LiFeYPo4 questions

[SunEagle]

6) VERIFY your charger's voltage with a FLUKE!

Does that mean I shouldn't use my Extech EX530 meter?

[PNjunction]

Here I am breaking my own rule about EV, but in the spirit of KISS, here's how to "bottom balance". Some might do this because it is not uncommon to drive the battery down to dangerously low levels, and with the application of high-current, one of the cells can easily go reverse polarity right under your nose. They want the cells to reach the LVD all at the same time.

Once again, balance does not mean trying to obtain the same capacity out of the cells, they merely want the VOLTAGES to be the same to help prevent a nasty reverse-polarity situation.

As solar users, you *could* do this, but then you leave your balance ragged at the top. You choose one method or the other, you don't do both.

1) If you have a decent charge in your bank already, you do a bulk discharge until each cell is roughly near 3v.

2) Using some sort of automation (hobby charger, programmable dc power supply that can deal with batteries etc), you take each cell down to about 2.6 to 2.75v depending on who you talk to. Some may place all the cells in parallel and do it that way, or others might do each cell individually still connected in series. It is just tedious, but in the wrong hands, if you aren't on the ball, you are in the DANGER zone of taking your cell too far.

3) You are looking for a "resting" voltage which implies that you may need to do this multiple times after the cells have rested for 15 minutes to several hours since they tend to recover voltage after a rest. Once they are fnally stable rested at 2.6 - 2.75 (stick to one value only), you can then begin a charge.

4) NOTE NOTE - I have often NOT seen the common procedure of using NO MORE than 0.01C to initially bring these low cells back out of the deep discharge knee to 3.2v before using the normal full current! I wonder how many EV'ers do a nice bottom balance, and then hammer their cells with too much current initially.

5) Once the VERY FIRST CELL hits about 3.5V - STOP. Take note of your charger voltage, and use this as your HVD. Again, to do this right takes cell level monitoring and you may not have an exacting top-balance. Just use the very first cell to reach 3.5v as your guide.

6) VERIFY your charger's voltage with a FLUKE! I can personally attest to my West Mountain Radio voltage being off a little bit. If you want precision, you'll verify that your charging gear is actually up to spec! With thousands of dollars at stake, there is no reason to use some cheap-o voltmeter.

Man, I did it now. Hope I didn't open the floodgates. But some solar guys ask me about this. I tell them not to under-design their banks in the first place!

[Sunking]

PNJunction,
Thanks for the help. I will look into those companies that you mentioned. Being new to all this, I agree that EVers seem to have a lot more on their plate when it comes to the whole system setup and potential problems.

How so?

Most DIY's use a Warp 9 and 13 electric motors. Both are 72 volt motors, and they use a single battery string of Calb 100 AH batteries with a BMS. Those with Deeper pockets just step up in voltage to 144 volts using the same Calb 100 AH battery. Even th ebig boys working in R&D use the same Calb 100 AH cell up around 400 to 600 volts using AC induction motors with VFD. Not much difference between 48 and 72 volts.

[PNjunction]

Then I stumbled across Thundersky 700ah cells (same yellow cells as Winston??). I can pick up these cells for $560/cell with 700ah from several different suppliers. Seems like one hell of a deal. Any thoughts on these?

Thundersky's pre-date the Winston/Sinopoly/Balquon split. Since that time of Thundersky, it is reputed that the more modern Winstons have higher manufacturing tolerances, improved chemical and build quality. Also the addition of Yttrium (The "Y" if LiFeYp04) is supposed to improve extreme cold weather performance.

IMPORTANT - some very old ancient documentation on Thunderskys show them as having / needing an initial charge to 4.2v per cell. DO NOT DO THIS. Likewise, 2.5v per cell as a discharge limitation is not something to use as an LVD. This is akin to setting a lead-acid LVD to 10.7v - too low, don't go there. Unfortunately, you may see these ancient specs in some modern commercial "drop in" lifepo4 replacements. I guess they didn't get the memo 4 years ago.

My inclination is to leave the old Thundersky's in the hands of experienced practitioners, not a first timer. Again, I have no personal experience with anything but GBS, but would be cautious about old stock Thunderskys - I'd have to verify the storage voltages myself to make sure that no cell is dead from improper high-heat storage or mishandling over the years.

Here is an example of a 12v / 400ah bank of Winstons undergoing testing for marine use. Note that this test goes WAAAY too far in discharge, something we should never experience:

https://www.youtube.com/watch?v=oJ8AfJsOhsM

While he took this down to nearly 100% DOD discharge, there is no indication that he limited current during the recharge at that low DOD!! If you go under 80% DOD, you should not apply any more than about 0.01C until the cells reach 3.2v again - whereupon you can apply normal charge current. So don't go below 80% DOD in the first place, and of course never leave a lifepo4 in a badly discharged condition for very long. Get to it asap.

Note that for my land-locked system, I'll be having none of that rat's nest, but simple pack-level maintenance as described here:

http://www.solarpaneltalk.com/showthread.php?17609-Simplified-lifepo4-charging-and-care

I don't think I will have a problem assembling the individual cells with my friend who is in the solar industry.

Cool. Just make sure he knows what he is doing, ie using a torque wrench set to the proper specs for the terminal connections, and also making sure that the terminals and links are clean before assembly. Last thing you need is a high-resistance terminal here, and having the bms mask the issue with incessant balancing, or melting the case around the terminals. Basic battery care really.

Do all battery cell banks need to be strapped to prevent expansion?

Yes. This prevents any side or top-torque breaking or twisting the cell terminals if the cells were just loosely assembled and sat down. Any significant movement without banding/strapping and game over. Tall cells operating under high-heat environments have been reported to slightly buckle or sag, and this is also bad, since you don't want any expansion within the case itself separating the anodes/cathodes away from each other. Under abusive over or undercharge practices, the cells can expand, but having them strapped and banded is not a cure for improper charging! Don't do that in the first place.

[asheville solar]

PNJunction,
Thanks for the help. I will look into those companies that you mentioned. Being new to all this, I agree that EVers seem to have a lot more on their plate when it comes to the whole system setup and potential problems. This is purely coming from an outsider who is newly looking into all this. I will be having a friend who is an installer of a local solar company help me with the whole system install.

I did want to get your advice or anyone else's on batteries. I have looked into the Calb cells with a 400ah rating. Not a bad price. Then I stumbled across Thundersky 700ah cells (same yellow cells as Winston??). I can pick up these cells for $560/cell with 700ah from several different suppliers. Seems like one hell of a deal. Any thoughts on these?

I don't think I will have a problem assembling the individual cells with my friend who is in the solar industry. Do all battery cell banks need to be strapped to prevent expansion? Naturally I will be wiring the BMS system for these. I am thinking about the Orion Jr. which seems like a quality manufacturer.

Thanks,
Chris

[Sunking]

Compared to us, they've got extremely high voltage / high current battery setups

Maybe not as much as you think. One of the most popular motors used in the DIY arena is a TransWarp 9 motor made by netgain and it is a 72 volt motor. Other popular motor are based on 140 VDC platform of controllers and Open Source BMS. Most use a single string of 100 AH LFP from Calb CA100FI Cell 3.2V 100Ah. Now some of the big guys in R&D I know who work for GE and another electrical manufacture run AC motors with with VFD up to 400 and 500 volts use the exact same Calb cell in a single string. Once I think has gone with two parallel strings but that is an exception.

The only thing really different is very brief discharge rates of 5C on EV's from a standing start to standing on the accelerator until you hit cruise speed, then discharge rates drop dramatically to less than 1C.

--Mod Note: Something about this post got it automatically classified for moderation. Probably the direct links to product. Since Dereck (Sunking) is a regular with an established reputations for impartiality, I am allowing this post.

[Sunking]

Compared to us, they've got extremely high voltage / high current battery setups

Maybe not as much as you think. One of the most popular motors used in the DIY arena is a TransWarp 9 motor made by netgain and it is a 72 volt motor. Other popular motor are based on 140 VDC platform of controllers and Open Source BMS. Most use a single string of 100 AH LFP from Calb CA100FI Cell 3.2V 100Ah. Now some of the big guys in R&D I know who work for GE and another electrical manufacture run AC motors with with VFD up to 400 and 500 volts use the exact same Calb cell in a single string. Once I think has gone with two parallel strings but that is an exception.

The only thing really different is very brief discharge rates of 5C on EV's from a standing start to standing on the accelerator until you hit cruise speed, then discharge rates drop dramatically to less than 1C. But the tech is the same.

--Mod Note: Something about this post got it automatically classified for moderation. Probably the direct links to product. Since Dereck (Sunking) is a regular with an established reputations for impartiality, I am allowing this post.

[PNjunction]

The amateur EV'ers have a nightmare on their hands - although it can be done.

Compared to us, they've got extremely high voltage / high current battery setups with oodles of interconnect wiring, sometimes the unnecessary use of balancing systems that fail or mask pending issues - the whole gamut.

It can be done - it took me a few years to somewhat agree with Jack Rickard on the basics, despite the entertainment.

Still, I always kept in mind the experiences of those in the marine crowd using lifepo4 house banks as a closer approximation to what we do. (ie, Cruisersforum, etc). And compared to them, we have it even easier than they do! Many are going on 5 years or more, but still we are watching out for real-world 10 year usage patterns. I'm on my way with my own.

[Sunking]

Exactly my point - that is not the application we use them for, where EV'ers routinely run 0-90% DOD or more,

I would debate that but get your point. From an EV POV especially DIY EV where they use the exact same cells you do run them more in the 90's to 100, down to 20-30% to hit the maximum cycle life. Users of commercial made EV's CPU will not let them drain them below a set point, but there are not many using LFP.

EV's charge differently than Solar. In Solar you have to use a Forced Charge Algorithm (basically the same as the BULK for FLA) to 90 to 95% SOC and terminate. EV's charge a bit slower, well sometimes slower, but they use Constant Voltage Current Taper Algorithm and charge terminates when current tapers down to 3 to 5% of C. That allows them to saturate from 95% where solar stops with EQ-Balance to full charge at 100%.

[PNjunction]

From the EV side of the house where they are used extensively for some time with a track record.

Exactly my point - that is not the application we use them for, where EV'ers routinely run 0-90% DOD or more, have failing balancing circuitry and other infrastructure points-of-failure related to vehicle operations.

Our house-bank needs are entirely different, and will lead a very sheltered life in a "fractional C" environment.

[Sunking]

I'm with you on the sizing, although my needs are only 2-day autonomy, if that - but my needs are relatively small. Still, any solar user would be foolish to design to go past 50% DOD, even though lifepo4 will easily do 80% DOD at full performance. My GBS batteries continue to amaze me.

When it comes right down to it, there is not much difference in usable capacity between FLA and LiFePo4. If you cycle both at optimum cycle life SOC they are pretty close to each other. Example let's say 100 AH. Al FLA you cycle from 100 to 50% for 50 AH. LiFePo4 80 to 20% or 60 AH usable.

I don't know where you got those claims from.

From the EV side of the house where they are used extensively for some time with a track record.

[PNjunction]

I have looked into two companies after finding them mentioned in a previous thread from this forum if I am correct. They are Alliance Renewable Energy Inc. and Electric Car Parts Company.
Possible choices would be Calb CA400FI new model at 400ah. $497/cell
Or similar Thundersky (Winston owned if I am correct) from Alliance Renewable Energy.

I was going to mention those guys, especially if you are dealing with USA stock. I can't speak personally for any of them. I just know I can get my hands on high quality Calb's and GBS rather quickly if needed.

Others commonly found:

elitepowersolutions
evsource
ev-power
electricautosports
gwl/power

CALB for most is considered top of the line, and the CA cells are great. However, don't go nuts over other models that offer higher and higher rates more applicable to the hardcore high-performance EV market - in other words, they have cells that can do 5-10C or more easily - but if you made say a 400ah bank of those, would you EVER draw 4KW continuously in a house-bank setup? No. So you'd be paying for a feature that you'd never use. Stick to the lower rate cells for house-bank duty.

This is a common mistake for those trying to adopt cylindrical high-rate cells for our "fractional C" application. They are paying through the nose for a current capability they will never achieve in either charge or discharge.

Unless you are a good DIY'er, I would definitely have them pre-built with all the necessary cell-links, banding straps, and the like. You don't want a large box of just individual cells showing up on your door unless you know how to deal with it.

[PNjunction]

PN FWIW sizing requirements for lithium are not much different that than FLA. With FLA minimum autonomy is 5 days, lithium minimum is 3 days to be equal in usable capacity.

I'm with you on the sizing, although my needs are only 2-day autonomy, if that - but my needs are relatively small. Still, any solar user would be foolish to design to go past 50% DOD, even though lifepo4 will easily do 80% DOD at full performance. My GBS batteries continue to amaze me.

One very positive aspect about lifepo4 that is sometimes overlooked is that their ability to operate in a partial-state-of-charge, even their preference for that, and the ability to go to 80% DOD affords a wide headroom latitude for poor solar conditions, unexpected loads, or minor goofs in actual capacity calculations. Had it not been for this aspect, I wouldn't have been interested in getting my own batteries and verifying the claims.

The cycle life claims for lithium thus far have proven to be false, and cold weather performance is very poor. Lots of stranded Leaf owners last winter.

I don't know where you got those claims from. Most of what I've seen from house-bank users has been extremely positive, although they have only been commonly available for about 5 years now - still a few years ago to get the calendric-aging affect accounted for. Also remember that the Leaf, (like the Boeing incident that also gets brought up), are NOT USING lifepo4 chemistry. Neither are they operating in the same conditions a typical house bank user is. Add to that incidental things like thermal management, wiring infrastucture, charging details etc can easily distract one from the real usage we would put them to. That whole discussion is a red-herring.

You know we're going to laugh about this one day when I hunt you down, and hand-deliver a set of lifepo4 cells and run them through some cycles emulating house-bank duty. I'd like to bring my video camera so I can post it on the net - the look on your face will be priceless.

[Sunking]

Energy usage in the evening would be normal house hold of three usage. Fridge, oven/stove top, lights, possibly washer and dryer, heat & air. I know that does not help without looking at each component separately and see what it's drawing.

OK I have not kept up with this thread, but you cannot run all that on battery. You can get the small stuff, but ovens, cook tops, dryers, air conditioning are very high wattage devices. Ovens pull 2500 to 4000 watts, cook tops 1500 to 4000 watts, electric dryers 4000 to 4500 watts, air conditioning can go even higher than all the others.

[asheville solar]

PN FWIW sizing requirements for lithium are not much different that than FLA. With FLA minimum autonomy is 5 days, lithium minimum is 3 days to be equal in usable capacity. The big advantage of lithium can be Peukert losses and weight, but weight i snot really an issue for off-grid unless we are talking a boat or RV. Myself personally I would not use lithium a sit is just still too expensive and unproven. The cycle life claims for lithium thus far have proven to be false, and cold weather performance is very poor. Lots of stranded Leaf owners last winter.

PN & Sunking,
Thanks for the helpful insights. For clarification, I will be still tide to the grid with the hybrid inverters. With the battery bank, it will only be used to supplement my energy usage at night. I would love to go off the grid but like you said, the expend would be very high. I very well may manage my cells with the BMS system as you mentioned. Purely on the discharge minimum if I understood and the Outback charge controllers for the max cutoff.

Energy usage in the evening would be normal house hold of three usage. Fridge, oven/stove top, lights, possibly washer and dryer, heat & air. I know that does not help without looking at each component separately and see what it's drawing.

I have looked into two companies after finding them mentioned in a previous thread from this forum if I am correct. They are Alliance Renewable Energy Inc. and Electric Car Parts Company.
Possible choices would be Calb CA400FI new model at 400ah. $497/cell
Or similar Thundersky (Winston owned if I am correct) from Alliance Renewable Energy.

Does anyone have any thoughts on either company or cells?

Thanks,
Chris