OffGridInTheCity Build

[Korishan]

Yeah, totally get that. I'm now at a balancing point myself. I have just enough panels, storage, and controllers to be good for through most of the night. I get 2 more controllers I can get better efficiency. But I also need to get bigger storage and more panels to truly make it through the night w/o being on grid. But then again, it won't last longer than a day w/o help.
And that's with 7kW of Panels and 16kWh of storage.

 

[Fukuokian]

Another good reason to regularly monitor your batteries... https://www.bbc.com/news/articles/c059gqml15go

 

[OffGridInTheCity]

Another good reason to regularly monitor your batteries... https://www.bbc.com/news/articles/c059gqml15go

Unless they hear you going in/out to do the monitoring! I'm going to have to find a new location for my powerwall

 

[OffGridInTheCity]

Self-discharging pack update....

Agh. In addition to pack #54 (used medical cells), pack #3 (new? ebay cells) has decided to self-discharge, even more aggressively. I hooked up an OPUS / paralleled these 2 packs to push charge into them simultaneously as I need to 'prop these up' every week these days.

It's interesting that even though paralleled, they show significantly different voltages - I would have thought they'd stay in sync better.
After charging yesterday, pack #3 = 3.68v but pack #54 = 3.74v per Batrium longmons....

It's been a sweet 7+ year / 2,400cycle run with only occasional balance tune-ups but now self-discharge seems to be accelerating on these 2 packs.

It makes me uneasy as it could be these packs are particularly affected (weak, IR)? and the bulk of the others are good for a while yet OR... I could be looking at 1/2 my powerwall descending into rapid self-discharge packs as they near 2,700 cycles over the next year. I just don't know.

 

[Korishan]

This is where active monitoring really shines. After 7years, pretty good run. You might be starting to look at system wide degradation, though. Time to start looking at alternative storage for replacement. At least you see the problem sneaking up instead of just Boom!!! no power available.

 

[kje]

Self-discharging pack update....

Agh. In addition to pack #54 (used medical cells), pack #3 (new? ebay cells) has decided to self-discharge, even more aggressively. I hooked up an OPUS / paralleled these 2 packs to push charge into them simultaneously as I need to 'prop these up' every week these days.

It's interesting that even though paralleled, they show significantly different voltages - I would have thought they'd stay in sync better.
After charging yesterday, pack #3 = 3.68v but pack #54 = 3.74v per Batrium longmons....
View attachment 33662

It's been a sweet 7+ year / 2,400cycle run with only occasional balance tune-ups but now self-discharge seems to be accelerating on these 2 packs.

It makes me uneasy as it could be these packs are particularly affected (weak, IR)? and the bulk of the others are good for a while yet OR... I could be looking at 1/2 my powerwall descending into rapid self-discharge packs as they near 2,700 cycles over the next year. I just don't know.

7 years!! Amazing. I would be proud.

What about trying my method? Desolder all cells in the cellpack in question and let sit for one or two days, then check the voltage on every cell. You should discover the self discharging cell. It could be only one cell dragging the rest down.

 

[OffGridInTheCity]

7 years!! Amazing. I would be proud.

Was hoping for 15yrs....

What about trying my method? Desolder all cells in the cellpack in question and let sit for one or two days, then check the voltage on every cell. You should discover the self discharging cell. It could be only one cell dragging the rest down.

My other projects (replaced flooring and painting the house) are finishing up and 'battery maint' is next.

I'm hoping to follow you're work and recover these packs and hope the other 54 packs in batteries #1 thru #4 are still good for another 1,000 cycles. But who knows.

 

[OffGridInTheCity]

This is where active monitoring really shines. After 7years, pretty good run.

Agreed! Batrium has been key for me to have confidence and has highlighted the bad packs in time to do something.

You might be starting to look at system wide degradation, though.

As you can see in the chart in post above this one, due the intervening years between battery builds, even if the 40% (54kwh) of the powerwall >2,000 cycles 'goes bad' rapidly I still have 60% (81kwh) of the powerwall at <2,000 cycles. 81kwh is enough to consume my PV if I expand the voltage operating range.

Time to start looking at alternative storage for replacement. At least you see the problem sneaking up instead of just Boom!!! no power available.

I haven't visualized a plan to move from Lithium-ion to LifePo4 and trying to mix.... not sure I want to go down that road.

Have been toying with the idea of Nissan Leaf cells - but 60-66% capacity compared to new specs puts me off and not sure they'd be any better than building with newish 18650 cells.

 

[Korishan]

I don't like the mismatched voltage range issue

The overall voltage range is almost the same for the system voltage

LifePo4 has a flat voltage curve and I operate (inverter on/off) by voltage rather than SOC

You can still do voltage for the cut off/on. There's 2 distinct curves that will start to change voltage values and that's when you trigger. At 3.45V (55.2) the cells will start to rise pretty quickly. So you can use that as a marker. With the bottom, it's around 2.8V (44.8V). These can vary a little bit depending on manufacturer and age of the cells.

not sure how to keep things balanced

Once they are top balanced, they will very rarely go out of balance. But what you can do is once a month you have a secondary charger that would be configured to push the cells higher than what your standard chargers do. So for instance, if you only go to 55.5V (3.46), then you set your balancing charger to 56.5V (3.53V) ((or even to 57V/3.56V)) and allow the BMS to do it's job. Because they are very slow to get out of wack, it wouldn't take long for them to balance off.
For mine, my top charging voltage is right towards the top of the range, 58.2V (3.63). But, I also only have 16kWh currently and need to store as much as possible before evening. Another 16kWh is on order, so that'll make a difference on the cap. I'll lower them down to 56.0V (3.50V) when they are put into production.
My bottom voltages are currently 46.4V (2.9). I could lower that a little and still be fine tbh. But under my loads it starts to bounce too much. With 32kWh possible, 44.8V would be fine as it wouldn't drop so fast with the added cells to hold it.

This is my current curves. The spike in Orange at 21hrs is when I was using the AirFrier. But you can see at about 22hrs the current drops, so it switched over to grid. But the discharge curve drops off pretty quickly after 46V, almost vertical. Of course, I'm also pulling close to 45A right then. Double capacity, split the load in half per battery bank, and that curve won't be so steep.
China needs to get on the ball and get that battery shipped! It's been 3 weeks already!

 

[OffGridInTheCity]

@Korishan - appreciate the detailed comment...

The overall voltage range is almost the same for the system voltage

Agreed. For me, it comes down to 1) voltage based inverter on/off rather than SOC based and 2) keeping things balanced.

You can still do voltage for the cut off/on. There's 2 distinct curves that will start to change voltage values and that's when you trigger. At 3.45V (55.2) the cells will start to rise pretty quickly. So you can use that as a marker. With the bottom, it's around 2.8V (44.8V). These can vary a little bit depending on manufacturer and age of the cells.

At the upper end.....
I currently let the 18650 go to 56.0v (4.00v/ion-cell) and I could go a little higher - up to 56.7v(4.05v/ion-cell). 56v/16 = 3.5v/lifepo4-cell and 56.7v/16 = 3.54v/lifepo4-cell

So you think top balance at 3.50v-3.54v/lifepo4-cell would be good enough?

At the lower end ....
I perceive 44.8v as too low for my 18650 packs. The lithium-ion pack voltages start diverging (>100mv) at 47v (3.36v/cell = 18650 knee) and would be uncomfortably wild by the time I hit 44.8v (3.20v/cell)

I wonder - do you think I could stick with a minimum of 48v inverter cut-off? (48v/16 = 3.00v/LifePo4-cell) or is this too hi for lifepo4 / would they get out of sync SOC wise with each other or 18650.

Once they are top balanced, they will very rarely go out of balance. But what you can do is once a month you have a secondary charger that would be configured to push the cells higher than what your standard chargers do. So for instance, if you only go to 55.5V (3.46), then you set your balancing charger to 56.5V (3.53V) ((or even to 57V/3.56V)) and allow the BMS to do it's job. Because they are very slow to get out of wack, it wouldn't take long for them to balance off.

I would want to use the Batrium longmons for balance - including lifepo4 cells. I could alter my operating cycle to force 56.0v -> 56.7v range as needed for top balance.

I would miss auto-level but I'm not clear how the middle of the of charge/discharge curve would work in practice between the 2 x chemistries when paralleled. I could switch to using 2 different Batrium cores - one for each type of chemistry and parallel the 2 groups.

I think it's the low cut-off 'by voltage' question above that is my biggest concern.

And in the background, I like 18650 cell level fusing as a 'safety measure' (at least in my mind) and of course Lifepo4 bricks are much more powerfull 'invididual units' compared to 18650 cells and one can't fuse at the smaller level.

 

[Korishan]

So you think top balance at 3.50v-3.54v/lifepo4-cell would be good enough?

Yeah, that's fine for LFPs. Once they are top balanced, they don't get out of wack very easily. Just periodically push them up to 3.6 to make sure they are top balanced, which wouldn't take very long.

do you think I could stick with a minimum of 48v inverter cut-off? (48v/16 = 3.00v/LifePo4-cell)

that's right at the top of the last/bottom knee. So it should still be fine. Right after that the voltage starts to drop pretty steadily until it hits the cliff around 45V (2.8).

 

[OffGridInTheCity]

This is the last year for the 30% tax credit for those of us in the US.

Went round and round in my thinking on 'next battery build' and ultimately decided to continue with 18650 cells as they just work smoothly for my situation / comfortable with them / setup to deal with them.

Ordered enough for 2 x 13.5kwh batteries - either to expand my powerwall or take a couple of batteries off to a remote property if we go that route.

Battery clearing house has these samsung 26J 2600mah Pink cells at 44c/cell (pretty good price) but shipping was astronomical - e.g. $988 on 780 modem packs. Don't understand the shipping - is this real FedEx prices or just a way to inflate the sales price?

(NEW SHIPMENT) Excellent Condition Samsung 26J 18650 3.6v 2.6Ah cells in green modem packs (4 cells per pack)

Great Condition Samsung 26J 18650 3.6v 2.6Ah cells in green modem packs (4 cells per pack) We just got a fresh shipment of these modem packs in, factory sealed. These are testing at excellent capacity. These cells have a 95/95 guarantee: "Specifically we guarantee that at least 95% of the cells...

www.batteryclearinghouse.com

Battery Clearning house says these are 95/95 - meaning 95% have 95% original capacity and I believe @Wolf got some a year or so back and didn't complain - so I'm hopeful they're as advertised.

Overall it comes out to 71.26c/cell. 1400cells * 71.26c = $998 + holders =$121 + longmons = $441 + other = ~$200 = $1760 for 13.5kwh or...
~$130/kwh.

 

[Korishan]

but shipping was astronomical - e.g. $988 on 780 modem packs. Don't understand the shipping - is this real FedEx prices or just a way to inflate the sales price?

Just for the single item it was $50s/h. A little steep for the few count

Overall it comes out to 71.26c/cell. 1400cells * 71.26c = $998 + holders =$121 + longmons = $441 + other = ~$200 = $1760 for 13.5kwh or...
~$130/kwh.

I ordered 16 Goshin LFP's for just under $1000 shipped for 16kWh. They come with the JK BMS (w/ touchscreen) and flexible busbars. Granted, that's no case. But I know how to work wood or weld up a case if needed.

 

[OffGridInTheCity]

Battery clearing house has these samsung 26J 2600mah Pink cells at 44c/cell (pretty good price) but shipping was astronomical - e.g. $988 on 780 modem packs. Don't understand the shipping - is this real FedEx prices or just a way to inflate the sales price?

Woke up to some good news - a refund on part of the shipping from Battery Clearinghouse!

Nice to find that Battery Clearinghouse is still a legitimate source of cells / business. This put's shipping of 3120 cells at $418 instead of $988 - more reasonable.

Brings things down from $130/kwh to ~$110/kwh.

 

[OffGridInTheCity]

WOW - this seems to be one of those rare - everything's going my way day. Another $247 off - bringing $110/kwh down to $101/kwh
Obviously, this only adds to confirmation of Battery Clearinghouse as a legitimate source of 2nd hand cells.

 

[Korishan]

They may have really awesome customer support, problem is for me their pricing is just too high. I just did a look through and their pricings are 2-3x or more than if I just bought directly from Alibabba (which they have US, or nearby, warehouses as well). I just can't see paying $400 for 3kWh of storage when I can get 8kWh from Ali for the same price, and those are brand new cells and includes a JK BMS.

 

[OffGridInTheCity]

They may have really awesome customer support, problem is for me their pricing is just too high. I just did a look through and their pricings are 2-3x or more than if I just bought directly from Alibabba (which they have US, or nearby, warehouses as well). I just can't see paying $400 for 3kWh of storage when I can get 8kWh from Ali for the same price, and those are brand new cells and includes a JK BMS.

You're thinking LifePo4? 18650 at 44c each for 95% original capacity / 2600mah cells seemed OK to me.

 

[Korishan]

Yeah, I was looking at LFP's. At this point in my venture, I'm not really interested in possible firecrackers

 

[OffGridInTheCity]

Yesterday was surprise good, today is surprise nasty from Batrium. Turns out I have to pay additional $ (not disclosed by Batrium when I ordered) on US import fees to get my order. Its ~10% more onto the order.

It's typical of Batrium (in my experience) to not care about their customers but not even mentioning the extra $ for entry to US on checkout. I've ordered 20 times over the last 7 years and it's never come up before - so it's a nasty surprise.

I know Trump has done tarrifs etc, and that's not my complaint. It's that Batrium didn't bother to estimate (can't be that hard for a competent business) or even mention this on checkout (**Import $ extra) so the real price is hidden! I found the lack of support on Batrium website after DHL notification that I owed money that this was my problem - not theirs.

Good hardware but they clearly (in my personal opinion) don't value customers - too bad I'm captive at this point.

 

[OffGridInTheCity]

Time to look at failing packs #3 and #54.

This year, packs #3 and #54 started loosing ground during charge consistently - causing a 100mv imbalance over 3-4 weeks. I've been limping along, manually charging them up each month but the time has come to see what's happening.

My packs are 260ah each, but made up of 2 x independent 130ah sub-packs. Using my iCharger X8, I did a capacity test on recent/extra/good/new packs to establish a charge/discharge capacity baseline.

The baseline packs (never cycled NCR18650A green cells) reported 120.2ah between 4.070v and 3.000v discharge at 13a / 0.1C. The iCharger X8 only goes up to 4.1v charging for some reason and after settling, the sub-pack is left at ~4.07v to start the discharge test. Thus I get 120ah instead of 130ah (sum of individual cells) during the discharge test.

Yesterday, pulled pack #3 and sub-pack #3a reported 80.3ah discharging from 4.07s -> 3.0v at 13a / 0.14C. That's ~66% of original capacity after 2,473 cycles / 7.5yrs. This pack is from my 1st battery deployed in May 2018. Cells from ebike packs from BatteryHookup and similar eBay ebike packs and presumably 'new'ish at the time. Tested 98-100% original capacity. I don't see any damage or broken fuse wires or corrosion.

Samsung ICR18650-22P - https://secondlifestorage.com/index.php?threads/samsung-icr18650-22p-m-cell-specifications.8117/
LGDAMF11865 - https://secondlifestorage.com/index.php?threads/lg-lgdamf11865-cell-specifications.2492/

-------
Next up, 1) test sub-pack #3b, 2) look for sub-pack level self-discharge and 3) individual cell discharge -> repair.