Hi there

I have a rack of 4 X Pylontech US2000 Plus batteries that were installed in Feb 2018.

They are connected to  2 X Mecer SOL-I-AX-5M inverters thru  48V battery fuses.

Solar panel strings run via 6 fuses from the roof then to the PV BUS Bar, 2 X 145v.

Backup power is a Honda 20i generator set to a 2amp input in the Master inverter.

Program 26 on the inverter is set to 53.2 volts for bulk charging.

In January of this year we arrived back from our Christmas/New Year break to notice the 2 batteries in the middle of the stack of 4 were not holding charge overnight.

A visual inspection of them and I could see swelling of both units. The unit at the top and bottom are operating normally.

We had to wait for the company we bought them from to open in the middle of January before we could log the fault.

They were eventually sent to Segen Solar for assessment and were rejected for overcharging. However these dates on the the "Event Data" spreadsheet correspond with the time we were away and most likely when the batteries started to fail.

They also occur when charging would not be possible given my property is north facing from a solar perspective and the generator is only used on overcast days.

Could these voltage spikes be as a result of the battery cells failing and if so how could this be proved?

I have attached the Event Data sheets if anyone would care to take a look and offer any advice.

Any and all thoughts and advice are greatly appreciated.

Stuart

 

Event Data 0159.xlsx Event Data 4669.xlsx

The BMS is supposed to disconnect the battery when it goes overvoltage. This definitely should be covered by the warranty.

Looking at the logs, it seems that the cell failure is evident before the overvoltage events - the lowest and highest cell voltages are very far apart at the start of the log and ~45% SoC, and should already cause a cell overvolt disconnect.

And there goes my theory, for the time being, that Pylon has a robust / reliable BMS. 

A BMS's mandate is to protect the chemistry. Over voltage and it should stay in protection until manual intervention. Short circuit and it should stay in protection until manual intervention. Overcurrent charge or discharge, and the bms should apply a time window of forgiveness before lifting the protection, but  after  many events in a short period of time and it should stay in protection. Over temperature, and the bms should stay in protection.  Cell min/max delta voltage over threshold and it should stay in protection. 

So you will want to start by asking Segen why the bms did not protect. Swelling is real bad. And share it with us, we also want to learn wether we can trust a battery or not. 

44 minutes ago, OzzyMozzy said:

Hi there

I have a rack of 4 X Pylontech US2000 Plus batteries that were installed in Feb 2018.

They are connected to  2 X Mecer SOL-I-AX-5M inverters thru  48V battery fuses.

Solar panel strings run via 6 fuses from the roof then to the PV BUS Bar, 2 X 145v.

Backup power is a Honda 20i generator set to a 2amp input in the Master inverter.

Program 26 on the inverter is set to 53.2 volts for bulk charging.

In January of this year we arrived back from our Christmas/New Year break to notice the 2 batteries in the middle of the stack of 4 were not holding charge overnight.

A visual inspection of them and I could see swelling of both units. The unit at the top and bottom are operating normally.

We had to wait for the company we bought them from to open in the middle of January before we could log the fault.

They were eventually sent to Segen Solar for assessment and were rejected for overcharging. However these dates on the the "Event Data" spreadsheet correspond with the time we were away and most likely when the batteries started to fail.

They also occur when charging would not be possible given my property is north facing from a solar perspective and the generator is only used on overcast days.

Could these voltage spikes be as a result of the battery cells failing and if so how could this be proved?

I have attached the Event Data sheets if anyone would care to take a look and offer any advice.

Any and all thoughts and advice are greatly appreciated.

Stuart

 

Event Data 0159.xlsx 139.23 kB · 4 downloads Event Data 4669.xlsx 137.54 kB · 1 download

Try contacting pylontech directly and hear what they say

54 minutes ago, BritishRacingGreen said:

And there goes my theory, for the time being, that Pylon has a robust / reliable BMS. 

A BMS's mandate is to protect the chemistry. Over voltage and it should stay in protection until manual intervention. Short circuit and it should stay in protection until manual intervention. Overcurrent charge or discharge, and the bms should apply a time window of forgiveness before lifting the protection, but  after  many events in a short period of time and it should stay in protection. Over temperature, and the bms should stay in protection.  Cell min/max delta voltage over threshold and it should stay in protection. 

So you will want to start by asking Segen why the bms did not protect. Swelling is real bad. And share it with us, we also want to learn wether we can trust a battery or not. 

Slighty off topic, but while we are questioning the integrity of BMSes, DIY builders please note that a standalone active balancer can and will destroy cells when it experience certain fault modes. I have seen 3 x 280Ah LFP cells damages as a result of balancer partial short circuit.  The partial short circuit depletes the cell over a period of time, many hours later its 0V . In this case, yes, the bms can see the huge delta and go in protection, but it cannot protect the downstream havoc. 

It must be noted that the balancer lies outside the protection region of the bms, as it is connected downstream from the BMS, towards the cells. 

EDIT: What would have helped was if we had a monitoring system that also reads protection status from bms. A number of push notifications in succession  could have saved the day for England. Even a loud buzzer.  Thats why I rate the work @Sc00bs and other are doing for HomeAssistant etc. 

Edited June 6, 20233 yr by BritishRacingGreen

Thanks @BritishRacingGreen but you give me too much credit, all the heavy lifting is done by the guys like @kellerza @Bloubul7 @slipx @stephanjoubert @Gary Waterworth and all the other guys that I have forgotten, I just try to join the dots 🙂 .

 

 

3 hours ago, BritishRacingGreen said:

EDIT: What would have helped was if we had a monitoring system that also reads protection status from bms. A number of push notifications in succession  could have saved the day for England. Even a loud buzzer.

It's pretty easy to use something like a Teensy and use it as a CAN sniffer. I use one to change the CAN messages to the inverter but also to publish readings to MQQT. You can then read in Node Red and make announcements over Alexa, publish to Pushover, e mail or Telegram for example.

Edited June 6, 20233 yr by Tinbum

Thanks for all the feedback, I have used the above observations in a response to Segen and will forward it once received.

I took the extended warranty on these just in case but perhaps there is more money in sales than honouring legitimate warranty claims given the current situation here?

1 hour ago, OzzyMozzy said:

Thanks for all the feedback, I have used the above observations in a response to Segen and will forward it once received.

I took the extended warranty on these just in case but perhaps there is more money in sales than honouring legitimate warranty claims given the current situation here?

Please come back with the reply from @SEGEN SOLAR

 

Funny enough i have had a issue with 2 US2000 batteries. I had to shut the system down because i knew i didn't have enough power to last the night and wanted to preserve the charge for the next morning to use. Shutdown at 34% SoC and startup 7 hours later showed 2 batteries at 4% SoC. Logs showed before shutdown the charge was at 34% and 4% the next entry hours later.

Contacting Pylontech they had me download and send them the logs. which i done over 2 days as they took so long for the first two batteries to download and bad weather meant i wanted to conserve power. They have said "We actively hope to solve your problem, but two of your batteries have been recorded as frequently overcharged and I have found that your inverter MPPsolar 7248Max is not in our compatibility list, these are signs of improper use and will not be covered by the warranty."

I responded staiting that the inverters are in fact Voltronic axpect rebadged which are on their supported list and that any over voltage should have been handled by the BMS to prevent any damage. The really annoying part of this though the over voltage in my case was done 3 years ago before i even brought the batteries and not a ongoing problem (no over volt in the 2.5 years i have had the units and only 4 other events in this period).

There response "Our battery's passive protection mechanism, MOSFET, is only activated when the battery parameters are extremely abnormal (Short circuits, high currents, high voltages, etc. that can damage the circuit and may cause safety events). The battery BMS only uploads its own voltage, recommend charge and discharge current  to the inverter,, and the time to stop chargind/discharging is controlled by the inverter. Therefore, if the battery is still being charged despite reaching the alarm value, the voltage at the inverter must be higher than 54V (otherwise there would be no voltage difference between the battery and the external power source and it therefore would not be charged) as well as inverter does not react to the data uploaded by the battery according to the preset logic."

So to me the high voltage (54v) that the batteries received over a 17 day period in 2020 were not a high enough value for their BMS to worry about so meaning a very low chance of a problem occurring, causing them to say the warranty is not valid is just wrong.

Will be watching to find out how they treat you.

I trust segen for nothing.

I have had 2 invertors warranty rejected. I will never buy anything that is sourced from them.

A friend of mine had this system installed at his office the other week and is very keen to see how Segen handles my warranty claim.

 

13 minutes ago, OzzyMozzy said:

A friend of mine had this system installed at his office the other week and is very keen to see how Segen handles my warranty claim.

 

MASSIVE bank there

6 hours ago, Webbie said:

There response "Our battery's passive protection mechanism, MOSFET, is only activated when the battery parameters are extremely abnormal (Short circuits, high currents, high voltages, etc. that can damage the circuit and may cause safety events). The battery BMS only uploads its own voltage, recommend charge and discharge current  to the inverter,, and the time to stop chargind/discharging is controlled by the inverter. Therefore, if the battery is still being charged despite reaching the alarm value, the voltage at the inverter must be higher than 54V (otherwise there would be no voltage difference between the battery and the external power source and it therefore would not be charged) as well as inverter does not react to the data uploaded by the battery according to the preset logic."

 

So to me the high voltage (54v) that the batteries received over a 17 day period in 2020 were not a high enough value for their BMS to worry about so meaning a very low chance of a problem occurring, causing them to say the warranty is not valid is just wrong.

Will be watching to find out how they treat you.

So if you have received an overvoltage alarm ever, you are screwed for life, no warranty

5 hours ago, OzzyMozzy said:

A friend of mine had this system installed at his office the other week and is very keen to see how Segen handles my warranty claim.

 

Should be very worried if he/she had bought from Segen. They seem to make up their own rules when it comes to warranty 

Edited June 7, 20233 yr by hoohloc

52 minutes ago, Dieter said:

So if you have received an overvoltage alarm ever, you are screwed for life, no warranty

not really, it depends on where you bought your Pylons from

Just to update you all This is the reply i got today from Pylontech

"Let me clarify that it is not because your battery has reported a fault that your 10 year warranty has disappeared, it is because the 54V has exceeded the maximum charge voltage indicated in the battery manual and this is considered improper use and therefore we cannot offer a warranty for subsequent failures, customers have the right to decide how to use their batteries, but our warranty can only be offered to batteries that have been used properly."

So searching for the warranty and manual i received with the batteries i go as i don't recall any of that.

Looking through the manual i got with the batteries there is no maximum charge voltage of 54V listed. In 2.2 Specifications on page 4 it does state charge voltage 52.5 ~ 53.5 (I take this as a suggested charging range) but certainly no mention of a maximum charge of 54 volts listed here but maybe it is implied 53.5V is the maximum?

On page 22, under trouble shooting 5.2 c) High Voltage: If charging voltage above 54V, battery protection will turn on. Solution Check whether voltage is too high or not, if it is, to change the settings on power supply side.

So the battery protection should have shut the batteries down if the voltage is over 54V and a manual reset should have been done to correct the problem is the way i see this. The fact this error occurred two or three times in a row, 4 seconds apart from one another, says no battery protection happened to make a person aware there was a issue to look into at all, as i would expect to happen, but because of this my warranty is voided according to them. This is the very reason why i think a warranty claim on the defective product is required more than ever.

I Don't know about you OzzyMozzy but i sure feel like i threw away 13K on these batteries. I Wish you better luck than what I have seen.

9 minutes ago, Webbie said:

I Don't know about you OzzyMozzy but i sure feel like i threw away 13K on these batteries. I Wish you better luck than what I have seen.

😮 That's a good price, even if it's 13K each. Well... at current pricing that I can see. How old are these batteries?

One gets the impression that battery warranties are a sales gimmic and are virtually impossible to claim against.

I have 4 pylontech us3000 batteries. They all experienced overvoltage, logged, not shutdowns/warnings, when being added to a stack. It takes them a day or 2 to stabilize and then they're all good.

If one looks at the wording of the warranties, there are so many reasons for them to reject the warranty, I don't see them ever honoring their warranty. I'd say any life you get out of a battery after 5 yrs is a bonus..Sad really.

Solar really only took off in SA on a large scale in 2018, so we are now seeing the first 5yr+ battery performance and warranty issues coming to the fore. I'm sure we will hear many more of these issues going forward.

I see a trend in inverter manufacturers insisting on selling their brand inverter+battery as a package, no other batteries allowed to be attached to their inverter.

Agree with @FixAMess, though as a disclaimer, I haven't been through any warranty claims. Having read the "terms" though, things like having to register the installation within a week, take pictures, use the right fuses and settings (which as others have pointed out, is difficult when there are 4-5 different datasheets and manuals saying different things for a single battery), and only then, they might, at their discretion, honor a "warranty".

It feels like many of these have enough escape hatches built in to avoid any claims if they deem fit, but honor claims soley to avoid bad publicity when it suits them.

Edited June 8, 20233 yr by reapster
Typo

Hi OzzyMossy

Off topic and I unfortunately cannot contribute more info on your issue. I would however appreciate to know how you got hold of the two event data sheets.

I have two UP5000 batteries linked to Victron MP2 and Cerbo GX (via the BMS-CAN cable) and want to get the same data for my batteries.

Feedback will be much appreciated.

The interesting thing from those event lists is that the actual over-voltage events are all logged after you reported the fault.

And in all likelihood, they were caused by the faulty batteries (rather than the other way around).

Inverters (even cheap ones) are usually very good at maintaining voltage limits.  Generally, the only time they will go over is when there is an instant reduction in load (like when the BMS on one battery in a parallel bank disconnects due to cell imbalance).

 

On 2023/06/06 at 9:59 AM, OzzyMozzy said:

I have attached the Event Data sheets if anyone would care to take a look and offer any advice.

Any and all thoughts and advice are greatly appreciated.

As I understand things they are claiming that the high voltages reported by the battery BMSes on the two batteries that failed were the cause of the failure, rather than a consequence of it.

I think there's an easy way to prove or disprove this claim. I'm assuming that all four batteries were connected in parallel, and therefore the voltage across all of them should be the same. If you were to extract the logs from the other two batteries that are still presumably fine, then that should tell you whether the inverter did in fact raise the charge voltage beyond 54V.

If there is something wrong with the inverter and it did attempt to charge the batteries beyond 54V for an extended period of time then Pylontech may be within their rights to refuse the claim. If it only happened once or twice, then I would say they need to honour it.

I am following this thread with interest, becase I am also sitting on a couple of Pylontech batteries that were accidentally overcharged beyond 54V during commissioning. It only happened once, due to misbehaviour of the inverter after a firmware update, and the BMS disconnected the batteries. However, I am fearful that Pylontech will refuse to honour the warranty if the batteries were to fail a couple of years down the road for an unrelated reason, simply due to this one incident that is in the event logs.

41 minutes ago, PierreJ said:

As I understand things they are claiming that the high voltages reported by the battery BMSes on the two batteries that failed were the cause of the failure, rather than a consequence of it.

I think there's an easy way to prove or disprove this claim. I'm assuming that all four batteries were connected in parallel, and therefore the voltage across all of them should be the same. If you were to extract the logs from the other two batteries that are still presumably fine, then that should tell you whether the inverter did in fact raise the charge voltage beyond 54V.

If there is something wrong with the inverter and it did attempt to charge the batteries beyond 54V for an extended period of time then Pylontech may be within their rights to refuse the claim. If it only happened once or twice, then I would say they need to honour it.

I am following this thread with interest, becase I am also sitting on a couple of Pylontech batteries that were accidentally overcharged beyond 54V during commissioning. It only happened once, due to misbehaviour of the inverter after a firmware update, and the BMS disconnected the batteries. However, I am fearful that Pylontech will refuse to honour the warranty if the batteries were to fail a couple of years down the road for an unrelated reason, simply due to this one incident that is in the event logs.

Might be worth updating your firmware to the latest versions as that will wipe all the event and cycle data and then you can start from a clean slate. I have been updating my firmware regularly so all that info is lost every time so they wont be able to use it against you/me in the future anyway. 

Something worth noting is a friend of mine had some issues with one of his 3000C modules throwing high voltage errors and disconnecting in the first week after commisioning (turned out to be a BMS issue reporting incorrect cell voltage of 4,1V) . The battery was sent in for warranty replacement only to be sent back saying there is nothing wrong with it.

But you can be sure all those events are stored in the logs for better or worse? If you install new firmware and you have many months of clean data i cant see how they would refuse a warranty,then again what if you install new firmware on a battery and it breaks shortly after ? would be interesting to see how they handle that .

Hi CarloB 

I'm so new to this forum I'm not even sure how to tag you correctly! The event data sheets were provided by Segen and as I'm off-grid and ignorant when it comes to all things Solar I can't offer any advice as to how to extract them.

Perhaps another another member could advise?

I would also appreciate any advise on how to do this as well as updating the firmware on Pylontech batteries.