Hey guys, I have 400ah (48v) batteries installed on a 5kva Axpert King, but noticed my battery capacity dropping very fast. 2 and a half hours after sunset, with the batteries fully charged, it went over to Eskom. 

I have attached a screenshot of my settings, unfortunately the video taken from the app is too big

Please help, 

Thanks

Johan

 

What was your load at that time, and which batteries do you have.? 

Ive attached a picture of my batteries (Sorry Im still new to this, so I thought you will understand it better). 

I have 4 in series, and 2 banks of 4 in parralel, giving me a 400ah backup (So 8 cells in total)

 

My average use was between 500 to 600W (around 11 amp)

Edited July 8, 20206 yr by JohanDbn

Here is my screenshot from watchpower at 8pm. Note my batteries were full at sunset

Edited July 8, 20206 yr by JohanDbn

8 hours ago, JohanDbn said:

Hey guys, I have 400ah (48v) batteries installed on a 5kva Axpert King, but noticed my battery capacity dropping very fast. 2 and a half hours after sunset, with the batteries fully charged, it went over to Eskom. 

How old is the batteries? How do you charge them? If you use panels, how many panels do you have? 

I am not sure about your back to grid voltage setting. 46 is too low in my opinion ( @Coulomb?) . I would not allow my battery bank to go down to 46volts. 

With Trojans and most other Lead acid batteries, if you reach 46 volts while drawing 500 watts, they are either undercharged or old and on their way out. 

LAstly, voltage is not an accurate indication of SOC, but in this case I worry a bit. 

Try and measure the voltage across each battery when they are at rest (no load and not charging). If there is a big difference on one or a few of them that may indicate a damaged battery.

1 hour ago, Jaco de Jongh said:

I am not sure about your back to grid voltage setting. 46 is too low in my opinion ( @Coulomb?) .

Sadly‚ 46 V is the default. I'd certainly not want to go below 48 V for lead acid; I'd want to stop at 50 V rested, which might be more like 48 V under load.

@JohanDbn, did you regularly let the battery discharge till it cut off during load shedding? The 42 V default cutoff for lead acid (10.5 V per 12 V battery module) is a theoretical state that should not be visited ever, except for coming up with a marketing capacity for the battery. Just a few excursions that low will ruin a lead acid battery.

That plus running down to 46 V before switching to utility has likely worn them out.

It's really sad how easy it is to ruin a lead acid battery, and how the specifications and inverter-charger defaults are designed for longer run-time (during warranty) and shorter life.

2 hours ago, Jaco de Jongh said:

How old is the batteries? How do you charge them? If you use panels, how many panels do you have? 

I am not sure about your back to grid voltage setting. 46 is too low in my opinion ( @Coulomb?) . I would not allow my battery bank to go down to 46volts. 

With Trojans and most other Lead acid batteries, if you reach 46 volts while drawing 500 watts, they are either undercharged or old and on their way out. 

LAstly, voltage is not an accurate indication of SOC, but in this case I worry a bit. 

The batteries are brand new. The last 4 were installed yesterday. As for the panels, I have 12 400W panels. I saw last night 46 is about 40%,also thought that too low

1 hour ago, Coulomb said:

Sadly‚ 46 V is the default. I'd certainly not want to go below 48 V for lead acid; I'd want to stop at 50 V rested, which might be more like 48 V under load.

@JohanDbn, did you regularly let the battery discharge till it cut off during load shedding? The 42 V default cutoff for lead acid (10.5 V per 12 V battery module) is a theoretical state that should not be visited ever, except for coming up with a marketing capacity for the battery. Just a few excursions that low will ruin a lead acid battery.

That plus running down to 46 V before switching to utility has likely worn them out.

It's really sad how easy it is to ruin a lead acid battery, and how the specifications and inverter-charger defaults are designed for longer run-time (during warranty) and shorter life.

This was the first night I ran them,so fortunately I would not have caused damage. You suggest I increase it? 

Based on the manual, I should be getting much more backup time than 2 or 3 hours. All I want is for these to carry me through the night on base load,looking at my installation that should be possible. 

2 hours ago, DeepBass9 said:

Try and measure the voltage across each battery when they are at rest (no load and not charging). If there is a big difference on one or a few of them that may indicate a damaged battery.

Disconnect each one as I test, or can I keep them connected? 

Did a check on the batteries. They all seem fine. Ranging between 13 to 14.5 volt all, and overall 55V, so seems the batteries are still hundreds. Could the inverter be reading wrong?

2 hours ago, JohanDbn said:

This was the first night I ran them,so fortunately I would not have caused damage.

OK, sorry for the scare. I hear so many stories of lead acid only lasting a few months; I assumed that this was yet another one of those.

Quote

You suggest I increase it?

Yes. Perhaps try 48 V. Unfortunately, that will cause it to switch to utility even quicker.

12 minutes ago, JohanDbn said:

Did a check on the batteries. They all seem fine. Ranging between 13 to 14.5 volt all, and overall 55V,

That's quite a variation, if they were all measured at the same time (or within a minute or two). It also can't be with no load and rested for an hour or several; that part is important. It's not vital to disconnect them, and I know that's a pain, but try to pick a time when the loads are very light and there is zero charging.

2 hours ago, JohanDbn said:

I saw last night 46 is about 40%,also thought that too low

The 46 V or the 40% too low? 46 V or 11.5 V per 12 V module, if rested, is a very low state of charge; they should not spend much time there at all. Like three times in their entire service life. BTW, the 40% reported by the inverter is a very rough estimate based solely on the battery voltage. The low DC cutoff voltage (setting 29, defaults to 42 V) is considered -2% SOC; for every tenth of a volt above that they give it 1% more SOC. So 46.0 - 42.2 = 3.8 V = 38 tenths ≅ 38% SOC. So it doesn't consider the age of the battery, temperature, load, etc.

2 hours ago, JohanDbn said:

The last 4 were installed yesterday.

So the first test was on a 200 Ah battery (4 x 200 Ah 12 V modules in series)? At 50% depth of discharge (DOD), that's 100 Ah (at the 20 h rate), call it 88 Ah at C/5, so with an 11 A load, you'd expect some 8h of runtime. Something is certainly off, even accounting for fridge start spikes and the like.

I'm wondering if some of the battery modules are older than others, or for whatever reason are at a vastly different state of charge. It might be necessary to charge the low ones first. I'd take them out of circuit for that, if at all possible. You don't want to overcharge some to bring up the others, since they are sealed battery modules.

Also, I notice that your battery modules call for 14.4 to 14.7 V charging; that sounds like AGM to me. I would change the Bulk/CV charging voltage to 57.6 V for summer; now that it's winder perhaps 58.0 V (if you're sure you won't forget to change it for summer). You should change the maximum charging current to about 100 A; this will help the inverter-charger decide when it's time to stop absorb charging. That should allow for a much better charge; it could be all that's wrong.

21 minutes ago, JohanDbn said:

Could the inverter be reading wrong?

It's certainly possible. You presumably have a multimeter; does the overall battery voltage match the voltage reported by the King?

Fortunately, there are ways to calibrate the inverter's battery measurement, but it's a bit of mucking about.

20 minutes ago, Coulomb said:

It also can't be with no load and rested for an hour or several; that part is important. It's not vital to disconnect them, and I know that's a pain, but try to pick a time when the loads are very light and there is zero charging.

I never tested with just the 4. I did my first battery test now, with all 8 and 400ah. Itwas at 100% charged (according to the inverter), and the overall voltage was pretty spot on with the inverter. Very low load during testing (about 300W)

 

23 minutes ago, Coulomb said:

You should change the maximum charging current to about 100 A;

I got told putting that too high could be detrimental to the batteries?

 

Could it be a faulty connection to the inverter even?

 

With lead acids in series (in fact any series arrangement of cells or batteries) you run the risk of getting an imbalance.

This imbalance will never correct itself and will just keep getting worse typically with the weakest bugger getting weaker and weaker...

That is why Victron have a LA battery balancer product and the midpoint monitoring feature in the BMV 702/712 so that you can get an early warning. 

A cheap solution is to get 4x voltmeters and hook them up (check calibration first) so that you can see at a glance what the voltage on each battery is. They consume negligible power so you could just leave them on permanently. Something like this will work nicely https://www.communica.co.za/products/dpm-dig-volt-meter-4-5-150v-red

 

I have (like Coulomb) heard so many tales of premature failures of LA batteries. One of my mates put in 4x 200AH before lockdown and (with correct charging parameters). A week ago he is finding that they look like they have lost a lot of capacity. Without a AH capacity checker it is difficult to check for him. I have seen some nice ones on AliExpress which would be very useful. 

2 hours ago, JohanDbn said:

Did a check on the batteries. They all seem fine. Ranging between 13 to 14.5 volt all, and overall 55V, so seems the batteries are still hundreds. Could the inverter be reading wrong?

With no load or charge and all in balance, they should settle at around 12.5 to 13V if charged, but there shouldn.t be more than about 0.1V difference between the individual batteries. This is pointing to a problem. I would charge the bank, and then disconnect it and let the batteries settle a bit, and then measure voltages. If one or more is much different to the others then that is your problem.

Just did another voltage check, with under 400W load:

Battery 1: 14.04V

Battery 2: 14.19

Battery 3: 12.9V

Battery 4: 12.85

 

Battery 5:  13.57

Battery 6: 13.57

Battery 7: 13.32

Battery 8: 13.55

(Battery 1 - 4 and 5 to 8 in series, the two sets in parallel, Battery 1-4 is about 3 days old, 5 to 8 is about 2 weeks old)

Also attached the info from when i tested

49 minutes ago, JohanDbn said:

Just did another voltage check, with under 400W load:

Battery 1: 14.04V

Battery 2: 14.19

Battery 3: 12.9V

Battery 4: 12.85

No way those battery modules are under load; while there is a <400 W load on the inverter, the battery modules are being charged. Modules 3 and 4 are markedly low voltage compared to the other two. They need to be charged separately.

49 minutes ago, JohanDbn said:

Battery 5:  13.57

Battery 6: 13.57

Battery 7: 13.32

Battery 8: 13.55

Even module 7 should be charged a little to catch up with the others in its string. Since this set is relatively even, you can see that they are no-where near full; it looks like they are floating at around 13.5 V per module (54 V, per the float voltage setting). This is your problem. If your installer won't fix it, get a car battery charger (not ideal, but it will suffice for this hopefully one-off purpose) and get them even. If you can find a 12 V load (I have some old car headlights with the filaments in series, with decent wires and crocodile clips and a ceramic mug to keep the heat and light away from me), it will be easier to being them all to around say 12.7 V rested than trying to get them all up to 12.9 V or whatever they go to when full. In other words, charge the low voltage ones and discharge the high voltage ones to "meet in the middle". I'd say you'll have to take them all out of service for this, separating the strings, though the connections between modules could stay.

2 hours ago, JohanDbn said:

3 hours ago, Coulomb said:

You should change the maximum charging current to about 100 A;

I got told putting that too high could be detrimental to the batteries?

That's true. But it says on the front of the battery module something like (forgive me if I misread it): "Constant current charge maximum 60 A". With two strings, that's 120 A max. I suggest cutting it back to 100 A because of the Axperts' well known tendency to overshoot target battery charge current. You'll likely never see more than 80 A anyway, and that would only be on good solar days an hour either side of noon. But as I mentioned, the maximum charge current is used to figure out the "tail" current (when the charge current falls to this value, the battery is pronounced full). It's basically the setting divided by 5. So 100 A / 5 = 20 A, or 10 A per string. Leaving it at 50 A means the tail current is 10 A or 5 A per string (hence per battery module). Ironically, having this figure too low could mean you overcharge your battery, because with fluctuating solar input, it might never get the charge current down that low, and stay in absorb mode all day.

I found the specs for my batteries. So looking at this, i see they have 6 cells, and cycle use is 2.4 *2.45VPC.

Does that mean my float should be (2.4*6) * 4 batteries in series = 57.6 ? @Coulomb

What should my bulk charge and float charge here be? Im planning on leaving them on charge for the weekend.

I have attached all the information I got....

 

wrt to software, have I got the latest, or should I upgrade?

 

Edited July 10, 20206 yr by JohanDbn

Any help guys? 

On 2020/07/11 at 1:10 AM, JohanDbn said:

I found the specs for my batteries. So looking at this, i see they have 6 cells, and cycle use is 2.4 - 2.45VPC.

Does that mean my float should be (2.4*6) * 4 batteries in series = 57.6 ? @Coulomb

Yes, that's correct.

On 2020/07/11 at 1:10 AM, JohanDbn said:

What should my bulk charge and float charge here be? Im planning on leaving them on charge for the weekend.

You should use 57.6 for the bulk/CV/absorb voltage setting (setting 26), and you can leave the float setting at 54.0 V.

On 2020/07/11 at 1:10 AM, JohanDbn said:

wrt to software, have I got the latest, or should I upgrade?

Main firmware version 71.92 is quite recent, probably more recent than any I have collected. I would not attempt to update it, unless you feel the need to fix the premature float bugs. In that case, you should consider patched firmware version 73.00e (but not factory firmware 73.00; there will be an issue with fault code 90 in 6 months).

@Coulomb I set it as you suggested, heres my latest findings. 

Top and bottom bank battery voltage:

Battery 1 : 13.11

Battery 2: 13.13

Battery 3: 13.50

Battery 4: 13.56, so these seems to have settled in decently.

At the time my panels lost solar, it was at 100% charge,53.3V. My average usage hovered around 500W. This was at 17:30.

At 19:30 , i was sitting at 42%, 46.4V

That means, at minimal load of about 500W(10A on 48V), i got 2 hours backup time, instead of the expected 18(ish) on 400ah with a switch to utility at 46V

Now also to note, the moment i went back to utility, the battery capacity  shot up to 90% within 2 minutes.

Now with the batteries fine, could this be a faulty inverter? Could the inverter be pulling crazy amounts of power? What else could this be?

Could it really be that the batteries discharge in 2 hours and recharges in 5 minutes at a 10a charge?

I have attached screenshots with the numbers, as well as my current settings. 

 

 

 

Edited July 14, 20206 yr by JohanDbn

On 2020/07/15 at 3:54 AM, JohanDbn said:

Battery 1 : 13.11

Battery 2: 13.13

Battery 3: 13.50

Battery 4: 13.56, so these seems to have settled in decently.

It's looking hopeful, but I'd like to see even better balance.

Quote

At the time my panels lost solar, it was at 100% charge,53.3V.

You seem to be putting too much faith into the SOC reading on the inverter. Any value higher than 42.0 + 10.2 = 52.2 V will be called 100%, even though that's only 13.05 V per 12 V module, which is just getting started if charging. So get a good SOC reading you need a battery monitor such as a Victron BMV, or observe the battery voltage near the absorb voltage (57.6 V) for well over an hour, until the charge current (while the battery voltage is near 57.6 V) falls to about 10 A. It would be great if you could watch your charge (middle, green) LED during the day, to see if it stops flashing before the above conditions are met (i.e. the battery is truly full).

It looks like we can lightly patch main firmware version 71.92 for Axpert Kings soon. The only hassle is we can't show a decent version number, as the version display is controlled by the MCU (display/BMS) processor, and it only accepts 4 numeric digits. I get a strong feeling that this is the basis of your problem.

As a point of interest, what is your display/MCU firmware version number? It doesn't seem to show on the Android app. It will show blow "U2" when pressing the up or down buttons on the inverter. [ Edit: oh, I see it's 02.40 from a Watchpower screendump. That's plenty recent enough. ]

Quote

At 19:30 , i was sitting at 42%, 46.4V

Actually, that's an average of 46.4/4 = 11.6 V, with presumably lightish load (500 W or so). I would guestimate that at more like 25% SOC.

Quote

That means, at minimal load of about 500W(10A on 48V), i got 2 hours backup time, instead of the expected 18(ish) on 400ah with a switch to utility at 46V

Hopefully, it's just because it's not charging properly. I'm short of time at present; did we try temporarily just increasing the float voltage setting (setting 27) to about 57.5 V? So then it's basically the same as the absorb voltage, so if it goes to float prematurely, it still charges? There is then a risk of overcharging your battery, but that seems to be the least of your worries just now. If runtime improves dramatically, then you know it's the premature float issue.

Quote

Now also to note, the moment i went back to utility, the battery capacity  shot up to 90% within 2 minutes.

No, the inverter's voltage-based estimate of the SOC shot up. Batteries change voltage dramatically when switching from discharge to charge (and back again); it's a sort of hysteresis. The inverter also doesn't take into account the effect of charging current on the battery voltage (so it ignores internal resistance as well as the hysteresis). The actual SOC will not have shot up, and your inverter isn't using insane power to fast charge the battery.

Now with the batteries fine, could this be a faulty inverter? Could the inverter be pulling crazy amounts of power? What else could this be?

Quote

Could it really be that the batteries discharge in 2 hours and recharges in 5 minutes at a 10a charge?

No.

Quote

I have attached screenshots with the numbers, as well as my current settings. 

Thanks. It often forestalls questions that I have.

After load shedding is over (is that being optimistic?), you should increase the voltage in setting 12 to say 48 V. That will increase the life of your battery by not discharging it too far after poor solar days. But leave it at 46 V for now, otherwise it's more difficult to compare situations.

Edited July 16, 20206 yr by Coulomb

@CoulombThanks for the reply. I changed float to absorb, and then one battery shot to 17V, and i promptly took it back down again. 

What I have noticed, is when I charged it via AC overnight, it did reach the absorb voltage. I then had better performance at night. It lasted me from 6pm to 4:30am. That clearly is not yet ideal, but far better than the 2 hours I got previously. To note, we also had the washing machine on while on batteries, which wouldve taken some standby time off, so its not a clear reflection on a constant amperage.

Noticed a 20% drop over4 hours, from 100 to 79, and then a major drop in an hour from that 79% to 65%. Bearing in mind the inverter may not read the SoC on the batteries correctly, the inverter is the one making the decision on when to go over to utility again, any idea what could be the reason for this drop? The load was not increased, to the contrary, the load actually decreased, as we were heading to bed. Could it be ambient temperature related?

To mitigate the float charge bug, i dropped my max charging to 10A, and waited an hour after sunrise to bump up themax charging amps. At last check it was hovering around the absorption voltage at 11 amps. If my knowledge of charging is correct, I should now start seeing a gradual drop into the float mode?

That should leave me at float voltage at sunset, and with (hopefully) no larger appliances running tonight, I can get back to you tomorrow on a more realistic standby time on an average of 10A/500W.

[Edit: It did go down, and now hovering on a trickle charge of 3A around float]

The night i only got 2 hours, I cannot remember seeing it going higher than the float. That may well be a setting issue, as I had max AC on 2A (SBL/UDC) and max charge on 50A, which if im right is the perfect setting to trigger the bug if it carries that into sunrise?

The floating bug issue.... Flashing the inverter to the one you linked to may be a downgrade on the firmware? Is that advisable , or should I rather just play around on the voltages to mitigate the floating voltage bug until the patch come out? Is there a way/site where I can see when this comes out? I did not see any downloads on the voltronix site.

 

5 hours ago, Coulomb said:

After load shedding is over (is that being optimistic?), you should increase the voltage in setting 12 to say 48 V. That will increase the life of your battery by not discharging it too far after poor solar days. But leave it at 46 V for now, otherwise it's more difficult to compare situations.

Hope for the best, prepare for the worst. 

As for poor solar days.... I live in Durban, what are those? lol

Edited July 16, 20206 yr by JohanDbn

1 hour ago, JohanDbn said:

I changed float to absorb, and then one battery shot to 17V, and i promptly took it back down again. 

Eek. It sounds like your battery modules are very out of balance.

Quote

What I have noticed, is when I charged it via AC overnight, it did reach the absorb voltage. I then had better performance at night.

Yes, the premature float bug doesn't show up with AC charging, unless you change the settings to unusual values; then it will happen even with AC charging.

Quote

It lasted me from 6pm to 4:30am. That clearly is not yet ideal, but far better than the 2 hours I got previously.

The rest might be due to the imbalance.

Quote

Noticed a 20% drop over4 hours, from 100 to 79, and then a major drop in an hour from that 79% to 65%.

So 79% on the guess-o-meter represents about 12.5 V per 12 V module, a moderately charged state, perhaps even around 79%. But 100% on the guess-o-meter is just over 13 V per 12 V module, which is a state that won't last long and will quickly drop to around 12.8 V (if fully charged) or lower. So that is probably the reason for the initial sharp drop.

65% SOC GOM is around 12.2 V per 12 V module. With a light load, that might represent some 50% real SOC. That's a pretty fast drop, and suggests to me that one or more of your modules is running out before the others.

See if you can find out which one(s) are sagging first, and try to charge them separately with an automotive charger of some sort. Unless it's a particularly beefy charger, it will take hours to shift the needle on those. After say 8 hours, let it rest an hour or two, and compare with the voltage of the others (preferably take one string out for this, so you have some rested modules to check against). If it's still below at that point, give it some more hours. You may have to do this for the second worst and even more. Once they are close (withing 0.1 V certainly), put them back in service and see if that improves things.

Quote

Could it be ambient temperature related?

Temperature certainly affects lead-acid battery voltages. I'm rusty, but I think the voltage should rise as the temperature falls.

Quote

To mitigate the float charge bug, i dropped my max charging to 10A, and waited an hour after sunrise to bump up themax charging amps. At last check it was hovering around the absorption voltage at 11 amps. If my knowledge of charging is correct, I should now start seeing a gradual drop into the float mode?

The current should gradually drop, yes, but the voltage (if the sun permits) should hover around the absorption voltage until the current falls to below 10 A (your 50 A from setting 02 (I hope I remembered that correctly) divided by 5. Then the charge current should briefly go negative, the battery voltage should fall to around the float voltage, and it should bump around that value for the rest of the day.

Occasionally, after a big load, it might go to zero charging for a minute, then back to bulk charging again, heading towards the absorb voltage again.

Quote

[Edit: It did go down, and now hovering on a trickle charge of 3A around float]

That sounds about right for maintaining the float voltage.

Quote

The night i only got 2 hours, I cannot remember seeing it going higher than the float. That may well be a setting issue, as I had max AC on 2A (SBL/UDC) and max charge on 50A, which if im right is the perfect setting to trigger the bug if it carries that into sunrise?

I don't understand this. You mention setting 11 (maximum utility charging current) set to 2 A, but this seems irrelevant as you had UDC (Utility Disabled from Charging). What is being carried into sunrise?

Quote

The floating bug issue.... Flashing the inverter to the one you linked to may be a downgrade on the firmware? 

Yes, it's downgrading the firmware. You can in this instance put it back, actually to a slightly later version than what you have now: Axpert King main firmware version 71.92. Or of course wait till the patched version comes out, but you have the factory firmware in case it gets delayed.

Quote

Is that advisable , or should I rather just play around on the voltages to mitigate the floating voltage bug until the patch come out?

I suggest you try and sort out the imbalance issue first.

[ Edit: I meant to add that a lot of the King firmware changes lately seem to be focused on lithium batteries with BMSs. Since you are running lead acid, these changes won't affect you, so regressing back to a lightly patched version of 71.80 should not cause you much concern. However, there is a small chance that the format of the data exchanged between the main and remote display firmware may have changed in some way that will affect you. But if that's the case, you could also regress the remote display firmware to version 02.00, an upload by user RichieRich. ]

Quote

Is there a way/site where I can see when this comes out? I did not see any downloads on the voltronix site.

Voltronic Power no longer provide a site to download firmware, after the clones came out with Voltronic firmware in them. Keep watching the files area of this forum: https://powerforum.co.za/files/ .

Edited July 16, 20206 yr by Coulomb