LiFePO4 Batteries: Discussion Thread for Those Using Them as House Banks

[T1 Terry]

This is going to be way off topic but I'll try and tie it back in at the end.

I know nothing of either of these sellers but to me the second one looks a better choice. the high charge voltage would be the most concern to me. I doubt if that BMS looks at the voltage of all 20 cells so at 3.65v average across all 20 cells there is a very good chance of at least one cell going over 4v and being held there for some time, this stuffs the cells completely. a safer charg voltage would be 3.45v per cell and a warning device to tell you if any cell goes over 3.8v so the charging can be stopped till that cell is discharged back to 3.4v.
The 44v disconnect is also worry because that is only an average of 2.2v per cell. I wouldn't want to take any cell below 2.8v, if any cell is dragged to zero and held there will also stuff a cell completely.
If you can open up the cell and pack and connect up 5 of these http://http://www.ebay.com.au/itm/5pin-JST-XH-Connector-Adapter-plug-silicone-4S-4-Cell-/130422865970?pt=Radio_Control_Parts_Accessories&ha sh=item1e5dcf0032 across 4 cells each and buy 1 or more of these http://http://www.hobbyking.com/hobbyking/store/uh_viewitem.asp?idproduct=17422 and more between the plugs and cell packs to keep them balanced.

to bring this back on topic, the charger/balancer from Hobby king will plug into the same plug as the Cell Logger 8 so if you have wired one of those in already one of these things would make doing that one a mth balance a plug in and walk away type affair. I wouldn't be fast but it would get there after a day or so, the much bigger capacity house battery sells make the job much longer than the small capacity cells they were designed for but if the pack is charged up and close this unit only needs to top them off.

T1 Terry

[LoveMyWoodie]

Wow-I read that at least ten times till I wrapped my head round it.

The idea I get is to -never over charge ,or over discharge-and life will be good.

The links didn't link to your eBay-but I read cell log and connections from the code.

Guess I will gamble on the cheaper second one and monitor it and cross my fingers and learn.Will be better than the ten SLA bricks in my bike and lighter too.

Cheers and thanks,

please continue,makes fascinating reading and so glad to see ICE have a run for the money.maybe we can stop fighting over oil and fill up our tanks from the sun.

[LoveMyWoodie]

Terry

Found the cell logger 8 from your description,

Cell Voltage Monitor & Logger 2

Right ? and now just need some connectors and monitor ?

Thanks again for giving me some time and help.

[T1 Terry]

Originally Posted by LoveMyWoodie

Wow-I read that at least ten times till I wrapped my head round it.

The idea I get is to -never over charge ,or over discharge-and life will be good.

The links didn't link to your eBay-but I read cell log and connections from the code.

Guess I will gamble on the cheaper second one and monitor it and cross my fingers and learn.Will be better than the ten SLA bricks in my bike and lighter too.

Cheers and thanks,

please continue,makes fascinating reading and so glad to see ICE have a run for the money.maybe we can stop fighting over oil and fill up our tanks from the sun.

Bugga, don't ya just hate that :lol: Google Hobby King, then T400W-V2 into that sites search box. It's a charger balancer.
The ebay one was just the plugs, 5pin JST-XH Connector Adapter plug silicone 4S 4 Cell, try looking that up in ebay.

Already power mine from the sun, 1800w of solar.

T1 Terry

[Hyprdrv]

Terry,
Yes thank you for the subject string. I to am eventually going LiPo and this is great info even though I'm probably going to have to read the whole thing a few times to get it to stick in my head. Could you talk a little bit about your Propulsion bank setup and any issues with going to a 144V system with Lipo's? I'm using the PakTrkr monitoring now for the AGM's. How about monitoring the 144V LiPo pack. Do you just string the BMS loggers together?
I currently have a 144V propulsion pack of AGM's and will be going to LiPo's in the next few years (my pack is healthy at this time) and would be interested in your input on this size pack (monitoring and issues) since you sound like your going that route as well with your RV. I do have a concern with my Genset and how to monitor the charging since this is a direct output from the genset and not a "charging" (I'm rectified to 144v "veriable") function.
Thanks again for the education from all you guys.

Steve in Solomons MD
Lagoon 410 SE

[T1 Terry]

Hi Steve,
About all this has to do with the topic is that mine will be both house batteries and propulsion batteries but maybe it should go over to the electric propulsion thread.

At the moment mine are all in 12v configuration for house power. Later I will change this to 48v packs in parallel for house power and series for propulsion, or in my case propulsion assist. I plan to monitor the 48v pack with 2 Cell logger 8 units, charge them as 48v and top balance only. I also plan to use balance gauges between the 48v packs, as long as all the packs are discharging evenly the led in the centre is illuminated but if the voltage between one pack and the next changes to led moves along bar. A quick and easy indicator that one pack has a cell that's struggling and will need attention. If the cell loggers indicate the cells in the lead pack are handling the load and the balance gauges say that all the packs are supplying evenly then a full BMS isn't needed. Expensive on contactors with 2 x 500amp contactors and 5 x 200 amp contactor for the pack alone without the drive contactors but much safer for the pack voltages being limited to 48v for the majority of the time and far easier to charge and maintain with the bonus of a mega house battery pack.

T1 Terry
EDIT: Just a correction regarding terminology just so misunderstandings don't creep in. These are Li batteries, lithium iron, not Lipo batteries, lithium polymer, very different characteristic batteries. In my opinion, lipo batteries are way to unstable for anything but they can be tolerated in very small packs for model planes and boats with all the safety precautions in place. These things even have special bags to charge them in so if they burst into flame they don’t burn the place down.

[sytaniwha]

Hi all,

I just got back from taking our boat from Langkawi (in Malaysia) to Phuket (in Thailand). We spent about 7 days doing it, day sailing between islands and enjoying a leisurely trip.

It was an interesting trip in terms of the LiFePO4 batteries. Here are some observations.

We typically ran the engine about 1-2 hours per day (into and out of anchorages), and ran under sail for about 10 hours per day, and were anchored for the remaining 12 - 13 hours.

Following the discussion here between myself & Terry 2 weeks ago, I adjusted our charge source voltages down a little (from 3.55V/cell to 3.45V/cell). I didn't notice any difference, so it was obviously a good choice.

During the motoring phase, our alternator (110A with a Sterling smart regulator which can put out about 90A on a fairly continuous basis) only needed to deliver on average about 20A to charge the cells, suggesting that they were close to fully charged anyway. (we were only using 50Ahr per day of our 400Ahr capacity).

We ran one of our fridges continuously, and we used about 4A on average (night & day), and the cells never dropped below 3.32V/cell. All 4 series cells remained on identical voltages (to 2 dec places) through the charge/discharge cycles.

One of the interesting side effects of using these cells is the response of our fridge. We have an Isotherm holding plate fridge with their proprietary "ASU" controller. Essentially it looks at the supply voltage and increases the compressor speed when it thinks the alternator or shore power is on (it's a 13.2V thresh-hold), which freezes the plate down. Below 13.2V, it runs in economy mode, trying to conserve energy to not run down the batteries. With SLA batteries, this worked as described, with the economy mode on all the time when sailing, and then the high speed mode switching on when the engine was turned on. Because of our high sailing time %, the high speed mode time often wasn't enough to keep the temp cold enough and we had to run the engine for longer than needed every few days to pull the fridge temp down again (this resulted in quite a variable fridge temp, which isn't good for food). With the LiFePO4 battery, it runs at the high speed most of the time whilst sailing without engine - i.e. the battery is almost always over the 13.2V threshold. Result - we have very cold beer and a very stable fridge temp.

Also, just another plug for Electric1's routine for ensuring all your cells are initially balanced - it works very well. Our cells remain totally in-sync without a BMS. This is because we were supplied with closely matched cells, we initially balanced them, and we are using just 4 cells in a house-bank application. Our cells continue to be within 0.01V of each other, even at the extreme ends of the charge/discharge curves.

I think these cells are absolutely perfect for sailing yacht house bank applications, and in fact require less maintenance and caring for than SLA batteries.

A great example of this is that we no longer leave our charger on when in a marina (we'll be away from the boat for 2 weeks now). The cells simply don't self-discharge at a rate worth worrying about, so if we aren't leaving systems on, there's no point leaving the charger on, and no fear that someone will unplug us.

Cheers,
Paul.

[T1 Terry]

Good to hear all is going well Paul,
Do you have any solar? Of the 4 RV house systems I've been involved with over here 2 are in fully time service and the other 2 are getting the feel for it. It's a big change to get the hang of after lead acid batteries, no problems if the batteries don't fully charge, the volatage barely drops, big loads seem to have little effect and they don't need float charging.
does your fridge use 4 amps continuous, 24/7, or only when it's actually pulling the temp down? Do you know how many Ah per 24 period it draws once it's down to operating temp?

T1 Terry

[sytaniwha]

At the moment we're about half way through a total refit, so we don't have all the equipment hooked up that I would like. That means that I can't monitor load levels very well - I basically guess averages from looking at the instantaneous current meter that we do have connected.

So the 4A I stated in the previous post was my "best guess" for our total average draw for everything over a 24 hour period.

The Isotherm fridge is a great unit, but it's sitting in an old cabinet with old & inadequate insulation. The cabinet is poorly built and way too big (about 175 litres), and so causes the Isotherm unit to run too much (plus we can't load the fridge full, which means it's inefficiently cooling air). Peak load when it is in high speed pull-down mode is about 8A, but I reckon it runs at an average of about 3A, consuming maybe 72 Ahr per 24 hours.

We have the materials for a galley rebuild which we will do in Jan. We're going to have 2 * 90 litre fridges, and a 40 litre freezer, all run off separate compressors. One of the fridges will also be able to run at freezer temp.s., thus giving us flexibility depending on where we're sailing (during ocean crossings it makes more sense to have more freezer space than fridge: during coastal work the opposite is true). Based on our energy consumption of the current fridge setup, I expect each of the fridges to use about 36Ahr per 24 hrs, and the freezer maybe 48Ahr (we'll see what actually happens in a few months - hopefully those are conservative numbers if the cabinets are well built).

Apart from the fridge, the only other steady consumers at the moment are our instruments and lights. (the autopilot isn't reconnected yet, but that will be another significant draw when it is).

I'll be putting back the battery fuel gauge and other monitoring items early next year, so hopefully I'll be able to post accurate figures then.

We do have solar cells, but they were destroyed when another boat ran into us and killed them when it was trying to leave the marina under adverse wind and tide conditions. We're awaiting the insurance replacements. Luckily no other major damage was done.

Cheers,
Paul.

[Marqus]

Originally Posted by T1 Terry

It's a big change to get the hang of after lead acid batteries, no problems if the batteries don't fully charge, .....
T1 Terry

Just to focus on that quote for a moment:

With a Lead-acid battery, if you cycle it between 50% and 80% SOC, never taking it to 100%, the battery chemistry is adversely affected over time - shortened life.

Now, with LiFePo, say due to circumstances (e.g. limited daily alternator time or whatever) you can only cycle the house bank from 20% up to 60% every day. So the LiFePo cells NEVER get to see a near-full SOC, or even substantially full SOC.

Can LiFePo chemistry hack such abuse without shortened life?


Note that what I am describing is NOT "shallow cycling" (eg. down to 60% then back up to 90%).

Instead, you could call it "bottom cycling" - you suck amps until the batt gets near its Low Voltage Cutoff, then you start charging, but never charge for long enough (due to practical circumstances) to get the SOC up high, only reaching 50 or 60% say. And you do this for months on end. (It will kill a lead-acid batt).

1) Anyone have manufacturer/research information that says such cycling will not hurt the battery's chemistry and lifespan?

2) Anyone have experience of a LiFePo battery adversely affected due to such cycling?

[Whimsical]

No experience with them yet but have lurked on the car forums for quite a while and some of them have the batts for a long time without keeping them charged while they build thier cars and have not reported problems.
I hope this is correct as i will be ordering mine soon and they will sit for 1 or 1 1/2 years before they are used

Mike

[sytaniwha]

Hi Marqus,

Here's my take & knowledge of what you describe.

Firstly, I don't believe that the situation will have an adverse effect on the life-span of the battery.

The reason it affects SLA negatively is that the SLA electrodes suffer from sulphation if they are not charged. Sulphation is the process of lead-sulphate crystals forming on the discharged portions of the electrodes. Although it is possible to recover from some instances of sulphation, you will never get full recovery, and if you leave the plates discharged long enough you'll get no recovery. Unfortunately for SLA, the sulphating process starts as soon as any portion of the plates become discharged.

One reason why yacht based house banks lose capacity quite quickly over time is that if you can't put in a full charge on a regular basis, the bit of the plates you're not charging sulphates, reducing the capacity. This compounds over time.

For SLA, the regime you talk about would be disastrous.

LiFePO4 doesn't suffer from sulphation, so from that aspect the key chemistry driver for not running the regime you say is not there. Is there another reason not to do it? I don't believe so.

The only support evidence in the real world that I have for this is from how our cells were shipped to us. It doesn't directly answer your scenario, but it does give some support to an answer.

CALB typically ship their cells at 60% SOC. We received our cells about 3 weeks after their manufacture date, and we then proceeded to do absolutely nothing with them for about 8 months (I didn't even check their voltage - silly, but true). After opening the shipping box to do a visual inspection, they remained in the box a unconnected.

When we eventually installed them and did an initial charge to establish balance, I put about 45% charge into them, and then did a full discharge, getting over 100% capacity out of them.

That says that they self-discharged by around 5% in 8 months, and that being left at about 60% SOC for that time had no adverse effects on capacity.

As I said, it doesn't directly answer your question, but it is in the same general area, and combined with the chemistry attributes leads me to postulate that your scenario should have no problems..

I know for a fact that if I had left activated LA cells for that long, they would have only been good to be door stops (that's why we went to LiFePO4 in the first place). That's why LA manufacturers ship their batteries "un-activated" around the world, and your local distributor "activates" them.

So, that's my input and experience. It'll be interesting to see what other people have found.

It's a good question.

Cheers,
Paul.

[T1 Terry]

These cells are stored and shipped at approx 50%SOC, that don’t self discharge below this point. I have 38 Thundersky LFP 90Ah cells that have been in the packing case for approx 2 1/2 yrs. Have charged 16 of them to give me a 360Ah 12v house battery. It took roughly the same amount of charge to bring them up to conditioning charge voltage as it did the 3 mth old Winston LYP 90Ah 16 cell pack. The only difference I can see is the last bit of charge to get to 100% takes a bit longer, maybe an hr at 20 amps trickling down to zero where the LYP cells go from 100 to zero in about 15 mins. Both packs deliver more than 100% of their rated capacity, both sets stay reasonably well in balance, the LFP set did muck around a little for the first week but bugger all really, I was just being pedantic really.
They take about a mth of continuous cycling to fully get their legs but it's really only the last bit of charge to 100% you would see any difference.
As far as the mid range cycling. From what I can gather from people who have been using these cells for over 3 yrs and from what I can gleen from what Jay Whitacre of the Carnegie Mellon Robotics institute says, these cell actually live even longer if not taken below 30% SOC and not above 90% SOC and will very happily float up and down anywhere within that envelope. The electrolyte heating that causes plate coating happens when the cells are pushed above 100% or below 0% but some degradation occurs below 20% and above 98%.
Because it's near impossible to tell exactly what percentage each cell is without taking it to one extreme or the other and then counting Ah from there the 30% to 90% figure give a comfortable barrier to work in.
If you were never to take the cells above 3.4v or below 3.2v they would possibly out last the user. Even though that's only a 0.2v range that represents between 30% SOC and 90% SOC, their voltage curve is that flat.
I've had mine down to 17% SOC with a 30 amp load on them and they still held above 13v.
If you have a good Ah meter and SOC meter you could float between these figures indefinitely, can't get much more care free than that, but, you would still need to do a charge to bring every cell above 3.5v every so often just to reset were 100% is. Any cell with a no load voltage above 3.45v is 100% full.
They do need to be secured well and the terminal bolt tensions check every so often and the tops of the cells should be kept clean to stop shorting between cells, same routine as lead acid batteries but these don't spit acid out. The reason for securing and firmly packing them together is to stop the cases from bulging if fast charging and to stop them moving between each other and undoing the terminal bolts.
Hope this helps.

T1 Terry

[senormechanico]

T1 Terry,

Would a maximum of 100 amps into a 200 aH pack be considered fast charging?

Actually I doubt if I could even attain that output with only a single belt on my alternator, but on a sunny day I'd be adding up to 20 amps from solar at the same time.

I have the battery bank bolted up with 1/8 inch aluminum plate and corner brackets on one side and a ply bulkhead on the other.

Thoughts?

[T1 Terry]

3CA or higher is considered fast charging so in your case that would be 3 X 200 = 600 amps. The old Thundersky and now Winston and Sinopoly spec 0.5CA as the charging and discharging they did their cycle testing with so in your case that would be 100 amps but I can't see the additional 20 amps from your solar having a detrimental effect. I think they were more concerned with constant current charging till the cut off voltage was reached and then switching to constant voltage charging, in this case it would have been 100 amps forced in until the terminal voltage reached 3.6v. but then constant voltage charging would take over without any current limit, possibly a problem on smaller capacity cells.
Both alternator charging and solar charging use constant voltage so with the max voltage set to 3.5v or less per cell there is no risk of over or under charging the cells and we don't have the capacity to develop big amps to possibly cause cell damage. I guess they had to have a standard testing method to give repeatable results and that was the one that suited the best. Winston make cells ranging from 40Ah to 10,000Ah, not sure how you would get the 5,000 amps to do the charging test but they do cut the max max rate to 1CA for those cells :lol:

As far as the mounting with the plate and corner brackets, sounds fine but use either nyloc nuts or locking nuts on the threaded rod as the tension required is so small that it's not enough to stop them vibrating loose.
I would recommend a cover lid of perspex or plexiglass type material so the risk of things being dropped across the terminals is minimised but you can still see the battery tops to ensure there are no signs of arcing from loose terminal bolts. The copper link plates discolour is there is any resistance at a connector so it's quick and easy to spot.
Clean both the terminal tops and copper link plates with a bit of emery cloth and apply something like Alminox to stop any corrosion because I'd imagine salt air and water would be about as corrosive to aluminium and copper as you could get. any where that handles gear for high voltage/current transmission will carry a suitable anti corrosion paste, if you know a linesman he will have a bit he could possibly spare you, you don't need much. It's a great product to use on all electrical terminals and connectors actually so it might be worth the $20 odd for a tube, there is a lot in a tube.

T1 Terry