Real world battery performance

[Saul]

Thanks for all the thoughts and opinions.

@Chloe - I first noticed something about 2 or 3 days after installation, The 2 batteries were performing at about 70% of expected capacity. I informed my installer and in consultation with Hubble, he tried to balance the 2 batteries but concluded that the one battery was faulty and hence returned it. This is the one that they say might be the BMS - I will get some feedback tomorrow hopefully and let you all know how it goes. 

Since that one was removed, my remaining battery is running at about 86%-90% of expected capacity. One of the reasons I started the post was to try gauge what performance other guys were getting as this was obviously low but I was unsure of how accurate the Sunsynk/Solarman measure was. I wanted to make sure that when my other battery was re-installed I knew what was reasonable to expect.

@Achmat, @Tariq (with Pylontechs) are getting pretty much 100%, @Kill_switchis getting above 100% with same setup as me and @leshen has indicated that he is also getting 100% on his AM2 installations (although measured with RIOT - I don't know if he has checked this on the Sunsynk or Solarman). @ThatGuyhas indicated that he is aware of a test unit that has underperformed. So, as it stands, it appears that my remaining battery is also underperforming, although not as dramatically as the 2 combined. I'm also not sure how the BMS of the remaining battery might have been affected after the other battery was removed but from the feedback so far, there does not appear to be any reason why I should not expect 100% on both batteries. I would expect this to become less over time, but certainly not on a new battery.

I have no issue with Hubble - all products can have faults at some stage, but it is how these are resolved that is important. If Hubble are able to resolve the issues with both batteries I will happily let you know and vice versa.

[Saul]

Hi @ThatGuy

I was hoping to be further down the resolution path before providing feedback. This is where I stand at the moment (please note that all communication with Hubble has been done by my installer):

- Hubble agreed to swap out the battery that they initially thought might just be the BMS. This is obviously a positive. However, they said the new battery should be ready early last week, but then on Wednesday said it had been delayed until early this week (but based on other threads I've seen, this seems optimistic). They did not confirm what the problem was with the initial battery.

- I will hopefully be in a position by the end of this week to provide feedback on the new battery performance (providing I have it reinstalled in the next day or two - which as mentioned, seems unlikely now).

- My existing battery continues to run at about 88% of expected capacity. I have asked my installer to convey this to Hubble but I'm not sure how strongly he has made this point to them. Once the new battery has been installed, we will see if this affects the existing battery (I'm not sure if the BMS has been affected by the old misbehaving battery). I know @HubbleLithium check this forum from time to time, so I've included them on this response so that they are aware that I have a concern (as do others) that their batteries are running below the expected capacity. 

So, my battery was returned to Hubble on 21 April, which means I have been without an item I paid around R25k for, for almost a month. I have moved from mildly annoyed to annoyed. Needing to tiptoe around my limited battery capacity with the new bout of loadshedding has not made me happier. As I mentioned before, I can accept that problems arise with any product, and even that there are issues around global supply of components, but the lack of clear communication and expectations management is disappointing. 

I will provide further feedback when appropriate.

 

 

[Saul]

@Brani @ThatGuy 

So, my new battery was ready on Tuesday as promised. My installer could only collect and install today. He told me when he arrived that Hubble had confirmed that the old faulty battery apparently had an inactive cell, which is why they replaced it. So, a definite positive there as they provided an entirely new battery -  only downside was the time to replace - as mentioned before, I have some sympathy given current circumstances, but felt the communication could have been better. Of course, the faulty cell is also slightly concerning, but apparently this is very rare.

As it was only installed today, I have had limited time to assess new performance. However, on limited data (using the method I described before), the 2 batteries I have look like they are performing at about 1kwH per 10%. This is better but still at only about 91% of expected performance, main caveat being the limited time to assess performance and reduce the effects of rounding.

I will monitor this more closely over the next few days to get a better handle on it and report back.

[Saul]

Update: 

I have monitored my system over the past few days with the replaced battery and I am still getting about 1kwH per 10% sometimes fractionally more. So I'd say about 92% of label capacity. This is mildly disappointing but probably only because other members have shown better performance with both their Hubbles and other brands. Prior to knowing that, I would have written it off as efficiency losses or Sunsynk measuring inaccuracy. If there are other Hubble owners out there getting similar performance, please let me know!  I'm not going to lose any sleep about it though but will watch it to see that it doesn't deteriorate (or at least, not quickly). Of course, if anyone at @HubbleLithiumhas any thoughts/comments on this, I would appreciate hearing them.

In terms of the overall post sales service from Hubble - I would probably say that it was on average positive. I never dealt with them directly, but their willingness to replace the faulty battery entirely was appreciated. There was a delay in the process as mentioned before, which was not great, but I am willing to accept that there were mitigating circumstances. I have not had any experience with other solar-related brands so I cannot compare Hubble to them.

 

 

[HubbleLithium]

Hi Saul,

Thank you for the information and writeup.

How are you measuring the battery? Are you measuring what the inverter gives you in output or you measuring what the battery gives you in KWh output?

All inverters have an efficiency rating. Most of them now say "Maximum" efficiency, however on average they lose 10~15% in efficiency when converting DC to AC. I would suggest measuring direct from the BMS to see what the battery is pushing out. The only way is to plug in a RIOT Cloudlink. It will graph it direct to the cloud and samples every 10 seconds.

Batteries design capacity is rated by what the battery can provide in kwh at their terminal output, not the inverter.

Below is a RIOT directly connected to the BMS and a inverter direct and sampling both at the source for 100% accurate sampling. We can also remotely see each cell level voltage on the cloud for the Hubble and cell balancing drifts etc.

You can see the RED line in the inverter data feeding power back into the grid.

[Saul]

@HubbleLithium

Thanks for the reply. As mentioned before, I am using the inverter to measure - the Sunsynk has fields which show "Battery in" and "Battery out" in kwH which effectively capture kwH used for charging/discharging. I do not know whether, for example, the "Battery out" measures the kwH before or after conversion to AC  and would have been happy to write off the lower than expected kwH reading as a loss through inefficiency of conversion.

However, other members of the forum, both those using Hubbles and those using other brands (all with the same inverter), have shown readings at 100% and even above (compared to mine of about 90-92%), which suggests that the value the inverter shows is before any efficiency loss. The most likely explanation to me is that my batteries are still underperforming slightly, which is mildly disappointing for brand new batteries. Whilst I appreciate the RIOT may be more accurate, I do not think it is reasonable to purchase it just to check this especially as I am able to compare using the same method with other forum members.

[Boerseun]

On 2021/05/27 at 10:11 PM, Saul said:

@HubbleLithium

Thanks for the reply. As mentioned before, I am using the inverter to measure - the Sunsynk has fields which show "Battery in" and "Battery out" in kwH which effectively capture kwH used for charging/discharging. I do not know whether, for example, the "Battery out" measures the kwH before or after conversion to AC  and would have been happy to write off the lower than expected kwH reading as a loss through inefficiency of conversion.

However, other members of the forum, both those using Hubbles and those using other brands (all with the same inverter), have shown readings at 100% and even above (compared to mine of about 90-92%), which suggests that the value the inverter shows is before any efficiency loss. The most likely explanation to me is that my batteries are still underperforming slightly, which is mildly disappointing for brand new batteries. Whilst I appreciate the RIOT may be more accurate, I do not think it is reasonable to purchase it just to check this especially as I am able to compare using the same method with other forum members.

Hi @Saul Seems very strange I have 2 x Hubble AM2 Batteries with a theoretical Capacity of 11KW or 11000 watt if you look at this screenshot of solar man you will note 

my battery was at 46% Capacity which is theoretically 5060 Watt left in the battery My system Recharged 6700 watt into the battery you can see it clearly on the screenshot 

this means either my battery has a larger capacity than 11 000 watt if you add 5060 to 6700 watt then we get 11760 watt However I must admit when my batteries were brand spanking new I only got 10 000 watt out of them the problem is each installation has its own factors that can affect efficiency and so on your inverter may be slightly less efficient or your batteries need more time to "run In" whatever the case may be I am very pleased with how my batteries have been performing for the past 3 Months sofar.

[Saul]

Thanks for sharing @Boerseun.

Since my one battery was replaced, my readings have been consistently at about 91%, which is equivalent to the 10000Wh you were originally getting. This is on discharging - when charging the implied capacity is pretty much at the expected 11kwH.  This is perhaps consistent with your experience - perhaps you are still getting less on discharging? Or maybe the higher than 11kwH on charging you are getting is consistent with 11kwH on discharge. I'd be interested in knowing this. 

As mentioned before, I would have previously written this off as efficiency loss - about 10% ,except for the fact that others with the same inverter were getting the full capacity on discharge (both with Hubbles and other batteries).  But I accept your point that there may be other factors which can affect this number. Overall, I am very happy with my installation - I can live with what may be slight underperformance or a less than optimal setup - I will just watch the battery performance closely to see if it deteriorates (or improves)!

 

 

[Boerseun]

14 hours ago, Saul said:

Thanks for sharing @Boerseun.

Since my one battery was replaced, my readings have been consistently at about 91%, which is equivalent to the 10000Wh you were originally getting. This is on discharging - when charging the implied capacity is pretty much at the expected 11kwH.  This is perhaps consistent with your experience - perhaps you are still getting less on discharging? Or maybe the higher than 11kwH on charging you are getting is consistent with 11kwH on discharge. I'd be interested in knowing this. 

As mentioned before, I would have previously written this off as efficiency loss - about 10% ,except for the fact that others with the same inverter were getting the full capacity on discharge (both with Hubbles and other batteries).  But I accept your point that there may be other factors which can affect this number. Overall, I am very happy with my installation - I can live with what may be slight underperformance or a less than optimal setup - I will just watch the battery performance closely to see if it deteriorates (or improves)!

 

 

I agree with you however I am getting a RIOT in place soon that will give me more accurate readings and a little more in depth information that one can analyze the true performance and losses. Give your system some time it may improve as mine did. 

[FixAMess]

10 minutes ago, Boerseun said:

I agree with you however I am getting a RIOT in place soon that will give me more accurate readings

A bit of a side question, Will RIOT work on any battery, I'd like to get accurate in/out power on my pylontechs?

[Boerseun]

4 minutes ago, FixAMess said:

A bit of a side question, Will RIOT work on any battery, I'd like to get accurate in/out power on my pylontechs?

The Guys at Riot are in constant state of development as far as I know you will get charge discharge rates capacity and so on from Pylontech and some basic Info not as detailed as the Hubble apparantly Pylontech is very protective on their code for their BMS etc the version of Riot Iam getting comes with a very accurate 3 phase meter that can be used to remotely monitor loads in stand alone mode and the Sunsynks MODBUS code will be added too it soon so I will be able to get far more accurate readings from the inverter and battery. So yes it will work with Pylontech but with more limited information on the battery. 

[MyproD]

Don't mean to bring up an old thread but I came across this and decided to have a look at the data.

I have 1x AM-2 battery on a Sunsynk 5kw inverter.

It looks like I use 13-14% per 0.5kWh

So 28% for 1kWh. 

Seems way out or is my maths out. 

This is all very new to me. 

 

@HubbleLithium Could you have a look and confirm? System is around a month old. 

 

 

 

 

 

Edited January 4, 2022 by MyproD

[Saul]

Best to measure it over a longer period -and a bigger SOC change -  small decimals can have a big impact, especially with smaller numbers. Looking at your numbers if you measure from 63% SOC to 51%, that's a 12% change for 0.6kwH. This converts to an effective capacity of 5kwH (1/.12 * 0.6), which is what I saw on mine and am still seeing 6 months later. So looks OK, but as mentioned, best to test it over a longer period.

[MyproD]

Thanks for the quick Reply Saul. 

I am new to all this.

I had a look at the plant and I see its charged to 99% from 25% and taken in 3.1kWh

 

Edited January 4, 2022 by MyproD

[Saul]

I'm not familiar with that view so it's hard to comment - but the calculation would suggest an implied capacity of 4.2kwH, which is not great, but makes me suspicious of those numbers given the earlier numbers you supplied looked much better.

If you can supply the same table as you did in your original post (but over a longer period), we should be able to get a clearer picture. 

[MyproD]

Hi Saul

Thanks!

So I charged the battery to 100% and set the inverter to use from 00.00 this morning. 

Looks like I am getting 3.0kWh from the battery from 100% to 20% 

[MyproD]

So, quick update on this.

I took the battery in as something wasn't right. i was getting around 2.8kWh from 100% to 20%.

Took it in yesterday and they called me today.

The BMS was showing 73% even though I unplugged it at 100% I have the logs and offered it to them.

Looks like the cells were not balanced and the inverter wasn't pushing enough power through to balance them once it got to 100% on the inverter side.

They said they tested a 2000 watt load for over two hours and it passed. 

Will see what happens...

[maxxis]

@MyproD

What was the final outcome? 

I'm now seeing a drop of about 10% in SOC by morning compared to where I was in early January. 

Normally at around 35 to 40% by 07:00 and now seeing 25% +-

Nothing changed at all in terms of consumption patterns.