jonescg's home battery storage system

[T1 Terry]

Not good news at all Chris, but as one door closes another will open, just make sure you get through it before it slams shut
Any talk over that side of the island about setting up a battery plant using the lithium we are digging out of the ground? Or aren't we actually digging it up yet or refining it here yet?

Very unsure about the next moves with our small business, so we might be joining the unemployment line as well. With all the China dramas evolving we can't be certain of a supply of cells into the future. At the moment work is very quiet because no one can get into the state with all the borders being closed.
Maybe you could start doing your own off grid and semi off grid systems using the knowledge you have gained setting up your system.

T1 Terry

[jonescg]

Well the guys at Electro.Aero are keen to have me on board, so I'm certainly interested in that. In the meantime i have a redundancy to live off for a while as I get my house solar and storage in order.

After finishing a big battery build for @Sutho (which I might put photos up for...) I finally had time to look at the AVASS cells and BMS.

I've opted for the EV-Power single wire loop BMS with a hysteresis control unit: https://www.ev-power.com.au/product/bcu-micro-08b/ and cell-top modules.

20200809_112816.jpg (187.23 KiB) Viewed 1213 times

I figure the BMS really ought to be simple, and the inverter should only need some voltage parameters like Vmax (full, stop charging) and Vmin (empty, stop discharging). The BMS will manage voltages outside the inverter's range, so if something bad were to happen and the inverter didn't stop when it was supposed to, the BMS would step in as a line of last resort.

So here's the setup:
Basically the single wire loop is monitored by the BCU which is in the grey box. It has a reset switch on it, but as I found out it can't be a momentary switch - it works more like a latching relay. The Gigavac contactor is powered by the BCU, and if the loop goes open circuit, the BCU waits 10 seconds before cutting power to the contactor. THis is to deal with any hysteresis due to brief heavy loads, but I can't see many of them hanging off the inverter.

20200815_163058.jpg (210.58 KiB) Viewed 1213 times

The 12 V power to run the BCU is supplied by a 48 V to 12 V DC/DC converter (the silver thing) and it's powered by the switched side of the main contactor. I simply don't trust this thing to not flatten the battery once the thing opens at <Vmin. So it needs to be 'jump-started'. That's what the big resistor is for. You can use a jumper but the capacitor spark is pretty wild. Once it's bridged you turn the switch on. If the loop is closed (all good) the contactor will close again.

20200815_163035.jpg (259.37 KiB) Viewed 1213 times

And finally when it comes time to assemble the battery proper, I'll do so on this little step and slide the whole show backwards into the box. It's a tight fit, but everything just fits.
Hope to install it and start work on the system once the rain stops.

20200815_165330.jpg (209.4 KiB) Viewed 1208 times

[jonescg]

Today was the big day - I spent most of the day assembling the battery, bolting the terminations down, carefully mounting the contactor etc. Satisfied with how that went, I pushed it back into the enclosure where it fit perfectly.

...only to realise I forgot to solder the BMS normally closed loop

And that's when the fun started. And by fun I mean an additional two hours of back pain as I had to completely disassemble the battery and re-assemble it. See, there is a 5 mm drop on the inside of the enclosure. Easy enough to get the front row of cells over the rise when pulling it back out, but the rear row got stuck on it. SO I had to pull everything apart and start again.

20200818_135803.jpg (261.4 KiB) Viewed 1170 times

20200818_135813.jpg (211.84 KiB) Viewed 1170 times

Anyway, it's done now, and looks great. I tightened up the compression clamps a little more this time, in the hope that if I ever need to pull it out again, it won't get stuck on the ledge.

HSB done.jpg (217.85 KiB) Viewed 1170 times

Currently charging using a 48 V charger to see that the BMS works (if it needs to work).

[jonescg]

Yep, it works

However I have one of my cells bypassing furiously when the pack is at 55 V, and it's the one with the higher DCIR. I might set the Goodwe to a Vmax of 56 V (which is about 55 V at the cell terminals) to prevent any nuisance tripping. And in the meantime try and top up the cells either side with a power supply or something.

[Cowardlyduck]

Looking great mate. I recently did a similar thing when finally finishing off my powerwall...forgot to properly mount the BMU temp sensor and only realised after battening it all up.

Will you be putting any labels on it? I haven't done mine yet, but according to AS_NZS_5139_2019 it needs a bunch of labelling to comply.

Cheers

[jonescg]

Yeah totally, gotta get my ES, PV and battery type stickers, as well as a copy of the MSDS for LiFePO4 cells. And three affidavits from a QC confirming the isolators were no more than 3 m away...

[Cowardlyduck]

And three affidavits from a QC confirming the isolators were no more than 3 m away...

Huh? That's news to me...I mean I knew the isolator needed to be located like that, but why the affidavits?

Cheers

[jonescg]

Sorry I didn't use sarcasm font.

[Cowardlyduck]

Sorry I didn't use sarcasm font.

That's a relief, got me worried there for a min thinking I actually need to get my setup inspected by someone....not that it's in anyway non-compliant, but still.

I'll be interested in what stickers you get and where from as I haven't done mine yet.

Cheers

[brendon_m]

The sparky I got to install my inverter left me with a sticker pack with all the stickers that I need for the labelling on my system. Looks like it was just a generic battery storage pack. I've seen them from solar panels wholesale suppliers like one stop warehouse
https://www.onestopwarehouse.com.au/inv ... uct_detail

[Cowardlyduck]

The sparky I got to install my inverter left me with a sticker pack with all the stickers that I need for the labelling on my system. Looks like it was just a generic battery storage pack. I've seen them from solar panels wholesale suppliers like one stop warehouse
https://www.onestopwarehouse.com.au/inv ... uct_detail

Thanks...didn't know such a thing existed. I was thinking I would have to get mine custom made.
Your post led me to search and find this gem:
https://aussiesolarlabels.com.au/solar-labels

Hopefully that helps everyone out.

Cheers

[francisco.shi]

Let us know how the inverter behaves. I have a Goodwe inverter and some times it sells electricity from the battery to the grid at night. I wonder if there is some setting that is not right. It generally happens when a load turns off. It seems the reaction time is very slow.

[jonescg]

Let us know how the inverter behaves. I have a Goodwe inverter and some times it sells electricity from the battery to the grid at night. I wonder if there is some setting that is not right. It generally happens when a load turns off. It seems the reaction time is very slow.

Yes I understand that it's got about a minute of lag between one action and another. Part of the requirements with the network operator - make all changes smoothly and carefully.
Also, it needs about 100 W at all times to power its brains. That 100 W comes from the grid, which makes sense - of all of the power sources it's hooked up to, the grid is there 24/7. I'm also going to be export limited to 3 kW (Western Power rules) but I doubt I'll be exporting very much at the end of a sunny day, especially not if I'm charging an EV.

My plan is to power the lights and house GPOs on the backup circuit and all other appliances on the normal grid connected circuits. So things like A/C, the shed, hot water system and EV charger will be on their own.

[brendon_m]

Also the backup circuit may not work properly (in backup mode) with a less than 150W load

[francisco.shi]

There is a 500w continuous load. It often sells for min.

Screenshot_20200820-135618_SEMS Portal.jpg (273.81 KiB) Viewed 1079 times

You can see here it starts taking energy out of the battery. I have seen it sell up to 2kwh out of the battery during times when there is no solar input.
Any positive in the yellow line is power sold. Any positive on the purple line is power out of the battery.

[Cowardlyduck]

Double check that's not just an artefact of the inverter's reporting mechanism.
My Solax shows similar spikes and overshoots:

Solax Output.JPG (40.18 KiB) Viewed 1068 times

However when I actually watch the batteries BMS (bluetooth) or the inverter display panel, I can see that it's not as generalised as my the reporting shows. In my case the Solax cloud reporting mechanism averages and reports every 5min, then tries to smooth the curve between points. This leads to it sometimes showing grid output at the same time as battery consumption, however when I check it via the BMS it isn't actually doing that.

Cheers

[francisco.shi]

When I check on the AGL website for the power sold it matches the reporting from the inverter.

Screenshot_20200820-181126_AGL Energy.jpg (264.17 KiB) Viewed 1067 times

Here is an example of it selling .1kwh over a 30min period. This picture is from the utility meter. If I look I can find some times up to .5kwh being sold over a 30min period at night. The panels generally don't make anything after about 4pm.

[Cowardlyduck]

TBH, that makes me want to avoid Goodwe inverters...but I don't have much experience with these yet. Just my Solax and Zever before that.

Cheers

[acmotor]

jonescg,
Looks like it is coming together nicely.
A few thoughts...
Did you have to order the cell top boards from ev-power set for 3.6V max for OV rather than 4V ? Or is the OV programmable ?
How do you plan to access the top of the cells in your cabinet to do manual voltage readings given the opting for a non cell voltage readout BMS ?

[nuggetgalore]

jonescg,

How do you plan to access the top of the cells in your cabinet to do manual voltage readings given the opting for a non cell voltage readout BMS ?

I was wondering that too.That cabinet is a tight fit.
I am glad that I can easily access the tops of my Avass cells to check from time to time that the BMS and shunts work as per design.

[jonescg]

jonescg,
Looks like it is coming together nicely.
A few thoughts...
Did you have to order the cell top boards from ev-power set for 3.6V max for OV rather than 4V ? Or is the OV programmable ?
How do you plan to access the top of the cells in your cabinet to do manual voltage readings given the opting for a non cell voltage readout BMS ?

Rod has changed the cell top boards so they start to pass current at 3.55 V, and the over voltage trip appears to happen at 3.9 V. Despite the tight fit, I can see the LEDs on all modules from the front of the cabinet, and if I need to check voltages I can still fit a hand in to place an alligator clip on each terminal.
My plan is to use the custom settings on the Goodwe so it regards the battery full voltage as 56 V and empty is 47 V. The inverter itself does coulomb counting which is good enough for me.

[francisco.shi]

I have never been able to get the inverter to do anything useful with the battery. No matter what settings I put the inverter stops at 44v. The top voltage seems to work ok. But the estimation of state of charge has nothing to do with reality. It looks more of a reflection of current going in or out of the battery. I have seen it repost 96% SOC when the battery is at 45v but outputting 3kw and the SOC go down as the battery charges.
Here is an example.

Screenshot_20200821-101214_SEMS Portal.jpg (251.12 KiB) Viewed 1045 times

The more current that goes into the battery the lower the SOC.
If you figure out how it works let me know.
I did try to get it to communicate using CAN bus but when I changed the battery type to one with BMS there was nothing going out of the CAN bus wires.

[jonescg]

Hopefully I can just use the default settings - http://www.solazone.com.au/wp-content/u ... ressed.pdf
I think it regresses to default if it doesn't get a BMS signal. I intend on setting the charge voltage to 57 V, float voltage back to 55 V and the Vmin at 47 V. It's basically treating the battery as a lead acid one with some softer settings.

[brendon_m]

In the manual it says if they lose the bms signal it assumes a SoC of 95%.

[jonescg]

Well the good news is that I've managed to balance it to the point where the BMS doesn't trip out while charging. For some reason, the Vmax of the charger I have here is set to 59.1 V, which is high to say the least. Still, it dropped the current right back to 0.8 A while maintaining 59 V, and while some cells were bypassing furiously, they never exceeded 3.9 V. So The lowest cell at this point was 3.55 V, so practically full. I might try a couple more charge-discharge cycles to ensure all red lights come on at the same time, as the charge current tapers down.

2020-08-22.jpg (256.24 KiB) Viewed 876 times

Also put the railing up on my patio - room enough to place a beer can. However I'm more pleased with how easy it was to bring 3.6 m lengths of timber home in the Ioniq...