Testing the ZnBr2 cells/chemistry zinc bromide

100kwh-hunter

Active member
Joined
Mar 2, 2019
Messages
1,355
Update 2: 1st of march 2020

How I gotstarted:
https://secondlifestorage.com/showthread.php?tid=8557
How to build:
https://secondlifestorage.com/showthread.php?tid=8772

This first post will be updated when needed with new test results and cells.
Goal is to find out:
1: there max charging volt.
2: self discharging.
3: what they can hold on power
4: c rate at charge and discharge.
5: behavior when connected in series
6: cycling test.
7: need for under or over charge/voltage protection
8: need for balancing.
9: how much it will take to get them full and how much they will give before empty.

The new cylindrical cells:
Cell one is happy to take 0.3 ah and give 0.3ah with every cycle it is going up, i used 50cc of electrolyte: 40gr of pure ZnBr2 to 100cc distilled water, the max that the water can take without tricks is 44 grams per 100cc(2 M max)
When beginning discharge the cell is at 1.80v after discharge the cell jumps back to 1.77v and i can do a second run
Cell one is going to have its fifth cycle and cell two his first atm.

Cell one holds ~35-40cc electrolyte( i spilled some electrolyte)

Average charge discharge rate 0.200ma
Average capacity 300mah
7 cycles

Cell two holds 50cc electrolyte.
Average charge discharge rate 0.200ma
Average capacity 550mah
Both at 1.80-1.82v
3 cycles

I think those are nice numbers

Max charging voltage seems to be 4.00v, best would be 3.5v, they react exactly as lead acid, but with a unlimited cycle life.
You can feed them 10ah, but they will only take what they want.
Cell one and two dimensions: 5cm in height and id 57mm.
Self discharging can be to 0.00 volt without damage.
There is so far no need for a bms or balancing.
This message will change from time to time when the numbers are cracked and narrowing down.
For now cell one: 1.80v 50cc of 1.75 M solution will give 350-400mah for the costs of 15% of a new 800 cycle lifeli ion 18650.
Against unlimited cycle life? you do this math...
Will update when i have some more numbers/statistics.

Building of this cell:
https://secondlifestorage.com/showthread.php?tid=8772
Look for things that look similar to a pipe :cool:

--------------------------------------------------------------
The cells below are with oasis and closed design, every experiment FAILED...
On power storage, taking and giving some mmmmmmmmmah, lets say femtoamp hours.
There was no flow of the electrolyte and there was to much pressure building up.
One week of charge at 0.02a and burning a led for 10 minutes.....
Above this alina, i will post test results of the new cylindrical cell without oasis, looks really better.....
------------------------------------------------------------


With foam and closed cells...failedon every aspect, the model was wrong
Cell specs as 26-02, with all the same dimensions: 7x5cm and 2 cm thick,
Charge and discharge are done with a 4a dc dc converter, zb2l3 and 3x yellow led (1.6v to 2.1v at 20mahx3=60mah)

Cell 1: No graphite foil and ~60-70% full of electrolyte.
Cell 2: No graphite foil and 100% full
Cell 3: With graphite foil and 100% full

Cell test results:
Cell 1:
Charged several times to 1.82 - 1.85v with 3.5v charge, discharged to 0.00-0.05v several times.
Left on the charger for three days, voltige was 2.15v after been disconnected from the charger for3 hours.
To weak to light up one yellow led by itself.

Cell 2:
Charged severaltimes to 1.82 - 1.85v with 3.5v charge,discharged to 0.00-0.05v several times.
Left on the charger for three days, voltige was 1.82vafter been disconnected from the charger for3 hours.
To weak to light up one yellow led by itself.

Cell 3:
Taking his initial charge:
3 hours at 2.5v, max reached voltige 1.30v: was standing on 1.30vfor over one hour.
Went to 3.6v charging, it takes 0.030ah now.
Adding the graphite foil seems to improve as a collector.
After 4 hours it took 0.200mah!
After adding 3 hours it took 0.125 still loading...
Graphite foil is the way i guess

Cell 4 and 5 failed also...
Not enough mah taking or giving, plus to much presseur.

Various tests:
Connected cell 1 and 2 in series to gain 3.65v and 3 leds in parallel.
Every full cycle, same results: 2.5 hour burning time, cell 1 ended at 0,15v and cell 2 at 1.65v, 3 hour charging in series at 7.2v at 0.025ah
Some dischargetests with cell 1 and 2 after discharge with leds was with a resistor of 5.1 ohm 5 watt to go to 0.00v and charge it back up again to 3.65v at 7.2v charge.
Tested cell 2 allone for charge at 5v and 7v...7v is a no go, the cell had pressure building up inside.

General results up to now:
It seems they must have dubble voltige before they want to take any power into them.
The more electrolyte is in them the more they can storage.
Most tests will be done at the same way at the same cell to get consistent results
With foam and closed cells...failedon every aspect, the model was wrong


If anyone has a idea, thought, comment, testing, behavior or whatsoeverplease do share.
Not that i don't like it, but please give a like if you want me to keep this updated and continue to share my work.
Thanks in advance, reading,understanding and the likes for appreciation for my work....this does not count as hobby anymore :cool:
 
When I watched the youtube video, the bloke said something about the maximum current for charging and discharging depending on the electrode surface area, maybe that's something you can take into account or test.

I think in general the most important battery spec to test is the overall Wh capacity. This will give you a general idea on how much cells you'd need to build a 10kWh powerwall. And as such you should be able to calculate the cost of such a power wall and compare with a powerwall built of 18650's.

I also wonder if with this battery chemistry an expensive bms is realy necessary. I think if you monitor individual cell voltages and have a system to switch battery off if there's an over or undervoltage, that would be enough...

If they are realy easy to build, then one shouldn't mind build a thousand of them...

I myself am a big fan of building a powerwall from "a lot of" lower powered cells. It makes for a more flexible design and easier to replace faulty cells if necessary.
 
@Ivo,
I totally! agree with you on 5.5 points out of 6 you stated, and let me clarify what i learned so far, specific questions need specific answers, in my opinion.

First: indeed it is the surface area, but he had a anode and a kathode submerged in the electrolyte= 80-100 mah charge/discharge per cm2.
My anode and cathode are not submerged, so 70% (no typo or calculation mistake) less surface area.
Then the concept of a closed cell(safety!!!): 20-40 mah per cm2.
How can this be: simple, the electrodes are both sides wet or in my case at one side wet.
His first wet cell had free floating 4.4cm of electrolyte (two uk thumbs, he is/was mistaking a lot about his measurements!)
Now the big problem of a free float anode/cathode cell is that the zinc crystals grow in a non logical way, the first cristal that touches the bromine solution, will engage a self discharge, hence why we use a phenolic foam.
To keep those two separated.
This is also the reason that those cells MUST BE HORIZONTALLY!
This is THE reason i think all those testing jars and standing cells ect are failed. look close to what the bromine and zinc are doing when under charge!!!!!!I wont even spend time to do those kind of experiments...i am not a schoolboy anymore.
Indeed you could say that the phenolic foam or a other filter/medium would prevent the crystals to go a non logical way, it will force it to go the linear horizontal way, with maximum use of the total surface area, before they reach/contact each other.
In all the (failed) experiments from 1885 til now there is one big difference: the medium!.
BUT it is not alone the surface area, also the amount of znbr solution, theoretically speaking a cell of (in my case) with a surface area of 35cm2 but 10 cm in height would give 5 times more capacity, this test is VERY HIGH ON MY LIST, VERY HIGH.....aka the next test and enlarge the total surface area.
Lead acid batteries have the same and a other problem: horizontal and the electrolyte, before it was a youtube hit i already was busy to make "new" lead acid batteries from dead ones, come on how simple you want to have it. Cycle life is just 600.
So this is all in my second test setup, third will be more specific to energy storage(not system yet!)

Quote of Ivo:
I think in general the most important battery spec is to test is the overall Wh capacity. This will give you a general idea on how much cells you'd need to build a 10kWh powerwall. And as such you should be able to calculate the cost of such a power wall and compare with a powerwall built of 18650's.

The overall capacity: incl the failed tests and ect: the capacities are way over every chemistry known to us right now, regarding storage!!, BUT we want to have 10 KWH in 5 sec to meet the demands of the surge for our oven, induction, cooking, boiler, ecar ect....This is the problem with znbr chemistry....you need A LOT of m2.
But if you can meet that surge you want, and you have ~7.5m2x0.75m over or stack the 7.5m2 in the way you like. you can have the surge, and 150kwh-170kwh in storage.
The costs compared with second handed 18650 to newly build znbr is not even 15% (if bought znbr in bulk)assuming it will last 3 years instead of 50 years, but you will need surface!!!!M2!!!!
And this test the cycle test...IF i get good results, i could have the proper equipment! and then we are going to test the real cycle life.

Quote of Ivo,
I also wonder if with this battery chemistry an expensive bms is really necessary. I think if you monitor individual cell voltages and have a system to switch battery off if there's an over or undervoltage, that would be enough...

With the results i have so far...so far that is.....no bms or balancing is needed, YET!...just seek the proper pv panels or adapt your battery voltage to the pv panels and you are done.....but don't forget the diodes and the fuses.
Remember i am still testing, for save, good and fast charge is double the voltage(NOT TRIPLE!!!)at 20-40mah per cm2, and they can be discharge to 0.00v without damage, up til my tests now!

Quote of Ivo:
If they are really easy to build, then one shouldn't mind build a thousand of them...

Now that is one of the problems to be solved, why build thousand of them if you can build a flow battery.....Costs my dear boy....costs...come on you are my neighbor you know our tax collector.....
You don't want to build a flow construction, if you want go ahead, prepare to pay a lot for materials and get disappointed in results.
Been there done that....
Simpler would be: theoretically atm: build a 30x30cm x 100cm in height will give you 1.82v and around 1500a storage, but a discharge rate of 10ah if you are lucky at 1.82v..
However the initial charge could take up to several months to never, what about discharging to 50% or 30% or to 0v, and start up again, you have one advantage, the inc and the bromine are still in there place, thanks to the medium.
And our stats aka the EU don't want it, cause it's to easy for diy....just like a lot of other successful endeavors of me..like biodiesel ect
So the optimum would be to discover the proper height and the best/biggest surface area.
A single cell of 1m2 by 5cm hiegh would give you any surge you want...if one cm2 can give you ~30mah what can 10.000cm2 do?
If 1cm3 ZnBr2 solution can give or store 1 to 3 amp and your cell holds 30.000cm3....at 1.82v...just questioning, theoretically speaking.

Quote from Ivo:
I myself am a big fan of building a powerwall from "a lot of" lower powered cells. It makes for a more flexible design and easier to replace faulty cells if necessary.
And this is exactly the point, you can repair a znbr cell just to deplete it to 0.00v start initial charge at cell level and you will be fine.

Thank you Ivo

Now that 0.5 point i disagree:
M2 and M3 do better calculations next time and get your numbers right,read and study more...but 5.5 out of 6 points would be on a so(you know what in mean..dutchies also) a 9.0 on vwo!!!!! well done Ivo.

Give my time to test, to prove the results i have read on so far (was a lot!) and improve, in the meantime:
Any thoughts, ideas or questions are welcome and really appreciated, maybe i am not aware of somethings and i have to learn extra.
Keep it coming.
Best
 
People like you can help the world become a better place!

I will follow it further, i will also build a ni fe cell, not right now, but i will do it.

If you need help on a mechanical, designing point, shout, I am a mechanical engineer designing stuff full time ?

Keep up the good work!
 
100kwh-hunter said:
But if you can meet that surge you want, and you have ~7.5m2x0.75m over or stack the 7.5m2 in the way you like. you can have the surge, and 150kwh-170kwh in storage.
The costs compared with second handed 18650 to newly build znbr is not even 15% (if bought znbr in bulk)assuming it will last 3 years instead of 50 years, but you will need surface!!!!M2!!!!
And this test the cycle test...IF i get good results, i could have the proper equipment! and then we are going to test the real cycle life.

Today I have been thinking about something.

Concerning the "punch" issue, I remember having seen a youtube video once. Check this one out:


Again it's from robert murray smith. He's talking about another battery chemistry, but I think the principle is the same here.

To summarize:

To deliver a punch, attach some kind of capacitor to the battery cell. Basicaly, the battery charges the capacitor, and when a load is attached, the capacitor is able to discharge quickly to get things going. In the mean time the battery can catch up with the load, or the surge load already has passed.

Maybe this is an idea you can do something with.

Regards

Ivo
 
Indeed, if this project is going to be successful i was indeed planning to hook up something like a big capacitor or some lifepo cells to catch the first punches for the surge.
This video is not one of his best, I think also this capacitor does need meet the requirements.
Would be a fun project to, witch of his projects are not fun to do :huh:
Probably the capacitor i am referring to is in the members only section.

First i am going toredesignhis blade type cell, the blade cell has to many problems, but the theory holds firmly ground.
One of the biggest problems is the venting hole to prevent pressure building up.
The venting hole can not be in one of the sides, cause you will get rid of the precious electrolyte.
To make a vent hole on top it can be done, by adding a tube, problem would be the graphite foil, bromine will attack immediately if overflow or a not good connection. Building wise it would bring to many challenges/problems.
His blade cells are also not entirely full.

Some quick specs from my side.
Loading to 2.50v does nothing.
Loading to 3.00v gives bearly 1.70-1.75v in the cell
Loading with 3.30v gives a nice exact 1.82v but merely no power stored.
Loading with 3.60v gives some stored power, depend on the charger time ~100mah at 35cm2.
But will bulge the cell (before it reaches it full potential?)
From what i have read to get a proper loading voltige you must have a minimum of 3.5v to max 5v, per cell with a charge discharge of 20-40mah per cm2. every cm3 could store 1.2a, but on storage there are no hard numbers!
First i want to prove some of those numbers.
After bulging the contact with the electrodes is reduced, you really must press it back for good contact.
Tripels the charge from 0.090ah to 0.250.
I am charging atm cell 4 and 5 at 3.40v to try out where the bulging is beginning.
From 2.50 to 3.00 in12 hours,to 3.30 in6 hours, to 3.40 in 6 hours, to 3.50 in 6 hours ect

Cell 3 is on its way to his 4th cycle, i will post results probably tomorrow. It is ~kinda promising?
The next design will be cylindrical and way easier to build, incl electrodes with a very easy venting way.
Wish i had a 3d printer with hdpe filament or with aother filament that could handle acid, but witch one?
 
Just some random thoughts/sketch
image_ihdamd.jpg
 
EXACTLY what I am intending to build....But without a lit...a tad more simpler....you will see, really small tad
And to think I made a stl file about it....3d printers will understand this


But you brought up a other challenge/question/problem.why not cut up the phonological foam. To give more room....for both and gain some.somehow I think this will work....and somehow I think this won't work...test and trail...thanks again for sharing your thoughts Ivo! I
I really appreciated this a lot, Ivo! Thanks



image_ngltzq.jpg

Charging values are write?
But they are good....


Yep, one is bulging...mah charge is going down.....there is a buble.
image_erhdsz.jpg
 
They did not make it throughout the night.....nothing to report...blade cells are not good, leaking everywhere.
build version 2 today.
https://secondlifestorage.com/showthread.php?tid=8772&pid=60170#pid60170

Testing results will be report on this topic, first post.
Thank for reading, any comment, thought, idea and whatever is very welcome.
Thanks for reading, best: Igor
 
Results so far for the new improved cell design:
Below 3.20v they will not take any charge, above 3.60v you will oxidize the zinc crystals.
Resulting in a to high resistance to give or take any power, meaning deplete them to 0v and start all over again.
Deplete them to 0vto start them fresh, give them 3 days and the zinc will dissolve back into the solution.
Going to test this a couple of times more for there max charge.

When they are done withdischargethey will go immediately back to 1.77v from 1.82v.
(Round two: 1.70v from 1.77v will give you ~80mah per 10cc.)
Every 10cc of 1.75 M solution will give you ~100mah.
No matter how big or small the cell is.(In my test cases 10cc 30cc 50cc and 100cc.)
All the cells have a id of 54 millimeters.
100mah per 10cc.
What is also interesting, the charge and discharge rate:
150-200mah charge, 100-150mah discharge.
All those numbers are keep on coming back!!!

Going to test with different electrodes:
Graphite foil, graphite rods, electrodes made of copper and carbon filled paper and ferrite rods.
Width and length.
Try to improve the electron flow.

With those numbers we can theoretically calculate what we need for a 12kwh storage.
28 jerry (25 liters)cans filled with 1.75 M solution, 2 M is almost not possible, or boil it of.
8 kilograms per jerrycan and 20 liters of distilled water.
Meaning we need 224kilograms of ZnBr2 and 560 liters of distilled water.
You can put 12 jerry cans on one square meter.

The cheapest ZnBr2 supplier out of 15 inquirieswas incl shipment ex vat or local tax (IF you are able to get this into your country) 1100 euro's, per 100 kilograms.
Drum with 200 liters of distilled water: ex shipping: 100 euro.
If we would go to 1.6 M solution, the costs for 10-11kwh storage, would be 300 for the water and 2200 for the ZnBr2
In lifepo4 same storage would be:3700e ex vat.
The ZnBr2needsalso a fullbms! for 28 to 30 cells per string for 48v to 50v

Conclusion: it works,it is cheaper, BUT you will have to deal with a very VERY low c rate.
You will need a lot of room, create your own bms, health hazard, ect

Protect yourself from metal fever (zinc oxide is also absorbed thru the skin, make absolutelyNOmistake about that!!!)
Metal fever is absolutely no joke, yes i was young and stupid once.
After a cell is charged, there is a cloud of bromine, this is a very poisonous liquid, evaporates very quick even at 0 celsius temps, don't let it escape, keep it under water, it is healthier to get copd from smoking.
Bromine will attack and permanently damage your organs, from nose, eyes and mouth to the other end.
Unless you hadthe proper education and have the rightstuff for this, don't go here.

I went further than a lot of published papers and experiments, at least i have some hard numbers!
Vacation well spent, i will not proceed to build a ess out of ZnBr2 cells.
Holy grail? sorry no, too much fuss to get the proper materials, building,controlling and maintaining the cell's.
Room for error is also to small.

Thanks for reading, tips,hints and thoughts that were shared.
 
Back
Top