Pros & Cons of Lithium-ion batteries for solar application

Re: Pros & Cons of Lithium-ion batteries for solar application

mtdoc wrote: »

I

The Low voltage and high voltage limits in the BMS would simply be a fail safe to prevent damage to your battery bank if the inverter and charger was not programmed correctly. No need for any direct communication between BMS and other components.

True as far as it goes. The BMS will often incorporate temperature compensated settings, as well as circuitry to interrupt the charge cycle if any cell goes over the temperature limit.
The CC should at least have a remote temperature sensor to measure the battery temp, as well as temperature compensation of its voltage setpoints which match the temperature coefficients for that Lithium chemistry, which may well be different than the corresponding coefficients for any Lead Acid cells.

Re: Pros & Cons of Lithium-ion batteries for solar application

Here is a link to some folks that are living in a bus with LiFePO4 batteries

http://www.technomadia.com/category/life-on-the-road/technology/lithium-ion/

And here is a response from MidNight Solar when I inquired about using said batteries with their charger

Aloha Charles,As long as the Li battery's BMS allows it to be a source as well as a sink, there should be no problems.Cheers,TomTom Carpenter Technical Sales 360-403-7207 Ext.150On 10/29/2012 1:07 PM, midni9@sadar.lunarpages.com wrote:

Re: Pros & Cons of Lithium-ion batteries for solar application

THis months home power has an article about Lithium Ions with the pros and cons. The big issue that stood out as a con was that the BMS's currently offered lack some critical features

Re: Pros & Cons of Lithium-ion batteries for solar application

BTW here is a chance at a Lithium Ion battery for cheap. Of course no BMS or TMS unless you can adapt the existing. Scrap the car out for parts which should cover purchase price and shipping and get your battery for free.

http://cgi.ebay.com/ebaymotors/130826602917

Re: Pros & Cons of Lithium-ion batteries for solar application

nut17 wrote: »

Hi Bill, I answered this question on my "newbie" thread, but in short, I have no automated monitoring system for cell balance, but now check manually about once per month and the 4 cells are always within 0.01v, which is even better than the supplier said they would be. Cell balance is apparently no longer an issue with the latest versions of this technology.

I think you found out something fundamental like I did. Don't treat your house bank like you would an electric car where eeking out the last drop of capacity may cause a balance issue in the first place as you try to hypermile it home!

I'm doing small-scale testing $$ with a low-current misapplication of Shorai Lifepo4 powersports batteries before committing to something much larger and the bms / no bms speculation goes on endlessly and so I put it to the test.

What I discovered is this - assuming that one purchases *quality* cells that are closely matched in internal impedance to each other, (you can order them this way I believe) and stays within both the voltage charging specs, and low voltage discharge specs, the balance is extremely close - until you operate in either one of the "knees" above or below the flat charge / discharge profile. Typically this is about 80 - 10% DOD.

I had to do some digging, but finally found a reference from Shorai stating why they don't use a bms. It is because they utilize quality matched cells, and designed the battery to be operated within the 10 - 80% DOD range on a regular basis. I have purposely taken my Shorais to the limit, and even when operated well into the knees (95% DOD to 14.8v overcharge - testing only, NOT recommended!) they are STILL relatively well balanced. Solution - don't operate at the extremes. It was only until later did I find that info from the factory and was pleased that it agreed with my own testing.

I have the factory balance charger. And a Fluke 87V for verification, which I consider the minimum for accuracy here. I also do solar charging with my Morningstar Prostar 15 pwm. Unless I abuse it by purposely draining an individual cell via the balance port, these things stay within an acceptable balance to me even when charged via the standard two + / - terminals.

I learned another trick - just don't charge them to 100%. Ie, just charging up to about 14.2 or so is fine, since there is very little power in the last part of the miniscule absorb - this will help to retain balance.

Here's the real kicker - in addition to the factory charger, I like to have a backup, and picked up an Optimate TM-291 Lithium charger specifically for what it does to handle balancing which fascinated me:

Just charge up to 14.3v, a little bit under a full charge, and repetetively cycle the voltage from about 14.1 to 14.3v - since the battery is not fully charged, this will allow those cells that are a bit low to come up on their own. Interesting! I'm doing that now on a pair of Shorais paralleled together so I can watch the balance delta over time and see how far it may diverge.

Again, quality matters so no el-cheapo duct-tape wrapped cells here. That's been one of my biggest questions and it may have been answered by this small test setup - a bms may just be covering up poor cell quality construction, and that the answer may actually lie in producing a good quality battery in the first place, and operating it within the 80-10% DOD window.

Stay out of the charge / discharge extreme knees, (an LVC may be handy here), use quality cells and chargers to begin with, operate within voltage specs - just like my beloved Pb batteries. By not treating my house bank like an electric car, but willingly giving up a miniscule amount of overall capacity, the whole balance issue may be overblown and just needs common sense monitoring now and then.

Ok, just my opinion, so take that for what it's worth - which may be far less than .02C.

Re: Pros & Cons of Lithium-ion batteries for solar application

Good point. The question is should the battery manufacturer have to build in safeguards that cover all the forms of charge / discharge abuse, or should that responsibility lay with the quality and maintenance of the bike or end user?

My beloved Odyssey tppl agm's didn't come with a bms, and although a faulty regulator may not immediately kill it, it IS damaging the battery. So one may not notice the effect immediately, but it will catch up with you later on. Still, the fault lies with the bike, not the battery's ability to handle abuse.

It is much like Rolls-Surrette or any other battery manufacturer demanding that one stay within specifications. To my knowledge, staying within spec is basically your bms!

Actually Shorai does have a *limited* overvoltage protection (not a true bms) so that you won't suffer immediate failure as long as you don't exceed 15.2v - to the best of my knowledge. However that is NOT a green light to live with a faulty regulator. It just gives you some time to fix the problem. Wide open reg failure is not part of that plan however.

Another manufacturer, AntiGravity doesn't use a bms. They use cylindrical A123 cells instead of prismatic, however I'm sure that they don't just grab cells from a plastic bucket and wire them together either. Unfortunately, I've seen camps on both sides basically rehash the old Odyssey flat-plate vs Optima spiral cell drama, which I've grown tired of. Choose the proper capacity and size for your application and be done with it. Toss in country-of-manfucturing drama, (usually misguided about the real facts), and the comparison with the automotive agm theater is complete. I grow tired of that drama and instead of watching, I'll buy BOTH and put them on the bench! Marketing vs reality becomes immediately apparent - both pro and con.

I don't want to turn this into a Shorai advertisement - I just picked them since I could get them over the counter to test with. However they don't just slap random cells together, but match them at the factory prior to assembly, which is part of the labor cost. There's more to the story but I'll leave that to the bike threads.

An additional clue to LiFepo4 battery care was demonstrated by the Optimate Lithium charger - when below 80% DOD, it applies a very low current (somewhere around 0.1C or so) until the battery reaches that stage, and then continues on with full current charging. It makes me wonder how many guys that are bottom balancing near the very limits of capacity are killing cells with an immediate application of full charge current at that low SOC? That dovetails nicely into how I revive an overdischarged Odyssey AGM with very low current and not trying to hammer them back to life - which doesn't work. Ok, different chemistry, but seems like an overall logical thing to do.

The Optimate also compares internal resistance, so it is much smarter than just basing logic on voltage. That's something I won't be able to do when I build my larger pack, but that clue of not hammering a LiFepo4 at under 80% DOD is something I'll adopt. The best bet is not to go further than that. I'm not trying to hype Optimate, but what it does do seems so logical I'll try to emulate that.

Re: Pros & Cons of Lithium-ion batteries for solar application

PNjunction wrote: »

What I discovered is this - assuming that one purchases *quality* cells that are closely matched in internal impedance to each other, (you can order them this way I believe) and stays within both the voltage charging specs, and low voltage discharge specs, the balance is extremely close - until you operate in either one of the "knees" above or below the flat charge / discharge profile. Typically this is about 80 - 10% DOD.

My understanding of battery balance is that it is the balance in SOC combined with differences in storage capacity, not the difference in voltage between the cells. Any imbalance in the SOC and storage capacity of the cells will lead to a difference in the cell voltages of differing magnitude depending on how unbalanced the cells are, their SOC and the difference is storage capacity.

If we have two perfectly matched cells with exactly the same capacity but one has a SOC of 60% and the other has 50% both will have the same cell voltage to within maybe 10 millivolts because they are in the middle of the very flat part of the charge curve but are very unbalanced. If we now charge them both slowly with 40% of their rated capacity the SOC of the cells will be 100% and 90%. At this point the cell voltages could be around 3.6 volts and 3.4 volts. If we put another 1% of charge into the cells you could get the voltages being 4.0 volts and 3.41 volts and so on. The same thing but in reverse happens if we discharge the cells.

The balancing process is simply a case of balancing the SOC of the cells. In the above example we balance the two cells by putting 10% more charge into the second cell than the first. After balancing, if we charge them from 50% to 100% of their capacity, they could both read around 3.6 volts and we could utilise a full 100% of their capacity (not that we would) if they are in series. This process is the same as Equalising Lead Acid batteries, the only difference being that Lead Acid batteries will self balance if we push the voltage high enough to start breaking down the water in the cells into Hydrogen and Oxygen whereas LiFePO4 batteries need to have the charge imbalance dealt with externally.

The main benefit i can see of having cells closely matched in capacity is that it means that the cell voltages will not diverge from each other over the whole range of SOC, think of the above example. I think one has to look at the extra cost and work out if it is worth it.

Internal cell impedance matching is probably not as important in off-grid systems with the low charge and discharge current requirements compared to electric vehicles with their high charge and discharge current requirements.

I had to do some digging, but finally found a reference from Shorai stating why they don't use a bms. It is because they utilize quality matched cells, and designed the battery to be operated within the 10 - 80% DOD range on a regular basis.

I agree with vtmaps on this, rather like fuses and other safety equipment the BMS will hopefully just sit there and never have to do anything but ...
As a bare minimum i would have something that monitors the voltage of the individual cells and will alarm if any go out of the comfort zone, this should be connected to something that disconnects the power source in the event of overvoltage or disconnects the load for undervoltage.

Just charge up to 14.3v, a little bit under a full charge, and repetetively cycle the voltage from about 14.1 to 14.3v - since the battery is not fully charged, this will allow those cells that are a bit low to come up on their own. Interesting! I'm doing that now on a pair of Shorais paralleled together so I can watch the balance delta over time and see how far it may diverge.

I can sort of see the logic in this, I would be interested in your results

the whole balance issue may be overblown and just needs common sense monitoring now and then.

I would agree with this, for my system I do it manually, for a friend who is not very technical I have set up an automatic system.

Re: Pros & Cons of Lithium-ion batteries for solar application

When it comes to charging, so are we. Development has already been established for many years, but most consumers and media don't know the difference between laptop lithium-cobalt, and much safer Lithium-Iron. Note the "iron" part. Part of the consumer holdback is that they just don't know the difference, and are scared of the word lithium. It would be much like us being scared of a "lead" battery, and not knowing that additional chemistries are either unstable, or in fact quite safe. I had friends deathly afraid of the newfangled "AGM" technology. Wouldn't touch it.

The problem is that one has to differentiate between RC, drag bike, EV, or other motive power usage, and typical low current (relatively) house-bank solar applications. One can get easily distracted by the problems that they face.

LiFepo4 charging is even easier than lead acid. For a 12v solar setup do this:

* Set your charge controller to about 14.1v, even though you *can* set it up to 14.4. Don't obsess about obtaining 100% charge by mandating that current absolutely drops to zero at the end of absorb. The 14.1v setting will give you some headroom for out of balance conditions at the top end.

* Use reasonable quality gear and batteries. Duct-tape or shrink-wrapped cells of unknown pedigree are out.

* Low current usage (<2C, lower even better) if you start out relatively balanced, will stay that way.

* Don't discharge past 80% DOD. That would be about 12.866v for a 12v pack, or 3.466v per cell.

* If you need *insurance* against mishaps, then a BMS that incorporates LVC, HVC, and balancing should be looked into. If you like hands-on maintenance, and use your own LVC, then invest in a quality multimeter, something of the accuracy of a Fluke 87V or better.

* An HVC is already covered by your controller / charger's top voltage limit.

* Look into LiFepo4-specific controllers for more accuracy. Genasun comes to mind, although high-quality controllers with good user configurability can also be used. (For simple Morningstar controllers, the tiny Sunguard 4.5a is already at 14.1v. SunSavers on the other hand would benefit by leaving the jumper set for "sealed", which is also about 14.1 / 14.2v. No need to use the 14.4v flooded jumper removal. Here we're concentrating on voltages and ignoring the Pb chemistry silk-screening.

* Don't use automotive trickle chargers unless you have verified their accuracy. For example, Schumacher speed chargers with their built-in high-voltage desulfation routine have boiled over some friends brand new Odyssey's and Optimas. I have tracked three different versions and didn't like what I saw. If they don't publish their voltages and charge routines, track them yourself with voltmeters and ammeters and discard them as necessary. Imagine what they'll do to a LiFepo4. But why waste time with unknowns. For small stuff, an Optimate TM291 Lithium, or perhaps the CTEK lithium charger is the much better bet than using automotive tenders especially if their performance is not known.

* And kids, DO NOT use a typical solar panel trickle-charger that you find at the automotive parts store, or behind your windshield without using a charge controller! No ifs ands or buts here as Lifepo4 does not like the 18-20v output when directly connected and it races into the final steep charging knee. This is immediate abuse.

* Don't cheap-out by wasting time and safety with too-good-to-be-true counterfeits or junk. Lastly, like all battery chemistries, they demand respect despite their toy-like weight.