LYFEYPO4 working temperatures

[karrak]

I actually agree with Sunking that you will probably get a greater lifespan from your battery if you charge it to 3.4 volts/cell and float it at that voltage rather than recharging several times a day to 3.45 volt/cell. This will keep your battery fully charged for the whole day. You would probably do even better to have the float voltage at 3.34 volts if this still keeps your battery full.

Barba, could you give us more details of your system, what is the power output from your solar panels, what are your daily power requirements, do you have grid as backup or a backup generator.

Simon

[Barba]

My system is as follows

1 dual axis sun tracker 720w
1 fix 7000w
1 inverter 8000w
Active balancer Kaitek ( it actively balances while discharging and when recharging)
16 cells 300Ah winston LYP

I know this capacity is very low but I wanted to test a small pack before going to a higher capacity.
My daily consumption in kWh is quite low in the summer and higher in the winter but limited to 8kwh a day in the winter.
With system I run my whole house .so I warm up and cool down the house with a heat pump and I use induction to cook. The energy consumption is so low due to a A+ energy class house.i use the grid as a back up

[Willy T]

Hello sunking
This is exactly what I did the first time but then I read on the forum that floating is a killing so I changed strategy with a capacity meter cutting the panels from the mppt recharger when I reach the 100% which for me was 3,45 and current below 0,25C.

What about the temperature? What is the best range?

Floating is a evolution with SunKing in the last couple months as soon as he realized Solar is not a Golf Cart that you plug into the grid at will.

[Barba]

Sorry gents but I had to cut the pictures a bit for uploading. But the content is still readable
The thing that surprises me is that they cycled the cell 13000 times charging it to 4volts every time.

This is out standing!

[karrak]

Wow, that is quite some system, makes our system with 1.2kW of fixed panels, 360Ah @24 volt Winston Battery and 4kW inverter look tiny. We are totally off-grid and don't have generator backup. Our daily requirements in winter are around 2-2.5kWh and in summer around 4-5kWh. Heating and most cooking in winter is done on wood heater, cooking in summer is induction and very efficient oven. We are situated around 33^o south and our climate is similar to the climate in southern California.

With so much solar power and grid backup I wonder if you need a larger battery, you probably get better value from LFP batteries if you work them hard and you have the grid to fall back on.

Thanks for posting the report, very interesting. It is a shame they didn't do a complete charge and discharge at the start and end to see if the total capacity of the battery had changed.

From what I have learned to date and understand the degradation of the cells is not only caused by cycling but also caused by parasitic reactions that occur regardless of whether the cells are being used or not. The cyclic ageing is greater with larger charge and discharge currents. The calendar ageing is dependant on the SOC/voltage the cells are at and increases fairly linearly with increase in SOC/voltage. Both these degradation mechanisms are dependant on temperature. I would think this dependance is logarithmic with a slow increase to a knee around 35^oC and then increases rapidly from there.

Simon

[Sunking]

From your post http://www.solarpaneltalk.com/showth...l=1#post154383 you give the following figures.

Well then lower it.

Yes, but they keep the voltage at the float voltage which will require then to start charging again if a load drops the voltage below the float voltage. If the load is greater than power the solar panels can supply the load will start drawing power from the battery. When the load is removed the charge controller will then recarge the battery back up to the float voltage.

Well duh. that is the way any Floating System works.

[Sunking]

Please give data to back up this statement.

It is common knowledge, but if you need a reference try this link______________http://batteryuniversity.com/learn/article/how_to_prolong_lithium_based_batteries ___________Then search the page for_____________ Most Li-ions charge to 4.20V/cell and every reduction in peak charge voltage of 0.10V/cell is said to double cycle life. For example, a lithium-ion cell charged to 4.20V/cell typically delivers 300–500 cycles. If charged to only 4.10V/cell, the life can be prolonged to 600–1,000 cycles; 4.00V/cell should deliver 1,200–2,000 and 3.90V/cell 2,400–4,000 cycles. ---------------------------------------------------------------------Jeff also talks about it in his lecture which you missed. There is no need to go above 3.4 vpc. Nothing but unnecessary stress going any higher.

[Barba]

Interesting article but all of it is referring to a LI-ION which is a different chemical from LIFEPO4 and which is different from LIFEYPO4.

If you guys open the pictures I posted earlier you can see cycles life at high DOD with charges up to 4,2 every time!

[PNjunction]

The "Y" in LiFeYPo4 is the addition of yttrium in the Winstons mix is to mostly improve cold-weather performance. Your issue is high heat.

The addition of small amounts of additional material to enhance certain operational conditions does not make them inherently too much different from following the standard lifepo4 routines. GBS adds a *touch* of manganese, but this does NOT make them LiNMc batteries for instance.

Yes, mentioning things like 4.2v voltages without being very specific about it, can lead to confusion for the newcomer coming from the world of laptop, flashlight and cells that are not lifepo4.

YET, you mention 4.2v as specified for Winston. This leads to the following confusion for newcomers:

That ability to hit 4.2v with lifepo4 has to do with TIME. Yes, you can hit them with 4.2v, but ONLY if you are charging at some ridiculously high rate -and for a very very short time. Such as with EV regenerative feedback bursts, and the like. We don't do that with solar.

This is the problem that many early amateur EV'ers faced many years ago and what most of us point out now:

Upper end voltages like 4.2, and low voltages like 2.5v are at the very high and bottom edge of immediate damage. They are on the spec sheet, BUT it does NOT mean that one can use those without knowing EXACTLY what they are doing - or at the very least warning users that these are "edge cases" that fit a very tiny specific set of needs. We don't have those needs. We aren't supplying very quick 10C bursts to our batteries.

Many early EV'ers prematurely damaged their large prismatics by not understanding how these specifications fit in the real world, and not in the lab. Those extreme voltages are LAB-specs under very controlled conditions, and not real-world specs where it is wise to be conservative.

Also consider that most manufacturer's life-cycle tests are just "beating the clock". That is, repetetive charge and discharge cycles are followed back to back, perhaps with a a short amount of rest like an hour in between. That period of charge/rest/discharge beats the clock where parasitic reactions don't have time to do their dirty work over the lifespan of years.

I have a feeling you are seeking / looking for some sort of holy-grail factor that elevates lifepo4 or Winstons in particular to some sort of elevated grace beyond the laws of physics.

[Barba]

I just want to know how many cycles I can expect cycling my batteries at that temperature and at that DOD.
Can anyone answer?
Is 0.05C cutting the charge at 3,4 volts @30C doing any bad in a long run?
Thx guys

[PNjunction]

I just want to know how many cycles I can expect cycling my batteries at that temperature and at that DOD.
Can anyone answer?

You, like me and countless other pioneers can only go by the manufacturers rapid charge / discharge cycle testing. There is not enough data yet to prove one way or the other if it will meet or beat the manufacturer's specs. What makes it even harder is that we are relatively low-current users, a special niche that uses cells primarily intended for the amateur EV / motive power market, so we can't go by their figures either.

Is 0.05C cutting the charge at 3,4 volts @30C doing any bad in a long run?

That is good! Because of the .05C charging rate, you are basically "absorbing" yourself up to 100% SOC. Actually about 95% SOC when you let the cells rest, and they fall back a little. Most would actually consider 3.45v under charge, and then under rest it settles back to 3.4v - but that is a full charge at your rate. Do you need a full charge? No.

Because of what appears to be your over-sized bank, you have the luxury of not even going that far! Why not stop it at 3.35v to stay well below the top - it appears you have plenty of capacity to do so.

At this time, the ONLY way to get a definitive answer that is not purely internet anecdotal, is to send your cells to Professor Jeffrey Dahn's group at Dalhousie University, tell them to use 88F, and have them place it on their testers, where they can actually measure the parasitic reactions that we as consumers have no way of measuring, and giving a much more exact prediction of expected cycle life.

I'm not sure they would accept scads of consumers wanting to do this, so this is purely an example of how we are pioneers in theis field - nevermind being in a very small niche subset of low-current solar users.

There are only 3 options: 1) watch it yourself and provide us with real-world feedback, 2) listen to anecdotal internet, EV or sales literature, or 3) send it off the Dalhousie.

I think you are giving yourself gray-hairs prematurely...

[PNjunction]

Since we're playing the voltage game with lifepo4, one thing not mentioned is the accuracy of your metering!

With your sensitivity to the whole thing, (that's a good thing in my opinion regarding lithium!) is your metering up to snuff? Do you have a calibrated standard against which you can compare to all your other on-board meters and see if you may in fact be 400mv off!

Grab a Fluke, Agilent, Gossen-Metrowatt or the like and this will ease your mind. OR, have your existing meter calibrated. Or, compare your existing meter to someone who has one of these high end units, and do compensation. Take any lithium cell (chemistry doesn't matter for this test), to another bench with a calibrated / reputable standard, and see if your meter is indeed accurate at the low-volts scale (manual or autoranging).

Voltages are thrown around in forums, but the real-world feedback can be inconclusive simply because we can't rely on the posters accuracy of his own gear unless they state it up front!

In my case I use a bevy of Flukes like an 87V, and even their own low end 11X line, *which compared to the 87v* was well within reason for my use with lifepo4. No shirt-pocket metering here, unless of course it was at least compared to something with a known reputation for accuracy.

[Sunking]

Using voltage to determine what state of charge a LFP batteries is a bit on the fruitless side. Even with a good lab quality 1% 4.5 digit Fluke won't even tell you a whole lot. At 3.3 volts you are in the 20 to 30% SOC range. Just .1 volts higher at 3.4 you are in the 90% SOC range. The discharge curve is so flat measuring voltage is not of much use, only an indicator. What I have learned and it works remarkable well to my surprise is Coulomb Counting. All that means is counting Amp Hours pumped into a battery, and counting AH out of the battery. Problem with Counting is it does take some re-calibrating from time to time, although in the EV world they have some pretty decent ones with battery databases built into them that account for efficiency errors. Lithium batteries are really the only type that Counting works because the charge efficiency is in high 90% range, and Peukert Losses are insignificant in Lithium batteries and non existent below 1C discharge rates which solar users would rarely ever do. Well I guess 12 volt RVer's might push 1C frequently.

[PNjunction]

Sunking I agree with you on many levels - but I like to take it to the street.

The op's original statement kind of proves my point "I charge to 3.4v at .05C" and stop. Discussion wise, this is accepted.

That's cool, but DOES HE really? As we know, there is a big difference between an error of say 3.3v, and 3.5. Is one placing blind faith in their BMS monitoring system, whatever that is, without once checking the voltage trigger hardware for accuracy? If you have multiple monitoring meters, WHICH one of them do you trust? Did I get the one monitoring system that slipped out the door with poor qc or suffered some sort of calibration error during rough shipping?

That's why I say that it is WISE with an expensive investment in lifepo4, to have at least ONE standard you trust against which all your other gear is compared or recalibrated to.

If one ignores our advice to be conservative and starts to play around in the deep knees, they BETTER have some trust in their monitoring gear. Establishing a good accurate standard is real peace of mind and provides more trustworthy data in forum feedback.

On the other hand, even though I often admonish that voltage is not a TRUE indicator of SOC, if your metering is accurate enough, you CAN follow a lifepo4 slope - even in the middle to make a relative reading stay relative, and not just be totally out of the ballpark - and dangerously so at the ends.

Ironically, to keep the errors in relative readings down, one should use an accurate meter to start with, especially if they are sharing those experiences with others. We just take it for granted when talking about it, but it could save a noob's butt and budget by making a lifetime investment in a good tool - that they trust.

[Barba]

Measuring tools were the first thing I bought to defend myself from wrong readings.
I use a lab monitor device which is accurate down to 0.00001V and accuracy of 0.001A.
It has the ability of counting ah with a shunt of 50 mv.
Bought it from a German supplier....

Anyway the interesting thing was resetting the ah counter only when I hit 3,45 volts at 0,25 C and then counting back from there down to 20% SOC and with a load of max 0.25 C and then see what was the resting voltage which was 3,25 v per cell.
So well above the 2,8v.

I want to see what happens if I recharge to 3,4 until the current drop to 0.05 and then count from there 80% DOD and see the final voltage of the cell...

What do u think?