STATUS of ESS/Storage Cells being purchased. (commodity cells)

Steve_S

Offgrid Cabineer, N.E. Ontario, Canada
Introduction:

I posted this into a response and felt that this should have its own thread as a point of clarification. There has been much debate, discusion and even theories, conjecture & postulation. All of this creates confusion due to many perceived points, real or otherwise. This is by no means the full and final word on the matter but rather more of an anchor in an unsteady sea. Some may disagree and that is perfectly fine. These are based on observations in real & practical use & application including some fairly extreme testing & even abusing of LFP (LiFePO4) cells. I will likely post a detailed thread about the Thrash Test Process I have been doing, warning that it's not for the light of heart as that progresses.

The figures represented are "General" averages. Always refer to the Manufacturers Specification Datasheet for the Cells you are working with. NEVER Assume they are all the same, they are NOT ! Even between different models from the same Manufacturer IE: Calb CA, CAM & SE models are LFP but very different.



I am sorry but this is more "to the point" bullet type points. Let me put this as plainly, clearly & simply as possible, as translation issues are considered. Some translation software can be overly creative shall we say. We have all suffered that so.

NONE of the EVE, Lishen or other brands of cells being sold as "Storage" or "ESS" or anything like that at these price points are MATCHED or even BATCHED.
NONE !

Vendors sort them by Voltage (typically 2.99 to 3.20 voltage - storage & shipping voltage) so that matches up. Then they double-check with a YR1030 or YR1035 to see if the standing IR (Internal Resistance) is the same. They in turn CALL THAT MATCHED ! In one sense, this is correct BUT it is misleading in the grand scheme of things, it is a play on Interpretation. BE AWARE OF THIS, you have been duly warned.

They will NOT match up during Charge or Discharge. They will deviate above 3.40V and below 3.10V without manually conditioning the cells by Bottom & Top Balancing. Top Balancing CAN improve the top side some but not the low side. This is also a very marginal gain related to the voltage curve.

You will get "RUNNERS". These are cells with variable IR. They will either RUN to 2.50V Fast from around 3.10V while the others are above 3.1V or RUN to 3.65V from 3.40V while the others are still at 3.850 and above. These Runners LIMIT the total battery pack capacity as they will force Low Volt Disconnect (LVD) or High Volt Disconnect (HVD) cutoff before the other cells reach that point. These are runners because they transition FAST. The "Runners" appear to be roughly 10% of cells sold. * Runners can be batched together (as a pack) if they match Top & Bottom giving a reasonable return, but it will be the "oddball" with tricky edges.

280AH cells can deviate as much as 280mv, or 1mv per AH capacity, either at the top or bottom. This seems to hold true to other sizes as well.

Manual Conditioning & preparation is ESSENTIAL for these COMMODITY CELLS. That is Bottom & Top Balancing.
Failure to do so, will not realize the full potential of the assembly.
* Nota Bene *
This will also result in havoc with SOC reading accuracy if not done.

Premium EV Grade, Grade-A Cells will cost 2X just to start PLUS the cost of Matching & Batching which can easily add $50+ per cell to the cost.

Fortunately: LFP optimal voltage is 3.2V per cell. The cells "as is" function well between 3.180 to 3.400 +/-. This is actually the primary curve for the chemistry.
- Below 3.00V is pointless, it is the end of the curve and only represents <5% of the AH Capacity.
- Above 3.50V is also only about 5% of actual capacity, Charge LFP to 3.65V till amps taken = 1A, within an hour the cell will settle between 3.55 to 3.50, this is NORMAL and not unexpected.
- Most folks, to ensure Long Lifespan & Cycle Life generally reserve off the top above 3.5 and below 3.0 anyways.

The ONLY way to verify IR and to manually test, sort & batch them is a process. An IR Tester such as this (the OEM, not a relabelled VAR) Real Four Wire Lithium Battery Internal Resistance Tester YR1035+ Nickel Hydrogen Lead Button Polymer Alkalinity|Battery Testers| - AliExpress is required. * Also available on EBay/Amazon, beware of knock offs, many are rebranded by VARS (value added resellers), Yoarea IS the OEM. (original equipment manufacturer)

Pre-Balancing Cells | Orion Li-Ion Battery Management System
High resistance cell | Orion Li-Ion Battery Management System

AN FYI EXAMPLE: Custom Job for matching cells (batteryspace.com) It is not a joke ! Matching & Batching = $$ due to time & resources required. Gear for doing large batches cost as much as a house !

⚠️ When discharging/charging cells for top/bottom balancing and after assembly, remember to "FIX" or "Clamp" the cells together. They will expand & contract up to 2mm per cell (normal) during their cycles. You do not want them to Puff or Bloat. REFER TO MANUFACTURER SPEC SHEET ! Typical recommendation is 12psi @ 50% SOC

TIGHTENING TORQUES:
- 70 kgf.cm (5.0 lbf.ft)
- 35 kgf.cm (2.5 lbf.ft)
- 7 to 12 kgf.cm (0.5 to 0.9 lbf.ft)

Preparation, the Devil in the Details:
Ensure that Cell Contact surfaces are clean of greases, waxes & oils (there is often cutting oil from tapping cells and waxes to protect from corrosion.
Also, clean your BusBars regardless of type, again Oils & Waxes can cause issues with readings.
GUARANTEE that there are no Burrs, Ridges or Edges on the busbars. The stamped ones all have a slight ridge. Made ones always have burrs.
-- File the ridges down or chamfer the holes slightly which is better.
You can use a very small amount of NoAlox/OxGuard on the cell contact surface to reduce corrosion potential. DO NOT GET ON THREADS ! It is a grease and will seriously affect PSI readings when tightening bolts. Stripping the thread = BAD.

NOTE, the above oils, waxes, ridges can and do affect readings, and can cause IR issues where cells posts will heat up due to poor contact. They can also create other side effects which are not immediately observable.

When Companies are Calculating the WH Rating of cells they use 3.20V which is the Nominal Voltage.
The Formula, 24V example: 3.2V X 8cells X 100AH = 2560 WH


lfp-voltage-chart-jpg.27632


Hope this helps, Good Luck.
Steve

IF WARRANTED, I would be willing to collaborate on an addendum related to doing Self Matching Cells. There are many "partial pieces" floating around on the forum and that should likely be compiled into one digestible "to the point" reference. That is presently beyond the scope of this particular thread.

Please, let us keep this ON TOPIC without side trips or hijacking. I do not want to be mean but once threads get cluttered with irrelevant meanderings they end up being near useless.
Thank You in Advance for your cooperation. :)
 
Last edited:

Steve_S

Offgrid Cabineer, N.E. Ontario, Canada
Links for some Known Good Benchtop Power supplies / Chargers which can be used for individual Cell preparation, Top Balancing and up to full pack charging.
There are several Brands/Models and with various price points. There are also some very good products out of China and elsewhere which people here are using but I am not personally familiar with them. Possibly someone may be able to provide some links for Known Good Benchtop Chargers.

SPECIAL NOTE:
A standard 100AH ESS cells are capable of 1C Charge & discharge rates (generally). It is generally recommended not to exceed 0.5C (50A) charge rate. A 100AH cell from 0% SOC (2.500V) to 100% SOC (3.650V) will take 1 full hour at 1C or 2 hours at 0.5C to charge in optimal conditions. Consider these factors when seeking a Power Supply. That is unless you are willing to waste days of time.
EV Graded Cells can go up to 3C or even 5C discharge & charge rates. These are significantly more costly!

Volteq variable DC power supply - Volteq - Reliable Regulated Variable DC Power Supplies
Mastech Variable DC Power Supply - Best Deals on Mastech Variable DC Power Supply (mastechpowersupply.com)
Multimter & Power Supply (tekpower.us)

Best Bench Power Supply For Electronics Hobbyists 2020 - Maker Advisor

⚠️ ALWAYS refer to the Manufacturer's specification sheet for Your Specific Make & Model of cells.
 
Last edited:

Steve_S

Offgrid Cabineer, N.E. Ontario, Canada

PING for @Will Prowse
I linked your video in here as I felt it is very relavent. I'm sure you may have another or two in your vast library that may also be suitable for inclusion. Additionally, I'd welcome your additional commentary. I feel it would be of great value to our membership here.
 
Last edited:

DJSmiley

Solar Addict
For powersupplies you can use a combination of a power supply and a CC/CV regulator.

On AliExpress there are some nice 30/40A converters available.

For example:
https://nl.aliexpress.com/item/32849390499.html will provide 40A with CC and CV options. ($20)
Can be used with a 12V powersupply to make a 40A charger when you set it to approx 14.2V
(You might have to replace the fan)

https://nl.aliexpress.com/item/32821840536.html will provide 30A with CC and CV options ($8)
Can be used with a 12V powersupply to make a 30A charger for topbalancing single cells

Combine both: Connect BMS, set HVD to 3.55V, charge with approx 14.2V till a cell hits HVD,
reconnect cells in series, topbalance them to 3.65V.

Another (more user friendly) option is the Riden RD6018, this is a 18A module with display and so on so you don't need seperate shunt/meters as you do with the above options. If you combine this with 2 pieces HP 450W power supply in series (12V/38A each) you can make a 18A fully adjustable power supply for <100 USD I think. You even can power 2 of those Ridens to make a dual output power supply (18A each output) for <$150
Output voltage of the Ridens is limited to the provided voltage, so if you use 2x12V in series the max output will be somewhere near 23-24V unless you modify them to output 14V each. You also can use 3 in series to make your powersupply to go up to approx 35V @ 18A
You also can use powersupplies for POE switches. For example a 1400W powersupply for the Cisco 6500 series will provide 27A @ 50V

For the primary powersupply: A cheap option is using old server power supplies. HP/Compaq uses for example 12V for their servers. You can get a 1000W ( 12V @ 83A ) supply for <$25 on Ebay. Plenty of 750-1000W units available, and they require only a wire/resistor (Depening on the model) to be used as a power supply.
(Note: Power rating generally is lower when powered by 100-120V, but still a massive amount of amps for the $$)

With some more modifications you can also disable the internal OVP, allowing it to increase the output voltage to 14V or so. (But i'm not sure if they have current limiting, or just shutdown/start hicking). But the above $20 unit will prevent that.
If you're really into electronics you can probably make something with an opamp which provides current-based feedback so the voltage will be automaticly adjusted down to stay within a certain range, making it a full 80A charger.

Some Cisco powersupplies also provide massive amps: the 1400W AC for the 4500 for example provides 12V / 115A (PWR-C45-1400AC )

If connecting multiple in parallel... you can get over 200A for <$75. And in that case the overcurrent protection probably will be your wiring ;)
 
Last edited:

michael d

Solar Addict
nice bench power supply 3010 has 3 4 digits readout of amps, volts, and watts; Variable cc cv 30 volt 10 amps. I set at 26.8 and was pushing about 9 amps into my BYD 5kwh 24-volt battery boxes. I charged 4 of them this way. as the voltage gets closer to the set voltage(CV constant voltage) on the bench power supply the amperage keeps decreasing. it started at about 300 watts. It works great and will not overcharge batteries. cost $62.69 tax and delivery to my farm in South Dakota via Amazon Prime (2-day delivery). it has fine and coarse adjustments for the voltage and also fine and coarse adjustments for the amperage settings.


I clamped the small alligator clamps from the bench power supply onto the 2 gauge cable lug (from the 175 amp Anderson plug on the BYD box - which I hydraulically crimped and heat-sealed) mounted to an insulated stud on a piece of plywood. watched it and periodically tested it with my cheap craftsman multimeter. worked great. plan to use this on the new 280Ah Lishen cells (along with a Riden 6012W bench power supply, and a 30 ampLiFePO4 charger. more details later on those.
the uninterruptible power supply project (aka power shed continues).

4 BYD box LiFePO4 24V 5kwh 200Ah battery pack with bms 130A LiFePO4
 

Attachments

  • 20201209_162544.jpg
    20201209_162544.jpg
    58.4 KB · Views: 36
  • 20201209_162153_Burst01.jpg
    20201209_162153_Burst01.jpg
    62.6 KB · Views: 36
  • 20201209_162358.jpg
    20201209_162358.jpg
    67 KB · Views: 36

DJSmiley

Solar Addict
Do you have the Riden already? A new version is now out for a few months (6018, 18A) and a 24A version is announced for next year.
If you combine that with a nice 48V POE supply (eg from an old switch) you can make a hell of a powersupply.
 

Steve_S

Offgrid Cabineer, N.E. Ontario, Canada
I just purchased & received a TekPower 15/40E, 15VDC, 40A capable bench supply.
I'm actually somewhat impressed with what I received and the material that came with it is very good. Connectors & all.

They are presently on sale for $189.99 USD, direct or through eBay & Amazon as well.
TekPower TP1540E DC Adjustable Switching Power Supply 15V 40A Digital Display (North American Model 120VAC input)

1540e_1.jpg


This will start getting a "workout" later on today. Will report if anything is amiss but am confident there won't be.
 

Bob142

Build more, learn more.
I just purchased & received a TekPower 15/40E, 15VDC, 40A capable bench supply.
I'm actually somewhat impressed with what I received and the material that came with it is very good. Connectors & all.

They are presently on sale for $189.99 USD, direct or through eBay & Amazon as well.
TekPower TP1540E DC Adjustable Switching Power Supply 15V 40A Digital Display (North American Model 120VAC input)

1540e_1.jpg


This will start getting a "workout" later on today. Will report if anything is amiss but am confident there won't be.
Looks identical to the one I've been using for top balancing. 15V 40A power supply from circuit specialists. $169.15 USD.

It's been great and stays cool as a cucumber putting out 30A set at either 14.6V or 3.65V for hours on end. I've used it on 20 EVE 280s so far without issue.

IMG_3492.jpg
 

Steve_S

Offgrid Cabineer, N.E. Ontario, Canada
Looks identical to the one I've been using for top balancing. 15V 40A power supply from circuit specialists. $169.15 USD.

It's been great and stays cool as a cucumber putting out 30A set at either 14.6V or 3.65V for hours on end. I've used it on 20 EVE 280s so far without issue.

View attachment 29866
TekPower is Made in China and likely is available rebranded. I got the TekPower because it was readily available and fast to get and has good reviews. Couple of folks here have them and they are happy with them, which was also a selling point (reassurance wise).
 

Bob142

Build more, learn more.
TekPower is Made in China and likely is available rebranded. I got the TekPower because it was readily available and fast to get and has good reviews. Couple of folks here have them and they are happy with them, which was also a selling point (reassurance wise).
I'm sure it's gonna be a great addition to your arsenal.
 

michael d

Solar Addict
Do you have the Riden already? A new version is now out for a few months (6018, 18A) and a 24A version is announced for next year.
If you combine that with a nice 48V POE supply (eg from an old switch) you can make a hell of a powersupply.
I have the Riden 6012w. of course, they make it bigger and better after you buy it. I had to buy the battery separate, the power cord separate, the banana plug alligator leads separately. I bought the 6012w and the power supply and the case together from one vendor. then found out all the other missing but required parts to assemble the Riden 6012w! sorry, about the long reply but I hope the Riden works out well for some stuff anyways. :) more on it later.
 

Steve_S

Offgrid Cabineer, N.E. Ontario, Canada
Gentle Reminder ;-)
Please, let us keep this ON TOPIC without side trips or hijacking. I do not want to be mean but once threads get cluttered with irrelevant meanderings they end up being near useless.
Thank You in Advance for your cooperation. :)
 

michael d

Solar Addict
Gentle Reminder ;-)
Just a small question? what does ESS stand for? Electric Solar System??? I am going specifically a 24-volt Electrodacus SBMS0 etc right now for the new LiFePO4 Lishen 280Ah 3.2V cells in a block-shaped bank of 16 (2P8S). I read and study daily so hopefully not to learn the hard way with the expensive LiFePO4 cells. they are a major expense. but I like the monitoring and balancing the Electrodacus SBMS0 offers. an I backed off a little from the 48-volt system for the moment. thanks in advance for any replies you provide to me. I know some of the threads in the forum are terribly long and you have to read and weed thru some comments that do not pertain. :unsure:
 

efbasham

New Member
This quickly expands to an "N-dimensional" space when you take into consideration temperature, and rate of charge (C/N) + rate of discharge. I'm still trying to form an apples to apples comparison when calculating $/KW-h in LiFePO4 because not all LiFePO4's are the same. Too many people are saying "I got a good deal on X brand 180A-h for ONLY $100" only to have the inevitable debater saying "Hah! I can get TWICE that much capacity for half the price in no-name brand blue wrapped cases so you got taken...." It's hard enough when name brand cells are 10-15% over their nameplate rating while the magical blue wrapped cells often struggle to even achieve their advertised rating.



CALB-LabTestOf4EAca180.jpg

This graph started out to illustrate the capacity given up as design margin or life expectancy gains on the low end - it's a 35 hour discharge of 4S4P CA180 cells (nameplate Vmax=3.65, Vmin=2.5) but discharged from 13.80VDC float until the first cell hit 2.800VDC. The center "0" line is really 3.000VDC but I moved it to magnify the 3.2-3.0V segment. Stopping at 3.0 VPC would have given up 49381 Amp-seconds in the area under the 3.0-2.8 VPC discharge band, or 7.6% of the nameplate (180Ah) rating. The next run will look at the top capacity between 3.45 and 3.65 plus compare charge rate.

Also, this test shows how closely spaced good quality cells are WITHOUT any active balancing. There are 4 lines in that chart that don't diverge until the bottom. I prefer the QNBBM bi-dir dc/dc converter, but it was removed for this test, which spanned 35 hours.

My battery testing process is based on using a Keysight 34970A data logger + matrix switch card & 6.5 digit DVM + HP 6050A DC electronic load w/ 6 x 60A modules. I have several HP CV-CC power supplies up to 1KW DC for charging, plus several inverter/chargers that don't output a DC waveform like a power supply making me wonder if the "real world" charging picture is better done using an inverter-charger rather than a Lab grade DC supply. I'm curious if other people have experience with varying the charge & discharge rate against an established discharge "standard" like say 20h or C/20? W-T makes mention of it often in their user's manual with sets of curves.

Given that every shop with blue film and aluminum sheet metal seems to suddenly be cranking out LiFePO4 cells in China, I'd be interested in defining a REPRODUCIBLE test standard for capacity and maybe overall quality, just like there is a test standard for 4-wire AC IR measurement. Once I define some reasonable limits, life cycle testing is next. I think most China mfgr's do not expect their end customers to be able to conduct cycle-related capacity loss measurements. We're starting to see climbing specs like "5000 cycles to 70%" supported by a downwards sloping graph that was straight outta Matlab, not Labview or Benchlink.

f at hplab dot com
 

efbasham

New Member
Below graph shows a zoomed in view of how fast terminal voltage drops below float level and approaches 3.300 VPC (y - axis zero point on this zoom view) - apparatus setup is in chart info & final full plot to arrive later today. By ~2500 seconds, each cell reaches ~3.301 VPC. No bi-dir dc/dc active balancers were used on this array. I'll be comparing the discharge energy at this slightly higher float voltage to the prior curve I posted when the first cell reaches 2.800 vpc. f @ hplab dot com

1607871778714.png
 

Bearpaw

Wading Trough 💩 To Excellence
Steve_S,
Thank you for this incredibly detailed and informative series of posts. Absolutely fantastic! Of course that caused an avalanche of a whole new line of valuable posts with even more information. Forget Marvel! You guys are super heroes in my world!
 

KenDan

Solar Enthusiast
I have a question on clamping cells together. This seems to be a point of contention for some folks.
Slight cell swelling/bloating is a result of internal out-gassing which I think is normal to a certain extent when under load. I don't think that clamping cells together under pressure will prevent this. The outgassing still occurs and will still cause some positive pressure inside the cell. I believe the outgassing is mostly absorbed or reversed when the cell is not under charge or discharge load. The amount of outgassing and swelling slowly increases over time as the cell has been through more charge-discharge cycles.

If the cells are partially constrained, they will still expand to a certain extent. Some of the force will then be transferred to the buss bars and battery terminals. I would think that in this situation it would be advisable to use flexible buss bars, or even short cables between cells to eliminate the stress on the battery terminals.

This brings up the question - why not just allow the cells expand and contract freely to relieve stress on the terminals? Allowing some space between cells also provides some airflow which could be beneficial.

I could be wrong - but I have long assumed that clamping of cells was done in EV applications to prevent then from rubbing on each other and possibly wearing through the protective shrink wrap and causing a short. In non-EV applications the cells are not shifting around so there is no risk of this happening.
 
Last edited:

Gazoo

Dumb Dumb
As far as EVE cells, if compression is done properly the cycle life is increased considerably. From 2000 to 3500 cycles. Compression helps to prevent delamination of the "jelly roll" which is housed inside of the pouch. I would think this applies to all prismatic aluminum cells.

I think there is no way to totally keep the cells from expanding and contracting. And because some of us have received undulated cells this becomes an even bigger problem. Therefore I agree the best and possibly only way to relieve terminal stress is to use braided busbars or short cables....IMO.

Anyways this is off topic and this discussion should be kept in one of the many other threads that is has or is being discussed in. Sorry Steve.
 

KenDan

Solar Enthusiast
As far as EVE cells, if compression is done properly the cycle life is increased considerably. From 2000 to 3500 cycles. Compression helps to prevent delamination of the "jelly roll" which is housed inside of the pouch. I would think this applies to all prismatic aluminum cells.

I think there is no way to totally keep the cells from expanding and contracting. And because some of us have received undulated cells this becomes an even bigger problem. Therefore I agree the best and possibly only way to relieve terminal stress is to use braided busbars or short cables....IMO.

Anyways this is off topic and this discussion should be kept in one of the many other threads that is has or is being discussed in. Sorry Steve.
Well, this was in direct response to a highlighted statement made in the original post. That's why I asked the question here...

"⚠️ When discharging/charging cells for top/bottom balancing and after assembly, remember to "FIX" or "Clamp" the cells together. They will expand & contract up to 2mm per cell (normal) during their cycles. You do not want them to Puff or Bloat. REFER TO MANUFACTURER SPEC SHEET ! Typical recommendation is 12psi @ 50% SOC"
 
Top