BMS connection cables with series / parallel lithium batteries

[MIKEJAMES]

New to this site, based in Wales UK.
With a 3 kW solar array, I wish to install 16 lithium batteries (8 in series x 2) and then in parallel giving 24 volts.
This was advice from reliable supplier who also supplied a JK BMS.
How do I connect the control cables to the batteries? The manual only advises batteries in series. Any advice appreciated.

 

[Steve_S]

Unless you have properly Matched & Batched Cells do NOT put the cells in parallel. MANY Vendors tell you they are matched but they are NOT ! Most often they are Voltage & IR Matched with a hand device like a Yaorea YR1035+ Battery Tester at STATIC Storage Voltage. Proper Matching & Batching is a long process where the IR (Internal Resistance) is tested at ever 0.50mv through the working voltage range (3.000-3.400) with calibrated charge & discharge. Batching puts cells with identical IR throughout the range together so that they perform equally during use. This is why EV's & Commercial ESS systems all use Matched & Batched cells.

You have 16 Cells. Assemble two 8S Simple Batteries each with their own BMS, and set the batteries in Parallel to a common DC Bus which the SCC & Inverter/Charger are connected to. By putting battery packs in parallel within a "Bank" you are effectively dividing the Load & Charge between the packs thereby reducing the stress on them and increasing longevity and doubling their output capacity handling as a result. This also provides for Fall Back Fault Tolerance because if one pack cuts off for any reason the other will continue to carry the load.

A consideration for this, is to ensure that each battery can handle the full load or charge anticipated for the system. So if you are building with 280AH cells then get 250A BMS' for each battery pack, collectively they can deliver up to 500A.

FYI a Panasonic 1200W Inverter Microwave uses 77A off a 24V battery, an Oster Coffee Maker with Thermal Carafe (no warmer element) uses 55A while brewing off a 24V system.

24V/250A = 6000W
6000W ÷ 120VAC = 50A
24V can safely and easily handle up to 5000W output, while allowing for Surges. If Above 4000W (normal draw) then you should consider 48V system.

Links to helps you along your path
NEW JK-BMS Manual V: 2.2 2022/07/22
General LiFePO4 (LFP) Voltage to SOC charts/tables 12/24/48V
Luyuan Tech Basic Lifepo4 Guide V1.0A

Hope it helps, Good Luck.
Steve_S

 

[MIKEJAMES]

Unless you have properly Matched & Batched Cells do NOT put the cells in parallel. MANY Vendors tell you they are matched but they are NOT ! Most often they are Voltage & IR Matched with a hand device like a Yaorea YR1035+ Battery Tester at STATIC Storage Voltage. Proper Matching & Batching is a long process where the IR (Internal Resistance) is tested at ever 0.50mv through the working voltage range (3.000-3.400) with calibrated charge & discharge. Batching puts cells with identical IR throughout the range together so that they perform equally during use. This is why EV's & Commercial ESS systems all use Matched & Batched cells.

You have 16 Cells. Assemble two 8S Simple Batteries each with their own BMS, and set the batteries in Parallel to a common DC Bus which the SCC & Inverter/Charger are connected to. By putting battery packs in parallel within a "Bank" you are effectively dividing the Load & Charge between the packs thereby reducing the stress on them and increasing longevity and doubling their output capacity handling as a result. This also provides for Fall Back Fault Tolerance because if one pack cuts off for any reason the other will continue to carry the load.

A consideration for this, is to ensure that each battery can handle the full load or charge anticipated for the system. So if you are building with 280AH cells then get 250A BMS' for each battery pack, collectively they can deliver up to 500A.

FYI a Panasonic 1200W Inverter Microwave uses 77A off a 24V battery, an Oster Coffee Maker with Thermal Carafe (no warmer element) uses 55A while brewing off a 24V system.

24V/250A = 6000W
6000W ÷ 120VAC = 50A
24V can safely and easily handle up to 5000W output, while allowing for Surges. If Above 4000W (normal draw) then you should consider 48V system.

Links to helps you along your path
NEW JK-BMS Manual V: 2.2 2022/07/22
General LiFePO4 (LFP) Voltage to SOC charts/tables 12/24/48V
Luyuan Tech Basic Lifepo4 Guide V1.0A

Hope it helps, Good Luck.
Steve_S

Many thanks Steve-S for your response. Your warning about LiFe PO4 280 AH cells in parallel concerns me. Only part of the Solar array is presently roof mounted and can be changed to 48v. Grid voltage here is 230v AC 50 cps so currents will be reduced and balancing problems reduced hopefully. Many thanks also for the Links. I have been a licensed Radio Ham for over 50 years but modern battery technology has passed me by. I hope I may feel to contact you again.
Regards Mike.

 

[Steve_S]

The Watts, Amps & Voltages from the solar array has little to do with the batteries on the other end of it. The Solar Controller (SCC) is what manages & handles that. Most SCC's can do 12/24/48 Volts Batteries but pending on what voltage of battery you have the SCC will have limits as to the Watts, Amps & Volts it requires to deliver the max charging amps it is designed to do.

You are a HAM OP so you should know more than the average person out there as a byproduct of your hobby. I have lived from Coast to Coast and from our most southern tip to the Deep Arctic (Iqaluit Nunavut) and everywhere uses 120/240VAC 60 Hz. I lived in Calgary & Edmonchuck for a few years and am well aware of Alberta Power systems as well.

My "warning" is just a fact of life which applies to all cells regardless of AH Capacity, it is something that needs general understanding and acceptance that many vendors "play numbers games" and sometimes will use ongoing misinterpretations & misunderstands to further their malarchy. Many are quite good at mis-representation and doing it in a way that they get away with it.

Example.
EVE LFP 280K are rated for 6000 Cycles to 80% Gross Capacity, some companies advertise 280AH cells to 10,000 cycles BUT to a 70% Gross Capacity remaining and they hype that their cells are good for 10,000 cycles... The "Industry Standard" is to 80% Gross Capacity remaining...

Many companies sell cells rated at 280AH and state you will get 280AH from those cells. The GOTCHA in this, is they will take B Grade cells and test them from 2.500 to 3.650 using the Gross AH to show that they will deliver that BUT in reality THIS IS FALSE and horribly misleading! A "Proper Grade-A" Cell will come out between 288-295AH Gross AH but 277-282AH Net AH.

All Batteries of ALL Chemistries have TWO Voltage Values.
- The Allowable Voltage Range, which is the voltage range that they can operate WITHIN without hard/damage/degradation. For LFP this is from 2.500-3.650 Volts per cell. The AH from 2.50-3.65 is the GROSS AH Rating and this is NOT what the cells will deliver in reality.
- The Working Voltage Range is from 3.000-3.400 Volts per cell, this is where the NET AH comes from which the cells will actually deliver. This is the primary reason that ALL LFP Nominal Voltages are stated at 3.200Vpc (50% SOC).
** All LFP will "Settle" post charge input, so even if you charge your cells to 3.650 and allow Amps Taken to drop down to 2A and fully saturate the cells, stop charge input and without ANY load at all, the cells will settle and within an hour you'll be seeing 3.500Vpc on average and another hour later they'll be settled to around 3.450Vpc +/- a wee bit.

LFP has an Extremely FLAT voltage curve from 3.000-3.400. Above 3.400 to 3.650 is the sharp cliff climb on the voltage curve and at most 10AH but usually about 5AH Gross Capacity and the cells will deviate from each other at these voltages as you are exceeding the working range. Below 3.000 is a cliff fall on the voltage curve and again at most represents 10AH but usually about 5AH of Gross Capacity.

Properly Matched & Batched cells that are Grade-A will deliver that 280AH from the Working Voltage Range of 3.000-3.400.
There is of course a Price Difference that goes with that and 95% of folks don't wanna pay that not realizing the down the road implications.

FYI:
I am 100% Offgriid solar powered up near Algonquin Park Ontario. I run off 24V with 1295AH/33kWh battery bank consisting of
1x 24V/105AH (Matched/Batched Grade-A),
2x24V/174AH (Used EV Grade LFP - Thin Prismatic) &
3x24V/280AH (2 Packs of B-Grade & 1 Pack with Grade-A Matched/Batched)
All managed by JKBMS' with 2A Active Balancing.
You can see more details here: About My System (updated NOV.2022) which is a tad out of date as I am making changes now and again in spring when I add the next Midnite Solar Classic-150 SCC and it's array of Q-Cell 395's.

Be careful of Wives Tales generated from Misinformation & misunderstandings... There is piles of it out there and most especially on Youtube ! So many "fools" call a Single Battery a "Bank" because it has several "cells" in the battery, but these people also think that a car battery is One Cell when it is not. A Bank is obviously 2 or more battery packs (regardless of the number of cells within each pack) in parallel.

Hope it Helps, Good Luck.

 

[MIKEJAMES]

The Watts, Amps & Voltages from the solar array has little to do with the batteries on the other end of it. The Solar Controller (SCC) is what manages & handles that. Most SCC's can do 12/24/48 Volts Batteries but pending on what voltage of battery you have the SCC will have limits as to the Watts, Amps & Volts it requires to deliver the max charging amps it is designed to do.

You are a HAM OP so you should know more than the average person out there as a byproduct of your hobby. I have lived from Coast to Coast and from our most southern tip to the Deep Arctic (Iqaluit Nunavut) and everywhere uses 120/240VAC 60 Hz. I lived in Calgary & Edmonchuck for a few years and am well aware of Alberta Power systems as well.

My "warning" is just a fact of life which applies to all cells regardless of AH Capacity, it is something that needs general understanding and acceptance that many vendors "play numbers games" and sometimes will use ongoing misinterpretations & misunderstands to further their malarchy. Many are quite good at mis-representation and doing it in a way that they get away with it.

Example.
EVE LFP 280K are rated for 6000 Cycles to 80% Gross Capacity, some companies advertise 280AH cells to 10,000 cycles BUT to a 70% Gross Capacity remaining and they hype that their cells are good for 10,000 cycles... The "Industry Standard" is to 80% Gross Capacity remaining...

Many companies sell cells rated at 280AH and state you will get 280AH from those cells. The GOTCHA in this, is they will take B Grade cells and test them from 2.500 to 3.650 using the Gross AH to show that they will deliver that BUT in reality THIS IS FALSE and horribly misleading! A "Proper Grade-A" Cell will come out between 288-295AH Gross AH but 277-282AH Net AH.

All Batteries of ALL Chemistries have TWO Voltage Values.
- The Allowable Voltage Range, which is the voltage range that they can operate WITHIN without hard/damage/degradation. For LFP this is from 2.500-3.650 Volts per cell. The AH from 2.50-3.65 is the GROSS AH Rating and this is NOT what the cells will deliver in reality.
- The Working Voltage Range is from 3.000-3.400 Volts per cell, this is where the NET AH comes from which the cells will actually deliver. This is the primary reason that ALL LFP Nominal Voltages are stated at 3.200Vpc (50% SOC).
** All LFP will "Settle" post charge input, so even if you charge your cells to 3.650 and allow Amps Taken to drop down to 2A and fully saturate the cells, stop charge input and without ANY load at all, the cells will settle and within an hour you'll be seeing 3.500Vpc on average and another hour later they'll be settled to around 3.450Vpc +/- a wee bit.

LFP has an Extremely FLAT voltage curve from 3.000-3.400. Above 3.400 to 3.650 is the sharp cliff climb on the voltage curve and at most 10AH but usually about 5AH Gross Capacity and the cells will deviate from each other at these voltages as you are exceeding the working range. Below 3.000 is a cliff fall on the voltage curve and again at most represents 10AH but usually about 5AH of Gross Capacity.

Properly Matched & Batched cells that are Grade-A will deliver that 280AH from the Working Voltage Range of 3.000-3.400.
There is of course a Price Difference that goes with that and 95% of folks don't wanna pay that not realizing the down the road implications.

FYI:
I am 100% Offgriid solar powered up near Algonquin Park Ontario. I run off 24V with 1295AH/33kWh battery bank consisting of
1x 24V/105AH (Matched/Batched Grade-A),
2x24V/174AH (Used EV Grade LFP - Thin Prismatic) &
3x24V/280AH (2 Packs of B-Grade & 1 Pack with Grade-A Matched/Batched)
All managed by JKBMS' with 2A Active Balancing.
You can see more details here: About My System (updated NOV.2022) which is a tad out of date as I am making changes now and again in spring when I add the next Midnite Solar Classic-150 SCC and it's array of Q-Cell 395's.

Be careful of Wives Tales generated from Misinformation & misunderstandings... There is piles of it out there and most especially on Youtube ! So many "fools" call a Single Battery a "Bank" because it has several "cells" in the battery, but these people also think that a car battery is One Cell when it is not. A Bank is obviously 2 or more battery packs (regardless of the number of cells within each pack) in parallel.

Hope it Helps, Good Luck.

 

[MIKEJAMES]

Thanks, quick reply whilst I absorb your notes. In all my 90 years I have never come across a technology with such a mass of " Old Wives Tales". It seems to be a "Scammers Dream " selling components to the DIY folk like me, although, as you say, my Ham experience has given an advantage in basic Electrial theory. Some distributors are great at selling but short on advice. My present SCC a Victron 100/50 does not operate at 48 v.
Great system you have.

 

[Steve_S]

The "Wives Tales" started several years ago and is pretty much all of the fault lies on terrible translation software, translating Simplified Chinese to English which generates the infamous "chinglish" that needs further decoding. We still see that in many docs & manuals currently in circulation with products.

Even the simplest of things when reading LFP Specification Doc's such as "Fixing cells" or Fixture Pressure... Translation = Binding Cells (not squeezing & compressing them). Bind them tightly so they do not shift in any way but do not squeeze them hard as that is bad for them. There are a couple of Way Over-babbled threads just on that one topic and when you read through a few hundred you can easily see how folks are "Interpreting" Manufacturer's docs and information. This is essentially to prevent any stress on the terminal posts and causing them to flex in any way, which is essential for any "Mobile Battery Pack" but also for a big standard ESS as the Prismatic cells do expand & contract a little due to Temperatures & State of Charge... It's funny that people cannot connect the "obvious" when faced with it... IE you CANNOT Charge LFP below 32F/0C Temp because the electrolyte begins to freeze & expand and by the time the cells can no longer discharge @-20C/-4F you are deep into frozen electrolyte. Now you know the answer to what happens when a fluid freezes...

I've edited a few BMS Manuals (JK being one of them) and a wide variety of reference docs for English Content and have also left a long trail of docs, manuals & guides related to this subject matter over the last few years which are in use in several sites now... always surprises me when people post my docs etc... At least it's all helping folks.

With regards to your Victron 100/50 that is a fine SCC and quite manageable thanks to Victron's software, and you can parallel other Victron SCC's to increase charge capacity.

As you saw, I use a Samlex EVO-4024 Low Frequency Inverter that can handle 12,000W Surge. This is also very popular with a lot of HAMS because it's good "clean" Pure Sine but they are not cheap. I had it run my 120V MIG Welder & 3HP Compressor without as much as a blink.

 

[toms]

New to this site, based in Wales UK.
With a 3 kW solar array, I wish to install 16 lithium batteries (8 in series x 2) and then in parallel giving 24 volts.
This was advice from reliable supplier who also supplied a JK BMS.
How do I connect the control cables to the batteries? The manual only advises batteries in series. Any advice appreciated.

If you have cells with significantly different states of charge it isn’t a good idea to put them in parallel and fully charge them (ie parallel top balance), other than that you need to realise that every prismatic LiFePO4 cell you have is made up of parallel sections.

If you buy two 100ah cells and connect them in parallel, it is chemically no different than buying one 200ah cell.

As for redundancy, if that is important to you ensure you have two completely independent systems.

For cell monitoring, in a parallel arrangement the monitoring cable can be anywhere on the parallel connection between the cells (typically on one of the cell connectors, but can be a separate connection)

With parallel cells, all cells in that group will perform as well as the worst cell. (in much the same way as an individual cell performs as well as its worst segment).

You are correct in saying there are a lot of wives tales surrounding LiFePO4. My first pack (over a decade old) is a parallel 2P16S setup. I see a pattern of people advising others on how great their system is - then disappearing when it fails a few years later.

The widespread useage of seperate packs coincided with the widespread use of ebike BMS units in residential off-grid power systems. These units (like the JK you have), aren’t reliable at high constant currents that can be seen in residential systems. This is avoided by using multiple packs and multiple BMS. If that applies to you, then that is your reason not to parallel cells. If you are using low ah cells, and low load currents compared to your BMS rating, parallel will be better - (less wiring, more compact, less components).

 

[MIKEJAMES]

The "Wives Tales" started several years ago and is pretty much all of the fault lies on terrible translation software, translating Simplified Chinese to English which generates the infamous "chinglish" that needs further decoding. We still see that in many docs & manuals currently in circulation with products.

Even the simplest of things when reading LFP Specification Doc's such as "Fixing cells" or Fixture Pressure... Translation = Binding Cells (not squeezing & compressing them). Bind them tightly so they do not shift in any way but do not squeeze them hard as that is bad for them. There are a couple of Way Over-babbled threads just on that one topic and when you read through a few hundred you can easily see how folks are "Interpreting" Manufacturer's docs and information. This is essentially to prevent any stress on the terminal posts and causing them to flex in any way, which is essential for any "Mobile Battery Pack" but also for a big standard ESS as the Prismatic cells do expand & contract a little due to Temperatures & State of Charge... It's funny that people cannot connect the "obvious" when faced with it... IE you CANNOT Charge LFP below 32F/0C Temp because the electrolyte begins to freeze & expand and by the time the cells can no longer discharge @-20C/-4F you are deep into frozen electrolyte. Now you know the answer to what happens when a fluid freezes...

I've edited a few BMS Manuals (JK being one of them) and a wide variety of reference docs for English Content and have also left a long trail of docs, manuals & guides related to this subject matter over the last few years which are in use in several sites now... always surprises me when people post my docs etc... At least it's all helping folks.

With regards to your Victron 100/50 that is a fine SCC and quite manageable thanks to Victron's software, and you can parallel other Victron SCC's to increase charge capacity.

As you saw, I use a Samlex EVO-4024 Low Frequency Inverter that can handle 12,000W Surge. This is also very popular with a lot of HAMS because it's good "clean" Pure Sine but they are not cheap. I had it run my 120V MIG Welder & 3HP Compressor without as much as a blink.

 

[MIKEJAMES]

Thanks for advice. Initially I will be installing a single string of 8 x LiFe PO4 280 AH cells in series connected to my Victron 100/50 SCC. using JK BMS. Hopefully I will gain some experience using these cells before attempting to connect another string in parallel. many thanks for assistance.

 

[MIKEJAMES]

If you have cells with significantly different states of charge it isn’t a good idea to put them in parallel and fully charge them (ie parallel top balance), other than that you need to realise that every prismatic LiFePO4 cell you have is made up of parallel sections.

If you buy two 100ah cells and connect them in parallel, it is chemically no different than buying one 200ah cell.

As for redundancy, if that is important to you ensure you have two completely independent systems.

For cell monitoring, in a parallel arrangement the monitoring cable can be anywhere on the parallel connection between the cells (typically on one of the cell connectors, but can be a separate connection)

With parallel cells, all cells in that group will perform as well as the worst cell. (in much the same way as an individual cell performs as well as its worst segment).

You are correct in saying there are a lot of wives tales surrounding LiFePO4. My first pack (over a decade old) is a parallel 2P16S setup. I see a pattern of people advising others on how great their system is - then disappearing when it fails a few years later.

The widespread useage of seperate packs coincided with the widespread use of ebike BMS units in residential off-grid power systems. These units (like the JK you have), aren’t reliable at high constant currents that can be seen in residential systems. This is avoided by using multiple packs and multiple BMS. If that applies to you, then that is your reason not to parallel cells. If you are using low ah cells, and low load currents compared to your BMS rating, parallel will be better - (less wiring, more compact, less components).

 

[MIKEJAMES]

Very many thanks for your guidance. Great response to my query.

 

[toms]

Thanks for advice. Initially I will be installing a single string of 8 x LiFe PO4 280 AH cells in series connected to my Victron 100/50 SCC. using JK BMS. Hopefully I will gain some experience using these cells before attempting to connect another string in parallel. many thanks for assistance.

If you already have one 8s string, it may be easier for you to replicate that rather than disassemble and add parallel cells.

Main reason for this is the fragility of the cell connections on the typical 280ah cells - the less you mess with them the better.

 

[MIKEJAMES]

If you already have one 8s string, it may be easier for you to replicate that rather than disassemble and add parallel cells.

Main reason for this is the fragility of the cell connections on the typical 280ah cells - the less you mess with them the better.

 

[MIKEJAMES]

Thanks for that Toms. Will remember it.

 

[scsn]

Hi
Sorry for adding this question, but it is about 2 BMS banks in parrallel . I have 2 3S BMS circuits each giving 11.1v @ 3.3A . I was thinking of having a diode at the outputs of each BMS and tie the outputs together . Having the diodes would prevent any voltage from feeding into the other BMS (and visa versa) , would this work?

Thanks
Scott