Battery Guarantees

[tillson]

A couple of recent posts have caused me to consider, what is a battery guarantee?

Some battery guarantees are quite measly, only lasting a few days, whilst other appear to be very generous, boasting 2 years. But what are they guaranteeing?

To the layman, lacking the rich knowledge of some of the members here, what do they believe they are getting when they purchase a battery? For example, one might be tempted by a battery boasting X Amp hours guaranteed for Y months in contrast to a battery offering X Amp hours, but for only Y/4 months.

From the table kindly supplied by Flecc via The Battery University, one brand of battery, if used within manufacturer’s guide lines, can be expected to lose a minimum of 30% of its capacity in two years, quite possibly possibly even more.

What would you expect from the various manufacturers and / or retailers if you presented your battery to them displaying a 30% reduction in capacity towards the end of its warranty period?

Bearing in mind, if the battery is sold with a tempting X Amp hours guaranteed for Y months, should it still be able to deliver this within the warranty period? To coin a phrase, I think that, “the man on the Clapham Omnibus”, would expect just that. If battery deterioration is known about only by a select few, (manufacturers and a few people with specialist knowledge) should the warranty be boasted in slightly less flamboyant terms? For example, X Amp hours guaranteed to deliver X *(0.66) Amp hours over Y months? (Sorry for the X and Ys, I am trying to be non brand specific)

Unless a manufacturer / retailer is prepared to exchange a battery exhibiting less than the advertised capacity at ANY time during the warranty period, then I think that the guarantees are mis-leading and should be amended. We don’t accept this with other products, so why should bike batteries be any different? In fact, my iPhone carries a guarantee to the following effect. A less than 50% degradation in battery capacity during the warranty period will not be considered in terms of a warranty claim. This may be disappointing, but at least it is made known in the iPhone literature.

To me, with the current situation, a 90 day guarantee may be just as good as a 2 year one. It could be that the manufacturer giving the 90 day guarantee is just being more realistic?

I have no particular beef with anyone over this issue, but I think that it needs to be made clear to existing owners and to prospective new owners EXACTLY what a battery guarantee is. I think that it is all too woolly at the moment and could lead to disappointment and a tarnished image of ebikes. I also hope that this post will stimulate some debate and opinions on the matter.

Could the retailers / manufacturers set out some parameters in terms of battery capacity degradation that they would accept for a warranty claim?

 

[Mussels]

I can't see it on the Wisper site but I think their warranty is for 80% capacity and two years, it might have changed since I saw that.

 

[flecc]

eZee have published a similar qualification on capacity versus life. I daresay all rsponsible manufacturers do have similar provisions, but the issue is probably considered too complex for the main retail market. After all, much of the world wouldn't even understand expressions like capacity, discharge rate, depth of charge etc., but those are intrinsic to any guaranteeing of battery life.

It really is a difficult issue which cannot be ruled on a solely time basis, partly due to the efficiency characteristics of our electric motors. Take these two bikes:

The first is used in a flat area on an intermittent basis, only part discharged and recharged each time. The current drain rate will be low since the motor will spend nearly all its life at near maximum efficiency.

The second is used for daily commuting on a hilly run, emptying the battery every day. The current drain rate will be high for much of the time while climbing since the motor runs near to it's worst efficiency point then.

Three years later the first owner will be happily using their bike and it won't have lost much capacity and range, but the second owner is very likely to be heading towards the end of life of their second battery.

Between those two is a whole range of usage conditions, and it's for all such reasons that batteries are considered consumables.

We have an illustration between yourself Tom, and John (Aldby). His Agattu is used intermittently, is about two years four months old and the battery is apparently performing well last he spoke of it recently, while your battery is questionable at one year of daily use.

As you can see, it's impossible to set a warranty without getting into technicalities, and that will leave most consumers baffled. However, it is desirable that manufacturers do have warranty standards that can be accessed by those interested.
.

 

[Straylight]

I think therefore that more effort needs to be devoted to motor design and internal gearing, in order to keep all motors running at or near peak efficiency, irrespective of the terrain they're used upon. We could then have a more level playing field with regard to what is normal battery life.

There's also the question of usable capacity. Manufacturers won't impose conservative charge/discharge limits via the battery's BMS/the bike's controller, as this would reduce the headline range for their product. I believe however, that to an average user, battery longevity would be more important than range.

 

[flecc]

I think therefore that more effort needs to be devoted to motor design and internal gearing, in order to keep all motors running at or near peak efficiency, irrespective of the terrain they're used upon. We could then have a more level playing field with regard to what is normal battery life.

Internal gearing that can change with conditions would certainly help hub motors to stay nearer to maximum efficiency. The Panasonic motor that Tom uses is always at maximum efficiency due to the way the system works, but still acts in the same way as hub motors with consumption due to the power phasedown conditions. Lower power is added when cruising at near the maximum assist speed but higher power is added at slower climb speeds, especially below 9.4 mph.

An additional problem with these crank drive systems is the gearing though. If owners up the gearing by changing the rear sprocket to get higher assist speeds, as many owners do, the battery gets more of a hammering due to working over a wider speed range and the highest current drain being extended over a wider range of speeds. This again makes warranty fixing difficult. I don't know if Tom's gearing has been raised.

There's also the question of usable capacity. Manufacturers won't impose conservative charge/discharge limits via the battery's BMS/the bike's controller, as this would reduce the headline range for their product. I believe however, that to an average user, battery longevity would be more important than range.

Ideally, given the current/performance demands, e-bikes should have very much larger batteries, 30 to 40 Ah typically, and with LiFePO4 advances and less tolerant BMS, much longer battery lives to as much as 10 years could be possible. But of course, two wheel bikes would struggle to accommodate such batteries. Cars like the Chevrolet Volt and BYD use substantial LiFePO4 batteries intended to last for ten years, solely because they can cope with very large battery sizes.
.

 

[Wisper Bikes]

I can't see it on the Wisper site but I think their warranty is for 80% capacity and two years, it might have changed since I saw that.

Hi Andrew, it's 75% and 2 years.

All the best

David

 

[Straylight]

As an aside, thinking about usable capacity, do the Ah/Wh ratings on current batteries take account of the low voltage cut-off, or are they theoretical, being based on running the pack to zero volts - which is of course impossible using a bike?

 

[tillson]

You’re right flecc, it is a very difficult area. In fact, when you consider all of the variables it becomes almost impossible to deduce expected battery life.

I am sure that under the right conditions, the Schwinn battery with its, laughable 90 day warranty, could last longer and give better performance than say a Panasonic battery with its, “2 year guarantee”. For me, this just illustrates the point that a warranty is pretty meaningless unless it is accompanied by some sort of specification setting out exactly what the battery is being guaranteed against. At the moment, it is just too open ended to mean anything.

I wonder if it is fair to make such a guarantee. Are we really able to deduce that the battery guaranteed for 1 year is better than the one which is guaranteed for 90 days, and that the one guaranteed for 2 years is better than all of them? I don’t think we can say that, but the , headline grabbing guarantees insinuate this.

Just turning to flecc’s latest post at 17:38. I have reduced the size of the rear sprocket to an 18 tooth, and this may well cause the motor to draw a grater current for a longer period than it would if the standard sprocket was used. However, this latest battery, which I have had since December 2008, has been used exclusively on the lowest power setting. The battery has certainly lost a significant amount of range over the winter and it doesn’t seem to have recovered in the much milder weather that we have had over the last few days. I don’t know if this is acceptable because I don’t know what acceptable is.

Perhaps a more transparent way to warranty a battery would be to factor in the known minimum degradation. For example, a 10 Amp hour battery could be marketed at a 5 Amp hour battery guaranteed for 2 years. Not as headline grabbing, but more transparent.

It would be handy if some of those in the trade, could provide guidance as to what they deem to be unacceptable battery degradation in terms of their warranties.

Edit: My post crossed with that of David's at Wisper. Just the sort of information that is needed. Thanks.

 

[flecc]

As an aside, thinking about usable capacity, do the Ah/Wh ratings on current batteries take account of the low voltage cut-off, or are they theoretical, being based on running the pack to zero volts - which is of course impossible using a bike?

I'm sure they are just the theoretical figure, since that has always been the general rule for all batteries. I can't imagine they'd break with convention.

For lead acid (SLA) on our bikes, the position is far worse of course, the Peukert effect meaning only about half the content is usable!
.

 

[flecc]

I am sure that under the right conditions, the Schwinn battery with its, laughable 90 day warranty, could last longer and give better performance than say a Panasonic battery with its, “2 year guarantee”.

i doubt the Schwinn li-ion would outperform the Panasonic battery under the same battery operating conditions, given the long term expertise of Panasonic in this area and the additional management these batteries get, but I take your point Tom, the warranty periods could easily be very misleading.

A to B magazine tested one of the latest nominal 10 Ah eZee batteries for capacity under different levals of power usage and recorded anything between 10 and just below 7 Ah, depending on the current draw set on the powerful Torq 1 motor. That could breach a "70% remaining" warranty on day one! This underlines what I've said above about using large enough batteries. With a 30 to 40 Ah battery, that Torq would have shown hardly any variation in battery capacity with differing current levels drawn by that motor.

Quite simply, all e-bikes are under-batteried, emphasising that we are overreaching technological advances in even having e-bikes at present. Maybe we should have continued with development of the former bicycle petrol motors, developing quiet tractable lightweight four-strokes to use until battery technology caught up. Something like that beautifully quiet purring motor in the little Honda small suitcase generator would be ideal, quieter than some e-bikes in fact.
.

 

[tillson]

That's an interesting point you make about the eZee battery and illustrates very well how difficult it is to make any sense of a battery guarantee.

 

[flecc]

Some added information.

eZee have given 70% DOD warranty over two years on their add-on LiFePO4 parallel battery. Their warranty on the Li-polymer batteries is for 70% over one year I seem to remember.

The Chevrolet Volt e-car I mentioned above only operates between 30% and 80% of battery charge, not much different from the Toyota Prius (20% and 80%). Its this that gives the 8 to 10 year battery life. By contrast, Wai Won Ching says his BMS uses 99% of the available charge down to 31.5 volts, hence the very short e-bike battery life, the cells chemically stressed to their maximum.
.

 

[Blew it]

As an aside, thinking about usable capacity, do the Ah/Wh ratings on current batteries take account of the low voltage cut-off, or are they theoretical, being based on running the pack to zero volts - which is of course impossible using a bike?

The claimed capacity in watt-hours refers to the useable content from 31.5 volts up to 39-42 volts.

 

[flecc]

The claimed capacity in watt-hours refers to the useable content from 31.5 volts up to 39-42 volts.

Every watt/hour rating I've seen is simply the battery nominal voltage times the claimed capacity, i.e. 37 v 10Ah = 370 W/h.

Of course the capacity can relate to the usable content within defined voltage limits, but since the capacity is invariably quoted for an unrealistically low rate of current usage, it doesn't have much practical meaning with lithium batteries. The A to B test of the eZee battery capacity I referred to above illustrates my meaning, few able to use a bike in a manner that realises the quoted 10 Ah.

As the current usage rate of a lithium battery increases, the voltage drops markedly, and the W/h ability with it of course. E-bikes with lithium batteries run at considerably below the nominal battery voltage for much of the time due to the inadequate motor-power/battery-size ratios that I've referred to above.

Of course other battery types like NiMh for example drop voltage under higher loads, but for a given rate, NiMh drops less voltage than lithium types.
.

 

[wurly]

Interesting.
Maybe you can answer my question flecc.
Is it possible then, by charging up a Ping battery to 100% (so all lights on the bms are illuminated) this would be, in fact, shortening it's life?
If so, without monitoring the charge going into it, how would it be possible to extend it's life to the maximum by keeping the capacity within the 30-80% limits. I was under the impression the bms was balancing cells at all times, so maybe there is no need to fully charge and probably better if i didn't....
Mel

 

[z0mb13e]

While the argument against is cost, it seems like a missed trick to me, that the BMS doesn't include a logging sub system that could monitor and log the battery usage over it's life time (or warranty period).

The information obtained would surely be invaluable in setting warranty policy and determining future battery capacity.

Even better if it was possible to hook this logging system up to the bike computer, embed a few accelerometers and you could get a pretty good picture of how the bike is being used.

 

[flecc]

Interesting.
Maybe you can answer my question flecc.
Is it possible then, by charging up a Ping battery to 100% (so all lights on the bms are illuminated) this would be, in fact, shortening it's life?
If so, without monitoring the charge going into it, how would it be possible to extend it's life to the maximum by keeping the capacity within the 30-80% limits. I was under the impression the bms was balancing cells at all times, so maybe there is no need to fully charge and probably better if i didn't....
Mel

Although it's possible to charge short Mel, it actually wouldn't gain much worthwhile. This could be partially due to still probably using tho other end of charge to empty which is equally stressful. However, another problem would still be the inadequate battery size problem that I've referred to above. The e-cars I've mentioned get their long battery life in two ways, first by having really adequate battery size in relation to their motor power, second by not using either end of the charge/capacity.

There's a hint at the relationship for the former with the conditions and lives of a couple mentioned. The Toyota uses between 20% and 80% of charge and has an 8 year battery warranty, the Chevrolet Volt uses between 30% and 80% of charge and 10 years is expected.

If you were able to use a 30 to 40 Ah Ping battery, size depending on the motor power you have, and also only used the centre 60% of charge, you could expect as much as a 10 year life. But of course a battery of that size and weight isn't practical on a bike, the car world is much easier for designers.
.

 

[flecc]

While the argument against is cost, it seems like a missed trick to me, that the BMS doesn't include a logging sub system that could monitor and log the battery usage over it's life time (or warranty period).

Definitely, but cost is probably the main reason why this isn't done, and also influenced by the foregone conclusion that e-bike battery life is limited anyway. Even the consumer "ignorance" factor may have been considered, they still wouldn't understand why one person's warranty is honoured at 18 months while another's isn't at only 14 months, though the monitoring could indicate why that was justified. Consumers tend to judge by the superficial information, understandable when the majority don't appreciate the complex technical issues involved.
.

 

[tillson]

While the argument against is cost, it seems like a missed trick to me, that the BMS doesn't include a logging sub system that could monitor and log the battery usage over it's life time (or warranty period).

I think that the Panasonic battery does include something like this. I remember being told when I exchanged one of my batteries that it would be sent back to Derby Cycles where they could look at the usage profile to see if there had been any abuse / abnormal use taking place.

Whilst ever the battery constitutes a third of the cost of the bike (a modest Panasonic powered bike) and has no specific guarantee (Wisper excused) it will always be a source of contention and disappointment.

This can’t be good for electric biking and makes me hesitate to recommend it to anyone. I find that you really have to look at the whole issue and what you want out of an electric bike before committing to one. I find this very difficult to explain to someone thinking of buying. These bike lie somewhere in the mix of a toy, a sort of keep fit device and a slow mode of transport. They certainly aren’t an economically viable way of doing any of these things, but if the mix suits you, then I suppose they are worth it. If the batteries could be sorted out with a clear concisely worded guarantee of some length, then I think ebikes would slot nicely into an economically viable means of transport.

 

[z0mb13e]

Interesting about the Panasonic battery... ever the tinkerer I would love the see this feature so I could download the data to see if I need to modify my cycling to extend the battery life.

Also a programmable speed controller would be interesting so I could play with the power delivery...

The Wisper battery guarantee was a big factor in my choosing Wisper. Also going for the biggest battery I could gives a certain peace of mind knowing that I am not really taxing it that much.