diy solar

diy solar

Weird overnight battery discharging

shavermcspud

Solar Enthusiast
Joined
Mar 12, 2020
Messages
567
I have a strange issue that I cant quite put my finger on, batteries are around six weeks old, 3x100Ah SLA in parallel, with a magic eye to check supposed charge status
initially I had the scc and inverter load connected to one of the batteries at the end of parallel connection until I read online that it should be taken from opposite ends of the parallel string, this doesn't seem to have made a difference

During the day it would appear the batteries are fully charging at 14.7v with plenty of current as it appears to tail off to float voltage and current around 12:30
Ive dismissed the SOC / battery capacity readout as this seems to be random at best.
After sun down with hardly any load the battery (ive disconnected the inverter to rule that out) voltage tails off and last night went down to 12.11v (50% soc?) this is with no load bar the SCC being connected to the batteries.
Ive got an ammeter / shunt in place which shows the SCC draws around 800Ma constant.

Ive either got a phantom load somewhere or my batteries are knackered already or I am reading this all wrong.

I have 10x 100 watt panels 2s * 5p, MPPT epever 40amp controller, 1000watt pure sine inverter (currently disconnected) I'm not running anything off the load terminals on the SCC
Screenshot_2020-04-11 Grafana - Solar.png
 
3x100Ah SLA in parallel,
Are these deep cycle batteries? Brand?

I think you need to get them charged up and disconnect everything. Measure the voltage of each battery (apart) to make sure that they are in fact charged. Then see if they self discharge quickly. Then put a small load on and see if any of them discharge quickly.

Sorry not much help, you need to isolate to identify the problem it seems.

Good luck. Check back in if you get results that are mystifying (or even if you find a problem).
 
  • Like
Reactions: Ped
Uhhh look at your inverter. When theyre going bad they start pulling loads like that.

EDIT: I see you ruled that out.
 
you say you have a 0.8A constant drain from the charge controller. From 20:00 to 6:00 (10 hours) this sums 8A. According to your green graphs let's say your battery capacity is 62,5% at 20:00 and 37% at 6:00.
IF, the Bank capacity is 300A, at 20:00 you have 187A remaining; and 111A at 6:00. This means the bank lost 76A overnight (at least between 20hs and 6hs). Please let me know if I'm doing something wrong, but this is what I understood:
- This is what can be observed when disconnecting the loads from the bank.
- In other words, you have [ Panels > SCC > shunt > Bank ].

If this is true, we have an inconsistency between the shunt and the graph. Right? Since 10 hours at 0.8A should be 8A. Against 76A.
I never tried had nor had a battery monitor at all (I'm pretty new to solar, so take that into consideration!) but I have a hunch that your battery monitor is speculating the battery capacity by measuring the bank voltage alone? What happens if you disconnect the solar panels AND the load for a whole day to see if the graph asymptotizes after a few hours.
Also, the monitor might have a depth of discharge function? Maybe it is set to 70% and the battery capacity is set to something smaller?

It's strange that the battery capacity and charging current do not mix well. If the bank capacity is at 100% the battery charging current should be very small, right? This is what makes me think that the battery monitor is measuring bank capacity according to voltage.

In any case, if I'm contradicted by more experienced users please disregard everything I said!
 
Reading the graph you can see the charge current drops with sunset. That tells you theyre not getting charged. At 1700 the solar day ends while its still pumping 30A.

Big giant no-no right there. The current going to battery should taper off before sunset ALWAYS.

Also 1kw should be producing 40A minimum this time of year. That controller is not big enough for that much solar. You need a 60A because summer output should be a hair over 50A .
 
looking closer we see the battery voltage drop to 13.8v right after noon, and the charge current ramp up to 30A at around the same time.

Thats not right. What is boost time set to? Should be 180 minutes at 14.6v with temp compensation. And with temp compensation you should be hitting higher than 14.7v unless its pretty warm outside. Float voltage should be 14.6v because youre not in storage and not floating so you dont want it kicking in at all. That tricks controller into being a 2 stage charger.
 
Hi all

Firstly I think I haver found one of the issues, one of the batteries although a SLA type has leaked, quite badly out of the casing at the top, not sure why this has happened but i have removed it from the system altogether, yesterday I charged the remaining two batteries with solar during the day and then put my automotive charger on until they were full, (my automotive charger went straight to float within 5 minutes indicating that the batteries were pretty much fully charged, i disconnected everything and left the batteries overnight, this morning they are reading 12.21 and 12.28 each. im thinking I have duff batteries here. the batteries are Shield MF27-1000 leisure battery link the date stamp says manufactured July 2019 and last charge sticker shows Nov 2019, maybe I have somehow ruined these during my use or a bad batch., altough looking back from date of install they havent ever been taken bellow 50% DOD

im going to look at all the other suggestions here and report back. the SCC is setup for the correct battery type, voltages etc as far as i can tell.

I just realised the above graph is from the wrong day, the spike at 30amp is due to my wife having her office equipment running that day.
 
Just as an update

I reconnected the two remaining batteries, the SCC and the inverter this morning at 6am, around sunrise, we have fairly decent sunshine at the moment , load according to the shunt/readout is between 650ma & 1.1amp, that includes the inverter on standby with no mains load and the raspbery pi running the graphing software

Upon connecting everything back up it was reading 12.26 volts, all these figures in the graph are from the SCC btw.
You can see the charge current increase and voltage increase, the capacity goes from 40 to 78 within an hour?? although I don't know how the SOC is calculated.

Historically the current will ramp up to 15 amps for a few hours and around 12 / 12.30 it will drop off to 2amp and stay there for the rest of the day until sundown. I'm fairly sure these batteries have had it but as to the cause I don't know.

solar 1.png

20200412_094518.jpg
 
Last edited:
Wont allow me to edit the above post.
But the batteries i reinstalled this morning that were at 12.2 volts are now showing they're at boost voltage and current already.
Sun has been constant all morning with zero clouds.

Charged started off about 7am and ramped up to 14a for 45 minutes from 9.30 to 10:15 and is now tailing off to 2amp @14.36 volts

confused.
solar 2.png

ive gone back to the USB cable and my pc to double check the settings with the provided software
solar 1.pngsolar 2.png

And just as typing this, it has gone to float already.
solar 3.png
 
Last edited:
Are these AGM? SLA can mean either. I am going to assume theyre flooded but sealed. IF theyre AGM, your voltage values are OK.

Ok so eq voltage is set to 14.6v....it needs to be at least 15.6v. It can need as high as 16.2v or so.

-Charge limit - 16v
-Boost and eq duration - set to highest. should be 180
-Boost voltage - 14.6v
-Float voltage - 14.4v

They need forced equalized for a couple days to be sure theyre shot or not. (Temporarily set boost voltage to 15.8v)

It works like this, when you discharge, an oxide film builds on the plates called sulfation. To get rid of it you must recharge to 100% asap. If you go days, weeks that oxide becomes permanent and you cant get rid of it. It blocks the chemical process and lowers the battery capacity. You will see extreme voltage drop offs and very fast climbing voltage reached under charging.

You should be able to get a warranty claim on them. Def do it.
 
Last edited:
Are these AGM? SLA can mean either. I am going to assume theyre flooded but sealed. IF theyre AGM, your voltage values are OK.

Ok so eq voltage is set to 14.6v....it needs to be at least 15.6v. It can need as high as 16.2v or so.

-Charge limit - 16v
-Boost and eq duration - set to highest. should be 180
-Boost voltage - 14.6v
-Float voltage - 14.4v

They need forced equalized for a couple days to be sure theyre shot or not. (Temporarily set boost voltage to 15.8v)

It works like this, when you discharge, an oxide film builds on the plates called sulfation. To get rid of it you must recharge to 100% asap. If you go days, weeks that oxide becomes permanent and you cant get rid of it. It blocks the chemical process and lowers the battery capacity. You will see extreme voltage drop offs and very fast climbing voltage reached under charging.

You should be able to get a warranty claim on them. Def do it.

Thank you Ped for your quick reply, they are sealed lead acid not AGMm I have just made those changes now, i presume the built in defaults are not as ideal as the manual makes out for Lead acid batteries

I think the 40amp in the screenshot refers to the charge controller rating and not the actual chagre current, Ive only even seen 38amps max and thats when we are also using power to run a load.
 
Last edited:
Sure. There seems to be another caveat at play here that's common with paralled batteries. charging unevenly since the primary battery was gassing. The others may have been starving while over charging the first.

Make sure youre connecting cables are good with an ohm test. They should be very large and short, 2awg minimum.

Also when you set controller voltages check the voltage at the terminals on battery 1 and 2. Adjust the controller until the lowest terminal voltage is at the target charge voltage.
 
Last edited:
Sure. There seems to be another caveat at play here that's common with paralled batteries. charging unevenly since the primary battery was gassing. The others may have been starving while over charging the first.

Make sure youre connecting cables are good with an ohm test. They should be very large and short, 2awg minimum.

Thank you for that, i have double checked all the cables with a meter, they're good, battery interconnects are 2awg as you recommanded and 20cm long, also to the charge controller but 40cm long, from batteries to inverter 40cm long, the longest cable run is from the branch split from the solar output to the SCC is 2 meters.

As i said originally I may have had them cabled incorrectly at first in terms or charge in and out. I had all the batteries in parallel and feeding in and out from one end of the battery group, incidentally the battery that leaked is the one that was feeding in and out of the parallel setup. ive since tapped the positive and negative off each end of the parallel group.
 
  • Like
Reactions: Ped
If out gassing occurs again try pulling temp coefficient down to -1 or -2, that will keep the voltage from climbing so far over the bulk setpoint.

I think youre right, the bank got way out of balance.
 
If out gassing occurs again try pulling temp coefficient down to -1 or -2, that will keep the voltage from climbing so far over the bulk setpoint.

I think youre right, the bank got way out of balance.

I really appreciate your help on this. Lots of learning to do on my part. probably should add, I got these batteries for free so not a massive issue if they're dead I am very tempted to go down the LiFepo4 route with built in BMS, cant get those for free though unfortunately but seems like a better solution all around plus with the better DOD, reduced weight and hassle.
 
If out gassing occurs again try pulling temp coefficient down to -1 or -2, that will keep the voltage from climbing so far over the bulk setpoint.

I think youre right, the bank got way out of balance.

Hi Ped, you are a genius, i changed the voltage and charge settings like you said, after two days just leaving the SCC connected with the new settings, the batteires have charged correclty, overnight the voltage stays at 12.65 which I interpret as fully charged. seems that did the trick. I even used the whole setup yesterday to run my pc all day drawing 36amps constant and still the batteries took a charge and were 12.65 overnight.
 
Awesome

If you want lifepo4 these are a killer deal and what I use. 1 is 20% more usable amps than the 2 you have now.

 
Last edited:
Back
Top