Newbie questions about LiFePO4 battery packs

[Peter VanDerWal]

> A Trojan T105 battery can deliver 185ah in 5 hours or 225ah in 20.
>

First off, unless you're driving really slow, or have lots of batteries on
board, even the 5 hr rate is too much.  Typical EVs drain the batteries at
about the 1 hr rate.
A T-105 has about 105 AH available at the 1 hr rate.

> 1) If full discharge = 225ah, but the battery can't deliver any more
> than 185 of that in one commute, does this imply that an EV can't
> discharge a battery much below 80% DOD in a single trip? In other words,
> if you charge the battery, unplug it and run until the car stops, does the
> battery still have 40ah of (inaccessible) charge remaining in it, or is
> that 40ah lost as heat generated by the rapid discharge?

Yes and no.  Much of the extra is wasted as heat.  However there is still
some energy left in the batteries.  Part of the problem is depletion of
the acid close to the plates, so you can wait a bit for the acid levels to
stabilize throughout the electrolyte. Plus you can still get more energy
out at lower amp discharges.

However, this is NOT what they mean by 80% DoD.  80% DoD means 80% of the
AH available AT THAT DISCHARGE LEVEL!!!  If you're averaging about 105
amps (the 1 hr rate) then you should only remove 80% * 105 AH or about 84
AH.

You'll note that 84AH is a LOT less than 225 AH.  If you take out 84 Amp
and let the battery rest and then try to pull current out at 105 Amps, the
battery voltage will quickly sag.  You'll have maybe another 25 AH
available before the voltage sags below 1.75 VPC (i.e. dead)

>> I'm also confused about packs constructed of the various types of
>> Lithium
>> Ion cells available. An available type of LiFePO4 cell is advertised as
>> 3.2v, 1100mAh, $3.20 ea in quantity.
>>  www.voltphreaks.com/ssl/catalog/product_info.php?products_id=31
>> www.voltphreaks.com/ssl/catalog/product_info.php?products_id=31
>>
>> At risk of embarassing myself with my incomplete knowledge:
>> a) to make a pack equivalent to the usable kwh in a pack of 24 t-105's
>> (144v@225ah*.8) requires 8000 of these cells??? (3.2v@...*.8=2.8wh
>> each)
>> That's $25,000.

It depends on if you are trying to match the 20hr discharge capacity of
the T-105s or the 1 hr discharge capacity.
Note above:  225 vs 84 AH

>>
>> b) clearly this pack would not be arranged as a single string. How would
>> it be set up?

You wire a whole bunch of cells in parallel (and hope they survive).  Then
you take these parallel groups and wire them in series.
If you wire 80 of these in paralle, you'll have a ~3.2V @ 88 AH module.
Wire 45 of the modules in series and you get ~144V @ 88 AH, a wee bit more
energy storage than the T-105s have at the 1 hr rate.
Of course you should take voltage sag into account, but I'm trying to keep
it simple.
Oh yeah, you'll probably want some kind of Battery Management System (BMS)
for the LiFePO batteries.  Something to make sure the voltage doesn't go
to high (on a cell by cell basis) during charge or to low during
discharge.

Also worht considering, the energy/power needed to get your car down the
road is proportional to the weight of the vehicle.  If you can save 1,000
lbs, it will take less energy to go the same distance.

>> c) Ian described these cells as a comparatively good value, but if my
>> understanding of their capacity is correct, they cost $0.91/wh. Yet, his
>> review of  http://zeva.com.au/tech/headway/ these headway
>> cells indicates that they cost $0.67/wh. In comparison, golf cart
>> batteries appear to cost between $.07 and $0.15/wh.

Don't forget the cost of the BMS for the LiFePos, you're probably looking
at somewhere around $20-$30 per cell/module.

At EV discharge levels, T-105s currently cost about $0.20 per Wh.  They
are pretty much the best bang for the buck available.


> Do lightweight/long
> life solutions really cost this much more?

Yup.  If it was cheap, we'd all be using them.

>> d) to get "6cyl" performance from a motor, are "high power" (as opposed
>> to
>> high capacity) cells really required?

Performance = Power
High Performance = High Power

A typical 120V 400 amp motor/controller will produce about the same power
as an old VW bug.
Step up to a 144V 1,000 amp controller (and some high power batteries) and
you're up to the level of a modern economy car.

If you want 6cyl performance you'll need either high voltage, very high
current or both.



_______________________________________________
For general EVDL support, see http://evdl.org/help/
For subscription options, see http://lists.sjsu.edu/mailman/listinfo/ev

[lumberjack_jeff]

Thanks to everyone who contributes to raise the collective knowledge. My knowledge probably needs raising to a degree greater than most. ;-)

I have a number of questions.

A Trojan T105 battery can deliver 185ah in 5 hours or 225ah in 20.

1) If full discharge = 225ah, but the battery can't deliver any more than 185 of that in one commute, does this imply that an EV can't discharge a battery much below 80% DOD in a single trip? In other words, if you charge the battery, unplug it and run until the car stops, does the battery still have 40ah of (inaccessible) charge remaining in it, or is that 40ah lost as heat generated by the rapid discharge?

I'm also confused about packs constructed of the various types of Lithium Ion cells available. An available type of LiFePO4 cell is advertised as 3.2v, 1100mAh, $3.20 ea in quantity.
www.voltphreaks.com/ssl/catalog/product_info.php?products_id=31

At risk of embarassing myself with my incomplete knowledge:
a) to make a pack equivalent to the usable kwh in a pack of 24 t-105's (144v@225ah*.8) requires 8000 of these cells??? (3.2v@1.1ah*.8=2.8wh each) That's $25,000.
b) clearly this pack would not be arranged as a single string. How would it be set up?
c) Ian described these cells as a comparatively good value, but if my understanding of their capacity is correct, they cost $0.91/wh. Yet, his review of these headway 
cells indicates that they cost $0.67/wh. In comparison, golf cart batteries appear to cost between $.07 and $0.15/wh. Do lightweight/long life solutions really cost this much more?

d) to get "6cyl" performance from a motor, are "high power" (as opposed to high capacity) cells really required?

I feel as a student in a foreign language feels when hearing something that indicates that what he thought he knew about the meaning of words is not the actual meaning.

My background is mechanical - not electrical. Can someone please point me to an online resource to make me less ignorant?

Thanks
Jeff

[Peter VanDerWal]

> Peter VanDerWal wrote:
>
>> A T-105 has about 105 AH available at the 1 hr rate.
>
> This is incorrect.
>
> The "105" in "T105" refers to minutes of reserve capacity at 75A (though a
> modern T105 has been improved to spec at 115min now).

I never said the 105AH had anything to do with the name T-105.  I just
happen to remember some real world discharge curves that indicated the 1
hr capacity was around 105 Ah and it stuck in my mind because of the
coincidence.

> Taking the published 5hr capacity of 185Ah and the 115min @ 75A reserve
> rating and plugging them into Uve's calculator yields a Peukert exponent
> of 1.357 and an estimated 1hr capacity of 121Ah.
>
>> > 1) If full discharge = 225ah, but the battery can't deliver
>> > any more than 185 of that in one commute, does this imply
>> > that an EV can't discharge a battery much below 80% DOD in
>> > a single trip? In other words, if you charge the battery,
>> > unplug it and run until the car stops, does the battery
>> > still have 40ah of (inaccessible) charge remaining in it,
>> > or is that 40ah lost as heat generated by the rapid
>> > discharge?
>>
>> Yes and no.  Much of the extra is wasted as heat.
>
> Actually, just no. ;^>

Oops, you're right I was thinking WHs.  However, I'm pretty sure the full
capacity of the battery (the remaining 100+ AH) is not available even at
lower discharge rates once you've drained it at the higher rate.
I've never tested this though...

> There are *no* "extra" Ah wasted as heat; the energy wasted as heat shows
> up as a reduction in the usable Wh, not as additional Ah that went
> somewhere else.  The lost energy manifests itself as a reduction in the
> terminal voltage, which reduces the amount of usable Wh of energy
> available due to removing some amount of Ah from the battery.  The losses
> are I2R; removing 100Ah @ 100A results in greater voltage sag, less usable
> energy (Wh), and more heating than removing the same 100Ah at 10A.
>
>> However, this is NOT what they mean by 80% DoD.  80% DoD
>> means 80% of the AH available AT THAT DISCHARGE LEVEL!!!  If
>> you're averaging about 105 amps (the 1 hr rate) then you
>> should only remove 80% * 105 AH or about 84 AH.
>
> Right, except that the 1hr capacity is about 121Ah, so 80% of this is
> 97Ah.

...calculated using Peukert and data from the 2hr and 5hr portions of the
curve.

If you have Bob Brant's book "Build your own Electric Vehicle", page 232
has a chart showing discharge curves for several different batteries. The
chart was developed from Trojan's data (they used to publish a lot more
info on their batteries)
It shows a 1hr capacity of approx 118 AH.

Take your pick, any of the numbers (Your's, mine, or Bob's) are as likely
to be correct for a given battery as any of the others.

If you don't have Bob's book, you can see the curve:
http://books.google.com/books?id=XcFbEX-de4kC&pg=PA232&lpg=PA232&dq=trojan+T-105+discharge+curves&source=web&ots=opH9E6txaO&sig=Bv7dKZB74qbn8BIcPBofRGVYAII&hl=en&sa=X&oi=book_result&resnum=8&ct=result

>> A typical 120V 400 amp motor/controller will produce about
>> the same power as an old VW bug.
>
> But not necessarily the same performance ;^>

True, the EV typically won't be quite as peppy as a bug because it's
heavier (1300 lbs of batteries will do that)

>
> For instance, my 120V EV uses a 450A controller and ADC 8" motor.  The
> stock engine in my car was rated 48HP @ 5100RPM and 57ft-lb @ 3200RPM.
> 450A into an ADC 8" motor results in about 72ft-lb of torque, and this is
> available starting near 0RPM.  By keeping the EV weight not too much above
> the original, the result is that my EV actually has superior driving
> performance than the original ICE (and an old VW bug ;^).

Have you actually raced a Bug?  I don't doubt you could take it in the
first 50 ft, but after that I think it'd be a toss up.
If your battery is at say, 50% DoD, I'd bet on the bug.
In a race from AZ to OR, I'd definitely bet on the bug :-)

>
>> Step up to a 144V 1,000 amp controller (and some high power
>> batteries) and you're up to the level of a modern economy car.
>
> Here again, the peak power numbers may be similar, but that doesn't
> necessarily mean the performance will be.  The difference in torque
> characteristics between a DC EV and ICE result in different HP/lb
> performance scales, as shown time and again by the EV drag racers.

I was giving the EV the benefit of this.  A new Ford Focus comes with a
140hp motor.  A Z1k with a 9" ADC a 144V pack of stiff AGMs will put out
maybe 110HP, but only when they are fully charged, and only for a few
minutes.

The continuous motor current on the Z1k is only 300 amps which works out
to  perhaps 50 HP (at 144v), again assuming a fully charged pack.



_______________________________________________
For general EVDL support, see http://evdl.org/help/
For subscription options, see http://lists.sjsu.edu/mailman/listinfo/ev

[txhokie4life]

I'm only a couple weeks ahead of you on the learning curve -- but I'll answer what I can

lumberjack_jeff wrote:

Thanks to everyone who contributes to raise the collective knowledge. My knowledge probably needs raising to a degree greater than most. ;-)

I have a number of questions.

A Trojan T105 battery can deliver 185ah in 5 hours or 225ah in 20.

1) If full discharge = 225ah, but the battery can't deliver any more than 185 of that in one commute, does this imply that an EV can't discharge a battery much below 80% DOD in a single trip? In other words, if you charge the battery, unplug it and run until the car stops, does the battery still have 40ah of (inaccessible) charge remaining in it, or is that 40ah lost as heat generated by the rapid discharge?

As I understand it -- you don't want to go beyond 80% DOD -- It can damage the batteries taking them down to 100% -- or at least shorten their life.  Also (what I have seen from LiFePO4 simulations) as you get to a point there is a steep
knee in the curve of supplied voltage versus remaining current capacity.  So you'll go from 80% DOD to 100% DOD VERY
quickly.

I'm also confused about packs constructed of the various types of Lithium Ion cells available. An available type of LiFePO4 cell is advertised as 3.2v, 1100mAh, $3.20 ea in quantity.
www.voltphreaks.com/ssl/catalog/product_info.php?products_id=31

At risk of embarassing myself with my incomplete knowledge:
a) to make a pack equivalent to the usable kwh in a pack of 24 t-105's (144v@225ah*.8) requires 8000 of these cells??? (3.2v@1.1ah*.8=2.8wh each) That's $25,000.

There are some devils in the details - but you basically have the concept correct -- there are some other "lossse" such
as IR drop and cell voltage droop that means you will need more than rated values.

b) clearly this pack would not be arranged as a single string. How would it be set up?

very simple really -- you add batteries in series to get Voltage, and in parallel to get current.

Think of them as LEGOs -- you have gang them up many ways, but in the end you'll need the same number of
individual cells. -- also note there is a need for Battery Management Systems (for any non Pb cells), and they may dictate
how you gather your battery sub-blocks.

c) Ian described these cells as a comparatively good value, but if my understanding of their capacity is correct, they cost $0.91/wh. Yet, his review of these headway 
cells indicates that they cost $0.67/wh. In comparison, golf cart batteries appear to cost between $.07 and $0.15/wh. Do lightweight/long life solutions really cost this much more?

What I am finding is your mileage will vary on all these specs.....   What I have researched and been told
by many on this forum is that is true that the Pb (lead-acids) will be cheaper,
but they are heavy, generally have less Energy density (read less range, lower Ahrs), and will likely have a shorter
lifespan than say LiFePO4.
However it seems that Pb can give you more short term Energy (read better 0-60 performance (i.e. Higher peak Amps)

d) to get "6cyl" performance from a motor, are "high power" (as opposed to high capacity) cells really required?

Probably true -- see the discussion I created "So what do I really need?"

I feel as a student in a foreign language feels when hearing something that indicates that what he thought he knew about the meaning of words is not the actual meaning.

My background is mechanical - not electrical. Can someone please point me to an online resource to make me less ignorant?

Amen -- once you get started -- you'll be way better off than me -- I'm Electrical and not mechanical :-)

good luck,

Mike




Thanks
Jeff

[ev-4]

> 1) If full discharge = 225ah, but the battery can't deliver any more than
> 185 of that in one commute, does this imply that an EV can't discharge a
> battery much below 80% DOD in a single trip? In other words, if you charge
> the battery, unplug it and run until the car stops, does the battery still
> have 40ah of (inaccessible) charge remaining in it, or is that 40ah lost
> as heat generated by the rapid discharge?

As I understand it, the energy is lost.

Bill


_______________________________________________
For general EVDL support, see http://evdl.org/help/
For subscription options, see http://lists.sjsu.edu/mailman/listinfo/ev

[Roger Stockton]

ev@... wrote:

> > 1) If full discharge = 225ah, but the battery can't deliver
> > any more than 185 of that in one commute, does this imply
> > that an EV can't discharge a battery much below 80% DOD in
> > a single trip? In other words, if you charge the battery,
> > unplug it and run until the car stops, does the battery
> > still have 40ah of (inaccessible) charge remaining in it, or
> > is that 40ah lost as heat generated by the rapid discharge?
>
> As I understand it, the energy is lost.

Nope, most of that unused energy is still in the battery.

There are a couple of things going on here:

- the *usable* energy available from the battery depends on the rate at which you try to take it out of the battery.  That is, the higher the discharge rate, the less usable energy that is available from the battery.  So, at some low discharge rate, such as 11.25A, a battery might be caapble of delivering 225Ah of usable energy, but it takes 20hrs to extract this energy.  In an EV, we typically discharge at much higher rates, such that less usable energy is available before the battery voltage under load falls to the value considered to indicate "empty" (1.75V/cell).  At 75A, this 225Ah battery might only deliver 140Ah.  The rest of the energy (225-140 = 85Ah) is mostly still in the battery, it is just that you can't take it out at 75A without the battery voltage sagging even lower.  If you reduce the discharge rate to a lower level, then you can remove some more energy from the battery before its votlage again falls to "empty".  You can keep reducing the discharge rate in t!
 his manner until the battery can no longer supply enough current to move the vehicle without the voltage falling below "empty".

- batteries tend to last longer if you don't drain them completely empty, so while a battery might be rated 225Ah (over 20hrs), you generally wouldn't want to use more than about 80% of this (180Ah over 20hrs) on a regular basis or the battery life will be shortened.  This battery might only have 140Ah available at a more EV-useful discharge rate of 75A, and so to limit the discharge to 80% one would only remove about 112Ah @ 75A.

Hope this helps,

Roger.


_______________________________________________
For general EVDL support, see http://evdl.org/help/
For subscription options, see http://lists.sjsu.edu/mailman/listinfo/ev

[Roger Stockton]

Peter VanDerWal wrote:

> A T-105 has about 105 AH available at the 1 hr rate.

This is incorrect.

The "105" in "T105" refers to minutes of reserve capacity at 75A (though a modern T105 has been improved to spec at 115min now).

Taking the published 5hr capacity of 185Ah and the 115min @ 75A reserve rating and plugging them into Uve's calculator yields a Peukert exponent of 1.357 and an estimated 1hr capacity of 121Ah.

> > 1) If full discharge = 225ah, but the battery can't deliver
> > any more than 185 of that in one commute, does this imply
> > that an EV can't discharge a battery much below 80% DOD in
> > a single trip? In other words, if you charge the battery,
> > unplug it and run until the car stops, does the battery
> > still have 40ah of (inaccessible) charge remaining in it,
> > or is that 40ah lost as heat generated by the rapid
> > discharge?
>
> Yes and no.  Much of the extra is wasted as heat.

Actually, just no. ;^>

There are *no* "extra" Ah wasted as heat; the energy wasted as heat shows up as a reduction in the usable Wh, not as additional Ah that went somewhere else.  The lost energy manifests itself as a reduction in the terminal voltage, which reduces the amount of usable Wh of energy available due to removing some amount of Ah from the battery.  The losses are I2R; removing 100Ah @ 100A results in greater voltage sag, less usable energy (Wh), and more heating than removing the same 100Ah at 10A.

> However, this is NOT what they mean by 80% DoD.  80% DoD
> means 80% of the AH available AT THAT DISCHARGE LEVEL!!!  If
> you're averaging about 105 amps (the 1 hr rate) then you
> should only remove 80% * 105 AH or about 84 AH.

Right, except that the 1hr capacity is about 121Ah, so 80% of this is 97Ah.

> A typical 120V 400 amp motor/controller will produce about
> the same power as an old VW bug.

But not necessarily the same performance ;^>

For instance, my 120V EV uses a 450A controller and ADC 8" motor.  The stock engine in my car was rated 48HP @ 5100RPM and 57ft-lb @ 3200RPM.  450A into an ADC 8" motor results in about 72ft-lb of torque, and this is available starting near 0RPM.  By keeping the EV weight not too much above the original, the result is that my EV actually has superior driving performance than the original ICE (and an old VW bug ;^).

> Step up to a 144V 1,000 amp controller (and some high power
> batteries) and you're up to the level of a modern economy car.

Here again, the peak power numbers may be similar, but that doesn't necessarily mean the performance will be.  The difference in torque characteristics between a DC EV and ICE result in different HP/lb performance scales, as shown time and again by the EV drag racers.

Cheers,

Roger.


_______________________________________________
For general EVDL support, see http://evdl.org/help/
For subscription options, see http://lists.sjsu.edu/mailman/listinfo/ev

[Roger Stockton]

Peter VanDerWal wrote:

> Oops, you're right I was thinking WHs.  However, I'm pretty
> sure the full capacity of the battery (the remaining 100+ AH)
> is not available even at lower discharge rates once you've
> drained it at the higher rate.
> I've never tested this though...

I haven't tested it either, but my understanding is that if one discharges at a high rate until the voltage reaches "empty", and then reduces the rate to the 20hr rate and continues discharging, the total Ah removed will be nearly the same as if the battery ahd been discharged at the 20hr rate the entire time.

The key point is that while the usable capacity at a high rate may be 1/2 of the rated 20hr capacity, this is not because the other 1/2 of the 20hr capacity has been consumed somehow.

> > Right, except that the 1hr capacity is about 121Ah, so 80%
> > of this is 97Ah.
>
> ...calculated using Peukert and data from the 2hr and 5hr
> portions of the curve.
>
> If you have Bob Brant's book "Build your own Electric
> Vehicle", page 232 has a chart showing discharge curves for
> several different batteries. The chart was developed from
> Trojan's data (they used to publish a lot more info on their
> batteries) It shows a 1hr capacity of approx 118 AH.
>
> Take your pick, any of the numbers (Your's, mine, or Bob's)
> are as likely to be correct for a given battery as any of the others.

I have Bob's book, but his data is from 1992, I would not rely on it to be completely representative of current production batteries, and I wouldn't count on getting the performance shown in his graphs even if you can locate a set of 16-year old T105s ;^>

My estimate may not be super accurate either, but at least it is based on published specs for product you can buy today, and appears to be supported by Bob's 16-year old data (one would expect present T105s to be at least as good as those in 1992, and indeed my estimate suggests a small improvement).

> Have you actually raced a Bug?  I don't doubt you could take
> it in the first 50 ft, but after that I think it'd be a toss up.
> If your battery is at say, 50% DoD, I'd bet on the bug.

And I'm confident you'd lose. ;^>  I know you're used to floodies, but I run AGMs, and only 450lbs of them, not 1300.  My curb weight is about 2000lbs.

'Course, why stop at 50%DOD, make it 100%DOD and we can both bet on the bug! ;^>

> In a race from AZ to OR, I'd definitely bet on the bug :-)

You and me both! ;^>

> >> Step up to a 144V 1,000 amp controller (and some high power
> >> batteries) and you're up to the level of a modern economy car.
> >
> > Here again, the peak power numbers may be similar, but that doesn't
> > necessarily mean the performance will be.  The difference in torque
> > characteristics between a DC EV and ICE result in different HP/lb
> > performance scales, as shown time and again by the EV drag racers.
>
> I was giving the EV the benefit of this.  A new Ford Focus
> comes with a 140hp motor.  A Z1k with a 9" ADC a 144V pack of
> stiff AGMs will put out maybe 110HP, but only when they are
> fully charged, and only for a few minutes.
>
> The continuous motor current on the Z1k is only 300 amps
> which works out to  perhaps 50 HP (at 144v), again assuming a
> fully charged pack.

The Focus you refer to *is* spec'd at 140HP... at 6000RPM.  It is also spec'd at 136ft-lbs @ 4250RPM.  An ADC 9" motor fed 1000A is expected to deliver about 237ft-lbs from near 0 RPM; and will still be developing at least 100ft-lbs around 3000RPM; this *will* affect the actual performance in favour of the EV.

I suppose all we can do is agree to disagree here; my experience suggests that this hypothetical EV would offer real world performance somewhat above that of the Focus, though I suppose if the Focus driver's typical commute involves dropping the clutch at 3000-4000RPM and winding the ICE somewhere north of 6000RPM before grabbing the next gear that it could be close ;^>

Cheers,

Roger.







_______________________________________________
For general EVDL support, see http://evdl.org/help/
For subscription options, see http://lists.sjsu.edu/mailman/listinfo/ev