Axpert 3KV battery sense issue

[Nic Holmes]

Good evening!

 

Forgive me if I have posted this in the wrong area of the forums, not much. of a forum person.

 

Anyway - let me get into the nitty gritty, this will be a long post but I feel that it needs a lot of information and data to be hopefully solved!

Almost exactly 2 years ago I posted in power forums about my Axpert 3KV 24V inverter not sustaining a load off battery (+/- 40min at 100W on 100Ahr batteries) - it was put down to the fact the batteries I had purchased were sub optimum and not actually able to supply their rated capacity - not being able to afford new batteries I wrote it off and just lived with it.

Fast forward to now.

My neighbour had solar installed (Axpert 5KV along with 200Ahr batteries), during the installation I chatted to the installers regarding batteries, they recommended the batteries they were installing next door (200Ahr AGM batteries from National Battery)

So since I had been saving up for batteries fro the past 2 years, I took the plunge. Shelled out 10k and took delivery of 2 x 200Ahr AGM batteries - this was end of Feb.

I installed them and set the inverter to the settings as specified in the battery manual, did a quick test to see that were working and left them be.

two weekends ago we had a power failure in our area - we were watching TV and had 2 lights on - a total of 161Watts. 40min later the low battery alarm starts and within 10min the power is out! I go over to the neighbour and they are drawing 550Watts and still going (granted their system is 48V, but also 200Ahr). they had power for the full 4 hours of the power outage in our area.

This lead me to do some testing and some digging.

I connected 3 separate volt meters to the batteries - excluding the inverter display and monitored them all on a fixed 200W load.

the results were as follows:

Inverter - Meter 1 - Meter 2 - Fluke 77

27.3.        27.3.        27.2.        27.3

25.1.        25.2.        25.1.        25.2

24.1.        24.8.        24.7.        24.8

23.9.        24.7.        24.6.        24.7

23.2.        24.0.        23.9.        23.9 (low voltage alarm now sounding)

This is all within 40min

Disconnected load and recovered voltage on batteries after 5min rest = 25.2V

 

Now that I am unsure of whether the previous batteries I purchased were indeed poor or if they were maybe damaged by the inverter - I remove the new batteries and put the old batteries back - which are now proper buggered.

To ease my mind, I take one of the new batteries, connect it to a 400W Trailboss inverter I purchased from outdoor warehouse some time ago and connect the same 200W load up. 7 hours later I turn off the load with battery voltage indicated at 11.2V - recovered voltage after 5min rest = 24.5V

I then take the second battery and repeat the test - 7 hours again I turn it off, similar voltage readings.

 

I believe there is a fault with how the Axpert is reading voltage and sensing current in order to determine battery level - with this I do not believe it is charging the batteries correctly either and possibly damaging them.

 

I have also noticed when looking for the firmware version this afternoon that only one version is being displayed on the display and not two like it should. Current version displayed is U1 02 51

 

Has anyone ever experienced anything similar?

Is there is a fix for this?

Any other pointers?

 

Cheers!

 

[Jaco De Jongh]

@Nic Holmes,  What do you use to charge these batteries and what is the charge settings set to? Absorb, float and allowed charging current?

[Nic Holmes]

11 hours ago, Jaco de Jongh said:

@Nic Holmes,  What do you use to charge these batteries and what is the charge settings set to? Absorb, float and allowed charging current?

The batteries are charged by the Axpert inverter, set to the battery manufacturers specified charging conditions.

Settings for batteries/charging are as follows

2 - 20A (maximum charging current for solar and utility chargers)

5 - USE (user defined battery conditions)

11 - 20A (maximum utility charging current)

12 - 23V (voltage point back to utility source)

13 - FULL (voltage point back to battery)

16 - CSO (solar priority charge)

26 - 28.8V (bulk charge as per battery label data)

27 - 27.4V (float charge as per battery label data)

29 - 22.5V (low DC cutoff voltage)

 

all tests done on the batteries were using same charging method (Axpert on board charger)

 

 

 

 

[Jaco De Jongh]

20 minutes ago, Nic Holmes said:

The batteries are charged by the Axpert inverter

Sorry I wasnt clear, do you use the AC side or do you use panels to charge, and if you use panels, how many and what size are installed?

[Nic Holmes]

2 hours ago, Jaco de Jongh said:

Sorry I wasnt clear, do you use the AC side or do you use panels to charge, and if you use panels, how many and what size are installed?

Solar panels are priority charge. there are 640W worth of panels. 2 x 320W

in evening the batteries are topped up with AC side of the charger.

 

[Jaco De Jongh]

25 minutes ago, Nic Holmes said:

in evening the batteries are topped up with AC side of the charger.

This leads to more questions, Do you discharge your batteries on a daily basis? Why do you need to top it up in the evenings? 

[Nic Holmes]

1 hour ago, Jaco de Jongh said:

This leads to more questions, Do you discharge your batteries on a daily basis? Why do you need to top it up in the evenings? 

The unit is configured to supply power to the house (lighting circuit and a few plugs) off solar energy during the day and utility at night except in the event of a power failure which then supplies off battery. (menu setting 1 - SOL)

 

So this can at times and most evenings actually lead to some energy off the batteries being consumed - requiring a small top up when utility power takes over from the solar at the point the panels no longer produce enough energy.

 

Just to clarify - all lighting is LED

Plug circuit powers the TV, wifi router and one floor lamp.

With all lights turned on, TV running and all ceiling fans turned on we can only just reach 500W... On average the load is no more than 200W

 

However I am not sure how any of this relates to the problem I am seeing when the batteries are full and the mains power is removed without any solar present (ie at night) and 40min later the inverter disconnects due to "flat batteries" which at the time are showing nearly a full volt difference in level between inverter and external volt meters?

[Jaco De Jongh]

1 hour ago, Nic Holmes said:

However I am not sure how any of this relates to the problem I am seeing when the batteries are full and the mains power is removed without any solar present (ie at night)

Let me try and explain. If your batteries are not fully charged, you might see the same behaviour as now. 

You mentioned you have 640whp on the roof, you can expect 80% of that (512wh)  to reach the batterys, but you have a load between 200 and 500 watt, so not much power is reaching your battery's in the day...

No lets go a bit deeper, you can not determine how full your batteries are by measuring volts and Axperts are known for inaccurate voltage reading especially on the lower end of the scale. What you would need to be sure of what is truly happening, is a battery monitor, without that everything we say here is speculation, but I am willing to guess that your battery's are not fully charged. 

Your system in general are a bit unbalanced and i would have liked to see more panels in the equation, or alternatively a way to ensure the batteries remains full.

Then lastly, the charging from the grid is terribly inefficient and on your size system it doesn't make sense to do it. Whatever you save on power during day, wont make up for the extra cost for charging from the grid. 

[Nic Holmes]

28 minutes ago, Jaco de Jongh said:

Let me try and explain. If your batteries are not fully charged, you might see the same behaviour as now. 

You mentioned you have 640whp on the roof, you can expect 80% of that (512wh)  to reach the batterys, but you have a load between 200 and 500 watt, so not much power is reaching your battery's in the day...

No lets go a bit deeper, you can not determine how full your batteries are by measuring volts and Axperts are known for inaccurate voltage reading especially on the lower end of the scale. What you would need to be sure of what is truly happening, is a battery monitor, without that everything we say here is speculation, but I am willing to guess that your battery's are not fully charged. 

Your system in general are a bit unbalanced and i would have liked to see more panels in the equation, or alternatively a way to ensure the batteries remains full.

Then lastly, the charging from the grid is terribly inefficient and on your size system it doesn't make sense to do it. Whatever you save on power during day, wont make up for the extra cost for charging from the grid. 

Ok that makes sense.

The Axpert 3KVa unit can only take 600W solar input, that is why there is only that many panels installed.

But here is the curve ball.

 

Firstly - the powered circuits in the house are lighting and TV - neither of which are used during the day, so the Solar energy is purely charging batteries. Maybe one or to ceiling fans are on during the day - thats less than 100W

The Axpert (albeit is bad at charging) sits in float charge for about 70% of the day.

Only from about 5pm onwards do we use more energy than the panels produce.

The idea behind the system was for load shedding and power outages, which it is not handling currently. However when first fitted 5 years ago on used 200Ah batteries it ran extremely well, lasting a good 8 hours with no power...

 

Secondly - then if the batteries are not being charged fully, how is that on a 160W load through the axpert I get 50min give or take run time. Allow batteries to charge again (on the axpert) and then put them on a 400W inverter and the same 160W load is powered for 7 hours and counting?

 

What I will do tonight is charge both the batteries using an Optimate 20Ahr smart charger until it says that the batteries are full, then put them back into the system and see how long it powers that same 160W load.

[Jaco De Jongh]

1 hour ago, Nic Holmes said:

Secondly - then if the batteries are not being charged fully, how is that on a 160W load through the axpert I get 50min give or take run time.

While this is happening, have you ever watch the discharge current? What does the Axpert report?

[Nic Holmes]

1 hour ago, Jaco de Jongh said:

While this is happening, have you ever watch the discharge current? What does the Axpert report?

No I haven't. I will be sure to set it to that screen before doing the test.

So I charged both the batteries now with Optimate charger - took less than 10min each before the charger said they were full at 13.4V per battery. Charger did not even enter bulk charge stage.

I don't think I will get to doing the test tonight, I will attempt it tomorrow morning and report back.

[Nic Holmes]

19 hours ago, Jaco de Jongh said:

While this is happening, have you ever watch the discharge current? What does the Axpert report?

Ok, so I have just connected everything back up with the new batteries.

These were charged up to full by a separate charger to the Axpert.

Solar and mains disconnected.

First I placed a 215W load onto the system and the battery discharge current was 13A (applying a bit of maths P=VI at 24V that equates to 312W - but since batteries are 25.1V thats more like 326W!!)

I then placed a 80W load onto the system and the battery discharge current was 6A (again P=VI at 24V equates to 144W, at 24.9V now thats more like 150W)

Finally I placed a 200W load onto the system and have left that connected to see how long before the disconnect happens.

But at 200W, discharge is 11A which at 24V should be more like 264W.

Now remember I got 7 hours out of one of these batteries (12V and 150W load) so I should at least see 5 hours at 24V...

 

 

[wolfandy]

When doing your conversions you need to keep 2 things in mind:

1. The inverter does not operate at 100% efficiency
2. The inverter has self-consumption, which it draws from the batteries. I don't know how much this is for the 3kVA Axpert, but for my 5kVA it is just over 40W

An example from my system: My PV generation is 3800W and my load is 1500W. My batteries are only being charged with just under 2100W, which means that I have a bit more than 200W 'system losses' due to self consumption and inefficiencies

[Nic Holmes]

1 hour ago, wolfandy said:

When doing your conversions you need to keep 2 things in mind:

1. The inverter does not operate at 100% efficiency
2. The inverter has self-consumption, which it draws from the batteries. I don't know how much this is for the 3kVA Axpert, but for my 5kVA it is just over 40W

An example from my system: My PV generation is 3800W and my load is 1500W. My batteries are only being charged with just under 2100W, which means that I have a bit more than 200W 'system losses' due to self consumption and inefficiencies

That would equate to a 95% efficiency in your system is pretty reasonable.

But that is not the problem I am seeing... Forget about how much solar is being generated or whether mains chargers are enabled.

Simply put, full batteries on a 200W load as in a UPS arrangement lasts an 8th of the time the same load on the same battery (1 x 12V instead of 2 x 12=24 - both 200Ah) on another inverter.

[Nic Holmes]

So on the 200W load drawing 11A from the batteries it took 2.5 hours before the low voltage alarm started beeping....

 

[Nic Holmes]

I have just run the inverter off batteries again now, after the earlier test was concluded and the batteries were left to be charged by remaining solar for the afternoon and then the Axpert charger.

 

150W load, 6A discharge current from the batteries and 32min later the low battery alarm was beeping...

[Coulomb]

On 2020/04/21 at 3:54 AM, Nic Holmes said:

To ease my mind, I take one of the new batteries, connect it to a 400W Trailboss inverter I purchased from outdoor warehouse some time ago and connect the same 200W load up. 7 hours later I turn off the load with battery voltage indicated at 11.2V - recovered voltage after 5min rest = 24.5V

If the above is not a typo, and you're not mixing module (nominally 12 V) voltages with battery (nominally 24 V, i.e. two modules in series) voltage, then that's a catastrophic discharge of the battery, down to an average of 5.6 V per nominally 12 V modules. It's not fair to compare that brutality to the Axpert, which will stop discharging (even if somewhat early) at a less blood thirsty ~22 V (out of nominally 24 V).

I don't know the 24 V inverter-chargers well. They may have a way of calibrating the offset as well as the scale of the battery voltage measurement. That would seem to make the Axperts line up better with the voltmeters. But the cost is that it will then discharge the battery even deeper, which is really bad for battery life.

The sad reality is that you can't take more than about 20% of the energy from a lead acid battery's capacity on a regular basis. If you do, you end up with very short life, perhaps two months.

[Chris Louw]

Like Coulomb stated above it is not wise to withdraw more than 20% of the power of all lead acid battery . Withdrawing 200 watt for 7 hours from a 12v 200 ah battery is not good practice . Your question is why does the battery last longer on the Trailboss inverter . The only way to find out is to get a meter to find out what power is really used . Have you measured the amps taken from the battery . This can be done using a clamp meter on the battery cable or Victron has a battery management system that can do this very good but the cost is about R2200 . You can also use a digital power monitor that works with a shunt PZEM -051 . The cost is less about R280 . This will give you an better idea of the power ( kWh ) you take out of the battery . If the power used is still a lot more on the Axpert there is some fault with your Axpert or installation . The PZEM can also measure the amount of power going into the battery when charging . Keep in mind that there are losses when charging . The best to do now is to keep the batteries fully charged to limit any more damage to your batteries . Get a specification sheet of the batteries so you know what is good practice with them .

[Nic Holmes]

7 hours ago, Coulomb said:

If the above is not a typo, and you're not mixing module (nominally 12 V) voltages with battery (nominally 24 V, i.e. two modules in series) voltage, then that's a catastrophic discharge of the battery, down to an average of 5.6 V per nominally 12 V modules. It's not fair to compare that brutality to the Axpert, which will stop discharging (even if somewhat early) at a less blood thirsty ~22 V (out of nominally 24 V).

I don't know the 24 V inverter-chargers well. They may have a way of calibrating the offset as well as the scale of the battery voltage measurement. That would seem to make the Axperts line up better with the voltmeters. But the cost is that it will then discharge the battery even deeper, which is really bad for battery life.

The sad reality is that you can't take more than about 20% of the energy from a lead acid battery's capacity on a regular basis. If you do, you end up with very short life, perhaps two months.

Yes - apologies that was a typo.

When a single battery was connected to a 12V inverter (Trailboss) a 200W load lasted 7 hours before a turned off the system with a battery voltage reading 11.2V which recovered to a voltage 12.5V after a few minutes.

Even at 20% discharge they really should last longer than 30min on 150W load though?

[Nic Holmes]

3 hours ago, Chris Louw said:

This can be done using a clamp meter on the battery cable or Victron has a battery management system that can do this very good but the cost is about R2200 . You can also use a digital power monitor that works with a shunt PZEM -051 . The cost is less about R280 . This will give you an better idea of the power ( kWh ) you take out of the battery . If the power used is still a lot more on the Axpert there is some fault with your Axpert or installation . The PZEM can also measure the amount of power going into the battery when charging . Keep in mind that there are losses when charging . The best to do now is to keep the batteries fully charged to limit any more damage to your batteries . Get a specification sheet of the batteries so you know what is good practice with them .

The unfortunate part of all of this is that it seems like I need to throw more money at it to find out why its not functioning the way I believe it should be.

Especially when compared to a similar installation next door... 

I do suspect there is a fault with my Axpert, but how to determine that exactly without spending nearly as much as a new unit I dont know...

Currently I have set the unit to run off Utilities constantly and only use the solar to keep the batteries charged for the even of a power failure - as I really cannot afford to trash another set of batteries!

Does anyone know if there are firmware upgrades for the 3KV units?

[___]

7 hours ago, Nic Holmes said:

The unfortunate part of all of this is that it seems like I need to throw more money at it to find out why its not functioning the way I believe it should be.

I know this is a lot of money, but it's a lot less than a new set of batteries and I really cannot recommend this enough: Get yourself a Victron BMV-700 (or even better, the 702 or the 712) so you can see the proper state of charge of that battery. A lead acid battery is a horrible old beast... I mean, they are ancient technology and have a lot going for them (such as being almost 100% recyclable), but what many people don't realise is that it takes a lot of time to charge them to 100% full. They absorb energy fast and without much effort until about 80% SoC, and then the charge current drops off. By the time many chargers goes to float, the battery is not yet completely full, and using the voltage to attempt a measurement is just useless between 80% and 100%.

With the BMV monitor you can also see how many amp hours have been used, and it even has a ttg (time to go) reading that estimates how long you have left before it is empty.

They go for around 2.2k (for the 700, a bit more for the 702).

Edited April 23, 2020 by plonkster