|
| Author |
Message |
Nobadays
Member
|
# Posted: 2 Jun 2024 14:48
Reply
I'm still trying to figure out why my new battery bank is going flat every night with a 40w draw.
I installed 4 x 6v/225ah Vmax Charge Tank batteries last October. They seemed to have no trouble running my Starlink (average draw of ~40w, snow melt turned off) through October and November. Sometime the end of December or in January the system started shutting down in the early morning hours until the sun would charge the bank back up. I know snow on the panels was a factor and fully expected them to come back to full performance as spring came and the snow left.
So far I have reset the SCC from "smart absorb" to a fixed time of 1 hour each day. I have also done 2 equalizations @ 29v for 1 hour and one equalization @ 30v for 1 hour. Left the light on last night, a steady 40w draw. Hit low battery cutoff, I'm set at 23.8v (slightly below 50%) around 5am this morning.
Here is my math, and confusion.
These batteries are 225ah, tied together to create 1 x 24v bank. This equals 6.2kwh. Since my cutoff is ~50% I still should have 3.1kwh. A 40w draw... I'm using 14 hours from end of day charging until beginning off day charging... by my calculations is 560wh. So theoretically I would have ~5 days.
What am I not seeing! Thank you!
|
|
Steve_S
Member
|
# Posted: 2 Jun 2024 16:08
Reply
You have done all the right things and Equalization should have certainly woke them up. The thing is, you have 4 in series and if just one has become unreliable for whatever reason, they will all suffer the consequence. I'll assume you checked all the terminals & connections to ensure they are clean & tight.
It may be prudent to check each 6V battery independently to see if one is going bad. Diagnosis can be a PITA.
|
|
Nobadays
Member
|
# Posted: 2 Jun 2024 16:31
Reply
Quoting: Steve_S It may be prudent to check each 6V battery independently to see if one is going bad.
Thanks Steve... yes all connections are good at the batteries. I don't recall checking at the fuse or at the inverter so I'll look at those.
I did check them, after a full charge, with a 6v/12v load tester. Unfortunately the 6v scale doesn't have values on it, just good or weak. All of the batteries were just barely into the "good" , pretty much right on the line. I checked voltage on each after the load test (Unfortunately I didn't take notes) and I recall one of the batteries dropped a bit more than the others.
So I'm guessing my math is good.
The next step is to fully charge again and do another load test and take notes. I do read that Vmax has good customer service so if perhaps I do have a bad battery they might replace it. Fingers crossed! I love my LiFePo4 but these batteries are in my unheated shop and I want to be able to use them all winter to run Starlink for access to my cameras.
I'm headed to Oregon for the last grandchild's graduation so going to be gone a couple of weeks. I'll play with them more when I get back.
|
|
gcrank1
Member
|
# Posted: 2 Jun 2024 19:09 - Edited by: gcrank1
Reply
Fwiw, on my inherited 12v system I had 4x 140?ah ups agm bats that had been poorly paralleled (thus abused).
I 'reconditioned' them all 2x and hooked up fresh, it worked decently for a while them started to fall off much like you describe. Sorted out the failed bat and ran on 3, then 2, then 1 as they each died (that is one advantage of parallel bats).
The bank will only work as well as the weakest battery.
Also, Lead Acid batteries, which includes AGM, want to be kept fully charged. Yeah, we can't do that with use, so we Have To recharge them fully asap after each use or they sulfate, which kills them.
I think you were depleting them slowly over time without getting a full recharge which has compromised at least one, maybe all.
|
|
Brettny
Member
|
# Posted: 2 Jun 2024 19:48
Reply
I take it they didnt always do this? I also think letting the system shut down at 50% is beating the snot out of them...especially in cold weather.
|
|
ICC
Member
|
# Posted: 2 Jun 2024 20:01
Reply
How cold does it get?
|
|
paulz
Member
|
# Posted: 2 Jun 2024 21:26
Reply
Quoting: Nobadays Hit low battery cutoff, I'm set at 23.8v (slightly below 50%) around 5am this morning.
And full charge night before?
|
|
Nobadays
Member
|
# Posted: 3 Jun 2024 03:15 - Edited by: Nobadays
Reply
These were all new batteries last fall. For the two months I was able to observe them and from what I could tell during December... the Starlink never went down, they were fully charging every day. Since we have been back, middle of May, fully charging every day. They were absorbing by 10:30am today, even after shutting down early this morning. There is 1000w of solar and a Victron SCC keeping them charged.
It was a fairly mild winter with low temps mostly in the teens from what I am told with only a few days dipping into negative numbers. Those negative days are almost always clear so charging happens.
They should only get down to the 50% mark when clouds or snow prevent charging. It is rare here that this is more than 3-4 days before its cleared up and charging. If snow covered the panels it may have hit the cutoff but then it wouldn't have came back on until the sun cleared the panels. The panels get minimum 6 hours a day of direct sun so they charge back up quickly.
I need to do some more load testing and see if one battery is weak, drawing the rest down. My understanding is AGM batteries are not as "tender" as FLA when it comes to drawing them down to 50% repeatedly....that said drawing them that low shouldn't have happened with a minimum of 3 days reserve and theoretically 5 days at the 40w draw.
Thanks for helping me kick around ideas and what to look for!
|
|
|
paulz
Member
|
# Posted: 3 Jun 2024 04:38
Reply
The dead winter months were tough on my bank too, fortunately the LFPs seem to have taken the low charging in stride. Spent many winter months at the cabin, and even just running the fridge and minor things I had to run the charging generator daily to keep the bank up. Maybe just the fewer hours of sunlight, lower sun angle or increased shade.
Lately I’m basking in solar power with no changes to the system. Haven’t run the charge generator in over a month, using way more juice (wash machine, toaster, microwave..,) without worrying about it, and still full charges by days end.
Makes me wonder if those winter days took a toll on your AGMs.
|
|
Steve_S
Member
|
# Posted: 3 Jun 2024 10:34
Reply
Sadly, the quality of Lead Acid / AGM have gone down quite a bit in the past 5 years, even the heavy Solar Batteries are not what they used to be... Thinner lead plates etc.. Car batteries took the worst hit in quality. None will take the abuse they used to be able to.
I have a couple of friends who work at "Total Battery Co" and they move a LOT of batteries. They deal with tiny utility to Huge Heavy forklift & commercial batteries and they tell me how bad things have gotten pretty much across the board. Their scrap piles have a lot of lead acid that are 3 years old and up, they say it used to be mostly 5+ years but not anymore.
Sadly, all it takes is 1 flaky battery in a bank to kill the bank. That single Baddie is also pulling down & harming the rest by keeping them low and there's the rub. The longer the sick batt is in the bank the more "soft harm" is being caused to the rest.
That is one of the reasons I replaced my FLA Bank of Rolls Surette S-550's with LFP. I had 1 6V go very bad which I did not notice quickly as I was chasing the quirks, by the time I discovered the one Bad Batt the damage to the rest was Fait Accomplis - essentially murdered $3500 worth of Heavy Lead.
The WORST PART !
That Heavy Lead cost me $ 3500 (Today's price $3,840 + taxes) for 445Ah/20h or 222AH useable capacity. Each of my 24V/280AH LFP Packs cost me $1400 CAD to my door.
These packs are DIY, so I put together the parts.
- 8x EVE 280AH V3 Cells (A-Grade Matched & Batched with test reports)
- 1x Heavy Steel Casing w/ 200A DC Breaker
- 1X 200A JK-BMS Inverter Edition with 2A Active Balancing, LED Display (detailed status etc). Can interact with Inverter/SCC etc. Also can control a warming element to keep Batt Temps above 32F (heating silicone pad is about $40)
-- Assembly Time is 4 hours after Top Charging new cells (which is time consuming)
NOTE: I import everything from my #1 Supplier in Shenzen (as do many others @ DIYS) so it takes about 45 days to get but includes all costs & fees etc. All of this is available in the US from other vendors but for a significant markup and none provide A-Grade Factory matched & batched cells that I am aware of. Impatience can be costly too.
Basically, it was all a Very Expensive lesson learned but then LFP etc was crazy insane $ when I first bought my batteries in 2015.
|
|
gcrank1
Member
|
# Posted: 3 Jun 2024 12:54
Reply
I have to wonder what you are using to determine the State of Charge, it may have been showing full charge when the bats were not full? That your bat tester showed them all barely into the good (color bars rather than a real voltage?) is telling.....my interpretation of that is they are all 'barely good'.
And even if your tester was a good voltage tester, a bat can read high on the scale but have no lasting beans to do the job.
As my bank slowly faded to not have a nights worth was my reality check; fact was that yes, they didn't have enough beans anymore. I knew they were old and fading so I didn't have to face my own denial, if they were pretty new like yours I would have tried to find every other possibility.
Ime, weak bat(s) are the cause of 'weak battery syndrome', especially with lead acid.
|
|
Nobadays
Member
|
# Posted: 3 Jun 2024 18:33
Reply
Quoting: gcrank1 I have to wonder what you are using to determine the State of Charge,
Batteries are floating @ 27.12v according to the SCC, this after an hour of absorption @ 29.00v.
When I load tested them I used a standard 100amp load tester, HF... prior to load testing I used a Klein MM to check voltage on each battery. That voltage compared favorably to what the SCC was reading.
Quoting: gcrank1 As my bank slowly faded to not have a nights worth was my reality check; fact was that yes, they didn't have enough beans anymore. I knew they were old and fading
My batteries have been in service less than one year. They shouldn’t be "fading."
This off the Vmax FAQ: "under Cyclic Service Life- a battery subjected to a 50% discharge every day, could last about 3 years or 5-6 years if discharged every other day, If discharged by 25% daily, battery could last 4 years and 7-8 years if discharged every other day."
I might have a bad battery in the bunch... I was able to get testing recommendations/procedure off the Vmax site so...
When I get back from Oregon/Arizona I'm going to dig into this issue again.
|
|
gcrank1
Member
|
# Posted: 3 Jun 2024 19:39
Reply
Agree, they 'shouldn't be', but the fact that you don't get the run time says otherwise.
Its like when you go to start the car and it barely turns over, then the click-click of the solenoid.
You check the battery voltage and its now down.
You charge it up and it reads good.
Try to start the car and barely turns over, then the click-click.
No beans 
|
|
Nobadays
Member
|
# Posted: 30 Jun 2024 13:28
Reply
Well I've been working on other projects and just got back to the AGM battery issue yesterday. Mainly because I ordered/received 2 new LiTime 12v/100ah, self-heating LiFePo4 batteries, and installed them.
I have built an insulated box around these batteries to help hold heat in as they charge/discharge and when the heaters kick on. These come pretty highly rated with several folks out there who have torn down and tested them. They have low-temp cut-off as many LiFePo4 batteries do but the difference is when the SCC begins directing a charge to the batteries if they are in low-temp-off, the incoming current is directed to the internal heating pads. When the batteries reach 41*F the BMS then resumes charging. No "battery juice" is used to heat, just the incoming power from the solar array.
On the AGM'S, I need to find someone with a 6v charger so I can independently charge and test the batteries. I did note that 2 of the batteries were seeping electrolyte. One was obvious with moisture around the negative post. The other I couldn't tell where it was leaking from but discolored plywood where it was sitting says it has leaked quite a bit. I know they hit 50% charge several times last winter. I'm hoping they did not freeze and crack when they were low.
|
|
ICC
Member
|
# Posted: 30 Jun 2024 13:39
Reply
Insulated boxes and those heaters should be pretty good.
AGM electrolyte seepage is not good 
|
|
Steve_S
Member
|
# Posted: 30 Jun 2024 13:49
Reply
I dunno if you can open the AGMS and check the "juice" inside. If it is discoloured and sorta tan/brownish then you got real issues. Seepage is an extremely bad sign and a cause for concern. If you want to see it happen, do an Equalization Run on them, they'll cook up & leak for you.
You'll quickly fall in love with your LiTime Batteries. BUT do be aware of the Charging profile and KEEP IT MODERATE !
Have a PEEK HERE:
https://diysolarforum.com/threads/steve_s-tech-my-final-config-for-midnite-classic-sc c-jkbms-w-large-bank-works-amazingly-well-includes-charge-profile-bms-config.42549/
|
|
Nobadays
Member
|
# Posted: 30 Jun 2024 15:11
Reply
Thanks for this Steve! I did lower settings a bit as the settings that came with the batteries seemed too high.
They recommend: charge/bulk/boost of 14.4 (28.8) I felt this was on the upper end for sure. I lowered it to 28.4 but was questioning that as being too high @ 3.55 per cell. I'll lower that to the 27.5.
Float wasn't mentioned in their booklet, I set at 27.2.
Rebulk wasn't mentioned either. I set at .5 volts, I'll raise that to your recommendation of 1.8v ( correct?)
Discharge setting pretty well mirror yours.
Thanks again!
|
|
Steve_S
Member
|
# Posted: 30 Jun 2024 15:44
Reply
Happy to help.
A LOT of folks ran into quirks, foibles & issues and I give the the settings to use, they pt them in and all of a sudden Nice & Polite behaviour 95% of the time. That 5% is usually because of something else.
SCC's, Inverters & worst of all AIO's have various settings with varied naming conventions but in general most have similar functions & settings but not necessarily all. I find it really depends on the "Quallity" of the gear. For instance a Midnite Solar or Victron SCC compared to a Cheapo like PowerMR or the like Value stuff.
|
|
ICC
Member
|
# Posted: 30 Jun 2024 15:57 - Edited by: ICC
Reply
Quoting: Steve_S I find it really depends on the "Quallity" of the gear. For instance a Midnite Solar or Victron SCC compared to a Cheapo like PowerMR or the like Value stuff.
Agreed!!
|
|
Nobadays
Member
|
# Posted: 30 Jun 2024 16:18
Reply
I've given up on cheap solar equipment. Our cabin is all Schneider and my shop (where the new LiFePo4 batteries are) is all Victron.
If these batteries function as well as the Chins we have in the cabin, I'll be thrilled.
One question while we have Steve here.... LiTime recommends when batteries are in series that every 6 months to hook positive to positive and negative to negative and leave for 12-24 hours to balance the batteries.
Should I be doing this with my 24v/200ah batteries I have 2s2p for my cabin 48v system? I admit since putting them in service 3 years ago I haven't done anything/any testing with them.
Thanks
|
|
ICC
Member
|
# Posted: 30 Jun 2024 19:15
Reply
My home with its shop is all Outback charging, inverting and is networked together for programming, controls and data observation. A mix of Midnite Solar and Outback lightning surge protection. It has basically run itself for almost 15 years.
My cabin used to be Midnite Solar with a Samlex PSW in erter. Now Victron charge controller with the Samlex PSW in erter. Midnite Solar lightning surge protection.
. I can understand the theory behind connecting LFP batteries in parallel to balance, but I have never felt the need. My home runs on a 48 volt bank of series/parallel large GBS prismatic LFP cells. Right now the maximum cell voltage variation is 0.008 volts. If it was much over 0.01 then I might start thinking about a top end rebalance. 6 years old battery bank. That is my thinking but I am not certain what is considered to be the very best practise.
My much newer cabin battery is running a Vdiff of 0.006.
|
|
Steve_S
Member
|
# Posted: 30 Jun 2024 19:35
Reply
Schneider, Victron all excellent Tier-1 product. If I was to start over again, it would all be Victron. Well except the batteries of course but they would have Victron Canbus and interact. All of my BMS' can do that +.
I will say this "plainly" I am not one to suggest using LFP packs in series. I am firm believer in using Native Voltage for a system. There are far too many possible issues that can happen.
! Sorry if you are reading this on a phone. 
Chins & LiTime are "fairly" good batteries from the reviews/teardowns I've seen and from what I have read on the forums from users. Like others they have different options/version, heated or not, bms apps, or rs485/canbus etc.
Now if you've put them in series properly (wire lengths and the wee details) and can view the BMS info via app that will tell you what's going on and if there is anything amiss. If there is, there are different things that can be done depending on the issue. The "Swap Them Around" is a Lead-ism left over, there is some merit with regards to "Troubleshooting possible issues" but that's about it with LFP.
IF you cannot use a BMS APP to view the information within, your only other option is to check each battery packs voltage at the terminals.
1) You need a good Multimeter that is at least 2 decimal accurate DC.
2) MARK-LABEL your batteries !!! LAMO just had'ta ! so you know what position etc
3) (with inverter/scc etc OFF) Check the voltage while still connected to other batts, then recheck after batts are all disconnected from each other. Do note the voltages.
3) Let them sit for an hour and retest voltages & note.
You will likely see a small difference between them when you disconnect them. It shouldn't be much, something like 0.10-15.
After an hour, any "surface charge" (simplified) will be gone and they may all be a bit lower but again not by much, but they will be varied by now... You will see which is weakest by it being the lowest. Maybe switching positions may sort that because they won't all be getting the exact same volts/amps during charge nor providing the same during discharge. This is due to the resistance of the wires and any other connectivity things.
Unlike FLA/AGM - Lithium Based batteries are more sensitive internally to the millivolt & milliohm. So calibrating / compensating for the voltage losses from SCC to batteries and back to the Inverter with careful measurements is important.
Say the SCC Reads that it's ending 28.0V (3.500vpc) but only 27.2V (3.400vpc)s actually reaches the pack, when done that battery will be at 92-94% at most of the working voltage max.
For the Inverter side, the inverse issue its to ensure that it shuts down "before" the batteries force it down. It is "not" good to bottom the LFP Packs, you stand a chance of bricking them. So you have to adjust the voltages for cutoff between the "actual" at batt terminal and at Inverter Terminal.
Believe it or not, it is genuinely the Devil in the Details but once you get it right, you won't have anything to worry about oor chech/maintain constantly anymore. QWuick peak on your system App and off ya go.
Batteries in Series:
A lot of unpleasant things can happen, all dependent on the BMS & other gear. The biggest issue is that if 1 Pack in a series "fails" for any reason the entire set drops too. That's OK but comes with a gotcha. Pending on the types of BMS used some can handle it ok while others cannot, unpleasantly so.
Scenario 1 discharge.
12V+12V=24V, Pulling 20A by inverter. Batt-1 cuts off, Batt-2 gets 24V demand/20A = Overload shutdown. BUT some BMS' may actually attempt to deliver that and pop their FETS (usually dumb BMS that cannot be programmed or set)
Scenario-2 charging
SCC is pushing 27.6V/60A
Batt=1 cuts off due to an OVP. Batt-2 get's hit with that 27.6V and cuts'off because it's over spec. OK. But again some BMS may take a bit longer or even attempt to take it and pop their FETS.
These 2 scenarios are actually the most common seen. Now the better companies like Chins/SOK will usually state on the battery - datasheet is they can be put in series, sometimes right on the label. The cheaper bargain basement ones "No Freaking way in hell do it".
Apologies for the Long Response (especially for phone people LOL like War & Peace)
Hope it helps, Good Luck.
|
|
Nobadays
Member
|
# Posted: 30 Jun 2024 21:24
Reply
Thanks Steve. The LiTime I have do not have an accessible BMS, no Bluetooth. They do however state the batteries can be placed in series, " up to 16 identical batteries ". That I would not do!!
I couldn't find, with a similar price point/ reputation, a 24v internally heated battery. This is why I went with 2 x 12v LiTime batteries. (Chins came to market with 48v batteries shortly after we had bought the 4 x 24v for the cabin system.)
So do you think paralleling like batteries occasionally has any merit for balancing them? Or as ICC has found, good cells with a good BMS kind of makes this unnecessary.
I used to test our Chevy Volt, MNC lithium once in awhile and never found more than .01 volt difference between the cells.
Thanks
|
|
Steve_S
Member
|
# Posted: 30 Jun 2024 22:36
Reply
If the battery packs are properly wired etc, its not likely they'll need balancing... check the voltages and that will tell the story. LFP can "wander" a bit internally depending on cell grade.
The Volt NMC Battery Packs were really good matched cells so that's not surprising. Voltages are a bit funky in comparison and eliminate or hampers some gear.
I run my battery bank with parallel batteries but it's all well matched and there's generally no more than 0.05V between the packs, sometimes a bit more under a heavy load.
|
|
Nobadays
Member
|
# Posted: 30 Jun 2024 22:43 - Edited by: Nobadays
Reply
Thanks! I’ll check voltage on my 24v batteries at some point before fall and see how they are doing. I know Chins has cell balancers on the BMS they use. Just need to see what if the 4 batteries are equal. I have them wired to busbars then wires from there to the inverter and SCC.
Thanks!
|
|
|
