There are two types of Lithium Batteries I can supply:
Just a note about the terms Lithium Ion and Lithium
Polymer: as far as I can determine lithium batteries were initially made
with a solution of electrolyte in a sealed casing: they were called Lithium Ion
type. Later on instead of using a solution a mixture was made using some plastic
polymer to bind the ingredients together into a more solid mass. Though that
solid mass still acts as an electrolyte for transfer of electrons from positive
and negative parts of the battery.
So if a battery is referred to as a lithium polymer it is referring to the
plastic matrix that holds the battery into a solid block ( it seems some iron
phosphate batteries have the polymer in sheets rather than a sold matrix). It doesn't tell you
anything about the type of lithium battery it is, simply that it is not a liquid
type lithium battery. There are three main different lithium chemistries ( these
days all will be polymer bonded). Magnesium oxide, cobalt oxide and iron
phosphate. All three types are used in ebikes, iron phosphate being the latest
type to emerge. Mobile phones tend to use cobalt oxide type as far as I know .
Some lithium battery cells are in a metal outer case, some are in a plastic
outer case ( very thin ), that plastic outer case can also be called polymer but
doesn't really mean anything in terms of if its a lithium ion or lithium polymer
battery.
There are two types I can provide for use in ebikes ( or other uses ):
1. cobalt oxide
2. iron phosphate
note: I've never used magnesium oxide type though they are being sold for use
on ebikes. I have read that manganese oxide is more dangerous type of battery
than cobalt oxide but I dont really know much about that.
There are pros and cons for each type of battery.
I think the main difference is that iron phosphate is safer than cobalt oxide.
Cobalt oxide under certain cirmcumstances can catch fire, whereas iron phosphate
the cell will be damaged but no fire.
There are also differences in weight, cobalt oxide being lighter. Below shows a
comparison of cobalt oxide on left and iron phosphate on the right.
The cobalt oxide in this case shows a 24v 11ahr battery pack, I have outlined in
black two cells which together make up 3.7v 11ahr. Those two cells are close to
the equivalent capacity of the blue iron phosphate cell on the right. The cobalt
cells on the left are not in metal cases but only have a plastic outer case.(
there are cobalt oxide cells that have metal outer casing) which are a little
heavier than the ones shown below.
The iron phosphate batteries I will supply will have plastic pouches rather than
metal outer cases in order to minimise weight, the iron phosphate cell below in
a metal case weighs 475g, the ones in plastic pouches weigh 340g considerably
lighter.
1. Cobalt Oxide Lithium Batteries:
The first lithium battery I tested and then sold a few units was cobalt oxide
type. At the time I had no idea it was cobalt oxide, my initial contact with the
manufacturer I had been very careful to confirm with them that it was iron
phosphate type I wanted and not cobalt oxide. So they had told me the battery
was indeed iron phosphate so I felt it would be relatively safe, but I tried to
be cautious about who I supplied the battery to, that they would be very careful
with them.
Not so long ago whilst discussing the batteries on 'endless sphere' forum it was
pointed out to me that the battery voltage was too high to be iron phosphates, I
subsequently asked the manufacturer to confirm that they were indeed iron
phosphate, this time they told me they are in fact cobalt oxide. So my apologies
for people who recieved those batteries believing them to be iron phosphate, I
also thought they were iron phosphate which turned out to be not the case.
There was in fact some good things to come out of it though. There have been no
fires occur with the batteries up to this point in time. There are probably
around 20 or so of those batteries being used by customers. Though I have also
learned over time that there have been many fires involved with cobalt oxide
type lithium batteries being used in ebikes ( there have also been many fires in
mobile phones and laptops where they are also used, a quick google search you
can find much information about lithium cobalt fires). Noteably the dell company
laptop fires got alot of attention and subsequently dell is working with valence
company (iron phosphate lithiums)I guess to make safer.
There is one ebike company that I know has had fires using lithium cobalt oxide
batteries which I've been fortunate enough to be able to have a play with their
batteries. The reason for the fires seem to have occurred during overcharging of
the batteries. The battery monitoring system ( bms board ) on those batteries
was very primitive in design. It seems the bms board is quite crucial ( along
with a charger that cuts off when the batteries are charged ) in keeping cobalt
oxide lithiums safe. If one cell in a battery pack were to , for some reason,
not charge up, it will bring down the voltage of the entire pack, the charger
may not turn off as it thinks the batteries are not charged ( the voltage isn't
high enough to cut off ), so the charger will keep going , resulting in
overcharging the good cells and resultant fire.
So the bms board needs to be able to recognise if one the of the cells is not
charging up and keeping the voltage of the pack low. A good bms board should be
able to protect the pack from overcharging . Unfortunately I dont know enough
about electronics to understand how this is achieved, whether it is done by the
bms board alone or also using the charger itself. Anyway, I think I have been
extremely fortunate in that the bms board used in the cobalt oxide batteries I
have supplied seems to have been a very advanced design and no fires have
occurred.
So given that the bms and charger look very good for the 24v 11ahr packs I will
continue to supply these batteries to people who are interested in them. But I
cant send them out by air freight, there will always be the risk of fire with
these batteries, it is something that cannot be ignored that under certain
circumstances the cells can still catch fire even with a sophisticated bms board
and charger.
Here are two scenarios where a fire could occur:
1. water gets into the pack and shorts out the cells. The bms board could not
stop a fire or shorting in this case. If the terminals of the cells were to get
enough moisture that they shorted a fire will result. ( lithium cobalt fires are
particularly hot flames and quite exothermic, objects within maybe 2 metres
maybe at risk of damage). Fortunately lithium cobalt does not require any
atmospheric gases to operate, so the cells can be completely sealed from the
outside air and risk of water/moisture causing a short is probably quite small.
2. a metal object punctures the cells and causes a short: lets say an accident
occured and a metal object punctured the case which holds the battery cells,
this could cause a short in the cells and consequent fire. Probably the only
precaution would be to put the batteries in a very strong metal case which may
survive a severe accident.
So I would be quite ok to still supply the 24v 11ahr cobalt oxide battery packs,
but the buyer would need to have very good technical skills in order to wire
them up to an ebike and the charger. It is not particularly simple to wire them
up and great care is needed to make sure all connections are secure and correct.
There have been a few incidents with the batteries I have supplied ( none of
which resulted in fires), below are the different incidents that have occurred:
( thankyou murphies law!)
1. nick in sydney: first bike ride, hit a bad pot hole at speed, both
batteries flew out of the battery box and bounced down the road ( to Nicks
shock!). Batteries undamaged working as normal, some minor damage to outer
case holding the batteries. The batteries have been protected extremely well by
the plastic box they come in. After having taken apart a box ( with much
difficulty ) I have found they have used fibreglass ( or similar material ) to
hold the batteries in place in the box.
The batteries cannot move at all inside the box, they cannot be taken out
without destroying the outer box. The bms board and board that connects the
battery terminals together have been designed with shock resistance in mind.
Where the bms board attaches to the battery connections board, the soldered
joints have been then glued with some silastic to hold the wires in place in
case of shock abuse. The bms board itself has a piece of rubber which sits
against it, further reducing the chance of impact damage. So overall I'm
very impressed with the construction of the batteries and battery box.
2. one customer ( to remain nameless!!) accidentally put 240v directly into the
batteries. There was no resultant fire, the bms board was damaged ( needed to be
replaced) but it protected the cells from recieving 240v.
3. one customer had not been aware of the need to keep maximum amps taken from
the batteries to an acceptalbe level. 15amps is about the peak amp draw the
batteries should have over the long term, 20amps they seem fine with it. The
customer had been using a 30amp controller and riding their bike very hard, so
its likely the batteries were supplying 30amps on a regular basis for prolongued
periods ( apparently the customer was having a blast!!). When I accidentally
found out that the battery was being used in this way I informed them it was not
a good idea. The bms board has a overamp cut off of about 34amps. But it is good
to know the batteries survived apparently without damage be used in that way for
some time. Whether or not their longevity has been effected I have no idea at
this stage.
4. there were a few bms boards that stopped working reason I have not been able
to determine. There were no resultant fires, the batteries basically just didn't
supply power anymore.
5. the batteries have been used in series: 1. 2 packs in series ( to give 48v
11ahr): no problems
2. 3 packs in series to give 72volt 11ahr ( really is about 85volt when charged
up), I have used this way without incident and other customers. One
customer though had two lots of there in series packs in parallel, and two bms
boards failed ( I'm not really sure of the circumstance of how/when they failed
though).
3. one customer has used 4 packs in series ( 115v 11ahr) without incident thus
far.I have not tried this as yet, but shall do when I have a suitable controller
to try it.
6. One customer reported that one cell had a lower voltage than it should
have had, I'm yet to ascertain if that is correct or not, but I think the pack
is still in use, so it may have been an incorrect report, not sure on that one
as yet.
Functioning of the bms board:
The bms board has quite a complicated circuit. It uses surface mounted ( very
small ) transistors/resistors/capacitors and mosfets to monitor the battery. The
cells are arranged in pairs ( two 5.5ahr cells together to give 11ahr). These
pairs are treated like one cell by the bms system. Each of these pairs is
connected to the bms board by a thin wire which acts as the charging wire and
also probably plays a role in determining cell voltages.
The mosfets are switches which can be turned on and off depending on what the
circuitry detects. Some functions of the bms are :
1. protect the battery from overdischarge: the mosfets will turn off the battery
when the pack reaches about 21.5volts. This had been confirmed simply when
riding the batteries till the cut off kicks in. Some loss of power can be
noticed when the pack is getting low , its not a huge loss in power but is
noticeable. At some point the pack will just turn off if it reaches 21.5volts.
2. protect the batteries from overdischarge: the bms will cut off power if the
amps reaches 34amps. This was accidentally determined one night i was riding
home is quite heavy rain ( after a few beers........silly me!!). The rain had
shorted out my wiring ( which was largely exposed to the elements ), at some
point the bike power cut out during the ride, the next day fortunately the drain
brain meter had recorded the max amp reading at 34amps. So that would have been
the point at which the bms kicked in. The batteries themselves were wrapped in
clear tape to protect from water, though water did get under the clear tape but
the plastic box has not let anywater get inside it. Batteries were undamaged
still using them.
3. the bms board may play some role in preventing overcharging, but there have
been no incidents of that occuring and I'm not sure what role it plays. It is
probably a crucial thing to understand as that is when most fires I've heard of
occur during overcharging. The chargers that come with the batteries will cut
off when the battery is fully charged. I wish I had more understanding of how
this is achieved, but i'm guessing its the charger is doing this without any
role from the bms board.
4. the bms board does have overtemperature cut off switches, I think is probably
the last line of defence, if a cell does get over 50degrees celcius the bms
board will shut off any charge current ( if its being charged ) or any outgoing
current ( if the battery overheated when being discharged for some reason).
There are three cut off switched, one on the bms board itself ( which is set in
resin ), and two on the sides of the top of the cells . The temperature switches
that are sitting against the cells should recieve any heat from the cells quite
efficiently as the outercase of each cell is very thin plastic. Some heat
transfer paste between the temperature cut off switches and the cells may be the
only thing i could see might be an improvement but the heat transfer paste may
react with the plastic cell outer membrane so maybe not possible to use it. The
temperature cut off switches are pushed inbetween the outer case of the pack and
a cell, so there is quite a bit of pressure on the temperature cut off switch
and the cell so heat transfer to the cut off switch should be quite good.
The reason i'm going into quite a bit of detail about the temperature cut off
switches is that I'm aware of them being retrofitted to one taiwanese companies
bms sytstem ( after quite a few fires occurred) in order to protect the
batteries against overcharge. But their attempt at a retrofit was a waste of
time, the switches were not making good thermal contact with the cells and they
were rated at too high temperature rating. Also I tested one of the temperature
sensors they used by puting a lighter flame directly onto it, it took about
30seconds of naked flame to get the switch to click, probably way to late to
protect against a fire.
Below: burned out remains of a cobalt oxide 24v 18ahr pack, initial cell to
catch fire has ignited adjacent cells. Fire was pretty much contained by an
aluminium outer jacket not shown in picture. Insert one shows retrofitted
temperature cut out switches which have not worked to protect the pack. Insert
two shows a basic bms board too simple in its design to be effective bms. (
batteries from edan company in taiwan picture taken in my backyard, fire
occurred elsewhere this is just remains , by the way these are not the cobalt
oxide packs I've been selling!!)
Below shows a good bms board ( used in the 24v 11ahr packs I have been using
and can supply ). I can also supply these bms boards if anyone is looking for
some, chargers also can be supplied with the boards. Bms board suits 24v cobalt
oxide lithium batteries. Should be suitable for use with a variety of ahr size
packs.
I am prepared to continue supplying the cobalt oxide 24v 11ahr packs. Only to
people who are very skilled with electrical wiring and have some knowledge on
lithium batteries and are prepared to take all precautions necessary to ensure
if a fire occurs that the batteries are kept in a place where no damage will
occur to persons or property. I can only send them out by sea freight, there is
no option to send these batteries by air freight due to risk of fire.
For more detailed information on the cobalt oxide batteries see this page:
http://www.users.bigpond.com/solarbbq/lithiums.htm
Iron Phosphate Lithium Batteries:
Hopefully will prove to be a very safe lithium battery for ebikes or
other electric vehicles. When in use the cells have a voltage of around 3.2volts
when fully charged the voltage can be somewhat higher ( about 3.1volts ).
Depending on the size of the cells used ( capacity ) the metal cases may have
more than one individual cell inside it. For example 5.5 to 6 amp hours seems to
be a basic size cell. So a pack of 11 to 12ahr will have two lots of 5.5 to 6 ahr cells in the one metal case. Similarly a 18ahr ( approx ) cell ( metal
casing) may have 3 cells inside it.
The iron phosphate cells can come in metal cases as shown below or in a plastic
pouch, the plastic pouch will be much more lightweight, but as yet I haven't
sourced or tested that type of iron phosphate cell. So at the moment the weight
is quite high due to the outer metal cases housing the iron phosphate cells.
The cells are then put together in series (and/or parallel) to form a battery pack. The pack
below consists of 8 lots of 3.2volt cells ( each cell in this battery below is
about 12ahr). The voltage of this pack is 8 times 3.2volts = 25.6volts ( it is
used as a 24volt pack). Similarly a 36volt pack will use 12 cells and a 48volt
pack will use 15 or 16 cells.
Iron Phosphate batteries can be substantially heavier than lithium cobalt
batteries in part due to the outer metal casing the cells are contained in. The
pack below weight is approx 5.5kg ( compare with about 3kg for an equivalent
cobalt oxide pack not using metal outer casing to house the cells).
I was under the impression that iron phosphate batteries would not require a
sophisticated bms board ( battery monitoring system) but I appear to have been
quite mistaken on that front. Each cell needs to be charged separately from the
other cells. In the picture below you can see quite alot of wires on the top of
the battery pack which are used to charge each cell individually.
The bms boards on iron phosphate batteries appear to be very complex. They
will have quite a few functions to protect the cells from the following:
a. over discharge ( too high a current being taken from the batteries)
b. over charge ( too high a voltage occuring during charging )
c. under charge ( too low a voltage occuring during discharge of batteries
during use i.e. low voltage cut off)
Temperature cut off switches seem to be used in cobalt oxide batteries but not
in iron phosphate batteries.
The bms boards use mosfets ( which act like relays ) which can turn off the
battery output or incoming charge if any of the conditions above occur.
These batteries are quite a new product to enter into the ebike market. There
are many companies now supplying them in china. The most well known brands are
by Valence Technologies and A123 company. Both of those companies have been
making for some time and produce cylindrical cells which can be put in series or
parallel to make up a battery pack. Unfortunately with both those companies is
not possible for me to get cells or packs from them as they only deal with large
companies. The A123 cells can be found in dewault brand powertools which
quite a few ebike enthusiasts seem to be using. The valence technology cells can
be seen in use in segway transporters and in new toyota prius.
There are many other companies making iron phosphate battery packs in china.
Some are only very recently making these cells and are still having problems
with some issues involving charging and bms boards. I'm not sure at this stage
how much difference in quality there is between maufacturers but will test a few
differerent ones and see the results.
I think really they are still very
experimental in nature ( excepting valence and a123 and maybe a couple of other
companies who have been doing for some time), so I'm still not all that
confident in how long will be the lifespan of the cells produced by newer chinese
companies, I guess thats something that remains to be seen.
There also seems to be a wide variation in what is the acceptable maximum
discharge rate of the cells from company to company.
It does appear though that iron phosphate wont catch fire if a short occurs
somehow in the cells. Recently, in a pack coming from china, one cell had
shorted somehow. The reason for the short was the positive charging wire on the
end cell had made contact with the outer negative metal casing of the cell, and
shorted the cell. I think it was extremely small odds of this happening but
having a cell which has been shorted was an accidental test if these batteries
will actually catch fire.
So it seems they dont catch fire! I'm not sure how hot the cell would have
gotten but there was certainly quite a bit of gassing as the cell had expanded
its metal case quite a bit. The metal cases are quite rugged and contribute to
much of the weight of the pack. The cell was part of a pack in a styrofoam box,
there was no evidence of melting or heat damage to the styrofoam box. The only
signs of significant heat build up is around positive terminal near where the
short occurred.
I am prepared at this stage to supply iron phosphate batteries. Voltages from
12v up to 48v, capacity from 6ahr upwards. But I will only supply to people who
have good electrical knowledge and who are well aware of the experimental nature
of these batteries. Iron phosphate can be sent out air freight but I would much
prefer to send out sea freight to be on the safe side.
A few other examples:
1. ebike shop in woy woy nsw burned down due to lithium fire ( overcharging
thermal runaway cobalt oxide type)
aware of two other ebike shops in australia had lithium fires
2. china: wuxian lithium battery on an ebike caught fire and injured the bike
rider, was on chinese tv news broadcast ( my workman relayed that one to me)
It is extremely unlikely that retailers of ebikes will be forthcoming about any
fires that have occurred, also manufacturers are in the same situation. I think alot
of businesses, manufacturers and retailers have jumped in very quickly with a
view to making alot of money quickly, but has been a high price to pay for it.
I think is worth keeping in mind.
Will put up new info below as I do experiments:
Whats inside an iron phosphate lithium cell:
The estimated widths below are not accurate just my guesstimates, the copper
layer is very thin, probably only a couple of microns,
the polymer/resin layer is even thinner probably not even 1 micrometer thick.
The iron phosphate layer is probably even thinner again,
and probably the reason why one cell make refers to its cells as nanophosphate,
the actual layer of iron phosphate is extremely thin indeed.
Sitting in the garage for a few days with the black iron phosphate layer
exposed to the atmosphere there were visible drops of water sitting on it,
I think it must be very hydroscopic material.
Internals of shorted cell: August 5th, 2007
During transport from China, it looks like a cell had shorted, the
outer casing is in contact with the copper of the cells inside, and also forms a
negative
terminal, somehow a positive wire had touched the case and shorted a cell, nice
to know the cell didn't catch fire!! ( it looked impossible for this to happen
looking at how the cells and wires are attached, but it happened anyway).
But the cell was rendered unusable, the cells inside the metal case had expanded
and expanded the outer metal case quite alot.
The outer metal case is sealed from the outside atmosphere, but even if
oxygen had got into the cell, the iron phosphate is supposed
to not catch fire. I will try an overcharge experiment on a good cell that is
exposed to the atmosphere soon and see what temperatures the cells
reach.
Opening up an individual cell:
Some pics below of whats inside individual iron phosphate cell ( will put some
pics up of insides of a cobalt cell when I get around to opening one up):
Each cell ( for example the ones in metal casings above) is made of smaller cell
units inside, the example above a 3.2v 12ahr cell, has 7 lots of smaller cells
inside it, so it cell would be approximately 3.2v 1.7ahr. Below left shows 4 of
the smaller cells. Picture on right shows starting to unroll the cell.
More pics of unrolling the cell.
there are two different metal backing sheets used. One is copper the other
i'm not sure what material it is, but its silvery in colour and looks like
aluminium but is probably not aluminium, not sure what it is. Also I noticed the
black deposit on the silvery metal and the copper have different properties, one
is extremely hydroscopic and attracts alot of moisture out of the air (visible
as water droplets on the black lithium iron phosphate material) thats the
silvery metal backing one, the copper backing one draws less moisture from
atmosphere so looks like they are not exactly same black material but not sure
what they are composed of. Guessing they are both some sort of lithium iron
phosphate material.
Below a few examples of lithium fires:
Reason I'm puting this info up is that anyone considering purchasing lithium
batteries should be well aware of the hazards and do their research before
purchasing.
Seems the ebike industry is full of people who will sell products they know are
dangerous and potentially can lead to loss of property or even life. I think
lithiums should be treated as highly experimental and potentially very
dangerous, especially cobalt oxide and mangnesium oxide varieties. Iron
phosphate are supposed to be safer, but it is even very early days for that
battery type, so I wouldn't completely discount the possibility of problems
arising with iron phosphates.
The number of fires that have occurred is completely unknown and will never been
known, I am aware of a considerable number of fires (probably around 20 or so
I'm aware of which would be a tiny fraction of those that have occurred, I
suspect the number is very large indeed).
One thing to be remembered though if buying lithiums: there are people selling
lithium batteries who are well aware that a fire is a very real possibility and
are still selling same lithiums that they know have caught fire with exact same
bms etc.
below NSW Australia: 2007: manufacturer of cell "edan" ( taiwan ), type: cobalt oxide
cause: probably poorly designed bms board. Bms board had been retrofitted
with
temperature cut out sensors ( after previous fires ), obviously not effective.
Seems fires occurred on the positive end
of the packs of cells in series, not sure of reason it would start at that end.
If fire was involved with discharging (or short circuit ) you would expect cell
to catch fire to be at the negative end where
electrons are leaving the battery ( higher energy density ). So it was probably
something to do with the charging of the battery,
perhaps an overcharge scenario.
( note: although batteries have positive and negative terminals, the electrons
flow from negative to positive,
as the early electricity researches chose which side to call positive and
negative, would have made more sense
around the other way, some historical reason for it anyway)
France: I think was 2005 or 2006 cause: same brand as above picture,
probably poorly designed bms board. Was earlier version of the bms board
involved with the fire above.
How could overcharging a lithium battery occur using smart lithium chargers??
If a cell in the pack has lost its voltage for whatever reason, perhaps some
manufacturing defect, then it will pull down the voltage of the whole pack,
as charging occurs the cell may not increase in its voltage, and the pack will
never become fully charged. So a charger may just keep on charging thinking
the pack is not fully charged. It may be the charger being used was not
sophisticated enough to determine that it should have stopped charging. I have
no idea
how the chargers determine when I battery pack is full. I have noticed that the
chargers I have been using will not start to charge a pack that has a faulty
cell,
but no idea how it determines there is a faulty cell. would like to find out
though, maybe something to do with resistance of the pack and voltage combined?
beware of spray ebike company:
more info