Struggling with battery capacity

[wgrippo]

So here is my data from today.
(1) about noon, PV charging lots of sun, the house is using 644W of power.
(2) 6pm, sun is mostly gone, but I opened the PV breaker, so no PV input, 216W of power draw.
(3) 7pm, still no PV input, and I turned of the AC breaker main to the house, and shut down the inverter

 

[wgrippo]

here it is in graph format

 

[anadiner]

anadiner, I didn't realize humidity had an effect. We are in San Felipe, Baja California, Mexico, 3 hours south of US border, right on the beach on the Sea of Cortez. 9 months out of the year we are dry desert and moderate temps, but in the summer the sea heats up and we are mostly between 65% and 80% humidity and temp in the 90s. But these are sealed batteries - what part of the system does humidity affect?

Causes internal decay. I have lots family in CA and yea dry. Im on east coast and at beach. Between humidity dampness getting water in everything (gas can, impossible to keep a salt shaker, etc,) then add to that high salt content.

Ppl here say lithium lithium ... Im just staying marine lead-acid. I know no matter what I get, humidity + salt air = damage any battery

 

[anadiner]

So here is my data from today.
(1) about noon, PV charging lots of sun, the house is using 644W of power.
(2) 6pm, sun is mostly gone, but I opened the PV breaker, so no PV input, 216W of power draw.
(3) 7pm, still no PV input, and I turned of the AC breaker main to the house, and shut down the inverter

View attachment 62666

What r u saying here? U ran house off solar during the day and it used 644w total? Wh?

Sun down. What r u saying? 216w shown as whats available? U didnt say tested batteries. Are they full?

U switched to grid. Will that charge batteries?

 

[GSXR1000]

FYI every time you run your batteries down to 10v you are killing them.. so let's say you have fine
done this 20 times.... major battery damage has occurred... and from the sound of things it is way more then 20 times

 

[anadiner]

FYI every time you run your batteries down to 10v you are killing them.. so let's say you have fine
done this 20 times.... major battery damage has occurred... and from the sound of things it is way more then 20 times

Yea. Said inverter would cut off during night from low voltage cutoff. I dont know what he/she has that set at.

This is one of the easy problems to me because: worked fine, changed nothing, something broke.

Wgrippo, has narrowed it down by testing and obviously the batts arent holding their charge.

Question is not why (cuz once damaged they wont) but how did it happen. Dont want this happen again.

Something caused batteries become inadequate in less than a yr in a 24,000wh system. I'd like to know answer also.

 

[MisterSandals]

I was thinking the batteries looked pretty good based on the voltages. I was expecting to see one in each series as a low voltage anchor.

So i checked the inter web thingy:
A 12-volt AGM battery typically reads 12.7V to 13.0V when full and unloaded...

According to this, all the batteries are still full. And on the very high end of full.

I think I'd like to see the voltages when the bank is at a much lower SoC before making a definitive assessment of the batteries.

 

[wgrippo]

I was thinking the batteries looked pretty good based on the voltages. I was expecting to see one in each series as a low voltage anchor.

So i checked the inter web thingy:
A 12-volt AGM battery typically reads 12.7V to 13.0V when full and unloaded...

According to this, all the batteries are still full. And on the very high end of full.

I think I'd like to see the voltages when the bank is at a much lower SoC before making a definitive assessment of the batteries.

This was right after sunset, so at this point they are pretty full. Yes, I don't see any huge differences either, except battery F is consistently a little lower than the others...

I will take some measurements tomorrow morning when the batteries are low.

 

[wgrippo]

Causes internal decay. I have lots family in CA and yea dry. Im on east coast and at beach. Between humidity dampness getting water in everything (gas can, impossible to keep a salt shaker, etc,) then add to that high salt content.

Ppl here say lithium lithium ... Im just staying marine lead-acid. I know no matter what I get, humidity + salt air = damage any battery

They are sealed batteries, in a small, mostly sealed, insulated, air-conditioned (during the day) room. I find it hard to believe humidity is causing my problem. There is no corrosion on the terminals - everything looks nice and clean.

Sure, outdoors in the summer we do have corrosion problems being right on the beach...

 

[wgrippo]

What r u saying here? U ran house off solar during the day and it used 644w total? Wh?

Sun down. What r u saying? 216w shown as whats available? U didnt say tested batteries. Are they full?

U switched to grid. Will that charge batteries?

No, I'm saying at the time of measurement, the power draw is 644W. Probably not important, just a reading I get from my inverter. Sun down same thing, at the time I measured the voltages the AC load was 216W.

I don't have grid, 100% solar. The nearest power pole is 42km north.

I have a generator, but the point of my post is to figure out why my batteries seem to be draining so quickly overnight. And to confirm my concern that this is happening, because the battery vendor insists I just need more batteries.

 

[circus]

I wonder if a big car battery load tester would spot a dead cell? Do strings of stringed lead-carbon cells have balance issues?

 

[Dacflyer]

here is something to consider,, refrigerators all have defrost cycles they go thru.. . usually 2x a day.. the defrost circuit can draw a lot of power, because it's a heater... maybe this is a cause??
I do know the defrost uses more power than the compressor..

But i too am thinking a possible battery issue has developed..
the batteries are not having even charges.. they are unbalanced.
Maybe take 1 string of line and charger the batteries individually with a good beefy charger m and see if that brings the voltages up.. then let them rest for an hour or so and see what the voltage is.. compare the reading with other batteries...

 

[wgrippo]

I wonder if a big car battery load tester would spot a dead cell? Do strings of stringed lead-carbon cells have balance issues?

I don't know for sure, but when I bought the batteries, I asked the guy I bought them from (supposed lead-carbon expert) about balance, and he said just hook them up and they will equalize as they charge.

I do have this load tester, it says it works on AGM. I use it mainly for starter batteries on motorcycles, for measuring CCA. I'll look into what else it can tell me...

 

[wgrippo]

here is something to consider,, refrigerators all have defrost cycles they go thru.. . usually 2x a day.. the defrost circuit can draw a lot of power, because it's a heater... maybe this is a cause??
I do know the defrost uses more power than the compressor..

But i too am thinking a possible battery issue has developed..
the batteries are not having even charges.. they are unbalanced.
Maybe take 1 string of line and charger the batteries individually with a good beefy charger m and see if that brings the voltages up.. then let them rest for an hour or so and see what the voltage is.. compare the reading with other batteries...

I wish this was a problem of finding where some unknown power draw was, but my inverter reports all of the power data. So I know how much power is being used at all times. For example:



I would love to find there is some other power draw that I'm unaware of and is not reported by the inverter, but I don't know how that could be. The batteries feed directly into the inverter, and the AC out goes directly into my electrical panel.

 

[740GLE]

So under a normal sunny day it’d be nice to know the carrying current going into the batteries, sounds like you have decent constant load.

My wild guess is that the SCC setting for bulk and foal maybe a bit low, thus never get fully charged up and go to float too soon.

 

[740GLE]

Reach out to the battery guy you got these from and see what a good bulk voltage should be, from the battery specs it does seem like the bulk voltage you have set up for 56v might be too low. At 58v that’s only 14.5v/per battery and that might be the difference.

For a 4500w array minus your load your only putting in 1000-1500w, divided by two strings you def won’t be over the charging current, I wonder if your current is too low going into the battery?

 

[Hedges]

I don't know for sure, but when I bought the batteries, I asked the guy I bought them from (supposed lead-carbon expert) about balance, and he said just hook them up and they will equalize as they charge.

I do have this load tester, it says it works on AGM. I use it mainly for starter batteries on motorcycles, for measuring CCA. I'll look into what else it can tell me...

I don't think a series string of AGM will equalize much, because we avoid charging at high enough voltage to equalize AGM. For FLA it is a normal, occasional, process.

Since you have two strings, you could isolate one string and then fully charge each battery of that string individually with an AC powered battery charger.

I have some older AGM automotive starting batteries which were getting weak and a bit low in resting voltage. (Automotive alternator regulator doesn't deliver correct voltage for charging AGM.) I fully charged them with automatic charging having AGM setting, then charged to 14.5V with 1A CC/CV power supply, then equalized with 15.5V 1A CC/CV. That did bring up rest voltage, but I haven't tested capacity or cranking amps.


"the battery voltage reached 56V by 10:30am, and stayed there for two hours until 12:30pm, when it dropped to 54V. It stayed right around 54V until around 5pm"

My AGM batteries (SunXtender) are supposed to be charged at least with 0.2C. If not, an extended absorption is recommended. Other people do say that lead-acid requires a long charge time.
It looks like your PV array and battery capacity could achieve that charge current, but might not depending on illumination and loads.

If you suspect charging issue came with change from PWM to MPPT (which would have had different parameters most likely, and of course differ in on/off cycles vs. steady-state voltage), then try swapping back in the PWM and see what happens.

Whatever SCC you use, ideally you would have excess PV capacity and regulate lead-acid battery charging to the optimum current. With Victron that can be done, using a battery current shunt and a Victron controller (not sure which model and p/n). It will tell SCC to increase/decrease output meeting inverter power draw while holding battery current steady.

 

[wgrippo]

Reach out to the battery guy you got these from and see what a good bulk voltage should be, from the battery specs it does seem like the bulk voltage you have set up for 56v might be too low. At 58v that’s only 14.5v/per battery and that might be the difference.

For a 4500w array minus your load your only putting in 1000-1500w, divided by two strings you def won’t be over the charging current, I wonder if your current is too low going into the battery?

Thanks, I just did. The battery guy recommends 57.4V bulk and 55.4V float. Will update those settings now...

Also, Growatt just got back to me and suggested raising my max charging current up from the default 80A, but the did not give a recommendation. Max setting is 160A, I think I will try 140A....?

 

[MisterSandals]

Max setting is 160A, I think I will try 140A....?

recommends 57.4V bulk and 55.4V float. Will update those settings now...

If you want to learn what the problem was, I'd change 1 thing at a time. But thats just me.

 

[Hedges]

It's just a math problem to extract correlation from multivariate test results

I have seen people say, "uncorrelated", but simply eyeballing two offset wiggly lines, I said, "Time delay from room temperature changing to response of the electronic component."
(Ok, I actually created an array of data in Excel with various offsets, and found one where there was high correlation between data sets.)

There is zero correlation between current in an inductor and the applied voltage. So "Causation does not imply correlation"