LIC Lithium Supercapacitors

peterperkins

100 W
Joined
Jun 6, 2008
Messages
196
I thought this might interest members on here.
I'm not aware of any other supercapacitor equipped road going vehicles in the UK apart from my MK1 Honda Insight.

The ageing old Nimh technology hybrid battery pack has been replaced with a Lithium Supercapacitor pack made up of 50 x 2200F Capacitors. They are low internal resistance, 50 milliohms for the pack, very efficient >90%, and effectively have an unlimited cycle life.

In my MK1 Honda Insight with tweaked electronics to manage them, it's a significant improvement in instant power over the old pack. They only store about 0.8ah but that's enough for rapid acceleration/braking with high efficiency.

Tech details and photos etc are on the below thread over at Insightcentral.net

http://www.insightcentral.net/forums/modifications-technical-issues/17282-super-capacitors.html

http://www.jmenergy.co.jp/en/products_cell_rami.html

View attachment 2

View attachment 1



https://en.wikipedia.org/wiki/Lithium-ion_capacitor

https://en.wikipedia.org/wiki/Supercapacitor

Might think about adding them to the MK1 Insight UK rally car.

http://www.oaktec.net/phdi/p1.nsf/supppages/oaktec?opendocument&part=6
 
Wow, that looks pretty cool. Thanks for sharing.

My old HCH has 215,000 miles on it but still working with the second pack.
 
50mOhm ESR for the cap seems pretty high to me. A single Zero module is <60mOhm and up to 5 in parallel for a motorcycle <12mOhm paralleled resistance.

To go from >200Wh/kg to 10Wh/kg would need at least one and ideally two orders of magnitude lower resistance to become tempting to use over LIBs even for purely regen storage.
 
It's <50mohm for the whole pack (50 caps in series) not counting interconnections.

It's actually probably less than that. The quoted esr for a single cap is 0.5mohm. :)
 
peterperkins said:
It's 50mohm for the whole pack (50 caps in series) not counting interconnections.

It's 60mOhm for the whole 28s string (116.4vdc peak) including business fuse and contactor on each of the modules in a modern Zero motorcycle. Dirt bikes hold two modules for 30mOhm when in parallel, and the S/DS up to 5 of them for 12mOhm.

We have different expectations for projects, I'm sure it's lower impedance than the stock Honda NiMH by a lot. Just doesn't seem very compelling if you're got modern cell options available.
 
The manufacturer says 0.5ohm per cap so that means 25mohm for my 50S pack,but i was erring on the side of caution and rounding it up..
I probably don't have the equipment to measure the ESR for the whole package inc contactors/fuse etc. But I might look at that..
 
The insight Mk1 gets a lot of it's fuel efficiency from the lean burn engine system.
The IMA side makes only a relatively small contribution to the total.
However when you consider the appalling SD rate of older nimh cells and the fact in average terrain you don't need long bursts of assist/regen.
(You just need enough really to rapidly get up to cruising speed + a bit more)

The inefficiencies (IR + SD losses) and conversion losses in the nimh setup are pretty bad.

Ultimately i may end up using them as a buffer pack with an old 50S A123 20ah pack I used for many years.
(It's still got about 17ah useable capacity but higher IR now)
The old A123 pack is a good voltage match and with some simple pwm switching to control (lowish current) flow back and forth they could work well together.
 
Hi Peter, i admire your work and skills in this field and i am impressed with your continued push to improve the hybrid system ...
I understand this was an experimental exercise, But other than a life expectancy far beyond any reasonable requirement, is there any real reason to consider this solution rather than the A123 pack (or similar) that you have previously used ?
there would seem to still be some significant cost, weight, size, capacity, (and even availability ?) issues with these S'caps compared to current Li cells.
$1000 worth of A123 26650's at 5 kg, would appear to give similar results and more capacity ??
 
Probably not is the simple answer.

But...
I had access to the caps at zero cost.
I'm interested in them as an experimental base for further research on the Insight IMA/Capacitor/Battery systems.
I enjoy tinkering and learning about stuff.
I might combine the caps with that old A123 50S pack i used before, it's got quite high IR now but still around 17ah capacity, so using the caps as a buffer might give me the best of both worlds and get more life out of the A123's.

I'm thinking of a couple of strategies to combine them.

1) Just connect the caps across the battery pack (after equalising the voltages!) and use them in parallel.
Let them sort themselves out.

Pros. Very Simple.
Cons. Does not use caps to maximum capacity. Reduces max IMA power availability. (reduces system max voltage)
Max A123 pack voltage is 50 x 3.65 (182.5V Max - 100V Min) (supercaps is 190v Max - 110v Min)

2) Keep caps as they are and use the old lithium pack as a low A constant charger feeding the caps steadily until they match the pack.

Pros. Reasonably simple using a bi-directional pwm switching circuit. A rate of say 5-10A would keep them pretty well charged and be easy on the ageing lithium A123 cells. Less peukert effect and better efficiency and higher voltage. The 50 x 3.2v nominals A123's would keep the cap pack upto around 160V, leaving headroom for regen.

Cons. More complicated. Extra weight.

I like 2. lets look at that a bit more closely.

Best perhaps to regarded the A123 pack as a simple energy store and use grid charging to keep it topped up.
However with bi-directional pwm controlled current flow from A123 to supercaps and vice versa might be a great combo.

In practise how would that work.

a) So if the A123 are charged and higher V than supercaps.
PWM1 turns on and charges supercaps at 5-10A rate until voltages match.

b) If the Supercaps are higher than say 180V (the balancing board voltage) then PWM2 turns off and bleeds current back into A123 pack at 5-10A rate until supercaps drop below 180V

Fairly simple to control that with voltage detection of the two packs and current counting/measurement to control current flow back and forth.
 
Back
Top