Hello,

This is a first time post so I hope you will bear with me. I had a system installer install a system a few months ago but the installer seems to have lost interest in helping me resolve a problem with the batteries not performing the way that I believed they should. I base my opinion on the results I get from the Centurion Solar raspberry pi which gives me the following charge information:

 

I don't understand most of what people are writing so I hope I can give the right information. I have a 5kW Mecer Inverter (I believe that this is an Axpert inverter) with 8 batteries, 2 x 48v banks. The batteries I have are:

Currently the settings on Inverter are:

Program 1: SBU

Program 2: 40A

Program 5: USE

Program 11: 40A

Program 12: 47v

Program 13: 54v

Program 16: CSO

Program 26: 54v

Program 27: 54v

Program 28: 45.6v

Program 32: 180

Program 33: EEN

Program 34: 56.4v

The problem that I am experiencing is that at night there is a maximum of between 500w and 1 kw being drawn at any given time and the batteries only seem to last for about 6 hours. They also seem to show this massive jump in state of charge of about 40% immediately when it is either discharging or recharging. Maybe that is how the Raspberry Pi is measuring it but I don't know if this is something I should be worried about. The batteries also never seem to achieve 100% SoC, which I thought they needed to.

My setup seems to be creating a situation where I seem to be charging the batteries more from utility power than from solar, which really doesn't make sense.

Please could someone tell me either that everything is okay, or else I am in the process of damaging my investment, or what other information I need to provide in order to help.

Thanks very much

Graham

 

 

 

 

 

Edited August 5, 20197 yr by GrahamICo

Welcome Graham.

Cannot help on the Axpert, but those Narada's, are they not UPS batteries?

Normally on a 5kva inverter ones needs a 300+ ah bank with a 4kw array to recharge AND power the daily loads, generally speaking.

Another point to ponder, you should not discharge most lead acid banks below 20% on a daily basis.
20% DOD / SOC of 80% is where most of them have their longest cycles advertised, like 4000 on Trojans.
Lithium banks like Pylontech has like 6000+ cycles to quite a low SOC / DOD.

If the Narada's are UPS batteries, then their cycles are quite low and you must stay on or above a SOC of 80% / 20% DOD to make them last.

Do you have a BMV battery monitor?

9 hours ago, GrahamICo said:

a maximum of between 500w and 1 kw being drawn at any given time and the batteries only seem to last for about 6 hours.

Lets look at this first. Nerada isn't keen on publishing Cycle life for there batteries , as @The Terrible Triplett mentioned above, Neradas were meant for UPS applications and are not the correct batteries for daily use. 

From THIS document you will see the the best life expectancy is at above 80% SOC , that will mean that you can only use 20% of the power, if you use more you will shorten the expected life of the batteries a lot. Now take into consideration the following, you have 360 Ah x 48 Volt available, that is   17280 wh and you can safely use 20% of that, that is 3456 wh. At 500 watt load, that will give you 6.9 hours of usage and with 1kw load it will give you  3.456 hours of usage. (No losses taken into consideration) . This correlates with your 6 hour time above. 

9 hours ago, GrahamICo said:

They also seem to show this massive jump in state of charge of about 40% immediately when it is either discharging or recharging.

This tells me they are looking at Battery voltage to determine SOC, its not accurate at all and not recommended. 

9 hours ago, GrahamICo said:

I don't know if this is something I should be worried about.

I will as the same Question as TTT above, Do you have a Battery monitor. If not please consider installing one as soon as possible. That is the only way you will be able to accurately determine the SOD of your battery bank, to protect your investment, this is needed. 

9 hours ago, GrahamICo said:

The batteries also never seem to achieve 100% SoC, which I thought they needed to.

This can be because they measure voltage, but to rule out other reasons, please supply us with the size of your PV array as well as your base load during the day?

If you dont have enough panels, and your load during the day is to high, you will not have enough power to recharge the batteries at day and end up charging from grid instead.  

The Narada NDT range aren't UPS batteries.  They're float batteries, meant mostly for C10 type of discharges in standby applications - typically telecoms use.  Jaco and TTT are right, they're the wrong batteries for solar applications.

I don't know the inverter settings, but it seems as though the OP floats at 54 V and have an EoD voltage of 45.6 V?  That's a EoD voltage of 1.9 Vpc - the datasheet doesn't even have data for that.  The highest EoD in the datasheet is 1.85 V, so I'll work with that.  You have two banks - that is a total of 48 cells. 

 

Load @ 500 W - that is 10.4 W/cell.  You should get more than 24 hours.

Load @ 1 kW - that is 20.8 W/cell.  That should give you closer to 20 hours.

At a EoD of 1.9 Vpc you'll obviously get less capacity, but still nowhere close to 6 hours.  Unless the EoD is the 47 V for program 12, which gives an EoD of 1.95 Vpc, which could explain the 6 hours.  If you want to test your batteries, set the EoD to 1.85 Vpc (44.4 V) and see what you get (try to keep the load constant).

 

How old are the batteries?  Speaking under correction, but most operators are using eiter the 190 or the 200, which means that the bulk that was/are being brought into the country at the moment are 190s and 200s.  I don't know who would bring in 180s, given that they're the same dimensions, very similarly priced (if not more expensvie due to import volumes) and of lesser capacity than the 190.  I won't be surprised if the 180s are old stock which were brought in before the 190s were released.  If they stood on a shelf for some time without their bi-annual charges, they might have been damaged.

If you've tested the batteries at a constant load to a known end voltage and they fail to provide the capacity they should, you could try your luck you can speak to Dartcom.  Afaik they have a sole import agreement with Narada, and try to find out when those batteries were brought into the country and who it was sold to.  That could give you an idea of whether your reseller had them on a shelf for too long.

32 minutes ago, Tacet said:

Load @ 500 W - that is 10.4 W/cell.  You should get more than 24 hours.

Load @ 1 kW - that is 20.8 W/cell.  That should give you closer to 20 hours.

Now this is interesting. What would your SOC be after 24 or 20 Hours? 

Empty

Basically the datasheet says that, from float to the voltage on the table, you should be able to carry X load for Y time.  Relatively easy to test if you have a constant current or constant power load bank (note that capacity increase with temperature, usually numbers are quoted at 25 degrees C).  You can derate that with 20 % to make up for the degradation of a battery over its lifespan (industry standard based on the Eurobat guide).  For a new battery, though, you can test at its full capacity, and it should be able to perform as specified. 

Below Narada's graph for OCV vs SoC.  So the numbers in their tables are all for full discharges to basically 0 % SoC.  But that is what these batteries were designed for - to stand on float for months, and to cycle once in a blue moon.  So is the OP's inverter setup for full discharges, or for partial discharges to conserve the batteries?  If it is setup to stop discharging at 1.9 Vpc (11.4 V per block), then its setup to do basically full discharges, and the batteries should give much more than 6 hours.  But if the cutoff is 47 V (1.95 Vpc /  11.75 V per block), then it is configured for dischargse down to somewhere between 10-20 % SoC (remember that OCV as per graph will be higher than voltage under load, which is what EoD is based on).

 

Either way, I pity the poor batteries, and I do suspect that they've already been damaged.