batteries abruptly stopped holding voltage under load...

Photowhit · January 2017

Other than the rapid change with the auto start, I would agree with Mike. Check those connections, perhaps some how things were changed when they added the auto start, disconnecting a string of batteries would do it! If the genny is connected directly to the battery bank, perhaps he disconnected one of the jumpers connecting the 2 strings?

Jabroni · January 2017

Estragon said:

If you don't already have them, I'd suggest you get a multimeter, an infrared thermometer, and a clamp-on amp meter. These are pretty basic tools you should probably have on hand even without this current problem.

With these tools, here is what I would do:

With minimal loads (no loads over ~4a for at least 1/2 hour) and charge controllers off, get a reference voltage and temperatures at the ends of each string of batteries, at the connection where the two strings are joined together, and the bussbar where the inverter and controllers connect. Write down these values.

If voltage drops more than about 0.1v as you move to connections more distant from the batteries, there is a problem with a connection or cable is undersized. Likewise, the connection temps should be very close to ambient. If all good so far...

Make sure AGS is off.

Add a known constant load like a heater (~1500w x 1.2 inverter tare÷25v=~72a). If you have a smaller known load like a string of lights use that.

When the system displayed voltage has dropped to ~23v check and note the above voltages and temps again.

Turn off the load.

Same as first test. Voltage may be sagging as you test but should still be fairly close if you can take them reasonably close in time. Temps should still be close to ambient.

If still all good check voltages across each battery +/-. If still good, grab a coffee and post results here. After 20mins check voltage at the end of each string again and note.

Turn controllers back on. Charge with genny if needed. Check voltage at the end of each string every 15mins and note. If you have the clamp amp meter also check the current into each string. Compare each reading to what your system is displaying.

When you get to absorb voltage turn your larger loads back on if need be.

N.B. This is what *I* would do. You'll need to assess your own comfort level working around live circuits and ability to do so safely.

Thank you.

I will buy this equipment but also get another well respected / referred solar guy out here to do this exact step-by-step process.

Another theory that was floated to me by a electrician was this, let me know if it sounds plausible:

I'm an electrician with minimal expertise with solar systems, but it sounds to me that when they added the auto start they put the voltage detector on the load side rather than the line side which is showing a much larger voltage drop than what is actually happening at your battery bank. The "trigger"wire should be on the battery side of the system, not the house side.

westbranch · January 2017

Quite possible... Tell us what you find
Where did the shunt get mounted? Do you have some pics?

Jabroni · January 2017

westbranch said:

Ahha, sorry I missed that you had 2 strings of 24V...
I see your main issue right now is the sudden drop off that arrived when the sun drops from the sky in December...
(My personal solution has been to cut back on optional consumption as needed or run the genny in the daytime so the battery can handle the overnight draw)
During the summer/fall I would wager that your PV was maintaining a goodly portion of the loads... come Dec. that is not happening as the battery is now depleted (Nov. probably was the shoulder/transition season...) and the solar charger is playing catch up, but it can never happen due to less incoming, so the gen has to be started....
Now we know that the settings are off for the start and stop as you outlined, good that you got that reset.
Now, hopefully the gen will start and stop only a few times a day since the batteries will get a far better charge, at a minimum along the lines of what they should have, that just needs to be fine tuned after the bank recovers with good long 3 stage charges.
Please keep us posted on the recovery actions you take. .

During the Summer/Fall, there wasnt any load that the PV couldnt handle during the day. And the loads at night were a minimal 150w at best (SOC would read about 92% at 9am on most days). Once in a blue moon I'd be out on site and the system would get some loads on it at night, and I'd use the generator to keep things charged up (albeit we're finding out now what a waste that was due to poor setup).

Even in November, I wouldnt say the solar was much help. The sun was already at a bad angle. I used the generator to get things back to 90+ SOC before letting the sun would do the rest. But again, between the jumpers being set wrong on the solar chargers and the geni not running in absorb for long enough, things werent getting charged all the way.

Im going to wait a few more days of "force charging" the setup before I try putting any loads on it, if at all. Im pretty gunshy at this point until Ive done more diagnostic work. I want to get the wiring theory sorted first. And then test the batteries to see if they are healthy. And then I must weight out if most of my performance loss has been due to the sharp drop in temperature.

I'll keep everyone posted for sure.

Jabroni · January 2017

just starting said:

Their are some nice blue Flame heaters. I don't know a lot about them may I suggest a rv furnace it might suit your needs and olny takes a small space. You can get a 10 amp 24volt to 12 volt converter to run it. But a 100# propane tank olny holds 2159400 btu so 30000 btu furnace on 1/4 time will last 11 days.and at 1.5 amp draw per hour at 24v with1/4 run time. + You could run 4" ducting where you want.

U got a link to what your speaking of? I'll check it out.

Jabroni · January 2017

mike95490 said:

But, I'm of the opinion that you have been deeply (90%) cycling your batteries all summer, and then the short days and cold weather finished them off,

Maybe Ive miscommunicated somewhere along this all. But during the summer the battery bank wasnt being used much, if at all. All of my heavy loads were during the day, in which the solar panel array kept things in float mode.

I never ran any AC units or needed heaters in the summer.

mcgivor · January 2017

@Jabroni said.i During the Summer/Fall, there wasnt any load that the PV couldnt handle during the day. And the loads at night were a minimal 150w at best (SOC would read about 92% at 9am on most days). Once in a blue moon I'd be out on site and the system would get some loads on it at night, and I'd use the generator to keep things charged up (albeit we're finding out now what a waste that was due to poor setup).

Sounds like the perfect storm, everything on the edge of failure but showing no obvious symptoms giving a false sence of security, then the cold, shorter days and higher demand came and exposed all the weaknesses.
Your plan to attempt to fully charge the battery is a good move, then with some knowledgeable assistance go over everything, then do a dynamic test, as static testing reveals little. A battery may have a voltage which indicates a high state of charge but as soon as a load is connected it collapses.

Jabroni · January 2017

mcgivor said:

Sounds like the perfect storm, everything on the edge of failure but showing no obvious symptoms giving a false sence of security, then the cold, shorter days and higher demand came and exposed all the weaknesses.

If anything, yeah November seems to have given me that false hope. Temps hadnt dropped so low (if that is the reason for the steep drop in performance) that the system couldnt perform and then December showed up and boom.

Im still crossing my fingers to find that the auto-start was improperly wired up. Or that there's a loose connection somewhere.

It wont cause me to go back to putting heavy loads on the system at night. But it will allow me to start trusting the auto-start once it's wired properly. As of right now everything is back to manual.

I'd love to be able to use the auto-start in the summer to handle things at night for short periods when and if I do put decent loads on the system. Just bypass the batteries altogether.

Jabroni · January 2017

westbranch said:

Quite possible... Tell us what you find
Where did the shunt get mounted? Do you have some pics?

Shunt?

I'll post pics of my entire setup this weekend. EVERYTHING.

You guys can pick it apart.

westbranch · January 2017

Shunt? http://www.midnitesolar.com/productPhoto.php?product_ID=101&productCatName=Accessories &productCat_ID=5&sortOrder=9&act=pc

and here is where it fits with all the other bits in a complete system... it fits into the NEG wiring...
http://www.midnitesolar.com/pdfs/classic_simple_flow_diagram.pdf

Estragon · January 2017

@Jabroni re shunt (voltage detector) location- It's a thought, but the problem is that if the location of the shunt is causing the reading to be way off, the reading is telling you the wire may be undersized or there may be a problem connection. You want beefy cables to carry your loads with minimal voltage drop from wire losses. Testing as described previously will help determine this.

What having the shunt wired wrong could do is mess with SOC calcs because the system needs to see what's going in/out ftom battery. Installed wrong it will see current going from panels and genny going to loads and think that battery is getting charged when it tsn't.

karrak · January 2017

I usually don't delve into threads on lead acid battery based systems as I don't have much experience with them but felt compelled to in this case.

To me this thread has highlighted a number of limitations with using lead acid batteries as storage in off grid system. Rather than go off topic I have started a new thread here http://forum.solar-electric.com/discussion/351566/have-lithium-ion-batteries-come-of-age-for-off-grid-systems

In the OP's case the installer should have known these limitations and have told @Jabroni about them as they place severe limitations on how you use the power from a lead acid battery.

Hopefully the battery has not suffered too much damage.

If @Jabroni needs to replace or maybe add another battery I would suggest looking at Lithium Ion batteries. Lithium Ion batteries would be quite happy with the sort of usage patterns than @Jabroni has described.

Simon

Jabroni · January 2017

Estragon said:

@Jabroni re shunt (voltage detector) location- It's a thought, but the problem is that if the location of the shunt is causing the reading to be way off, the reading is telling you the wire may be undersized or there may be a problem connection. You want beefy cables to carry your loads with minimal voltage drop from wire losses. Testing as described previously will help determine this.

What having the shunt wired wrong could do is mess with SOC calcs because the system needs to see what's going in/out ftom battery. Installed wrong it will see current going from panels and genny going to loads and think that battery is getting charged when it tsn't.

The cables I have connecting all the batteries are thick as hell. I'll post pics this weekend.

Jabroni · January 2017

karrak said:

I usually don't delve into threads on lead acid battery based systems as I don't have much experience with them but felt compelled to in this case.

To me this thread has highlighted a number of limitations with using lead acid batteries as storage in off grid system. Rather than go off topic I have started a new thread here http://forum.solar-electric.com/discussion/351566/have-lithium-ion-batteries-come-of-age-for-off-grid-systems

In the OP's case the installer should have known these limitations and have told @Jabroni about them as they place severe limitations on how you use the power from a lead acid battery.

Hopefully the battery has not suffered too much damage.

If @Jabroni needs to replace or maybe add another battery I would suggest looking at Lithium Ion batteries. Lithium Ion batteries would be quite happy with the sort of usage patterns than @Jabroni has described.

Simon

I got the batteries I did because I wanted low maintenance. I knew I would be paying more. And I knew I would get less life out of them in the grand scheme of things. But there are aspects of my facility that require things to be as low maintenance as possible.

I dont know what my next set of batteries will be. But Im all up for suggestions and I'll be asking you guys. I do know what I'll be asking of them though. I need zero maintenance. And I need to be able to run everything on my facility at night if need be. For at least 6 hours. But we'll dive into that the next go around when Im actually making the switch.

A big part of this is I never thought I'd be spending so much time out here during the winter. But then again, I never thought I'd have a rash of break-ins & vandalism that would require me to be here full-time guarding the place (6 break-ins in less than a month, $40k in damage, caught the culprit once, local sheriffs still wont touch the matter, it's been VERY frustrating). I dont want to go into too much detail on this really negative stuff Ive been dealing with this winter, it's way off topic. But the point is, I was never planning on spending more than a week or so at a time every few months out here. And the majority of those visits would be in the summer when the PV could easily take care of these loads.

The system runs perfectly during the summer for what I need it to do. It's a unconventional set-up. And I have ZERO doubts that it could've been designed better. But there are specific reasons why Ive gone with the unbalanced (and seemingly backwards) PV-to-battery ratio. That was intentional and it wouldnt have worked the other way around.

Granted, Im a laymen in all this. So if I got a 830 amp hour battery bank and the installer tells me half of it is useable, Im going to assume that I can put a 3000w load on it for 4 hours and then just recharge things and everything is going as it should. So long as the low voltage isnt shutting the system off prematurely, I wouldnt know any better. And up until December, that wasnt a problem. As Ive found out from these forums, that is wrong. Even before the AGS issues brought up the issues that Im experiencing, damage was being done to the batteries. Yet once again, according to the "professionals" they said I'd wear the batteries out sooner but they never warned me that I was causing permanent damage that could ruin the battery bank overnight. Things seemed to be working fine and the AGS was only supposed to make things convenient for me.

Ultimately, the AGS issues led me to catch the poor set-up on the controller (float, absorb time, amp hours) and the charger jumpers. So that's a silver lining right there.

Next on this list is
(1) verify the voltage detector is wired correctly
(2) battery temperature might be a issue that Im just going to have to wait out until spring arrives
(3) Im going to have a well recommended solar tech go thru the steps that Estragon posted

South Africa · January 2017

FWIW, just adding another idea.

I chose a small battery bank for a specific load with a slightly larger array to power selected things daytime. Battery backup is for purely at night if there is a power failure.

To ensure the batteries never go below a set SOC, I use the built in relay that is switched by the pre-set SOC in a Victron BMV Monitor. When SOC reaches X, the relay in the BMV Monitor switches which in turn switches larger 220v relays from solar power to utilities. Once batteries are re-charged next morning, i.e. SOC = Y, BMV relay switches again and I am back on solar power.

Did it like that to ensure batteries are always charged in case of power interruptions.

By using mechanical 220v relays there is a 1 second break between solar and utilities power so I need a online UPS. which turned out to be a saving grace. With electronic relays that switching delay can be as low as 10ms, fast enough that no break is detected between swapping sources. System works, UPS is a blessing in my case so maybe later.

Any case, more on point of this thread. The BMV's relay could also be use to trigger the start of a generator.

Measuring the SOC of batteries is quite an art.

WaterWheel · January 2017

A quick thought on using SOC meters as your only way to see the charge state of your battery bank.

Even with a well placed shunt SOC meters are only an estimate and the discrepancy can be cumulative.     With my system I trusted the SOC meter without checking SGs (something you can't do with AGMs) for about 7 weeks once when the system was only a few months old.        When I did check SGs after a 3 hr absorb and the SOC meter showing 100% the SG readings were 1.240 meaning my batteries were only about 81% charged (this was during the winter).        I monitored things for a few days with no charge controller changes and found that even though the SOC meter was showing 100% every sunny winter afternoon the SGs showed that I was actually undercharging the batteries a bit each day.

With AGMs it can be hard to check things but with time I've learned through SG checks that I need to adjust the charge controller settings every few days or weekly just a tad to keep the SGs above 1.267 or 95% SOC actual at the end of absorb.      Winter time requires a bit higher charge voltage ect.

Using only the SOC meter is not a good way to monitor battery state of charge.        I don't use AGMs but if I did I'd probably occasionally take all loads off them for 4+ hours and read the voltage with a good meter to verify that they were being properly charged.       I've found that ending absorb amps is a good indicator but you've got to be around to check the meter near the end of absorb.         For my 790 amp bank if the ending amps is around 11 amps my SGs will be around 1.266 (94% SOC).      I occasionally change my absorb time but I also set my charge controller to stop absorb if the battery charge amp draw drops below 8 amps to protect my batteries from overcharging.

Marc Kurth · January 2017

The ability to log voltage data is easy with many modern charge controllers. External data loggers can also be added. Good/bad trends show up very quickly with just a few battery bank voltage data points per day.

Having a snapshot of battery bank voltage every 4 hours is a powerful tool for diagnosis AND........ when discussing performance issues for warranty claims with a battery manufacturer.

Voltage data points can eliminate a whole lot of inaccurate guesswork.......

Marc

westbranch · January 2017

And with those voltage readings, if you have Watthrs expended and watthrs returned you soon get a sense of the state of your battery. For example if you used 400Whrs and put back in 600Whhrs in a day and your voltage when charging is finished and again when the battery rested 3 hrs, you will have some good information to proceed from... iff the voltage did not drop, then you probably have a healthy battery, if it did not hold it is failing, or there are other issue at play.... that is when you ask the Forum...

Jabroni · January 2017

Here are complete picswestbranch said:

Shunt? http://www.midnitesolar.com/productPhoto.php?product_ID=101&productCatName=Accessories &productCat_ID=5&sortOrder=9&act=pc

and here is where it fits with all the other bits in a complete system... it fits into the NEG wiring...
http://www.midnitesolar.com/pdfs/classic_simple_flow_diagram.pdf

Here is a bunch of pics of my system. I couldnt find a shunt. Maybe I missed it. Either way, fire away yall. Let me know if there is any closer pics you need to see.

Image: https://us.v-cdn.net/6024911/uploads/editor/bs/g85rijqkmact.jpg

Image: https://us.v-cdn.net/6024911/uploads/editor/9w/t3f6m6ovbtdu.jpg

littleharbor2 · January 2017

Too large of photos. Taking too long to load. Can you delete and resubmit in a smaller format.

westbranch · January 2017

Jabroni, what are the 3 black connections on the battery cell in the back row / right side?
it loos like the negative post (blue colour cap)

mcgivor · January 2017

Looks like Stevie Wonder and Ray Charles wired the system.
Battery terminals need to be cleaned, otherwise it's very hard to see/trace where anything goes.

Jabroni · January 2017

westbranch said:

Jabroni, what are the 3 black connections on the battery cell in the back row / right side?
it loos like the negative post (blue colour cap)

I believe they have to do with the MagWeb monitoring software. But Im not positive.

Jabroni · January 2017

mcgivor said:

Looks like Stevie Wonder and Ray Charles wired the system.
Battery terminals need to be cleaned, otherwise it's very hard to see/trace where anything goes.

No kidding. It's embarrassing. Thankfully you cant see all the missing screws and stripped out screws that held the cover plate on.

It's actually a miracle the system has lasted this long imo.

mcgivor · January 2017

One point that I noticed is the voltage reading at 29.84V, was it really cold when the picture was taken?

Photowhit · January 2017

Jabroni said:

westbranch said:

Jabroni, what are the 3 black connections on the battery cell in the back row / right side?
it loos like the negative post (blue colour cap)

I believe they have to do with the MagWeb monitoring software. But Im not positive.

They are Magnum's battery temperature sensors, they should be moved to terminals near the middle of the battery bank in my opinion stacking like this isn't helping to make a good connection. Magnum does say to attach it to the first batteries neg terminal. I can't see where this is mounted would make a huge difference, though often there are other items attached to the positive side. I guess Magnums don't share information? and need one for each charger? 1 for each charge controller and one for he inverter charger?

Estragon · January 2017

Looks like a shunt could be mounted on the back of the box under the plate separating the battery+ breaker in front from the bat- behind. You can just make out top of and bottom of shunt labels.

Looking at those battery posts I don't think it would take much more than a little wiggle by the installer to make an interconnect worse and cause the sudden worsening of your problem that you saw coincident with AGS install.

Estragon · January 2017

Also, it's hard to tell from the pic but the second from left battery on bottom row + connection bolt doesn't seem to extend past the top of the nut. Did the bolt snap? Crud under the lug?

Jabroni · January 2017

mcgivor said:

One point that I noticed is the voltage reading at 29.84V, was it really cold when the picture was taken?

Yes.

During the day it's about 40-45 F

And at night it's about 30 F

Jabroni · January 2017

Estragon said:

Also, it's hard to tell from the pic but the second from left battery on bottom row + connection bolt doesn't seem to extend past the top of the nut. Did the bolt snap? Crud under the lug?

Good eye.

Top half of the bolt is snapped. The installer noticed it the last time he was out here to diagnose the AGS issue. When and how it happened is unknown. But the installers are the only people to have ever laid hands on this system so they no doubt are responsible for it. There are so many little issues like that it's quite embarrassing.

I escalated the issue and went straight to the owner of the company who sold me this system. It just happens that his younger brother is the installer. We had a nice chat on the phone and here's basically the bullet points:

he is well aware of his younger brother's dismissiveness & passive aggressiveness when dealing with clients
he acknowledged that his brother has a tendency to do sloppy install work
he is not happy with the way the problem was diagnosed (or lack thereof... anyone who shows up with a ready printed quote trying to sell me more batteries has pretty much made up their mind on the diagnosis before even arriving on site)
he is perplexed at the sudden loss of battery performance but at the same time, wasn't dismissing my accounts of how it went down. We actually had a productive conversation trying to diagnose the issue where he was actually trying to figure the issue out while acknowledging the AGS install coincidence. And in turn, acknowledged that it's possible that the AGS wasnt installed properly and thus getting a wrong voltage reading
he also acknowledged that the ambient temperatures could be the culprit, being that the batteries were so new and had seen such little use up until November
he didnt think the BMK & charge settings were that big of an issue but did find it "interesting" that his brother claimed I had requested those settings. Which of course is blatantly false. I had no idea what the geni charge settings were until troubleshooting with the Magnum rep

All in all, he didnt blindly defend his brother. It was quite the opposite. All the frustrations I described in dealing with the installer was something the older brother knew I was being honest about because he has the same issues in dealing with the younger brother. So oddly enough, this actually helped my credibility with the owner.

He's pulling the younger brother off the project permanently and sending someone else out to review the work. In addition, Im still having a outside tech come review the work as well. Which has been a pain in the butt being that my local is so remote it's not easy to get appointments.

I'll keep everyone posted as this progresses.

batteries abruptly stopped holding voltage under load...

Comments

Categories