Lithium cells what's better; capacity or Discharge current ?

[UFO]

In the process of building a Battery pack, what is the most thing to consider when buying 18650 lithium cells ?
Nominal capacity or Discharge current ?
Let's say for a eBike project with a hub motor of about 4Kw, with 72V / 20W or more system :?
Exemple: Is is better to have 18650 cells with Nominal capacity of 3300 mAh with discharge current of 10A or cells having Nominal capacity of 2500 mAh with a discharge current of 20A ? :?

Thanks for the help...

 

[Punx0r]

depends on your needs

 

[riba2233]

it depends on your needs, if you are going to build low capacity pack then you have to pick high discharge cells, and vice versa. You should start with - how many Wh do I need, or can fit/afford?

 

[UFO]

it depends on your needs, if you are going to build low capacity pack then you have to pick high discharge cells, and vice versa. You should start with - how many Wh do I need, or can fit/afford?

Ok let's say that the configuration would be for a road / off-road mountain bike with a lot of power, for a range of about at least 50 km, that could reach the 80 to 90's km/H.
I have a Crytalyte 5403 for the project, and a 2004 Giant DH Comp bike , and for long term, i could afford over 140 cells if needed for range. For the moment, with the knoledge i got, it's not obvious to figure it out...

Looking like this bike, and that kind of power in the trails.

 

[Punx0r]

OK, so you want cheap, powerful and high-capacity

 

[UFO]

OK, so you want cheap, powerful and high-capacity

I do not mind people laughing about my poor electric knoledge...
But i never talked about cheap stuff, i just want to buy the right things the first time !
Because as most experienced people know; "Buy cheap, buy twice"... :wink:

So besides jokes, does anyone could give me a little clue on the main question for the project stated above
Is is better to have 18650 cells with Nominal capacity of 3300 mAh with discharge current of 10A or cells having Nominal capacity of 2500 mAh with a discharge current of 20A ?
Sorry to ask again, but this could help me to choose from tons of different cells on the market

 

[neptronix]

Depends entirely on your application and it's power demands, and the amount of range you want.
There are plenty of setups where a super putzy 1-2C rated 3500mah 18650 cell will do great for.
There are setups where only something like RC Lipo or A123's prismatic 20AH cells will work.
And then room in between.

What kind of range are you looking for?

 

[neptronix]

Go look at some maximum C rates of the battery you have your eye on. Cut that maximum C rate in half or less in order to have a safe discharge rate where you are not blowing a lot of your watt hours into heat, making your battery a space heater.

I always go with cutting that figure by a fourth.
So if i have a 10C cell, i want an average load of 2.5C.

If i have 3kW of sustained power, i want a battery that can output 3kW sustained easily.
With a 4C rated 18650 cell, i want to run it at 1C, therefore i need 3kwhrs worth at a minimum, which is very difficult to fit on a bike.
With a 10C rated cell, i want to run it at 2.5C average, so i'm going to need 1.2kwhrs worth of battery at a minimum.

A higher density battery pack is great, but the higher density you have, the less C rate you have to play around with.

Get out a calculator and figure this out. You can use the ebikes.ca simulator with the 540x motor as well to figure out power consumption, range, etc.

 

[UFO]

Go look at some maximum C rates of the battery you have your eye on. Cut that maximum C rate in half or less in order to have a safe discharge rate where you are not blowing a lot of your watt hours into heat, making your battery a space heater.

I always go with cutting that figure by a fourth.
So if i have a 10C cell, i want an average load of 2.5C.

If i have 3kW of sustained power, i want a battery that can output 3kW sustained easily.
With a 4C rated 18650 cell, i want to run it at 1C, therefore i need 3kwhrs worth at a minimum, which is very difficult to fit on a bike.
With a 10C rated cell, i want to run it at 2.5C average, so i'm going to need 1.2kwhrs worth of battery at a minimum.

A higher density battery pack is great, but the higher density you have, the less C rate you have to play around with.

Get out a calculator and figure this out. You can use the ebikes.ca simulator with the 540x motor as well to figure out power consumption, range, etc.

Thank you for your answer. I will try to digest all that information in the tiny brain i have...
With a bit of research i should understand this concept. Even if it's not all that clear, it sure sound logic !

 

[neptronix]

No problem.
Division and multiplication in a calculator, coupled with graphs that show how different cells perform at different C rates will tell ya everything you need to know, though

If you want to calculate the watt hours per kg ( energy density figure ) of any cell, it's:
watt hours in cell / kilograms of cell.
Or watt hours of pack / kilograms of pack.

So if you want to look at a 44 gram, 11whr nominal 18650 cell, the math is:
11wh / 0.044 = 250whrs/kg

If you don't know the watt hours of a cell, the watt hours is:
Nominal voltage x amp hours.

Do keep in mind that watt hours delivered are also a factor. Any battery that is used at higher than a fourth of it's maximum C rate is going to give you significantly diminished watt hours in the form of lower amp hours and lower nominal voltage. It gets worse, the further you go towards the maximum C rate. You can lose up to 10% of your available watt hours with something like an RC Lipo. A newer 18650 is more heat tolerant, so it can lose even up to 17% of it's power. Maximum C rates are based on thermal limits, not discharge efficiency.
The higher the C rate you run, the more your battery pack is effectively less energy dense, hotter, and suffering in terms of longevity. Which is why many packs have their cycle lives rated at something like 1/10th the maximum C rate. Fun little trick by lithium battery sellers.
Which is why i recommend always putzing a cell out, relative to it's C rate. A fourth of the maximum is always a recipe for good performance, cell life, and no issues with heat. Since an 18650 is rated for a higher maximum C due to it's heat tolerance, i'd say that that ideal would be less than a fourth.

watt hours per kg is very closely related to the space that a battery will take up.
Example: in a typical triangle bag, for a 150whrs/kg battery, your maximum battery is 25ah 46v ( 12S ) if using turnigy 20C lipo.
If you have the 190whrs/kg 10C multistars, suddenly you can fit 40AH 46v ( 12S ) in the same space.

Same goes for 18650s, though i don't know exactly how many cells you can fit into the same space, so i can't give you volumetric figures on that.

 

[UFO]

No problem.
Division and multiplication in a calculator, coupled with graphs that show how different cells perform at different C rates will tell ya everything you need to know, though

If you want to calculate the watt hours per kg ( energy density figure ) of any cell, it's:
watt hours in cell / kilograms of cell.
Or watt hours of pack / kilograms of pack.

So if you want to look at a 44 gram, 11whr nominal 18650 cell, the math is:
11wh / 0.044 = 250whrs/kg

If you don't know the watt hours of a cell, the watt hours is:
Nominal voltage x amp hours.

Do keep in mind that watt hours delivered are also a factor. Any battery that is used at higher than a fourth of it's maximum C rate is going to give you significantly diminished watt hours in the form of lower amp hours and lower nominal voltage. It gets worse, the further you go towards the maximum C rate. You can lose up to 10% of your available watt hours with something like an RC Lipo. A newer 18650 is more heat tolerant, so it can lose even up to 17% of it's power. Maximum C rates are based on thermal limits, not discharge efficiency.
The higher the C rate you run, the more your battery pack is effectively less energy dense, hotter, and suffering in terms of longevity. Which is why many packs have their cycle lives rated at something like 1/10th the maximum C rate. Fun little trick by lithium battery sellers.
Which is why i recommend always putzing a cell out, relative to it's C rate. A fourth of the maximum is always a recipe for good performance, cell life, and no issues with heat. Since an 18650 is rated for a higher maximum C due to it's heat tolerance, i'd say that that ideal would be less than a fourth.

watt hours per kg is very closely related to the space that a battery will take up.
Example: in a typical triangle bag, for a 150whrs/kg battery, your maximum battery is 25ah 46v ( 12S ) if using turnigy 20C lipo.
If you have the 190whrs/kg 10C multistars, suddenly you can fit 40AH 46v ( 12S ) in the same space.

Same goes for 18650s, though i don't know exactly how many cells you can fit into the same space, so i can't give you volumetric figures on that.

Yesss sir ! That's what I call a clear explanation...
If i get it right, ruffly said, to permit to go at the 4th of the C rate, i would fit some extra cells on the bike, to keep that C rate in a safe range.

 

[Kepler]

Get your head around C ratings and it will become very clear what cells you need for the job.

Take a Samsung INR18650-25R for example. Rated at 2.5AH . So this means when discharging at 2.5A, it is discharging at 1C.
However this cell is good 20A continuous and as such has a max continuous C rating of 8 (20/2.5)

You are looking for high powered setup of around 4 kW. So at 72V, this means you will be drawing around 55A to achieve 4 kW
We now can see be by placing 3 Samsung INR18650-25R in parallel, you now have a max 60A to play with and could get away with a 60 cells as a minimum (20S3P)

However this would be a measly 540Whrs (3 x 2.5ah x 72V) and no where near the range you are looking for. Also you would be working the cells hard at wide open throttle conditions.
Range wise, You need about 1000W to sustain 45kph. At this discharge rate you have 30 minutes out of your battery and as such will cover just over 20kms.

You have 50km range on your wishlist which means that you need around 1350 Whrs to achieve this range based on the 1000W / 45kph estimate.

1350 Whrs in a 72V configuration equates 18.75Ah. Divide this by 2.5Ah cell capacity means you need to have 8 cells in parallel. So the pack required is a 20S8P. A total of 160 cells are required for this setup.

With an 8 cells in parallel, you have a theoretical discharge rate of 160A at the batteries maximum C rating. With a max discharge of 55A estimated for this setup, the 25R could be considered overkill on paper however the extra discharge capacity headroom means you will get long life and minimal voltage sag out of the pack.

Personally, I would consider the Samsung INR18650-25R cells to be a solid contender for your needs.

 

[neptronix]

Actually, if you do the math on this 3P pack, you get less watt hours than you think.

Attached is the Samsung 25R cell spec sheet, which is actually one of the most detailed sheets i've ever seen ( kudos samsung )

You've got 60 cells. Each cell is rated 9.38whrs. Nominal voltage is 3.61v. That's 562.8 watt hours, yeah?

Samsung tells us that the cell will give 8.25whrs at 8C, and each cell will be 3.27 volts nominal.
Therefore 60 x 8.25 = 495 watt hours.

At 0.2C continuous, your pack is 72.2v nominal.
At 8C continuous, your pack is 65.4v nominal.

If you follow my 1/4th C rule, and you still want 4kW output, you want a 2C discharge rate to maximize your C rating.

At 2C continuous, your cells are 3.54v nominal ( not too bad of a drop from 0.2C, right? ), so your pack is 70.8v nominal. Your pack will deliver 535.2 watt hours of energy. Not bad.
But if you want 4kW, you're gonna need a 2kw-hr pack.. which could probably be stuffed into the triangle of a typical bike frame, so that's no big deal.

The whrs/kg figure is very good on the samsung 25Rs ( until you weld the cells up of course ), better than turnigy multistar for sure. but it can drop below 200whrs/kg pretty readily at the highest discharge rates, making a higher powered cell like a multistar a better choice.

Basically, how energy dense your battery is totally depends on how hard you thrash it.

 

[Kepler]

Fair comment on the battery capacity changing depending on what C rating you are drawing at. I still think 1350Whr is in the ballpark though. At 55A the pack is drawing at a rate of around 2.8C. However 55A would only used in small bursts so most of the time the pack would be below 1C and operating in its nominal voltage range.

 

[neptronix]

Yeah, i'm thinking in terms of the nominal wattage mostly. Short bursts are fine.
Does suck when you are at peak controller load climbing a long hill and you see your battery voltage dive and your speed drop dramatically though.

Depends on how much hill climbing, offroading, bearing down harsh winds etc i guess.

 

[UFO]

Get your head around C ratings and it will become very clear what cells you need for the job.

Take a Samsung INR18650-25R for example. Rated at 2.5AH . So this means when discharging at 2.5A, it is discharging at 1C.
However this cell is good 20A continuous and as such has a max continuous C rating of 8 (20/2.5)

You are looking for high powered setup of around 4 kW. So at 72V, this means you will be drawing around 55A to achieve 4 kW
We now can see be by placing 3 Samsung INR18650-25R in parallel, you now have a max 60A to play with and could get away with a 60 cells as a minimum (20S3P)

However this would be a measly 540Whrs (3 x 2.5ah x 72V) and no where near the range you are looking for. Also you would be working the cells hard at wide open throttle conditions.
Range wise, You need about 1000W to sustain 45kph. At this discharge rate you have 30 minutes out of your battery and as such will cover just over 20kms.

You have 50km range on your wishlist which means that you need around 1350 Whrs to achieve this range based on the 1000W / 45kph estimate.

1350 Whrs in a 72V configuration equates 18.75Ah. Divide this by 2.5Ah cell capacity means you need to have 8 cells in parallel. So the pack required is a 20S8P. A total of 160 cells are required for this setup.

With an 8 cells in parallel, you have a theoretical discharge rate of 160A at the batteries maximum C rating. With a max discharge of 55A estimated for this setup, the 25R could be considered overkill on paper however the extra discharge capacity headroom means you will get long life and minimal voltage sag out of the pack.

Personally, I would consider the Samsung INR18650-25R cells to be a solid contender for your needs.

Wow... Does all these informations are free ? How much do i owe you ? Thanks !
Before i this putting this post i was quite lost, but i came up with a couple choices of cells, and from what i could see that was used in similar projects, i could see that the configurations was more looking of over 100 cells (like picture attached), and even some with 280 cells. So I was kind of going in the middle, like 160. So if you suggest about the same as a safe option, maybe some invisible spirit lead us to this thought...
So with a pack of 160 cells 20S 8P, i would be ok for a nice ride... :wink:
From what i saw, with a bit of shopping, it's more affodable than i thought it could be at first. 8)
So i came up with a couple preliminary choices with the knoledge i had before this post, and i found a couple potential cells.
Here they are (not necessary in the priority order) :

- LG ICR18650HE2 , at 2500mAh rated at constant current of 20A (full data sheet attached)

- Samsung INR-18650-35E
Nominal capacity lt. Manufacturer 3500mAh (2.65V)
measured at 2A to 3V 3171mAh
Chemistry NCA / Li (NiCoAl)
positive FlatTop
Charging current 2 A
Discharge current 10.5 A
Charge voltage V 4:20
Discharge voltage 2.65 V
Internal resistance Ri 28 milliOhms

But your Samsung INR18650-25R looks good too. What do you think about those 2 last listed ?

 

[UFO]

Already replyed

 

[UFO]

Yeah, i'm thinking in terms of the nominal wattage mostly. Short bursts are fine.
Does suck when you are at peak controller load climbing a long hill and you see your battery voltage dive and your speed drop dramatically though.

Depends on how much hill climbing, offroading, bearing down harsh winds etc i guess.

Thanks neptronix, my last reply to kepler refers also about the excellent help you provide me. I can see that it's people like you guys that this forum makes sense, by helping the community.
I feel like i am going back to school, but this time, it's fun...
Not knowing how to put my last reply in this one, you'll have to look back to give me your thoughts about my suggestions.

 

[Kepler]

Wow... Does all these informations are free ? How much do i owe you ? Thanks !

No problems. Happy to help out.

If you were looking at 2kW peaks, I would say the Samsung INR-18650-35E would be a great choice especially since you could drop down your cell count by around 30% for the same Whr capacity. However as you are looking at 4 kW peaks, my advice would be to stick with - LG ICR18650HE2 or the Samsung INR18650-25R cells.

 

[LockH]

Wish only to add here:
https://en.wikipedia.org/wiki/Kinetic_energy

Kinetic energy
In physics, the kinetic energy of an object is the energy that it possesses due to its motion. It is defined as the work needed to accelerate a body of a given mass from rest to its stated velocity.

In other words, the more mass ("weight") you're carrying around and need to accelerate and push against gravity ("up hills")... well, it's like building in extra Wh/km as "standard overhead".

One way to reduce battery size/use is just to "slow down" as velocity through air increases energy required dramatically as speeds increase. So the alternative is to go more "aero". Like the velomobile:

Bon Chance mon ami.
L

 

[Voltron]

Off topic, but that thing looks amphibious! Am I seeing a flip down propeller?

 

[LockH]

Off topic...

HUH? Thread Subject "Lithium cells what's better; capacity or Discharge current ?" I assumed this related to energy consumption? So, thread title should read "How to waste more energy faster"?

 

[spinningmagnets]

You have provided many of your performance wishes for this battery pack, but I think it will be most helpful to know what size you can fit on your frame and battery containment. Have you calculated how many 18650 cells would fit as a maximum?

The triangle pack shown open above has 100 cells on their side (as an example)...

 

[Voltron]

I meant MY reply was off topic.

 

[LockH]

I meant MY reply was off topic.

Oooops. Sorry. Didn't want to bring sex into this. "What Happens When a Kayak has Sex with a Velomobile?"
http://www.travellers-content.co.uk/what-happens-when-a-kayak-has-sex-with-a-velomobile/