can't be sure that all cells will (at least initially) accept at the same rate or come to the same voltage with a given input. Although, according to a conversation with the rep, the balancing will be at smaller currents as the bank gets worked in, at least initially it'll take some time for the top cell to bleed off enough to come into equalization with the other cells. Trickle charging (~1A) is what's been recommended to me.
Pumping high current into an unbalanced bank will overload the shunts or cause the top cell to go over the max voltage, damaging it.

Fair enough. Which size/type did you pick up? Where did you source them from?

That's why the BMS is rigged to interrupt the charger if it senses a cell imbalance. (usually a relay) Putting a relay in line with an alternator output, however, can have the usual consequences if it opens under load.

Of course. That's why the Genasun product has you route the alternator's field wire through the controller board... attenuating or shutting off the field before or at the same time that it throws the contactor to break the charging circuit. That's a big difference between their's and Elithion's BMS offering, it seems.