Cheap 4kwh LiFePO4 batteries from Battery Hookup

[HMR]

Just got back from 2 weeks vacation and a lot changed while I was away. David Poz reported 3 KwH capacity in his 8 batteries, others here on this forum came in between 2.7 - 3.5 Kwh capacity, Will Prowse got indignant and talked with Batteryhookup owners and others and got them to change their advertised capacity and offer some form of compensation to those of us that bought these BYD batteries from them. Wow! The power of communication and community at work. I called batteryhookup today and suggested that in compensation for them not being 4 - 4.5 KwH as advertised I get two free units and I pay the shipping cost of $350 and that would square things up. Bryan at batteryhookup readily agreed and everyone is satisfied and happy. Now I will have 6 BYD units each at about 3 KwH for 18 total KwH which meets my 16 KwH design target. I am prepared to assume the risk of how long they will last in operational conditions. That brings my entire battery bank cost down to $130 / KwH. Thanks to batteryhookup for being a very responsible and responsive business partner.

I got a chance during my vacation to read the entire series of Nordqyst LFP files on the internet as people here on the forum suggested. One of these files came out very strongly showing that research data supports the theory that LFP batteries definitely develop memory effect capacity limitations due to incomplete charging and discharging. Repeated full discharging and full charging will eventually eliminate these reduced capacity limitations. Would those on this forum that do not agree with the Nordqyst contention on memory effects offer some proof of an opposing view? Thanks.

 

[nebster]

I got a chance during my vacation to read the entire series of Nordqyst LFP files on the internet as people here on the forum suggested. One of these files came out very strongly showing that research data supports the theory that LFP batteries definitely develop memory effect capacity limitations due to incomplete charging and discharging. Re

We need more data. If you (or the Nordkyn site) have links to studies, that would be helpful.

Some of us cannot seem to re-create this effect. But it may be slow to develop, or specific to a particular charging strategy, or something else.

 

[HMR]

Sorry about the misspelling. This is the Nordkyn article that references the memory effects for LFP. --> memory effects . The reference is to a Japanese research article listed in the reference section. This discussion is about half way down into the article. Let us know what you think.

 

[fastline]

I think some of the facts are being ignored here. They were advertised originally as 4-4.5kwh. that means when you get them, they should hit that mark. It is a bad business practice to just change the capacity AFTER many have already purchased and the real capacity is realized.

I firmly believe these batteries were cycled HARD and FAST in their original installation as evidenced by the massive heat sinks, fan cooling, and visual evidence on busbars of discoloration. I even have one where the plastic has melted. I think providing the "they just need cycled" argument is an avenue to excuse the findings in hopes to move customer complaints down the road far enough that returns or compensation is not an option.

There is at least one user that is logging them with 2 months of cycling and no substantial improvements in capacity.

In our tests so far, all cycles are showing nearly identical results AND there is evidence of self discharge which is further proved with how much power it takes to charge one! Someone else needs to log that but we are showing about 2x the power to 'charge' them vs the 'discharge capacity'. That is because the self discharge is counteracting the charge.

We are not horsing around with solar, MPPT, etc, we are directly charging and discharging with grid power to expedite testing, which is most certainly costing us time and money!

 

[HMR]

I think some of the facts are being ignored here. They were advertised originally as 4-4.5kwh. that means when you get them, they should hit that mark. It is a bad business practice to just change the capacity AFTER many have already purchased and the real capacity is realized.

I firmly believe these batteries were cycled HARD and FAST in their original installation as evidenced by the massive heat sinks, fan cooling, and visual evidence on busbars of discoloration. I even have one where the plastic has melted. I think providing the "they just need cycled" argument is an avenue to excuse the findings in hopes to move customer complaints down the road far enough that returns or compensation is not an option.

There is at least one user that is logging them with 2 months of cycling and no substantial improvements in capacity.

In our tests so far, all cycles are showing nearly identical results AND there is evidence of self discharge which is further proved with how much power it takes to charge one! Someone else needs to log that but we are showing about 2x the power to 'charge' them vs the 'discharge capacity'. That is because the self discharge is counteracting the charge.

We are not horsing around with solar, MPPT, etc, we are directly charging and discharging with grid power to expedite testing, which is most certainly costing us time and money!

Well, I am going to be an advocate here. Like Jason has been saying all along: its not an absolute standard here but a relative one. Yes these are used batteries and yes the sellers should have done better at describing their actual capacity. But face it, its a hard thing to do. It took this user community 4 months to do their capacity tests and report them. These people are in business and they cant afford that. Yes, they reported optimistic estimates of capacity for these batteries, but when proven wrong, to their credit they stepped up to the plate and were willing to compensate buyers. That is a hell of an honorable response in these USA in 2020. I personally applaud them!
As to whether we can expect increased capacity with more cycling, I think the jury is still out on that issue.

 

[Picasso]

Only one vendor has been stepping up. techdirctclub.com / bigbattery.com still shafts me.

 

[Maast]

In our tests so far, all cycles are showing nearly identical results AND there is evidence of self discharge which is further proved with how much power it takes to charge one! Someone else needs to log that but we are showing about 2x the power to 'charge' them vs the 'discharge capacity'. That is because the self discharge is counteracting the charge.

Where was the self discharge and reduced round trip efficiency posted? I've got 13 of these modules and have been following all the BYD posts closely and I must have missed this one.

 

[fastline]

The self discharge has been noted by others. The coulombic efficiency is something we are looking at and does not appear another single sole has looked into it. The reality is if a cell will not take a full charge or exhibits self discharge effects, the actual capacity within the battery will be substantially less. We first did probably what others are doing which is to get a full charge, then drop the hammer on discharge. I would encourage others to get to a full state of charge, then delay discharge for 24-48hrs. This is painting a completely different model for us on these batteries.

Li batteries should have very high coulombic efficiency so this is very telling about the internal degradation. It is going to suck wildly when people are paying to run a generator to charge these and not even realize they have to pump 2x the watts in. In short, it is like filling a pool with the drain still open!

I can assure these batteries were design for very high C charge/discharge based on design and ANYONE that posts anything above 3kwh capacity better be willing to prove the discharge rate. In some checking, I think these batteries were probably "rated" at the .5C or even 1C discharge rate as they were designed and intended to be pushed well beyond that. Obviously if you throw a .00005C discharge at them, the numbers will 'look' impressive.

 

[Supervstech]

My response from BH, just came in a few minutes ago... those guys must be swamped with requests.

THANK YOU WILL for pushing this issue!

 

[Picasso]

Where was the self discharge and reduced round trip efficiency posted? I've got 13 of these modules and have been following all the BYD posts closely and I must have missed this one.

my weakest pack only seems to blow / miss about 7 ah. aka what I put in vs what I get out in a cycle, this pack has been getting daily cycled now for a few weeks. this weak pack is just doing 75ah ish per cycle. I cant put in or take any more out of it before I get cell dive or the same cell sky high on charge.

 

[Supervstech]

On mine, if I charge them to 3.5V per cell, they drop to 3.34V per cell overnight.

cannot get them to reach and hold 3.4V even! That’s a good chunk of storage right there!

 

[Supervstech]

my weakest pack only seems to blow / miss about 7 ah. aka what I put in vs what I get out in a cycle, this pack has been getting daily cycled now for a few weeks. this weak pack is just doing 75ah ish per cycle. I cant put in or take any more out of it before I get cell dive or the same cell sky high on charge.

What voltage do you shoot for?

 

[fastline]

I will mention so as not to mislead, these batteries, as many other chemistries, will exhibit some "surface charge" so we expect some roll off but typically this will NOT take 2x the charge input just to get to 3.65V/cell. I would have issue with anyone reporting voltage values with a $5 meter because resolution and accuracy are important. We use several Fluke DMMs and analyzers that are certified and accurate.

It is also important to mention that only 10% of the SOC pushes up into the voltages we discuss here, BUT for the sake of determining actual capacity, and all the issues concerning these batteries, we have determined this is needed to see what we have here!

 

[HMR]

Summary of charge/discharge cycles of two of my six BYD batteries. I am charging up until the first cell goes over 3.5 v (where I set my charge controller, ToolKit RC M8S, to cut off). My charge rate has been either 5 amp or 8 amp. I discharged until the first cell goes below 2.8 v. My discharge rate has been 29 amp. My measurement instrument is a 100 amp shunt meter. Here are my results so far:

Battery Number one
Cycle 1: charge: 3.54 KwH; 133 Ah discharge: 2.76 KwH; 110 Ah Disch/Charge Eff: 78%
Cycle 2: charge: 3.18 KwH; 120 Ah discharge: 2.94 KwH; 116 Ah Disch/Charge Eff: 92 %
Cycle 3: charge: 3.19 Kwh; 120 Ah discharge: 2.87 KwH; 114 Ah Disch/Charge Eff: 90%

Battery Number 2:
Cycle 1: charge: 2.64 KwH; 99 Ah discharge: 2.29 KwH; 91 Ah Disch/Charge Eff: 87%
Cycle 2: charge: 2.77 KwH; 104 Ah discharge: 2.57 KwH; 101 Ah Disch/Charge Eff: 93%
Cycle 3: charge: 2.79 Kwh; 105 Ah discharge: 2.56 KwH; 101 Ah Disch/Charge Eff: 92%

Average Useable capacity = 2.7 KwH Average Disch/Charge Eff: 89%

Typical voltages at end of charge cycle: battery voltage: 27.3 VPC range: 3.37 – 3.54 (3.5 HVC)
Typical voltages after 24 hrs: battery voltage: 26.6 VPC range: 3.31 – 3.34
Most importantly: my batteries hold 26.6 volts VPC 3.31 – 3.34 for 3 weeks (longest time of observation so far) without any decrease in battery voltage or VPC.

Conclusion: I am getting app. 2.7 KwH with 89% efficiency and no detectable self-discharge leakage per battery. I plan to have 6 of these BYD in parallel at 24 v with app. 16.2 KwH storage for a total cost of $ 148 per KwH. I am very pleased with these results. My main remaining unknown now is how long these BYD batteries will perform at this capacity with gentle handling; ie. discharge until lowest cell reaches 2.8 v and charge until highest cell reaches 3.5 v at sub-C rates. Is there any competing LFP battery currently available that costs less than $148 per KwH?

 

[Darkstar]

Summary of charge/discharge cycles of two of my six BYD batteries. I am charging up until the first cell goes over 3.5 v (where I set my charge controller, ToolKit RC M8S, to cut off). My charge rate has been either 5 amp or 8 amp. I discharged until the first cell goes below 2.8 v. My discharge rate has been 29 amp. My measurement instrument is a 100 amp shunt meter. Here are my results so far:

Battery Number one
Cycle 1: charge: 3.54 KwH; 133 Ah discharge: 2.76 KwH; 110 Ah Disch/Charge Eff: 78%
Cycle 2: charge: 3.18 KwH; 120 Ah discharge: 2.94 KwH; 116 Ah Disch/Charge Eff: 92 %
Cycle 3: charge: 3.19 Kwh; 120 Ah discharge: 2.87 KwH; 114 Ah Disch/Charge Eff: 90%

Battery Number 2:
Cycle 1: charge: 2.64 KwH; 99 Ah discharge: 2.29 KwH; 91 Ah Disch/Charge Eff: 87%
Cycle 2: charge: 2.77 KwH; 104 Ah discharge: 2.57 KwH; 101 Ah Disch/Charge Eff: 93%
Cycle 3: charge: 2.79 Kwh; 105 Ah discharge: 2.56 KwH; 101 Ah Disch/Charge Eff: 92%

Average Useable capacity = 2.7 KwH Average Disch/Charge Eff: 89%

Typical voltages at end of charge cycle: battery voltage: 27.3 VPC range: 3.37 – 3.54 (3.5 HVC)
Typical voltages after 24 hrs: battery voltage: 26.6 VPC range: 3.31 – 3.34
Most importantly: my batteries hold 26.6 volts VPC 3.31 – 3.34 for 3 weeks (longest time of observation so far) without any decrease in battery voltage or VPC.

Conclusion: I am getting app. 2.7 KwH with 89% efficiency and no detectable self-discharge leakage per battery. I plan to have 6 of these BYD in parallel at 24 v with app. 16.2 KwH storage for a total cost of $ 148 per KwH. I am very pleased with these results. My main remaining unknown now is how long these BYD batteries will perform at this capacity with gentle handling; ie. discharge until lowest cell reaches 2.8 v and charge until highest cell reaches 3.5 v. Is there any competing LFP battery currently available that costs less than $148 per KwH?

Where did you buy your BYD's?

What are you using for BMS?

Have you done any balance charging on the packs?

 

[HMR]

Where did you buy your BYD's?

What are you using for BMS?

Have you done any balance charging on the packs?

BatteryHookup.com
For capacity testing I am not using a BMS. My ToolkitRC M8S charge controller takes care of HVC. My BattGo cell monitor sounds alert for LVC. I know from experience now that it takes app. 3.5 hours at discharge rate of 30 amp to get close to LVC of 2.8 v on my worst cell. So I just manually watch it and cut it off.
The TookitRC M8S is a balance charger and keeps all my cells in balance on the way up until they start going into the knee and I set it to cut off when the first cell goes over 3.5v. It works quite well automatically. I wish I had an automatic way to cut off at the low end also so I dont have to be so diligent.
I am moving on right now to investigate which BMS to buy - probably a Chargery one like SteveS has so diligently explored.

 

[davissd]

BatteryHookup.com
For capacity testing I am not using a BMS. My ToolkitRC M8S charge controller takes care of HVC. My BattGo cell monitor sounds alert for LVC. I know from experience now that it takes app. 3.5 hours at discharge rate of 30 amp to get close to LVC of 2.8 v on my worst cell. So I just manually watch it and cut it off.
The TookitRC M8S is a balance charger and keeps all my cells in balance on the way up until they start going into the knee and I set it to cut off when the first cell goes over 3.5v. It works quite well automatically. I wish I had an automatic way to cut off at the low end also so I dont have to be so diligent.
I am moving on right now to investigate which BMS to buy - probably a Chargery one like SteveS has so diligently explored.

How did you wire your ToolkitRC M8S ? Can you post some picks?

 

[Darkstar]

BatteryHookup.com
For capacity testing I am not using a BMS. My ToolkitRC M8S charge controller takes care of HVC. My BattGo cell monitor sounds alert for LVC. I know from experience now that it takes app. 3.5 hours at discharge rate of 30 amp to get close to LVC of 2.8 v on my worst cell. So I just manually watch it and cut it off.
The TookitRC M8S is a balance charger and keeps all my cells in balance on the way up until they start going into the knee and I set it to cut off when the first cell goes over 3.5v. It works quite well automatically. I wish I had an automatic way to cut off at the low end also so I dont have to be so diligent.
I am moving on right now to investigate which BMS to buy - probably a Chargery one like SteveS has so diligently explored.

I am considering trying the chargery as well. Is anyone getting their orders filled for chargery stuff lately? I ordered an ANTbms from chargery's retailer IC GOGOGO 2 months ago and still no sign of it.

David Poz uses what seems to be a single chargery 16T to manage an 8 pack 48v battery. He has the individual cells paralleled to each other so I think that is how he accopmplishes using a single BMS.

I have 14 byd packs here and the first 2 that i have balance charged to 3.6V/cell are holding at 27.7v overall.

 

[HMR]

I have 14 byd packs here and the first 2 that i have balance charged to 3.6V/cell are holding at 27.7v overall.

How did you "balance charge" your 2 BYD's and get all your cells up to 3.6 v without one or more going overvoltage in this knee area? and why are yours staying at VPC of 3.6 when most of the rest of us are reporting an quick drop out of the knee area back down to 3.31 - 3.34 stable VPC as soon as we stop charging?

 

[HMR]

How did you wire your ToolkitRC M8S ? Can you post some picks?

OK. will do. but give me a couple of days. Got to go now.