Planing CanView and BMS+ conversion with lithium

Yep, it is pvc insulation. Somebody mentioned to me that the end of the pouch cell envelopes are not insulated and hence can short out. Maybe this is not the case with these A123 cells. Anyhow I have used this insulation to avoid/prevent shorting.

The C rates are small, no bigger that 120 amps (3C) which only happens momentarily, 40-50 amps about 1C on average. Regenerative braking charges up to 100 amps (2.5C)for about 5-10 secs before dropping out when HVC reaches 3.65volts.

Thermal management - where I am located the battery pack will not go any lower than 5 deg. Celcius and will not go much higher that 40deg.C. So maybe using aluminium plates between each cell to remove heat would be a good idea but I do not think it would be necessary. I guess the cells would age faster as you had mentioned in other post.

Long term mechanical integrity - I think you may need to expand on this as I do not know exactly what you mean.

I agree with what Planetaire says in his post that compression of the cells will not save your cells if they are overcharged and I do not see much point in them. At most they may prevent the cells from splitting and leaking electrolyte. At least that is what has been my experience with 20AH RFE pouch cells. I do not see much point in the compression at the low C rate.

I am planning to plaster the internal walls of the big "thunderbox" with polycarbonate sheeting and silicon sealant. I will try to make each cell group box air tight and the main box also to prevent contaminants from entering. This is important from a point of view of preventing a fire and possibly oxidizing.

I would not think there would have been much chance of corrosion for the metals I chose. Anyhow I can get this paste I mentioned from a previous post which would prevent oxygen from entering the connection.

http://api.viglink.com/api/click?format=go&key=53c9d790d7758c6d332c629ed0accb8b&loc=http://priuschat.com/threads/planing-canview-and-bms-conversion-with-lithium.99420/page-42&v=1&libid=1361863572182&out=http://ev-power.com.au/webstore/index.php/ev-parts/electrical-terminal-grease/noalox-anti-oxidant-joint-compound.html&ref=http://priuschat.com/threads/planing-canview-and-bms-conversion-with-lithium.99420/page-43&title=Planing CanView and BMS+ conversion with lithium | Page 42 | PriusChat&txt=noalox&jsonp=vglnk_jsonp_13618636687704

The supplier ran dry on the supply of these A123 20AH cells so some had the full tabs and some had a tab with a zipper fuse but the tab was shortened to about 10mm so I had no choice but to clamp some of the zipper fuse. Some have no markings. You can see a photo of a A123 cell with a zipper fuse a few posts further up this thread. I try to use at least two buddy cells with zipper fuses on each group.
Maybe a crimper which makes a zipper fuse can be purchased? Do you know of any?
I also will be using a 100A HRC fuse (house mains service fuse 100A used in Australia) to electrically split up the pack in the middle (or as near as possible) if a short occurs.

Just anderson plugs SB120 - 120A 600V. For each group and complete pack has a anderson plug which can easily be disconnected. Obviously it is up to the user to know that they should never be plugged/unplugged under load.

Thanks.

 

Do these cells have pressure release valves? pressure can build up in LiFePO4 cells and cause them to bulge and the plates etc to separate . .increasing cell internal resistance. Pressure plates stop this physical deformation from occurring.. At least with other cells .. Have not actually used A123 pouches .. Do the have a good technical reference and handling documentation ?

 

Does this mean you do NOT have 70 cells as a MATCHED battery set from at the factory ?

 

They don't have any valve. So if any "abnormal chemical reaction" append they will inflate. If they can not they will leak. So if pressing them please use a spring.
If they can they will have a huge thickness (like a balloon) but still stay closed.
And yes, if inflated, their internal resistance will increase.

But why pressing them ?
As say Lopezjm we use them at a "quiet" average amperage (40-50A). It is near 1C.
In my car cell temp increase is 10°C during a trip with full (90%) cell discharge, and only the cells that are near the middle of the packs.

 

My Question is what do A123 recommend in there documentation for these cells?

 

It depends what you mean by "matched". Ideally you want all your cells to come from the same production batch of cells as their internal resistance would be very closely matched if not almost identical. However my cells were probably stripped from an existing battery pack and maybe they came from different battery packs and their internal resistance may be slightly different. But all my cells are genuine A123 20AH cells and should meet the A123 specs.

 

Some cylindrical cells have the can tied to anode or cathode potential, others do not. A123's 26650 can is at cathode potential, the 32113 can is floating. I'm not aware of a pouch cell that has its bag strapped to either electrode. A123's pouch cells are fully insulated from both electrodes via multiple mechanisms (though a high-impedance breach of these mechanisms, leading to high self-discharge or decrepitation of the pouch, is a possibility with the "recalled" cells). A123 specs allow for a self-discharge rate of <2% SOC/month. That said, putting a physical barrier between cell groups is good practice for multiple reasons--electrical isolation being just one.

Your C-rates are conservative, so this would suggest air-cooling is sufficient. Low temps, you're OK (lucky you--this is something OEM's do not have the luxury to ignore). However, your max ambient operating temperature is enough to give pause, even at these C-rates. A few more degrees (which you should expect to come from the cells themselves), and you're into the possibility of mild amounts of electrolyte breakdown.

Now we get into why we want (need) stack pressure. You should never design a pack thinking "these cells will never see abuse, so I don't have to plan for that". Occasional, mild abuse is almost inevitable in an automotive application. A123's cells are remarkably abuse-tolerant and their design is intended to make the best of that. Sooner or later, small amounts of gas will form inside many of your cells. If stack pressure is sufficient, any gas that's formed works its way to the margins of the cell, outside of the electrode stack, where it does little to no harm to the cell's performance. Absent that pressure, the gas can build up between layers of electrodes and separator ("mechanical integrity"), where it will diminish performance much more dramatically. If the abuse gets severe enough, the seal and roll on the edge of the pouch will open. This is the design intent of the cell--the roll and seal are the de facto "safety vent", and for that to break open in the intended range, stack pressure is also necessary. I couldn't find a spec for recommended pressure. I don't think the level is super-critical, but I would shoot for 10-30 psi.

You're on the right path from an isolation/environmental protection standpoint, but I worry about heat management for a module in a sealed box. You don't even really have proper air-cooling at that point. I would seriously consider finding a way to push (clean) air through those boxes, at least above a certain temperature.

Aluminum can corrode pretty readily, and you should always assume water will be assaulting all connections. You will also have Cu in the anodes. The grease looks good--I would definitely use it. Remeber that in a HV system, tiny impedances add up to hot spots and performance degradation in a hurry. A dozen milli-Ohms can make a big difference.

Partial cell fuses are better than none at all. Sorry, I don't know of a tool for puching fuses into tabs, but I imagine your machining skills are good enough to reproduce the dimensions of ours.

Standard practice would integrate your center-pack fuse into a service disconnect which can be removed from outside the pack prior to any work on the pack. This insures the pack is always split when you're inside it. (I'm sure you already know this, others reading may not.)

I would also put a fuse in line with each module's Anderson connector, and at each far end of the pack, next to the contactors. This is also more-or-less standard design around here.

 

When cells are manufactured, even within a batch run, there are variations in capacity and IR .. So each cell is individually tested, cycled and performance data recorded and logged. Then cells are selected, matched and packaged (boxed) in Battery sets( to customer order) according to these performance and test result data to ensure the cells within a battery set are all as close as possible to each other in cell characteristics. This matching process is very important to the longevity and balance of a Battery pack.

 

In this case, I would change "is" to "can be". Pack performance will be limited to the performance of the weakest series element in the pack, but in this application I think that is as far as the consequences should go, barring a bona fide cell defect.

 

Oh, while we're on the subject of fusing...I also recommend fusing your V sense lines at a nice, low current.

 

The minBMS comes with a PTC resettable fuse. There is a PTC fuse on the PCB in line with each sensor wire.

There would be no point in adding further fuses to the sensor wires.

Thanks,

John

 

Just out of interest. What are the other reasons?

That is a good explanation.


OK, however spring loading is required so that the cells can expand to a point rather than burst at the seams and when the cell/s cool down (gas cools down and then contracts to a liquid?) the springs will push the cell back together. Is this theory correct?

I thought about using two S/S316 plates (one at each end) pressed together by two metal straps but it is not spring loaded. However my PVC insulation will allow for some cell expansion during any abuse.

I guess I could install a cooling fan/s pushing air into the box and controlled by a thermostat. However I would prefer to keep it airtight and install a couple of large heatsinks on the lid or front of main box with cooling fans to draw out heat. Anyhow I will install a thermostat just to see how hot it gets inside the box and see the results.
Keeping the main box airtight is a good objective, it prevents/avoids corrosion, dust build-up and so on that may cause a fire. I also have a auto-fire extinguisher from my old Enginer kit that I can install inside the box.

Thanks.

John

 

Just offhand:
1) Physical barrier between cells makes propogation through the pack of catastrophic cell failures more difficult.
2) Aids in sinking heat out of the module evenly. Were all cells just pressed together, the cells at the center of your module would get considerably warmer than those at the ends.
3) Helps distribute stack pressure evenly across the whole surface of all cells.
4) In your case, it puts space between those bulky joints where cell groups are joined together.

We use a combination of barrier materials. E-coated steel plates for rigidity, isolation, and heat sinking, and foam pads for evening out the stack pressure while making room for temp sensors (see below). The plates at the ends of our modules are thick, strong plastic.

The whole idea is to prevent all but the slightest expansion--as in so slight as to not be visible to the naked eye. I wouldn't try to finesse this with springs. I cited the module design in the Rimac Concept One in another thread and they use springs, but the springs look very heavy and stiff. I doubt the design is meant to allow for cell thickness expansion. If you get enough gas pressure to rupture a cell, you have a serious problem. Good stack pressure will bring this to your attention sooner than later. Note that cylindrical cells have their cans to perform this function. An overcompressed jelly roll in a cylindrical cell will be rejected as the cell will develop problems from the extra space between the jelly roll and the can.

A classic case of competing objectives. Monitoring your cell temps is a good way to figure if you need to take further steps or not. If you don't already have some, try to put some temp sensors down inside the modules, right on a cell at center stack--and also at the ends if you've got the bandwidth to work with. Good designs monitor both air temps and cell temps. On-board fire suppression is a nice touch. What agent does the system use?

 

I was thinking of adding a small cooling fan at the top of the box where the heat would gather and add some air filters to the air inlet and outlet.
I think the extinguishing agent is some sort of foam. See photos of it with the end end cap removed.

 

Neat. These are not as widely used as they perhaps should be in end-user packs. I wonder if you can get them with Halotron in them....

I should mention that from a serviceability standpoint, your design is awesome. Most prismatic designs make replacement of single cells a PITA if not impossible. Working with grey market cells as you are, your ability to do so with relative ease may well come in very handy someday.

 

Thanks for the thumbs up. I cannot take all the credit of the design. A lot of it is based on the design by Planetaire. You would need to read the whole thread which would be a marathon performance if you did. Similarly my five cell group boxes will slide into the main box and the Anderson plug will slide into another Anderson plug inside the main box which is bolted on a right angle polycarbonate plate which is bolted to the main box bottom.
I will also install a service fuse (to split the battery pack in half) at the front of the main box so it can be removed before opening the box. Maybe I can mechanically interlock the service fuse with the front lock so it can only be opened/unlocked once the service fuse has been removed.

 

Nice craftmenship and has real application. What type of BMS is in one of the images?

 

Centralised miniBmS from Clean Power Auto - Home, they are no longer available for sale but the distributed miniBMS is still available.

 

I am with this thinking as well. Compression while required on cells. Overcharged and overcharging of RFE 20ah packs is real concern .

 

Update on my new 40Ah A123 battery pack.

I have been using the new 40AH battery pack for about two weeks now in ICE kill mode or Forced EV mode. It looks good. I have done some data logging, plotting and graphs using my Duinomite Mega Canview V4.0.
I have removed the old frankenstein 4 KWH x 2 Enginer PHEV kits.
For thermal management I am using one thermostat, the sensor is located in cell group 3 in the middle cell but siliconed to the edge of the middle cell as the sensor is too big. I have observed over a few weeks by inspecting the thermostat LCD screen that the battery temperature measured by the thermostat does not rise any more than 14 degrees Celcius above the ambient temperature. The relay output of the thermostat is N/C and is wired in series with the miniBMS HVC/LVC loop wire. The thermostat can be programmed to alarm on temperature either too hot or too cold. At the moment I have it programmed to alarm if too hot (=>40 deg.C).