Beginner questions series

Tostaki

Member
Joined
Sep 20, 2017
Messages
61
As part as my DIY powerwall journey, I have some questions.
As it may help others, I will try to organize them.

Working with used laptop battery packs

- From your experience, what is the dead cell ratio found ?

- Is there some brands, models of cells to be avoided ?


Building packs

- For a 48v pack, my calculation gave me 13S setup (48/3.7). I saw you are with 14S setups. Why this difference ?

- I heard about double fusing. I am not sure to understand this one quite well. Is there any link to explain it ?

- What is your recommendation of power per pack and why ? (I mean 5 kWh, 10, 15, ...)

- Is there a lower limit for cells capacity not to go for ?

- What if mixing cells of various capacity in the same pack ? Does the lower one drive the whole pack ?


Thank you
 

DarkRaven

Active member
Joined
Sep 2, 2017
Messages
1,235
Tostaki said:
- From your experience, what is the dead cell ratio found ?

Dead as in 0V? About 5% in my last batch with another 5% having very low capacity.

Tostaki said:
- Is there some brands, models of cells to be avoided ?

No, all name branded cells are generally good. Only the unbranded/noname cells are difficult. Not necessarily because they are bad (they might be, but that's not 100%) but because you don't know anything about them.

Tostaki said:
- For a 48v pack, my calculation gave me 13S setup (48/3.7). I saw you are with 14S setups. Why this difference ?

The rated system voltages (12V, 24V, 48V) aren't fixed voltages, but ranges. 12V is about 10.5V to 15V, for example. And the higher you go up in voltage the lesser is the impact of a single cell added in series.
Example: 1S is 3.7V, 2S is 7.4V. The second cell doubles the voltage. 13S is 48.1V, 14S is 51.8V. The 14th cell adds only 7% to the total voltage. The relative change in voltage is so small that it still fits within the 48V system range. Personally I consider a 14S system fitting for 48V, but 13S may work as well, depending on your environment.

Tostaki said:
- I heard about double fusing. I am not sure to understand this one quite well. Is there any link to explain it ?

Double fusing means fusing both terminal of a cell or battery or any voltage source really. Technically this isn't necessary because you don't need to open the circuit on both sides to stop current flow.

Tostaki said:
- What is your recommendation of power per pack and why ? (I mean 5 kWh, 10, 15, ...)

No recommendation in general, depends on your system.

Tostaki said:
- Is there a lower limit for cells capacity not to go for ?

That is up to you to decide and again depends on your system.

Tostaki said:
- What if mixing cells of various capacity in the same pack ? Does the lower one drive the whole pack ?

The difference should be as small as possible. The more current you want to draw from the pack the closer the individual capacities should be. For low current applications it isn't as important to have them closely matched.
 

SimonW

Member
Joined
Aug 26, 2017
Messages
89
Tostaki said:
- From your experience, what is the dead cell ratio found ?
This will depend on what you get and how old they are and how they've been treated.
I was only able to get old leftover packs after all the good stuff had been picked out by people before me...
Out of ~730 cells collected to date, I got:
~9% that were rusty/corroded
~14% that were at zero volts
~10% that were below 1v
So about 1/3 gone already, the rest were OK to capacity test..
I'm 60% done testing and have a further 5% dead cells (not taking a charge/overheating etc)

Again, depends on what you get, their age, etc etc
You will very likely do better than me!

- Is there some brands, models of cells to be avoided ?
Sanyo (red cells) seems to have a bit of a bad rep for a high percentage of heaters, I can't confirm that as I'm yet to test my Sanyo's, leaving them 'till last.

- I heard about double fusing. I am not sure to understand this one quite well. Is there any link to explain it ?
Both ends of the cell, one at the positive end, one at the negative end.
Some people do it, some people don't.
I will be fusing one end only, doesn't matter which.
 

daromer

Moderator
Staff member
Joined
Oct 8, 2016
Messages
5,564
- From your experience, what is the dead cell ratio found ?
40% above 2Ah 20%dead rest in middle

- Is there some brands, models of cells to be avoided ?
Those not working.

- For a 48v pack, my calculation gave me 13S setup (48/3.7). I saw you are with 14S setups. Why this difference ?
13s to low. The low voltage limit goes way below whats nominal. 14s fits alot better and can be used from top to bottom


- I heard about double fusing. I am not sure to understand this one quite well. Is there any link to explain it ?
Gues double fusing as in both positive and negative? Its just a matter of safety.

- What is your recommendation of power per pack and why ? (I mean 5 kWh, 10, 15, ...)
Cant answer. It depends on your setup. I got 100kWh total right now.

- Is there a lower limit for cells capacity not to go for ?
Its up to you to set. My limit is 2Ah on the secondary bank and on the one im just adding as booster its 1700 right now. The 1700+ is though just play around pack and i dont consider them as a part of the main packs.


- What if mixing cells of various capacity in the same pack ? Does the lower one drive the whole pack ?
As long as you mix them properly in the packs and spread them:



I have tons of videos about the matter and also check out Peter (HBpowerwall) He also have TONS TONS of easy to view videos
 

Tostaki

Member
Joined
Sep 20, 2017
Messages
61
Thank you all for those replies. It helps. Thanks
I spent a lot of time on your videos (@hbpowerwall & @daromer) I will continue.

tostaki said:
- What is your recommendation of power per pack and why ? (I mean 5 kWh, 10, 15, ...)

I was unclear with that question. Let say you have a 14S50P pack. With a 2Ah capacity per cell that makes 100 Ah & 5 kWh battery. To add more capacity, you can either go 14S100P or 2x(14S50P). What is the best option ?
 

DarkRaven

Active member
Joined
Sep 2, 2017
Messages
1,235
It makes no (electrical) difference, 2x14S50P and 14S100P is essentially the same. Once you have started with 50P packs you probably want to continue and don't switch to 100P half way through building your battery. Going 100P right from the start, yeah, why not. Just remember that smaller packs are easier to handle and maintain if some work is needed.
 

Korishan

Moderator
Staff member
Joined
Jan 7, 2017
Messages
6,395
Either/Or. If you go with 100p, you should increase all ur packs evenly together. This keeps them balanced. If adding a second string, you just need to make sure you are equal voltage. So your second string could be 50p or 100p or something else.
 

WallBender

Member
Joined
Sep 16, 2017
Messages
41
It seems most people opt for the 14s100p, it's less banks to manage for balance.

If you had a balancing bms for each battery pack, having two overall might be nice for being able to disconnect one and still run off the other.

I'm pretty sure the main concern when expanding the power wall is keeping things balanced.
 

daromer

Moderator
Staff member
Joined
Oct 8, 2016
Messages
5,564
Larger packs = less work with balancing.

BUT

I opt for n+1 strategi. Doing that you can take out 1 string if needed for maintenance due to balance issues or such. It cost some more but you save some downtime (Yes im an IT guy)

The size of packs also depends on what you got when you start. If you got enough cells to do planned n+1 or just 1 then you build the packs based on what you got.
 

Hanssing

New member
Joined
Jun 6, 2017
Messages
16
darome said:
I opt for n+1 strategi. Doing that you can take out 1 string if needed for maintenance due to balance issues or such.

Why not just downgrade to 13S whilst doing maintenance on a pack?
It's all just settings in the inverter/balancer.
Quick interruption of power-supply, whilst removing a pack, reconfigure, and then running.

Identifying which cells in a bad pack are self discharching ore rebuilding it can take some time, right.

Having two strings, you loose 50% capacity doing maintenance on a full string
Downgrading to 13S for a while only looses you 7%

Edit - In your case the. answer is clear, you have constrainst on S-nr. due to mixed setup of LiFePo/LiIon but for singel-chemestri setups the mode could be S-1 instead of n+1 :)
 

Terry

Member
Joined
Jul 18, 2017
Messages
58
Hanssing said:
darome said:
I opt for n+1 strategi. Doing that you can take out 1 string if needed for maintenance due to balance issues or such.

Why not just downgrade to 13S whilst doing maintenance on a pack?
It's all just settings in the inverter/balancer.
Quick interruption of power-supply, whilst removing a pack, reconfigure, and then running.

Identifying which cells in a bad pack are self discharching ore rebuilding it can take some time, right.

Having two strings, you loose 50% capacity doing maintenance on a full string
Downgrading to 13S for a while only looses you 7%

Edit - In your case the. answer is clear, you have constrainst on S-nr. due to mixed setup of LiFePo/LiIon but for singel-chemestri setups the mode could be S-1 instead of n+1 :)

Or, your other option is say for example I am building a 14s120p, trick is on first string u build 15 packs of 120 cells, the 15th pack doesn't actually go into the build, but kept for a quick, "hot swap" if u will if u need to pull one for maintenance. Now I will also be sticking with 14s120p for each additional string , so for the second and further subsequent strings, I only.need to build the 14 packs, cause I wall always have the backup replacement, ready to go
 

daromer

Moderator
Staff member
Joined
Oct 8, 2016
Messages
5,564
Doing 13s by going down could work. But i dont see the reason doing that way at all. Thats purely or emergencies. Its better to have 1 pack in spare. You still need those cells when 1 pack goes bad so why not build it with n+1 string instead so you dont have to take the system down.

You only loose that 50% during the maintenance. You can still run the system without having any downtime. Bring out the string. Swap out the pack and back into the string. No downtime and its done within an hour as long as you got the +1 and have a pack to spare :)

In my case we talk about perhaps 20% at most if i remove 1 string since i got that many strings. Since my system is built to last for a week with low sunlight i have no problem having it down for that time.

With that said this approach all depends on how much cells you got. For many the 14 to 13s might be better sollution...
 

Tostaki

Member
Joined
Sep 20, 2017
Messages
61
This discussion turns highly interesting. Working in IT, I understand the Daromer's strategy. The 13S fallback is a solution too. But maybe the spare pack with one hour down time is the compromise.

To be decided once my cell harvest I really on it's way (just warming up)
 

Korishan

Moderator
Staff member
Joined
Jan 7, 2017
Messages
6,395
You also have to remember pulling the pack and going to 13s will drop string voltage. If for what ever reason your packs are low to begin with, your run time is now severely diminished. Having the 14s swap pack allows you to maintain voltages while doing the work needed.
You could even set up a system where there is virtually no downtime by allowing a secondary plug in for each pack for the spare. So, for a few minutes (or less, depending on how the setup is and how fast you are), the pack being replaced and the secondary pack are in full parallel mode.

This, of course, does have it's downsides as well. But, any build/design will have pros/cons
 

wim

Well-known member
Joined
Apr 6, 2017
Messages
560
Also think a spare pack is the bestway to go.... just have to find a trick to "equalize" the spare pack with the otherpacks voltagebeforeputting it inthere.
Maybe not so much if only one string in use, but with multiple strings... things might get hot :p
 

Korishan

Moderator
Staff member
Joined
Jan 7, 2017
Messages
6,395
Hmmm, good point. I wonder, could it be possible to have some diodes in place that would be activated when the pack is partially in place? This would allow for the other pack to be pulled out without getting an inrush. Then fully connect the replacement pack to disengage the diodes. Now, this would mean that all packs have these.

Just an idea. Things to think about. Don't wanna deviate from the OP's subject, though ;) Perhaps anew thread is in order to further talk of this subject.
 

jdeadman

Well-known member
Joined
Mar 28, 2017
Messages
931
that's what your BMS is for. If you are just taking out one cell the BMS will see that it's Higher than the rest and turn on the balancing. Now if you were putting this into a parallel setup you could use a current limiting resistor to slow down the current when there is a difference.
 

Korishan

Moderator
Staff member
Joined
Jan 7, 2017
Messages
6,395
Top