batteries abruptly stopped holding voltage under load...

Jabroni · January 2017

Hi there. Im having a issue with my solar setup. Maybe you guys here have some insight that could help me out.

- 6000w in solar panels
- 4k inverter
- (2) Magnum charge controllers
- Magnum remote controller
- (8) Full River AGM batteries http://www.fullriverbattery.com/product/batteries/DC400-6
- Honda 7k generator

System was running fine for about 9 months. (mostly always in float, very few cycles during summer months, appx 200-300 cycles at most have been done on these batteries)

This past November it's been getting used a lot more. Typical load on the system would be between 2000-3000w. I'd have to run my generator twice per day to recharge the system. In between charges I'd get about 5 hours out of my batteries with that load.

I had my solar company install a auto-start system in early December so I wouldnt have to get up at 5am to turn the geni on every morning.

Here's where things went wrong...

Before the installer had hit the end of my driveway, the auto-start kicked on due to low voltage. The geni ran for 30 minutes, the voltage went back up, then auto-shut down the geni.

15 minutes later, the voltage was low again, the auto-start kicked on. And so it repeated like that. At one point the geni had kicked on 40 times in one day and I was worried about ruining my generator, so I disabled the auto-start.

Same issue persisted. I can no longer put loads of 1200w or more onto the system without the voltage dropping extremely fast and the system going into protection mode.

This problem started immediately to the day/hour that the auto-start was installed.

The installer is claiming that the batteries are shot. But he came out to diagnose the problem 2 days ago and couldnt find a problem with the batteries (using a handheld volt tester) much less a short/ghost load in the system.

Im wondering if by some chance there is a way that Im getting false voltage readings to the controller? And thus kicking on the auto-start/auto-protection shut down. Somehow related to the auto-start because this started happening when the auto-start was installed. Not days or weeks later. The same day.

The day the installer showed up to diagnose the problem, after not being able to figure it out, he had a $5000 invoice already printed up ready to go. Trying to sell me new batteries. My question to him was "what if the problem persists?" He claims it wont, and that the problem must've existed before the auto-start was installed, "I just didnt notice it". That's funny, I think I wouldve noticed my system shutting itself down 40x per day.

So here I am, at a total loss for explanations. Hoping maybe some experts here can help.

just starting · January 2017

Is this 48 v

Jabroni · January 2017

just starting said:

Is this 48 v

24v.

Sorry, I shouldve stated as much.

just starting · January 2017

Agm are very picky about the charge profile they can be overcharged very easy can you tell us about the charge profile? Bulk asorb and float. I'm just asking what the more knowledgeable will ask.

Jabroni · January 2017

just starting said:

Agm are very picky about the charge profile they can be overcharged very easy can you tell us about the charge profile? Bulk asorb and float. I'm just asking what the more knowledgeable will ask.

I do not know enough about these sorts of systems to know the charge profile.

I can tell you this much, these were the settings that the installer set it too:

Auto-Start on State of Charge: originally set to 60% and to shut off at 85% (assuming the panels could do the rest)
Auto-Start on Voltage: originally set to 23.8v and to shut off at 28.8 (which the installer later admitted he shouldve set to absorb, but it's not the root issue here)
Auto-Start on Load: originally set to 3600w or 30amps

Ive got the SOC to auto-start currently at 80% and shut off at 95% as to protect the batteries. But again, this function doesnt ever see use. The voltage drops way too fast.

I lowered the Voltage auto-start to 23.2 and have tried it as low as 23 to see if when under load, the system would hover in the low 23's. It doesnt. I just continues to drop. Prior to the auto-start being installed, the voltage would be in the high 23s or low 24s while under load for several hours.

And I never put a 3600w load on the system so that function has never had to kick in.

Typically in the past, the voltage would correlate with the SOC. Now? They dont correlate at all.

DConlyGuy · January 2017

well doing the math you have about 830ah at 24 volts and since your using 3k watts at 24 volts that 125 amps and hour so in about 5 hours you would need to recharge them again

Jabroni · January 2017

DConlyGuy said:

well doing the math you have about 830ah at 24 volts and since your using 3k watts at 24 volts that 125 amps and hour so in about 5 hours you would need to recharge them again

Correct. That isnt where the issue lies tho.

The issue is that instead of needing to charge the batteries 5 hours later (as was the case all of November) the voltage drops in as little as 15 minutes and then the auto-start kicks on. Runs for 30 minutes. Then shuts off. Then the load drops the voltage down again and the cycle repeats itself all day.

If I shut the auto-start system off, I still only get 15 minutes before the voltage is so low that the entire system shuts down (to protect the batteries). This only started behaving this way IMMEDIATELY the day the auto-start was installed.

Prior to that, I could go the 5 hours on a load of 125 amps.

So now Im sitting here wondering how/why the auto-start could be giving me possibly false low voltage readings?

DConlyGuy · January 2017

well i will crossing my fingers for you that the battery's are not! already toast, 3k load on such a small battery bank is really hard on the battery's

Jabroni · January 2017

DConlyGuy said:

well i will crossing my fingers for you that the battery's are not! already toast, 3k load on such a small battery bank is really hard on the battery's

When the installer came to diagnose the problem, and after we ruled out a ghost load or short, the next target was the batteries.

We disconnected them all from each other and ran individual voltage tests under 100w loads. None of the batteries were "bad" or in need of immediate replacement.

A 3k load may be hard on the batteries, Im not one to argue that. But doesnt it seem HIGHLY coincidental that the day the auto-start is installed that all of a sudden the batteries cant hold their voltage (as far as the controller is reading)? That's why Im sitting here zero'd in on the auto-start. Im inclined to think something about it changed the way voltage was being read. I dunno if that's even possible but that's all I can deduce.

The damn company that installed the system is trying to say that my batteries were already shot and I just wasnt noticing it before. Which is ludicrous.

just starting · January 2017

My bank Is olny 60 amps more and I have 1/4 the panels. I pull about 3.5 total kw every 24 hours. My voltage is at 24.4 in the morning. I think you should load test the batterys individually. You could also check the temp on each battery while charging with a inferred laser temp gun and see if one is hot

just starting · January 2017

Where are you located?

just starting · January 2017

Or charge it and with no loads and shut down for the night check the voltage after it rests all night with no loads or charge.

Jabroni · January 2017

just starting said:

My bank I olny 60 amps more and I have 1/4 the panels. I pull about 3.5 total kw every 24 hours. My voltage is at 24.4 in the morning. I think you should load test the batterys individually. You could also check the temp on each battry while charging with a inferred laser temp gun and see if one is hot

We did load test each battery individually.

None of them were ''bad" or showing need of immediate replacement. And trust me, had they been, the installer wouldve gladly suggested as much. That seems to be their solution to everything, spend more money.

Jabroni · January 2017

just starting said:

Where are you located?

Northern California. And yes, it's brutally cold here right now. But that was the case in November as well.

Jabroni · January 2017

just starting said: Or charge it and with no loads and shut down for the night check the voltage after it rests all night with no loads or charge.

At night there is a minimal 4-5 amp load on the system. By morning the voltage is 24.12v

That isnt the test youre asking for, but it's close. Does that tell you anything?

With everything turned off, there is a 1-2 amp load on the system, which is the solar system itself.

just starting · January 2017

So I think what is happening Is you are pulling a big load witch will run down the batterys very quick so if you pull 125 amps per hour your bank drops voltage consistently with the amp draw. So your voltage keeps dropping. Its perkuts law. You should read it on wiki. There is no real way to check SOC on a agm. Olny resting voltage. And a load test. When you put a set draw on the battery for so many min and then check voltage against the draw and the end volts. It will tell you the capacity of the battery.

just starting · January 2017

I don't think a 100w load would be enough maybe a 100amp load for a few min. But I'm not sure. I got my batterys used for 1500$ I'm switching to flooded after this bank.

westbranch · January 2017

your link to specs say ~230 Min @ 75A that is 3.8Hrs at 75A, with that rating
the Discharge table shows that you can run for 2 hrs at 123A or ~ 125A so at running them for 5 hrs you have severely damaged them ,,, I hope they can be recovered...
Cold temps would lower the amount you can safely remove...

Jabroni · January 2017

just starting said:

So I think what is happening Is you are pulling a big load witch will run down the batterys very quick so if you pull 125 amps per hour your bank drops voltage consistently with the amp draw. So your voltage keeps dropping. Its perkuts law. You should read it on wiki. There is no real way to check SOC on a agm. Olny resting voltage. And a load test. When you put a set draw on the battery for so many min and then check voltage against the draw and the end volts. It will tell you the capacity of the battery.

I got a basic understanding of perkuts law.

But again, the 125amp draw was never a problem until the autostart was installed.

Even if I disable the auto-start, I cant get 15 minutes out of the batteries without the voltage reading so low that the system goes into protection.

Had I never installed the auto-start, I would think my system would still be performing as did before.

Jabroni · January 2017

westbranch said:

your link to specs say ~230 Min @ 75A that is 3.8Hrs at 75A, with that rating
the Discharge table shows that you can run for 2 hrs at 123A or ~ 125A so at running them for 5 hrs you have severely damaged them ,,, I hope they can be recovered...
Cold temps would lower the amount you can safely remove...

Ok. So are you suggesting that after a months worth of abuse, the batteries took a dump on the day the auto-start was installed?

Let's say I was abusing the batteries, their performance would've dropped off gradually. Maybe even steeply. But instantaneously the day the auto-start was installed? That's too much coincidence for me.

I'd like to zero in on the auto-start and how it possibly could be causing a false voltage reading or maybe some sort of setting was changed on the remote controller that is causing a false reading.

just starting · January 2017

Were the batterys new? What is the date code? Did they sit before being installed?

just starting · January 2017

That way when I kill it I'm not out 5000$ olny 1500$

BB. · January 2017

Can you put a volt meter across each battery during that time before autostart kicks in? It is possible that you have one weak/bad cell (or even wiring connection) that is dropping the DC bus voltage.

If you have a DC current clamp meter (like this Sears unit), measure the current from each string (2 parallel strings?).

http://www.sears.com/craftsman-digital-clamp-on-ammeter/p-03482369000P

If you do not have a bad cell, it sounds like it could be a poor electrical connection somewhere in the battery/bus cabling.

-Bill

Jabroni · January 2017

just starting said:

I don't think a 100w load would be enough maybe a 100amp load for a few min. But I'm not sure. I got my batterys used for 1500$ I'm switching to flooded after this bank.

Im just going by what the installer's handheld voltage meter was showing.

Jabroni · January 2017

just starting said:

Were the batterys new? What is the date code? Did they sit before being installed?

New. I dont think they sat.

As Ive said, they performed fine for months. I dont know what a data code is but here is the exact model: http://www.fullriverbattery.com/product/batteries/DC400-6

Jabroni · January 2017

BB. said:

Can you put a volt meter across each battery during that time before autostart kicks in? It is possible that you have one weak/bad cell (or even wiring connection) that is dropping the DC bus voltage.

If you have a DC current clamp meter (like this Sears unit), measure the current from each string (2 parallel strings?).

If you do not have a bad cell, it sounds like it could be a poor electrical connection somewhere in the battery/bus cabling.

-Bill

He was using a DC current clamp meter I think. That's how he was getting the 6+ volt readings. He couldnt find a bad cell. But that was with all the batteries disconnected from one another.

I think there is a bad battery/bus cabling connection as well. But he was unable to find one. Forgive me if I dont trust his work. So I called up Magnum to trouble shoot with them...

I spent 2 hours with their customer support. I wanted to understand the auto-start better and what could be the issue.

We went thru every setting on the controller and also some of the solar charging hardware.

(1) other than the kind folks on this forum, they were the first person Ive spoken too to NOT dismiss my claims of the system not performing immediately after the auto-start was installed. They were very perplexed at the coincidence of it all.

(2) that said, they could not fathom how the auto start would have any effect on voltage readings.

(3) so then this is when it got "fun". We went thru all the settings on the controller and here are the mistakes found (some of which you already alluded too...

- AGS stop was set to 28.8 (shouldve been set to FLOAT)
- Absorb was only set to 2.0 hours (shouldve been to 4.0 hours)
- and here's the REALLY big one, the BMK was set to 790aHrs (so the controller this entire time has been charging for a smaller battery bank than was already here and for too short a absorb time... my actual battery storage is 830aHrs)
- and lastly, the jumpers on the solar charger controllers were set wrong, and were set for a much smaller solar array than I currently have

So yes, the batteries have been taking a beating. During the summer due to the jumper settings and since November when the generator was being used heavily since the absorb as set wrong. And apparently the entire time since the BMK has been wrong all along. Magnum had some settings and charging advice that they thought could try to salvage the batteries a bit but cleaning off the plates (forced charging) and so Im going to try it.

This is mostly a installer setup error, and the resulting damage to the batteries not being charged fully.

That said, Im still hung up on the coincidence of it all. The performance drop off was too sudden and was immediate after the auto start was installed. There just seems to be something else too it. Im not saying the batteries havent taken a hit. But if their is a problem related to the autostart, new batteries arent going to solve this issue.

Jabroni · January 2017

just starting said:

Is this 48 v

To be honest, Im not sure at this point.

Can you help me figure that out?

All I know is that with I look at my remote controller, the system shows that it should be in the 24-30v range depending on DC input levels.

just starting · January 2017

I was assuming with the 6 K of panels I would have went 48v less copper ,higher volt array. Smaller wire for higher voltage.Its 24 v system for sure. IMHO, you should have some one teach you how the settings are and what they are for, you have a lot invested in your setup it's nice. A lot of power, depending on your location.

Photowhit · January 2017

Jabroni said:

just starting said:

Is this 48 v

To be honest, Im not sure at this point.

Can you help me figure that out?

All I know is that with I look at my remote controller, the system shows that it should be in the 24-30v range depending on DC input levels.

You can just check the inverter, it will be voltage specific. If it's not a direct read, get the model number.

Assuming a 24 volt bank (I would say it should be a 48 v, but a 24v is viable) The 'abuse' is not as bad as suggested. You have 2 strings and the people saying you have abused them aren't taking this into account. You are only drawing 1/2 of the current from each string. First thing is to check all your connections and the wiring gauge. 3000 watts at 24 volts is about (3000/24=) 125 amps, This is a substantial draw and I suspect it is voltage sag that is creating the generator. If you will read amperage from the link you shared, the batteries are rated for 340ah at a 5 hour draw, 125a x 5hr = 625ah, 2 strings of 340ah = 680ah battery bank (THIS IS WHAT YOU HAVE IF YOU ARE DRAWING THAT AMOUNT OF ENERGY!).

Large constant draws of this size will create a voltage sag. It's always hard to figure out where you are at with AGM batteries. You could do a full load test, but if you just disconnect everything when the batteries reach 23 volts (or where ever you have the auto start set and let them rest for an hour or 2 and then see where the voltage levels out, I suspect it will give you something much higher.

If you find the voltage returns, I would fist be sure you have used very stout wiring for your connections. I would be sure the 2 strings are connecting correctly for even resistance.

Once you have the number that the voltage returned to, to use as a 'real' SOC, I would keep adjusting you Low voltage disconnect until the 'real' SOC number reflects what you want, starting at 50% capacity or 30% or 65%...

Other things to consider, are the batteries cold, I suspect with the huge 'push pull' they are staying warm, but at 32 degrees you lose about 15% of capacity. Loose connections as well as undersize wiring can contribute to low voltage readings.

Jabroni · January 2017

Photowhit said:

Jabroni said:

just starting said:

Is this 48 v

To be honest, Im not sure at this point.

Can you help me figure that out?

All I know is that with I look at my remote controller, the system shows that it should be in the 24-30v range depending on DC input levels.

You can just check the inverter, it will be voltage specific. If it's not a direct read, get the model number.

Assuming a 24 volt bank (I would say it should be a 48 v, but a 24v is viable) The 'abuse' is not as bad as suggested. You have 2 strings and the people saying you have abused them aren't taking this into account. You are only drawing 1/2 of the current from each string. First thing is to check all your connections and the wiring gauge. 3000 watts at 24 volts is about (3000/24=) 125 amps, This is a substantial draw and I suspect it is voltage sag that is creating the generator. If you will read amperage from the link you shared, the batteries are rated for 340ah at a 5 hour draw, 125a x 5hr = 625ah, 2 strings of 340ah = 680ah battery bank (THIS IS WHAT YOU HAVE IF YOU ARE DRAWING THAT AMOUNT OF ENERGY!).

Large constant draws of this size will create a voltage sag. It's always hard to figure out where you are at with AGM batteries. You could do a full load test, but if you just disconnect everything when the batteries reach 23 volts (or where ever you have the auto start set and let them rest for an hour or 2 and then see where the voltage levels out, I suspect it will give you something much higher.

If you find the voltage returns, I would first be sure you have used very stout wiring for your connections. I would be sure the 2 strings are connecting correctly for even resistance.

Once you have the number that the voltage returned to, to use as a 'real' SOC, I would keep adjusting you Low voltage disconnect until the 'real' SOC number reflects what you want, starting at 50% capacity or 30% or 65%...

Other things to consider, are the batteries cold, I suspect with the huge 'push pull' they are staying warm, but at 32 degrees you lose about 15% of capacity. Loose connections as well as undersize wiring can contribute to low voltage readings.

Im fairly certain this is the model inverter I have: http://www.magnum-dimensions.com/renewable-energy-products/inverter-chargers
It's the MS4020PAE model. And yeah, it says 24v.

While we're going over the specs, here are the solar chargers I have http://www.magnum-dimensions.com/renewable-energy-products/controllers
I have 2 of the PT-100's (not sure why since they are rated to handle up to 6600w)

On to your other points, yes appx 100-125 amps was my constant load. And I understand that over time, the voltage should continue to drop under these loads. Dropping in 15 minutes tho? That seems odd considering that was never the case pre auto-start install.

As for the battery temperature, Im going over the Magnum data for the month of November. Average battery temp was about 20 degrees celsius. By November 26th the avg temp each day was around 12 degrees (and the fall off was pretty abrupt). For the whole month of
December the batteries were averaging around 10 degrees.

The ambient temperature in the shop where the batteries are houses was pretty much the same temps. Anywhere from 50 degrees fahrenheit in November to now 30-40 degrees in December/January. Could this be the cause of the sudden drop on performance? Would that mean that once things warm up in the Spring time that the batteries will be able to hold their voltage under load again? I suppose this would account for the "coincidence" Im so hung up on.

Jabroni · January 2017

just starting said:

I was assuming with the 6 K of panels I would have went 48v less copper ,higher volt array. Smaller wire for higher voltage.Its 24 v system for sure. IMHO, you should have some one teach you how the settings are and what they are for, you have a lot invested in your setup it's nice. A lot of power, depending on your location.

I appreciate the help and the kind words. I posted this same situation/issue over on another solar forum and didnt get anywhere close to the reception I received here.

https://www.solarpaneltalk.com/forum/off-grid-solar/batteries-energy-storage/general-batteries/340328-batteries-wont-hold-voltage/page4

Some of the members were more interested in proving how smart they were / how ignorant on this topic I am than actually helping. I know Im ignorant on this subject matter compared to someone where it's their hobby. My ignorance was never up for debate haha.

And by and large they misread the situation as they were convinced my entire setup was useless and so poorly designed that it was not salvageable. Which again, had nothing really to do with the topic at hand.

And just to be clear, since on that other forum it seems to have been a huge focal point, while Im running an industrial facility on my solar system, I am NOT running industrial heating. They were convinced I was using WAY more energy than a typical household. This isnt the exact model heater I run but it's very similar http://www.homedepot.com/p/Lifesmart-1500-Watt-6-Element-Infrared-Room-Heater-with-Oak-Cabinet-and-Remote-LS-1000X-6W-IN/204308189?cm_mmc=Shopping|THD|G|0|G-BASE-PLA-D27E-Electrical|&gclid=CjwKEAiA79zDBRCgyf2FgeiY-CESJABzr0BMOh37T-RMbVKZu0LyUSLAguYN4DpTKK9e1z1Ns37AkhoCDwrw_wcB&gclsrc=aw.ds

Correct me if Im wrong, but that isnt an unusual heater to have plugged into one's small apartment. My small living quarters at my facility is exactly 600 sq ft. I run that heater and sometimes a 2nd oil heater (just like this one http://www.homedepot.com/p/Pelonis-1500-Watt-Digital-Oil-Filled-Radiant-Portable-Heater-with-Remote-Control-NY1507-14A/206021790 ) to stay warm. That's where the loads were coming from. Yes, 3000w at times. But nothing that unusual for a home setting in my experience.

batteries abruptly stopped holding voltage under load...

Comments

Categories