Lipo Winter Failures

[Wayne Giles]

In a a post on 'Loong 5S 4300 Lipo Tests' I quoted:- " I believe that Max currents + cold packs = Overstress + damage." and in reply to a query from DH106 I said I would precis the argument from a recent lecture. I think it justifies a new thread as other peoples experiences would help.
People who fly in winter eg competition soarers, experience a very high lipo failure rate which have become known as "Winter Failures" and I recently gave a joint lecture with Bob Smith, part of which attempted to explain why we think this happens and thus enable flyers to avoid it.
Rather than re-write it in precis I have tried to extract the section which deals with the problem below. It is only a theory, but there is some strong supportive evidence; eg when Bob carried out his 100cycle discharge tests on 13 packs, he only had one complete pack failure and one partial failure, so 2 in a total of 1300 discharges. Critically, the packs were all contained in a temperature controlled environment at about 35deg.cent. In testing the cycling equipment I had 2 packs fail in approx. 12 - 15 cycles at about 15 deg.cent.. The difference was the temperature

Backgound info from early in the lecture is that the heat generated in the pack is caused by the current going throughthe ESR (Equivalent Series Resistance) of the pack and is equal to the current squared x the ESR and gives an answer in watts of heat dissipated. The ESR of a pack varies with temperature and has a negative slope so that it is highest at low temperatures:- (Click on graphs to enlarge)
[img]

[/img]
This slide shows how ESR varies between three different lipo packs over an ambient temperature range.
If the packs are operated at 15C, (33A) then every 30 milliohms of ESR will cause a drop of 1V at the connector. You can see that the best pack is about 16 milliohms at 25*C, so that the volt drop is 0.5V. The worst pack at the same temperature is 55 milliohms equivalent to a drop of 1.8V. At 5*C this volt drop increases to 3V, which is why in winter you launch a model at full throttle and it cuts out after 10 - 15 seconds. Whilst you are collecting the model the heat dissipated spreads through the pack so that by the time you re-launch the model the ESR has fallen enough to hold the volts up above 3V/cell.

[img]

[/img]

This slide shows the losses dissipated as heat in three different packs discharged at 20C and at three different temperatures.
As we know that the power lost as heat in the pack ESR is I²R, we can calculate and compare the losses for the three packs at various temperatures. Looking at the two extremes for 20C, the best pack at 35*C dissipates 19 Watts, whereas the worst pack at 5*C dissipates a horrendous 174 Watts, which is just not sustainable. To put this in context, 174 Watts would boil the equivalent volume of water from 5*C in less than 2 minutes.
The heat dissipated in a poor pack will reduce the ESR and hence increase the voltage as the discharge progresses. This produces a characteristic sag in the discharge curve, which we think damages the pack.
A constant current power discharge shows it clearly on a poor pack.

[img]

[/img]

This is pack “B” from the ESR curves and the lower plot shows a typical “voltage sag” where the heat dissipated in the pack reduces the ESR and causes a voltage rise. Bob alluded to this in a recent article and we believe that it causes permanent damage to the pack and is the reason for the high rate of pack failures in winter months when the starting temperature is lower. The pack shown here is rated at 22C but, if our belief is correct, it is obviously overstressed even at 10C and 5*C and just about OK at 25*C.
The current above which this sag occurs depends on both temperature and how good the pack itself is, i.e. on the value of the ESR at the start of the discharge.
I have tried to quantify what is an acceptable power loss in a Lipo pack based on many discharge runs and think it is in the region of 5 – 6 Watts/Ah/cell. Using this assumption; which may be pessimistic or optimistic, I have calculated the max current that can safely be taken from the three packs on a CONTINUOUS basis, at 5*C, 20*C and 35*C

[img]

[/img]

This slide shows the max current which may safely be taken from the three packs at different temperatures. You can see that it varies from 22A at the lowest temp. and the worst pack up to 56A for the best pack and highest temperature.
We are aware that these currents are never used continuously throughout a flight in practice, and our results look pessimistic. Nevertheless, manufacturers CLAIM that the packs are rated for continuous discharge and it is the only practical comparison we can make.
No manufacturer, to my knowledge, suggests derating at low temperatures or qualifies their "C" ratings. Our theory to explain "Winter Failures" is only based on our findings and is very much a theory, but the circumstantial evidence is starting to look overwhelming.

Of the three packs compared on the first graph (ESR v Temp) , "A" is a 25C Loong, "B" is a well known 22C pack and "C" is a pack I tested for Rob and was rejected as a possible catalogue item.

Wayne Giles

[mikeb]

This is very interesting. With ni-cd cells, the charging process is endothermic, while for ni-mh cells it is exothermic. They are reverse for discharge. This is why ni-mh cells get hotter, when charging, than ni-cd cells. I have not seen any reference to which way round, if either, lipo cells are. Clearly, due to the ESR, they generally get warm for both charge and discharge. If lipos are endothermic on discharge, then the initial current flow on discharge will also have have a cooling effect, which would make the voltage sag worse!

Mike.

[Wayne Giles]

Mike,

I believe only Nicads are endothermic on charge which is why the negative slope is so steep at the fully charged point compared with Nimh. I think Lipos are neutral or close to it as the charging efficiency is >99%. Although the temp. rise must be due to the sum of any electrochemical action (+ or -), and the heat generated by the current through the ESR, the former is likely to be so swamped by the latter as to be of no significance.

Rob was unaware of the info. posted but has since told me that he noticed the rise in complaints about early lipo failures last winter.

If anyone has had direct experience, it would help if they could POST DETAILS ON THE FORUM regarding any cold weather lipo failures, not just Loongs. If the theory is correct, then I would expect the lowest ESR packs (Kokam and Loongs) to show the lowest failure rates.

Wayne

[Andy2No]

It's worth bearing in mind that if you store fully charged lipos at room temperature, they start to lose capacity - permanently. For long term storage, I think the recommended voltage is 3.95V per cell, but for storing cells fully charged, close to 0 degrees C (but not freezing) is best.

As I don't like to have to replace lipos too often, I store mine in sandwich boxes in the fridge. I then take them to the field in an insulated picnic bag (thick carrier bag sort, inside my picnic hamper transmitter bag). I still do the same thing in winter and haven't had any problems with cells failing.

I think the main reason for that is probably because I always set planes up to use no more than half the C rating of the batteries, e.g. running a 20C 1800mAh at 18A max, not 36A max. I also time my flights and measure cell voltages afterwards to make sure I'm not hammering my packs, and adjust the time if the cells go below about 3.8V. Below 3.7V means likely damage, so I aim a bit higher. I also try to avoid taking off at full throttle unless it's absolutely necessary for a particular plane. It rarely is.

People flying with Ni based cells have much more trouble flying in winter - the packs don't fail on them but the performance is terrible if they're too cold, to the point of barely flying, or only flying very briefly.

It's pretty clear that lipos can warm themselves up, but it takes a bit of time for the heat to spread evenly through the pack. If your models are set up to draw high currents and you launch at full throttle, that would be a problem.

Generally, launching at full throttle is not necessary, or even a good idea. The best way is to use enough throttle and no more - release the handbrake on your car, put it in neutral, stand behind it and throw yourself against it as hard as you can. Did it move? No. Did it hurt? Yes. Now try just leaning your weight on it and pushing gently. I think you'll find that works a lot better.

Trying to take off at full throttle is inefficient because the pitch speed is a lot higher than the airspeed, unless you're facing into a stiff wind. It's worth experimenting to see what's a good throttle position for take off, and not giving it full whack straight away.

[whitemice]

rtgh

[moyesboy]

I have yet to experience a complete winter failure but I have experienced lipos underperforming in the cold.

The informatin here about keeping fully charged lipos cool to improve life is in conflict to information elsewhere. My charger is from FMA in the USA and their lipo chargers actually have a "cold weather mode" and a "charge to store" mode. They state that in cold weather the capacity of the lipo is reduced andso it should not be fully charged. The cold weather mode cuts off at 90% capacity (I'd need to research the voltsper cell).
The theory is that a fully chargedlipo at room temperature will actually get damaged if you let it get cold. They recommend keeping charged packs at over 10 degrees c. Letting them get cold is similar to over charging them and damages them. This conflicts with the suggestion for keeping fully charged packs in a fridge though "whitemice" doing that seems to be getting along ok.

It is possible that what is best for one brand is not best for another I guess. The FMA cells are Worley Parsons cells but they had the same recommendations for the Kokam cells they sold before.

Long term storage at about 60% capacity seems to be universally agreed. This is what charge to store does on my charger and this is how all lipos seem to be shipped (they get very cold in warehouses and in air freight).

In the cold what I do is I have a foam lined box (camera case type) that I keep my lipos in and on a cold day I stick a microwaveable pet cimfy pad in there (hot pad for the cat to sit on). Then still the lipos from there into my heli and it flies like its a warm day. I've had cold lipos that barely lift the heli off the dack but will fly full on 3D in the warm. If I ever find the lipo got cold I don't persist in the hope it will warm itself up but I stick it in my pocket for a while before trying again.
I have 2100 3 cell packs that have been stored fully charged at room temeperature for 2 years and still fly the heli the same as ones that are less than year old so they don't seem to be going off. I now have some 5s4200 that I'm trying to keep stored at 60% as they wereso expensive...

[john cole]

Long-term storage: I've seen recommendations for storage at much lower charge-states.
CADEX recommend 40% charged / 60% discharged (which equates to about 3.75 volts per cell)
http://www.buchmann.ca/article23-page1.asp
but Sony recommend storage almost fully discharged at 3.3 volts per cell. The following URL downloads the Sony Lithium Ion Rechargeable Battery Technical Handbook as a .pdf but note you need the Japanese character-set to read it properly, even though the text is in English. Best guide to Li-Ion batteries I've come across (with good explanations about why over-charge is so dangerous and over-discharge destroys the cell):
http://www.sony.com.cn/products/ed/battery/download.pdf
Note that destructive over-discharge is said to be at MUCH lower voltages than 3.3 volts per cell. The copper in the cell will migrate and form a short-circuit, but only below about 1 volt per cell. That is over-discharge.

I've seen references to not letting the batteries "freeze". I'm not sure what people mean by this, but it's nothing to do with water turning to ice! LiPos CANNOT contain water, as the water would be electrolysed (to hydrogen + oxygen) by the internal 3.7 volts potential. The electrolyte can be taken down way below zero Celsius, and the Sony manual gives operating data down to minus 20 degrees Celsius. According to Sony, "The electrolyte is a solution consisting of a mixture of solvents including carbonate esters, in which is dissolved lithium phosphate hexafluoride."

Provided the LiPo is not charged above about 4.2 volts per cell, none of the Lithium is in the metallic state (which is why LiPos are correctly classed as Lithium Ion cells). It will either be present as Lithium Phosphate Hexafluoride in the electrolyte or as Lithium Cobaltate (or a mixture of this and Nickelate and / or Managanate) in the cathode / Lithium Hexacarbide in the anode - in all cases in the ionic state. Above 4.2 volts per cell the Lithium can plate out as metal on the anode. That's when they get most dangerous, as overcharging can lead to rupture, and the exposed metallic Lithium will spontaneously ignite.