Troubleshooting

[Midrover170]

Hey all -

So, nothing much good to report after spending a day at the cabin. System had a full week to charge up with ideal conditions (minus it being November) and I'm still far from a full charge. Here's a recap:

• 4 panels at 780 total watts (24v)
• TriStar TS-45 (PWM) w/sense wires and RTS
• GoPower! GP-SW 2000/24
• 4 UB8Ds (12v) @250Ah, 500Ah total

Here's what happened when I arrived:
I got to the cabin just before dark and turned the inverter and breaker on. Like last weekend, I started losing volts rapidly. In just over an hour, I went from 27.2v (my Float setting) to 25.6 v, where I kind of stabilized. System is programmed to shut off at <25.4v, so not great. I was only using LED lighting during this. Around midnight I went back to check and with no load the voltage actually crept up a hair. Battery sense was reading "valid" with no faults or alarms. Last stat is from my Ah reading. Over the last 4-5 days, the system claimed to have received 33Ah from the panels.. if I'm reading that right.

And today:
While the sun was out today and during PWM phase, the CC was hitting much higher than it's programmed to. I have absorption set at 14.6v, and the CC was hovering around 29.9v any time the sun was shining. For some reason, this trips my inverter because it thinks it's getting more than 30v. Battery temp was around 55 degrees and the CC was reading 73.

Basically, I'm at a loss as to what to do. All these questions are flying around my brain while I try to protect my investment:

• Is the higher-than-programmed CC voltage due to something the RTS or sense wires are triggering, and why would a feed with less than 30v trip an inverter that can handle 30v? The "battery" reading on the TS-45 is about .1 higher than the "battery sense" reading; for example, at one point today "sense" read 29.92 while the "battery" reading said 30.03. During that high PWM cycle, battery temp was reading 55 degrees and the CC was at 73.
• Is my main issue a battery bank too large for my PV? If I need more, I have two 12v panels not being used. Is it possible to series those and have them feed into the 24v system? These extra panels are HIGHTEC HS-M-B-150's which were donated about a year ago. Specs:
  • PMax 150
  • Vmpp 18.21
  • Impp 8.20
  • Open Circuit Voltage - 22.92
  • Short Circuit Voltage - 8.76
• Does it make more sense to take one of the battery strings offline, especially given my really modest loads?
• Or, does it sound like I simply have bad batteries?

I don't know if I'm in a pickle, but it kind of seems like it. THANKS for any help. I have to wait until Monday to call the supplier so I'm anxious to say the least.

[KMac]

What's 100% state of charge supposed to be on those? What you end with on float isn't it. If I understand it, you are still at 25.6 when the inverter and small loads are on. That is 12.8v per 12v battery. Do you have a state of charge chart? For mine, 12.7v (25.4v) is 100% charged.

I have Trojan flooded, and if mine are at 12.7v (25.4v), with inverter on, and any sort of load (at room temp), then I'm happy. When I show up to the cabin it might be higher than that (25.9v or so), but that is just excess that burns off very quickly and isn't a sign of a problem. Float voltage is not 100% state of charge voltage. It is more.

[Midrover170]

What's 100% state of charge supposed to be on those? What you end with on float isn't it. If I understand it, you are still at 25.6 when the inverter and small loads are on. That is 12.8v per 12v battery. Do you have a state of charge chart? For mine, 12.7v (25.4v) is 100% charged.

http://www.mrsolar.com/content/pdf/Universal/UB8D.pdf

[KMac]

I know little about batteries and less about AGM, but a quick Google Search of "AGM state of charge chart" led me to an article saying AGM 100% starting charge is 12.8v, which is what you have when it settles down even with a slight load (inverter and LED).

In sum, I don't see that as any sort of a problem. Someone please correct me if I am missing something here.

The remote temperature sensor located on your batteries gives the CC information so that it can adjust the charging voltage. My flooded Trojans will charge at over 30v in winter. In summer it is much lower. The programming you did of the CC included a number for how much the adjustment should be from the referenced temperature. You can measure the temperature and calculate it by hand, and that should be the same as the "target voltage" on the CC live data page. Are you saying the charging voltage is higher than that target voltage?

[KMac]

The inverter trips because, as you wrote, the terminal voltage is over 30v (battery sense voltage is irrelevant it this). My inverter also has a ridiculously low high-voltage shutoff. And it is not adjustable. So, I just lowered the max charge voltage and since it is float charging most of the week I assume they get plenty full. In any event, it only does it when really cold and charging voltage peaks over the high-voltage shut off of the inverter. Not a big deal with my flooded acid batteries. Again an AGM expert might chime in as to whether you could do the same.

[Midrover170]

OK, thx KMac. When talking to Universal Battery, they said 12.7 (25.4) was 70% and to avoid anything below that. It does look high after viewing some AGM SOC charts, but maybe unique to this battery? 12.8 being full seems like a really thin margin for me. I might have to lower my max charge as well. I'll wait and see if someone with more AGM experience weighs in.

[KMac]

The manufacturer info you linked to had something akin to the state of charge charts. It is the approximate residual charge based on voltage. As you know, this is temperature dependant too, hence the range. Also, that is measured open circuit, meaning NO loads, not even that big 2,000 watt inverter. So, turn it all off for several hours and check voltage then, and then adjust your math for battery temperature, and compare to that chart, and I suspect you will be spot on and your batteries fine.

[Midrover170]

Thanks. I sure hope so.

[Mike90250]

Either the batteries are sulphated from chronic under charge, and just can't deliver. or they have been overcharged so much @ 30V there is little active material left.
I would think 30V on an AGM would be fatal

[Midrover170]

Either the batteries are sulphated from chronic under charge, and just can't deliver. or they have been overcharged so much @ 30V there is little active material left.
I would think 30V on an AGM would be fatal

Hey (again) Mike. Setup of this system has been interesting.

These came out of the box about three weeks ago. I don't think the batteries have received over 30amps in their lifetime. I've monitored them closely: No gassing, no smell, nothing visible. CC is reassuring on that regard as well.

It seems as though I have plenty of panel to charge them up, but that voltage just drops off even under the slightest load. Possible that they set on a shelf way too long? If I call the vendor and gripe about it on Monday, I just want to make sure I'm not overlooking or missing something. And, worst case, if I have to drag 600 pounds of battery back to town, I want to make sure it's warranted.

[Mike90250]

Then you have to do a load test on the system. What's the C/20 spec on the batteries ( on the spec sheet near the 250ah spec- make sure it's the c/20 rate, not c/5 or something) Then apply a compatible load on the inverter (a bank of incandescent lights ) and see how long it takes to get down to 50%

So if the 12V, 250ah battery is rated at c/20 the discharge rate you are looking for is going to be 12 amps for 10 hours. Since they must be in some series/parallel, split the parallel strings apart, and test each string by itself, record the voltage of each battery at start, and every 30 minutes under load.

To get a 24V, 10A drain, you need 240 watts plugged into the inverter. (10a is close to the 12a spec, and there will be a bit extra from inverter losses)
If you make it to 10 hours then test the other bank, and record voltages again, You will find the dud batteries this way. You are only looking to discharge 50%
Now you have data to take to the battery store with you.

[KMac]

Again I ask, what is 100% charged voltage? On the spec sheet chart you attached 12.8v or a bit above seems to be it.

Nothing you described sounded like bad batteries from sulfation (they are new) and not overcharged. I think you are concerned when you watch the surface charge burn off, but it burns down to a bit over 12.8v, and look at the chart you sent, it appears that's fine.

Do a sort of load test and I think you will see that surface charge is irrelevant, there is voltage sag whenever under load, and they will bounce back up in an hour or two to a proper range when load is off.

[Sunking]

Like last weekend, I started losing volts rapidly. In just over an hour, I went from 27.2v (my Float setting) to 25.6 v, where I kind of stabilized..

So WTF is your problem? 25.6 OCV is 100% plus SOC. It does not get any better than that.

[Midrover170]

Christ on a cracker. That's what I'm doing here, trying to LEARN. I've never understood these forums where the most frequent commenters are also the most angry. If you don't like helping others out, why do it?

When I have an mfg. tell me 25.4 is 70% in addition to a "60-80%" charge alarm as soon as I turn on the inverter, that doesn't seem right. Call me crazy, an idiot, whatever you like, but I don't seem to be throwing false alarm.

[KMac]

I guess the most direct style got the point across. Be grateful as he knows what he is talking about. Look at the chart. Ignore loss of surface charge volts. The mfg. Rep. v. The spec sheet. Okay, follow the chart. Try some big healthy loads and burn some power and see what you get. Good luck.

[Sunking]

Well Cheese n Rice lets get you educated.

First forget voltages, they do not have much meaning which was your 3rd big mistake. Your first blunder is using AGM, and second blunder was using 12 volt batteries in parallel.

OK let's start with State of Charge voltages aka SOC. Completely meaningless on a working system. Tells you ABSOLUTELY NOTHING. I can take a battery., a good battery, fully charged and make it read 10 volts. By your logic and your Inverter logic a completely dead battery. Complete nonsense. I can take the exact same battery, discharge it completely and make it read 13.0 volts and you and your Inverter will think it is fully charged. Again complete nonsense.

So what is SOC voltage good for. It is only a relative indicator on a fully rested battery. By fully rested I mean disconnect the battery from everything, put in a room at 75 degrees and wait for 24 hours before measuring the voltage. Then it is only a Ball Park indicator, and even if the voltage measure OK does not mean you have a good battery. Anyway not sure where are you are getting your info from but for AGM battery here is the chart.



OK AGM batteries do have applications in a Battery systems. However they must be justified. This will really piss you off. By the time the batteries reach end of life, you over paid 400% more than you had to. AGM batteries cost twice as much as FLA, and only last half as long. You paid $1800 to $2000 for your batteries, ratther you use them or not will need replaced in two or 3 year. For $800 to $1000 you could have bought good FLA batteries that last 4 to 6 years.

Lastly you are stuck inside a 12 volt toy box. Batteries are 2 volts, not 12 volts. Last thing you want is to use parallel batteries. That will significantly cut your battery service life in half. If you needed 24 volts @ 500 AH you should have bought 500 AH batteries. There is no such thing as a 12 volt 500 AH battery. If they exist you could not lift them. You should have been looking at 4, 6, or 8 volt 500 AH batteries. Problem is you are stuck inside that 12 volt toy box. So next time you buy batteries here in a year or two try a good 7 year battery like Rolls S-550 6 volt 550 AH They will run circles around your batteries.



FWIW a 24 volt AGM battery 100% SOC = 25.6 volts @ 80 degrees F or 26 degree C. As temp goes up, so does voltage. As temps goes down voltage follows as you can see in the charts. What voltage did you say you measured?

[KMac]

One more thing to keep in mind. Batteries are limited in duration both by cycles and by years. For a weekend only cabin you will cycle them, what, 100 times a year? At that rate they will die of old age, not cycle number limitations. So use them on the weekends. Don't be stingy or worried. They will die but more from age than use in a weekend scenario.