Hi all - yes click-bait like topic for good reason.

Before I decided to write this particular post I also scoured this forum and noted that various users are having quite frankly unacceptable levels of SOC issues with all kinds of batteries even Pylontech and other "reputable" brands - so I'm starting to think that the majority (i.e. not all but 50-60%+) of the batteries and new makes and models that seem to land on our shores every other day, are simply not suitable for our unfortunate and insidiously unique power problems here in SA - I'm talking about frequent loadshedding, multiple times per day for 2-6 hours at a time, depending on where you are fortunate or unfortunate to be in our country.

Those that would like to follow previous long CFE discussions can view them  here  and I have a very long (sorry) but complete view of my installation, how I use it and the dreaded SOC drop from 40%-10% in 5 minutes or less or worse still 40% to 0% almost instantly here

For those that don't to go read them, a very summarised version is bullet pointed below which doesn't give you 100% context - so if you're going to ask questions or tell me I'm 'stupid' without referencing important info in the original post(s) - be warned - I'm called the TheMafiaMan for a reason...🤣

• 8kw DEYE inverter
• 2x CFE 5100s batteries setup as 30% restart 20% battery warning and 10-15% Shutdown. Final Battery setup will be 20kWh of storage via 2 more 5.12kWh batteries to be added in 2023 still.
• The system is protected by 200amp keto fuses and each board has surge protection as well - the house sits behind the inverter because A) I want clean pure sinewave power to all my plugs, lights and screens and B) fridges and appliances with inverter type motors are all better protected that way too. Your electronics will thank you.
• I charge the batteries @40amps so roughly 2kWh have experimented with 20amps and even 75amps - for my use case for now 40amps is my sweet spot giving my inverter head room to still run other higher draw loads when the batteries are charging.
• No solar currently, only used as a storage system to facilitate work from home and a "normal" quality of life (8.1kWh solar array is being commissioned)
• 6kva diesel Genset (dedicated generator room for Genset and fuel - already wired in with manual change over) just finishing off sound isolation, ventilation and then will automate change over with the dry contact and Gen feed-in on the DEYE Inverter
• Finally Eskom connection will capped off once solar is in and run in my own "Islanding" setup where I am my own "Utility provider".

Let me start by pointing out that my supplier for all my equipment mentioned is Solar Advice - solar.co.za and that they have been only helpful, support providing and quick to execute where CFE keep dropping the ball - silver lining they will be sorting out my problem without fanfare - the way a reputable business should and I keep recommending them over and over.

From my side I agreed to go through not one but 2 firmware updates with the "overseas - team" at CFE, resulting in about 75% of my frustrating 2+ weeks of irritability.

Let's get on with it - two plus weeks of frustration, hair pulling and plenty, and I do mean plenty of potty mouth later - I now have the latest CFE-5100s firmware on both of my batters and the great news is...

I am at least 50% worse off than before, cannot make it through 2hrs of loadshedding with 10.24kWh of storage and less than 2kWh load, and those that are good at math that means? Yup 79-80% and then 3% 3-5min later, and then off. Unacceptable by any standards - I mean are there a few AAA batteries inside my two 5.12kWh LiFePo4 units.....???

But this doesn't really paint the whole picture so let’s start in Nov 2022

• November 2022 system is commissioned and installed.
• Batteries are charged from +-56% to full over about 3hrs at 30amps or roughly 1.67kWh - for the Mathematically/Electrically inclined, 1602watt to be precise - DC wattage and AC wattage are not 1:1 parity 30amps at 53.4v is 1602watt - I was not charging my batteries at 6.6kw - if you think that, then this thread is not for you!
• load$hitting had not hit its full stage 6 stride at that time, so think mixes of stage 2/3, 4 and 6.
• Batteries were being used and run down to about 80-85% in stage 2 over 2 hours with mixed loads ranging from 1-2.5KWh and 55-60% in stage 4 or 6. We had no interruptions and not a single error or problems.
• We had a 12hrs long load$hitting event twice in the Dec/January and batteries hit 41% with a very OCD (me)controlled 350-450w/h load on these specific occasions – again no problems.
• Then we got into full stage 6 load$hitting 3 times a day - yep, I live in an area where we got 9-12hrs per day, day after day.
• Finally but irrelevant to the SOC issue like many other CFE and Raytech battery owners I had the 100%-90% cycle which I didn’t particularly have a problem with – go read the long post I made, to understand why – batteries should not sit at 100% for hours and days on end – I hope you have a fire extinguisher near – more info about that in my long ramble as well – go read it!

This is when the problem reared itself for the first time between 40-38% left in the tanks and the inverter would either throw anything from a F56DC_VoltLow_Fault usually followed by but no exclusively F58BMS_Communuication_fault to even one or two F59AC_V_GridCurr_High Fault – for this discussion the most frequent by an overwhelming margin is the F56DC_VoltLow_Fault just as or about when the system hits 11/5% or 3% whichever the Chinese battery lotto gods deemed fit that day and then off.

The First update which required the inter-battery comm cable to be disconnected - took less than 3 min OTA (over the air) – each battery has a WiFi Module and independent battery monitoring software that can only seem to report voltage correctly as confirmed with a Voltmeter - but is useless at doing anything else.

I was instructed to cycle the batteries a few times between 10/15% and 100% and Monitor it.

The looping 100-90% discharge had changed to a 100-95% loop but again refer to my first long post (yes this current one will probably become known as the longest waffle on this forum – hopefully I’m a little entertaining at the very least) I still didn’t take issue with that.

1 week later the batteries now don’t cut out at 40% but now “gracefully” (intended sarcasm) go from 80-60% in 3min flat, then on to 50% at a leisurely pace and then when they hit 43% descend rapidly form there to 30/20/10/5/3% and then dead in 5-10min.

CFE stop responding in our happy little WhatsApp group containing expletives of me telling them everything from "they don’t know anything about BMS implementation, firmware and code writing" to how generally useless they are in responding to real issues – this goes on for the rest of the week.

They then announce how sorry they are ("Sorry Dear"??? I'm not anyone's darling/dear), that there WhatsApp wasn’t allowing them to log back in but that they had triumphantly a “new” firmware that would sort out all my trouble 😊

Disconnect comm between the two batteries again, never hear form them for another day but during the early evening get plunged into darkness because the batteries are flat and now to add to my annoyance charge from 0-100% in 45min going form 40-80% in 5min because can you guess why? They’re completely unbalanced – remember the inter-battery comm cable is disconnected waiting for the firmware update that never happened, and I now spend the next 2 days manually discharging each battery – via voltage control not BMS (by now it should be apparent to you that the BMS is $hit) and then charging each one individually, paralleling them again and reconnecting comms etc.

Firmware the 2^nd finally happened on Tuesday this week – remember the one that will solve all my trouble😊 – batteries turn off after a successful OTA - Firmware update complete.

100-95% cycle now replaced by 100% and no cycling – hmm not too happy about that but I have other things to do, like work… and then Load$hitting that night 6pm - and 18min later plunged into darkness – 100% to flat - yes my troubles are truly over!

By now some of you probably have an idea as to what is the root cause of at least one of the problems – their BMS is beyond $hit – how do I know? Read on…

With torch in mouth while hurling profanity (I am quite talented you can tell) I disconnect everything to a safe point, reconnect only the master battery i.e., 1 of them and start everything up – switch to voltage monitoring mode to bypass the BM$hit (see what I did there) and apply a 4.2kWh load which runs for a whole 50min before the battery enters a fault mode and the solenoid disconnect trips – Thank god that actually works. Inverter shows voltage dropped below 47.9 (my own setting) and therefore the battery is at 5% or less.

Wash, Rinse, repeat – do the same with battery 2, the slave – and run for about 70min with a mix of 2/3.5 and 1.1-kWh loads during that time and then click - solenoid disconnect trips at 47.8v.

This all after the BM$hit declared that they where both flat to my poor DEYE inverter almost 2hours prior. By now I'm sure the inverter has developed or is developing schizophrenia.

Luckily by this time the power is back on and both batteries are at 1-2% each so I reconnect, re-parallel the batteries and connect the inter-battery comms and let the unit charge up in BMS mode (shocked I'm using the BMS again? There is a method to my madness) @40amp so 2.01kWh and monitor.

3hrs later we’re at 61% - great that is more or less correct, but then it all goes to $hit once again as 61-83% happens in about 10min and then 83-100% took 45min further – confirming my “rubbish BMShit” hypothesis in both directions conclusively. I did this 2 more times in both directions to confirm. Confirmed – in case you’re wondering - exactly repeatable.

I worked out on some spreadsheet – because I have time to kill like that – TheMafiaMan remember. At best, I can get about 6.5 maybe 7kWh of usable power out of my 10.24kWh storage but not without a massive, probably warranty voiding and time consuming workaround as you've seen above. Sorry, not good enough – I allow for 10% DoD as per the manufactures spec or maybe that should be claims. I paid like many people, a lot of money for storage, one of if not the most expensive single components in your system and I will not accept 60-70% and of that completely unreliable capacity – neither should anyone.

Anyone who has read this far – congratulations – but you must also by now agree, enough and no more. Which is exactly what I communicated to SolarAdvice – who have been privy to my WhatsApp “consult” with CFE and agree that I’ve exhausted all options.

I have options that I will be exercising and that is getting rid of CFE off my property once and for all – but this is basically a warning to all and sundry.

STAY AWAY from the CFE product – A properly working, reliable and trustworthy BMS in a LiPo, LiFePo4 and the like is non-negotiable.

You shouldn’t have to wonder if you can make it through 2hrs of loadshedding because the data you get from one of the most expensive items in your alternative power solution is “wonky” at best.

Unless you like dancing with fire – literally - and have a few chemical fire extinguishers on hand – you should have at least one with any large Lithium-xxxx-insert-your-particular-battery-chemistry-here anyway, but right now – for me, these represent a substantial electrical and fire risk to my property and the vehicles in my garage. I’m 100% not prepared to roll the dice on either side. Besides the fact that if you’ve followed my rant this far (thank you) this is not a workable use case and why I’ve had to postpone the Solar Array until I have reliably working storage solution.

As my intro eluded, this seems to be a far wider reaching issue cropping up on other battery systems too – so like others I’m left wondering what my options for a reliable battery brand are. One that can provide a battery system suitable for the SA market with our unfortunately unique power situation that actually works…

My Short list so far is 10-15kWh, 1C discharge rate and 10yr warranty

• SolarMD’s 14.3kWh battery.
• Freedom Won 15/12 battery -  but not so sure I'm onboard with Local-is-lekker - Solar MD are supposedly also local but I know of a couple of installations    with both brands that are more than 12-18 months old and have had zero issues with the batteries, so I'll need to investigate some more -  there is at  least one horror story which seems like a 5 year old model with 470+ cycles that is basically dead and FW cannot offer replacement cells because they're EOL(end of life) according to the comment placed this guy purchased 5 x 30/24 units which he must have sold organs for   
• LG Chemistry – I don’t believe I will be able to source or afford it – at the moment rated worldwide as one of if not the best, but a 16kWh unit is almost the cost of a Tesla Powerwall think €9,377.00 sourced international with shipping and then add what ever bend-over-tax when it gets to SA and I'll most likely be at R220-240k+ and that’s without the Tesla’s already built-in inverter. You also need an installer with LG Chem Installer Certification and I've only found the 6.5kWhunit locally but no pricing from SegenSolar who to date have not responded to email or phone calls - I might try again.
• DEYE – I’m hearing good things, but they are also expensive – R38k-R43k for 6.14kWh (R6.1-7k per 1kw vs R5.2k I paid) - in Fairness if you pair it with a DEYE inverter they extend your inverter warranty to 10 yrs from 5yrs to match the 10yr battery warranty,  in essence you're paying for the 5yr extra warranty on your inverter so this is looking like an option but only if I can get a higher capacity single module between 10-15kWh will I consider this Sunsynk which is DEYE vice versa have a 15.97 LFP wall mount unit for about R96k so that is priced in my opinion more inline but I'm not sure if you get the same inverter warranty extension for a DEYE inverter because you do get an extra 5yrs on the Sunsynk invert...

I’m not really interested in Hubble for personal reasons (but feel free to change my mind - also in the same boat with local is lekker - but I see that actual capacity seems to be questionable), anything with ESS in the name – I think CFE and few others are all the same 5.12kWh battery in fact wrongly or rightly so they’re unheard of in Europe, the US and Australia, both large markets for Batteries, which aren’t accepting inferior LiFePo4 and the like batteries onto their shores and we should be asking ourselves why.

The Only exception here is AlphaESS which SolarAdvice use to carry but no longer do and in hindsight while I was told the systems cannot be paralled to each other the 5kWh/10kWh Storage & Inverter Combo Smile5, is well documented and supported in Australia, Singapore and in some parts of Europe and the paralleling issue can easily be worked around by creating separate 5kW circuits and running separate subsystems - if I had to do it all over again which I cannot afford to do now, I would go this route because I have an honest Sparky and that means I would have split my solar array into 4x 2.025kw Arrays (2xMPPT per inverter) feeding 2 separate 5kw circuits with 20kWh of storage between the two and cost wise, would have come out about the same as my 8kw Inv. 20kWh(eventually) storage and 8.1kw Solar Array - However, I wouldn't attempt this type of setup with the majority of unscrupulous 'electricians/installers' I have had the displeasure of dealing with over the years.

Rant over - For now...

 

 

 

Edited June 19, 20232 yr by TheMafiaMan - NotMuffinMan
important info for context

Hi

Thanks for this comprehensive sharing of your pain...

I have a 5KW Deye + CFE-5100 installation since Nov 2022 (panels to be added coming week).  Much the same battery experience.  After a lot of to-and-fro with the installer/supplier and several FW updates I was still miffed with the performance (any load beyond ~45% SOC causes battery to drop out, and then charges from 0% to 70% and then jumps to 100%).  Eventually the installer swapped out the battery with a new CFE-5100.  Now the same issues are resurfacing.  Plus I often get the very annoying "F56 DC_VoltLow_Fault" error message on the Deye with a high load on and SOC < 40%.

I fully agree that the BMS is super bad.

I will have to attempt to convince the installer to replace the battery with another brand...  Do you have an opinion on Greenrich, Dyness, Shoto brands?  I see they also have a 5.32kWh Sunsynk in stock for R47k...

Hi Jak

My hope is to inform and educate (in my limited way) especially around the batteries available in SA - if I had the money to run an independent lab I'd be testing batteries all day and I think the South African public would be shocked (no pun intended ) to see how they perform in various scenarios.

It's scary how repeatable this very issue is with the CFE-5100s battery - what you've outlined here is exactly my very issue.

As for recommendations I cannot comment on Greenrich - forum long timers will need to step in and correct me but they're fairly new I think,  12-15 months and it seems like feedback is mostly positive - Shoto seems to me at face value to very similar to CFE - personally I'm not going to go that route - Dyness I've always struggled to find large sets of info and large user bases of, but touch wood - I believe them to be very reliable.

Personally given the warranty extension DEYE is offering on a DEYE inverter when you purchase their own battery its looking like an option - but an expensive one - for me I would have to spend R20k more (1/3 of my solar array cost) than I originally did to get albeit a slightly larger storage system 12.28kw vs 10.24 and the extended warranty on my 8Kw DEYE inverter but one thing I've not mentioned to date on why I chose the CFE ones besides the 10yr warranty was the mounting solution I opted for and the space I have to work with - I'm using CFE's vertical wall mount which allows me to vertically hang two CFE-5100s back to back in a really small foot print - my setup and space allowed for 8 such batteries to occupy less than 1.5sqm of wall real-estate for an eventual 40.96kw storage system in the smallest space today of its kind. 

• If I now opt for the DEYE 6.14 kw unit I'll be limited to 3 batteries maximum and 18.42kw of storage but at 25kw (5x 5.12kw) of storage price when compared to CFE
• Solar MD I'd be limited to 2 units but would get close to 28kw which I can live with and not break the bank.
• Sunsynks 15.97 LFP 2 units as well and then have about 30kw of storage eventually

Or

• Sunsynks 10kw which I can fit 3 of space wise and also get 30kw - this one makes more sense to me when we talk redundancy, pack failure and swap outs if ever needed

Bottom line is this - I'm in for a far greater expense either way - but rather now this early in and while I have the option of RMA'ing the CFE's and getting something else - I have a feeling dictated by finances and space its most likely going to be 2 x 6.14kw DEYE batteries...if you've been quoted R47k for the Sunsynk then rather go for the 6.14kw DEYE they range currently between R38k-R43k - if that's an option for you I'd go that route and make sure your Inverter warranty gets and benefits from a 10 Warranty and call it a day - and like I'm saying in a lot of my posts a 4.5-9kg dry chem / ABC fire extinguisher - always better safe than sorry. Cheers.

Edited June 16, 20232 yr by TheMafiaMan - NotMuffinMan

Hi

I also like the wall bracket that can hang two CFEs vertically, but I'd much rather have a robust / reliable BAT+BMS combo than worry about mounting and space issues! Do you know whether the Sunsynk batteries also come from the Deye factory (like the Sunsynk inverters do)?  And whether the Deye warranty extension will apply if you add a Sunsynk battery to a Deye inverter?

PS. I see that my installer/supplier doesn't list CFE batteries anymore...  I suppose they simply stock whatever they can lay their hands on at any given point in time.
PS 2.  I'd love to build my own battery some day when I'm grown up 😉
PS 3. Thanks for the fire extinguisher reminder! I've recently seen a video of a LiIon fire on an electric scooter, quite scary.

100% agree the mounting flexibility means nothing if the BMS/BAT is unreliable just unfortunate for me as i have dedicated space and will not be able to expand over time as i wanted to - this morning I did another test to qualm my paranoia with my two batteries and well this image shows it best

100%-7% on a 5.27kw constant load in 25min then back from 7% to 100% with a 40amp charge in 55min...flaky as all hell - and I'm running my DEYE in BMS Battery mode. As you can see the latest update I got from CFE now pins the batteries at 100% - I think after 1hr or possibly some other defined factor I'm not privy too, it does stop taking a charge and then discharges at about 30-40watt - not sure what the nominal discharge rate would be sitting idle.

I'll know on Monday if the Sunsynk batteries are from DEYE and if the inverter warranty extension applies - I think it won't but best to check my facts...even if it doesn't I might consider spending the extra to get the higher capacity but I'd prefer slightly smaller capacity and multiple batteries for redundancy if there was a failure of a pack etc. I'll let you know what i find out.

Hello, so the solution to your problems is very easy.

1. Setup your battery in your inverter by Voltage, NOT by SOC. Unplug the comms cable to the BMS. Check the manuals of your battery for the recommended battery voltage settings for Bulk, Float and Equalization voltages and program that into your inverter. 

2. Ensure that you fully charge your batteries every now and again. That is what the "Equalization" function in your inverter can be used for. On the Deye/Sunsynk inverters you can program it to do equalization every x number of days. I would say once every 3 or so days. This ensures the battery gets fully charged and whatever voltage "drifting" that happened in the last 3 days has been rectified.

Once you do the above two steps you will no longer have the issues you are describing. Thank me later.

UPDATE: So I see your battery manual only states the Max Charge Voltage as 56V. I did check and they use LFP type cells, so it would be safe to assume the best voltages for your battery should be:

Float V: 53.6

Absorption V: 55.2

Equalization V: 55.2

Shutdown V: 47

Low Batt V: 48

Restart V: 52

Edited June 17, 20232 yr by HendrikBigChief

@HendrikBigChief Thank you for your very detailed info very much appreciate.

For some further clarity from my side - I'm not running the inverter based on SOC but rather BMS Battery Mode - CFE and I've checked with some other battery providers will not honor a warranty on an LFP battery run in Voltage Managed mode - Its universally accepted especially in the EU market that you should not run specifically LFP batteries without a BMS - Especially in the manner you've described where 2 or more are involved and there is for obvious reasons no comm connection between the 2 batteries.

To be clear I have already run as part of testing and fault finding both batteries individually in Voltage Managed mode to get to the bottom of my issues and the fact remains that on a 2.5-3kw load these batteries assuming for any efficiency/conversion losses and even generous ones, should be able to run a 2.5kw constant load for 110minutes - I've allowed in the calculation a 10% DoD - in the tests I ran even at 2kw, I could not get more than 55min of runtime - at which point the CFE's low voltage solenoid can be heard triggering a contactor disconnect and then go into a low voltage fault mode or roughly 4-5% charge left at 48.7v - already above your 47v suggested Shutdown value.

I used your exact numbers except for shutdown at 48v and low batt trigger at 48.7v but my batteries are clearly foobaared as I cannot even get them to 48.6v. I'm not going to give CFE ammunition to possibly wrangle out of my warranty as I know they store values for BMS bypass - it's not worth me risking the best part of R52k and its clearly not they way most LFP if not all should be run - again you're more than welcome to disagree with me, but my family and properties safety is much more important.

• I have this all documented for the supplier and they agree that CFE have basically proven they cannot fix it - I'm not going to keep accepting firmware updates that are progressively and clearly becoming more and more unreliable.
• I'm not going to run 10-20kWh of LFP batteries in non-BMS mode - sorry I consider that unsafe - you're more than welcome again to not agree with me - but I've dealt with a large Lithium based fire before and since this is my personal residence I'm not willing to risk the worst case scenario The BMS must do its job and I will seek out a battery manufactured that has a proven reliable one - even if that takes some time - I do have a diesel Genset all wired into the property so I'm not completely SOL for alternative power.
• Unless you have a further updated manual for the CFE5100s I've 2-3 different versions spanning 2022, and they do not provide those numbers as you're rightly so expected to run them with the BMS

Again - appreciate your detailed input and help - but I'll be RMA'ing these batteries and hopefully replacing them with something that works correctly - in BMS mode

Edited June 19, 20232 yr by TheMafiaMan - NotMuffinMan

So this is my last ditch effort at trying to work with CFE before I throw the towel in completely on these batteries...

My printed manual and some of the online versions I found show the parallel connection of 2 and 4 of the batteries as per Fig1

Fig1

This afternoon while digging and digging and digging some more to get some last ditch info I stumbled across another CFEnergy site and downloaded a manual dated 20230321 and found this diagram for parallel connection which looks like this Fig2 - the main difference being that the positive of the first battery and the negative of the last battery are what go to the invert - as apposed to the positive and negative on one battery going to the inverter and the two being bridged in the same way.

Fig2

Anyone who has these CFE-5100s batteries 2 or more @JAK, can you please confirm what is the correct connection method or if it makes no difference? Mine are currently connected as per fig 1 but this might now be incorrect...though I have no errors or faults so maybe its nothing - I've also confirmed after watching a very long training video conducted here is SA with CFE that:

A) The white units which have serial numbers starting with 1416x are from 2022 - so that's good its not an old battery from 2021 - and mine are 14168xxxx which denotes batch 8 from around Aug/Sep and its is definitely a v3 of the 5100s and has the newest version of their BMS - apparently v1 and v2 where different.

B ) in the video an installer or reseller mentions or part answers a question about the cables as per what I'm asking above and its even more unclear whether FIG2 is for series connections as he sort of confirms then back tracks on with some mumble about that not being correct...or that you wouldn't get 200ah and only 100ah blah blah- frankly I really don't care - he confirms how clueless most installers are and that they use us as guineapigs.

FFS!!! which of any of this is actually correct or valid and can someone take a picture of there CFE units so I can compare - please.

Thank you in advance.

Edited June 18, 20232 yr by TheMafiaMan - NotMuffinMan

@TheMafiaMan - NotMuffinMan, my understanding of @HendrikBigChief's post is that you're making changes to how the inverter interprets battery SOC.  By no means will the inverter disable the battery BMS!  So even when the bat-inverter interface is set to voltage mode, the BMS will still continue to do its thing (which in the case of the CFE appears to be shoddy...).  No LFP battery should ever be operated without a BMS.

It's however not clear to me what the difference in lifetime of an LFP battery will be when operated with an inverter in voltage mode as opposed to BMS mode.

10 minutes ago, TheMafiaMan - NotMuffinMan said:

Anyone who has these CFE-5100s batteries 2 or more @JAK, can you please confirm what is the correct connection method or if it makes no difference?

Sorry, I only have one...

Thanks @JAK

 

14 minutes ago, TheMafiaMan - NotMuffinMan said:

and mine are 14168xxxx

Same here: 1416802xxx

2 minutes ago, JAK said:

@TheMafiaMan - NotMuffinMan, my understanding of @HendrikBigChief's post is that you're making changes to how the inverter interprets battery SOC.  By no means will the inverter disable the battery BMS!  So even when the bat-inverter interface is set to voltage mode, the BMS will still continue to do its thing (which in the case of the CFE appears to be shoddy...).  No LFP battery should ever be operated without a BMS.

It's however not clear to me what the difference in lifetime of an LFP battery will be when operated with an inverter in voltage mode as opposed to BMS mode.

Thanks Jak, yes I know it doesn't "disable" the BMS but it does bypass the BMS "rules" for the Pack to a certain extent and I'm just not going to do it because that alone proves the BMS is faulty or badly calibrated which means I cannot rely on it to do its primary job correctly

@JAK do you agree though that the two manuals contradict one another in terms of parallel connection diagrams? Or am I just being paranoid?

Fully agree.  I'm also not willing to accept inferior performance after paying R30k+ for an item.  Will post once I've made some progress with convincing the supplier to take the CFE back...

2 minutes ago, TheMafiaMan - NotMuffinMan said:

I cannot rely on it to do its primary job correctly

Also my reason to be apprehensive about doing this is that when you run them via the inverter in voltage monitoring mode the BESS app (which is mostly rubbish) cannot communicate with the battery at all (until you go back to BMS mode) and provide any battery details, no individual cell voltages, SOC, AMPS in or out - I actually think in this case the BMS is "asleep"...

17 minutes ago, TheMafiaMan - NotMuffinMan said:

do you agree though that the two manuals contradict one another in terms of parallel connection diagrams? Or am I just being paranoid?

Electrically there is no difference between the two diagrams.

@JAK  I was of the same opinion but I’ve been staring at this nonsense over the last week for so long I wanted to see if I was missing something - thanks.

55 minutes ago, TheMafiaMan - NotMuffinMan said:

@JAK  I was of the same opinion but I’ve been staring at this nonsense over the last week for so long I wanted to see if I was missing something - thanks.

It should be figure 2. It's to do with the cable lengths to each individual battery from the inverter. This way they are all the same so batteries should charge and discharge more evenly.

@TheMafiaMan - NotMuffinMan

One more thing you can try is to leave it in BMS lithium mode, but then in your "System Work Mode" screen, ensure you have a period of time every day where the battery is fully charged via grid, this will ensure that the SOC is correctly calculated by the BMS. So for example after the whole day of sunshine, at 6pm, tick grid charge and set battery level to 100% until say 11pm.

Once you do this, let it run a day, then please post a chart here with the Battery Voltage, Battery Current, SOC and Battery Power. I will then be able to diagnose if your issue is fixed.

The tech is not perfect. People have high expectations of the stuff, but the reality is that batteries are not perfect, conditions are not perfect and the firmware running on the BMS is not perfect. I doubt you will find batteries that perform much better. Multiple battery brands have the same issues.

Calculating the exact correct SOC is nearly impossible with lithium batteries (the voltage is nearly flat from 20% to 80% SOC). There are many many factors that influence it such as battery chemistry, battery age, battery internal resistance, temperature, charge and discharge rate, individual difference between cells, luck, etc. SOC is based on an algorithm that approximates the SOC, but it can be wildly wrong. Charging to 100% once a day resets the algorithm in a good way.

Also, the way we use the batteries in SA is not really what they were intended for, we not only use them as solar batteries, but we also use them multiple times a day as backup for load shedding. 

Edited June 18, 20232 yr by HendrikBigChief

1 hour ago, TheMafiaMan - NotMuffinMan said:

Also my reason to be apprehensive about doing this is that when you run them via the inverter in voltage monitoring mode the BESS app (which is mostly rubbish) cannot communicate with the battery at all (until you go back to BMS mode) and provide any battery details, no individual cell voltages, SOC, AMPS in or out - I actually think in this case the BMS is "asleep"...

Nope, the BMS is alive, it is still doing cell balancing and checking for over and under voltage as well as over charging and logging info. In a way the battery is still communicating with the inverter as the inverter reads the voltage from the battery and uses that in combination with the amps pulled to determine what the battery is doing.

2 hours ago, JAK said:

It's however not clear to me what the difference in lifetime of an LFP battery will be when operated with an inverter in voltage mode as opposed to BMS mode.

Personally I believe running it in BMS mode will shorten the life dramatically in the way it is working now for two reasons.

1. You are currently fully discharging the poor battery every time the BMS miscalculates the SOC as higher than what it actually is and then the system crashes when the true SOC is actually 0. Very bad for any battery.

2. Lithium batteries don't like a high float voltage, many BMS'es specify a float voltage that is much too high, that is why I specified the lower float in my first post that is known to be good for lithium.

On 2023/06/18 at 7:14 PM, HendrikBigChief said:

BMS. So for example after the whole day of sunshine, at 6pm, tick grid charge and set battery level to 100% until say 11pm.

Thanks @HendrikBigChief I'm running in storage only no solar currently.

 

On 2023/06/18 at 7:14 PM, HendrikBigChief said:

Also, the way we use the batteries in SA is not really what they were intended for

Agreed.

On 2023/06/18 at 7:14 PM, HendrikBigChief said:

Calculating the exact correct SOC is nearly impossible with lithium batteries (the voltage is nearly flat from 20% to 80% SOC).

You keep referring to SOC (I beg to differ on the correct calculation thereof but anyway)- I'm referring to the actual usable Kw I'm getting out of the batteries in BMS or Voltage mode being 6kw at best for 10.24kWh of storage (the center of my gripe to begin with)

60% of my installed capacity after 7 months of frankly light use? - not acceptable! I've calculated multiple times with loads I know to be correct and constant and can monitor via other means and not just the battery app, inverter logging or DEYE Cloud/Solarman so that I can verify the data. 

On 2023/06/18 at 7:14 PM, HendrikBigChief said:

One more thing you can try is to leave it in BMS lithium mode, but then in your "System Work Mode" screen, ensure you have a period of time every day where the battery is fully charged via grid, this will ensure that the SOC is correctly calculated by the BMS. So for example after the whole day of sunshine, at 6pm, tick grid charge and set battery level to 100% until say 11pm.

I've experimented with this before "time of day usage" I have no solar so not really relevant - I can confirm that this has been tried before as a possible problem solving approach and made no difference - for example allow the batteries to go to 60% with solar but while being Grid tied do a final run up to 100% late to make sure you're topped up properly (I understand the rational )- my intention and detailed below is to be 100% of grid I will not be grid tied at all - but in a Storage solution only as i'm currently setup I don't have the luxury of entering say stage6 with 60% left at any time of the day, especially when the power doesn't come back on in some instances for another 1-3hrs

On 2023/06/18 at 7:46 PM, HendrikBigChief said:

You are currently fully discharging the poor battery every time the BMS miscalculates the SOC

I never said I was doing that as my normal operation - I suggested at the beginning of my thread people should got back and read my first post and the original one I referred  to because it seems you've not done that other than insist multiple times now that BMS are bad and that I run my system in Voltage mode and thank you later. I won't be running in V mode - I did that previously only to confirm the data and for trouble shooting 2 weeks ago during the firmware updates to corelate data I had collected. 

On 2023/06/18 at 7:46 PM, HendrikBigChief said:

Lithium batteries don't like a high float voltage, many BMS'es specify a float voltage that is much too high, that is why I specified the lower float in my first post that is known to be good for lithium.

Hmm...why would any manufacturer have a BMS that specifically is going to shorten the life of battery(s) that they have to warranty for 10yrs if that is going to as you're suggesting almost definitely lead to them failing prematurely?

Why have a BMS at all then? See where I'm going with all this?

There are numerous applications, devices and systems that have existed long before LiFePO4/LFP & prismatic cells and the like and those BMS systems have all worked correctly in my experiences - I have lithium battery based systems 10+ yrs old that still work to within 80-90% of spec...

This all paints a very bad picture of batteries in SA at the moment and confirms my suspicions all along and as you've mentioned and rightly so that none are really suitable for the multiple loadshedding events we have - They are vastly exacerbated in a Storage only setup like mine.

That a real issue for me and the type of Money myself and others are forking out.

I've outlined previously but will touch on again the energy production route I'll be following:

• 8.1kw solar array (Still to be commissioned) - in ideal conditions, no clouds, no panel shading, ambient of between 25-35c etc. etc. I can comfortably produce 25-30units each day - an array this size can peak at 45+ units a day - I use a 4.8 hour factor to calculate possible peak and exclude the trailing in and out power that would be created from 7am-9am and 3pm-5pm which on a good day could account for another 8-12 units (over the conservative 25units) in my specific array - easily getting me over my worse case 20-25 units a day I need - so I've erred on the most conservative side possible.
• My battery storage is only meant to supplement 
  • poor production numbers between 9am and 4pm
  • and at night 10pm onwards where my base load never exceeds 700w
• I have embedded diesel generation - 6.5kw and fuel for weeks
• Using my program logic which I've implemented before, I can run the Genset between 8am and 8pm at night that steps in to: 
  • Supplement load if the weather is kak, or if storage is below 40% or solar generation is below a threshold I set or all of them combined
  • I can top up storage (batteries) if it is below 60% @4pm in winter and 5-6pm in summer (I can vary this window up to 8pm at night reasonably)
• In all these scenarios I'd most likely keep to my 40amp charge rate (which I've used) and well enough below the 0.5c recommended on most batteries if you want to actually take care of them.

We'll agree to disagree on you not believing in running a BMS - as I've stated for me this is simply non negotiable and for LFP poses a significant system risk - insurance companies would have a field day with 60%+ of the installations and so called "professional" ones I've seen.

The BMS is the heart of or the ticking time bomb in an LFP pack and it must 100% work - I'm non-negotiable on it. If I have to go through 10/20/30+ battery brands to find one that actually works according to spec - so be it - I'll burn diesel when they don't - simple.

The manufactures data on all of these batteries (and I've studied 10 of dozens of them by now) are disingenuous at best to flat-out lies, hidden in small print where some bother to disclose it with others simply stating numbers that go against the laws of physics / science and in this case chemistry. 

I'll pick on SolarMDs 14.3kWh pack (which I am considering) the real number in their specs which is the only one worth reading is 13kWh usable at 0.3C - that's a measly  3.9kWh load (for me) and I'm pretty sure that if I ran that pack at 3.9kw that I would not get 3h33min of runtime - SolarMD i'm up for testing.

Just like my CFE packs which can at best deliver 3kWh out of 5kWh @ supposedly 1C - they like to obfuscate the numbers and bullshit on their labels but these 5.12kWhw batteries can only deliver 0.5C semi-reliably which means you need to keep load at below 2.4kw to get anywhere near getting 3.5-4kWh of use out of a 5.12kWh pack -

Naturally I've tested this and @ 1.7kWh load the battery discharge curve is fairly linear to about 30% then tappers more seriously as the packs voltage drops to 50-49v - I don't know about you but I didn't pay R55k to have 1.7kw on tap and their technical spec says I shouldn't have to accept that dreary performance either. More so i'f i'm going to expand to 20kWh or possibly look at 2x 12/15kWh packs

I've seen numerous sites, heard advice given by resellers, installers and electricians alike (all self confessed "experts") that have for example said that 2 (or more) x 0.5c batteries means you now have a 1C system or 2C system (4 packs) etc.- this is categorically false - you have a larger storage pool but you're still governed buy the 0.5C rating.

Bottom line - for the average consumer out there this discussion and many others like it wouldn't even vaguely resonate with them or allow them to know what to follow - or do - so they do what most consumers do - follow the guidelines in the specs or rather the specs the installer regurgitates off of the manufactures data.

I'm calling out that specific data, in most cases, grossly misleading and at the very least contradictory.

Edited June 19, 20232 yr by TheMafiaMan - NotMuffinMan

22 hours ago, Tinbum said:

It should be figure 2. It's to do with the cable lengths to each individual battery from the inverter. This way they are all the same so batteries should charge and discharge more evenly.

Thanks @Tinbum the cable lengths are identical so it should then be as per fig 2 - very confusing the manuals vary and add or remove info for the same version battery 3-4 months apart 

@TheMafiaMan - NotMuffinMan

I am now very curious about what is happening on your system, can you send me the chart of this day:



But can you add the Battery Voltage and Battery Current to the chart. If those parameters are not available, go to the chart under the Device menu in Solarman. That chart has more parameters available.



Remove Production and Consumption. I only need Batt Voltage, Batt Current and SOC. That should tell us what is going on. Please also copy the whole day. I need to see the full crash and charging parts until SOC = 100%. Can you also take a screenshot where you hover where SOC=100% and the battery current is as low as possible. I want to see what the Batt Voltage is in that case.

I need to see the values like so: