On-grid domestic battery storage

joefizz

EVandPV said:

I've actually been having other issues with the new battery.

I've been monitoring them with the Batterview app and one morning I got up to find it was down to 0% SOC while the other 2 were still at 50%(ish).

The voltage was indicating that it should've been about the same SOC as the other 2.

So I contacted Pylontech and have been trying to correct the issue but I can't now get it to charge to more than 87% SOC or 51v even though the SOH is at 100% according to the Lux web portal.

I'm giving it one more grid charge tonight before sending it back.

Try going back to where you were with the original batteries and get them up to 100% again on both levels (you could try both individually but probably not necessary)

Disconnect those and then set it up just with the new battery and give it maybe a day or two of working and then get it back up to 100%

When all level, connect the three of them together and try using different ones as 'master' battery for comms, try favouring the new one first.

Its a long iterative process to go back and retry but give it a go, most of us have plenty of time on our hands at the minute ;-)

Its most likely a comms issue between the batteries, particularly if they have different firmware. I dont know offhand but its likely new firmware will be backwards compatible with comms (although they _should_ be using the same protocol but having worked in the area years ago its amazing what people think they can do with protocols) but the older batteries may not be fully forwards compatible.

EVandPV

joefizz said:

Its a long iterative process to go back and retry but give it a go, most of us have plenty of time on our hands at the minute ;-)

I'm classed as a "key worker" so don't quite have as much time as I'd like.
Appreciate the advice though, thanks Joe.
I've given the new battery (I say new but it's 2nd hand dated 2017) 3 full cycles but it just won't get above 87% for some reason. I agree it's possibly BMS related though as the cells all look good in Batteryview and are all balanced. 
My 2 original batts are US2000 Plus and the new one is US2000B, the main difference other than the firmware version seems to be the max DOD which is 80% on the B and 90% on the Plus batts. According to Pylontech, these 2 versions of Pylons are totally compatible and you can  mix and match them no problem.
I hooked them all up together again last night and grid charged them in the hope that the 2 good batts might pull the other one back in line but no such luck.
They're discharging at the moment and and seem to be behaving so I'll see what happens when I charge them again tonight during the Octopus Go period (we've had very little pv today due to cloud).
The problem batt is number 3 in this screenshot.

joefizz

EVandPV said:

I've given the new battery (I say new but it's 2nd hand dated 2017) 3 full cycles but it just won't get above 87% for some reason. I agree it's possibly BMS related though as the cells all look good in Batteryview and are all balanced. 

Is the 87% max with just it connected solely to the inverter?

EVandPV

joefizz said:

EVandPV said:

I've given the new battery (I say new but it's 2nd hand dated 2017) 3 full cycles but it just won't get above 87% for some reason. I agree it's possibly BMS related though as the cells all look good in Batteryview and are all balanced. 

Is the 87% max with just it connected solely to the inverter?

Either on it's own or connected to the other 2. 
I'd be surprised if a battery less than 3 years old could lose THAT much capacity unless it was treated really badly.

joefizz

EVandPV said:

Either on it's own or connected to the other 2. 
I'd be surprised if a battery less than 3 years old could lose THAT much capacity unless it was treated really badly.

I have been saying that since I started posting in this forum. It is possible, particularly if doing 2 daily cycles (solar plus charging from mains at night) and was the only battery connected. Its only 13% after all and 5 year warranty at 80% without the 2 daily cycles would point in that general direction.

Depends on the description really, its not that out of limits but if it was 90% Id say it would be closer to the mark. Id let it run for a week or so (depends on what your return conditions are) to see if it gets up around that and would lay off the double charging if you can to give it a chance to balance out overnight.

EVandPV

joefizz said:

EVandPV said:

Either on it's own or connected to the other 2. 
I'd be surprised if a battery less than 3 years old could lose THAT much capacity unless it was treated really badly.

I have been saying that since I started posting in this forum. It is possible, particularly if doing 2 daily cycles (solar plus charging from mains at night) and was the only battery connected. Its only 13% after all and 5 year warranty at 80% without the 2 daily cycles would point in that general direction.

Depends on the description really, its not that out of limits but if it was 90% Id say it would be closer to the mark. Id let it run for a week or so (depends on what your return conditions are) to see if it gets up around that and would lay off the double charging if you can to give it a chance to balance out overnight.

Yeah, I was thinking maybe 90-95% would have been nearer the mark at that age and I could live with low to mid 90's.
It's odd that the inverter is reporting 100% state of health too.
I'm only grid charging at the moment as it's been cloudy and we've had little spare pv.
I have another week to send them back and to be fair, the retailer has been pretty helpful offering a replacement or full refund.
Don't think I could justify the cost of another brand new Pylon.

Edit: Down to 41% at the moment so at least the problem battery hasn't crashed to 0% again as it did the other day when the other 2 reached 50%.

EVandPV

Well, bit of improvement this morning at 90%. 
The one change I made yesterday was increasing the discharge cut-off from 20% to 25% as the new battery was getting pulled just under 20% by the other 2 so I'm wondering if this made the difference ??
Still don't get why it didn't recover when charged on it's own though ??
The weather forecast looks a bit rubbish again tomorrow so I'll probably do another grid charge again tonight and see how it goes.

joefizz

EVandPV said:

Well, bit of improvement this morning at 90%. 
The one change I made yesterday was increasing the discharge cut-off from 20% to 25% as the new battery was getting pulled just under 20% by the other 2 so I'm wondering if this made the difference ??
Still don't get why it didn't recover when charged on it's own though ??
The weather forecast looks a bit rubbish again tomorrow so I'll probably do another grid charge again tonight and see how it goes.

It could have made all the difference. BMSs are complicated but sometimes have really simple assumptions in the algorithms, its like the comments I made about batteries in 7-10 years time getting to around 50% usable where the BMS will spend all its time shifting charge around so accelerating the demise. With one battery it will just sort that out and get it to its optimum and then when connected to others will try to balance them all out. It will then depend on the assumptions the bms makes (like for instance all batteries being the same, installed at the same time etc etc) so will shift charge around and end up the algorithm does more harm than good (to use a melodramatic turn of phrase!).

Better to go back to first principles each time and get each individual battery sorted out then connect up to others. You might fight that doing this once or twice a year is beneficial, particularly after a low charge regime or period of high activity (thinking spring and autumn here myself).

I have a refresh charger here for my AA batteries, which I occasionally use, it works by going through I think 3 full charge discharge cycles to work out the maximum mAh left in each battery. I then write these values on the side of the battery and keep the sets of 4 matched up so they are all close to each other.

So might be worth cycling each individual battery from supposed max to the cut off a few times just to 'refresh' each one.

Just on the inverter state of health, it will depend on what it classifies as state of health. I mentioned it a few times over on the electric car thread, what it says is the 'state of health' is different to the state of charge which is different to the maximum available capacity which is different to the maximum useable capacity. Its not quite the same to compare it to mobile phone batteries but thats something a lot of people will be familiar with, It might say 100% charged but can drop very quickly and then at 22% charge the phone turns off. You could be seeing something similar with the change in the cut off percentage.

Its only really when you have your batteries installed in your system that you can make any judgement call on it. Once you have it all stabilised again over the summer, thats where the extra battery will help even out the  loads and reducing the cycles on each individual battery, so if you only drop to say 40% overall you are keeping all the batteries in the optimum charge band and also keeping the bms 'right' as it were.

chamelion

Hi folks,
I've managed to get some partial integration with solcast, if anyone feels like experimenting and want some tips!

What I have so far:
* Updates of actuals sent to solcast, so that they can tweak the predictions for my site
* Cumulative solar estimation in kwh for the next day. At the moment, this is sending me a telegram notice daily at 8pm, and I can use a telegram bot I have to update the variables in octolux such as required state of charge. This last step is manual.

What I don't yet have
* Building the automated logic to have octolux adjust variables such as required SOC, cheap_charge etc based on the next day solar prediction and energy rates. I have all the data flowing in automatically, just haven't had the chance to map out every scenario so as to take actions from this data. 

I likely won't do the latter just yet - but if you are a smaller energy consumer then this step is a looooot easier as you have fewer scenarios. In my case with my house baseline at over 2.5kwh, at the moment it doesn't reaaaally make sense to automate the last bit, as everything generated is consumed by me or the batteries. 

buglawton

There was a lot of publicity about used Nissan Leaf batteries being re-purposed as on grid storage. How come no consumer level products have resulted? Should be a lot cheaper than a Powerwall. Thinking of a way to keep the fridge and gas CH running if there are power interruptions during the upcoming Zombie Apocalypse.

Edit: Looks like there is something, Nissan X-Storage circa £3-£4k. Its not clear if this uses a new or used battery. No reviews anywhere so not sure if it's real or marketing vapour. If there was a way of abstracting a decent 240v supply from your parked Leaf, I guess that would also be a good solution for occasional power cuts.