OffGridInTheCity Build

ALL NEW - Battery Finder Search for 12/24/36/48v or by capacity www.batteryfinder.net
New & used Batteries, Solar, accessories, LifePo4 cells & more 5% Coupon "Powerwalls" www.batteryhookup.com
OffGridInTheCity said:
Can adding 10% capacity to a pack can 'cover up' a slow self-discharge? - is this even theoretically possible? Meaning the extra capacity takes in charge and 'covers' the self-discharge during the discharge part of a cycle.
Click to expand...

Maybe it can 'cover up' 10% of the voltage sag of the pack? I believe the self-discharger cell will continue to drag the pack down and I imagine it'll have negative effect on the cells around in the long run. I would rather try to solve the problem by changing the cell(s) in question.

No more than one cell is needed to have voltage sag on a pack. I found out this when I made my battery #1 and discovered voltage sag on the 120p pack.

You could try this approach for this first pack and hope the other 13 packs will continue for another year or two. It doesn't take too much effort and time. 😊

I like this trial and error, investigating and problem-solving method. 👍☀️
 
I agree with @kje again. I think supplementing the sagging pack with additional cells will help to mask the problem. This is particularly so if the self-dischargers (assuming that is the issue) are very slow. It's also going to depend on things you have limited control over -- like depth of discharge on any given day, recharging capacity from your solar on any given day, etc. If your entire battery is regularly cycling between 70% and 90% SOC, for instance, you may not have enough voltage drop across the entire battery on a day to really notice much sag in a single pack. If it's cloudy for a couple of days, or you have very high use for some reason, you might be more likely to see it as the overall battery discharges further and the weak pack has no opportunity to recharge or "hide" in the data. Just thinking out loud. An easy test might be to disconnect your array for a day and just draw on the battery. I know your data tracking is good enough to have some granularity on individual pack voltages across the span of 24+ hours. It might show you what's going on. It would also be a fairly easy experiment to do "before and after" adding the additional parallel cells, to evaluate with at least some control how much of a difference that solution makes.

Cheers, John
 
OffGridInTheCity said:
Just don't know and don't have a clear 'vision' in my mind
Click to expand...
You and me both are looking with our minds eye through a fog.
I think you may remember my Battery#3 with which I had an issue similar as yours. https://secondlifestorage.com/index.php?threads/wolfs-powirwall.7804/post-91056
In this case I replaced the pack with a "new" replacement and the result was excellent.
As you know I do not use "helper cells" to adjust the Ah of each pack as my testing is very rigorous and I use a strict IR and mAh go no go rule on each of my cells. I found no serious issues with the "droopy" pack after testing each cell and comparing them to there original values. Also no self dischargers where found
I would have used these cells again. So in my case I am also stumped.

Coming out of winter and just now starting to see close to 100% charged, this is what my voltage differences from batter1 to battery 4 look like.
No adjustment or anything has been done to these batteries or for that matter individual packs.
Battery#1 varies between 10 mv to 60 mv
Battery#2 varies between 20 mv to 80 mv
Battery#3 varies between 20 mv to 60 mv
Battery#4 varies between 7 mv to 30 mv
Which ends up with an average of around 55mv for the last 7 days. I can live with that.
Now we all would like to have our batteries be like battery#4, but why is the v diff so flat on this battery?
The only thing that is different on this battery#4 is all the LGGBM261865 cells were produced between July, August and November of 2018.
Whereas the LGFBM261865 cells in battery#3 had productions dates from November 2017 to March 2019 could this
17 month difference be the answer. IDK, but it is worth investigating.
I'll leave it at that.

Wolf

1744170623049.png
 
Wolf said:
You and me both are looking with our minds eye through a fog.
I think you may remember my Battery#3 with which I had an issue similar as yours. https://secondlifestorage.com/index.php?threads/wolfs-powirwall.7804/post-91056
In this case I replaced the pack with a "new" replacement and the result was excellent.
Click to expand...
Yes I remember which is why I wasn't jumping in with both feet to dissemble the sagging pack. Also, I felt bad for you and was secretly relieved it hadn't happened to me - till now at at least :)

Even with the 10p additional capacity, pack #54 is already falling behind during the CC->CV part of the charge curve. It's not discharging as the lowest but falls behind as the battery nears full charge of 4.0v/pack. For example, after today it peaked at 3.94v while the next lowest groups are 3.95v and 3.96v - that's a 30-40mv lag after just a few days!
1744419732478.png


Will give it a little more time but....

I've decided I might as well start building a replacement pack. I have a set of old (4-6yrs) 2nd hand cells (Pink Samsung ICR18650 2600mah and Blue LGDAA31865 2200mah) that tested ~75-85% original capacity 5 yrs ago and a set of 'never cycled' 100% capacity (Green NCR18650A 3000mah) from the most recent battery build. The NCR18650A is the same brand as battery #4 cells but the Samsung/LG packs play well with the other packs and would be a good use to try to use them up.

What to do....

Rather than fall into analysis paralysis I'm going to use @DG98 car analogy and think of this battery as 60% worn tires with 1 tire (bad pack) prematurely bald. Rather than put on a new tire I'll start with a replacement 'worn tire' using the older/lower-capacity cells so I have 'worn tires' all around. Remember that battery #4 started at ~86% original capacity and has 2,000+ cycles - so integrating a 'worn tire' pack may work and degrade gracefully with the rest.

I'll see if I can do some capacity tests and analysis of bad pack in battery #4 as part of this and maybe get some insight into what happened to #54 and try to get a sense of degradation after 2,200 cycles.
 
Last edited:
Started testing the 150 old cells to build a 'worn tire' pack.... Apparently I stored them at full charge and they all show 4.05v to 4.11v which got me excited; however, they're showing 8-12% capacity loss compared to the original tests. Worse, the OPUS IR test is showing 120mOhm to 200mOhm across the board. I know OPUS IR is fuzzy but my newer cells show 60mOhm so this is 2x-3x worse.

Sigh - These 'old' cells from the last 7 yrs have degraded too much and it's time to toss them.
1744485432739.png


On to using the remaining 'newer' cells to build a spare pack...
1744485694992.png


I've processed aprox 18,000 cells and this means I'll be out of cells for the 1st time since I started this powerwall - not sure how to feel about that.
 
Last edited:
OffGridInTheCity said:
Started testing the 150 old cells to build a 'worn tire' pack.... Apparently I stored them at full charge and they all show 4.05v to 4.11v which got me excited; however, they're showing 8-12% capacity loss compared to the original tests. Worse, the OPUS IR test is showing 120mOhm to 200mOhm across the board. I know OPUS IR is fuzzy but my newer cells show 60mOhm so this is 2x-3x worse.

Sigh - These 'old' cells from the last 7 yrs have degraded too much and it's time to toss them.
View attachment 33251

On to using the remaining 'newer' cells to build a spare pack...
View attachment 33252

I've processed aprox 18,000 cells and this means I'll be out of cells for the 1st time since I started this powerwall - not sure how to feel about that.
Click to expand...
Maybe it's time for a proper IR tester? 😊
 
Random Update: Batteries need attention!

I had a 7s7p DIY in an APC 1500 with Chargery BMS that I've used for years with no trouble. About 2 months ago, I unplugged the APC to clear out the room to do flooring. The other day I realized - it's been a couple of months, I should see what's up with that battery / disconnect the BMS.

Wow - battery ruined in just 2 months. Here's a reading of the parallel cells - notice #3 and #5 won't register. Note: #1 and #2 are >3.0v because I charged them up from 2.x. Then I tried to 'revive' #3 for several hours but it's 0v and continues rapid (seconds) self-discharge. I tried paralleling in a 4p pack to 'jump it' + hours on OPUS to no avail.
1749311694606.png

So much for the Chargery BMS protecting the battery! I love Chargery in design but they have been a nightmare all around.

It just highlights that unlike solar panels - batteries require attention! BMSs do not always do a good job! and in general, not good to leave them unattended for long unless you prep them for storage.

I have 3 more 7s7p for APC 1500(s) prepped and they sit OK at 3.7v'ish for years with no trouble.
 
Last edited:
Update on pack #54: Continuing to loose ground / needs ~7% touch up charge every month now.

I'm in the upper voltage range (3.8v <-> 4.0v) of my daily cycle this time of year. Up at this voltage range, pack #54 doesn't like to charge up from 3.8v -> 4.0v as fast as the other packs. Interestingly at the lower range 3.6v <-> 3.8v pack #54 seems to behave OK.

I continue to have no clue / don't understand. Is it possible there is higher charge resistance at 3.9v->4.0v range? or bad connection contributing to charge resistance at 3.9v or... is it possible there's actually 'cell memory' for some 18650?

In any case, this battery is in a difficult to reach location so I'm just adding 20ah of charge once-a-month and will get back to it this winter when all my other home projects are on hold due to weather. Hopefully it won't degrade very fast.
 
Last edited:
I think partial self-discharge from 4.2v to 4.05 is acceptable and even normal after a month of storage.
When i test cells i put them away fully charged at 4.20v for 4 weeks, If a cell self discharges below 4.00v then i will get rid of it.

Thinking of it, i have some cells by the tester that were stored before my mom died in november, i will let you know how they are doing.
 
  • Like
Reactions: kje
While the APC unit was disconnected, there was no charge current being produced to keep the battery charged. This allowed the BMS to slowly discharge the battery over time. It's also possible that the UPS itself was also slowly discharging the battery even though it was "turned off". There's always parasitic losses.

So yes, it would have been a good idea to disconnect the battery from the unit while it was in storage. This would have drastically reduced the possibility of the battery being damaged.

Considering a Chargery BMS was being used, there's a lot of extra circuitry to keep powered.
What's interesting is that 2 sections are not being read. Or is that be design?
 
italianuser said:
So, you're saying that it would have been better to disconnect the battery from the circuit after disconnecting the APC, right
Click to expand...
Yes! Or at least monitor closely (maybe check each week) and disconnect before anything went below 3.0v/cell.

Keep in mind I had it wired with Chargery BMS and relay (to shut off power to APC). Maybe Chargery shut the relay off but did not disconnect 'itself' and took cells all the way down. 2s of the 7s was at 0v and the other 5s were at 0.Xv - 2.Xv - so Chargery took or allowed the batter to destroy itself. Disspapointed in Chargery as well as myself.
 
Last edited:
Highest PV produced on the 15.2KW PV array to date: 92.6kwh for yesterday (Jul 14).
1752590007902.png


Here in Southern Oregon, we're in that time of year (Jul and Aug) where we routinely go over 100F hi / no clouds at all. Reached 105F hi the last couple of days and the AC (heat-pump) power load has grown to be just a little greater than max PV - so it's letting the system produce max PV without throttling due to max battery charge.

On the 13th the "IKwh: 80.6" (e.g. Inverter Kwh) is what it took in actual power to cool/run the house for 24hrs at 105F. The system/inverter is reporting ~83% efficient which means that 92.6kwh PV * .83 = 76.8kwh power available... not quite enough for 80.6kwh used by the house for 24hrs of Jul 13 cycle.
 
Last edited:
  • Like
Reactions: kje
OffGridInTheCity said:
Highest PV produced on the 15.2KW PV array to date: 92.6kwh for yesterday (Jul 14).
View attachment 33438

Here in Southern Oregon, we're in that time of year (Jul and Aug) where we routinely go over 100F hi / no clouds at all. Reached 105F hi the last couple of days and the AC (heat-pump) power load has grown to be just a little greater than max PV - so it's letting the system produce max PV without throttling due to max battery charge.

On the 13th the "IKwh: 80.6" (e.g. Inverter Kwh) is what it took in actual power to cool/run the house for 24hrs at 105F. The system/inverter is reporting ~83% efficient which means that 92.6kwh PV * .83 = 76.8kwh power available... not quite enough for 80.6kwh used by the house for 24hrs of Jul 13 cycle.
Click to expand...
Man I miss 100+ degrees. One of the last summers I was in Austin was 90 consecutive days over 100 with 60 of those over 105.
 
Random Post.....

We've had a run of >105F days and this has gradually exceeded daily PV production drawing down the powerwall. Just fooling around - I'm letting it go lower than normal to see if I can make it thru without grid-assist and it hightlighted some weak packs....

It's 7:19am as I post this, with a modest ~2100w load which is typical for early morning. A few minutes ago my >10mv alarm when off on the dashboard and it shows Pack #2 and #3 have started to 'drop rapidly' once they reached 3.42v/cell level.

These packs are right at the main + power take-off of the overall powerwall and these acks are among the oldest at 2374 cycles and.... but who knows what's happening. Curious that's its only these 2 out of the 140 packs overall.
1755095014609.png


Note: The dashboard updates every 5 minutes (130mv) vs Batrium which is 'live' (140mv) so the pack has dropped 10mv within the last 5 mins....

About 50minutes before PV goes positive and things start charging - we'll see :)
 
Last edited:
Pretty interesting. Considering they are the oldest, they are probably starting to loose capacity just from sheer usage. Where those before you were doing more IR scrutinizing?
 
Korishan said:
Where those before you were doing more IR scrutinizing?
Click to expand...
Yes! Never checked! These cells were from 'new' 20cell ebike batteries from BatteryHookup back in the day (2017) and was my 1st battery.
Notice that all of the 1st 14packs (1st battery with these ebike cells) don't balance that closely.

Operationally I get nervous to see this rapid drop after 3.41/pack to 3.38 in <hour - indicating the discharge knee - as this could very well drop <2.8v/pack in a few hours and get into 'risky' territory. This puts ~48v my lowest, unattended, operating floor.

2hrs of charging / 12.7kw of incoming PV and packs #2 & #3 are pretty much recovered. Powerwall back to 60mv max difference between packs.
1755105804191.png


Maybe I'll dive into this in winter (when I'm limited to indoor projects) along with tackling pack #54's self-discharge.
 
OffGridInTheCity said:
Operationally I get nervous to see this rapid drop after 3.41/pack to 3.38 in <hour - indicating the discharge knee - as this could very well drop <2.8v/pack in a few hours and get into 'risky' territory
Click to expand...
Yeah, totally agree. Might be time to pull those out and replace with more balanced cells, ones that have been scrutinized via IR stats as well.
 
Random post - HEAT = A LOT of Heat-Pump kwhs as the heat-pump get's strained/at-it's-limit.

When the summer hi(s) go from 80F to 107F, the power consumption goes up an extra 45kwh in 1 day on my 4-ton whole house heat-pump for 2600 sq ft home. I suspect 107F 'strains' the heat-pump causing inefficiency = sharp spike in power as at I know from past experience that at 115F the heat-pump can't keep the house at 72F. So 107F is near the max capability of the heat-pump.
1755620053379.png

This is what makes 100% off-grid / solar so challenging. One might be going along OK during normal situations but the extremes highlight cracks. My powerwall is big enough to cover the losses of PV vs consumption over a 3 day heat-wave but if 107F were to become the norm I'll need more PV to be 100% off-grid during summer.

Max PV Aug 14th at 92.5Kwh * 83% overall efficiency -> AC = ~77kwh max consumable power at these temps. In the spring I got a 96Kwh PV day and I attribute the temps + accumulated dust as the drop from 96 -> 92.

In 2019 when I had the heat-pump installed, I asked for 5-ton but the HVAC company refused with their typical BS of air flow or whatever. I'm sorry I backed down now as perhaps the power consumption would not spike at 107F if the unit was bigger/more-capble and thus stay in a better efficiency zone. Not to mention that 4-ton can't keep house at 72F in 115F - the house rises to 74F/75F'ish.
 
Last edited:
You must log in or register to reply here.
Back
Top