Parallel Battery Strings - worst solution ever?
HTML-код
- Опубликовано: 26 авг 2024
- As an addition to my last video, I want to dig a bit deeper into the topic of having battery strings/banks connected in parallel.
Lots of you guys left amazing feedback and shared your experience with that. Is there really a concern of paralleling battery banks?
Orion BMS - Strings, Parallel Cells, and Parallel Strings (17page PDF)
www.orionbms.c...
The Battery Box 2.0!? Or what shall we do?
• The Battery Box 2.0!? ...
Please subscribe and join me on my journey of the Off-Grid Garage!
For more information, please check out my website with links to all the batteries, materials, devices and products I have used in my videos: off-grid-garag...
New to AliExpress? Use my referral code below to register and earn US$19 store credit. What a great start! a.aliexpress.c...
The 'buy me a beer 🍻' link is here (on the website, scroll all the way down):
off-grid-garag... Thank you for your support!
**** These are the batteries you are looking for ****
🔋➡EVE LF280 LiFePO4 cells 3.2V/280Ah: off-grid-garag...
🌞➡Solar, Controllers, Inverters and accessories: off-grid-garag...
🔌➡Batteries, charger and accessories: off-grid-garag...
⚡➡Cables, inverters and electronics: off-grid-garag...
🔧➡Tools, analysers and testers: off-grid-garag...
Take a sneak peak at the Off-Grid-Garage in the Victron VRM World:
vrm.victronene...
My Tesla referral code (free 1500 SuperCharger km / 1000mi for you and me)
ts.la/stefan96963
(Thank you very much to everyone who have used it so far!)
I like your 2 separate battery approach for a couple of reasons. #1 With 2 BMSes you will double the balancing current. #2 If there is a problem with a cell and the BMS shuts the battery down, you will still have power from the 2nd pack, eliminating a Blackout situation.
Those reasons are exactly why I chose to use that method. Unfortunately, I had already bought 4S BMSs before deciding to go with a 24V system (instead of 12V). So for the time being, I've got two BMSs per 8-cell bank instead of one. It's just like having two 12V batteries in series (per bank), so it works fine. I'd most likely fry my 8-cell active balancers if I tried hooking them up to this set-up, but I haven't seen much need for it anyway.
Having that modularity and ability to add/remove capacity or perform series battery set maintenance as needed without taking your entire system down is key. Sounds like you have a great plan and your videos going deep in the weeds on most subjects have taught me a ton. Thank you.
Nice video, one minor critique on your comments around 5:30. You are correct in that there won't be much current going between the banks if the voltages are similar and the banks are just connected to each other. And you are correct in that the voltage of a bank with a higher SOC will want to rise more quickly than one with a lower SOC while charging. But current will still not flow from the higher SOC bank to the lower one while charging is still active. That isn't where the problem is. The bank with the lower SOC will receive most of the current from the charger simply because the bank with the higher SOC will have lower current acceptance (it takes a higher voltage to make it accept more current, but since the banks are tied together it can't get that higher voltage).
The problem occurs when the charger turns off or the change rate changes. That is when you will start getting currents going from one bank to the other.
There is an easy way to describe this (it also describes the micro current cycling you talk about later). Lets say you have two banks with mismatched SOCs. What will happen when you charge the banks up is that the charger will eventually manage to push the voltage up, but then when the charger turns off or the charge rate changes significantly, the settling rate for the two banks will be different and will cause current to flow from one bank to the other as they settle down.
Similarly, if discharging the ganged banks with a load, and your load then turns off (or changes significantly), the cells in both banks will 'recover' some voltage... but if the banks are mismatched, they will recover at different rates and again will push some current from bank to another. And this happens EVEN IF the banks are fairly well matched.
Thus you wind up with micro-cycling between the banks.
--
Also, if you didn't see it... with the number of strings you have, blocking diodes are now MANDATORY. They are not optional any more. You absolutely need them for safety and shouldn't depend on your circuit breakers and fuses for that function. Long-term reverse voltage stress on the panels is just not good for them. Even though a solar cell is basically a diode, it is one designed to generate electricity and not designed to have good reverse-breakdown-voltage characteristics.
Also the voltages coming off of the panels are fairly high and an inductive spike can spike the voltage high enough to hit the break-down voltage of the panel cells, causing a fire in a short-circuit situation. While unlikely, it is even possible for the inductance to cause the voltage spike high enough to arc the breaker and/or fuse, turn into plasma, and not break the connection. Hence why blocking diodes are now mandatory.
-Matt
Both approaches are valid and have positive and negative aspects. Do what you want to do. I went with parallel banks, and each individual bank can handle the full load. No regrets, and easy to increase run time by adding another battery. But don't worry, either one will work well, but since BMSs have a tendency to fail, more redundancy was my choice.
You have made your decision, great. Just as a side note Pylontech use the parallel string method in there battery systems. Each of the us3000 battery frames I am using is operating at 15S and parallel connections are made at the front of each frame. Each frame has its own BMS, shunt and internal cut off. The only difference is each frame's BMS is connected via an RS485 data cable so BMS data is shared to each frame as well as the MPP inverter charger I am using, so I assume each us3000 frame knows what the parallel frames are doing. I have been toying going your direction but as I have already invested in the Pylontech solution its far easier for me to add a few more Pylontech frames to increase the storage capacity.
Hey I'd love to see pictures of this.
Brilliant... Very cogent arguments for either potential setup. It just goes to show, that there is no actual Right or Wrong answer, it's all dependant on the individuals needs and priorities.
I have 16S2P and two sets of 16S1P. Both configurations have pros and cons. Just to note some advantages of the 2p system:
- easier to match individual cell capacities (combine the strongest+the weakest, then 2nd strongest and 2nd weakest etc...), this way you can evenly distribute differences in capacities.
- You just need one BMS
- cons: you need more busbars
I find you sharing very helpful and thought provoking. I have just received 32 310Ah cells from China and am waiting for the lithium battery tester to arrive to check them. After seeing this video it has made me change my mind back and forth between two 1P16S and one 2P16S layouts. As both seem to work and you have opted for the former I will go for the latter and report back if I come across any issues. It will save me having to buy and therefore wait for a second BMS. Keep up the excellent work. We are all learning lots care of your efforts.
I do not know your application, but I suggest two parallel strings with their own bms, current monitoring and their own inverter. The inverters need to be designed to synch with each other. This is for life safety systems (medical, sailing, etc.) Since each bms, current shunt and inverter only handles half the power, the expense of the electronics can be lower than 2x. Obviously if one string faults, some loads may have to be shed, but the critical ones can be supported.
My application is a bit unusual as I am using a small trailer to transport surplus energy from an off grid system back to my home. Not a critical application so no problem with a 2P16S configuration. My main problem at the moment is that the cells are well below specification and I am hoping to get a resolution through Alibaba. I am very glad I bought a EBC-A40L tester. It has enabled me to prove that the supplier has sent cells which are not up to spec.
@@nigelcharles511 who was your supplier?
Andy, I’ve had 2 1p16s 280ah packs connected in parallel for several months and they stay perfectly balance. They both charge and discharge at the same rate. I’ve turn off one pack for several hours to manual unbalance them and by the next day they are again perfectly balanced . I’m using Daly smart bms and all the cells are balance to within a few millvolts
Each pack has its own Daly BMS, correct? Did you have to do anything to the BMS to make it work in parallel or was it ok from the factory? I am planning on building one just like what you mentioned and appreciate the info.
@@cwildmann I did reduce the high voltage cutoffs settings to 3.65 . However, an update, I noticed one battery was charging and discharging faster. Removed all the busbars and cleaned and now working well again. Now looking for the best anti oxidation paste and will redo again.
Following this with great interest. I have recently built my 8s 24 v battery and am also now thinking of doubling in size. Learning a great deal and love the format
What can you run on that?
@@mcorrigan6729 it has a total of 4,800 watt hours capacity, of which I can safely use around 80% (without stressing the batteries). I run most of my loads through the 3kw inverter, but also have a few 12v light strings. I also have a back up petrol generator, and some independent solar lights. Its part of my off grid garden / cabin system, so just run lights, fans, water pump, charging computer/phones etc, a rice cooker, and some electrical tools during the day (angle grinder, circular saw, etc).
@@jonathanhampshire8405 could you possibly send me potos of your setup? or a list of which components you are using. I swear I have spent probably 20 hours reading about different aspects but have not been able to grasp it for some reason. I ordered 8 3.2v 280ah lifepo4 batterys, like these because thats all I can really afford, but would like to do a 24v battery that can power a small cabin setup that will be almost identical from the sound of it. I appritiate any advice you may have. With all of the information out there it has gotten slightly.. confusing. It sounds like you have this setup down pretty well. Also, how many watts do you have in terms of the solar?
@@mcorrigan6729 I am no expert and have learnt by doing. probably spent more like 200 hours trying to work out my current system! I also have a friend who helps with all the wiring, fuses, etc. My basic set up includes: DIY 24 v LifePO4 battery using 8 3.2v 200ah cells plus 150amp Daly BMS (4,800 watt hours), 3,000 watt 24 v inverter, Epever Tracer 40amp charge controller, a 24 to 12v converter to the load output of the charge controller, a Victron smart shunt to help monitor the battery State of Charge and history, and 3 x 390w 24 v solar panels (1,170 watts) connected in parallel. As noted, a friend has helped with all the wiring, fuses, etc as I really don't know much about that. Will Prowse's RUclips channel is a good learning resource if you haven't yet seen it. Good luck!
Hi Andy, your channel gave me the courage to purchase cells (only 4) for RV installation. I got them,finally, a couple of weeks ago, & now have an operational system. Thanks
Love your channel Andy!!! I will be starting a channel soon as well! I have 3 strings of 16- 200ah LiFePo4's and 3 JK BMS I am putting together for my 16x7 tiny house on wheels I have a MPP LV 6548 im hooking up to the 27KW bank, its all replacing an AIMS 24v 14KW system I currently have run for nearly 3 years un interrupted! 2 deep freezers and a full electric kitchen with a bunch of big appliances!, Air conditioning, electric heat, 50 inch TV and many fans! all running off 8-260 watt solar panels! we have been all over the south and western US 2 years now without a hiccup!! Lovin it!!
Keep up the great work Andy!!!
Awesome exactly my thoughts on the subject and what I am going to try
I connected 26 14s10p (18650 cells) packs in parallel. No problems with that. Each pack has its own BMS and active balancer.
Great, thanks for sharing.
hey zertyhc is there anyway we could email about your system? I just picked up qty 12 packs with a 18650 cells in a 40p13s configuration. I'm looking to take the 12th pack and change the other 11 packs to be 40p14s which should help my inverter battery voltage range. So I need 11 bms's but struggling to deicde which brand to go with. I'm struggling to find anyone doing a similiar system to mine.
The one advantage in parallel is if you have some cells that always tend to run a bit high (happy) and another that’s a bit low(sad) you may be able to pair them up so you get a better average. That way the BMS’s balancer doesn’t have to work as hard and you get a better finishing voltage.
That's my theory as well. I've heard others say this isn't a good idea, but I wasn't impressed with the reasoning.
Thank you SO much Andy. You keep answering technical questions i have. Awesome effort dude! Dave from Zimbabwe!
Thanks a lot David. In fact, these were all questions I had myself and I could not find answers to, so I made these experiments myself.
That is super interesting that you’re going this route.
I will be watching for your results :)
A further advantage to smaller paralleled strings vs. one huge paralleled cell bank: Physically moving them. Say I have a bunch of whatever cells, I can move them from one application to another if they're paralleled each with own BMS. JBD bms with UART is a good choice for this. I have a front and rear paralleled 10s packs from Leaf modules, in a 1970's GE Electrak garden tractor. A lesson I've proved from the tractor: keep the main +- cables identical lengths. I did not with with layout of the tractor and I do see eddy currents, a couple of amps? for some 15 min? after charging and discharging, since one bank has less cable resistance (I might redo it to improve). I plan to do 4 paralleled 14s (7 modules) for a house battery bank, and these will also be able to be moved to an RV as needed. Downsides are extra complexity of multiple bms's and disconnects (all smaller and cheaper though), but the flexibility and redundancy I think could be worth it.
Thank you for sharing, Dale, much appreciated. Yes, you're right, I didn't think about moving things around which is much easier with strings in parallel than having just one big monolithic block of cells. For example when upgrading to a bigger enclosure, I can do this string by string without loosing power.
I agree on your logic of paralleling strings. It's pretty much what people have been doing with car batteries !! They can't access each cell, and so they parallel 2 strings ! Great videos .
This topic is very interesting, i have the same question in mind. How to parallel.
I'm in proces of building a small house for vacation, and i will start with a 48V 100Ah batery pack, but in time i will grow up the capacity and numbers op solar panels.
Waiting for next results.
You're doing great jov, thanks for this channel.
Sorry for bad English I'm from România
You can parallel by connecting cells in parallel first (positive to positive and negative to negative) and then connecting parallel packs in series into a single battery to give you the desired voltage. Or you can forgo the initial paralleling and just build individual batteries (cells connected in series), and then combine the batteries in parallel using a busbar, as the content creator in this video is doing.
@@firpofutbol i have 16 cells 100Ah in series at this time, I will see when upgrade time come the solution.
Thanks for reply, regards
061021/1838h PST 🇺🇸 061121/1138 Brisbane. Thank you immensely for discussing such an intelligent subject on paralleling LiFePo4 batteries. There has been no such discussion on YT, or may be some peripherally, skimming on this, very important issue. But no one knew what was being discussed.
After downloading the ORION BMS, going through it and listening to your lecture; it became crystal clear as to how to Parallel LiFePo4 batteries, correctly and safely. Imagine 33 batteries being paralleled with its own dedicated BMS! Flabbergasting, indeed, but logical.
Thank you for your lecture on this subject, you have opened up and cleared my doubts.
Stay safe and 73s…
Boat builders use parallel strings (see Mastervolt, Victron). Best to split the DC bus into load and charge. Overloading switches off the charge bus, but your load will continue. Undervoltage will switch off the load DC bus, but keeps the charging up/connected.
If you use parallel strings, split DC bus probably makes not much sense (as one of more other strings are always connected, when only one string disconnects).
Yes, that makes sense. You also need a bit of advanced control systems then to have your load/charger connected and disconnected under certain conditions.
I'm running 6 byd battery strings in paralell each has its own bms no problems . really no other way to do it when you connect premade batterys.
Thank you.
I am looking at 8 (500ah 3.2v) cells making a 1000ah 12v bank but could not find a bms to fit a single bank, so I was going to make two banks of 500ah with 2 bms's and connect them to a common busbar.
I was not sure if I should fit 2 shunts before making the connection to the busbar.
The reason for the set up is infrequent charging and drawing higher loads over short periods (230a for 3 minutes) as well as a small load continuous.
The only thing I was definately advised to do was make all the leads before the common connection to the busbar identical lengths.
I look forward to your next video, theory is fine, but practical is better.!
Hey Andy, please number your videos, so it is easier to find the previous video you often mention. Thanks for the great show, and manny amps to you!
👍 thank you!
i also stick on separate banks, because a bad bms can not Kick me out of buisness.
best wishes from Germany.
Another consideration: many of these "discussions" revolve around cylindrical cells... Like 18650s with a limited capacity (in comparison to these big 280Ah cells) and thus limited charge/discharge rates.
If you have 3p 280Ah cells, where will you find a BMS capable of 900+ amps charge/discharge? How many solar charge controllers can pump 200+ amps into your battery (0.25c)?
Once you move away from small cylindrical cells, it just makes sense to build 1p banks with discrete BMS boards per bank. I don't know that you could even find a BMS capable of matching a 2p, 3p, 4p.... pack with such huge capacities.
Of course, if you're building a pack from 3400mAh cells, can you imagine the number of BMS boards you'd need to power a house if you used 1p banks? Phew!
Economics is the real deciding factor in both scenarios. Either method works.
Thank you for using my comment in this video. I don't remember what BMS manufacturer said it, but they said I could not parallel BMS units. Maybe another topic.
I'm running a 3 x 1P18S system, with one Batrium BMS (blockmon). It sure makes for some nice graphs but I concluded this week that I will need active balancers on each of the strings to help out on the higher and lower end of the charge cycle. It Has more to do with the used cells the alibaba rip off has send me then the setup itself.
Im looking forward on your experience with the cheap balancers while charging, mine are still in the mail :)
They have sent you used cells???
I just got another 8s and 16s balancer delivered yesterday from Hankzor (for more testing). They send out that stuff very quick.
@@OffGridGarageAustralia sadly yes, they keep denying they are used but the evidence is undeniable, glue residue, scratches and dents. I ordered 158 cells together with some other poeple. Alibaba denied the initial refund, got a 3000$ refund in the end because I wrote a bad review. I guess the price should have been a clue, we payed around 10k for 158 cells excluding taxes.
Hi, I'm now reading your comment. Do you use Batrium WatchMon CORE with 54 blockmons connected in series? What capacity cells are you using? Batrium blockmon balancing isn't enough? Do you use shuntmon too?
I'm asking so much because I'm deciding what to build:
1. 16S2P battery with WatchMon CORE +16 blockmons +1 ShuntMon or
2. two 16S1P in parallel with WatchMon CORE +32 blockmons +1 ShuntMon
Thanks Andy, very well explained - looking forward for your upcoming battery V2.0. Cheers
I had a 4s4p and had a problem and 8 cells died together and 8 cells lived to be used another day. Now I have a 5s4p pack, higher voltage, higher energy, same peak current (1300A) and it should survive as it won't be overcharged at the voltage the first try was experiencing (18.2V). I had a full-time capacitor balance board, and no bms, no fuses, hard to fuse for over 1000A anyway :-D Aside from the over charging, the balance board got salty and wet and failed or possibly got cooked by the overly high voltage, and that's what lead to two banks staying at an acceptable voltage and 2 banks being absolutely fried/boiled to death :-D So from that extreme torture test, I would conclude that the paralleled cells do indeed walk in step and fight in unison and die side by side, or live together through strength in numbers. Having said all of that, My other comment stands, I'll seriously consider having 4 independent banks, however I can't think of a valid use case beyond a 2-way split, so there's that.
Oh, wow, what an experience! How did they die, just didn't work any more or did you have a RUD?
@@OffGridGarageAustralia What's a RUD? The cells seemed to have leaked electrolyte out and upon inspection later after finding it wouldn't start from the pack any more I found that all 8 dead ones were open circuit, accepted no current, no matter the input voltage. I suspect they're dry inside. Could they be rehydrated? Not sure, but don't really care, either. It was kinda hard to tell, though, as the salt water basically shorted out the cells at one end anyway and the resulting corrosion made them quite ugly. However upon disassembling and testing some of the 8 survivors were some of the ugliest, but I got them cleaned up and they're in service doing just fine right now :-)
Perhaps the best "Andy" solution is to just have a shunt on each battery in addition to the system-wide shunt you have out in front of the busbars. You could even wire the shunts into a battery monitor display with the polarities reversed so it acts more like a "tachometer" for the battery banks, telling you which direction the flow of energy is biased. Even better would be having the pi or an arduino independently monitoring the shunts and recording data and packaging it up for VRM to consume.
Thank you so much for your videos. I love the way that you present the information, and the topic is something that I'm very much into. Cheers!
Same as parallel drop in style lithium batteries (e.g. battleborn). Each has built in BMS.
Thanks very much for your efforts to explain so much about the different types of arrangements for batteries. I have just bought sixteen batteries and two BMS's to build two 24v banks and parallel them. I am replacing an old set of GM batteries in my boat with LifePO4 batteries I bought on eBay. I am not really very advanced in my understanding of these systems but luckily I have found your channel and have now discovered how to build up my battery. Your vid on programming the Victron controller was also useful as I have a Victron controller on my boat's 1500 watt solar system. I have a very cheap inverter on my system , 24v to 240v 1500 w and am wondering if I would make any significant gains in efficiency by replacing it with a better quality inverter. The original controller was also a cheapy and it frizzed out after about six months which is when I replaced it with the Victron which I love because of the bluetooth connection , the original controller had no bluetooth . So thanks again I will watch a lot more of your vids as they are a great source of knowledge for me.
I'm getting ready to build a 2P16S using 280Ah LifePo4 cells. I will use a 100Amp BMS (all from battery hookup). This will replace a DIY 18Kwh pack (48v) that I made from 18650 cells a couple years back. I figure that if I was able to run 120 cels (LiPo 18650 = 120P14S) in parallel for 2.5 years without ANY issues at all, then a 2P configuration using LifePo4 should have no issues at all.
What could go wrong, I like it
This is also how the software industry evolved from single large systems to multiple instances/containers running in parallel.
I think this parallel string idea is for sure the right way to go in the future as it makes things much more modularized (easier to service and adding or reducing cells) and as long as the BMS is cheaper then it will be perfect.
For now I think the BMS world hasn't evolved in to this modularized mind set yet, because what they could make is there is a simple/cheaper submodule (voltage current sensor + relay) for each string then have a central management unit that can wire multiple sub modules.
Maybe the demand of personal home/RV battery system never has the need to frequently changing capacity or servicing cells so perhaps there is no market pressure on to this "distributed" architecture so it might never hit the market 😁
Consider creating a dual of your entire system down to the inverter. Hopefully your inverter can take a synch input so that the inverters can be paralleled to the load. That way you never have to worry about taking one down for maintenance - you just might have to shed some loads. In fact you might want to routinely take them off line in low demand periods to balance the individual strings.
Not sure if the inverter likes to have another one in the network. Has anyone tried with an off grid system?
@@OffGridGarageAustralia Query in with Victron to see if two inverters in parallel can synchronixe. Not technology challenge, but a simple design challenge. They have the tech to synchronize 3 phase, so it is defintely possible. Thanks for the reply.
@@donr416 I'll see what I can find. I think they are been connected to each others to sync but I will definitely do some research on that. Thanks.
Andy, This is really great information and well presented. I live on a boat - 1850 w array 8X 6 volt 330 AH wet cells 48 volt . When they die I am going to LIP and you are getting me up to speed. Thanks
Andy, I think you have made the right decision to build two separate banks with their own separate BMS and fusing protection before paralleling them. If you parallel the cells first, then you are creating a battery with a humongous amount of short circuit current potential which makes it very difficult to provide adequate fault interrupting capability. The individual 280ah cells are already capable of producing very high amounts of short circuit current. Paralleling the individual cells makes it even worse. If you parallel the individual cells you had better use at least Class T fuses with high interrupting capability to protect the bank from a catastrophic fault.
In my opinion, fault current interrupting ratings of protective devices is a subject that seems to be overlooked or ignored in most DIY battery builds that I've seen on the internet and it is a serious safety consideration that needs to be addressed. All too often, I've seen cheap fuses or circuit breakers used to protect these lithium battery banks. There can be serious consequences with these batteries if the catastrophic protection devices are not capable of breaking the maximum fault current available when called upon.
Good stuff, I think that will work well for you. Although my 24 cells are 6P4S from new and I've never had an issue if i got a new batch I would build a second battery. I've even thought of splitting my pack into smaller packs and buying more BMS's
All parallel banks could be at different voltages but as long as each series string has its own bms the bms’s will limit the current to safe levels in both charging and discharging between each bank.
I love that you've done the research and have decided to begin doing your own experiment. I'm looking forward to seeing what the results are. I'm betting that it will work perfectly!!!
I think it is very unfortunate that all of the engineering advice of the problems, complexities, and dangers of paralleling strings provided by the ORION BMS writeup, was "thrown out the window" because of some anecdotal evidence of success with parallel strings. A single or even multiple examples of success with parallel strings simply does not negate sound engineering warning of that practice.
Andy, I already have a double battery setup as you describe and I have split the charge controllers between the batteries, so there is little current flow between the batteries or you are using up their cycle life just moving power back and forth between them. Keep up the good work. Love your channel.
Orion's claims are misleading. If parallel anything you always will have balancing between them. Between 2 cells (in 1 pack) or between 2 packs. But with 2 packs you monitor all cells and with 1 pack you monitor only the two cells average. Also you can buy cheap Renogy shunt with display to monitor each pack's power consumption (no need to buy an expensive Victron). I always tell people to test the 20-80% and 30-70% balance of 2 cells and measure the current flowing. Because of the flat discharge curve only minimal. So if you put 2 cells parallel it will take a long time for them to balance out. Luckily you use lower charging V, so you will not float one pack full when the other is still charging. This helps balance out the 2 packs easily. Only one thing you have to pay attention and that is the current. If 2 packs online you have 2C charging and discharging rate. If one stops (like BMS stops it) then you have to get it lower to 1C max. Or you use only 1C and there is no problem :)
Agreed. They read like someone talking to DIY'ers telling them they need a degree to work with lithium, "and here's why". Many of their premises are non-issues or at least less than significant impact. You don't NEED all the little bells and whistles that would be nice to have, but unnecessary for a functional system. And those can be added a bit at a time. I've been in power at AT&T for 20 years (recently retired) and I've worked with every kind of battery chemistry, including molten sodium batteries (sodium-nickle). Kinda like telling a farmer he needs a certified mechanic to install a start battery on the tractor and that "duct tape and bailing wire" will get you killed (that's a John Deere joke). Maybe. But isn't that our choice?
1C in a 280Ah 51,2V battery means over 14kW of power!
Andy uses only a 2,4kW power inverter
@@zarbonida For now ... but later he wants to connect the house to the off-grid system.
No, they are not misleading.
Parallel strings balance only the strings, not the cells inside.
With 2 packs you need 2 string BMS and one master BMS - total of 3 BMS.
And you will need 3 shunts also.
To put 2 cells paralel you need to cycle test their capacity and sort them in bins.
Now you build packs only from a single bin and connect them at equal SOC (State of Charge).
And Orion talks especially about that - cascaded failure occur because you lack all of this costly things, and someone does not pay attention.
In automated system there must be no need for someone to pay attention at it, so it doesn't fail.
@@TexasVeteranPatriot Cascaded failure is pretty significant impact.
And if you wish, you can go without all they recommend, but instead capacity testing and binning of the cells is needed.
On top of that life cycle testing is usually done too - cell is cycled until certain cycles.
Also, talking about AT&T is completely out of context here.
They do not buy random undocumented chinesium on Ali.
If they don't do all capacity and life cycle - the OEM does it for them.
How would two banks of different capacities behave? I have 4 202ah and 4 85ah lifepo4 cells on their way, each with their own bms (intended for different purposes). But could they theoretically be banked together? Are there any tests you can show with your 100ah cell with the 280ah to show why it would or would not work?
Following
The higher capacity bank will be the keeping the voltage up during discharge and down during charge. This will cause most of the current flowing from/to the higher capacity one while voltage will be equal over paralleled banks.
@@MarkusIngalsuo so the load could use the power from the lower capacity battery as needed where available, and the smaller battery would 'match' the voltage state of the larger battery? Any danger of harm to either the smaller or larger banks?
@@sl0rion, yes, both batteries will have equal voltage but the bigger capacity one will take most of the load because it's heavier to push/pull.
@@sl0rion, no damage will occur if the currents do not exceed the ratings. This can be kept in shape with fuses of suitable size.
Andy, since you go trough that path, is it possible to measure the energy swap between the packs along the day? the idea is to have a graph/map of internal exchanges and calculate the losses (if possible). Perhaps it will give you one or two fantastic videos... Thank you again for all the remote experience and explanations you provide.
Andy,
I don't want to waste your time, but I have a technical question, like everyone.
In this very useful video you connect the 2 battery banks in parallel.
How many Amp's you can use in total?
Is it like BMS 1 + BMS 2 = total please?
Regards.
Willy
Lithium batteries have a bms built in--.is it a problem paralleling two or three batteries with a bms in each.
Glad you found the right solution for your system. If you are curious about the stray charging currents between the individual banks, just have two shunts, one per bank and that way it's easy to monitor. This will satisfying your thirst for knowledge And Beer lol.
I was just downloaded and read that Orion BMS - Strings, Parallel Cells, and Parallel Strings, particularly pg 15 "Parallel strings with 1 charger per string and single discharge DC bus."
As I'm sitting here I got to thinking what if you had 2x discrete off grid systems instead since if you had say 800ah in 2x strings you probably need 2x 4000w inverters paralleled?
Why not have 2x discrete systems each with their own solar panels, charger, inverters, BMS, and inverter?
The cost will undoubtedly be more but only one extra charger, more wiring and higher installation time/cost is needed. Thing is I don't know much about off grid but I will because I'm building an off grid house soon on my 52ac bush block that has no power.
What I don't know is if 2x inverters can combine synchronized AC output but since many 48v inverters are designed to be paralleled maybe they can - maybe they can't, I dunno?
Even if they can't, then you could wire the house with two independent systems. One for the lights, one for the power points (or however you want to divide it).
Generator autostart may need a diode or possible just 2x relays to isolate the start inputs but that seems comparatively simple to do.
For such a project I would use two Victron Multiplus II inverters which also have AC input for a generator if you need it. They can be synchronised and work together to power the entire house.
Sure, you can go with separate batteries but as you said, additional costs and cabling effort. If you parallel the battery banks you can always take one offline and run the house still from the other one (or two).
I’m looking forward to seeing the results of parallel batteries. I haven’t seen this topic covered by other RUclipsrs. I plan to do the same using (8) 280 ah EVE cells and (2) BMSs to create (2) 280 ah 12v batteries down the road for an RV build.
My plan is exactly like yours, but with (8) 100 ah cells for (2) 100 ah 12v batteries.
👍 great Andy, I think you are right, parallelling cells will never reveal bad cells. If you look at your fellow aussie HBPowerwall he is using al lot of parallel cells and now and then needs to dig into issues with them (thermal camera, fuses etc). So a parallel crate with BMS will cover that, and your power will not easilly switch off while the cells are massive, small cells will be overloaded much easier, the 280Ah is huge. If you use the right fuse and size of wire it will be a good solution and never forget a maintanable one. So good thinking Andy and thanks again for sharing. I was wrong with the huge Amperes which flow when parallel the cells as also the BMS will monitor the progress and will kill the connection when something goes wrong. Also paralleling cells will give you 3.2V with ~600A in 1C, that's huge when you make mistakes, that will easier burn than a closed box in series. Have fun building the new one!
I have a 14s144p, and you don't need one BMS per 1p. You figure out how much you need based on how much difference you have between the 1s and the maximum charge rate. If you don't have enough and see the cells are not balancing fast enough, then you just throw on one more BMS parallel to the existing BMSs.
I agree with your decision to run multiple BMS's... I am running 5 BMS's. Also, I would suggest stripping the insulation from the BMS wires after moving them on top of the bus bars. LOL
VERY new to parallel batteries. My system 2S 4P (12V lead acid x 2 = 24V x 4P)
Does it matter if I create 'branch' in my parallel string?
I have tied connecting 2 24V pairs to one of the 24V pairs for ergonomic reasons.
24 P 24 P 24 P 24
P
24
I have a 500w 12v wind turbine... marsrock mppt hybrid controller 1500w wind and 1500w solar. This mppt controller came with a battery balancer but it seems to only be used for 24v application (series wiring). Can having a 12v battery wired in parallel effect the mppt controller?
This was simply an excellent analytical discussion, I applaud your thinking. Also kudos to your many followers on your channel, it seems you have a very sophisticated, intelligent following. Need to send you a few beers, I often join you when studying you videos.
Nice Story to Come, thanks for letting us join the adventure.
Good job as always 🙏🏾👏🏾🇬🇧🇯🇲
062021/0422h PST 🇺🇸 062021/2122h Brisbane 🇦🇺 Addendum to my below comments. I went through the OrionBMS supplement on paralleling LiFePo4 batteries. The more I read, the more I get confused. Then again your comments, contradicting the Orion statements. Now, I’m like, should I stick to “Classic” paralleling of 8 Prismatic cells? Or should I wait for your test results in few weeks? I’d rather wait for it.
My simple question is why the publicity from Orion about the “dangers” of Paralleling the cells? Orion also emphasize the use of a Qualified EE to perform paralleling the cells. I incorporate every safety measures, redundant safety measures even, like CBs, ANL fuse blocks and switches (on single 4S my system) Paralleling the same configuration with similar safety measures, will cause safety issues? I’m flabbergasted. I ask, (Orion) why such fuss? The Battery assembler from Nevada 🇺🇸 has JBD BMS built-into their batteries and has no objection to paralleling them. A family in Idaho 🇺🇸 has installed 8 of those batteries in 24V configuration
in parallel and running with no issues for over a year.
I think, your approach to seek the truth about paralleling is much appreciated and is looking forward to the results. Please let’s know. 73s…
i have similar build requirements for a go cart. as proper weight distribution is needed i seek a setup that will allow the use of multiple individually bms protected 16s3p packs in parallel going to a fused buss bar and then controller. i fear that when charging or discharging if one pack disconnects due to a voltage fault then the rest of the packs will operate and become out of balance from the pack thats disconnected. my fear is that the bms may at some point will clear the flault when im unaware of it, reconnect parallel and experience extreme balancing currents. i feel as though i need a bms that can communicate with the others to prevent this? please what is your opinion?
Thanks
Thanks for your videos. For people who are reluctant to mount batteries in parallel (obviously LifePo4 do not have the problems of lead-acid ones) say that Batrium has a system of connected BMSs that can control different batteries at the same time (I think up to 7) and connect them in parallel. It is quite an expensive system, but it is one of the most advanced that I know of; although the technical advice is with subscription (something that I dislike and I decided not to buy them). If I decided to increase the capacity of my battery (it is exactly the same as yours, but with a 200A JBD BMS) I would also choose the option of making a whole battery and connecting them in parallel. Although the rest of the components are not of the quality of yours (I did not put Vitrom) Sorry if it is not understood correctly, my English is very limited and this is a machine translation.
Hi, what is the best way to connect 5 or 6 gel batteries of 100AH for a 12v system? They do not give me warranties if I connect them in paralel... Thanks!
All same brand and capacity? Should be no problem. How do they know you had them in parallel in case of a warranty?
@@OffGridGarageAustralia yahh, how do they know... but yes, 4 X bat with 100ah each, same dá-te, same brand... can I connect them directly with Cooper bars, positivo in one side and negative in opposite side of the bars? Many thanks for your reply.
Here is another comment. 😄 You don't need a 2 pole breaker per battery. You can just put it on the positive side. When you said "it counters the BMS when one bank charges to the other bank" is not correct. You are actually assisting the BMS and that is how some of the more efficient BMS work instead of bleeding current out as heat through a resistor, they use the power to charge up other parts of the battery. The BMS is concerned about equalizing the voltages across the different cells so you don't overcharge any individual parallel bank which causes fires. Also the current from one bank to the other bank is OK as long as your bus bar can handle all the currents combined so where you put the charger will be part of that equation. Currents going from left to right will cancel currents going right to left, but currents going left to right plus having a charger going left to right will add the two currents in a small section of your bus bar and will heat up if your bar is not sized properly to handle the current. Even if the charger was on the left side, that would first start charging the left bank then the right bank. Think of current as water, and the physical height as the voltage and when you open the gates, wherever the water flows is how the current would have acted. I'd be interested in seeing your prior video, but I don't know which one it is. I do parallelize my banks. The only concern is if you are bottom balancing and so how much difference the cells are would become a concern, but you just have to make sure your BMS can handle the difference. If you top balance, and make sure the low cutoff voltage is set higher than the lowest capacity cell, you're good. But most cells bought around the same time will mostly degrade at the same rate and after years, they will most likely still be very close to each other. Mine still are after years. and LiPoFe batteries have even more cycles before they degrade than other Lithium chemistries. My memory says somewhere around 1000-4000 cycles for LiPoFe
I have run the 280ah 16s 2 strings in parallel using the daly bms. I would be curious if the JK bms can also work in parallel mode
You can use any BMS for that, even different once for each bank.
LFP cells need to be paralleled, the more the better. Balancing will never be required. Especially when cells getting older, used and uneven. As LFP cells are always in a interest of reaching a resting state, you are actually helping them to achieve their goal by creating bigger capacity on cell level. Parallel strings are absolute waste of money, and just doing the opposite of what the cell wants. My own pack with degraded, heavily used cells would never work with parallel strings, but with all cells in parallel, I don't even need to touch anything with a balancer to keep them happy. The 14s1p example is a LiIon pack, a totally different story...Du hast das Glueck 32 wunderbare, neue Zellen zu haben. Die wollen als Paar zusammen sein :)
Hola , saludos desde Panamá, siempre me pregunte porque UD no armó un sola batería grande e vez de 3 pequeñas ya que UD tiene un Tesla y este solo tiene una gran batería hecha de muchas baterías en paralelo-serie y no pasa nada .
Sir can you explain me the searis connected lithium phosphate battery is imbalance cells how can we balance it ?
I have a question:
I have two battery packs. Each pack is 48volts 16s 272ah. I want to connect both banks in parallel using two 200 amp Daly BMS with parallel modules on both.
Yes, will work perfectly.
Congrats with the 10000 subscribers. As an afterthought on your video about parallel batteries I thought of the following. On my boat I have a LiFePo bat. Parallel to a LA battery. To avoid current flowing between the batteries I installs two blocking diodes between the batteries. One at each batterie. At least this avoids current flowing from ones batterie to the other.
Best regards,
Tom
Thanks Tom, Yes that is a very good example if you parallel different chemistry batteries. They would just charge and discharge themselves all the time otherwise.
Would you mind sharing your LFP & LA parallel specs? Nominal V and capacity of each? And what rated blocking diode you're using? Thx!
@@tylerb9808 seen your request a little late sorry for that. I am using 50A blocking diodes between both batteries and the load. These blocking diodes can bought on AliExpress. La battery is 120Ah and the LiFePo is 100 Ah. Voltage levels are typical for both resulting in the LiFePo having the highest voltage usually. Let me know if you want more info.
Your decision of separate packs, each with their own BMS is best when you don't have matched cells. Especially when you are paralleling older cells you have used for a while. Periodically check the current distribution between the two strings, either with each BMS if you trust their accuracy, or with a clamp on ampmeter. If you get less the 20% difference in current balance you can consider it good.
Parallel individual cells should be closely matched. This typically needs to be done by manf. that can track battery origin batch and verify their balance.
Recall all the issues people have with achieving low terminal contact resistance with the aluminum terminals. Put your 4 point resistance tester to use to ensure you don't have a higher resistance contact which will really screw up the balance causing one string to supply a lower percentage of total inverter current.
Thank you.
Hi, can you still charge the battery to full cap. & how long would it take?
if the inrush current is too high , the BMS will cut or the breaker will trip
In before the hate messages, please try it and get a good sense of the paralleled current transfer. I have found that my 1p6s batteries paralleled to 9 strings passes significant amounts of current at times. As you said, it is the top and bottom end that matters. Every battery passing current to the one low string can hit 6-7 amps on a battery that is otherwise passing near no current as it tries to keep the voltages up or opposite when it is trying to keep the voltages down.
To do it again I would group the batteries by resistance and match high resistance batteries one per parallel string to do 9p6s.
You need a lot of shunts. :-) Thank you for what you do. There is probably no perfect solution.
where do you get your circuit breakers?
I don't mean to be the nervous Nancy here but looking at that table with those cells on it makes me think of the dog coming in and bumping against the table, those cells are going to hit the floor. 🤔🤨🧐
Nothing like this. This is a solid massive workbench made from 150x150mm hardwood bolted to the concrete floor. It does not move a single bit.
Dogs are all well behaved, they don't come into the garage until I call them. That's the rule.
Sir yes sir, and thank you for all your time and work. You are a wealth of knowledge and I'm learning a lot from you. cheers!
@@ryanaf2023 thank you
As always with high energy, if it goes sideways....it will do so spectacularly.
What ever you do, the cells are so blue...
Fully empty? Hell? Dunkel? Dunkhell? What the hell....
If you start with paralleled strings with seperate BMSs, you at least have the chance to monitor what is going on, and can in case of serious problems still parallel the cells.
But everything that uses pre-built packs of cells parallels the strings, so why should it not work? And if you have a voltage in the flat area of the curve, you might still encounter the same problems at cell level, but cannot monitor what is going on.
Are you allready planning the next garage for all the new panels to charge all your batteries?
I think the guys at Orion Bms are wrong when they say that the best configuration is 2p16s and that two series of 16 connected in parallel is not good. I think it's exactly the opposite. Think of it when two cells are connected in parallel, the total current in that circuit is divided into two parts that will never be equal because neither cell is identical, not to mention the fact that they cannot be monitored and balanced individually. If one of the two cells starts and degrades over time, we will not know this and will circulate through the other cell a current much higher than half the total current in the circuit, as would be ideal.In the end, apart from the fact that the two cells will degrade quickly, the total capacity of the battery bank will decrease.I suffered this at the lead battery bank wich I use now. I gave up 2p8s and switched to two series of 8 connected in parallel, but I think it's too late and some of the batteries in my battery bank are compromised .I will not repeat the same mistake with the new LiFePo4 batteries wich I ordered and I am waiting to receive them.
In the configuration with two series of 16 connected in parallel, at least we have the possibility to monitor and balance each cell separately with active balancer or bms. And if the 16 cells in each series are balanced correctly, we will have the same total voltage on each from the two series and automatically through each series will flow half of the total current in the circuit.
And that's why they build one of the best BMSes - because they are wrong.🤣
Deseando ver el final. Gracias por tus enseñanzas
This is your comment reminder to take a break brother. Go have a coffee and your choice of flakey pastries :D
Haven't uploaded a video for 4 days. I'm on deprivation...
I put 4× 80 amp diodes on each battery and supply them from there own mppt charger , no circulating currents
the problem is with more cells in parallel the more hassle the bms will have keeping the cells balanced now while they new i doubt this is much of a problem but as the cells start to age i think they will need to be balanced more often with only a low dump load capacity to discharge the other voltage bank it will struggle to bring that bank down to that of the others
Maybe get another Victron SmartShunt to have one after each banks BMS? To monitor what happen with the individual banks? At least in the beginning?
Well, could have a 3rd for both also?
Each BMS has their own shunt anyway, so I can monitor that through these. Will be interesting to see what will happen.
@@OffGridGarageAustralia yes true. I have a Daly BMS, and it not show drawing current so exactly , I think it need to draw around 2A before it start show the draw current , under that it just display 0. For example, when I have my led light connected, it still show 0 A. I guss yours better.
where is part 2 ?
That is part 2. Part 1 is here: ruclips.net/video/zSMiB4_kVYg/видео.html
Another great video Andy thank you
How do I get ahold of you? I went to your website, and tried finding an email or something? Had some cool stuff to share with you I know you’d love!
I have a 16S lifepo4 and a 14S lion that are paralleled together, each has its own BMS, the charge and discharge curve are very different to each chemistry, the overall charge voltage is 56.6 volts, these set of batteries have been married together for over 2 years, the only thing to watch is both banks are the same voltage when married
Ross, I note that you have also built some pretty impressive LTO battery banks using a modified form of Stuart Pittaway's DIYBMS. Possibly this is worthy of consideration for Andy's project? Looks good to me but I'm not really qualified to comment! So, over to you.... if you think it could be a candidate in Andy's solution I (and hopefully others) would welcome your comments. Cheers Chris
Here in the U.S we have available a number of "drop in", 12-volt-nominal LiFeP04 batteries which are designed to replace lead acid batteries, primarily in mobile applications (RVs of various flavors, boats, etc.). Battleborn is probably the most well known of these.
The individual drop-in batteries vary in construction and type of cells but are always internally in some form of series string to get the 12 volt nominal voltage (there are also 24-volt "drop-ins"). The "drop-ins" also always have an internal BMS (if they don't have an internal BMS they can't really be considered "drop-in").
When more than one of these batteries are used in an installation (which is usually, based on my observations) these batteries are paralleled (which, when using a battery with internal prismatic cells, exactly duplicates the setup described in the video, albeit at 12v) and, sometimes, put in series strings and then paralleled. E.g., in my caravan (travel trailer in U.S.-dialect) I have six 105Ah, 12-volt-nominal batteries (each with 4 prismatic cells internally). The six batteries are arranged in 3 series strings which are paralleled for a 315Ah, 24-volt-nominal bank.
These sorts of installations have been going on in the U.S. for several years and I've yet to see any real complaints nor have I observed more than the usual number of smoking holes in the roads around here :-)
How many times you will have one battery bank disconnected? well, if you carry on DIY-ing this stuff intensively, this might happen quite often :D
Once he's done fiddling with details, which I think he is mostly done now that he has decided to copy-paste his current battery box setup, he won't need to disconnect things anywhere near as often once he's done changing what he needs to change to accommodate parallel batteries. His next major disconnect will likely be when he adds another electrical box to house the battery bus bar, the dual DC breakers for each battery and the battery cable inlets since his current one is too small for all of that extra wiring and hardware.
I've worked with dual parallel batteries my entire life. One isolated one night while the other gets consumed, and the opposite the next night. That way there's always at least 50% available the next morning in the worst case scenario. That exact system is in need of a LiFePO4 upgrade and I've been planning it for a while. However depending on which cell sizes are available when the time comes to order, I was thinking of ordering 16 cells instead of 8 for 2 banks. I guess that's still on my mind, but now I'm thinking about 4 banks instead of 2 :-)
Mal wieder sehr informativ! Danke!
Ahahah
Good executive call. 👍
That's the way mate!
And a correct and possible one also.
Already seen three LFP battery banks in parallel each one with individual BMS, and no worries.
Even saw Battery banks of LFPs being paralleled with different chemistry NCAs and working like a charm for years now... So... Go for it!
1p 16s and don't forget 1b..(one beer) 🍺🐸
I am happy with my 2P4S LF280 configuration for my sailboat, but I seriously considered 2 separate 4S batteries. Regarding parallelizing or separate batteries, consider that companies like Renogy and Battle Born make LiFePO4 drop in 12V replacements with each battery having a 10 year replacement warranty! Many people upgrading marine systems(and RV folks too) have built systems with 8 - 12 100AH drop in batteries. I would have gone this route with 5 100AH BattleBorn batteries but for the cost. My 560AH battery cost me under $1k US total whereas 5 BattleBorn batteries were close to $5k US.
I just realized that comment where I included a url link had disappeared. For example, my comment for the last video suggesting to use plywood to make a battery box on your rolling dolly has disappeared. I had many suggestions based on my battery box build with plywood and 1/8" thick FRP wall paneling. I included a link to my sailboat blog showing the build. I think adding the link killed the whole comment. My fault.
The cells inside these BattleBorn batteries are extremely precisely built with exact capacity.
They also life cycle test the technology - they age the same.
To do this with random undocumented chinesium cells you need to buy 5 times more than needed and do capacity tests.
Than if lucky you will find ones which are close enough.
Without life cycle test...
And here comes the same cost, as you can see...
I'd be curious to see if the capacity is affected in any way. Maybe you could run 5 or so charge and discharge cycles and then do a combined capacity check with atleast 0.5C to see if it you get the expected combined capacity.
Hi Andy, I have not built anything yet, but I would go for multiple strings of 1p16s, its more scalable! Anyways its a pleasure watching your videos and learning things. Should I say things or stuff?
Not only should the BMS Cable be on top of the bus bar...you should also strip the insulation off of the cable when you hand wrap it around the post! LOL